ADVANCES IN
Immunology EDITED BY FRANK J. DIXON Scripps Clinic and Research Foundation La Jolla, California ASSOCIATE EDITORS
K. FRANKAUSTEN LEROYE. HOOD JONATHAN W. UHR
VOLUME 45
ACADEMIC PRESS, INC. Horcourt Brace Jovonovich, Publishers Son Diego London
New York Berkeley Boston Sydney Tokyo Toronto
COPYRIGHT 0 1989 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. San Diego, California 92101
United Kingdom Edition published by ACADEMIC PRESS LIMITED 24-28 Oval Road. London NW 1 7DX
LIBRARY OF CONGRESS CATALOG CARD NUMBER:
ISBN 0-12-022445-3
( a k . paper)
PRINTED IN THE UNITED STATES OF AMERICA 8 9 9 0 9 1 9 2
9 8 7 6 5 4 3 2 1
61-17057
ADVANCES IN IMMUNOLOGY, VOL. 45
Cellular Interactions in the Humoral Immune Response ELLEN S. VITETTA, RAFAEL FERNANDEZ-BOIRAN, CHRISTOPHER D. MYERS, A N D VIRGINIA M. SANDERS Department of Mlcrobiology, Unlverslty of Texas Southwestern Medical Center. Dallas, Texas 752SS
I. Introduction
The immune system has evolved primarily to combat infection by pathogenic organisms. It is characterized by its virtually infinite repertoire of specificities, its highly specialized effector components, its complex regulatory mechanisms, and its mobility. In contrast to most other organ systems, the immune system is not confined to a single site in the body: immunocytes and their secreted molecules traffic within and among lymphoid organs and various body compartments. Hence, a highly complex system of communication has developed among the various cell types in the immune sytem. One important mechanism of communication is the requirement for interactions among cells for the activation and differentiation of resting B lymphocytes into antibody-secreting cells. These cellular interactions involve both cell/cell contact and the release of mediators (cytokines) that can act in either an autocrine or paracrine fashion on cells both within and outside the immune system. In the present review we will discuss the interactions between T and B cells and the role of accessory cells and cytokines in the generation of specific antibody responses. The first portion of the review is a historic perspective (Section I), followed by a summary of present-day concepts (Sections 11-V): in the final section we speculate on how the different components of the immune system might function in mvo (Section VI).
A. MODELS OF T CELL/BCELLINTERACTIONS In 1966, Claman and his co-workersprovided the first direct evidence that T and B cells interact in the generation of an antibody response to sheep red blood cells (SRBCs)* (1). Lethally irradiated mice were *Abbreviations used: ABC, antigen-binding cell; AIDS, acquired immune deficiency syndrome; APC, antigen-presenting cell; BCGF, B cell growth factor; BCDF-y, B cell differentiation factor for IgG,; BSF-1, B cell stimulatory factor-1; BSF-2, B cell stimulatory factor-2; C, constant region; CML, cell-mediated lysis; CSF, colonystimulating factor; DNP, dinitrophenyl; DTH, delayed-type hypersensitivity; EAF, 1 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
2
ELLEN S. VITETTA ET A L .
injected with thymocytes or bone marrow cells from normal or immune syngeneic donors. The cell transfer was followed by challenge with antigen (SRBC) and the spleen cells from the recipient mice were assayed at various times for the secretion of hemolytic antibodies. Such experiments showed that neither thymus-derived nor bone marrow-derived cells could, on their own, elicit an anti-SRBC response. However, when the two types of cells were present in the same recipient, an antibody response was elicited. Studies by Davies et al. (2) and Mitchell and Miller (3-7)confirmed and extended these conclusions and established that the antibodyforming cell precursors were derived from the bone marrow population and that thymus cells could recognize and react specificallywith antigen, but did not, themselves, produce antibody. It was suggested that the thymocytes or their mature progeny played a role in helping the B cells to differentiate into antibody-producing cells. With the advent of in vitro tissue culture techniques developed by Mishell and Dutton (8, 9) and Marbrook (10, 11), it was possible to further elucidate the roles of different cell types under more controlled experimental conditions. Using in Vitro culture, Mosier and his colleagues (12-14) first demonstrated that when spleen cells were separated by virtue of their ability to adhere to plastic, neither the adherent nor nonadherent population of cells could, on their own, make an antibody response. The cells in the adherent population required for an antibody response were macrophages, while the cells in the nonadherent population were primarily T and B cells. Experiments carried out by many investigators confirmed that T and B cells were required for both primary and eosinophil activation factor; EBV, Epstein-Barr virus; EDF, eosinophil differentiation factor; FCS, fetal calf serum; FDC, follicular dendritic cell; F I X , fluorescein isothiocyanate; HGF, hybridoma growth factor: HIV, human immunodeficiency virus: HSA, human serum albumin; ICAM-1, intercellular adhesion molecule-1; IFN , interferon; Ig, immunoglobulin; IL, interleukin; Ir, immune response (gene); b,dissociation constant; KHF, killer helper factor; KLH, keyhole limpet hemocyanin; LAF, lymphocyteactivating factor; LAK, lymphokine-activated killer cell; LFA, lymphocyte function associated; LPS, lipopolysaccharide; LT, lymphotoxin; MABC, memory ABC; MAF, macrophage-activating factor; MHC, major histocompatibility complex: mAbs, monoclonal antibodies; MTOC, microtubule organizing center; NK, natural killer cell: OVA, ovalbumin; PC, phosphorylcholine; PCT-GF, plasmacytoma growth factor; PFC, plaque-forming cell; PGEp, prostaglandin E,; PKC, protein kinase C; PMA, phorbol myristate acetate; PNA, peanut agglutinin; PPD, purified protein derivative; r, recombinant; R, receptor; RAMIg, rabbit antimouse Ig; SDS, sodium dodecyl sulfate; sIg, surface immunoglobulin; SN, supernatant; SRBC, sheep red blood cell; Tc, cytotoxic T cell; TCGF, T cell growth factor; TcR, T cell receptor; TD, thymus dependent; Th, T helper cell; TI, thymus independent: TNF, tumor necrosis factor; TNP-ABCs, trinitrophenylantigen-binding cells; TNP-MABCs, memory TNP-ABCs; TRF, T cell replacing factor; B,T suppressor cell; TT, tetanus toxoid; V, variable region.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
3
secondary antibody responses and that macrophages served as accessory cells (7, 12, 15). B. INTERACTIONS BETWEEN T AND B CELLS SPECIFIC FOR DIFFERENT DETERMINANTS O N THE SAME ANTIGEN (HAPTEN-CARRIER EFFECT) An important but initially perplexing observation was the demonstration of the “hapten-carrier’’ effect. It had been known for many years that immunization with a small nonimmunogenic molecule or hapten was effective only when the hapten was coupled to a “carrier” molecule which was immunogenic. It was demonstrated that cooperative interactions between distinct lymphocytes specific for carrier and haptenic determinants were essential for the development of an antihapten antibody response. The first direct evidence for cooperative participation of two cells with distinct determinant specificitieswas obtained by Mitchison (16). Spleen cells obtained from syngeneic donor mice that had been immunized with a hapten-carrier conjugate secreted antihapten antibodies following challenge with a homologous haptencarrier conjugate, but not with a heterologous conjugate containing the correct hapten but another carrier. In contrast, when spleen cells from donors immunized with a hapten-carrier conjugate were mixed with spleen cells from donors immunized with another carrier, a good secondary response could be obtained using the hapten conjugated to the second carrier. Thus, cells specific for the second carrier helped the hapten-specificB cells to make an antihapten response. It was later shown by Raff and colleagues (17, 18) that the carrier-specificcooperating cells, or helper cells, were thymus derived, whereas the antihapten antibodyforming cells were bone marrow derived. These experiments were made possible by the identification of a marker on thymus-derivedcells called 8 (now Thy-1), and the development of an antibody against 8, which, in the presence of complement, could lyse the thymus-derived cells. By eliminating 8 cells from cell mixtures, it was demonstrated that they were responsible for carrier specificity and for cooperating with the B cells in the elaboration of an antibody response to the hapten. The phenomenon of cooperation between carrier-specificT cells and haptenspecific B cells was also demonstrated independently by Rajewsky et al. (19), who immunized rabbits with a hapten-carrier conjugate and observed that the animals made a significant antihapten antibody response if they received a supplemental intervening immunization with free, unconjugated carrier. It was further demonstrated that the intervening immunization with the carrier primed a second helper T cell population, which could then cooperate with the hapten-specific B cell +
4
ELLEN S. VITETTA ET A L .
population in responding to the hapten-carrier conjugate. While subsequent in vilro studies went a long way in defining the cell types involved in the linked response to hapten and carrier, they also pointed to the vast array of artifacts involved in obtaining antibody responses in vitro as opposed to in vivo. Numerous papers focused on the roles played by the constituents of the tissue culture media, fetal calf serum (FCS), plastic ware, and the addition of reagents (such as 2-mercaptoethanol) in obtaining optimal antibody responses in m’tro (9, 20-24). During this time, immunologists were faced with the frustrating problem that it was often difficult to repeat experiments among different laboratories because of technical differences in the culture systems. Nevertheless, with time, it became evident that the basic tenet of T cell/B cell collaboration involving hapten-specific B cells and carrier-specificT cells was correct, albeit with some qualifications. For example, experiments carried out by Miikela and his associates (25, 26) and others (27-31) established that an increasing density of repeating antigenic epitopes correlated with decreasing dependency on T cells, and that epitope density affected not only the magnitude of the response, but the isotype of antibody secreted as well. Additional variables that influenced the in vilm antibody response included the physical nature of the carrier (particulate or soluble), virgin versus memory cells, and differences among mouse strains (32, 33). C. THEROLEOF T CELLSON THE QUALITY OF THE ANTIBODY RESPONSE As discussed above, experiments carried out in the 1960s demonstrated that T cells are required for B cells to respond to most antigens. It was further demonstrated that T cells were also required to induce the progeny of the B cells to switch from the secretion of IgM to IgG antibodies (34-42). Furthermore, other experiments showed that T lymphocytes were involved in affinity maturation of the antibody response and that the more “T independent” the response, the poorer the affinity maturation (39, 43-46). It was concluded that T cells play an essential role not only in the activation of resting B cells, but also in isotype switching and affinity maturation of this response. Thus, T cells can regulate the levels and affinity of serum antibodies of different isotypes. Different Ig isotypes are essential for providing the most effective means of eliminating a given type of pathogen, e.g., viral, parasitic, or bacterial. Finally, T cells play a major role in generating memory B cells which can respond more effectively to subsequent challenge with antigen (47).
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
5
D. SUPPRESSION VERSUS ENHANCEMENT OF THE IMMUNE RESPONSE Gershon and Kondo (48, 49) were the first to develop the concept of suppressor T cells. In general, suppression appeared to be nonspecific (50, 51); however, there were claims of antigen-specific suppressor T (Ts) cells (52-55). Subsequently, additional mechanisms for down regulating immune responses were suggested, including antibody-mediated suppression generated by immune complexes (2, 56-68) or antiidiotypic antibodies (69-71). It is now apparent that at least some types of suppressor cells can be distinguished from helper T (Th) cells by virtue of the expression of different surface markers. More recent evidence suggests that suppression may be mediated, at least in certain cases, by soluble factors elaborated by the suppressor cells that can act on helper cells and, possibly, on other immunocytes as well (72). The mechanisms by which Ts cells manifested their effect were controversial and have still not been well defined. These observations led to experiments in which it was demonstrated that supernatants (SNs) from activated T cells could, in some instances, substitute for T cells in enhancing or suppressing the immune response (73). In retrospect, much of the confusion concerning suppression can be accounted for by two findings: (1) different subsets of T cells secrete a large array of different lymphokines which exert both helper and suppressor activities (74, 75); (2) soluble factors act primarily on cells which had been activated by a T cell/B cell cooperative interaction and which can induce the partially activated B cells to differentiate and, in some cases, can suppress this differentiation (50, 51). E. THEROLEOF THE MAJORHISTOCOMPATIBILITY COMPLEX IN T CELL/BCELLINTERACTION A large number of experiments in the 1970s established the relationship between immune response (Ir) genes, encoded by the major histocompatibility complex (MHC), and the ability of T and B cells to interact in the elaboration of an immune response (76-79). Molecules encoded by the MHC, in particular the I region of the MHC, influenced the ability of T cells to interact with macrophages and B cells (76, 78, 80-84). This implied that the products of Ir genes (now class I1 MHC genes) were involved in cellular interactions between T cells and B cells and indicated that in order for a T cell to recognize a B cell or an antigenpresenting cell (APC), recognition of class I1 molecules was necessary. T cell receptors (TcRs) involved in recognizing class I1 molecules were either the same or different from those recognizing antigen. In the last
6
ELLEN S. VITETTA ET AL.
few years, the basis for genetic restriction has been elucidated and the concept has emerged that a single TcR can recognize both the antigenic moiety and the MHC molecule on the surface of an APC (85, 86). It has also been shown that accessory molecules on both T and B cells influence the interaction of the two cells. F. ANTIGEN-SPECIFIC RECEPTORS ON T AND B CELLS Perhaps no issue in immunology has had a more colorful and controversial history than the elucidation of the antigen-specific receptors on T and B cells. Studies in the early 1960s demonstrated that B cells express antibody molecules that served as antigen-specific receptors (87-97). It was initially thought that the antigen-specific receptors were the same as the major class of immunoglobulin (Ig) in the serum, e.g., IgG (97). However, studies in the early and mid-1970s established that the major antigen-specific receptors on the B cells are monomeric IgM (98) and IgD molecules (99), both of which share the same antigen specificity (90, 100, 101) and idiotype (102-104), and each of which contains a C-terminal transmembrane domain responsible for anchoring the receptor in the membrane of the B cell (105-109). It was later reported that B cells can express other classes of Ig on their surface depending upon their state of differentiation (110-112) and that these receptors can also express transmembrane segments (112). Furthermore, B cells often express more than one isotype of surface Ig (sIg) (110-112). Based on previous studies, it was assumed that the Ig receptors on B cells are responsible for recognizing the haptenic portion of a hapten-carrier conjugate, whereas, the T cells express a receptor of a different specificity that recognizes portions of the carrier. For a long time, there was heated controversy concerning the nature of the TcR which divided the immunologic community into two camps, i.e., those who were convinced that the TcR was Ig (113) and those who were not (114). In 1980, pioneering experiments by Hedrick et al. (115) and Yanagi et al. (116) demonstrated that the receptor expressed on the T cell was indeed different from that expressed on the B cell. The major TcR consists of disulfidebonded a and @ chains, each having a M , of 45,000-55,000. The a and @ chains express transmembrane and cytoplasmic domains. It was further shown that the two chains of the receptor can be divided into variable (V) and constant (C) region segments (117-119) which show structural similarities to V and C domains of Ig (116, 120-122). Indeed, the TcR is encoded by V, D, and J segments of DNA, in a manner analogous to the Ig receptor on the B cell (123). The a@ dimer of the TcR recognizes both antigen and portions of the MHC molecule (123-125). Finally, it was shown that other surface molecules on the T cell
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
7
[such as cluster-designation (CD)3, CD4, CD8, lymphocyte functionassociated antigen (LFA)-I]interact with the crp complex or with molecules on the B cell (125, 126). More recently, 6 and y chains of the TcR have been identified but their role in antigen recognition has not been fully defined, even though T cells expressing 6y receptors have been identified (127-130). Most importantly, T cells, in contrast to B cells, do not recognize soluble native antigen, but recognize “processed’ antigens (131-136) or endogenously synthesized antigens (137) that are bound to the surface of APCs, e.g., macrophages, dendritic cells, Langerhans cells, B cells, or tumor target cells. In the case of exogenous antigen, the presented form of the antigen arises as a consequence of a series of intracellular processing events after the native antigen binds to and is internalized by the APC (138). It is thought that these processing events involve degradation and/or denaturation of the antigen, recycling of fragments of denatured antigen to the surface of the presenting cell, and association of the fragments with class I1 MHC-encoded molecules. Recent studies supporting this concept will be presented later in the chapter. G. THEANTIGEN-BRIDGING MODELOF T CELL/BCELL INTERACTION The original model postulated by Mitchison (16) to describe associative recognition of antigen by T and B cells suggested that receptors on B and T cells recognize the hapten and carrier portion of the native antigen, respectively, and, hence, antigen bridges the two cells. Following bridging by antigen, the T cell delivers activating signals to the B cell. This model has now been revised to include the current-day notion that the B cells bind the haptenic portion of the antigen, internalize the haptencarrier complex, degrade the carrier, and present peptide fragments of the carrier in association with a class I1 molecule to specific T h cells. Th cells also recognize peptides of the processed carrier (taken up nonspecifically)on the surface of a macrophage or other accessory cells (138). In fact, both Th and B cells may interact on the surface of a macrophage where the B cells recognize the native hapten and the T cells, the processed carrier. This becomes important in the physiological setting of a lymphoid organ where T cells and B cells are concentrated in different portions of the organ (139). In this system, it is attractive to postulate trafficking of cells through the lymph node to a common site where antigen is present in sufficient concentration to bind to the relevant cells and where these cells can interact. In the following sections of this review, we discuss the mechanisms underlying the hapten-carrier effect, T cell/B cell interaction, the role
8
ELLEN S. VITETTA ET AL.
of Th cells in inducing affinity maturation and Ig class switching, genetic restriction, T cell suppression, and the role of accessory cells and soluble mediators in T cell/B cell interaction and activation. Furthermore, the role of the MHC will be discussed both in the context of processing and presentation of antigen to T cells by APCs and in the context of T cell/ B cell interaction. The focus of this review will be on the mechanisms involved in the interaction between T and B cells, and the functional outcome of these interactions. Ii. Antigen Processing and Presentation
A. REQUIREMENT FOR PROCESSING Between 1970 and 1980, experimental findings related to antigen processing and presentation established that (1) T cells are required for the generation of antibody by the differentiated progeny of activated B cells responding to TD antigens; (2) B and T cells do not recognize the same epitope on a TD antigen; (3) the antigenic epitopes recognized by T and B cells must be physically linked on the antigen in order for the antigen to elicit an antibody response; and (4) T cells cannot bind native, soluble antigen, but have receptors for “processed” antigen associated with class I1 molecules on APCs. These findings suggested that APCs must first bind native antigen and, at a later time, present some other form of this antigen to a T cell in an antigen-specific, MHCrestricted manner. As a result of the interaction of the T cell with an APC, the T cell becomes activated. The term “presentation” can be used to define the capacity of APCs to express altered forms of antigen to T cells. Antigen “processing” describes the steps by which an APC converts native antigen to a form which can be recognized by a T cell. Processing may involve proteolytic degradation, denaturation, or modification of the antigen. Processing also involves the association of antigenic fragments with class I1 molecules and the expression of these complexes of processed antigen and class I1 molecules on the surface of APCs. The earliest studies to correlate antigen catabolism with presentation were carried out by Ziegler and Unanue (140), who exposed macrophages to the bacterium Listeriu monocytogenes. Following extensive washing of these treated macrophages, Listenu-primed T cells were added to the cultures and were allowed to bind. Unbound T cells were decanted from the cultures and the depletion of antigen-specificcells was assessed. These experiments showed that (1) a processing period of approximately 1 h is required before the T cells can bind to the antigen-pulsedmacrophages;
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
9
(2) antigen presentation is decreased after treatment of the APCs with lysosomotropic agents: (3) presentation is inhibited when the macrophages are fixed prior to exposure to antigens; and (4) after the APC has been exposed to antigen for at least 1 h , subsequent fixation or treatment with lysosomotropic agents no longer affects their ability to present antigen. These experiments also demonstrated that the binding of T cells to antigen-presenting macrophages is “restricted” by class I1 molecules and that the native antigen is ingested and catabolized by the macrophages.
TYPES OF APCs B. DIFFERENT The macrophage was the first APC to be identified. Monocytes/ macrophages exist both in the circulation and in the tissues, and together with polymorphonuclear leukocytes (PMNs) form the first line of defense against foreign pathogens. Macrophages have the ability to engulf particulate and soluble antigens, to degrade them, and to present antigen fragments to T cells (138). More recent studies have demonstrated that in addition to macrophages, other cells can also act as effective APCs. In general, all cells capable of presenting antigen to T cells constitutively express class I1 antigens on their surface or can be induced to do so. These cells include dendritic cells (141-144), Kupffer cells (145), Langerhans cells (146-148), vascular endothelial cells (146), Schwann cells (149), astrocytes (150), thymic stromal cells (151), human dermal fibroblasts (152), B lymphocytes (153), and human class 11+ T cells (154). Although the role.played by these APC cells in mu0 is not completely understood, evidence from a number of laboratories suggests that macrophages, B lymphocytes, and dendritic cells are the major APCs present in lymphoid tissues. Steinman and co-workers (141-144) have demonstrated that dendritic cells are very efficient APCs. In fact, their studies suggest that the dendritic cell may be a more important cell than the macrophage for the initial activation of resting T cells in viuo. Another special type of APC is the follicular dendritic cell (FDC) (155, 156). This cell is unique in that it carries immune complexes of native antigen on its surface in a nonprocessed form. In vivo, the FDC may serve as a reservoir for native antigen which can be subsequently bound and processed by other APCs and, in particular, B cells. Since 1980, a number of groups have demonstrated that B cells can process and present antigen to T cells both in mlro and in mvo (153). This demonstration supports the earlier hypotheses of antigen bridging in T cell/B cell interactions. Once it became clear that B cells could bind native antigen, process it, and present processed fragments of
10
ELLEN S. VITETTA ET AL
antigen to T cells in an MHC-restricted manner, the mechanism of cognate interaction between B and T cells became clearer. In the remainder of this section we focus on antigen processing and presentation by B cells, a comparison of the ability of B cells versus other APCs to present antigen, and proposals concerning the roles of the different APCs in viuo. C. B CELLSAS APCs Benacerraf first hypothesized that B cells were major APCs (157). This was deduced from two findings: (1) B cells express high densities of class I1 molecules, which were thought to be involved in T cell/B cell communication, and (2) since the initiation of an antibody response requires the “bridging” of T and B cells, it would be logical to presume that B cells should function as APCs. The first report to clearly demonstrate that this was the case was that of Chesnut and Grey (158). Using rabbit antimouse Ig (RAMIg) as an antigen and rabbit Ig-specific T cells, they circumvented the problem of the low frequency of B cells which could specifically bind to a purified protein antigen, because RAMIg can bind to all sIg+ B cells while antigen binds to
CELLULAR INTERACTIONS IN T H E HUMORAL IMMUNE RESPONSE
11
of irradiating the presenting B cells (which also proliferate as a consequence of T cell-mediated activation). Under these conditions, normal splenic B cells depleted of T cells and macrophages were unable to present KLH to T cells except at very high antigen or cell concentrations. In contrast, both the BAL and the macrophages were very efficient APCs and induced significant levels of IL-2 secretion from the T cells at 100-fold lower cell numbers. When normal primed T cells (instead of T hybridoma cells) were used as responding cells, the normal B cells were, again, very poor APCs for KLH. Although the BAL were better APCs than the B cells, they were 50 times less efficient than splenic adherent cells. Two issues were raised by the aforementioned experiments. (1) There were marked differences in the capacity of different B cells to present antigen to T cells and this appeared to be dependent upon their state of activation; i.e., resting B cells were less effective than activated or neoplastic B cells. (2) There was a dichotomy between the ability of B cells to present antigens which they could bind specifically via sIg, e.g., RAMIg, versus nonspecifically, e.g., KLH and ovalbumin (OVA). This latter point raised two further questions: (1) whether presentation of antigen by resting B cells was physiologically relevant under conditions where the B cells did not express sIg receptors specific for the antigen in question, and (2) whether resting B cells could process irrelevant antigen (as opposed to RAMIg) as effectively as macrophages. Between 1983 and 1985,a number of papers were published with conflicting views as to whether resting B cells could act as APCs. Cowing and co-workers(163,164)and Grey and co-workers(131,165-167)showed that resting B cells could not act as APCs, while Parker and associates (168,169) and Pierce and colleagues (170)demonstrated that they could. This controversy was partially resolved by Ashwell et al. (171)who showed that resting B cells could present antigen, but that this event was radiosensitive. This was subsequently confirmed in several reports (163,172, 173). In earlier experiments by some groups, the resting B cells had been irradiated prior to using them as APCs. Although studies have now confirmed the exquisite radiosensitivity of the presentation process in resting B cells, this may not be the entire explanation for the lack of antigen-presentingactivity reported previously. In 1985,Krieger and co-workers (167)irradiated B cells with 500 rads and subsequently examined them for their ability to act as APCs following Percoll density gradient fractionation into cells of low and high density. Only the low-density cells were capable of eliciting a significant T cell response. These results conflicted with those of Parker and co-workers (169, 174), who separated B cells on the basis of density using a
12
ELLEN S. VITETTA ET A L .
countercurrent elutriation system and found that the dense cells could act as APCs. The difference in the two studies may be attributed not to irradiation of the B cells, but to differences in the antigens used. In the experiments of Krieger and co-workers (167), OVA was used as the antigen, while in the studies of Parker and colleagues (169, 174), RAMIg was used. Thus, the difference in efficiency of uptake of these two different antigens, which were (RAMIg) or were not (OVA) bound to the resting B cells via sIg, undoubtedly influenced the ability of the B cells to act as effective APCs. Hence, antigens bound to sIg on resting B cells may be more effectively internalized and/or routed to different intracellular compartments than are antigens which are taken up by fluid-phase pinocytosis. Furthermore, fluid-phase pinocytosis is probably a relatively inefficient process in resting B cells and would not result in uptake of large enough quantities of irrelevant antigen unless the concentrations of antigen in the medium were very high. In contrast to resting B cells, large or activated B cells may pinocytose antigen more efficiently.
D. ANTIGENPROCESSING A number of studies have addressed the issue of what intracellular events are involved in “antigen processing” and how these events compare in B cells and macrophages. The picture that has emerged, although not yet complete, suggests that antigen processing takes place intracellularly, is time- and temperature-dependent, and is inhibited by agents that interfere with lysosomal function. In addition, processing involves recycling of antigenic fragments to the surface of the APC. Finally, B cells are more effective at binding and presenting specific antigen because of their ability to capture antigen via sIg molecules. However, processing events in macrophages and B cells may be very similar, although it is possible that B cells process sIg-bound antigens in endosomes while macrophages process antigens in lysosomes. In addition, antigen taken up by fluid-phase pinocytosis in B cells may also be routed to lysosomes. 1. Intracellular Events
Different investigators have established that antigen processing is an intracellular event, since metabolic inhibitors, inhibitors of secretion, and inhibitors of lysosomal enzymes can prevent processing. Furthermore, cells which cannot internalize antigen, or cells from which bound antigen is removed by enzymes, cannot present antigen. In this regard, Metezeau et al. (175) demonstrated that anti-Ig antibodies bound to B cells are rapidly internalized, but that the endocytic process is blocked either by incubation at 4OC or by treatment with metabolic inhibitors
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
13
such as azide and 2-deoxyglucose. Furthermore, Fab’-anti-Ig is routed primarily to endosomes while divalent anti-Ig is routed to lysosomes. Hence, the degree of cross-linking of sIg molecules on B cells may determine how antigen-sIg complexes of different sizes are routed intracellularly. The approach used by Unanue and others (176-178) has been adapted to analyze internalization of antigen by B cells (131). i.e., cells are treated with antigen at 4OC and then washed extensively. After warming the cells to 37OC, they are fixed to prevent further internalization. Cells are then treated with proteolytic enzymes and the amount of antigen released from the cell surface is determined. Any remaining cell-associated antigen is assumed to be intracellular. Results from such studies confirmed that specifically bound antigen is rapidly internalized while nonspecifically bound antigen remains on the cell surface (131). A similar approach has also been used to measure the intracellular transit time and reappearance of processed antigen on the cell surface of macrophages and B cells. The first studies of this type were carried out by Ziegler and Unanue (140) using macrophages pulsed with Listenu. This study established two important points: (1) light fixation of APCs (1% paraformaldehyde) blocks antigen processing, but does not inhibit antigen presentation after processing has ocurred; and (2) macrophages require 30-60 min after pulsing with antigen before they can present it to T cells. In this case, antigen presentation was measured by the direct binding of T cells to macrophages. More recently, Lanzavecchia and co-workers (179) have analyzed the kinetics of antigen processing and presentation by antigen-specificB cells by measuring Ca2+fluxes in the T cells; Ca2+fluxes occur very rapidly once a T cell binds to antigen on an APC. Using a tetanus toxoid (TT)specific B cell line and specific antigen, it was confirmed that a processing period of 60 min was required before the APCs would present specifically bound antigen to T cells. Taken together, the results of Ziegler and Unanue and Lanzavecchia demonstrate that a period of 30-60 min is required before antigen bound to either B cells or macrophages can be presented in a form which the T cell can recognize. Information concerning intracellular routing of antigen has been obtained by using inhibitors of proteolysis or agents which raise the pH of lysosomal compartments (e.g., NH&l or chloroquine) and inhibit the activity of lysosomal enzymes. These agents block antigen processing and/or presentation and imply that processing occurs in lysosomes and involves proteolysis. However, it should be noted that both chloroquine and N h C l have multiple effects on cells, such as interfering with recycling of intracellular vesicles and fusion of endosomes with lysosomes
14
ELLEN S. VITETTA ET A L
(180-182). In order to understand how these inhibitors effect processing, Ziegler and Unanue (183) treated macrophages with 0.01 M NH4C1 or 0.1 mM chloroquine and found that this treatment had a negligible effect on antigen binding and ingestion. Likewise, chloroquine or NH&l added after antigen processing had a minimal effect on the binding of T cells to presenting macrophages. However, when the macrophages were treated with lysosomotropic agents prior to treatment with antigen, both presentation and catabolism of antigen were reduced by greater than 50%. The results of these experiments suggest that NH4Cl and chloroquine do not effect T cell/macrophage interactions, but do effect antigen degradation and/or recycling of antigen to the cell surface. The effect of chloroquine on antigen presentation by B cells and macrophages was studied by Grey and co-workers(131, 165). Their results demonstrated that 40 p M chloroquine is sufficient to totally ablate antigen presentation by B lymphoma cells and macrophages from peritoneal exudates. Furthermore, they demonstrated that chloroquine only inhibits antigen processing and not presentation; once an antigen had been processed by B lymphoma cells, subsequent addition of chloroquine has no effect on the presentation of this antigen, even though a second antigen added following chloroquine treatment can no longer be presented. These and similar findings (184) using B cells must be interpreted with caution in light of recent observations that B cells are far more sensitive to the toxic effects of lysosomotropic amines than are macrophages. In this regard, C.D. Myers and E.S.Vitetta (unpublished data) have shown that concentrations of chloroquine as low as 0.1 mM cause irreversible damage to B cells in 3-4 h. Furthermore, Nowell and Quaranta (185) demonstrated that incubation of B cells with 1.0 pM chloroquine or 10 mM NH4Cl inhibits transport of newly synthesized class I1 molecules to the cell surface, resulting in the intracellular accumulation of nascent molecules. Considering the major role of class I1 molecules in antigen presentation, the notion that chloroquine effects only lysosomal catabolism of antigens should be reevaluated. It is important to note that monensin (131) and xyloside (186), two agents which block antigen presentation, prevent the transpot of newly synthesized proteins to the cell membrane.
2. Processing by Macrophages versus B Cells Although the studies mentioned above demonstrate that both B cells and macrophages are able to process antigens and present them to T cells, it was not clear whether B cells and macrophages handled antigen differently. In this regard, Grey and colleagues examined the fate of radiolabeled proteins bound to the surfaces of B lymphoma cells or
CELLULAR INTERACITONS IN THE HUMORAL IMMUNE RESPONSE
15
macrophages (165). Conversion of the radiolabel from acid precipitable to acid soluble was used as a measurement of protein degradation. Using radiolabeled RAMIg and KLH as antigens, macrophages handled them in a similar fashion; over the course of 20 h, SO-SO% of the proteinassociated radioactivity initially bound to the cells was released into the supernatant (SN) in an acid-solubleform, and 20% in an acid-precipitable form. In contrast, in the B lymphoma cells, the processing of RAMIg and KLH was different. Radiolabeled RAMIg was processed with kinetics similar to those reported for macrophages; 50 % of the protein was released in 20 h in an acid-soluble form and 20% was released in an acidprecipitable form. Analysis of antigen fragments on sodium dodecyl sulfate (SDS)polyacrylamide gels showed that both macrophages and B lymphoma cells degraded intact RAMIg (M,= 150,000) into fragments of 50,000. However, when the B lymphoma cells were pulsed with low concentrations of radiolabeled KLH, almost 90% of the labeled protein remained cell bound and acid precipitable. These findings confirmed earlier results that B cells can degrade a specifically bound antigen, e.g., RAMIg, in a manner analogous to that observed for macrophages. However, antigen not bound to sIg is not readily internalized or degraded. Chesnut and co-workershad previously compared processing of antigen by B cells and macrophages (131) using the secretion of IL-2 by T hybridoma cells as an assay for antigen presentation by B lymphoma cells or macrophages. Their studies demonstrated that paraformaldehyde furation effectively blocks presentation of antigen by B cells within 45 min after its addition to the cells. Thus, some time-dependent event is required prior to presentation. Antigen which had bound to the cells at 4OC could be entirely removed by pronase treatment immediately after binding, but became pronase-resistant after the cells were warmed to 37OC, suggesting that antigen is rapidly internalized. Both macrophages and B cells are able to present high concentrationsof either native or denatured OVA to an OVA-specific T hybridoma cell and this presentation can be blocked by either lysosomotropic agents or carboxylic ionophores, e.g., chloroquine or monensin. Taken together, these findings provided further evidence that any antigen bound to a macrophage or B cell in high enough concentrations can be internalized and denatured, and that B cells do not have to take up antigen via sIg if the antigen is present in high enough concentrations to be internalized by fluid-phase pinocytosis. E. ANTIGEN PRESENTATION BY ENRICHED POPULATIONS OF ANTIGEN-SPECIFIC B CELLS Although nonspecific pinocytosis of antigens by resting B cells is generally inefficient as compared to the same events in macrophages, B cells have a significant advantage over macrophages in capturing
16
ELLEN S. VITETTA ET A L
specific antigen via sIg. It could be argued that presentation by B cells of antigens which do not bind via sIg is an in vitro artifact resulting from the high concentrations of both antigen and cells used in culture. Thus, in a physiological setting, B cells might present only their specific antigen to T cells. As mentioned above, studies by Grey and co-workers (153) as well as Parker and co-workers (168, 169, 174) have examined presentation of antigen by B cells using the “polyclonal” antigen, RAMIg. In both cases, the specific antigen (RAMIg) was presented at 10,000-fold lower concentrations than was nonspecific antigen (normal rabbit Ig). The first studies to use an entirely antigen-specific system were those of Abbas and associates (172, 187-190) and Lanzavecchia (184, 191, 192). Abbas and colleagues purified trinitrophenyl-antigen-binding cells (TNP-ABCs) from the spleens of either primed or unprimed mice on TNP-gelatin (172, 187). When the gelatin was warmed and the bound cells were recovered, 20-40oJo could bind to TNP. These cells were analyzed for their capacity to present either TNP-derivatized or native poly(g1ycine-leucine-phenylalanine) (GL4) to GL$-specific, I-Edrestricted T hybridoma cells. As previously demonstrated using RAMIg as antigen, TNP-ABCs could present TNP-GL4 (as compared to GL4) to T cells at least 1000-fold more efficiently; presentation was I-Edrestricted. Likewise, the importance of antigen uptake by sIg on the B cell was demonstrated by blocking (TNP-GL4) uptake with anti-Ig or TNP-human serum albumin (HSA) (187, 188). Further studies by the same group demonstrated that TNP-ABCs from unprimed mice had properties similar to those of TNP-ABCs from primed mice (172). The hapten and carrier specificity of these responses were further demonstrated by using both OVA and GL4 as carrier molecules, TNPspecific and fluorescein isothiocyanate (F1TC)-specificB cells, and TNPor FITC-haptenated antigens (172). Finally, Abbas and co-workers(172, 187-190) have used TNP-ABCs to demonstrate that presentation of antigen by the TNP-ABCs is specific at low antigen concentrations (when antigen binds to sIg) and nonspecific at high concentrations (when antigen is not taken up by sIg molecules). As described previously, Lanzavecchia (184) utilized TTspecific human B cell clones which had been transformed by Epstein-Barr Virus (EBV) to present antigen to TT-specific T cell clones. The TT-specific B cells were able to present their antigen at 104-foldlower concentrations than were cells from EBV-transformed B cell lines not specific for TT. Likewise, when a second irrelevant antigen was used, purified protein derivative (PPD), the TT-specific B cells presented antigen only at high concentrations. Lanzavecchia (192) further demonstrated that TT-specificB cells are capable of presenting antigen at concentrations lower than the
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
17
dissociation constant (Kd)of the sIg on the B cells, suggesting that only a very small fraction of the sIg molecules is necessary to take up the amounts of antigen necessary for subsequent presentation. These findings led to the proposal of the “vacuum cleaner” model of antigen presentation in which B cells “concentrated” small amounts of antigen from their environment, processed it, and presented fragments (at a high enough concentration) on their surface to initiate interaction with specific T cells (192). BY ENRICHED POPULATIONS OF F. ANTIGENPROCESSING ANTIGEN-SPECIFIC B CELLS
The same approaches described above for the study of presentation by antigen-specific B cells have been used to analyze processing. Lanzavecchia (192) has shown that presentation of antigen by TT-specific human B cell clones follows many of the same patterns described for normal splenic B cells and RAMIg. Hence, B cells present specific antigen at very low concentrations and presentation requires a period of time for processing. Processing can be inhibited by fixation or by treatment with 10 p M chloroquine. In these experiments, the B cells were EBV transformed and were therefore radioresistant. Unfortunately, no direct comparison of presentation of antigen and anti-Ig has been reported in this system. One direct approach to the analysis of intracellular events occurring in antigen processing is the measurement of protein degradation by the B cell. C.D. Myers and E.S. Vitetta (unpublished results and Ref. 193) have used TNP-ABCs purified by rosetting spleen cells from normal mice with TNP-horse erythrocytes (194, 195). From 60 to 80% of the TNP-ABCs bind specific antigen, and some of the properties of these cells are summarized in Table I (196). Although TNP-ABCs can degrade both anti-Ig and TNP-proteins, the kinetics of degradation are different. TNP-ABCs treated with anti-Ig degrade proteins in a manner similar to that previously reported for both unpurified B cells and macrophages. TABLE I GENERAL CHARACrERISTlCS OF ANTIGEN-SPECIFIC B CELLS 1. 2. 3. 4. 5.
From 60 to 90% of the cells bind TNP (194, 195, 369) From 80 to 90% of the cells are sIg+ (194, 369) Greater than 95% of the cells are in Go (369) Greater than 92% are small B cells (as determined by cell volume) (369) Of the cells, 1/2 grow in response to dextran sulfate plus LPS (194) and 1/15 secrete IgM in response to T h cells and TD antigen (369, 453)
18
ELLEN S. VITETTA ET A L .
However, when specific antigen is used, degradation is more rapid and reaches a plateau at about 4 h. The percentage of antigen released into the culture medium as acid-solublefragments is only 30% (of the antigen bound) as compared to 60% when anti-Ig is used. It is important to note that the kinetics of release of acid-solublefragments of antigen into the medium parallel the kinetics of antigen presentation as determined by the formation of specific T cell/B cell conjugates (197) (to be discussed later). OF ANTIGEN FRAGMENTS WITH CLASS I1 G. INTERACTION MOLECULES The experimental approaches thus far described suggest that protein degradation and denaturation are important steps in antigen processing. In addition, the requirement for expression of class I1 molecules by APCs suggests that there is a relationship between processed antigen fragments, class I1 molecules, and antigen presentation. An approach used by many groups to demonstrate that APCs present fragments of antigen has been to fix cells, to incubate them with either native or partially degraded proteins, and to evaluate their ability to present proteins versus protein fragments to T cells. Allen (198) has reviewed data based on similar approaches and described three different types of antigen: (1) those which require proteolytic degradation or cyanogen bromide fragmentation before they can be presented by fixed cells; (2) those which require only reduction of disulfide groups to allow presentation by fixed cells; and (3) those which can be presented as native antigen by fixed cells. Other groups have described protein structures which require different degrees of degradation prior to presentation by fixed cells (133, 199-202). Using fmed cells or artifical membrane systems such as those described by Watts and McConnell(203), a number of groups have analyzed the structure of “minimal peptides” required for antigen presentation by APCs (84, 132, 136, 204-211). Schwartz et al. (82, 212-219) analyzed synthetic peptides of cytochrome c for their ability to stimulate T cells (when presented on APCs). By using a series of synthetic peptides which differed by one amino acid, and analyzing both the antigen and cell concentrations required for efficient antigen presentation, it was proposed that specific portions of the antigen fragment interact with either class I1 molecules on the APCs and/or the TcRs. Minimal peptide structures required for antigen presentation and binding of antigen fragments to class I1 molecules have been analyzed by Berzofsky and co-workers (220-224), who have demonstrated a common structural theme for many of these peptides, i.e., the majority of antigenic fragments recognized
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
19
by T cells form an a-helical structure with an amphipathic nature. This structure, which was initially described for a fragment of sperm whale myoglobin, has been postulated to be important in immunodominance, i.e., the selection of T cell epitopes. This could be achieved by one of two mechanisms: (1) the amphipathic nature of the peptide may be specifically required for the binding of an antigen fragment to a class I1 molecule; or (2) since amphipathic fragments have a hydrophobic face, they may associate with the cell membrane, effectively producing a reservoir of fragments for class I1 molecules to bind to, either independently of the TcR or in its presence. The second possibility is supported by the data of Falo et al. (225, 226), who demonstrated that APCs treated with phospholipase (following antigen processing) lose their ability to present antigen but that the expression of class I1 molecules on the cell surface is unaffected. These data suggest that phospholipase treatment removes the antigen fragments from the surface of the APCs because of their weak binding to the cell surface. Alternatively, data from the same group (225) suggest that in vim processing includes the addition of a lipid moiety to antigen fragments, resulting in a covalent, phospholipase-sensitive attachment to the membrane. In an attempt to demonstrate that antigenic fragments can bind directly to MHC class I1 molecules, Babbitt et al. (135, 227) purified class I1 molecules from cells and performed equilibrium dialysis in the presence of synthetic peptides of antigen. In this and similar studies by Grey and colleagues (134, 136, 206, 228, 229) it was demonstrated that peptides had a measurable affinity for class I1 molecules and that binding was MHC-restricted with regard to the class I1 antigen used. However, since the K d of peptides and class I1 molecule is relatively low ( M), questions concerning the physiological relevance of the specific binding of antigenic peptides to class I1 molecules were raised. Hence, it is likely that the concentration of specific antigen fragments on the surface of APCs is too low to allow effective binding to class I1 molecules. Nonetheless, class I1 molecules could bind to peptides in a specific intracellular compartment (e.g., endosomes), where high concentrations of both peptides and class I1 molecules would be present. In this regard, experiments by Cresswell (230) demonstrated that endosomes (in which transferrin receptors are recycling) fuse with other endosomes carrying nascent class I1 molecules in transit between the Golgi apparatus and the cell surface. If antigen fragments are generated in similar endosomes recycling to and from the cell surface as suggested by several studies [(183, 193, 231, 232) and C.D.Myers and E.S.Vitetta, unpublished observation], it is possible that class I1 molecules and antigen fragments may associate and recycle to the surface in the same vesicles.
20
ELLEN S. VITETTA ET A L
This would provide the high local concentration of antigenic fragments necessary for binding to class I1 molecules. As mentioned above, this hypothesis is supported by the finding that chloroquine inhibits antigen presentation at concentrations which have no effect on antigen degradation by B cells, but which inhibit the transport of class I1 molecules to the cell surface [(185) and C.D. Myers and E.S.Vitetta, unpublished observation]. In any case, it is possible that random, transient, low-affinity interactions between class I1 molecules and peptides could be stabilized by simultaneous binding to TcRs. OF CLASS I AND CLASSI1 MOLECULES AND H. A COMPARISON THEIR ROLEIN ANTIGENPRESENTATION Bjorkman et al. (233, 234) have produced an X-ray crystallographic model of an MHC class I molecule and, from this, have hypothesized its three-dimensional structure. The class I molecule has a large cleft on its upper surface which might be the antigen-binding site. After the assignment of all the amino acids in the sequence of the class I molecule, there remained electron-dense material, apparently residing in the proposed antigen-binding cleft. This material may represent bound antigen peptides, although this has not been directly demonstrated. Nevertheless, if this material does represent antigen, it would suggest that most, if not all, class I molecules on the cell surface already have antigen fragments bound to them with sufficient affinity so as not to be removed during purification and crystallization. In any case, these findings have been extrapolated to class I1 molecules, for which hypothetical model structures have now been produced (235). In this regard, it is important to mention certain differences between antigen presentation by MHC class I and class I1 molecules, since the data described here have been extrapolated to antigen presentation in the context of class I1 molecules on APCs. Class II-restricted antigen presentation invariably involves CD4+ (human) or L3T4+ (mouse) cells, which usually function as T h cells (236), although there have also been reports of class II-restricted, CD4+/LST4+ T cytotoxic (Tc) cells (237). CD8 (human) Lyt-2+ (mouse) cells are invariably class I restricted and usually function as Tc or Ts cells (236), although there are also reports of CD8+/Lyt-2+ cells that can provide help for B cells (238). Furthermore, CD4+, class I-restricted cells have also been described (239). In general, there is a difference in the nature of the antigen presented in the context of class I or class I1 molecules on APCs. As described above, class II-restricted cells are primarily involved in the recognition of antigen which has been endocytosed by the cell, processed in endosomes or lysosomes, and then recycled to the surface (240, 241). Mechanisms by +
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
21
which these antigenic fragments could become associated with class I1 molecules have been described in previous sections. In contrast, many class I-restricted T cells recognize antigen that has been synthesized by the cell (e.g., viral proteins or minor histocompatibility antigens) (240, 241). The interaction between newly synthesized antigens and class I molecules could occur during synthesis and assembly of both molecules. This possibility is supported by the finding of material in the proposed antigen-binding cleft of the crystallized class I molecules analyzed by Bjorkman and colleagues (233, 234, 242). However, a direct demonstration of the binding of purified class I molecules to viral or other antigens has not yet been carried out. OF ANTIGEN PRESENTATION in Vitro I. A SUMMARY
The results of studies carried out over the last decade have shown that B cells can serve as APCs and that they are most efficient at presenting antigens which they bind via sIg. A period of time for processing is required before antigen is presented to T cells. During this time, the antigen is denatured and/or proteolytically cleaved into small peptides, some of which are presented, but most of which are released into the culture medium. Very little is known about the intracellular mechanism(s) by which this proteolytic cleavage occurs or in which intracellular sites it takes place, although both endosomes and lysosomes have been implicated. However, fragments are in some manner recycled to the surface and become weakly associated with class I1 molecules. This information is summarized in the model depicted in Fig. 1. Receptors on T cells bind to the complex of antigen and class I1 molecules and this binding results in T cell activation. The activated T cells then activate the antigen-presenting B cells with which they specifically interact. J. THEB CELLAS
APC in Vivo Although it has been demonstrated that B cells can act as APCs in vitro, their role in vivo remains less well defined. Several lines of evidence indicate that B cells play a major role in antigen presentation, leading to T cell activation in mvo. Inaba and colleagues (141, 142) and Lassila et al. (243) have demonstrated that dendritic cells are critical for antibody responses to TD antigens in both mouse and man. Although it has not been definitively demonstrated that dendritic cells process antigen, it is possible that they bind fragments of antigen released by macrophages. Whatever the case, dendritic cells are the only APCs capable of eliciting a response when unprimed, resting T cells are used in m*tro(141, 142, 243). These results suggest that dendritic cells may be the first APCs responsible for activating T cells from unprimed AN
22
ELLEN S. VITETTA ET A L .
d
k L l *
FIG. 1. Antigen processing and presentation by B cells. (a) Antigen: in B cells, antigen uptake is mediated by sIg. At very high antigen concentrations, antigen uptake may occur by pinocytosis without specific binding to sIg; this leads to polyclonal activation of the antigen-presenting B cells following their interaction with T cells. The source of antigen could be soluble protein or immune complexes. These could be in solution, on the surface of FDCs or icoccomes, or on macrophages. (b) sIg: sIg on B cells plays a role in the specific binding of antigen and increases the efficiency of antigen capture. sIg (or fragments of it) may remain bound to antigen during processing and affect the peptides presented by protecting certain epitopes or making other epitopes more accessible to proteases in intracellular compartments. Cross-linkingof sIg on the B cell results in the induction of early activation events. Increased levels of class I1 molecules could also increase the efficiency of presentation. Finally, changes in the metabolism of the cell may result in the activation of enzymes involved in antigen processing, recycling, and presentation. (c) Intracellular routing: antigen is internalized into vesicles and is either processed in the endosomes and/or lysosomes. ( d ) Extent of antigen degradation: it is unclear how much antigen degradation is required for antigen presentation. In some situations, peptides of 8-10 amino acids are presented, while in others large fragments of several hundred amino acids may be presented without further processing. It is possible that a lipid or peptide may be attached to the antigen fragment to facilitate binding either to class I1 molecules or to the cell membrane. (e) Interactions between class 11 molecules and processed antigens-source of class I1 molecules: class I1 molecules may be derived from the cell surface or may be newly synthesized. In this instance, nascent class I1 molecules trafficking from the Colgi apparatus may interact with antigen fragments generated in the endosomes. Some fragments of antigen may have an affinity for the plasma membrane, either because of their amphipathic nature or because of the attachment of a lipid. These fragments would then recycle to the cell surface where they would be present at a high enough concentration to interact with class I1 molecules. The interaction of peptides and class I1 molecules may be further stabilized by the binding of TcRs to the class Wantigen complex.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
23
animals and that B cells (or macrophages) might present antigen only to activated, partially activated, or memory T cells. The general approach that has been used to analyze the importance of B cells as APCs in mvo has been to deplete mice of B cells (or B cells of certain specificities) and then to analyze the capacity of these mice to generate activated T cells. Using anti-p-treated mice which are deficient in B cells, Janeway et al. (244) demonstrated a requirement for B cells in the generation of T h cell activity in lymph nodes. These results were confirmed and extended by Ron et al. (245), who treated mice from birth with anti-p to deplete B cells and subsequently primed these mice with antigen plus adjuvant. Popliteal lymph nodes were then removed and the capacity of the T cells to proliferate in response to antigen on APCs was analyzed. T cells obtained from anti-p-treated mice were incapable of proliferating in response to antigen, even though they maintained their mitogen responsiveness. In addition, the APC activity of cells from the anti-p-treated mice was normal, as demonstrated by their ability to stimulate T cells from immunized mice. In contrast, the reciprocal experiment using T cells from anti-p-treated mice and normal mouse APCs was negative, demonstrating that the deficiency was due to a lack of T cell priming. These results were confirmed and extended in a second series of experiments which were carefully controlled for the presence of contaminating cells and clearly demonstrated a requirement for B cells in the priming of T cells in mvo (246). One alternative interpretation of these experiments is that a defect exists in the handling of antigen by macrophages or dendritic cells in anti-p-treated mice. Unless this is the case, results of the above experiment seem to contradict the conclusions of Steinman and associates (141, 142) that dendritic cells are the only cells which can activate a naive T cell. An alternative hypothesis which is supported by the work of Ron and Sprent (247) is that dendritic cells initiate the activation of naive T cells, but that B cells are necessary to further stimulate or clonally expand these T cells. Abbas and co-workers(248) have also analyzed cells from anti-p-treated mice in an antigen-specific system. In their experiments, TNP-ABCs were adoptively transferred into the treated mice in order to prime T cells in vivo. This approach permits a more direct assessment of the role of B cells in priming T cells, since it is possible to compare specific versus polyclonal T cell responses. A high level of antigen-specific reconstitution was found. Nevertheless, in such experiments there was a small but reproducible T cell response to an unrelated antigen which was attributed to nonspecific binding of antigen to B cells of irrelevant specificity (Lea,not sIg mediated) and subsequent presentation to T cells. However, these data could also support the notion of an initial round of
24
ELLEN S. VITETTA ET A L .
presentation of antigen by dendritic cells followed by clonal expansion of those T cells which can interact with, and activate, antigen-presenting B cells. The antigen-presenting capacity of germinal center B cells has also been examined by Tew and co-workers (249), who studied the fate of injected antigen in immune mice and demonstrated that the antigen was very rapidly bound to circulating antibody to form immune complexes which were transported to the draining lymph nodes. In the nodes, complexes were found on the surface of FDCs. These complexes were not processed, and with time the tips of the dendritic processes pinched off and formed small immune complex-coatedbodies (icoccomes). These icoccomes were endocytosed by both B cells and macrophages, which subsequently appeared to act as APCs in the germinal center. In a very recent study, germinal center B cells were primed in muo, highly purified, and used as APCs in Vitro. It was found that 1-8 days following secondary inoculation of mice with OVA, germinal center B cells were still capable of presenting antigen to T cells in the absence of additional antigen pulsing. Therefore, germinal centers may retain antigen for long periods of time using FDCs or icoccomes as antigen reservoirs. It remains unclear whether this process occurs in a primary response where immune complexes are limiting, and what role this pathway plays in the activation of T cells in mvo. 111. Helper T Cells
In the preceding section we discussed how B cells, macrophages, and other APCs process and present antigen, in association with class I1 molecules, to T cells, leading to their activation. Activated T cells play a fundamental role in the regulation of humoral and cellular immune responses through direct cell/cell contact and/or release of lymphokines. Three types of functionally distinct T cells have been described: Th cells, Ts cells, and Tc cells. T h cells mediate a variety of immunological functions and generally regulate the immune response. Th cells mediate delayed-typehypersensitivity (DTH) and help for B cellmediated antibody responses, and are involved in the activation of Tc cells. Ts cells generally down regulate the immune response and Tc cells are responsible for killing virus-infected cells, tumor cells, and other cells expressing foreign antigens on their surface. Over the past 15 years, T cell subsets which mediate specific functions have been identified and their interactions have been studied. The picture that has emerged is complex and incomplete, but supports the contention that (1) T cells mediate their responses by both cell/cell contact and by secreted
CELLULAR INTERACHONS IN THE HUMORAL IMMUNE RESPONSE
25
lymphokines; (2) activation of T cells is related to the type of antigen and its route of administration; and (3) the regulation of T cell activation and proliferation is controlled in part via lymphokine feedback circuits. The delineation of subset-specific T cell surface markers by Cantor and Boyse (250) contributed three important concepts to immunology: (1) functional heterogeneity exists among T cells; (2) functionally different T cells can be distinguished based on their expression of cell surface markers ( T cells that help antibody responses are of the Ly-1+2-3phenotype in mice and CD4+ in humans, while T cells that mediate cytotoxicity or suppression are Ly-1-2 3 or L3T4- in mice and CD8+ in humans); and (3) diversification of T cell function takes place before, and independently of, antigenic exposure. +
+
A. T h CELLHETEROGENEITY Studies by Marrack and Kappler (251) and Janeway (252) suggested that additional heterogeneity existed among Th cells that provided help to B cells for antibody production. A number of studies (244, 253-257) later confirmed that T h cells were functionally heterogeneous and that two main types could be distinguished based on the type of help provided to B cells (251, 253, 254, 257). In general, one type of T h cells was MHC-restricted, carrier-specific,and augmented antibody responses via cognate interactions with B cells. The other type of T cell cooperated with B cells through the generation of “nonspecific soluble factors” and could generally activate multiple clones of B cells, irrespective of their antigen specificity. Both types of T h cells acted synergistically.A number of characteristics distinguished these two types of T h cells. Tada et al. (253) reported that T h cells of the first type were nonadherent to nylon wool and did not react with antisera raised against molecules encoded or controlled by the I region (class I1 antigens) of the MHC. The other type of T h cell was nylon-wool adherent and reactive with such antibodies. Imperiale et al. (257) reported that T cells (generated from the lymph nodes of KLH-primed mice) cultured at low density corresponded to the first type, whereas the second type was preferentially generated in cultures of higher density. In the last few years, with the development of techniques to clone and propagate functionally active and antigen-specific Th cells in vitro (258), their heterogeneity has become even more apparent. Kim et al. (259) studied the ability of a series of antigen-specific I-A-restricted T h cell clones to help B cells that secreted T-15 idiotype-positive or T-15 idiotypenegative antibodies. They reported the existence of four types of clones: the first two induced B cells to secrete antiphosphorylcholine (PC)
26
ELLEN S. VITETTA ET A L .
antibody, although the resultant antibodies induced by the two clones varied in the proportion of T-15 idiotype-positive antibodies; the third type of clone could induce B cell proliferation but not antibody secretion and actually inhibited the Ig secretion induced by other types of clones. The fourth type of clone was autoreactive, inducing B cell proliferation and antibody secretion in the absence of antigen. It has recently been reported that the pattern of lymphokines secreted by clones of established T h cells can delineate two types of functionally distinct cells (Table 11). Mosmann et al. (75) studied a large panel of clones of LST4+ MHC-restricted T h cells raised against a variety of antigens in a number of mouse strains. Virtually all clones fell into one of two groups, depending on the pattern of lymphokines secreted after mitogenic or antigenic stimulation and on the type of “help” they provided to B cells. Cells from one type of clone, designated as Thl, secreted IL-2, interferon-y (IFN-y), and IL-3, but not IL-4, and some Thl clones (but not others) could help B cells. The other type of clone, Th2, secreted IL-4, IL-5, and IL-3, but not IL-2 or IFN-y, and provided excellent help to B cells to induce the secretion of IgGl and IgE. This study not only demonstrated that functional subsets of T h cells could be distinguished based on their pattern of secreted lymphokines, but supported the notion that the differential secretion of lymphokines was at least partially responsible for the functional differences between the two types of Th clones. Subsequently, the list of lymphokines produced by each type of T h cell has been extended (260) (see Table 11). An alternative classification, proposed by Bottomly and colleagues (259, 261, 262), divides T h cells into “helper” and “inflammatory” subtypes. These TABLE I1 LYMPHOKINES SECRETED BY DIFFERENT MURINEThl AND Th2 CLONESa Lymphokine IL-2 IFN-./ LT IL-4 IL-5 IL-3 GM-CSF TNF LT
Thl
Th2
+ + +
-
-
-(?)b
+ + + +
+ + + +
f ( < Thl) -
OFrorn Cherwinski et al. (260). + (263).
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
27
two subtypes are for the most part equivalent to the Th2 and T h l subtypes, respectively, described by Mosmann et al. (75, 263). Finally, it should be noted that rare clones have been described which secrete lymphokines usually produced by only one or the other type of clone (263a; R. Fernandez-Botran, unpublished observations). Thus, it remains possible that either the delineation of all clones into T h l or Th2 types is generally, but not always, the rule, and/or that T h l and Th2 clones represent T h cells “frozen” at different stages of differentiation, i.e., that normal T h cells of one type can give rise to T h cells of the other type. Hence, these rare clones may represent “transitional” cells. B. FUNCTIONAL DIFFERENCES BETWEEN T h l AND Th2 CELLS Evidence has accumulated that clones representing the T h l and Th2 types of T h cells defined by Mosmann et al. (75) are functionally different. Cher and Mosmann (264) reported that only Thl cells can mediate antigen-specific DTH reactions. Other studies have analyzed the relationship between the nature of B cell help and the type of lymphokines secreted by T h cell clones (75, 260, 261, 264). As mentioned above, in the original paper describing the classification of T h cells into Thl and Th2, both subtypes were able to provide help to B cells for antibody responses, although Th2 clones were better helper cells than Thl clones. In this regard, Killar et al. (261) studied three functional subtypes of T h clones based on their ability to induce an in vitro antihapten antibody response to hapten-carrier conjugates. They reported that clones that failed to produce IL-4 could not provide help to antigen-specific B cells. If the IL-4-secreting T cells correspond to the Th2 cell type (helper), then the T h cells that failed to provide help would be of the Thl type (inflammatory), based on their failure to produce IL-4 (and their ability to secrete IFN-y). In contrast to these findings, Stevens et al. (265) reported that some clones of Thl cells can induce TNP-ABCs to secrete IgM, IgGS, and IgGz, antibody. Thus, there is a discrepancy with regard to the role of Thl cells in providing help to B cells that will subsequently secrete antibodies. Two possible reasons for this discrepancy are as follows: (1) IFN-y (secreted by activated Thl cells) often provides negative signals to B cells (266-271); it is unclear why IFN-y is sometimes, but not always, inhibitory. It is possible that different clones of Thl cells secrete different amounts of IFN-y and that the high secretors prevent B cell help; (2) B cells which respond to Thl clones may represent distinct subsets (e.g., memory versus virgin B cells) or cells at different stages of activation. In light of the last point, Killar et al. (261) used small splenic B cells, Stevens et al. (265) used TNP-ABCs prepared by rosetting, and Mosmann et al. (75) used LPS-activated B cells. A number
28
ELLEN S. VITETTA ET AL
of studies have now suggested that the ability of Th cells to help B cells is, in fact, dependent on the activation state of the B cells. Bottomly and colleagues (259, 261, 262) have reported that both types of clones (ThMdammatory and Th2/helper) are capable of mediating polyclonal B cell responses, whereas only Th2/helper clones are able to mediate antigen-specific responses. Moreover, they have reported that antigenspecific responses are elicited from small B cells, whereas polyclonal responses are elicited from large B cells. Taken together, the above studies suggest that clones of Th2 cells, possibly because of their ability to secrete IL-4 and IL-5 (260, 261), are probably responsible for the major portion of the helper activity mediated by T h cells. The reason for the conflicting evidence concerning the helper activity of Thl clones remains to be clarified.
MARKERSOF THE DIFFERENT T h CELLSUBTYPES C. SURFACE Although the characteristic pattern of lymphokine secretion can be used as a functional “marker” to distinguish murine T h l and Th2 cells (75, 260), no cell surface marker has been described that is present exclusively on one type (Thl or Th2) of murine T h cell. However, antibodies have been reported to distinguish between functionally different T h cells in several species. Arthur and Mason (272) reported that the antibody MRC-OX22 could separate rat Th cells into those responsible for the majority of IL-2 production (OX22+) and those responsible for B cell help (OX22-). MRC-OX22 antibodies recognize an epitope on the common leukocyte antigen or T-200 molecule (CD45R) (273). Based on recent reports, antibodies recognizing epitopes on human (274) and mouse (263, 275) CD45R molecules also appear to distinguish between functionally different T h cells. Birkeland et al. (275) have used two monoclonal anti-CD45R antibodies recognizing epitopes on the T-200 molecule to stain normal T cells and clones of T h l and Th2 cells in an indirect immunofluorescence assay. The different subtypes of clones stain to different degrees, i.e., Th2 clones stain with greater intensity than do Thl clones. In addition, another anti-CD45R antibody has been described which distinguishes functionally different murine T h cells based on their density of the reactive T-200 epitopes. “CD45R-low”and “CD45R-high’T cells produce IL-4 or IL-2 and IFN-y and correspond to Th2 and Thl cells, respectively (263). Recently, Clement et al. (276) have studied the relationship between two human CD4 + T h cell subsets defined by reactivity with monoclonal antibodies (mAbs) directed against the CD45R (T-200)molecule. They have reported that the CD45R+ and CD45R- subsets, corresponding to “suppressor-inducer” and T h cells, respectively, are not composed of fully mature and functionally stable
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
29
T cells, but that they represent cells at different stages of maturation in a postthymic activation-dependent differentiation pathway. Although the ontogenetic relationship between CD45R+ and CD45R- T h cells is not known, these findings are not inconsistent with the notion that antigen or other factors may cause the conversion of one T h cell type into another. If this were the case, there would be a predictable sequence of lymphokines secreted during an immune response and these lymphokines would regulate the order of Ig isotypes secreted and the types of effector cells activated.
D. THEPROLIFERATIVE RESPONSE OF CLONES OF T h l AND Th2 CELLS T h cell clones do not constitutively secrete lymphokines but are dependent upon antigenic or mitogenic stimulation to do so (75, 260, 277). Another important consequence of T h cell activation after antigen presentation is the acquisition of a state of “competence,” which is characterized by the responsiveness of the T h cells to the proliferative effects of some of the secreted lymphokines, particularly IL-2 (277-279) and IL-4 (277, 280). T h cells are stimulated by the binding of either anti-TcR antibodies or antigedclass I1 (on an APC) to their receptors (138, 281). This results in the induction of lymphokine secretion and the expression of cell surface receptors (R) for IL-2 (IL-2R) and IL-4 (IL-4R) (282, 283). The secreted lymphokines act via autocrine and paracrine mechanisms to induce the proliferation and expansion of the activated T h cells as well as proliferation and differentiation of the B cells which they help (277, 280, 284-287).
E. REGULATION OF THE ACTIVATION OF T h l AND Th2 CELLS The different types of T h cells appear to differ in their sensitivity to the stimulatory or inhibitory effects of a variety of lymphokines (IL-4, IL-1, and IFN-y) (280, 287, 288), raising the possibility that lymphokinemediated circuits regulate the activation, expansion, and eventual down regulation of each particular type of T h cell. In this regard, IL-2 was originally defined as a lymphokine which mediated the proliferation of activated T cells (289), and IL-4 was defined as its counterpart for activated B cells (290). Following the generation of a mAb against IL-4 [llBll (291)], and the cloning and expression of IL-4 cDNA (292, 293), a number of laboratories reported that IL-4 could also mediate the proliferation of some T cell lines (293-297) or resting T cells [in combination with phorbol myristate acetate (PMA)] (298). Moreover, it was demonstrated that proliferation was not mediated through the production and utilization of IL-2, since neither anti-IL-2 nor anti-IL-2R
30
ELLEN S. VITETTA ET A L
antibodies were able to block the IL-4-mediated proliferation (277, 294, 296). The delineation of the different Th cell subsets by Mosmann et al. (75, 260) and the characterization of their lymphokines prompted investigations of how antigen-induced proliferation was mediated in Th2 cells, which were reported to lack the ability to secrete IL-2. A number of laboratories (277, 285, 286) reported that when stimulated in vitro by antigen on APCs, Th2 cells could proliferate in response to IL-4 in an autocrine fashion. In Thl cells, IL-2-mediated proliferation was also an autocrine event (285). Fernandez-Botran et al. (277, 280) reported that activated Th2 cells could also utilize IL-2 as a paracrine growth factor and that their responsiveness to both IL-2 and IL-4 increased after antigenic stimulation. These results suggest that the proliferation of one T h cell type can be modulated by lymphokines secreted by the other T h cell type. Kurt-Jones et al. (288) and Greenbaum et al. (287) have studied IL-2and IL-4-mediated proliferation of “activated’ T h clones. They reported that both clones respond to IL-2, but that Th2 cells were more responsive to IL-4 than were Thl cells. Thl cells did not respond to IL-4 except in the presence of suboptimal concentrations of IL-2 (287) and Th2 cells required the presence of IL-1 in order to respond to IL-4 (287, 288). IL-1 had no effect on Thl cells. In a recent report, Fernandez-Botran et a2. (280) reinvestigated the responsiveness of both types of T h cells to IL-2 and IL-4 as a function of the time elapsed after antigenic stimulation. Both Thl and Th2 cells respond to IL-4 and IL-2 shortly after antigenic stimulation and responsiveness to both lymphokines decreases with time after activation. Thl clones lose their responsiveness to IL-4 more rapidly and to IL-2 more slowly than do Th2 clones. In agreement with previous reports (287, 288), Th2 cells are more responsive to IL-4 than are Thl cells. When both IL-2 and I L 4 are added to cultures of Thl or Th2 cells, the proliferative response of both cell types is synergistic, suggesting that the presence of both lymphokines might be required for an optimal response. No requirement for IL-1 on the proliferation of Th2 cells in response to either IL-2 or IL-4 was identified, suggesting that there might be heterogeneity among Th2 cells with regard to their requirement for IL-1. Thl-derived IFN-y also has a regulatory effect on the proliferative response of Th2 cells. IFN-7 is generally inhibitory for IL-4-mediated activities on B cells (266-271). IFN-y also inhibits the IL-2- and IL-4-mediated proliferation of Th2, but not T h l cells (280, 299). Table I11 summarizes the responsiveness of the two T h subsets to different lymphokines. In summary, a number of mechanisms might exist in vivo whereby the initial stimulation of one or both types of Th cells could be amplified
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
31
TABLE 111
RESPONSIVENESS OF Th SUBSETS TO DIFFERENT LYMPHOKINES' Lymphokine
Thl
Th2
IL-1 IL-2 IL-4 IFN--/
++ +
-
+ ++ ++ +
-
Assay (Costimulation with IL-2 and IL-4) (Proliferation) (Proliferation) (Inhibition of IL-2- and IL-4-mediated proliferation)
"From Fernandez-Botran e t al. (280).
and subsequently down regulated. T h cells would become stimulated after antigen presentation by APCs, leading to lymphokine secretion and the expression of receptors for these lymphokines. The secreted lymphokines would then act via autocrine or paracrine mechanisms to induce the expansion of clones of activated T cells and proliferation and differentiation of activated B cells. Eventually, when antigen concentrations decrease, T cell activation would become self-limiting by down regulation of surface IL-2Rs and IL-4Rs and by the cessation of lymphokine secretion. In addition, it has become clear that lymphokines produced by one T h cell type (i.e., IFN-7) can down regulate the growth of the other T h cell type (280, 299). F. T h l AND Th2 CELLSin Vim The characterization of the two types of T h cells has been carried out in vi'tro using established clones of cells (75, 260, 261, 264). It is therefore possible that variables in the techniques used for isolating, growing, and cloning T h cells might select clones that are not representative of their in mvo counterparts. On the other hand, since the phenotype of these T h clones is stable and since most clones fall into one of the two categories described by Mosmann et al. (75), it is likely that they are representative of their normal counterpart in vivo (263, 300, 301). Indeed, studies correlating the phenotype of normal T cells (CD45R-low and CD45R-high) with their pattern of lymphokine secretion and other functional characteristics provide strong evidence that analogs of T h l and Th2 cells exist in vivo (263, 275). However, as mentioned previously, it is still not clear whether Thl and Th2 cells represent two different lineages which differentiate from a common precursor, or whether they represent different stages of the T h cell in the same differentiation pathway, although there is evidence to support the latter possibility (276). Powers et al. (302) have used limiting dilution cultures to analyze the frequency of IL-2- and IL-4-secretingprecursors in naive
32
ELLEN S. VITETTA ET AL.
and antigen-primed mice and have found that most freshly isolated Th cells produced IL-2 following primary stimulation. IL-4 production was detected only after further in Vitro stimulation. Similarly, Swain et al. (301) studied the generation of Thl and Th2 cells in vitro from freshly isolated T cells and reported that, in contrast to T h l cells, Th2 cells were dependent on priming before they secreted detectable levels of IL-4 or IL-5. They have suggested that precursors of Th2 cells require a maturation event before they can be stimulated to secrete IL-4 or IL-5. The delayed development of IL-4-producing cells (Th2) after repeated antigenic exposure suggests a potential role for Th2 cells in the generation of immunologic memory and/or in the immune response against antigens that are persistent or frequently encountered (302).
G. T CELL-MEDIATED REGULATION OF ISOTYPE SWITCHING IN B CELLS - MOLECULAR AND CELLULAR CONSIDERATIONS Since the early 1970s, it has been recognized that T cells influence the class of Ig secreted by B cells in response to TD antigens (34, 303-310). IgM and IgG3 responses are relatively thymus independent, whereas IgGl, IgGz, IgA, and IgE responses require T cells and their secreted lymphokines. These conclusions were deduced from experiments in which animals that were either carrier primed or lacked a thymus were injected with antigen. Carrier-primed mice made significant IgG responses whereas athymic n u h u mice did not (35, 311, 312), implying that carrierprimed T cells were necessary for switching. It was subsequently reported that an IgG response could also be induced in the presence of an “allogeneic” effect (74, 313-315). In contemporary terms, this finding suggests that T cells and their lymphokines play an important regulatory role both in the secretion of IgG antibodies and in the generation of immunological memory. During the past decade, numerous experiments have elucidated the organization and expression of the Ig constant region (CH) genes on the chromosomes of mouse and man (316). Not only has the 5 ’ -3’ order of the CH genes’been assigned in both species, but some of the molecular events underlying isotype switching have been identified. Thus, B cells can switch, at the DNA level, from the production of a functional transcript encoding contiguous V, D, J, and CM segments (exons) to VDJ juxtaposed to other downstream CH genes (317), which in the mouse consist of y3, yl, y2b, y2a, E , and a (5’43’). In most cases, switching involves deletion of intervening intronic segments of DNA as well as intervening exons encoding other CH genes. Hence, by switch recombination and deletion, a new constant region gene is juxtaposed to VDJ. Initial reports suggested that proliferating B cells switched
CELLULAR INTERACTIONS IN T H E HUMORAL IMMUNE RESPONSE
33
sequentially and stochastically from the expression of 5’ CH genes to other 3’ CH genes. However, Isakson et al. (318) first demonstrated that lymphokines strongly influenced the class of Ig secreted by LPSstimulated B cells, implying that switching was not entirely stochastic with regard to CH gene organization (i.e., 5’-3‘). More recent experiments have implicated T cell-derived lymphokines as mediators of directed switching to other isotypes of Ig as well (75, 268, 284, 318-322). To date, most experiments have used anti-Ig or mitogenstimulated B cells to study lymphokine-mediated switching. As described in a previous section, during the interaction of antigenspecific T and B cells, lymphokines are released by the activated T cells and these lymphokines can act not only on the cells involved in the interaction, but also on other cells as well. As noted earlier, Thl cells secrete IFN-y and IL-2, and Th2 cells secrete IL-4 and IL-5. With regard to these four lymphokines, a number of laboratories have demonstrated that B cells cultured with these lymphokines and with the mitogens LPS or purified protein from Salmonella typhimurium (STM) can selectively secrete certain classes of Ig (75, 268, 284, 318-323). For example, STM or LPS and IL-4 induce B cells to secrete IgGI (318, 323) and, at higher doses of IL-4, IgE (268, 322). IFN-y induces LPS-stimulated B cells to secrete IgGza(321), and IL-5 induces LPS-stimulated B cells to secrete IgA (300, 324). Therefore, at least with respect to IgGl, IgGea, IgA, and IgE in the mouse, the aforementioned lymphokines clearly play a major role in inducing B cells to secrete the isotypes indicated. A recent study by Stevens et al. (265) has placed some of these observations into the context of T cell/B cell interactions involving subsets of Thl versus Th2 cells which secrete lymphokines involved in class switching (265, 325). Using two KLH-specific T cell clones of the Thl or Th2 types, it was shown that the interaction between a TNP-ABC and a Th2 cell (that secretes IL-4 and IL-5) resulted in the production of IgGl, whereas interaction between a TNP-ABC and a Thl cell (that secretes IL-2 and IFN-y) resulted in the production of IgGza. During the interaction of TNP-ABC with Thl cells (which normally induce production of IgGz,), the addition of exogenous IL-4 resulted in the production of IgGl. Conversely, if B lymphocytes were incubated with Th2 cells (which normally induce the production of IgGl), the addition of an exogenous IFN-y failed to induce the production of IgGza but inhibited the IgGl-inducing effects of the Th2 cells. Since an individual TNP-ABC is capable of interacting with either a Thl or a Th2 cell (326), it can be postulated that the interaction between a B cell and a T cell (Thl or Th2) induces a state of responsiveness to many, but perhaps not all, lymphokines. Taken together with the results obtained using
34
ELLEN S. VITETTA ET AL.
mitogen plus lymphokines, these results suggest that the class of Ig secreted by a B cell is dependent upon the type of T cell with which it interacts and the lymphokines which the activated T cell subsequently secretes. Although the relationships between T cells and their lymphokines and the molecular events involved in isotype switching have not yet been related in a cause-and-effectmanner, some insights have emerged from studies utilizing the Th2 cell-derived lymphokine, IL-4. IL-4 can induce mitogen-activated murine sIgG - B cells to secrete increased levels of IgGl and IgE and decreased levels of IgGs and IgGzb (268, 318). The enhanced secretion of IgGl is accompanied by an IL-4-mediated increase in the transcription of y l genes (327), and an increase in the precursor frequency of IgG1-secreting cells (328), suggesting that IL-4 induces sIgM+ B cells to undergo a heavy-chain class switch from IgM to IgGl. Although both mitogen and IL-4 are necessary for the switch to IgGl secretion, IL-4 is required very early in the culture, several days before switch recombination can be detected (318, 329). The mechanism by which IL-4 regulates isotype switching is unknown, but recent studies on the molecular basis of isotype switching may provide some clues. Analyses of murine hybridomas and normal murine B cells have provided strong support for the hypothesis that switch recombination is a directed event which occurs within the tandemly repeated elements of the same switch region on both the productive and nonproductive alleles (330-333). This directed switch might be due to the induction or activation of switch recombinases specific for each switch region. Recent work suggests that isotype switching is regulated at the level of accessibility of the switch regions to a common switch recombinase (334-337). Stavnezer et al. (338, 338a) have demonstrated in the IgM+ 1.29 B cell lymphoma line that the specificity of class switching correlates with hypomethylation and transcriptional activation of the corresponding germ line CH gene prior to switching. Recently, IL-4 has been shown to induce transcription of germ line Cy RNA in normal B cells before switch recombination access (338b). This induction is dosedependent and markedly uninhibited by IFN-7 (338b). Yancopoulos et al. (339) have also shown that Abelson-transformed pre-B cell lines, which switch spontaneously in Vitro to IgG2b, produce small germ line (“sterile”)y2b transcripts prior to switch recombination. Most recently, it has been shown that LPS induces the synthesis of germ line y2b transcripts in Abelson-transformedpre-B cell lines and in normal splenic B cells prior to y2b class switching (340). IL-4 inhibits the induction of these transcripts in splenic B cells (341). This correlates with the
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
35
previous observations that IL-4 inhibits the secretion of IgGzb by LPSstimulated B cells (342). These studies have suggested an accessibility model for the regulation of isotype switching (342a), i.e., that the initial events in the process involve the opening up of the chromatin in the switch region, followed by transcriptional activation and eventual recognition of DNA segments by switch recombinases. Whether transcriptional activation actually directs specific recombination or is a manifestation of a more accessible chromatin structure remains to be determined. Nevertheless, the accessibility of switch regions is probably affected by DNA-binding proteins (343) that recognize elements in the switch regions with partial or perhaps complete isotype specificity. It will be important to identify and characterize such regulatory proteins and their corresponding sequence elements and to understand how they might be regulated by external stimuli such as lymphokines and mitogens. It is also possible that some lymphokines (e.g., IL-5) might have no effect on the switching machinery, but might instead induce secretion of Ig in activated postswitch B cells (those that already express sIgG or sIgA). Future studies will be aimed at relating T cells and their lymphokines to the molecular events leading to early changes in CH-encoded DNA and to switch recombination. Eventually, these concepts must be placed in a physiological setting to take into account how the Thl and Th2 cells that secrete different lymphokines are activated in vivo and how activation is related to the nature of the antigenic stimulus (e.g., parasite, virus, or bacterium). Finally, the anatomical site in which the T h cells reside must be considered. IV. Physical Interaction between T and 6 Cells
A. CELLULAR INTERACTIONS IN THE IMMUNE RESPONSE A large body of work spanning nearly two decades has established the basic requirements for, and consequences of, the interactions between immunologically relevant cells such as T h cells with macrophages and Tc cells with their target cells. The major findings are summarized as follows: 1. Resting T cells from unprimed mice express clonally distributed receptors which bind to processed antigen in association with class I or class I1 molecules (344). The class I or II/antigen complex must be presented on the surface of an APC.
36
ELLEN S. VITETTA ET A L
2. The antigen-specific, MHC-restricted binding of resting T cells to APCs leads to their activation, the release of lymphokines, and proliferation and expansion of the T cell clones (289). 3. Although any cell expressing the correct MHC-restricting elements and antigen can act as an APC, the major APCs in the immune system are dendritic cells (345, 346), macrophages (347, 348), epidermal Langerhans cells (349, 350), and B cells (162, 171, 351). 4. The binding of T cells to APCs is cation and temperature dependent (352,353) and may involve not only TcRs, but also other cell surface adhesion molecules on both cells, strengthening the binding affinity between the APC and the T cell. 5. The APCs and T cells (in the presence of antigen and MHCrestricting elements) form tight interactions (352,354-357) which, at least in the case of the Tc cell, facilitate the secretion of pore-forming proteins. 6. The activated T cells detach from the APC after a period of time and are able to initiate effector function, e.g., activated T h cells support antibody responses by resting B cells (358). 7. After their detachment from APCs, activated T cells can be maintained in a proliferating state by endogenously produced or exogenously added lymphokines (277, 280, 289). 8. Eventually, the activated T cells return to a resting state, probably as a consequence of both the depletion of lymphokines necessary for growth and the discontinued stimulation by APCs (278, 280). However, these resting T cells can be activated once again with antigenpulsed APCs (258). The requirements for the interaction between T and B cells are, in general, similar to those described above. Recent studies on T cell/B cell interaction are described in detail below. We will not address the issue of interactions among B cells or the activation of Ly-1 B cells and their possible involvement in autoantibody production. For discussions on these topics, the reader is referred to recent reviews (359-362). +
B. MODELSYSTEMS TO STUDY THE PHYSICAL INTERACTIONBETWEEN T AND B CELLS Both T and B cells are required for an immune response to a thymusdependent antigen (1, 2, 5, 16, 19). This observation led to the suggestion that the T and B cells physically interact, perhaps by antigen bridging. The first description of physical interactions between antigen-specific T and B lymphocytes was reported by Marrack et al. (363). Radiolabeled antigen-specific, I-region-restrictedT cell hybridomas were cultured with
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
37
adherent antigen-pulsed B cell lymphomas or hybridomas. Nonadherent cells were removed and the percentage of T cells attached to the adherent monolayers was estimated by measuring radiolabel remaining in the culture wells. Adherence of T cells to B cells was both antigen specific and MHC restricted and occurred within 15 min. Two additional experimental models have made it possible to quantify the number of T and B lymphocytes which adhere to each other and have facilitated the elucidation of factors required for conjugate formation. Experiments by Kupfer et al. (364) showed that the formation of conjugates between B hybridoma cells (LK35.2) and antigenand class II-specific cloned T h cells (DlO.M.1) required antigen presentation by the B hybridoma cells and MHC-restrictedinteraction between the two cells. Cells of the D10 clone were specific for conalbumin and IaKdeterminants. The LK35.2 cells expressed both IaKand Iad determinants but were not conalbumin specific. In the presence of high concentrations of conalbumin, the LK35.2 cells were able to nonspecifically bind the antigen and present it to the D10 cells. When LK35.2 cells and D10 cells were mixed, centrifuged, and resuspended, tight physical interactions or conjugates were formed; the number of cells capable of interacting was quantitated microscopically. The formation of antigenspecific, MHC-restricted conjugates induced reorientation of the microtubule organizing center (MTOC) inside the helper T cell to face the site of T cell/B cell contact. When conjugates formed in an antigennonspecific, MHC-nonrestricted manner, random reorientation of the MTOC was observed. Sanders et al. (197) employed a fully antigen-specific system, utilizing a highly enriched population of B cells specific for TNP (TNPABCs) and T cells specific for the carrier portion (KLH or OVA) of the TNP-carrier. The T hybridoma cells recognized processed antigen in the context of I-Ad molecules on the antigen-presenting TNP-ABCs. Because of the differences in size between the T hybridoma cells and the normal B cells, conjugates could be studied at the light and electron microscopic levels. In this way, two types of interactions were described. One type involved the interdigitation of one or more microvilli from each cell type (i.e., loose interaction). Loose interactions occurred in the absence of antigen and, hence, were not immunologically specific. Antigen-nonspecific, MHC-nonrestricted interactions between T cells and macrophages have also been reported (354, 356, 358, 365-367); the rate of dissociation of cells in these conjugates is rapid (354, 365, 368). The second type of interaction between the T and B cells involved an extensive area of close contact between the two cells (i,e., tight interaction). Tight interactions occurred in an antigen-specific, MHC-restricted
38
ELLEN S. VITETTA ET AL.
manner and the number of conjugates and the degree of close contact between the two cell types increased as a function of time. OUTCOME OF T CELL/BCELLINTERACTION C. THEFUNCTIONAL As reported by Sanders et al. (197), the functional importance of physical interactions between TNP-ABCs and T hybridoma cells is supported by two findings: (1) in the presence of the appropriate TD antigen (197), the T hybridoma cells induced the differentiation of TNP-ABCs into antibody-secretingcells; and (2) the number of TNP-ABCs forming tight interactions with T cells (197) was similar to, or greater than, the number of TNP-ABCs differentiating into antibody-secreting cells (369). These observations suggest that those B cells that form tight conjugates with T cells differentiate into antibody-secreting cells. Definitive proof of this hypothesis will require isolating conjugated versus unconjugated B cells to compare their ability to secrete antibody. FOR T CELL/BCELL D. BIOCHEMICAL EVENTS REQUIRED INTERACTION
1. Antigen Processing
As discussed in the previous section, the physical interaction between T and B cells is dependent upon the processing of native antigen by the B cells and presentation of processed, antigenic fragments or denatured antigen in association with class I1 molecules to TcRs. Following exposure of the B cells to antigen, a period of several hours is required before the B cells are able to conjugate to T cells. This suggests that antigen processing plays a rate-limiting role in conjugate formation. Supporting this hypothesis are the following data: (1) chloroquine, a lysosomotropic agent, inhibits the conjugation of TNP-ABC and T cells by 40% (197); (2) chloroquine inhibits MTOC reorientation in 50% of T cells that form conjugates (370); and (3) the rate of release of soluble radiolabeled antigen fragments from antigen-pulsed TNP-ABCs (C.D. Myers and E.S. Vitetta, unpublished observations) follows kinetics similar to those of conjugate formation between TNP-ABCs and T hybridoma cells (197). 2. Time and Temperature Conjugation between T and B cells is temperature dependent; a maximal number of conjugates forms at 37OC, fewer conjugates form at room temperature, and no conjugates form at 4OC (197). This implies that an energy-dependent event is involved. Once the B cells have processed antigen, their physical interaction with T cells occurs rapidly;
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
39
a maximal number of T cell/B cell conjugates forms within 5 min (197). Rapid adherence of antigen-specific T hybridoma cells to monolayers of adherent B cell lymphomas and hybridomas has also been reported (358). In contrast to the rapid rate of adherence, the rate of dissociation of specifically conjugated T and B cells is slow, i.e., the half-life is 4 h (V. M. Sanders, J. W. Uhr, and E. S. Vitetta, unpublished observations). E. INTRACELLULAR EVENTS WHICHOCCURDURING T CELL/BCELL INTERACTION As discussed above, an important event that follows specific T cell/ B cell interactions is reorientation of the MTOC in the T cell to face the site of contact with the B cell (364, 370). This event is reminiscent of the reorientation of the MTOC and the Golgi apparatus in Tc cells interacting with target cells (371). As shown for the Tc cell/target cell interaction (371), reorientations direct secretory vesicles from the Golgi apparatus toward the area of cell/cell contact. It has been hypothesized that the reorientation of these intracellular components in the T h cells that conjugate to B cells may serve a similar function (364, 370). These changes may direct the release of lymphokines into the intracellular space between the conjugating cells such that small amounts of lymphokines can effectively stimulate the interacting B cells because of their close proximity. Hence, local transfer of lymphokines might not effect neighboring cells. A concomitant localization of the membrane-associated cytoskeletal protein, talin, occurs under the membrane of the T h cell participating in specific conjugation (370). A rearrangement of talin is associated with cell interactions that involve the membrane-associated proteins of the integrin family (372), e.g., LFA-1. The integrin family includes molecules which act as receptors for extracellular matrix molecules involved in cell/cell adherence (372); these molecules have a common /3 subunit (95 kDa) and a variable a subunit (372). F. MONOGAMOUS NATURE OF T CELL/BCELLINTERACTION
The tight physical interactions that occur between normal T and B cells invariably involve one B cell and one T cell (326). In contrast, a large T hybridoma cell sometimes binds to two different B cells. The monogamous nature of B cells was reported in recent experiments using TNP-ABCs which could bind to either Thl or Th2 cells (326). In competition studies, the number of cells of one type of T h cell (e.g., Th2) was kept constant in the presence of increasing numbers of the other type of T h cell (e.g., Thl). Although the same antigen-specific B cell could bind to either a Thl or a Th2 cell, during the course of these
40
ELLEN S. VITETTA ET AL.
studies it was observed that if one T h cell had bound to a B cell, another Th cell (of the same or different T h cell type) was unable to do so. One possible explanation for the monagamy of B cells is that membrane-associated class I1 molecules are clustered at the interaction site between the T and B cells and would not be available for interactions with other T h cells. However, two experiments have excluded this possibility: (1) immunofluorescence staining of T cell/B cell conjugates has demonstrated that class I1 molecules remain relatively uniformly distributed on the B cell surface (V. M. Sanders, J. W. Uhr, and E. S . Vitetta, unpublished observations); and (2) when TNP-ABCs were conjugated to antigen-specific T h cells, they could not conjugate at the same time to alloreactive T h cells (which recognize only class I1 molecules) (326). This indicates that clustering of class I1 molecules at the interaction site does not occur to any great degree and is not responsible for the monogamy of B cells. It is possible, however, that other accessory molecules on B cells do cluster at the interaction site (e.g., LFA-1) and are not available to facilitate interactions with other T h cells. In any case, this does not rule out the possibility that, after the dissociation of the TNP-ABC from the T h cell, the TNP-ABC can subsequently bind to another Th cell [which is the case for T h cells and macrophages (368, 373) and Tc cells and target cells (357)l. G. CELLSURFACE Ig AS AN ACTIVATING MOLECULE As noted earlier in this review, because B cells express monoclonal sIg molecules, antigen-specific B cells are more efficient than macrophages in capturing specific antigen and presenting it in a processed form to T cells (169, 184, 187). However, the role of sIg per se in activating B cells has remained controversial. Anti-Ig stimulation of splenic B cells or antigen stimulation of TNP-ABCs induces a change in phospholipid metabolism (374-377), a decrease in membrane potential (378, 379), CaZ+ fluxes (380), and an elevation in the levels of mRNA for the c-myc protooncogene (381). These findings imply that B cells can be stimulated via sIg in the absence of T cells. However, stimulation of resting splenic B cells (382) with antigen alone is insufficient to induce their proliferation or differentiation into antibody-forming cells. The differentiation of resting TNP-ABCs into antibody-secreting cells requires the presence of T h cells in addition to antigen (383). In this regard, Julius et al. (384) cultured resting B cells with T h cells specific for B cell membrane molecules other than sIg [i.e., the male antigen or H-2 (class I) molecules]. In such experiments, antibody secretion did not occur unless anti-Ig antibodies were added to the cultures. These results showed that two signals were required for the activation of resting B cells. One signal was derived from the interaction between T h cells
CELLULAR INTERACTIONS
IN THE HUMORAL IMMUNE RESPONSE
41
and B cells and the other was provided by anti-Ig antibodies. In a study by LoCascio et al. (385), SRBC-specific “resting” B lymphoma cells, CH12, were cultured with T h cells which were specific for H-2 (class I1 molecules) on the CH12 cells. The B cells cultured with Th cells secreted antibody in the presence, but not in the absence, of specific antigen, supporting the conclusion that both sIg and cell/cell contact are involved in the induction of an antibody response. H. CELLSURFACE MOLECULES INVOLVED IN T CELL/BCELL INTERACTION Cell surface molecules involved in the formation of conjugates between T and B cells include L3T4, LFA-la, Thy-1, and TcRs on the T cell, and class I1 molecules and LFA-la on the B cell. This conclusion has been deduced from the observation that mAbs directed against these molecules inhibit the formation of conjugates in a concentrationdependent manner (197, 386). Furthermore, immunofluorescence staining shows localization of L3T4 and TcRs on the T cell at the site of specific conjugation with the B cell (387). Finally, antibodies directed against these molecules modulate both in vi2ro and in vivo functional responses of the B cells. 1. Class 11 Molecules
As discussed earlier, it has been postulated that antigenic fragments processed by APCs associate with class I1 molecules (235) and are recognized by MHC-restricted, antigen-specificT h cells (76, 78, 80-84). Hence, the production of antibody in response to a TD antigen is restricted by class I1 molecules on B cells (76-79, 388). In this regard, the experiments of LoCascio et al. (385) also demonstrated the necessity for class I1 compatibility between B cells and T cells activated by TD antigens and, in addition, implicated class I1 molecules as transmembrane signaling molecules. However, the state of activation of the B cell can determine its requirements for a class 11-restrictedT cell/B cell interaction in the induction of an antibody response (389-391). Resting B lymphocytes require a class 11-restrictedinteraction whereas activated B cells require only the lymphokines produced by T cells (previously activated by another APC in a class 11-restrictedmanner). In addition, recent observations have challenged the absolute requirement for the involvement of class I1 molecules in the activation of resting B cells (392-395); these findings will be discussed in more detail in a later section. A number of approaches have been used to show that class I1 molecules are involved in the events associated with the initial physical interaction
42
ELLEN S. VITETTA ET A L .
between antigen-specificT and B cells. Kupfer et al. (364, 370) have shown that specific reorientation of the MTOC occurs when class 11-compatible, as opposed to class 11-incompatible,B cells conjugate with T cells. Likewise, Sanders et al. (197, 386) reported that anti-class I1 antibodies inhibit the formation of antigen-specific tight interactions between T and B cells. Modulation of the level of expression of class I1 molecules on B cells has been accomplished and these cells have been evaluated for their ability to conjugate to T cells (396). The results of such experiments extended earlier observations that the density of class I1 antigens on B cells is directly related to the frequency of successful T cell/B cell interactions, leading to the activation of B cells (397-399). Hence, pretreatment of TNP-ABCs with IL-4 increased by twofold both the density of class I1 molecules on the TNP-ABC and the number of antigen-pulsed TNP-ABCs capable of conjugating to T cells (396). The same twofold increase in the number of IL-4-pretreatedTNP-ABCs differentiatinginto antibody-secretingcells was observed (282, 396). Thus, both the presence and the density of class I1 molecules on the B cell surface are important in determining the number of successful T cell/B cell interactions that occur. There is controversy as to whether the level of glycosylation of class I1 molecules on a resting B cell is a relevant factor in the association of T and B lymphocytes (163, 400). In one case, Frohman and Cowing (163) reported that class I1 molecules on resting B cells versus adherent APCs were glycosylated to different degrees and that this level of glycosylation explained the ability of adherent APCs, as opposed to resting B cells, to present antigen more effectively. By altering the level of glycosylation of surface molecules on resting B cells by neuraminidase treatment (which removes terminal N-acetylneuraminic acid residues), resting B cells became more effective APCs even though they remained in a resting state. The authors concluded that changes in the level of glycosylation of class I1 molecules on resting B cells are responsible for enhanced interactions between T and B cells. However, they did not prove that class I1 molecules, rather than other cell surface glycoproteins, mediated the observed effect. In this regard, and in contrast to the above conclusions, Krieger et al. (400) showed that planar membranes, into which class I1 molecules obtained from resting (“highly glycosylated’) and activated (“poorly glycosylated”) B cells had been inserted, could be pulsed with antigenic fragments and used as artificial APCs. Levels of glycosylation were indirectly evaluated by the ability of neuraminidase to enhance the ability of the intact B cells to present antigen. Results showed that both types of class 11-containingplanar membranes were equivalent in their antigenpresenting capacity. These studies suggest that glycosylation of cell surface
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
43
molecules other than class I1 molecules may be important in T cell/B cell interactions and the subsequent antigen-presenting capacity of B cells. The role of class I1 molecules as signal-traducing molecules in B cells has remained controversial (401). The major approach that has been used to address this issue has been to utilize anti-class I1 antibodies to stimulate B cells. Although some monoclonal anti-class I1 antibodies can stimulate resting B cells in a T-independent manner, treatment of mitogen-stimulated B cells can result in either inhibition (402-404) or stimulation of antibody secretion (405). Anti-class I1 antibodies also inhibit both anti-Ig-induced B cell proliferation (406) and differentiation induced by antigen and T cells (407, 408). Cambier and colleagues (409,410) have shown that anti-classI1 antibodies can stimulate the translocation of protein kinase C (PKC) from the cytoplasmic pool to a nuclear membrane-associatedpool, suggesting that class I1 molecules have a signaling function. This observation may also explain the inhibitory effects of anti-class I1 antibodies on mitogen-induced B cell responses, since a class 11-inducedtranslocation of PKC to the nuclear membrane would deplete cytoplasmic stores of PKC, which are usually translocated to the plasma membrane pool after mitogen stimulation (411). In summary, class I1 molecules on B cells are involved in the MHCrestricted interactions between antigen-specific T and B cells and may be involved in the transduction of activating signals to the B cell. Modulation of the density of class I1 molecules on the B cell influences both the number of B cells capable of conjugating to T cells and those B cells differentiating into antibody-secreting cells. 2 . LFA-1
Involvement of LFA-1 in cell/cell adhesion has been demonstrated primarily in the interaction between Tc cells and target cells (412, 413). Recently, LFA-1 has been implicated in interactions between T and B cells (197). LFA-1 is a heterodimer consisting of an a! subunit of 180 kDa noncovalently associated with a subunit of 95 kDa (414, 415). LFA-1 is expressed on resting and activated T and B cells (416). Springer and colleagues (417, 418) have reported that the intercellular adhesion molecule-1 (ICAM-1) may be the receptor for the LFA-1 molecule. An antibody directed against the ICAM-1 molecule inhibits LFA-1-dependent adhesion events and identifies ICAM-1 on many cells of hematopoietic and nonhematopoietic origin. T D antibody responses can be inhibited by anti-LFA-1antibodies (419-422) and it has been postulated that this is due to an effect of LFA-1 on intercellular interactions. The relative importance of LFA-1 as an adhesion molecule may be related to the avidity of the interaction between the antigedclass I1 molecule and the
44
ELLEN S. VITETTA ET AL.
TcR. Golde et al. (423) demonstrated that LFA-1 is strongly involved in the interaction between T cells and APCs only if the avidity of the TcR for processed antigedclass I1 molecules on the APC is low. These conclusions were deduced from experiments in which cells from a panel of T hybridomas (which released IL-2 in response to antigedclass 11) were treated with anti-LFA-1 antibodies. The ability of anti-LFA-1 antibodies to inhibit IL-2 release was inversely related to the binding avidity of the TcR to antigedclass I1 molecules on the APCs. Along the same lines, when TNP-ABCs from unprimed mice were used in the T cell/B cell conjugate assay, pretreatment of either the T cells or TNPABCs with anti-LFA-la antibody inhibited the formation of conjugates (197). In contrast, when TNP-ABCs from primed mice were used, pretreatment of the T cells, but not the TNP-ABCs, with anti-LFA-la antibody inhibited the formation of conjugates (386). A discussion of these results is presented in a later section of this review. Anti-LFA-1 also inhibits cluster formation between autoreactive T cells and neuraminidase-treated resting B cells (424), an observation which is consistent with the idea that autoreactive interactions of this type may be of low affinity and hence may require accessory adhesion molecules. A possible signaling function for LFA-1 has been reported by Mishra et al. (425). In this study, a mAb recognizing LFA-la (G-48) mimicked many of the actions of IL-4 on B cells. G-48 induced increased levels of expression of class I1 antigen on resting B cells, augmented the proliferation of anti-&stimulated B cells, and in insolubilized form, induced IgGl secretion in LPS-activated B cells. Of several anti-LFA-la antibodies tested, G-48 was superior in its stimulatory functions (W. Lee and E. S. Vitetta, unpublished observations). In this regard, this particular antibody also enhances, rather than inhibits, conjugate formation (V.M. Sanders and E. S. Vitetta, unpublished observations). A similar enhancement of conjugate formation was observed if TNP-ABCs were treated with IL-4 in the presence of antigen prior to conjugation with T cells (396). These data suggest a possible signaling function associated with antibodies directed against this particular epitope of LFA-la. In summary, LFA-1 participates in T cell/B cell interactions. The degree of participation of LFA-1 in these interactions may be influenced by not only the avidity of the TcR for antigedclass 11, but also by the ability of B cells to process antigen and express high densities of class I1 molecules. LFA-1 may also be involved in the transduction of activating signals to the B cell. 3. L3T4 (CD4) L3T4, the murine equivalent of human CD4, is a T cell-associated surface molecule which may be involved in the recognition of class I1
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
45
molecules on the surface of APCs (426-429), and, as such, plays an important role in strengthening interactions between T cells and other cells (430). In vitro, L3T4 plays an essential role in both T h cell activation and T cell-mediated activation of B cell proliferation (426-428). In an attempt to elucidate the role of L3T4 in functional responses of T cells, it was shown that anti-L3T4 antibodies inhibit conjugate formation between T and B cells (197, 386). In addition, when conjugates between T and B cells form, L3T4 localizes to the site of interaction, as does the TcR (387). Marrack et al. (430) have shown that the requirement for L3T4, like LFA-1 (423), is inversely proportional to the avidity of the TcR for the antigedclass I1 complex on the APC. Thus, L3T4 functions primarily as an adhesion molecule: its natural ligand is probably the class I1 molecule on APCs (426-428). Recent data suggest that L3T4 may also have a signaling function. It has been postulated that L3T4 plays a direct role in the inactivation of T cells. A number of workers have reported that antibodies directed against L3T4 molecules inhibit T cell activation by conconavalin A or anti-TcR antibodies in the absence of TcR recognition of antigen in association with class I1 molecules (431-433). Thus, it has been suggested that L3T4 may function as a negative signaling molecule when the T cells interact with cells lacking the appropriate antigenic fragment in association with class I1 molecules. During such a nonspecific interaction between a T and B lymphocyte, the binding of L3T4 (on the T cell) to its putative ligand (class I1 molecules) on the B cell in the absence of an interaction between the TcR and antigedclass I1 may cause the T cell to dissociate from the B cell so that it can “sample” other B cells that might express the relevant antigedclass I1 complex. In summary, L3T4 participates in T cell/B cell interactions by serving primarily as an adhesion molecule. L3T4 may also mediate signals to T cells and be involved in mediating dissociation between APCs and T cells. 4 . The T Cell Receptor
The majority of T cell receptors are composed of disulfide-bonded a and 0chains, each of which has a molecular weight of 45,000-55,000 and both of which have transmembrane and cytoplasmic domains. Other T cell receptors are composed of y and 6 chains, but it is not yet clear how these receptors function. The a0 chains of the TcR recognize processed antigen and class I1 molecules on the surface of APCs (132, 134, 136, 206, 227). This is probably the key interaction which initiates specific conjugation between a T cell and a B cell. Kupfer et al. (387) have shown that TcRs localize at the site of contact between T and B cells. Crosslinking of TcRs with antireceptor antibodies in metroactivate T cells and
46
ELLEN S. VITETTA ET A t
result in the directed release of lymphokines to the site of clustered TcRs (434). In addition, Taplits et al. (435) have shown that the secretion of enzyme-containing granules from T h cells exposed to anti-TcR antibodies occurs in an MHC-restricted, antigen-dependent manner and correlates with T cell proliferation. Taken together, these findings suggest that during the close interaction between T and B cells, clustering of TcRs results in T cell activation followed by the directed release of lymphokines and/or enzymes into the interaction site. 5. Thy-1
Thy-1 is expressed primarily on T cells although there is a report that activated B cells can also express Thy-1 (C.M. Snapper and W.E. Paul, personal communication). Involvement of the Thy-1 antigen in T cell/ B cell interaction was suggested by the finding that anti-Thy-1antibodies inhibit conjugate formation (197). In contrast to their ability to block conjugate formation, anti-Thy-1 antibodies activate T lymphocytes directly (436-439). In addition, Kroczek et al. (440)reported that antiThy-1 antibodies increased the concentration of intracellular Ca2 in T cells. Hence, anti-Thy-1 antibodies can activate T cells, but prevent T cell/B cell interaction. It would be informative to add anti-Thy-1 antibodies to cultures of T and B cells to determine whether specific antibody production can be inhibited. For example, it could be postulated that anti-Thy-1would inhibit the activation of resting B cells by T D antigens (by inhibiting T cell/B cell conjugation). At the same time, anti-Thy-1 might enhance the secretion of antibodies by lymphokine-responsiveactivated B cells (which had already undergone an interaction with T cells), since the T cells which had been activated by the anti-Thy1 antibody would be secreting lymphokines. +
I. CELLCONTACT VERSUS LYMPHOKINES IN T CELL/BCELL ACTIVATION The role of cell/cell contact versus the role of T cell-derived lymphokines in the activation of B cells by TD antigens has been addressed in several ways. In Vitro studies by Corbel and Melchers (441) have demonstrated that LPS-activated B cells can be induced to secrete antibody by the addition of soluble mediators. LeClerq et al. (442, 443) have shown that resting B lymphocytes can be stimulated to proliferate and produce antibody in the presence of T cell-derived lymphokines alone. In an attempt to distinguish between the requirement for cell/cell contact and the requirement for soluble mediators in the stimulation and differentiation of resting B cells into antibody-secreting cells, a
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
47
number of model systems have been developed. Melchers and colleagues (389, 444, 445) were among the first to study the early events involved in B cell activation by TD antigens. Blast transformation of small, resting B cells required antigen-specific, MHC-restricted interactions between B and T h cells. Differentiation of B cell blasts into antibody-secreting cells required antigen-nonspecific,MHC-unrestrictedsoluble factors produced by T h cells which had interacted specifically with MHCcompatible adherent cells in the presence of specific antigen. In this regard, Singer and Hodes (446, 447) reported that B cells expressing the cell surface antigen, Lyb-5, responded to T cell-derivedlymphokines alone. In contrast, Lyb-5-negativeB cells required MHC-restricted T cell interactions. These results indicate that some cell contact-mediated event was required to induce responsiveness of the Lyb-5-negativecells to T cellderived lymphokines. However, the relationship between Lyb-5-positive and Lyb-5-negative cells remains unclear. In addition, activation of Lyb-5-positive cells is MHC restricted in other experimental systems (398, 448, 449). Noelle et al. (383) cultured TNP-ABCs with antigen-specific, MHCrestricted T lymphocytes and antigen for 24 h. The T cells were then eliminated with anti-Thy-1antibody and C’ and the remaining B cells were cultured with lymphokines. Only the B cells which had interacted with the T cells became responsive to T cell-derived lymphokines and differentiated into antibody-secreting cells. These results indicate that resting TNP-ABCs must be activated by T cells and antigen via cell/cell contact prior to becoming responsive to lymphokines. However, lymphokines released at the contact site during the interaction could be solely responsible for the initial activation event. This possiblity was addressed by Krusemeier and Snow (450), who cultured TNP-ABCs with antigenspecific T cells that had been treated (prior to interaction) with cyclosporin A. This pharmacologic agent inhibits the release of lymphokines from activated T lymphocytes (451, 452). It was demonstrated that although cyclosporin A-treated T lymphocytes were unable to release lymphokines, they could physically conjugate to TNP-ABCs. TNP-ABCs which had interacted with cyclosporin A-treated T cells were unable to differentiate into antibody-secreting cells in the absence of exogenously added T cell-derived lymphokines, but were able to differentiate when specific lymphokines were added to the cultures. Thus, cell contact could be distinguished from lymphokine utilization. The development of lymphokine responsivenessby B cells cultured with cyclosporin A-treated T cells was partially blocked by the addition of anti-class I1 antibodies, suggesting that MHC-restricted interactions of TNP-ABCs with T cells render the TNP-ABCs responsive to lymphokines.
48
ELLEN S. VITETTA ET AL.
Rasmussen et al. (262) compared the requirements for specific versus polyclonal B cell activation by Th2 (IL-4- and IL-5-producing) cells. Hapten-specific antibody responses were observed only when the small B cells had physically interacted with T h cells. Polyclonal responses were observed when large B cells were cultured with lymphokines in the absence of T cell/B cell contact. The requirement for physical contact was determined by culturing cells in chambers separated by a nucleopore membrane. Thus, T cells could be physically separated from B cells without interfering with the diffusion of lymphokines and antigen. AntiIL-5 antibodies completely abrogated both the specific and polyclonal antibody response. This result suggests that IL-5 is essential for differentiation of resting or activated B cells and that T cell/B cell contact is critical for the induction of a state of IL-5 responsiveness in small, resting B cells. The characteristics of antigen-specific, MHC-restricted interactions between T and B cells are summarized in Table IV and a model is presented in Fig. 2.
CHARACTERISTICS
OF
TABLE IV ANTIGEN-SPECIFIC, MHC-RESTRICTEDFORMATION OF T CELL/BCELLCONJUGATES
Characteristic ~~
~~
Reference ~
1. Frequency: B cells forming conjugates with T cells (1/15) Frequency: B cells secreting antibody (1/15) 2. Antigen specific, MHC restricted 3. Antigen processing by B cell required 4. Time and temperature dependent 5 . T cell surface molecules involved: L3T4 LFA-la Thy-1 TcR B cell surface molecules involved: I -A LFA-la 6.
7.
8. 9. 10.
sk
T cell-associated L3T4, TcR, and talin cluster at the T cell/B cell contact site Class I1 molecules and sIg do not cluster at the contact site Reorientation of MTOC in the T cell Monogamous interaction of B cells A single B cell can conjugate with either a Thl or a Th2 cell
(197) (369) (197, 364) (197, 370, 386) (197, 370) (197, 386, 387, 424) (197, 386, 424) (197, 386) (387) (197, 364, 386, 396) (197) (424) (370, 387) (197, 326) (364, 370) (326) (326)
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
49
ICAM-1
B Talin antigenklass 11 complex
FIG. 2. Physical interactions between T and B cells. (Top) T and B cells involved in a “loose” interaction which occurs in the absence of antigen. Cell surface molecules do not cluster at the interaction site and no reorientation of the MTOC or talin occurs in the T cell. This type of interaction may involve cell surface adhesion molecules on one cell and their respective receptors on the other. (Bottom) A “tight” interaction which is antigen specific and MHC restricted. Class I1 molecules and LFA-1 on the B cell, and L3T4, LFA-la, TcR, and Thy-1 on the T cell, are involved in this interaction. T cellassociated LST4 and TcR cluster at the contact site and a reorientation of the MTOC and talin in the T cell occurs to face the interaction site with the B cell. These types of interactions are monogamous for the B cell since a single B cell conjugates to only one T cell at a time.
1. The Conjugation of Antigen-Specqic Memory B Cells to T Cells A large number of studies (47) have demonstrated that the differentiated progeny of memory B cells, as compared to virgin B cells, secrete large amounts of non-IgM antibody and that memory B cells themselves are more readily activated by smaller quantities of antigen and by fewer T cells. Furthermore, following antigenic stimulation, the latency period
50
ELLEN S. VITETTA ET A L .
for antibody secretion is shorter and the avidity of the antibody which the activated memory cells secrete is higher. These observations have raised questions concerning the mechanisms by which memory cells versus virgin cells interact with, and are activated by, T cells and antigen. Two technical advances have facilitated the characterization of the initial interaction between T cells and either antigen-specific virgin (TNP-ABC) or memory (TNP-MABC) B cells. The fmt technique involves the enrichment of TNP-ABCs and TNP-MABCs by virtue of differences in the avidity of their sIg molecules for specific antigen (194, 369, 453). The second technique involves the characterization and quantitation of the number of enriched antigen-specific B cells which physically conjugate to antigen-specific T hybridoma cells (197). Together, these two techniques have established that different B cells (e.g., virgin versus memory) differ in their requirements for physical conjugation to antigen-specific T cells. Physical and functional characteristics of both TNP-specific B cell populations are summarized in Tables I1 and V. Differences in conjugate formation between antigen-specific T cells and TNP-ABCs or TNP-MABCs are summarized in Table VI. When the number of B cells forming conjugates was correlated to the concentration of antigen to which the B cells were exposed prior to conjugate formation, TNP-ABCs, as compared to TNP-MABCs, required 10-fold higher antigen concentrations for half-maximal conjugate formation. This finding is not surprising, since TNP-MABCs express sIg receptors of a higher affinity for antigen than do TNP-ABCs (369). Therefore, these TNP-MABCs are better able to capture antigen and present it to T cells. In addition, the higher levels of expression of class I1 molecules on TNP-MABCs (369) probably contribute to the overall efficiency of antigen presentation at low antigen concentrations.
TABLE V CHARACTERISTICS OF TNP-MABCs
AND
TNP-ABCsa
1. TNP-MABCs can be prepared by lowering the TNP density fivefold on the red
blood cells used for rosetting the TNP-ABCs 2 . TNP-MABCs, but not TNP-ABCs, can be activated following adoptive transfer into irradiated, carrier-primed mice 3. TNP-MABCs, as compared to TNP-ABCs, express sIg receptors of a higher affinity for antigen 4. TNP-MABCs express higher levels of class I1 antigens than do TNP-ABCs 5. TNP-MABCs secrete high-affinity IgG after activation by a T D antigen and T h cells in Vitm, whereas TNP-ABCs secrete primarily low-affinity IgM “From Yefenof et al. (369).
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
TABLE VI DIFFERENCES IN CONJUGATE FORMATION BETWEEN T CELLS TNP-ABCS OR TNP-MABCS‘
51
AND
1. TNP-ABCs require a higher antigen concentration for conjugate formation than do TNP-MABCs 2. TNP-ABCs require more time to process and present antigen than do TNP-MABCS 3. LFA-1 molecules on TNP-MABCs, as opposed to those on TNP-ABCs, are not involved in conjugate formation ~
“From Sanders et al. (386).
TNP-ABCs require more time than TNP-MABCs to process and present antigen to T cells (386). When the number of B cells forming conjugates with T cells is correlated to the time during which B cells are exposed to antigen prior to their conjugation with T cells, at least 4-6 h is required for the TNP-ABCs to form a maximal number of conjugates. Presumably, this period of time is required for maximal processing and reexpression of antigen on the surface of the TNP-ABCs. In contrast, TNP-MABCs require only 2 h of exposure to antigen for maximal conjugate formation with T cells. The decreased time required for the TNP-MABCs to acquire the ability to conjugate to T cells may again reflect the increased affinity of their sIg receptors for antigen. Thus, TNP-MABCs may bind and take up antigen more efficiently than TNPABCs. Preliminary studies with the lysosomotropic drug, chloroquine, suggest that antigen processing may differ between virgin and memory B cells (386). Nanomolar concentrations of chloroquine added during the time of antigen processing by TNP-MABCs can effectively inhibit the ability of these cells to conjugate with T cells. In contrast, the same concentrations of chloroquine only partially inhibit the ability of TNP-ABCs to conjugate to T cells. These results also suggest that different intracellular compartments are involved in the processing of antigen in virgin versus memory B cells. As determined by the ability of antibodies against LFA-1 molecules to inhibit conjugate formation, studies by Sanders et al. (197) demonstrated that LFA-1 molecules on T cells, as well as those on TNP-ABCs, are involved in conjugate formation. In contrast, pretreatment of TNPMABCs, as compared to pretreatment of T cells with anti-LFA-la antibodies, did not inhibit conjugate formation. This finding suggests that LFA-1 molecules on TNP-MABCs are not involved in conjugate formation, even though TNP-MABCs express the same levels of LFA-1 on their surface as TNP-ABCs (V. M. Sanders and E. S. Vitetta, unpublished
52
ELLEN S. VITETTA ET A L .
observations). A more likely explanation is that the interaction between memory B cells and T cells is of a much higher avidity and, consequently, the requirement for LFA-1 molecules on the B cells is not as stringent as for the interaction between virgin B cells and T cells. Similar results have been described for primary versus secondary Tc cells which interact with target cells (454). These data suggest that the differences between the activation requirements of virgin and memory B cells may be explained, in part, by differences in the efficiency with which the B cells specifically interact with T cells. 2 . MHC- Unrestricted Th Cell-Dependent Activation of Resting B Cells
As presented in the preceding section, T h cell-dependent activation of resting B cells involves a sequence of events which includes the sIgmediated uptake and processing of antigen by the B cell, and the subsequent presentation of antigenic fragments in association with class I1 molecules to Th cells during direct T cell/B cell interaction. Antigen plus class I1 molecules bind to the TcR, resulting in both cell/cell contact and aggregation of TcRs (387). This results in the activation of the T cell, which then releases lymphokines toward the contact site with the B cell (434). Thus, the main role of antigedclass I1 complexes on B cells is to bind to and aggregate the TcRs. As mentioned earlier, it is not clear whether class I1 molecules per se are involved in signaling B cells, although many workers have concluded that the activation of resting B cells requires both stimulation of B cell surface determinants (e.g., class I1 molecules) during contact with the T cell and exposure to lymphokines, which mediate growth and differentiationof both the T and B cells. In contrast to MHC-restricted, antigen-specific activation of B cells, resting B cells can be induced to differentiate into antibody-secreting cells in a polyclonal fashion in the presence of histoincompatible T cells and high concentrations of antigen (455). These findings question the absolute requirement for the binding of TcRs to class I1 molecules on B cells in order to activate them and suggest that the activation of T cells by aggregation of TcRs results in the release of T cell-derived lymphokines. Subsequently, interactions between activated lymphokine-secreting T cells and resting B cells lead to B cell activation. Three recent findings have supported the latter hypothesis: (1) Hirohata et al. (456) cocultured human T cells (stimulated with immobilized antiCD3 antibodies) and highly purified peripheral blood B cells under conditions which allowed cell/cell contact. The immobilized anti-CD3 would aggregate the TcRs. When the T cells were activated and cocultured
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
53
with the B cells, polyclonal antibody secretion was induced. These results demonstrate that activated T cells provide all the necessary signals to induce B cell differentiation in the absence of antigen by a mechanism requiring direct cell/cell contact. (2) In a report by Owens (394), an anti-TcR antibody, F23.1, which recognizes the Vg8 region of the TcR, was attached to wells of microtiter plates. When T cells were added to the wells in the presence of histoincompatible splenic B cells, a polyclonal antibody response was elicited. The requirement for T cell/B cell contact was supported by two findings-(a) SNs from F23.1-activated T cells alone were unable to support B cell differentiation, and (b) B cells separated by a membrane from F23.1-activated T cells could not be activated. (3) Julius and colleagues (392, 393) used two different approaches to show that activation of resting B cells by T cells can be achieved by prior aggregation of TcRs. In the first (392), resting B cells were cultured with histoincompatible resting T h cells in the absence or presence of the F23.1 antibody. The Th cells were either F23.1-positive or F23.1-negative. The F23.1-positiveT cells, but not the F23.1-negative T cells, were able to support B cell differentiation in a manner comparable to that of LPS. In addition, an antibody response of similar magnitude was obtained if the Th cells were first cultured with irradiated histocompatible APCs and antigen and then added to cultures containing histoincompatible B cells. These results further support the notion that the activated T h cells can activate B cells in the absence of a class 11mediated signal. Any cell contact involved would therefore be dependent upon cell surface molecules other than class I1 molecules (e.g., LFA-1, L3T4). In the second approach, TNP-ABCs obtained from mice transgenic for Sp-6p.x genes (393) could be bridged to histoincompatible T cells with TNP-conjugated F23.1 antibodies. In such experiments, the TNP-ABCs were induced to differentiate into antibodysecreting cells in the absence of MHC-restricted interactions. When T h cells and B cells were cocultured with unconjugated divalent F23.1 antibody, but not univalent F23.1 antibody, an equivalent antibody response was obtained. This again suggests that aggregation of TcRs is the critical requirement for T cell activation and the subsequent induction of B cell differentiation. To explore this observation further, histoincompatible B and T cells were cultured in the presence of TNP conjugated to univalent fragments of the F23.1 antibody. Since this situation promoted B cell differentiation, whereas the addition of unconjugated (no TNP) univalent fragments of F23.1 antibodies did not, it was concluded that sIg (on the B cell) could effectively aggregate the TcR during antigen bridging and that recognition of class I1 molecules on the B cell is not necessary for induction of B cell differentiation.
54
ELLEN S. VITETTA ET A L .
Taken together, the above studies demonstrate that it is not obligatory for the TcR to deliver an activating signal to the B cell via recognition of class I1 molecules. However, these studies have not ruled out the possible requirement for other B c e l l - d a t e d molecules (e.g., LFA-1)in cell contact or B cell activation. One explanation for these results is that weak interactions between B cells and activated T cells can be mediated by LFA-1 or other adhesion molecules. In the presence of high concentrations of lymphokines (secreted by the activated T cells), these weak interactions might bypass signals normally provided by interactions with class I1 molecules. In support of this hypothesis, and as discussed in a previous section of this review, LFA-1 has been implicated in B cell signaling in other models (425). This hypothesis could be further explored by characterizing the molecules involved in the formation of conjugates between activated T h cells and histoincompatible B cells. For example, if the histoincompatible B cells could not conjugate to the preactivated T h cells in the presence of anti-LFA-1,etc., this would implicate these molecules as mediators of cell contact and perhaps signaling. In support of this hypothesis, interactions between T and B cells were artificially induced in the absence of hapten-carrier complexes by using TNP-modified, antigen-specific T h cells and TNP-ABCs (457). Proliferation was induced in both T and B cells and differentiation was induced in most B cells. B cell activation was not MHC-restricted, but could be blocked with antibodies directed against class 11, L3T4, and LFA-1 molecules. Although not yet tested, it is possible that the lymphokine-containingSN from the activated T cells in the presence of antibodies which bind to LFA-la [e.g., G48 (425)] might activate the B cells. It is now evident that signaling of resting B cells via class I1 molecules can be bypassed when activated T h cells are present and T cell/B cell contact can occur. This may account for “bystander effects” frequently observed when T cells and B cells are cultured under conditions of noncognate help (79, 389, 390, 458). An analogous situation may occur in the local environment of an activated lymph node and explain how “natural” antibodies appear in the circulation, i.e., resting B cells not specific for the antigen in question would be activated by contacting Th cells activated by other antigens. However, in this case, activated Th cells would arise as a consequence of class II/antigen-TcR interactions: the B cell bystander effect would be MHC-unrestricted, class II-independent, and lymphokine-mediated. V. lnterleukinr
Our knowledge of the physiochemical and biological properties of interleukins (cytokines) has grown exponentially since the initial reports that soluble products from activated T cells (lymphokines) could induce
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
55
B cells to secrete Ig after antigenic stimulation (74, 459). This has been due to the availability of mAbs, new cell culture techniques, and the molecular cloning of interleukins. Nonetheless, attempts to understand how interleukins regulate cellular activity, especially in vivo, have been complicated by the fact that their biological activities are usually extremely complex and pleiotropic (460-462). This complexity is exemplified by reports that (1) some lymphokines can act at different stages of development on a given cell type inducing different effects, some of which are inhibitory and some of which are stimulatory; (2) most lymphokines can act on cells from different lineages, including nonhematopoietic cells; (3) one lymphokine given in combination with another lymphokine can have both synergistic and antagonistic effects on different target cells; (4) in vi'tro activities of lymphokines from one species cannot be easily extrapolated to in vzvo models in the same species or even to in m'tro models in another species; and ( 5 ) with a few exceptions, the receptors for different lymphokines have not been well characterized. It is therefore possible that receptors for different lymphokines interact, share subunits, or bind other lymphokines with low affinity (460, 463-465). If so, this would call into question the physiological relevance of pleiotropism, particularly in Vitro, where large amounts of lymphokines are often added to cell cultures. In vivo, because of anatomical compartmentalization of different cells and the low concentrations of circulating lymphokines, some of the effects observed in vitro may not be physiologically relevant. The purpose of this section is to review our current knowledge of the biological properties of interleukins/cytokines that play a role in T cell/ B cell collaboration and activation. The discussion will focus primarily on in vitro systems in the mouse. Table VII lists the biological properties of some of the cytokines, the majority of which are secreted by activated Thl or Th2 cells, with special reference to their effects on T cells, B cells, and accessory cells. As discussed in the previous section, one of the consequences of T h cell stimulation by antigen, presented in the context of class I1 molecules on an APC, is the release of lymphokines (75, 260, 277). Such lymphokines have effects not only on cells immediately involved in cellular interactions, but also on other lymphoid and nonlymphoid cells as well (460, 461). In a physiological context, these pleiotropic effects presume that the activated T cell (or its lymphokines) can traffic to other areas of the body and that other target cells (not involved in the initial collaborative interaction between T cells and APC) express lymphokine receptors and perhaps class I1 antigens. This has not yet been proved to be the case in vivo. In addition, cells which are not directly involved in the immune response are able to secrete additional lymphokines that
TABLE VII MULTIPLEEFFECE OF CYTOKINE~ ON MURINET CELLS, B CELLS, AND ACCESSORY CELLS Lymphokine"
Sources
Biological activities
IGl -Two distinct but structurally related molecules have been cloned, IL-la! and IL-lfl (576, 577); [endogenous pyrogen (573), leukocyte endogenous mediator (574). lymphocyteactivating factor (LAF) (575)l
Monocytes, macrophages, B cell lines, T cell lines, fibroblasts, keratinocytes, Langerhans cells, astrocytes, endothelial cells, NK cells (reviewed in Ref. 571)
Effects on lymphocytic/myeloid cell precursors Maturation of pre-B cells (578, 579) Synergy with CSFs on immature precursors (571, 609) Increased synthesis of CSFs (571) Costimulatory activity on thymocytes (575) (reviewed in Refs. 570 and 572) Effects on B and T cells B cells Increased proliferation of B cells stimulated with anti-Ig + IL-4 (580) Increased proliferation of B cells stimulated with anti-Ig + dextran sulfate (581) Increased proliferation of antigen-stimulated B cells (484, 582) Increased PFC formation with IL-2 (527) T cells Costimulatory activity for growth of activated T cells (reviewed in Refs.570, 572, 609, 287, 288, and 583-585) Lymphokine release from activated T cells (reviewed in Refs. 570, 572, and 610-612) Costimulatory activity for IL-4 and IL-2-induced proliferation on Th2 but not Thl cells (280, 286-288) Effects on accessory cells (macrophages) (reviewed in Refs. 570 and 572) Increased synthesis of PGEz (613) Induction of cytotoxicity
IG2- [T cell growth factor, (TCGF) (reviewed in Refs. 289 and 522)]
Activated Thl cells (75, 260); Tc cells (614, 615)
IGS - [Mast cell growth factor, P-cell stimulating factor, multicolony stimulating factor, panstimulating hemopoietin (reviewed in Ref. 586)]
Activated T cells (Thl and Th2) (75, 260, 586), WEHI - 3 B (myelomonocytic cell line) (586, 587)
Increased migration Increased synthesis of IL-1, CSFs, and other cytokines Induction of release of collagenase and acute-phase proteins Effect on lymphocyte/myeloid precursors Growth and differentiation of thymocytes (523, 616, 617) Effects on B and T cells B cells Proliferation of B cells stimulated with anti-Ig + LPS (533) Proliferation and differentiation of antigenstimulated B cells (527-530) Anti-SRBC PFCs with IL-1 and IFN-y (527) Induction of IgM secretion (514, 530, 618) Differentiation with anti-Ig + IL-4 plus a second T cell-derived factor (516) T cells Induction of proliferation of resting T lymphocytes (high doses) (549, 550) Growth of activated T cells and thymocytes (both Thl and Th2) (reviewed in Reti. 289, 522, 523, 277, 280, and 288) Cytotoxic T lymphocyte generation (524-526) Effects on accessory cells Increased monocyte cytotoxicity (human) (554) Effects on lymphocytic/myeloid precursors Growth and differentiation of multipotential stem cells and myeloid progenitor cells (granulocytes, monocytic, eosinophilic, erythroid, megakaryocytic, and mast cells) (reviewed in Refs. 586 and 588-592) Growth of pro-B, pre-B, and pro-T cell clones (594, 595) Other Growth of mast cell lines (496, 591, 592) (continued)
TABLE VII (Continued) MULTIPLE EFFECTS OF CYToKINEs ON MURINET CELLS, B CELLS, Lymp hokine"
IL--[B cell growth factor-1 (BCGF I), B cell stimulatory factor-1 (BSF-l), T cell growth factor I1 (TCGF 11) (293, 619)]
AND
ACCESSORY CELLS
Sources
Biological activities
Activated Th2 cells (75, 260), mast cell lines (620)
Effect on lymphoid/myeloid precursors Growth factor for fetal thymocytes (496-498) Growth of thymocytes in the presence of IL-2 (499) Enhanced erythroid, granulocyte, monocyte, and mast cell colony formation in combination with erythropoietin, G-CSF, M-CSF, and IL-3, respectively (481, 482) Induction of cytotoxic precursors from thymocytes (496) Effect on B and T cells B cells Induction of class I1 MHC molecules on resting B cells (469, 470) Increase in size of resting B cells (472, 621) Induction of FccR on resting B cells (476, 477) Preparation of resting B cells to proliferate to antiIg (471, 472) Increased ability to conjugate with T cells (396) Synergy with Type I1 TI antigens (484) Increased secretion of IgGl and IgE by LPSactivated B cells or B cells stimulated with T cells and antigen (268, 284, 318, 319, 321, 473, 474, 485) Decreased IgGs, IgGs,, and IgGebsecretion (284, 318, 321, 474, 485) Increased IgA production with IL-5 (300, 324) T cells Proliferation of resting T cells with PMA (298)
IG5- [B cell growth factor I1 (BCGF 11) (501, 502), T cell-replacing factor (TRF) (500), killer helper factor (KHF) (508). eosinophilactivating factor (EAF) (506, 623)]
Activated Th2 cells (260, 300, 301)
Increased IL-2R expression in the presence of PMA (622) Growth factor for activated T cells (Thl and Th2) (277, 280, 285-28a, 294-296) Cytotoxic T lymphocyte activity (493-495) Effects on accessory cells Increased class I1 expression on macrophages (478-480) Increased tumoricidal activity (478-480) Growth factor for mast cells with I L 3 (293, 294, 475) Effects on lymphoid/myeloid precursors Eosinophil colony formation from bone marrow cells (proliferation and differentiation of eosinophils) (503, 506, 507) Differentiation of thymocytes into Tc cells (with IL-2) (508) Effects on B and T cells B cells Proliferation of normal activated B cells and B lymphoma cells (BCLI) (301, 500-502, 504, 514, 515) Increased IL-2R expression on activated B cells or B cell lines (509-511) Increased IgM and IgG secretion by activated B cells (300, 301, 500-502, 504, 510, 512, 514, 515, 517) Enhanced IgGl and IgE secretion with IL-4 (300, 517) Increased IgA secretion by LPS-activated B cells and lymphoma cells (with IL-4) (300, 320, 624) (continued)
MULTIPLE EFFF.CE
TABLE VII (Continued) MURINET CELLS. CELLS,
OF CYTOKINES ON
Lymphokine"
Sources
IG6- [Interleukin HPl (596, 597), plasmacytoma growth factor (PCT-GF) (598), B cell stimulatory factor-2 (BSF-2) (human) (60S), IFN-82 (human) (601, 602), 26-kDa protein (human) (600, 601)]
Activated T cells, macrophages (P388D1) and other cells (596, 598, 599)
AND
ACCESSORY CELLS Biological activities
Effects on lymphoidhyeloid precursors Support of colony formation by normal hematopoietic precursors in the presence of other factors (CSFs) (599, 625) Costimulatory activity on thymocytes (similar to IL-1) (599)
Effects on B and T cells B cells Proliferation and Ig secretion by activated normal B cells and lymphoblastoid lines (596-598, 603-604)
Hybridoma/plasmacytoma growth factor (596-600) T cells T cell activating factor (599, 606, 607) Cytotoxic T lymphocyte activity with IL-2 and IFN-7 (605) or in primary MLC (606) Cofactor for the expression of IL-2Rs (human T cells) (608)
IFN-y - [Macrophageactivating factor (MAF) W8)l
Activated Thl cells (75, 260) and Tc cells (615)
"Synonymous terms are given in brackets.
Effects on B and T cells B cells Differentiation factor for B cells (527, 565, 566) Induction of IgG2, production in the presence of LPS (265, 321, 568, 569) Decreased IL-4-induced activities, e.g., Ia expression (267, 269, 270); size increase (267); anti-Ig proliferation (266); FceR expression (476); IgG, and IgE secretion (268, 271, 300, 321) T cells Inhibition of IL-2- or IL-4-induced proliferation of Th2 but not Thl cells (280, 299) Effects on accessory cells Increased Ia expression (559-562) Increased FcyR expression (562) Induction of antimicrobial and tumoricidal activity (563) Enhancement of TNF and LT functions (564)
62
ELLEN S. VITETTA ET AL.
might also affect the function of B and T cells. In general, although the release of lymphokines is a direct consequence of antigenic stimulation, the effects of such lymphokines are not antigen specific. It should be noted, however, that stimulation of certain T cells has been reported to result in the release of soluble antigen-specific helper factors (466, 467). The role played by these factors in T cell/B cell responses is highly controversial, and will not be discussed in this review. A. Th CELLSUBSETS AND B CELLACTIVATION As described in the previous section, a number of studies (75, 261, 263, 300, 462) suggest that the Th2 subset is more effective in providing help to B cells because of the secretion of IL-4 and IL-5. In contrast, the ability of Thl cells to provide help to B cells is more controversial (75, 261-263, 300, 462). Although some clones of Thl cells can provide help to B cells (75, 265, SOO), it is generally accepted that they are not as effective as Th2 cells, probably due in part to their production of IFN-y, a known inhibitor of several B cell functions (266-271), and their failure to secrete IL-4 and IL-5. Sanders et al. (326) have recently demonstrated that virtually all antigen-specificB cells that can physically interact with T cells in an antigen-specific, MHC-restrictedfashion can do so with either type of T h cells. Thus, if B cell activation and differentiation can be induced by different Th subsets, secreting different lymphokines, then at least two different pathways of B cell activation should exist. B. B CELLACHVATIONBY Th2 CELLS-THEROLES OF IL-4 AND IL-5 1. IL-4
“Resting”Th2 cells do not secrete IL-4 or IL-5 (75, 260, 277). In 2-4 h following antigenic stimulation, IL-4- or IL-5-specific mRNA can be detected in Th2 cells (260). Lymphokine activities can be detected in culture SNs within 4-6 h, reaching a peak after 24 h (277). No differences have been observed in the pattern of IL-4 secretion by T-286 cells (a Th2 cell line) when they are stimulated with antigen on different APCs (e.g., splenic adherent cells, peritoneal macrophages, or TNP-ABCs) (277). However, as noted in an earlier section, TNP-ABCs are much more effective at capturing antigen at lower concentrations and, hence, make outstanding APCs when low doses of antigen are used in vitro (158, 169, 172, 184, 187, 468). a. Effects of IL-4 on B Cells. IL-4 was first described by Howard et al. (290) as a lymphokine that was different from IL-2 and that could synergize with anti-p antibodies to induce the proliferation of resting
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
63
B cells. It later became evident that IL-4 had a variety of effects which could be manifested on either resting or stimulated B cells. For example, IL-4 induced the increased expression of class I1 molecules on small, resting B cells (469, 470), and it prepared these cells to proliferate in response to anti-Ig (471, 472). These results suggested that IL-4 was a “progression”factor rather than a “growth” factor for resting B cells. IL-4 can also induce the increased secretion of IgGl after stimulation of resting B cells with LPS (284, 318, 473, 474) and is identical to the B cell differentiation factor for IgGl (BCDFyl) described by Isakson et al. (318, 473). It was later found that high concentrations of IL-4 also induced the secretion of IgE in B cells cultured with LPS (268). After the cloning and expression of IL-4 cDNA by two separate groups (292,293), it became evident that the activities of IL-4 were even more complex. IL-4 was reported to induce the proliferation of T cells (277, 280, 285-288, 294-296,298) and mast cell lines (293, 294,475), to induce the expression of FCEreceptors (CD23) on resting B cells (476) or B cell hybridomas (477), and to act on several other hematopoietic lineage cells, including mature macrophages (478-480) and precursors of macrophages, granulocytes, megakaryocytes, and erythrocytes (481, 482). From this brief description, it is clear that IL-4 can modulate the B cell response at many points in its differentiatiodactivation pathway. Its ability to do so may be programmed into the B cell when it is still in Go. How does IL-4 work during T cell/B cell collaboration? After cognate interaction between a Th2 and a B cell has occurred, secreted IL-4 could act in combination with putative signals provided by the physical interaction itself to activate the B cell and induce its proliferation. In fact, it has been reported by DuBois et al. (483) that a cloned T cell line which does not secrete IL-4 can induce the proliferation of B cells only if exogenous IL-4 is provided. IL-4 could also act on resting B cells that have not yet undergone cognate interactions with T cells to induce increased expression of class I1 molecules and prepare them to interact with T cells. In this way, IL-4 may enhance the efficiency of T cell/B cell collaboration. For this to happen, one would presume that the T cell with which the B cell interacts has already been activated by another APC and is secreting IL-4 before it interacts with a B cell. In this regard, Sanders et al. (396) have reported that treatment of small TNP-ABCs with IL-4 enhances the ability of such cells to physically interact with Th cells, an effect probably mediated by increased expression of class I1 molecules. The enhanced ability of IL-4-treated B cells to form conjugates with T h cells, correlated with a higher plaque-forming cell (PFC) response in treated versus untreated cells. Stein et al. (484) reported that IL-4 can act as a costimulant in
64
ELLEN S. VITETTA ET AL.
the proliferative response of B cells to type I1 thymus-independent (TI) antigens. Thus, IL-4 also acts on B cells by preparing them to proliferate after interaction with anti-Ig (471, 472), antigen (TI/type 11) (484), or antigen and T h cells (TD antigens) (483). IL-4 acts on resting B cells (subsequently cultured with LPS) to induce the secretion of IgGl and IgE and to decrease the secretion of IgGs (268, 284, 318, 319, 321, 473, 474, 485), indicating that IL-4 is an “early-acting” factor that can program “later events.” Although it is not yet clear whether B cells stimulated during cognate T cell/B cell interaction behave similarly to LPS-activated B cells, it is plausible that if IL-4 is secreted during cognate T cell/B cell interaction it should “prime” the involved B cell and perhaps other B cells to subsequently secrete IgGl. In fact, both Snapper and Paul (329) and Bergstedt-Lindqvist et al. (486) have reported that IL-4 can program uncommitted B lymphocytes to switch to IgGl and IgE production. The decisive requirement for IL-4 in TD antigen-induced IgGI production in Vitro has been confirmed by Stevens et al. (265); Th2, but not Thl, cells (which do not secrete IL-4) induce IgGl secretion in TNP-ABCs; IgGl secretion can also be induced if Thl-activated B cells are cultured with exogenously added IL-4. Th2, but not Thl, cells induce IgE responses (75, 260, 300). This activity can be inhibited by anti-IL-4 antibodies, demonstrating that IL-4 is required in IgE production. IL-4 is also involved in IgE production in vivo in mice infected with the helminth Nippostrongylus bmsiliensis (487). Although IL-4 programs the secretion of both IgGl and IgE, Snapper et al. (322) have suggested that their production is differently regulated, based on the requirement for higher IL-4 concentrations to induce a polyclonal IgE response. 6. Effects of IL-4 on T Cells. A number of laboratories have reported that in addition to its effects on B cells, IL-4 also stimulates the proliferation of several T cell lines (277, 280, 285-288, 294-296) or resting T cells in combination with PMA (488). IL-4 also functions as the autocrine growth factor for antigen-stimulated Th2 cells (277, 285, 286). Although it has been reported that resting T cells express receptors for IL-4 (489-492), Fernandez-Botran et al. (277, 280) have observed that “resting” Th2 cells do not proliferate in response to IL-4, but become responsive to the growth-promotingeffects of IL-4 after antigenic stimulation. These results suggest that IL-4 induces proliferation only in cells that have been previously stimulated by antigen or mitogen. It is not clear whether IL-4 functions as an “activation” factor to prepare resting T cells to proliferate in response to antigen and APCs (in the same manner that it prepares resting B cells to proliferate in response to antiIg), or whether it acts as a “true” growth factor for T cells.
CELLULAR INTERACTIONS IN T H E HUMORAL IMMUNE RESPONSE
65
In the context of T cell/B cell collaboration, it would be expected that following the activation of Th2 cells and subsequent lymphokine secretion, IL-4 would drive the activated Th2 cells to proliferate. Expansion of the T h cell pool would occur for 24-48 h, during which time the IL-4 would be secreted (277, 283). The responsiveness of the Th2 cells to IL-4 would slowly decrease in the absence of additional stimulation by antigen (280). IL-4 has additional effects on T cells. For example, IL-4 acts as a helper factor for the generation of Tc cells in unprimed and primed mice (493-495). In addition, IL-4 has been reported to act on thymocytes, stimulating their proliferation and differentiation (496). IL-4 synergizes with PMA in the generation of Tc cells from resting T cells (495) and has growth-promotingactivity on activated murine thymocytes (496-499). c. Effects of IL-4 on Other Cells. IL-4 induces increased expression of class I1 molecules on macrophages and enhances their cytotoxic activity (478-480). This effect is reminiscent of the effect of IL-4 on the expression of class I1 antigens on B cells. IL-4-treated macrophages also present antigen more effectively to T cells (480). Another important activity of IL-4 is its growth-promoting activity on mast cell lines in the presence of IL-3 (293, 294, 475). Mast cells mediate immediate hypersensitivity reactions by releasing histamine and other vasoactive substances. Thus, IL-4 not only acts by enhancing the humoral side of the hypersensitivity responses (increased IgE production), but enhances the effector phase (mast cell activation) as well. Finally, IL-4 has been reported to have effects on hematopoietic cell precursors (481, 482). Peschell et al. (481) and others observed that although IL-4 is not a colony-stimulatingfactor (CSF), it synergized with other CSFs such as granulocyte (G)-CSF, erythropoietin, monocyte (M)-CSF, a megakaryocyte-promoting activity, and IL-3 to stimulate colonies of several lineages, such as granulocytes, erythrocytes, monocytes, megakaryocytes, and mast cells, respectively. 2. IL-I
In the first reports that SNs of activated T lymphocytes could replace many of the functions of the T cells themselves, particularly the induction of Ig secretion by B cells, these activities were termed “T cell replacing factors,” or TRFs (74, 459, 500). Takatsu et al. (500) first established a T cell hybridoma, B151K12, which secreted a factor which induced B cell differentiation and was different from other known lymphokines. Swain and Dutton independently described a factor, originally called (DL) B cell growth factor (BCGF) (501) and later BCGF I1 (502), which induced the proliferation of normal B cells cultured with dextran sulfate
66
ELLEN S. VlTETTA ET A L .
or the leukemic B cell line, BCLl. Further experiments demonstrated that this factor was different from IL-4 and that it could not synergize with anti-Ig to induce the proliferation of normal B cells (502). It was later demonstrated that TRF had BCGF I1 activities copurified, suggesting that TRF and BCGF I1 might be the same lymphokine (503-505). It also became evident that TRF/BCGF I1 had even more complex biological activities. Sanderson et al. (506, 507) found that a lymphokine called eosinophil differentiation factor (EDF) and BCGF I1 were identical. More recently, other activities have been ascribed to TRFIBCGF II/EDE These activities include the ability to increase IL-2R expression on antigen-stimulated thymocytes, inducing them to become Tc cells in the presence of IL-2 (508), and to enhance IL-2R expression on activated B cells (509-511). The cloning and expression of cDNA for TRF/BCGF II/EDF (503, 505) confirmed that all three activities were mediated by a single molecule, which has now been designated IL-5. IL-5 is secreted by activated Th2, but not Thl, cells (260), although this point remains controversial (263). a. Effect oflL-5 on B Cells. Although both IL-4 and IL-5 have multiple biological effects, IL-5 in contrast to IL-4, appears to act primarily on activated B cells (512, 513), inducing their proliferation and differentiation into IgM-secreting cells. Thus, in the context of a cognate T cell/B cell interaction, IL-5 should act on B cells already activated by antigen, IL-4, and T h cells and induce these activated B cells to proliferate and secrete IgM (300, 301, 500-502, 512, 514). In this sense, IL-5 might induce both antigen-specific and polyclonal responses (262, 325). IL-5 is a good candidate for the “nonspecific” activity associated with the non-MHC-restricted,nylon-wool-adherentclass of T h cells described in earlier studies (253, 254). In addition to its ability to induce IgM secretion, IL-5 has also been reported to induce differentiation of TNPABCs into IgM- and IgG-secretingcells (500, 515). IL-5 has been reported by Mosmann and Coffman (283) and Murray et al. (324) to have enhancing effects on IgA secretion by LPS-stimulated B cells. It is not yet clear whether IL-4 and IL-5 act sequentially on the same B cells or whether different B cells respond to each lymphokine. Nonetheless, the fact that IL-4 acts primarily on resting B cells, whereas IL-5 acts primarily on activated B cells (512, 513), suggests that the presence of both lymphokines would result in enhanced proliferation and differentiation of B cells activated by T cells and antigen. IL-5 synergizes with IL-4 in a number of experimental models (300, 514, 516, 517). Recently, another interesting biologic activity of IL-5 has been described. Harada et al. (510) have observed that recombinant (r)IL-5 can induce the expression of IL-2 receptors (Rs) and responsiveness to IL-2 in B cells from dinitrophenyl
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
67
(DNP)-carrier-primed mice. It could be postulated that the biological significance of the induction of functional IL-2Rs by IL-5 on B cells is that such cells would then acquire responsiveness to IL-2, a lymphokine secreted by Thl cells. This would provide a means of regulating Ig production by both subsets of T h cells and suggests that the temporal order of lymphokine activities on B cells might be IL-4 IL-5 IL-2. In addition to the B cell stimulatory effects of IL-4 and IL-5, as noted earlier, LeClercq and colleagues (442, 518) have reported that certain clones of T h cells secrete a lymphokine, termed B cell activation factor (BCAF). BCAF can act on small, resting B cells, inducing polyclonal proliferation and differentiation in the absence of T h cells. Whether production of BCAF is restricted to a particular T h subset (Thl versus Th2) and what the relative frequency of this subset is remain subjects of investigation. Nonetheless] this lymphokine is another candidate for the “nonspecific,”non-MHC-restrictedhelper activity described in earlier reports. 6. Effect of IL-5 on T Cells. No effects of IL-5 on the proliferation of resting or activated T h cells have been reported. However, IL-5 can induce the generation of Tc cells from peanut agglutinin-positive (PNA+)thymocytes in the presence of IL-2 (508). This activity is probably mediated by the IL-5-mediated induction of IL-2R expression on thymocytes followed by acquisition of responsiveness to IL-2. c. Effect of IL-5 on Other Cells. As previously mentioned, IL-5 is identical to the EDF described by Sanderson (503, 506, 507). I L 5 induces the formation of colonies of eosinophils from bone marrow cultures and induces their proliferation. Eosinophils express FceRs and are capable of binding the Fc portion of IgE, after which they release a number of mediators of inflammation. In addition, the importance of eosinophils in the antibody-mediated killing of parasitic organisms has been demonstrated (519-521). In conclusion, IL-5 can potentiate B cell proliferation and differentiation and, in conjunction with IL-4, is involved in local immunity (via IgA production). IL-5 might also play a role in the immune responses to certain parasitic organisms.
-
-
C. B CELLACTIVATIONBY Thl CELLS-THEROLES OF IL-2 AND IFN-7 1. IL-2
a. Effects of IL-2 on B Cells. IL-2 (originally T cell growth factor, or TCGF) was described as an activity in the SN of mitogen-activated T cells that induced the proliferation of T cells (289, 522). The role
68
ELLEN S. VITETTA ET A L .
of IL-2 in the proliferation of activated T cells or T cell lines and in the development of Tc cells has been extensively studied (289, 522-526). The role played by IL-2 in B cell proliferation and differentiation is less clear. Early studies (527-530) suggested that IL-2, in combination with other factors, acted synergistically to induce a primary anti-SRBC response in vitro. These helper factors were later identified as IFN-y and IL-1 (527). In contrast, other studies suggested that B cells lacked IL-2Rs and could not respond to IL-2 (279, 531, 532). More recently it has become clear that, although resting B cells express very few, if any, I L - ~ R sat , least some activated B cells express functional IL-2Rs (532-537). Zubler et al. (533) reported that B cells stimulated with a mixture of LPS and anti-Ig expressed IL-2Rs and responded to IL-2, while B cells stimulated with LPS alone did not. Activated B cells (537) can respond to low doses (0.1-10 U/ml) of IL-2 and this effect is mediated via high-affinity IL-2Rs (533, 535-538). In other studies using higher doses of IL-2 (100-1000 U/ml), a greater range of responsiveness by B cells has been demonstrated. Nakanishi et al. (516) reported that although IL-2 did not display significant “TRF” activity at doses lower than 50 U/ml, it induced proliferation and IgM secretion in cultures of anti-Ig-stimulated B cells at doses >lo0 U/ml and in combination with IL-4 and a second T cell-derivedfactor. It has been suggested that the effect at high IL-2 concentrations is not mediated via “classical” high-affinity IL-2Rs (538, 539). In this regard, a number of laboratories (540-546) have reported that high-affinity IL-2Rs (& = 1 X lo-” M) are composed of at least two different subunits, both of which can bind IL-2, but with different affinities. One (p55) is the “classical”Tac antigen (human). It binds IL-2 M). The other subunit (p70-75) with low affinity (& = 1 X M). Together, p55 binds IL-2 with intermediate affinity ( K d =1 X and p75 form a high-affinity IL-2R (Kd = M) (547, 548). Some T cell lines express only low-affinity (p55) or intermediate-affinity (p75) IL-2Rs (540-544). Moreover, the presence of intermediate-affinity IL-2Rs (p75) on resting T cells (549-551), lymphokine-activated killer cells (LAK) (550), natural killer (NK) cells (550, 552), activated human B cells (537, 538, 551), and macrophages (551) has been reported. Although most of the biological effects of IL-2 are thought to be induced via receptors of high affinity (p75 + p55) (278, 279, 553), it has been reported that receptors of intermediate affinity (i.e., p75) can transmit signals after binding IL-2 (547-552). The activation of NK and LAK cells is probably mediated via p75 (550, 552). Saiki et al. (538) and Tanaka et al. (537) have reported that large human tonsillar B cells or B cells stimulated with Stufihylococcus aureus (Cowan I) (SAC) express
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
69
p75 and can proliferate and differentiate into Ig-secreting cells in the presence of high concentrations ( >lo0 U/ml) of IL-2. Although it is not clear whether munne B cells resemble human B cells in this regard, these results suggest that resting B cells might express intermediate affinity IL2Rs under certain conditions and that they might be able to proliferate and/or differentiate in response to high concentrations of IL-2. When B cells are activated by certain stimuli, the expression of p55 might be induced on cells already expressing p75, resulting in the formation of high-affinity IL-2Rs (547, 548). Such cells would be expected to respond to IL-2 at much lower concentrations when compared to cells expressing only p75. The previously described IL-2R-inducing activity of IL-5 thus becomes very important. It could be hypothesized that IL-5 would induce the expression of p55 on B cells that already express p75, leading to the expression of high-affinity IL-2Rs. According to this hypothesis, B cells would have to first encounter IL-5 (secreted by a Th2 cell) before they could respond to IL-2 (secreted by a Thl cell). Thl cells would therefore amplify responses to IL-5. On the other hand, at least some Thl clones can activate B cells via cognate interaction and in the apparent absence of Th2 cells or their SNs. Whether T cell/B cell interaction can substitute for IL-5 in rendering B cells responsive to IL-2 or whether some Thl clones do secrete IL-5 and IL-2 (262, 263) remains to be determined. In conclusion, IL-2 appears to have both proliferative and differentiative effects on B cells, although additional lymphokines might be required for the optimal expression of such activities. In addition, the relative concentration of I L 2 is probably important in determining which B cells will be most responsive to 1L-2, although large (or activated) B cells are the most likely candidates. b. Effect of IL-2 on T Cells. The proliferative activity of IL-2 on T cells has been well documented (289, 522, 523). In the case of Thl and Th2 cells, clones of both types respond poorly to IL-2 when they are in a "resting" state (277, 280). However, after antigenic or mitogenic stimulation, responsiveness to IL-2 is greatly enhanced (277, 280). This is paralleled by an increase in the expression of IL-2Rs as determined by immunofluorescence or by the binding of radiolabeled IL-2 (278-280, 553). IL-2 is an autocrine growth factor for Thl cells in vitro (285). Interestingly, although Th2 cells utilize IL-4 as their autocrine growth factor, following stimulation they upregulate their levels of IL-2% in much the same manner as Thl cells (277, 280). In fact, IL-2 is better than IL-4 as a growth factor for Th2 cells (280, 285, 286). Such observations suggest that for an optimal Th2 response to occur, IL-2 would have to be present and that, in vivo, IL-2 would play an important role in the proliferation of activated Th2 cells. As noted above, responsiveness
70
ELLEN S. VITETTA ET AL.
to IL-2 and IL-4 decreases with time after antigenic or mitogenic stimulation, rendering the T h cells unresponsive in the absence of further stimulation (280). c. Effect of ZL-2 on Other Cells. In addition to its effects on T and B cells when used at high concentrations, IL-2 augments the cytotoxicity of human monocytes (554) and the activity of NK and LAK cells (550, 552). As mentioned earlier, these results suggest that IL-2 might be acting on such cells via p75 (550, 552). They also show that the effects of IL-2 are not confined exclusively to lymphocytes. 2 . ZFN-y
a . Effect of IFN-y on B Cells. IFN-y, a lymphokine secreted by activated Thl (75, 260), Tc (555), and NK cells (556), has a variety of roles in the immune response. IFN-y has both antiviral and antiproliferative activities (557). It also plays a major role in macrophage activation (558-564) and it regulates the activities of cells involved in both humoral and cellular responses (557). IFN-y synergizes with IL-2 and IL-1 in the generation of anti-SRBC antibody responses (527) and promotes Ig secretion by resting or activated murine and human B cells (527, 565, 566). These observations suggest that IFN-y has a differentiation function (565, 566). Paradoxically, recent studies have suggested that IFN-y has potent inhibitory effects on B cell activation, proliferation, and differentiation (262, 266-270, 321). In this regard, IFN-y is inhibitory for IL-4. For example, IFN-y inhibits the IL-4-mediated expression of class I1 antigens (267, 269, 270) and CD23 molecules (FceR) (476) on resting B cells. IFN-y also has inhibitory effects on the proliferation of B cells stimulated by anti-Ig and IL-4 (266) and affects B cell differentiation by inhibiting the IL-4-mediated secretion of IgGl and IgE (268, 271, 300, 321). At the same time, it promotes the secretion of IgG2, (265, 567-569) in LPS-stimulated or antigen plus Thl-stimulated B cells. IFN-y-containing SNs from T h l cells (or inflammatory T cell clones) appear to have a suppressive effect on the Ig production of Th2-stimulated B cells (263, 300). Stevens et al. (265) recently reported that Thl, but not Th2, cells induce lgGza secretion by antigen-specific B cells after antigen stimulation and that this activity can be partially inhibited by antibodies against IFN-y. Interestingly, IgM secretion was not affected by IFN-y. Zn uiw, IFN-y inhibits the secretion of IgGl and IgE in mice treated with anti-IgD antibodies (569). IFN-y also enhances IgGza secretion in mice immunized with Brucella abortus (569). These results suggest that, in addition to its pleiotropic effects on other cells, IFN-y can regulate B cell function by acting both as a B cell differen-
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
71
tiation factor, especially for the synthesis of IgGza and as an antagonist of the early IL-4-mediated effects on B cells. It could be hypothesized that, in vivo, the isotype(s) of Ig ultimately secreted after stimulation by a particular antigen will be determined by the relative numbers of activated T h l versus Th2 cells and the secretion of IL-4 versus IFN-y. Antigens that induce predominantly IgGza responses might preferentially activate Thl cells and antigens that induce I&, and/or IgE might activate Th2 cells. The molecular mechanisms underlying the antagonistic effects of IFN-y on the IL-4-induced activities on B cells are not clear. However, binding studies have demonstrated that IFN-y does not compete for IL-4 in binding to IL-4Rs (489-492), indicating that these two lymphokines do not compete for the same surface receptor. 6. Effect of IFN-y on T Cells. In addition to its antagonistic effects on the IL-4-induced responses of B cells, IFN-y has now been reported to be inhibitory to the IL-4- and IL-2-mediated proliferation of Th2, but not T h l , cells (280, 299). Fernandez-Botran et al. (280) reported that the IFN-y-mediated inhibition of Th2 proliferation was only partial (at doses below 100 Wml) and was mediated by an antiproliferativeeffect rather than a decrease in cell viability. These results suggest that IFN-y can regulate the action of Th2 cells not only by selectively decreasing their proliferation, but by antagonizing the effects of IL-4. c. Effect of IFN-y on Other Cells. As already mentioned, IFN-y plays a major role in the activation of macrophages for cell-mediated cytotoxicity and antimicrobial activity (559-564). In addition, IFN-y induces the increased expression of class I1 molecules and FcyRs on macrophages (559-562). Moreover, IFN-y-treated macrophages secrete increased levels of IL-1 (570), suggesting that IFN-7 probably enhances the APC and accessory cell function of macrophages. Finally, IFN-y might also be involved in direct lymphokine-mediated cytotoxicity of certain types of cells, since it has been shown to synergize with tumor necrosis factor (TNF) and lymphotoxin (LT), other lymphokines secreted by Thl cells (557, 564). In conclusion, Th2 cells support humoral responses, as well as both the humoral and cellular components of immediate hypersensitivity reactions and immune responses against certain parasites. The activities of Thl cells (IL-2, IFN-7) are related more closely to the development of immune responses against viruses, other intracellular organisms, and cell surface antigens to which cell-mediated cytotoxic responses are elicited.
72
ELLEN S. VITETTA ET A L .
D. OTHERINTERLEUKINS 1. IL-1
IL-1 is a polypeptide secreted by a variety of cells, including monocytes, endothelial cells, and epithelial cells (570-572). Monocytes are still considered to be one of the major sources of this lymphokine (571, 572). A number of agents can induce macrophages to secrete IL-1, e.g., immunological stimuli (activated T cells or their products, immune complexes, and complement components) and microbial stimuli (endotoxin, exotoxins, yeasts, viruses, etc.) (572). IL-1 is probably the most pleiotropic lymphokine, having one of the widest range of activities and cellular targets. One of its activities, pyrogenicity (573), was described as early as 1948; IL-1 was originally called “endogenous pyrogen.” A number of other activities, independently described, copurified with leukocyte endogenous pyrogen. Among those activities was a stimulant of hepatic acute-phase protein synthesis, called leukocytic endogenous mediator (574), a substance from activated monocytes that, together with mitogens, induced thymocyte proliferation [lymphocyte activating factor (575)], and several other substances stimulating cartilage resorption, muscle wasting, collagen and fibronectin synthesis (570-572), etc. After the cloning and expression of cDNA for IL-1 (576, 577), it was established that these and other activities were mediated by the same molecule. IL-1 exists in two biochemically distinct but related forms known as IL-la and IL-lP [(576, 577); reviewed in Ref. 5711. Although amino acid homology between these two forms is limited, both bind to the same cellular receptor and are functionally equivalent. In this section, we shall limit our discussion to the activities of IL-1 on cells of the immune system, especially as they relate to the activation of T and B cells. a. Effect of IL-1 on B Cells. IL-1 has the ability to promote the maturation of B cell precursors (578) and pre-B cells (579). Thus, IL-1 induces the expression of x light chain and membrane Ig in a pre-B cell line that produces only cytoplasmic p chains (579). IL-1 synergizes with other lymphokines or antigens to induce mature B cells to proliferate and/or differentiate (reviewed in Refs. 484, 527, 570, 580-582). For example, Liebson et al. (527) showed that IL-1, in combination with IFN-y, synergized with IL-2 in the induction of an anti-SRBC antibody response. Howard et al. (580) reported that IL-1 potentiated the proliferative response of B cells cultured with IL-4 and anti-Ig, particularly at low cell densities. Stein et al. (484) demonstrated that IL-1 synergized with IL-4 in combination with type I1 TI antigens, in the same manner. Thus, IL-1 can act as a cofactor to potentiate the proliferative responses
i
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
73
of B cells to a variety of stimuli. Although it is still not clear whether IL-1 is a critical factor in B cell responses, it can act in concert with other lymphokines to potentiate their proliferation and/or differentiation. b. Effect of IL-1 on T Cells. IL-1 also acts as a cofactor, along with other stimuli such as antigen, mitogen, and other lymphokines, to stimulate T cell proliferation (287, 288, 575, 583-585). IL-1 has been reported to induce IL-2 secretion and IL-2R expression by activated T cells (570, 572, 583-585). More recently, some Th2 cell lines have been reported to require IL-1 in order to proliferate in response to IL-4 (287, 288, 585). On other Th2 cell lines, IL-1 is not absolutely required, but can synergize with other lymphokines such as IL-2 and IL-4 (280). Such results confirm the role of IL-1 as a cofactor in T cell proliferation and also suggest that there may be heterogeneity among Th2 cells with regard to their absolute requirement for IL-1. Alternatively, the requirement for IL-1 might be related to the state of activation of the T h cell. For example, the need for IL-1 might increase as the state of activation of the T h cell decreases with time after antigenic or mitogenic stimulation. Thus, Kupper et al. (286) reported that antigen-stimulated D.10G4 cells proliferate in response to IL-4 in the absence of IL-1 shortly after stimulation, whereas IL-1 was required to support IL-4-mediated proliferation later on. In contrast, IL-1 does not act as a cofactor for IL-2/IL-4-mediated proliferation of T h l cells (280, 287, 288). Kurt-Jones et al. (288) have shown that Th2 cells express more high-affinity IL-1R than do Thl cells. These results seem paradoxical in light of earlier reports that IL-1 acted by inducing IL-2 release from T h l cells, since Th2 cells, which secrete IL-4 and not IL-2, are more responsive to IL-1. c. Effectof I L l on APCs. Macrophage function is profoundly altered by IL-1. The multiple effects of IL-1 on macrophages include the induction of the synthesis of prostaglandin & (PG&), IL-1, CSFs, other cytokines, and collagenase and the induction of increased cytotoxicity (571). In terms of antigen presentation, it is not clear whether IL-1 is required for T cell activation, although under certain conditions the addition of IL-1 seems to restore the ability of macrophages to induce T cell activation after the depletion or inactivation of APCs (570). IL-1 secretion is only one aspect of accessory cell function; it is possible that different T cells or different activation signals may vary in their dependency on IL-1 stimulation. 2 . IL-3
IL-3 is a lymphokine secreted by activated T cells (75, 260, 586) and some myelomonocytic cell lines such as WEHI-3B (586, 587). Both Thl and Th2 cells secrete IL-3 (75, 260).
74
ELLEN S. VITETTA ET A L .
a. Effect of IL-3 on Nonlymphoid Cells. In contrast to the other lymphokines discussed in this section, the major targets for IL-3 are immature myeloid progenitor cells, rather than mature lymphocytes. IL-3 induces the proliferation and differentiation of granulocytes, macrophages, and erythroid, megakaryocytic, and mast cell precursors (586, 588-592). Several mast cell lines or other hematopoietic cells are dependent upon IL-3 for their growth (586, 591, 592). The isolation and expression of IL-3 cDNA has confirmed that the multiple activities on a variety of hematopoietic cell lineages are mediated by IL-3 (591, 592). b. Effect of IL-3 on T and B Cells. The effects of IL-3 on cells of the lymphocytic lineage are controversial. It has been suggested that IL-3 is able to support the in m2ro growth of cells from bone marrow that will generate T and B cells when injected into irradiated recipients (593). It is not clear from these results whether IL-3 acts on multipotential stem cells committed to the lymphoid lineage or whether it programs the cells to enter such a lineage (593). It has been reported by Palacios et al. (594), however, that IL-3 can support the growth of a population of B cell precursors, but not mature B lymphocytes. In addition, a number of clones with characteristics of pro-B and pre-B cells that are dependent on IL-3 for continuous proliferation have been established (306, 595). Similarly, cultures of clones expressing low levels of surface Thy-1 and Ly-1 antigens are able to give rise to mature T lymphocytes after in mvo transfer (prosy cell clones) and to proliferate in response to IL-3 (595). In summary, in addition to both its activities on myeloid precursors and mature mast cells, IL-3 could be involved in the early stages of the maturation of T and B cells. These properties implicate IL-3 (in combination with other factors) as a potential link between the immune response, hematopoiesis, and the recruitment of effector cells such as monocytes, granulocytes, and mast cells.
3. IL-6 The multiple activities of IL-6 were independently described before it became evident that they were mediated by the same lymphokine. In the mouse, these activities included a hybridoma growth factor (HGF, or interluekin HP-1); a T cell-derived factor that stimulated the proliferation of B cell hybridomas and plasmacytomas (596, 597); a factor derived from the murine macrophage cell line, P388D1, which also supported the growth of hybridomas and plasmacytomas (598); and a growth factor for hybridomas produced by a variety of other cells, e.g., endothelial cells, peripheral blood cells, and fibroblasts (599). In the human, similar activities have also been described and related to a protein designated as a 26-kDa protein (600, 601), IFN-02 (600, 602), and B cell
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
75
stimulatory factor (BSF)-2 (603). BSF-2 induced Ig secretion in B cells and lymphoblasts and had potent growth-promoting activity for hybridomas and plasmacytomas. Cloning and expression of the cDNA for IL-6 (602, 603) has confirmed that these multiple activities are mediated by the same molecule. IL-6 is secreted by a large variety of cells and has pleiotropic activities on a number of cellular targets. Moreover, the activities of IL-6 resemble those of IL-l, including pyrogen activity and an activity mediating costimulation of thymocytes (599). a. Effect of IL-6on B Cells. The activity of IL-6 as a growth factor for hybridomas and plasmacytomas has already been mentioned. In the human, IL-6 is a late-acting factor that promotes Ig secretion by activated B lymphocytes and lymphoblastoid cell lines (603, 604). In activated, normal B cells, IL-6 induces Ig secretion without proliferation. In contrast, it is a potent proliferative factor for lymphoblasts (604). Taken together, these results suggest that IL-6 acts on B cells after IL-4, IL-5, and IL-2. b. Effect of IL-6on T Cells. IL-6 acts as a cofactor for the generation of Tc cells from murine thymocytes in the presence of IL-2 and IFN-./ (605) and enhances differentiation of murine Tc precursors in primary allogeneic mixed-lymphocyte cultures (606). IL-6 also acts as a T cellactivating factor, promoting the induction of IL-2 expression by T h cells after mitogen stimulation (599, 607), or as a cofactor, inducing T cell proliferation after stimulation with mitogen or anti-TcR antibody (606). In addition, IL-6 is a cofactor required for the expression of IL-2R and IL-2 responsiveness by anti-CD28-activated human T cells (608). Although activated Th cells secrete IL-6, it is not clear whether its production is restricted to a particular T h cell type (Thl or Th2). c. Effect of IL-6on Other Cells. IL-6 can synergize with the IL-3 to promote the growth of hematopoietic progenitor cells (599, 607). Thus, IL-6, like IL-1, may function as a link between immune responses and inflammatory responses. Thus, its pleiotropic activities range from the enhancement of antibody secretion, T cell activation, and generation of Tc cells to mediating fever, acute-phase protein production, stimulation of hematopoiesis, and recruitment of effector cells. E. CONCLUSIONS In Sections I11 and IV, we discussed the two types of functionally different T h cells and the role of lymphokines in the regulation of T cell/B cell function. Several conclusions can be drawn: (1) The functional differences between the two T h cell subtypes can be accounted for, at least in part, by the lymphokines that each type of T h cell secretes after antigenic stimulation. (2) Lymphokines are pleiotropic, acting on
76
ELLEN S. VITETTA ET A L .
different cell types and on cells not directly involved in the immune response. (3) Different lymphokines act at different stages of T cell/B cell development, although overlaps are common. (4) One lymphokine can induce a particular cell type to express receptors for a second lymphokine, thereby making cells responsive to the effects of the second lymphokine. ( 5 ) Alternatively, a particular lymphokine can antagonize the effects of another lymphokine on the same cell. (6)Lymphokines can act on lymphoid and/or myeloid cell precursors, regulating their maturation and recruitment. Although Thl versus Th2 cells have different effects on B cells and other cells in d r o , their role in vivo is less well understood. Nonetheless, studies in vivo have confirmed the requirement for IL-4 and IFN-y in IgE (487) and IgG2, (569) responses, respectively. In light of our current knowledge of T h cells and lymphokines, it is probable that as the immune system has evolved, it has developed a very complex set of interactions among its components that enable it to respond efficiently to a similarly complex variety of antigenic insults (i.e., extra- and intracellular organisms, malignant cells, and soluble toxins). Hence, the nature of the antigenic stimulus probably determines the nature of the immune response against it. Therefore, different types of antigenic challenges might preferentially activate a particular type of T h cell, i.e., Th2 cells seem more efficient at providing help to B cells (humoral immunity) and, because of the roles of IL-4 and IL-5 in IgE and IgA production, they may be important in hypersensitivity, immunity against certain parasites, and local immunity; Thl cells seem to be more efficient at mediating DTH reactions and cytotoxic responses against cellular targets or intracellular organisms. Again, lymphokines such as IFN-y and IL-2 may be involved. In Fig. 3 we have schematically summarized the activities of the lymphokines secreted by Thl and Th2 cells with special reference to their role in B cell activation, proliferation, and differentiation and the functions of effector cells. Additional experimentation is required in order to understand how different T h cells are generated and activated in vivo. It remains to be determined how activated Thl or Th2 cells and/or their lymphokines induce B cell activation and how lymphokines produced by one T h cell subset regulate the activities of the other T h subset or its lymphokines. Such information would not only be invaluable in the understanding of how immune responses are normally regulated, but also on how certain aspects of the immune response (i.e., hypersensitivity, antibody production, and cytotoxicity) could be potentially suppressed or enhanced in mvo for therapeutic purposes.
--
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
MATURATION/ RECRUITMENT
ACTlVATlOE
Mast calls
t
Mami CSIIS Eoslnophlli
-
i
PROLIFERATION DIFFERENTIATION
Eomlnaphlls
-?
t
t
I
I
77
EFFECTOR CELLS
1
!
Autocrina, pomsrina
Q
Mymlo/manacyllc
6-
! I
@
Bone Marrow Precursor
i
; +?
I
FIG.3. Regulation of the immune response by T h cell subsets and their secreted lymphokines. B cells can be specifically activated by physical interactions with Thl or Th2 cells. Alternatively, large B cells can be nonspecifically activated by lymphokines. Of the lymphokines released by ThZ cells, IL-4 can function by inducing B cells to express higher levels of class I1 and CD23 molecules, inducing their entry into cycle (GI), enabling them to proliferate to a variety of stimuli and programming them to secrete IgG, and IgE following stimulation. IFN-7, secreted by Thl cells, can antagonize all the effects of IL-4 and can instead program the B cells to secrete IgG2,. IL-5 acts on cycling B cells to induce their continued proliferation, differentiation, and expression of IL-2Rs. IL-2 (from Thl cells) could also act on such cells via IL-2R to induce their proliferation and differentiation. IL-6 functions as a late-acting factor, stimulating proliferation and/or Ig secretion by activated B cells and plasma cells. IL-3 plays a role in the maturation of bone marrow precursors and immature lymphoid cells, although the latter is controversial. IL-1 and IL-6 can act as cofactors in the presence of other stimuli or lymphokines to enhance maturation, activation, and proliferation of T and B cells. Finally, these lymphokines also have effects on the maturation, proliferation, and activation of nonlymphoid cells such as mast cells, eosinophils, or macrophages, which have important roles in the effector phase of immune responses. The effector cells induced by each T h cell subset and their activities are listed in the far right-hand column.
78
ELLEN S. VITETTA ET A L
VI. Cellular Interactions in Wlvo-A Synthesis
In this section, we combine well-known facts about the immune system and how it responds to different pathogenic organisms with a contemporary view of the cellular interactions and effector mechanisms involved. Much of this discussion will be highly speculative and certainly controversial. Nevertheless, in our view, it is exactly this type of synthesis that is sorely lacking in most reviews on the cellular basis of immunity. As a result, both immunologists and nonimmunologists often have great difficulty (and we are no exception) in understanding and devising strategies to treat dysregulations of the immune system. In this regard, there is perhaps no better example of our ignorance in relating the etiology of an infectious disease to the immune system than in the case of acquired immune deficiency syndrome (AIDS). AIDS is a generally fatal viral disease which continues to be of great concern and confusion to immunologists. We know that the causative agent of the disease is the human immunodeficiency virus (HIV). There is a growing body of literature on the structure of the virus, its heterogeneity, the organization of its genome, and the mechanism by which it enters cells. There is also an increasing number of papers describing the many cell types that are infected with virus and the final decay of the immune system in overt AIDS. However, there is no clear understanding of how the infection progresses from latency to overt disease and how cellular interactions per se play positive versus negative roles in maintaining the latent state versus causing a collapse of the immune system. Thus, a major goal of cellular immunology is to bridge the gap between understanding the immune response to infectious organisms and its successful application to prevention and treatment of disease. This will become possible only when we understand how the immune system works at its most basic level and how the various components of the system interact and are regulated. For the sake of brevity, we will use only a few specific examples in our discussion and will not attempt to reference the many well-known and important papers that provided their factual basis. For this, we refer the reader to one of the many excellent microbiology/ immunology textbooks. Although this synthesis is incomplete and speculative, it brings to light the issues which we must understand and address over the next decade if we are to successfully wage a battle against infectious disease. Figure 4 depicts a hypothetical model of the major cellular interactions involved in the generation of a humoral antibody response. Although many aspects of the model must be further clarified, the object of this discussion is to place the events depicted in the model in the setting of
CELLULAR I N T E ~ C T I O N S IN THE HUMORAL IMMUNE RESPONSE
79
a lymphoid organ where anatomical restraints will determine to what extent and by what mechanisms cellular interactions occur. More importantly, this must be viewed in the context of the whole animal, e.g., the trafficking of lymphocytes, other effector cells, and antibodies in the generation, maintenance, and down regulation of the immune response. For the individual to mount an effective immune response against a particular pathogen, a number of cellular and molecular events must be tightly regulated. These include the isotypes of antibody secreted by the progeny of the activated B cells, the generation of memory cells, and the activation of effector cells. An attractive working hypothesis to explain how cellular interactions are initiated and regulated is that the type of antigen and its route of entry will determine the site and mechanism by which it is metabolized by the host. This, in turn, will influence the types, numbers, and perhaps the sequence in which Thl and Th2 cells are activated. Once activated, these T h cells can then regulate, via cell contact and/or release of lymphokines, the isotype of antibody secreted by specific B cells as well as the additional types of effector mechanisms generated, i.e., mast cell activation, recruitment of eosinophils, macrophage activation, DTH, and the generation of Ts and Tc cells. ORGAN A. THELYMPHOID After local invasion of the body by pathogens or other antigens, cells in the draining lymph nodes interact to initiate an immune response. Within the lymph nodes, the B cells are sequestered in follicles, while the T cells (which enter the node through postcapillary venules) take up residence in paracortical areas. Soluble circulating antigen, antigen from the tissues, and antigen engulfed by macrophages enter the lymph node through the afferent lymphatics or through the capillaries. In the node, antigen is ingested by macrophages at the periphery of the cortex, in the germinal centers of the follicle, and in the afferent sinuses, diffuse cortex, and medulla. In contrast to macrophages, dendritic cells may not process antigen efficiently, but may bind antigen fragments released by macrophages. Antigen not ingested by APCs collects in the primary follicle and is retained on the FDCs in close proximity to the follicular B cells. There are several major mechanisms by which antigen from the blood and lymph can initiate cellular interactions in the lymph node. (1)When antigen concentrations are high, the macrophages within the nodes are very effective at ingesting organisms, degrading them in lysosomes, and either presenting antigenic fragments to T h cells or releasing antigen fragments which then bind to dendritic cells or B cells. Because of their
80
ELLEN S. VITETTA ET AL.
efficiency in taking up foreign material, macrophages may play a predominant role both as APCs and as a source of released antigen fragments. (2) The dendritic cells in the follicles very effectively capture immune complexes and retain this material on their plasma membrane for significant periods of time. Small antigen-containing portions of the membrane of the FDCs pinch off and presumably can be taken up by B cells in the follicle. FDCs serve as reservoirs for native antigen and hence may be the major source of antigen for the follicular B cells. It is also possible that FDCs present paucivalent epitopes to B cells in a multivalent form, since high antigen concentrations on the surface of FDCs would be operationally multivalent. This would result in more effective cross-linking of sIg molecules on B cells. (3) In contrast to macrophages, antigen-specific B cells have an advantage in capturing antigen because of their clonally expressed sIg receptors and their proximity to FDCs. Hence, when antigen concentrations are low, the B cells may be the more efficient APCs. Furthermore, the B cells may play a role in clonally expanding T cells activated by dendritic cells. Taken together, mechanisms 1-3 above suggest that, in the simplest model, B cells would bind native antigen and T cells would bind processed antigen on the surface of clustered macrophages and dendritic cells at the periphery of the germinal centers. Here, clonal expansion of both the B and T h cells should occur after antigen-specific, MHC-restricted interactions. The mechanism by which the antigen-specificT cells travel
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
81
across the paracortical areas to the germinal centers is not known, but this could be mediated by chemotactic factors secreted by APCs. After interactions between T h cells, B cells, and APCs, the activated T h cells secrete lymphokines that are responsible for the growth of both T cells and B cells as well as for the differentiation of the B cells into antibody-producing cells. Fully differentiated plasma cells migrate to the medulla of the lymph node and secrete antibody into the efferent lymphatics. Antibody secreted from all the regional lymph nodes is collected in the lymphatic system and enters the blood stream via the thoracic duct. In addition, memory cells in the node are probably FIG. 4. T cell/B cell interactions in the immune response; a synthesis. ( a ) Highaffinity sig molecules on B cells capture specific antigen. B cells process antigen, recycle fragments to their surface, and present fragments in association with class I1 molecules. B cells undergo early activation events (altered phospholipid metabolism, changes in membrane potential, increase in size, and increase in expression of class I1 molecules). (b) At the site of infection, or in the draining lymph nodes, antigen concentrations will be at their highest. Therefore, although antigen-specific B cells may not be available to capture antigen, other APCs (such as dendritic cells, macrophages, and Langerhans cells) will do so. These cells will take up antigen nonspecifically, process it, and present antigen fragments in association with class I1 molecules on their surface. In the case of dendritic cells, they will take up antigen fragments released by macrophages. FDCs will bind immune complexes. Resting T cells will be activated and then bind to specific antigen-presenting B cells. Note: T cells which become activated will either go directly to step c or become memory T cells capable of going on to step c at a later time. (c) B cells expressing antigenic fragments in association with class I1 molecules interact with T cells (from step b) expressing TcRs specific for the complex of antigen fragments and class I1 molecules. “Tight” interactions involve various T and B cell-associated surface molecules in addition to class I1 molecules and TcRs, i.e., LST4, LFA-la, and Thy-1. Specific interaction induces reorientation of intracellular components associated with the release of soluble proteins. ( d ) The nature of the antigen probably determines the type of Th cell which is activated. Interaction induces Th cells to secrete lymphokines and up regulates IL-2Rs and IL-4Rs, and cells then become responsive to their growthpromoting effects. Physical interaction also induces B cells to become responsive to the lymphokines secreted by the Th cells. Although lymphokines are nonspecific, their effects are exerted primarily on those cells involved in the antigen-specific response via close cellular interactions. (e) Signals mediated by physical interactions and secreted lymphokines induce B cell activation, proliferation, and differentiation. The programming of switching may be directed by some of the released lymphokines (i.e., IL-4). (Jand g) Secreted lymphokines can also act on cycling B cells, inducing their proliferation and differentiation, thus resulting in an antigen-nonspecificpolyclonal response. (h) Primed (memory) B cells process antigen and present fragments in association with class I1 molecules to T cells. These B cells can focus antigen which is in soluble form or antigen in the form of immune complexes which are presented to them on the surface of FDCs, icoccomes, or macrophages. Physical interaction with T cells is basically similar to that of unprimed B cells except that conjugation occurs very rapidly and requires lower concentrations of antigen. Resting memory T cells may be able to interact directly with B cells and may not require activation by dendritic cells.
82
ELLEN S . VITETTA ET AL.
generated in the germinal centers by the dividing B cells. Perhaps when the supply of antigen is exhausted, cycling B cells return to their Go state and either remain in the follicles or leave the nodes in the efferent lymph and travel to other areas of the body (e.g., bone marrow). Factors influencing the migration of memory B cells to other locations may include the induction of homing receptors and/or the numbers of memory cells in the nodes, i.e., for a node to return to its resting state, memory cells would be released into the lymph. The isotype of antibody produced by plasma cells (e.g., IgG versus IgA) is probably controlled by the types of memory cells and Th cells in the different lymphoid tissues which are activated. The T cell-derived lymphokines from Thl or Th2 cells could mediate switching or induce B cells already expressing other isotypes to differentiate into plasma cells. Lymphokines would play a major role in regulating the secretion of different isotypes of antibody. Hence, IL-4 and IL-5 (secreted by Th2 cells) would be involved in IgGl, IgE, and IgA antibody production, while IFN-y (secreted by T h l cells) would be involved in IgGza antibody production. The growth and terminal differentiation of plasma cells may also require IL-6. For the Thl and Th2 cells to interact in regulating the B cells, lymphokines from one type of Th cell could up regulate receptors for lymphokines secreted by the other type of T h cell. Since evidence is mounting that Thl and Th2 cells might represent different stages of the same lineage, it is possible that low levels of antigen (on APCs) activate T h l cells and that Th2 cells require repeated antigenic stimulation to be activated. Alternatively, Th2 cells might reside near the follicles while Thl cells would cluster around macrophages in paracortical areas. Furthermore, antigen could induce Thl cells to become Th2 cells. This would imply that the type of T cell activated would depend on the dose and/or rate of clearance (versus persistence) of antigen as well as the location of the different types of T cells in the different areas of the node. For example, in the case of minor viral infections (where antigen levels in the circulation would initially be low) Thl cells, Ts cells, and Tc cells might be preferentially activated, whereas in the case of parasites (where antigen persists), Th2 cells may be activated (or T h l cells would be induced to differentiate into Th2 cells).
RESPONSES AGAINST BACTERIAL ANTIGENS B. THEIMMUNE Bacteria are highly diverse and can enter the host by many different routes. Hence, the type of immune response to bacteria may be dependent on the portal of entry into the body. For example, in the gut and mouth, bacteria must penetrate mucosal or epithelial surfaces to enter the bloodstream. Any type of immune response which prevents
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
83
attachment of bacteria to mucosa or epithelial cells will effectively prevent systemic infection. On the other hand, it is desirable not to damage the protective lining of the gut, mouth, etc. IgA that does not fix complement can fulfill these criteria by preventing attachment of bacteria in the gut without causing tissue damage. Hence, the generation of IgAsecreting plasma cells would be highly desirable in Peyer’s patches. Bacteria that do gain entry into the bloodstream are cleared by macrophages or Kupffer cells or are lysed by circulating opsonizing antibodies (e.g., IgM, IgG) and complement. As a consequence of ingestion of antigen, macrophages can secrete a wide variety of mediators which influence both immunocytes and other cells. Hence, in the case of bacterial infections, three types of immunity are probably operative: (1) Phagocytosis of bacterial microorganisms by macrophages or neutrophils occurring both prior to the generation of an antibody response or, more efficiently, after the coating of these organisms with antibody and/or complement components. The opsonization and digestion of bacteria by macrophages results in destruction of the organism and also allows the macrophages to present fragments of bacterial antigens to T cells. (2) B cells might also be rapidly activated in a polyclonal manner by endotoxin or polysaccharides from the bacterial cell wall. (3) Th cells would be activated by B cell clones presenting TD epitopes from bacterial fragments or bacterial products, e.g., toxins. In the mouse, because antibody-mediated phagocytosis of bacteria, bacteriolysis, and the prevention of attachment of bacteria to mucosal and epithelial surfaces can be carried out by particular antibody isotypes, the secretion of IgGl and IgA antibodies may be dependent upon IL-4 and IL-5 secreted by Th2 cells. Thus, activation of Th2 cells is probably of major importance in the generation of antibodies which protect the host against systemic bacterial infections. In contrast, IgG3 responses (e.g., to bacterial polysaccharides) may be relatively thymus independent.
C. THEIMMUNE RESPONSE TO VIRALANTIGENS Viruses, unlike most bacteria, are obligate intracellular pathogens; their extracellular concentrations may be low in the early stages of infection. Furthermore, during cellular interactions in general, virus might be transferred from one cell to another without reaching high concentrations in the blood. Viruses can be neutralized or eliminated by a wide variety of immune mechanisms, including antibody-mediated neutralization (IgM, IgG, or IgA), complement-mediated neutralization, opsonization by macrophages, and cell-mediated cytotoxicity by Tc cells or NK cells. However, until viral antigens are synthesized and expressed on infected target cells or until virions are present in sufficient quantity
84
ELLEN S. VITETTA ET A L .
in the circulation to be ingested by macrophages, degraded, and presented to T cells, humoral immunity, as well as the generation of Tc cells, would not occur. Assuming that sufficient antigenic exposure did occur, class I-restricted Tc cells or NK cells activated by IL-2-secreting Thl cells may play a major role in killing cells either expressing viral antigens and/or actively producing virions. Since Thl cells also secrete IFN-7, this should facilitate IgGza secretion by B cells and, in the mouse, IgGza appears to be protective in the case of some viral infections. Furthermore, IFN-.)I(also secreted by Thl cells) is an effective antiviral agent and is involved in the activation of macrophages. In contrast, in other viral infections where IgA or IgGl play major roles, Th2 cells would be activated by antigen and APCs.
D. THEIMMUNE RESPONSE TO PARASITES Parasites differ enormously in their organization, complexity, life cycles, and pathogenic mechanisms. From many studies in experimental animals, it has been established that different effector mechanisms are active against different parasites and that it is often difficult to predict which mechanism is the predominant one in a particular parasitic infection. Studies on immunity against parasites have suggested that in many cases the humoral antibody response plays a protective role. In the mouse, the two Ig isotypes most often involved in mediating antiparasitic responses are IgGl and IgE. IgE also plays a major role in antiparasite immunity in humans. Among the effector cells involved in immunity to parasites, eosinophils play a fundamental role in mediating parasite killing through interaction with parasites coated with specific antibody of the IgE and IgGl isotypes. This interaction results in the release of eosinophil granules, which contain substances toxic to parasites. Activated macrophages and neutrophils are also capable of killing parasites and, in some instances, mast cells are also involved. Based on these effector mechanisms, which involve IgGl and IgE antibodies, one would predict that the Th2 cells (which secrete IL-4 and IL-5) play a central role in the regulation of humoral immunity against many parasites. In particular, the persistence of parasites might be sufficient to induce the secretion of the large amounts of IL-4 necessary for the induction of IgE secretion (IgE secretion, in contrast to IgGl secretion, requires very high levels of IL-4 in mitogen-driven systems). IL-5could act by recruiting eosinophils, which would be further activated by factors derived from IL-4-activated mast cells. Mast cells would mediate parasite killing through their interaction with the IgE- and IgG1-coated organisms. IL-5 might also induce IgA secretion in local environments where parasites reside (e.g., gut).
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
85
In addition to protective antibodies, the host can defend itself against some intracellular parasites by the development of DTH responses and the activation of macrophages, both of which correlate with protective immunity and healing of parasite-induced lesions. Thus, Thl cells may be involved in these responses, since Thl cells mediate DTH and secrete IFN-y, which is a major macrophage-activating factor.
E. CONCLUDING REMARKS Based on the considerations described above, information from three areas of research will be important to increase our understanding of which cellular interactions are involved in the protective immune response. These areas include (1) the effect of dose, route, and type of antigen on the activation and regulation of subsets of Thl and Th2 cells: (2) the mechanism by which anatomical barriers influence cellular interactions in vivo; and (3) the cellular and molecular mechanisms by which B cells interact with and are regulated by Thl and Th2 cells and their lymphokines to generate specific antibodies of the appropriate isotypes. The disciplines of cellular and molecular immunology represent powerful tools for further elucidating mechanisms involved in cellular interactions. However, the major insight necessary to further our understanding of the immune response in both physiological and pathological states in the intact animal will involve more information about the pathogens themselves, the nature of protective immunity to different types of pathogens, and the activation and regulation of the immune network to elicit these protective responses. ACKNOWLEDGMENTS We are indebted to Dr. Jonathan Uhr, Dr. Michael Berton, Dr. Mark Till, Ms. Alexis Bossie, and Ms. Tracy Stevens for helpful comments and insights concerning this manuscript. We thank all the fellows and students, past and present, who have worked with us and influenced our thinking. We thank Ms. G.A. Cheek and Ms. N. Stephens for their patient and heroic secretarial help. We thank the editors of this volume for providing us with the painful and unusual opportunity of updating (and backdating) our data base and then putting our thoughts together. We apologize to any individuals whose work we inadvertently omitted from this review and its list of citations (address all complaints to E. Vitetta). The experiments which went on while we were writing this review are supported by NIH Grants AI-11851, AI-12789, and AI-07824 and were carried out by Ms.D. Bryant, Ms. E. Carlton, Mr. Y. Chinn, Mr. J. Harris, Ms.S. Joyner, Ms. L. Le, Ms. M.-M. Liu, Ms. M. McCarthy, Ms. S. Shanahan, Ms. R. Summers, Ms. L. Trahan, and Ms. N. Yen. Although we have not yet had time to analyze the results of these experiments, we hope that they do not disprove too much of what we have presented herein. Finally, for those of you who have read this entire review, we hope we have stimulated your thinking as much as its writing has stimulated ours.
86
ELLEN S . VITETTA ET AL.
REFERENCES 1. Claman, H . N.,Chaperon, E. A., and Triplett, R. F. (1966). Proc. SOC. Exp. Biol. Med. 122, 1167. 2. Davies, A. J. S.,Leuchars, E., Wallis, V., Marchant, R., and Elliott, E. V. (1967). Trunspluntution 5, 222. 3. Miller, J. F. A. P., and Mitchell, G. F. (1967). Nature (London) 216, 659. 4. Miller, J. F. A. P., and Mitchell, G. F. (1968).J. Exp. Med. 128, 801. 5. Mitchell, G. F., and Miller, J. F. A. P. (1968).J. Exp. Med. 128, 821. 6. Mitchell, G.E, and Miller, J. E A. P. (1968). Roc. Nutl. Acud. Sci. U.S.A. 59, 296. 7. Nossal, G.J. V., Cunningham, A., Mitchell, G. F., and Miller, J. F. A. P.(1968). J. Exp. Med. 128, 839. 8. Dutton, R. W. (1967). Adu. Immunol. 6, 253. 9. Mishell, R. I., and Dutton, R. W. (1967).J. Exp. Med. 126,423. 10. Marbrook, J. (1967). Lancet 2, 1279. 11. Marbrook, J., and Haskill, J. S. (1974). Cell. Immunol. 13, 12. 12. Mosier, D. E., and Coppelson, L.W. (1968). Roc. Nutl. Acud. Sci. U.S.A. 61, 542. 13. Mosier, D. E. (1969).J. Ex@. Med. 129, 351. 14. Mosier, D. E., Fitch, F. W., Rowley, D. A., and Davies, A. J. (1970). Nature (London) 225, 276. 15. Hodes, R. J., and Singer, A. (1977). Euz J. Immunol. 7 , 892. 16. Mitchison, N. A. (1971). Euz J. Immunol. 1, 18. 17. Raff, M. C. (1970). Nature (London) 226, 1257. 18. Raff, M. C., Nase, S., and Mitchison, N. A. (1971). Nature (London) 230, 50. 19. Rajewsky, K., Schirrmacher, V., Nase, S.,and Jerne, N. K. (1969).J. Exp. Med. 129, 1131. 20. Click, R. E.,Benck, L., and Alter, B. J. (1972). Cell. Immunol. 3, 264. 21. Click, R. E.,Benck, L., and Alter, B. J. (1972). Cell. Immunol. 3, 156. 22. Maizels, R. M., and Dresser, D. W. (1977). Immunology 32, 793. 23. Shiigi, S. M., and Mishell, R. 1. (1975).J. Immunol. 115, 741. 24. Shiigi, S. M., Fujikawa, L. S.,and Mishell, R. I. (1975).J. Immunol. 115, 745. 25. Makela, 0.(1970). 7kunsplunt. Rev. 5, 3 . 26. Makela, O.,and Cross, A. M. (1970). Rag. Allergy 14, 145. 27. Aird, J. (1971). Immunol. 20, 617. 28. Kontiainen, S. (1971). In&.Arch. Allergy Appl. Immunol. 41, 729. 29. Rubin, B. (1972). Euz J. Immunol. 2, 5. 30. Rubin, B. (1973). Euz J. Immunol. 3 , 26. 31. Klaus, G. G. B., and Cross, A. M. (1974). Cell. Immunol. 14, 226. 32. Makela, O.,and Mitchison, N. A. (1965). Immunology 8, 549. 33. Makela, O.,and Mitchison, N. A. (1965). Immunology 8, 539. 34. Taylor, R. B., and Wortis, H. H. (1968). Nature (London) 220, 927. 35. Miller, J. F. A. P., Basten, A,, Sprent, J., and Cheers, C. (1971). Cell. Immunol. 2, 469. 36. Katz, D. H., and Benacerraf, B. (1972). Adu. Immunol. 15, 1. 37. Cheers, C., and Miller, J. F. A. P. (1972).J.f i p . Med. 136, 1661. 38. Torrigiani, G. (1972).J.Immunol. 108, 161. 39. Mitchell, G. F. (1974). Contemp. Top. Immunobiol. 3, 97. 40. Braley-Mullen, H. (1974).J. Immunol. 113, 1909. 41. Romano, T.J., and Thorbecke, G. J. (1975).J. Immunol. 115, 332. 42. Kong, A. S., and Morse, S. I. (1976).J. Immunol. 116, 989.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
87
43. Katz, D. H . , Paul, W. E., Goidl, E. A., and Benacerraf, B. (1970).J. Ex$. Med. 132, 261. 44. Baker, l?J., Prescott, B., Stashak, l? W., and Amsbaugh, D. F. (1971).J. Immunol. 107, 719. 45. Gershon, R. K., and Paul, W. E. (1971).J. Immunol. 106, 872. 46. Taniguchi, M., and Tada, T. (1974). J. Ex$. Med. 139, 108. 47. Feldbush, T. L . , Stunz, L. L., and LaFrenz, D. E. (1986). In “B Lymphocyte Differentiation” (Cambier, J., ed.), p. 135. CRC Press, Boca Raton, Florida. 48. Gershon, R. K., and Kondo, K. (1970). Immunology 18, 723. 49. Genhon, R. K., and Kondo, K. (1971). Immunology 21, 903. 50. Dutton, R. W. (1972). J. Exp. Med. 136, 1445. 51. Dutton, R. W. (1973).J. Exp. Med. 138, 1496. 52. Kapp, J. A., Pierce, C. W., Schlossman, S., and Benacerraf, B. (1974).J Exp. Med. 140, 648. 53. Burns, F. D., Marrack, P. C., Kappler, J. W., and Janeway, C. A., Jr. (1975). J. Immunol. 114, 1345. 54. Bottomly, K., Kaye, J., Jones, B., Jones, F., 111, and Janeway, C. A,, Jr. (1983). J. Mol. Cell. Immunol. 1, 42. 55. Nakauchi, H., Ohno, I., Kim, M., Okumura, K., and Tada, T. (1984). J. Immunol. 132, 88. 56. Uhr, J. W., and Baumann, J. B. (1961). J. Ex$. Med. 113, 935. 57. Uhr, J. W., and Moller, G. (1968). Adv. Immunol. 8, 81. 58. Rowley, D. A., Fitch, F. W., Axelrad, M. A., and Pierce, C. W. (1969). Immunology 16, 549. 59. Pierce, C. W. (1969).J. Exp. Med. 130, 365. 60. Feldmann, M., and Diener, E. (1970). J. Ex$. Med. 131, 247. 61. Chan, P. L . , and Sinclair, N. R. (1971). Adv. Exp. Med. Biol. 12, 609. 62. Sinclair, N. R., and Chan, P. L. (1971). Immunology 21, 967. 63. Sjogren, H. O . , Hellstrom, I., Banal, S. C., and Hellstrom, K. E. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 1372. 64. Diener, E., and Feldmann, M. (1972). Transplant. Rev. 8, 76. 65. Feldmann, M., and Diener, E. (1972). J. Immunol. 108, 93. 66. Feldmann, M., and Nossal, G. J. V. (1973). Transplant. Rev. 13, 3 . 67. Baldwin, R. W., Price, M. R., and Robius, R. A. (1973). Int.J Cancer 11, 527. 68. Wright, P. W., Hargreaves, R. E., Banal, S. C., Bernstein, I. D., and Hellstrom, K. E. (1973). Proc. Natl. Acad. Sci. U.S.A. 70, 2539. 69. Eichmann, K. (1975). Eur. J. Immunology 5, 511. 70. Owen, F. L . , Ju, S. T., and Nisonoff, A. (1977).J. Ex#. Med. 145, 1559. 71. Owen, F. L . , Ju, S. T., and Nisonoff, A. (1977). Proc. Natl. Acad. Sci. U.S.A. 74, 2084. 72. Dorf, M. E., and Benacerraf, B. (1984). Annu. Rev. Immunol. 2, 127. 73. Schimpl, A., and Wecker, E. (1975). Z?ansplant. Rev. 23, 176. 74. Schimpl, A., and Wecker, E. (1972). Nature (London) 237, 15. 75. Mosmann, T. R . , Cherwinski, H., Bond, M. W., Giedlin, M. A., andcoffman, R. L. (1986).J. Immunol. 136, 2348. 76. Kindred, B., and Shreffler, D. C. (1972). J. Immunol. 109, 940. 77. Katz, D. H., Graves, M., Dorf, M. E., DiMuzio, H., andBenacerraf, B. (1975). J. Ex$. Med. 141, 263. 78. Sprent, J. (1978). J. Exp. Med. 147, 1142. 79. Marrack, P., and Kappler, J. W. (1980). J. Ex$. Med. 152, 1274.
88
ELLEN S. VITETTA ET AL
80. Kappler, J. W., and Marrack, P. (1976). Nature (London) 262, 797. 81. Kappler, J. W., Skidmore, B., White, J., and Marrack, P. (1981).J. Exp. Med. 153, 1198. 82. Hedrick, S. M., Matis, L. A., Hecht, T. T., Samelson, L. E., Longo, D. L., HeberKatz, E., and Schwartz, R. H. (1982). Cell (Cambridge, Mass.) 30, 141. 83. H a s h , K., Kubo, R., White, J., Pigeon, M., Kappler, J., and Marrack, I? (1983). J. Exp. Med. 157, 1149. 84. Shimonkevitz, R.,Kappler, J., Marrack, P., and Grey, H. (1983).J.Exp. Med. 158, 303. 85. Matzinger, P. (1981). Nature (London) 292, 497. 86. Dembic, Z.,Haas, W., Weiss, S . , McCubrey, J., Kiefer, H., von Boehmer, H., and Steinmetz, M. (1986). Nature (London) 320, 232. 87. Naor, D., and Sulitzeanu, D. (1967). Nature (London) 214, 687. 88. Byrt, P., and Ada, G. L. (1969). Immunology 16, 503. 89. Dwyer, J. M., and MacKay, I. R. (1970). Lancet 1, 1199. 90. Warner, N. L., Byrt, P., and Ada, G. L. (1970). Nature (London) 226, 942. 91. Basten, A., Miller, J. E A. P., Warner, N. L., and Pye, J. (1971). Nature (London) 231, 104. 92. Davie. J. M., and Paul, W. E. (1971).J. Exp. Med. 134, 495. 93. Davie, J. M.,Rosenthal, A. S., and Paul, W. E. (1971).J. Ex#. Med. 134, 517. 94. Humphrey, J. H., and Keller, H. U. (1971). In “Developmental Aspects of Antibody Formation and Structure” 0. Sterzl and I. Riha, eds.), p. 485. Academic Press, New York. 95. Unanue, E. R. (1971).J. Immunol. 107, 1168. 96. Raff, M. C., Feldmann, M., and De Petris, S. (1973).J. Exp. Med. 137, 1024. 97. Warner, N. L. (1974). Adv. Immunol. 19, 67. 98. Vitetta, E. S., Baur, S., and Uhr, J. W. (1971).J. Ex$. Med. 134, 242. 99. Melcher, U., Vitetta, E. S.,McWilliams, M., Lamm, M. E., Phillips Quagliata, J. M., and Uhr, J. W. (1974).J. Exp. Med. 140, 1427. 100. Ada, G. L., and Byrt, P. (1969). Nature (London) 222, 1291. 101. Golan, D. T., and Borel, Y. (1972).J Exp. Med. 136, 305. 102. Salsano, P., Froland, S . S., Natvig, J. B., and Michaelson, T. E. (1974). Scand. J. Immunol. 3, 841. 103. Goding, J. W., and Layton. J. E. (1976).J. Exp. Med. 144, 852. 104. Yuan, D., Uhr, J. W., and Vitetta, E. S. (1980). J. Immunol. 125, 40. 105. Melcher, U., and Uhr, J. W. (1976).J. Immunol. 116, 409. 106. Vassalli, P., Tartakoff, A., Pink, J. K. L., and Jaton, J.-C. (1980).J. Biol. Chem. 255, 11822. 107. Alt, E W., Bothwell, A. L. M., Knapp, M., Siden, E., Mather, E., Koshland, M., and Baltimore, D. (1980). Cell (Cambridge, Mass.) 20, 293. 108. Rogers.J., Early, P., Carter, C., Calame, K., Bond, M., Hood, L., and Wall, R. (1980). Cell (Cambridge, Mass.) 20, 303. 109. Early, P., Rogers, J., Davis, M., Calame, K., Bond, M., Wall, R., andHood, L. (1980). Cell (Cambridge, Mass.) 20, 313. 110. Teale, J. M., Lafrenz, D., Klinman, N. R., and Strober, S . (198l).J.Immunol. 126, 1952. 111. Durkin, H.G., Bazin, H., and Waksman, B. H. (1981).J.Exp. Med. 154, 640. 112. Yuan, D., and Vitetta, E. S . (1983). Mol. Immunol. 20, 367. 113. Marchalonis, J. J., and Cone, R. E. (1973). 7+ansplant. Rev. 14, 3.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
89
114. Vitetta, E. S., Bianco, C., Nussenzweig, V.,and Uhr, J. W. (1972).J Ex$. Med. 136. 81. 115. Hedrick, S . M., Cohen, D. I., Nielsen, E. A., and Davis, M. M. (1984). Nature (London) 308, 149. 116. Yanagi, Y . , Yoshikai, Y., Leggett, K., Clark, S. P., Aleksander, I., and Mak, T. W. (1984). Nature (London) 308, 145. 117. McIntyre, B. W., and Allison, J. P. (1983). Cell (Cambridge, Mass.) 34, 739. 118. Kappler, J., Kubo, R., Haskins, K., Hannum, C., Marrack, P., Pigeon, M., McIntyre, B., Allison, J., and Trowbridge, I. (1983). Cell (Cambridge, Mass.) 35, 295. 119. Meuer, S. C.,Acuto, O., Hercend, T., Schlossman, S. F., and Reinherz, E. L. (1984). Annu. Rev. Immunol. 2, 23. 120. Hedrick, S . M., Nielsen, E. A., Kavaler, J., Cohen, D. I., and Davis, M. M. (1984). Nature (London) 308, 153. 121. Fabbi, M., Acuto, O., Smart, J. E., and Reinherz, E. L. (1984). Nature (London) 312, 269. 122. Hannum, C. H., Kappler, J. W., Trowbridge, I. S., Marrack, P., and Freed, J. H. (1984). Nature (London) 312, 65. 123. Kronenberg, M., Siu, G., Hood, L. E., and Shastri, N. (1986). Annu. Rev. Immunol. 4, 529. 124. Marrack, P., and Kappler, J. (1987). Science 238, 1073. 125. Allison, J. P., and Lanier, L. L. (1987). Annu. Rev. Zmmunol. 5 , 503. 126. Weiss, A.,Imboden, J., Hardy, K., Manger, B., Terhorst, C., andStobo, J. (1986). Annu. Rev. Zmmunol. 4, 593. 127. Brenner, M. B., McLean, J., Dialynas, D. P., Strominger, J. L., Smith, J. A., Owen, F. L., Seidman, J. G., Ip, S., Rosen, E,and Krangel, M. S. (1986). Nature (London) 322, 145. 128. Bank, I., De Pinho, R. A., Brenner, M. B., Cassimeris, J., Alt, F. W., and Chess, L. (1986). Nature (London) 322, 179. 129. Pardoll, D. M., Fowlkes, B. J,, Bluestone, J. A., Kruisbeek, A., Maloy, W. L., Coligan, J. E., and Schwartz, R. H. (1987). Nature (London) 326, 79. 130. Bluestone, J. A., Pardoll, D., Sharrow, S. 0.. and Fowlkes, B. J. (1987).Nature (London) 326, 82. Colon, S. M., and Grey, H. M. (1982).J.Immunol. 129, 2382. 131. Chesnut, R.W., 132. Shimonkevitz, R.,Colon, S., Kappler, J. W., Marrack, P., and Grey, H. M. (1984). J. Immunol. 133, 2067. 133. Streicher, H. Z., Berkower, I. J., Busch, M., Gurd, F. R., and Berzofsky, J. A. (1984). R o c . Natl. Acad. Sci. U.S.A. 81, 6831. 134. Buus, S., Sette, A., Colon, S . M., Jenis, D. M., and Grey, H. M. (1986). Cell (Cambridge, Mass.) 47, 1071. 135. Babbitt, B. P., Matsueda, G., Haber, E., Unanue, E. R., and Allen, P. M. (1986). R o c . Natl. Acad. Sci. U.S.A. 83, 4509. 136. Sette, A., Buus, S., Colon, S., Smith, J. A., Miles, C., and Grey, H. M. (1987). Nature (London) 328, 395. 137. Matzinger, P., and Bevan, M. J. (1977). Cell. Immunol. 29, 1. 138. Unanue, E. R. (1984). Annu. Rev. Immunol. 2, 395. 139. Hood, L. E., Weissman, I. L., and Wood, W. B. (1978). “Immunology.” BenjaminKummings, Menlo Park, California. 140. Ziegler, K., and Unanue, E. R. (1981).J. Zmmunol. 127, 1869. 141. Inaba, K., Steinman, R. M., Van Voorhis, W. C., and Muramatsu, S . (1983).
90
ELLEN S. VITETTA ET A L
Proc. Natl. Acad. Sci. U.S.A. 80, 6041. 142. Inaba, K., and Steinman, R. M. (1984).J. Ex$. Med. 160, 1717. 143. Inaba, K., Witmer, M. D., and Steinman, R. M.(1984).J.Exp. Med. 160,858. 144. Inaba, K., Granelli Piperno, A,, and Steinman, R. M. (1983).J. Exp. Med. 158, 2040. 145. Rubinstein, D., Roska, A. K., and Lipsky, P. E. (1987).J. Immunol. 138, 1377. 146. Hirschberg, H.,Braathen, L. R.,andThonby, E.(1982). Immunol. Rev. 66,57. 147. Inaba, K., Schuler, G., Witmer, M. D., Valinsky, J., Atassi, B., and Steinman, R. M. (1986).J. Exp. Med. 164, 605. 148. Shimada, S., Caughman, S. W., Sharrow, S. O., Stephany, D., and Katz, S. I. (1987).J. Immunol. 139, 2551. 149. Wekerle, H., Schwab, M., Linington, C., and Meyermann, R. (1986). Euz J. Immunol. 16, 1551. 150. Takiguchi, M., and Frelinger, J. A. (1986).J. Mol. Cell. Immunol. 2, 269. 151. Ransom, J., Fischer, M., Mercer, L., and Zlotnik, A. (1987).J. Immunol. 139, 2620. 152. Umetsu, D. T., Katzan, D., Jabara, H. H . , and Geha, R.S. (1986).J.Immunol. 136, 440. 153. Chesnut, R. W., and Grey, H. M. (1986). Adu. Immunol. 39, 51. 154. Lanzavecchia, A., Roosnek, E., Gregory, T., Berman, €!, andAbrignani, S. (1988). Nature (London) 334, 530. 155. Schnizlein, C. T., Szakal, A. K., and Tm,J. G. (1984).Immunobiology 168,391. 156. Schnizlein, C.T., Kosco, M. H., Szakal, A. K., and Tew, J. G. (1985).J. Immunol. 134, 1360. 157. Benacerraf, B. (1978).J. Immunol. 120, 1809. 158. Chesnut, R. W., and Grey, H.M. (1981).J. Immunol. 126, 1075. 159. Bergholtz, B. O.,and Thorsby, E. (1979). Scand.J. Immunol. 10, 267. 160. Hiramine, C., and Hojo, K. (1980).Int. Arch. Allergy Appl. Immunol. 61,329. 161. Kammer, G. M., and Unanue, E. R. (1980). Clin. Immunol. Immunopathol. 15, 434. 162. Chesnut. R. W., Colon, S. M., and Grey, H. M. (1982).J.Immunol. 128, 1764. 163. Frohman, M., and Cowing, C. (1985).J. Immunol. 134, 2269. 164. Cowing, C., and Chapdelaine, J. M. (1983). Proc. Natl. Acad. Sci. US.A. 80, 6000. 165. Grey, H.M., Colon, S. M., and Chesnut, R.W. (1982).J. Immunol. 129,2389. 166. Bekoff, M., Kubo, R.,and Grey, H. M. (1986).J. Immunol. 137, 1411. 167. Krieger, J. I., Crammer, S. F., Grey, H. M., and Chesnut, R. W. (1985). J. Immunol. 135, 2937. 168. Tony, H.P., Phillips, N . E.. and Parker, D. C. (1985).J.
[email protected]. 162, 1695. 169. Tony, H. P., and Parker, D. C. (1985).J. Ex$. Med. 161, 223. 170. Jelachich, M. L., Lakey, E. K., Casten, L., and Pierce, S. K. (1986). Euz J. Immunol. 16,411. 171. Ashwell, J. D., DeFranco, A. L., Paul, W. E., and Schwartz, R. H. (1984).J f i p . Med. 159, 881. 172. Abbas, A. K., Haber, S., and Rock, K. L. (1985).J. Immunol. 135, 1661. 173. Ashwell, J. D., Schwartz, R.H., Mitchell, J. B., and Russo, A. (1986).J Immunol. 136, 3649. 174. Gosselin, E. J., Tony, H. P., and Parker, D. C. (1988).J. Immunol. 140, 1408. 175. Mttkzeau, P., Elguindi, I., and Goldberg, M. E. (1984). EMEO J. 3, 2235. 176. Ellner, J. J., Lipsky, P. E., and Rosenthal, A. S. (1977).J. Immunol. 118,2053.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
91
177. Ellner, J. J., and Rownthal, A. S. (1975). J. Immunol. 114, 1563. 178. Unanue, E. R. (1978). Immunol. Rev. 40. 227. 179. Roosnek, E., Demou, S., Corradin, G., and Lanzavecchia, A. (1988).J.Immunol. 140, 4079. 180. Ohkuma, S., and Poole, B, (1978). Roc. Natl. Acad. Sci. U.S.A. 75, 3327. 181. Tietze, C., Schlesinger, P., andStahl, P. (1980). Biochem. Biophys. Res. Commun. 93, 1. 182. Gordon, A. H., Darcy Hart, P., and Young, M. R. (1980). Nature (London) 286, 79. 183. Ziegler, H. K., and Unanue, E. R. (1982). Proc. Natl. Acad. Sci. U.S.A. 79,175. 184. Lanzavecchia, A. (1985). Nature (London) 314, 531. 185. Nowell, J., and Quaranta, V. (1985). J. Ex#. Med. 162, 1371. 186. Rosamond, S.,Brown, L., Gomez, C., Braciale, T.J., andschwartz, B. D. (1987). J. Immunol. 139, 1946. 187. Rock, K. L., Benacerraf, B., and Abbas, A. K. (1984).J. Exp. Med. 160, 1102. 188. Rock, K. L., Haber, S. I., Liano, D., Benacerraf, B., and Abbas, A. K. (1986). Euz J. Immunol. 16, 1407. 189. Lichtman, A. H., Tony, W. P., Parker, D. C., and Abbas, A. K. (1987). J. Immunol. 138, 2822. 190. Liano, D., and Abbas, A. K. (1987).J. Immunol. 139, 2562. 191. Lanzavecchia, A. (1986). Cum Top. Microbiol. Immunol. 130, 65. 192. Lanzavecchia, A. (1987). Immunol. Rev. 99, 39. 193. Myers, C. D., and Vitetta, E. S. (1988). In “Antigen Presenting Cells: Diversity, Differentiation and Regulation” (L. B. Schook and J. G. Tew, eds.), p. 127. Alan R. Liss, New York. 194. Snow, E. C.,Vitetta, E. S., and Uhr, J. W. (1983). J. Immunol. 130, 607. 195. Myers, C. D., Sanders, V. M., and Vitetta, E. S. (1986).J Immunol. Methods 92, 45. 196. Myers, C. D., Sanders, V. M., Yefenof, E., Oliver, K. G., Uhr, J. W., and Vitetta, E. S. (1987). In “Methods in Enzymology” (G. Di Sabato, ed.), Vol. 150,p. 316. Academic Press, New York. 197. Sanders, V. M., Snyder, J. M., Uhr, J. W., andvitetta, E. S. (1986).J. Immunol. 137, 2395. 198. Allen, P. M. (1987). Immunol. Today 8, 270. 199. Buus, S., and Werdelin, 0. (1986). Acta Pathol. Microbiol. Immunol. Scand. 94, 17. 200. Buus, S., and Werdelin, 0. (1986).J. Zmmunol. 136, 459. 201. Allen, P. M., and Unanue, E. R. (1984).J. Immunol. 132, 1077. 202. Guillet, J.-G., Lai, M.-Z., Briner, T. J . , Smith, J. A., and Gefter, M. L. (1986). Nature (London) 324, 260. 203. Watts, T.H., and McConnell, H. M. (1987). Annu. Rev. Immunol. 5, 461. 204. Berkower, I., Buckenmeyer, G. K.,and Berzofsky, J. A. (1986).J.Immunol. 136, 2498. 205. Berkower, I., Matis, L. A., Buckenmeyer, G. K., Gurd, F. R., Longo, D. L., and Berzofsky, J. A. (1984).J. Immunol. 132, 1370. 206. Buus, S., Colon, S., Smith, C., Freed, J. H., Miles, C., and Grey, H. M. (1986). Roc. Natl. Acad. Sci. U.S.A. 83, 3968. 207. Coeshott, C.M., Chesnut, R. W., Kubo, R. T., Grammer, S. F.,Jenis, D. M., and Grey, H . M. (1986). J. Immunol. 136, 2832.
92
ELLEN S. VITETTA ET AL.
208. Lakey, E.K.,Casten, L. A., Niebling, W. L., Margoliash, E., and Pierce, S. K. (1988).J. Immunol. 140,3309. 209. Falo, L. D., Jr., Sullivan, K., Benacerraf, B., Mescher, M. F., and Rock, K. L. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 6647. 210. Allen, P. M., Strydom, D. J., and Unanue, E. R. (1984). Proc. Natl. Acad. Sci. U.S.A. 81, 2489. 211. Naquet, P., Ellis, J., Singh, B., Hodges, R. S., and Delovitch, T. L. (1987). J. Immunol. 139, 3955. 212. Schwartz, R. H., Fox, B. S., Fraga, E., Chen, C., and Singh, B. (1985). J. Immunol. 135, 2598. 213. Hannum, C. H., Matis, L. A., Schwartz, R. H., and Margoliash, E. (1985). J. Immunol. 135, 3314. 214. Suzuki, G., and Schwartz, R. H. (1986). J. Immunol. 136, 230. 215. Quill, H., Carlson, L., Fox, B. S., Weinstein, J. N., andSchwartz, R. H. (1987). J. Immunol. Methods 98, 29. 216. Carbone, E R., Fox, B. S., Schwartz, R. H., and Paterson, Y. (1987).J. Immunol. 138, 1838. 217. Fox,B. S.,Q d , H., Carhn, L., and Schwaxtz, R. H. (1987).J Immuml. 158,3367. 218. Fox, B. S.,Chen, C., Fraga, E., French, C. A., Singh, B., andschwartz, R. H. (1987).J. Immunol. 139, 1578. 219. Kovac, Z.,and Schwartz, R. H. (1985).J. Immunol. 134, 3233. 220. Berzofsky,J. A., Comette, J., Margalit, H., Berkower, I., Cease, K., and DeLisi, C. (1986). Curx Top. Microbiol. Immunol. 130, 13. 221. Cease, K. B., Berkower, I., York Jolley, J., and Berzofsky, J. A. (1986).J. Ex$. Med. 164, 1779. 222. Spouge, J. L., Guy, H. R., Cornette, J. L., Margalit, H., Cease, K., Berzofsky, J. A., and DeLisi, C. (1987).J. Immunol. 138, 204. 223. Margalit, H., Spouge, J. L., Comette, J. L., Cease, K. B., DeLisi, C., and Berzofsky, J. A. (1987).J. Immunol. 138, 2213. 224. DeLisi, C., and Berzofsky, J. A. (1985).Proc. Natl. Acad. Sci. U.S.A. 82, 7048. 225. Falo, L. D., Jr., Haber, S. I., Herrmann, S., Benacerraf, B., and Rock, K. L. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 522. 226. Falo, L. D., Jr., Benacerraf, B., and Rock, K. L. (1986).Proc. Natl. Acad. Sci. U.S.A. 83, 6994. 227. Babbitt, B. P.,Allen, P. M., Matsueda, G., Haber, E., and Unanue, E. R. (1985). Nature (London) 317, 359. 228. Buus, S., Sette, A., Colon, S. M., Miles, C., and Grey, H. M. (1987). Science 235, 1353. 229. Buus, S . , Sette, A., and Grey, H. M. (1987). Immunol. Rev. 98, 115, 230. Cresswell, P. (1985). PToc. Natl. Acad. Sci. U.S.A. 82, 8188. 231. Werdelin, O.,and Buus, S. (1988). In “Antigen Presenting Cells: Diversity, Differentiation, and Regulation” (L. B. Schook and J. G. Tew, eds.), p. 93.Alan R. Liss, New York. 232. Lin, J., Berzofsky,J. A., andDelovitch, T. L. (1988).J.Mol. Cell. Immunol. 3 , 321. 233. Bjorkman, P.J.. Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J . L., and Wiley, D. C. (1987). Nature (London) 329, 506. 234. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., and Wiley, D. C. (1987). Nature (London) 329, 512. 235. Brown, J. H., Jardetzky, T., Saper, M. A., Samraoui, B., Bjorkman, P. J . , and Wiley, D. C. (1988). Nature (London) 332, 845.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
93
236. Swain, S. L. (1983). Immunol. Rev. 74, 129. 237. Mizuochi, T., Golding, H., Rosenberg, A. S.,Glimcher, L. H., Malek, T. R., and Singer, A. (1985).J. Exp. Med. 162, 427. 238. Swain, S. L. (1981). R o c . Natl. Acad. Sci. U.S.A. 78, 7101. 239. Chen, B. P.,DeMars, R., Hank, J. A., andSonde1, P. M. (1988). Cell. Immunol. 115, 363. 240. Morrison, L. A., Lukacher, A. E., Braciale, V. L., Fan, D. P., and Braciale, T. J. (1986).J. Ex$. Med. 163, 903. 241. Morrison, L. A., Braciale, V. L., and Braciale, T. J. (1988).J.Immunol. 141,363. 242. Davis, M. M., and Bjorkman, P.J. (1988). Nature (London) 334, 395. 243. Lassila, O.,Vainio, O., and Matzinger, P. (1988). Nature (London) 334, 253. 244. Janeway, C. A., Jr., Murgita, R. A., Weinbaum, E I., Asofiky, R., and Wigzell, H. (1977). R o c . Natl. Acad. Sci. U.S.A. 74, 4582. 245. Ron, Y.,De Baetselier, P., Gordon, J., Feldman, M., and Segal, S. (1981). EuzJ. Immunol. 11, 964. 246. Ron, Y.,De Baetselier, P.,Tzehoval, E . , Gordon, J., Feldman, M., and Segal, S. (1983). EuzJ. Immunol. 13, 167. 247. Ron, Y.,and Sprent, J. (1987).J. Immunol. 138, 2848. 248. Kurt-Jones, E. A , , Liano, D., HayGlass, K. A., Benacerraf, B., Sy, M.-S., and Abbas, A. K. (1988).J. Immunol. 140,3773. 249. Kosco, M. H., Szakal, A. K., and Tew, J. G. (1988).J. Immunol. 140, 354. 250. Cantor, H., and Boyse, E. A. (1975).J. Exp. Med. 141, 1376. 251. Marrack, P.,and Kappler, J. W. (1975).J.Immunol. 114, 1116. 252. Janeway, C. A., Jr. (1975).J. Immunol. 114, 1394. 253. Tada, T., Takemori, T., Okumura, K., Nonaka, M., and Tokuhisa, T. (1978). J. Exp. Med. 147,446. 254. Waldman, H. (1977). Immunol. Rev. 35, 121. 255. Moorhead, J. W. (1978).J. Immunol. 120, 137. 256. Swierkosz,J. E.,Marrack, P., and Kappler, J. W. (1979).J.Ex$. Med. 150,1293. 257. lmperiale, M. J., Faherty, J. F., Sproviero, J. F., and Zauderer, M. (1982). J. Immunol. 129, 1843. 258. Kimoto, M., and Fathman, C. G. (1980).J. Ex$. Med. 152, 759. 259. Kim, J., Woods, A., Becker-Dunn, E., and Bottomly, K. (1985).J.Ex$. Med. 162, 188. 260. Cherwinski, H. M.,Schumacher, J. H., Brown, K. D., and Mosmann, T. R. (1987).J. Ex#. Med. 166, 1229. 261. Killar, L., MacDonald, G., West, J., Woods, A., and Bottomly, K. (1987). J. Immunol. 138, 1674. 262. Rasmussen, R.,Takatsu, K.,Harada. N., Takahashi, T., andBottomly, K. (1988). J. Immunol. 140, 705. 263. Bottomly, K. (1988). Immunol. Today 9,268. 263a. Noelle, R.J. et al. (1989).J. Mol. Cell. Immunol. (in press). 264. Cher, D. J., and Mosmann, T. R. (1987).J. Immunol. 138, 3688. 265. Stevens, T. L., Bossie, A., Sanders, V. M., Fernandez-Botran, R., Coffman, R. L., Mosmann, T. R., and Vitetta, E. S. (1988). Nature (London) 334, 255. 266. Mond, J. J., Finkelman, F. D., Sarma, C., Ohara, J., and Serrate, S. (1986). J. Immunol. 135, 2513. 267. -&bin, E. M., Mond, J. J., Ohara, J., and Paul, W. E. (1986).J. Immunol. 137, 1573. 268. Coffman, R. L., and Carty, J. (1986). J. Immunol. 136, 949.
94
ELLEN S. VITETTA ET A L .
269. Mond. J. J., Carman, J., Sarma, C., Ohara, J., and Finkelman, F. D. (1986). J. Immunol. 137, 3534. 270. Oliver, K., Krammer, P. H., Tucker, P. W., and Vitetta, E. S . (1987). Cell. Immunol. 106, 428. 271. Reynolds, D. S., Boom, W. H., and Abbas, A. K. (1987).J. Immunol. 139, 767. 272. Arthur, R. P., and Mason, D. (1986).J. Ex$. Med. 163, 774. 273. Spickett, G. P., Brandon, M. R., Mason, D. W., Williams, A. F., and Woollett, G. R. (1983).J. Ex$. Med. 158, 795. 274. Rudd, C. E.,Morimoto, C., Wong, L. L., and Schlossman, S. F. (1987).J. Exp. Med. 166, 1758. 275. Birkeland, M. L., Metlay, J., Sanders, V. M., Fernandez-Botran, R., Vitetta, E. S., Steinman, R. M., and Pure, E. (1988).J. Mol. Cell. Immunol. 4, 71. 276. Clement, L. T., Yamashita, N., and Martin, A. M. (1988).J.Immunol. 141,1464. 277. Fernandez-Botran, R., Sanders, V. M., Oliver, K. G., Chen, Y. W., Krammer, P. H., Uhr, J. W., and Vitetta, E. S . (1986).Roc. Natl. Acad. Sci. U S A . 83, 9689. 278. Cantrell, D. A., and Smith, K. A. (1983).J. Ex#. Med. 158, 1895. 279. Robb, R. J., Munck, A,, and Smith, K. A. (1981).J. Ex#. Med. 154, 1455. 280. Fernandez-Botran, R., Sanders, V. M., Mosmann, T., andvitetta, E. S. (1988). J. Ex#. Med. 168, 543. 281. Alcover, A., Ramarli, D., Richardson, N. E., Chang, H.-C., and Reinherz, E. L. (1987). Immunol. Rev. 95, 5. 282. Vitetta, E. S . , Bossie, A., Fernandez-Botran, R., Myers, C. D., Oliver, K. G., Sanders, V. M., and Stevens, T. L. (1987). Immunol. Rev. 99, 193. 283. Mosmann, T. R., and Coffman, R. L. (1987). Immunol. Today 8,233. 284. Vitetta, E.S.,Brooks, K., Chen, Y. W., Isakson, P., Jones, S . , Layton, J., Mishra, G. C., Pure, E., Weiss, E., Word, C., Yuan, D., Tucker, P., Uhr, J. W., and Krammer, P. H. (1984). Immunol. Rev. 78, 137. 285. Lichtman, A. H., Kurt-Jones, E. A., and Abbas, A. K. (1987).R o c . Natl. Acad. Sci. U S A . 84, 824. 286. Kupper, T., Horowitz, M., h, E, Robb, R., andFlood, P. M. (1987).J.Immunol. 138, 4280. 287. Greenbaum, L. A., Horowitz, J. B., Woods, A., Pasqualini, T., Reich, E. P., and Bottomly, K. (1988).J. Immunol. 140, 1555. 288. Kurt-Jones, E.A., Hamberg, S., Ohara, J., Paul, W. E., and Abbas, A. K. (1987). J. Ex$. Med. 166, 1774. 289. Smith, K. A. (1980). Immunol. Rev. 51, 337. 290. Howard, M., Farrar, J., Hilfiker, M., Johnson, B., Takatsu, K., Hamaoka, T., and Paul, W. E. (1982).J. Exp. Med. 155, 914. 291. Ohara, J., and Paul, W. E. (1985). Nature (London) 315, 333. 292. Noma, Y., Sideras, T., Naito, T., Bergstedt-Liudquist, S . , Azuma, C., Severinson, E., Tanaba, T., Kinashi, T., Matsuda, F., Yaoita, Y., and Honjo, T. (1986). Nature (London) 319, 640. 293. Lee, F., Yokota, T., Otsuka, T., Meyerson, P., Villaret, D., Coffman, R., Mosmann, T., Rennick, D., Roehm, N., Smith, C., Zlotnik, A., and Arai, K. I. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 2061. 294. Mosmann, T. R., Bond, M. W., Coffman, R. L., Ohara, J., and Paul, W. E. (1986). Proc. Natl. Acad. Sci. U S A . 83, 5654. 295. Grabstein, K., Eisenman, J., Mochizuki, D., Shanebeck, K., Coulon, P., Hopp, T., March, C., and Gillis, S. (1986). J. Ex#. Med. 163, 1405. 296. Fernandez-Botran, R., Krammer, P. H., Dimantstein, T., Uhr, J. W., and Vitetta,
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
95
E. S. (1986).J.Exp. Med. 164, 580. 297. Severinson, E., Naito, T., Tokumoto, H., Fukushima, D., Hirano, A., Hanna, K.,and Honjo, T. (1987). Eul: J. Immunol. 17, 67. 298. Hu-Li, J., Shevach, E. M., Mizuguchi, J., Ohara, J., Mosmann, T., and Paul, W. E. (1987).J. Exp. Med. 165, 157. 299. Gajewski, T. F., and Fitch, F. W. (1988).J. Immunol. 140, 4245.
300. Coffman, R. L., Seymour, B. W., Lebman, D. A., Hiraki, D. D., Christiansen, J. A , , Shrader, B., Chenvinski, H. M., Savelkoul, H. F., Finkelman, F. D., Bond, M. W., and Mosmann, T. (1988). Zmmunol. Rev. 102, 5. 301. Swain, S. L., McKenzie, D. T., Dutton, R. W., Tonkonogy, S. L., andEnglish, M. (1988). Immunol. Rev. 102,77. 302. Powers, G. D., Abbas, A. K., and Miller, R. A. (1988).J. Immunol. 140,3352. 303. Rajewsky, K.,and Rottlander, E. (1969). Int. Arch. Allergy APpl. Immunol. 36, Suppl., 405. 304. Memll, D., Hartley, T. F., and Claman, H. N. (1967).J. Lab. ClikgMed. 69,151. 305. Minden, P., Grey, H. M., and Farr, R. S. (1967). J. Immunol. 99, 590. 306. Palacios, R., and Steinmetz, M. (1985). Cell (Cambridge, Mass.) 41, 727. 307. Greenwood, E C., Hunter, W. M., and Glover, J. S. (1963). Bi0chem.J. 89,114. 308. Corradin, G., Chesnut, R. W., and Grey, H. M. (1979). Ric. Clin. Lab. 9,311. 309. Gearhart, H. L., Pierce, S. K., and Payne Bose, D. (1977).J. Chromatop. Sci.
15, 480. 310. Mosier, D. E., Mond, J. J., and Goldings, E. A. (1977).J. Immunol. 119, 1874. 311. Schlegel, R. A. (1974). Aust. J. Ex$. Biol. Med. Sci. 52, 471. 312. Burns, W. H., Billups, L. C., and Notkins, A. L. (1975). Nature (London) 256, 654. 313. Rubin, A. S.,MacDonald, A. B., andcoons, A. H. (1974).J.Exp. Med. 139,439. 314. Farrar, J. J. (1975).J. Immunol. 115, 1295. 315. Rubin, A. S., and MacDonald, A. B. (1976). Immunology 30, 69. 316. Shimizu, A., Takahashi, N., Yaoita, Y.,and Honjo, T. (1982). Cell (Cambridge, Mass.) 28, 499. 317. Max, E. E. (1984). In “Fundamental Immunology” (W. E. Paul, ed.), p. 167. Raven Press, New York.
318. Isakson, P. C., Pure, E., Vitetta, E. S., and Krammer, P. H. (1982).J.Exp. Med. 155, 734. 319. Coffman, R. L., Ohara, J., Bond, M. W., Carty, J., Zlotnik, A., and Paul, W. E. (1986).J. Immunol. 136, 4538. 320. Coffman, R. L., Shrader, B., Carty,J., Mosmann, T. R., and Bond, M. W. (1987). J. Immunol. 139, 3685. 321. Snapper, C. M., and Paul, W. E. (1987). Science 236, 944. 322. Snapper, C. M., Finkelman, F.D., and Paul, W. E. (1988).J.Ex$. Med. 167,183. 323. Bossie, A., Brooks, K. H., Krammer, P. H., and Vitetta, E. S. (1987).J. Mol. Cell. Immunol. 3 , 221. 324. Murray, P. D., McKenzie, D. T., Swain, S. L., and Kagnoff, M. F. (1987). J. Immunol. 139, 2669. 325. Boom, W. H., Liano, D., and Abbas, A. K. (1988).J Ex$. Med. 167, 1350. 326. Sanders, V. M., Fernandez-Botran, R., Coffman, R. L., Mosmann, T. R., and Vitetta, E. S. (1988). Proc. Natl. Acad. Sci. U.S.A. 85, 7724. 327. Yuan, D., Weiss, E. A., Layton, J. E., Krammer, P. H.. andvitetta, E. S. (1985). J. Immunol. 135, 1465.
96
ELLEN S. VITETTA ET A L .
328. Layton, J. E., Vitetta, E. S., Uhr, J. W., and Krammer, P. H. (1984).J. Ex#. Med. 160, 1850. 329. Snapper, C. M., and Paul, W. E. (1987).J. Immunol. 139, 10. 330. Radbruch, A., and Sablitzky, F. (1983). EMBO J. 2, 1929. 331. Radbruch, A., Muller, W., and Rajewsky, K. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 3954. 332. Hummel, M., Berry, J. K., and Dunnick, W. (1987).J.Immunol. 138, 3539. 333. Winter, E., Krawinkel, U., and Radbruch, A. (1987). EMBO J. 6, 1663. 334. Rabbits, T. H. (1978). Nature (London) 275, 291. 335. Yaoita, T., Kumagai, Y., Okumura, K., and Honjo, T. (1982). Nature (London) 297, 697. 336. At, E W., Rosenberg, N., Casanova, R. J.. Thomas, E.,and Baltimore, D. (1982). Nature (London) 296, 325. 337. Shimizu, A., and Honjo, T. (1984). Cell (Cambridge, Mass.) 36, 801. 338. Stavnezcr-Nordgren, J., and Sirlin, S. (1986). EMBOJ. 5, 95. 338a. Stavnezer, J . , Radcliffe, G., Lin, Y.-C., Nietupski, J., Berggren, L., Sitia, R., and Severinson, E. (1988). R o c . Natl. Acad. Sci. U.S.A. 85, 7704. 338b. Berton, M., Uhr, J. W., and Vitetta, E. S. (1988). Proc. Natl. Acad. Sci. U.S.A. (in press). 339. Yancopoulos, G. D., De Pinho, R. A., Zimmerman, K. A., Lutzker, S. G., Rosenberg, N., and Alt, F. W. (1986). EMBOJ. 5, 3259. 340. Lutzker, S., and Alt, F. W. (1988). Mol. Cell. Biol. 8 , 1849. 341. Lutzker, S.,Rothman, P., Pollock, R., Coffman, R.. and Alt, F. W. (1988). Cell (Cambridge, Mass.) 53, 177. 342. Isakson, P. C., Pure, E., Vitetta, E. S., and Krammer, P. (1982). In “B and T Cell Tumors” (E. S. Vitetta and F. Fox, eds.), UCLA Symp. Mol. Cell. Biol., p. 391. Academic Press, New York. 342a. Alt, F. W., Blackwell, T. K., De Pinho, R. A., Reth, M. G., and Yancopoulos. G. D. (1986). Immunol. Rev. 89, 5. 343. Singh, H.,Sen, R., Baltimore, D., and Sharp, P. A. (1986). Nature (London) 319, 154. 344. Marrack, P., and Kappler, J. (1986). Adv. Immunol. 38, 1. 345. Steinman, R. M., Kaplan, G., Witmer, M. D., and Cohn, Z. A. (1979).J. Exp. Med. 149, 1. 346. Sunshine, G. H., Katz, D. R., and Feldmann, M. (1980).J.Ex$. Med. 152, 1817. 347. Mosier, D. E. (1967). Science 158, 1573. 348. Rosenthal, A. S., and Shevach, E. M. (1973).J. Exp. Med. 138, 1194. 349. Stingl, G., Katz, S. I., Clement, L., Green, I., and Shevach, E. M. (1978). J. Immunol. 121, 2005. 350. SilberbergSinakin, I., Gigli, I., Baer, R. L., and Thorbecke, G. J. (1980). Immunol. Rev. 53, 203. 351. Glimcher, L. H., Kim. K. J., Green, I., and Paul, W. E. (1982).J. O p . Med. 155, 445. 352. Martz, E. (1975).J. Immunol. 115, 261. 353. Lipscomb, M. F., Ben-Sasson, S. Z., Tucker, T. T., andUhr, J. W. (1979). Eur. J. Immunol. 9, 119. 354. Lipsky, P. E.,and Rosenthal, A. S. (1974).J. Ex$. Med. 141, 138. 355. Nielsen, M. H., Jensen, H., Braendstrup, O., and Werdelin, 0. (1974).J. Exp. Med. 140, 1260. 356. Lipscomb, M. F,, Ben-Sasson,S. Z., andUhr, J. W. (1977).J. Imrnunol. 118, 1748.
CELLULAR INTERACXIONS IN THE HUMORAL IMMUNE RESPONSE
97
357. Zagury, D.,Bernard, J., Thierness, N., Feldman, M., and Berke, G. (1975). Euz .J Immunol. 5, 818. 358. Swierkosz,J. E.,Rock, K., Marrack, P., and Kappler, J. W. (1978).J.Exp. Med. 147, 554. 359. Braun, J., Citri, Y., Baltimore, D., Forouzanpour, F., King, L., Teheranizadeh, K., Bray, M., and Kliewer, S. (1986). Immunol. Rev. 93, 5. 360. Herzenberg, L. A., Stall, A. M., Lalor, P. A., Sidman, C., Moore, W. A., and Parks, D. R. (1986). Immunol. Rev. 93, 81. 361. Sherr, D. H., Braun, J.. and Dorf, M. E. (1987).J. Immunol. 138, 2057. 362. Uher, F., and Dickler, H. B. (1988).J. Immunol. 140, 1442. 363. Marrack, P., Skidmore, B., and Kappler, J. W. (1983).J. Immunol. 130, 2088. 364. Kupfer, A., Swain, S. L., Janeway, C. A., Jr., andsinger, S.J. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 6080. 365. Lipsky, P. E.,and Rmenthal, A. S. (1973). J. Ex$. Med. 138, 900. Brendstrup, O., and Pedersen, E. (1974).J. Ex$. Med. 140, 1245. 366. Werdelin, 0.. 367. Ziegler, H. K., and Unanue, E. R. (1979).J. Exp. Med. 150, 1143. 368. Ben-Sasson, S. Z., Lipscomb, M. F., Tucker, T. F., and Uhr, J. W. (1978). J. Immunol. 120, 1902. 369. Yefenof, E.,Sanders, V. M., Snow, E. C., Noelle, R. J., Oliver, K. G., Uhr, J. W., and Vitetta, E. S. (1985).J. Immunol. 135, 3777. 370. Kupfer, A., Swain, S. L., and Singer, S. J. (1987).J. Ex$. Med. 165, 1565. 371. Kupfer, A., Dennert, G., and Singer, S.J. (1985).J.Mol. Cell. Immunol. 2,37. 372. Buck, C. A., and H o d t z , A. E (1987). Annu. Rev. Cell Biol. 3 , 179. 373. Ben-Sasson, S. Z., Lipscomb, M. F., Tucker, T. F., and Uhr, J. W. (1977). J. Immunol. 119, 1493. 374. Kriz, M. K., Vitetta, E. S.,and Sullivan, T. J. (1986).J. Immunol. 137, 478. 375. Grupp, S. A., and Harmony, J. A. K. (1985).J. Immunol. 134, 4087. 376. Myers, C. D., Kriz, M. K., Sullivan, T. J., and Vitetta, E. S. (1987).J.Immunol. 138, 1705. 377. Grupp, S. A., Snow, E. C., andHarmony, J. A. (1987). Cell. Immunol. 109,181. 378. Monroe, J. G., and Cambier, J. C. (1983). J. Exp. Med. 157, 2073. 379. Monroe, J. G., and Cambier, J. C. (1983).J. Immunol. 131, 2641. 380. Braun, J., Shaafi, R. I., and Unanue, E. R. (1979).J. Cell Biol. 82, 755. 381. Snow, E. C., Fetherston, J. D., and Zimmer, S. (1986).J. Exp. Med. 164, 944. 382. Sell, S., and Cell, P. G. H. (1965).J. Ex$. Med. 122, 423. 383. Noelle, R. J., Snow, E. C., Uhr, J. W., and Vitetta, E. S. (1983). Proc. Natl. Acad. Sci. U.S.A. 80, 6628. 384. Julius, M. H., von Boehmer, H., and Sidman, C. L. (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 1989. 385. LoCascio, N.J., Haughton, G., Arnold, L. W., and Corley, R. B. (1984). Proc. Natl. Acad. Sci. U.S.A. 81, 2466. 386. Sanders, V. M., Uhr, J. W., and Vitetta, E. S. (1987). Cell. Immunol. 104,419. 387. Kupfer, A.,Singer, S.J., Janeway, C. A., Jr., and Swain, S. L. (1987). R o c . Natl. Acad. Sci. U.S.A. 84, 5888. 388. Jones, B., and Janeway, C. A., Jr. (1981). Nature (London) 292, 547. 389. Andersson, J., Schreier, M. H., and Melchers, F, (1980). Proc. Natl. Acad. Scz. U.S.A. 77, 1612. 390. Julius, M. H., Chiller,J. M., andSidman, C. L. (1982). EuzJ. Immunol. 12,627. 391. Ratcliffe, M. J., and Julius, M. H. (1982). Eur. J. Immunol. 12, 634. 392. Julius, M. H., and Rammensee, H. G. (1988). Eur. J. Immunol. 18,375.
98
ELLEN S. VITETTA ET A L .
393. Julius, M. H., Rammensee, H. G., Ratcliffe, M. J., Lamers, M. C., Langhorne, J., and Kohler, G. (1988). Euz J. Immunol. 18, 381. 394. Owens, T. (1988). Euz J. Immunol. 18, 395. 395. Riedel, C., Owens, T., and Nossal, G. J. (1988). Eur. J. Immunol. 18, 403. 396. Sanders, V. M., Fernandez-Botran, R., Uhr, J. W., and Vitetta, E. S. (1987). J. Immunol. 139, 2349. 397. Henry, C., Chan, E. L., and Kodlin, D. (1977).J. Immunol. 119, 744. 398. Bottomly, K., Jones, B., Kaye, J., and Jones, E D. (1983).J.Exp. Med. 158, 265. 399. Bekkhoucha, F.,Naquet, P., Pierres, A., Marchetto, S., and Pierres, M. (1984). EuzJ. Immunol. 14, 807. 400. Krieger, J., Jenis, D. M., Chesnut, R. W., and Grey, H. M. (1988).J. Immunol. 140, 388. 401. Cambier, J. C., and Ransom, J. T. (1987). Annu. Rev. Immunol. 5, 175. 402. Niederhuber, J. E.,Frelinger, J. A., Dugan, E.,Coutinho, A., and Shreffler, D. C. (1975).J. Immunol 115, 1672. 403. Forsgren, S., Pobor, G., Coutinho, A., and Pierres. M. (1984).J. Immunol. 135, 2104. 404. Clement, L. T., Tedder, T. E,andGartland, G. L. (1986).J. Immunol. 136,2375. 405. Palacios, R.,Martinez-Maza, O., and Guy, K. (1983). Roc. Natl. Acad. Sci. U.S.A. 80, 3456. 406. Bonagura, V. R., Agostino, N.. Crow, M. K., and Pernis, B. (1985). Cell. Immunol. 96, 442. 407. Pierce, C. W.,Kapp, J. A., Solliday, S. M., Dorf, M. E., and Benacerraf, B. (1974).J. &p. Med. 140, 921. 408. Frelinger, J. A., Niederhuber, J. E., and Shreffler, D. C. (1975). Science 188, 268. 409. Chen, Z. Z.,McGuire, J. C . , Leach, K. L., and Cambier, J. C. (1987).J. Immunol. 138, 2345. 410. Cambier, J. C., Newell, M. K., Justement, L. B., McGuire, J. C . , Leach, K. L., and Chen, Z. Z. (1987). Nature (London) 327, 629. 411. Chen, Z.Z.,Coggeshall, K. M., and Cambier, J. C. (1986).J. Immunol. 136, 2300. 412. Davignon, D., Martz, E., Reynolds, T., Kurzinger, K., and Springer, T. A. (1981). Roc. Natl. Acad. Sci. U.S.A. 78, 4535. 413. Pierres, M., Goridis, C., and Goldstein, P. (1982). Eur. J. Immunol. 12, 60. 414. Kurzinger, K., and Springer, T. A. (1982).J. Biol. Chem. 257, 12412. 415. Sanchez-Madrid, F., Nagy, J., Robbins, E., Simon, P., andspringer, T. A. (1983). I. Ex#. Med. 158, 1785. 416. -Kurzinger, K., Reynolds, T., Germain, R. N., Davignon, D., Martz, E., and Springer, T. A. (1981).J. Immunol. 127, 596. 417. Rothlein, R., Dustin, M. L., Marlin, S. D., and Springer, T. A. (1986). J. Immunol. 137, 1270. 418. Makgoba, M. W.,Sanders, M. E., Ginther Luce, G. E., Dustin, M. L., Springer, T. A,, Clark, E. A., Mannoni, €?,and Shaw, S. (1988).Nature (London) 331, 86. 419. Davignon, D., Martz, E., Reynolds, T., Kurzinger, K., andspringer, T. A. (1981). J. Immunol. 127, 590. 420. Fischer, A., Durandy, A., Sterkers, G., and Griscelli, C. (1986).J. Immunol. 136, 3198. 421. Tedder, T. F., Schmidt, R. E., Rudd, C. E., Kornacki, M. M., Rizt, J., and Schlossman, S. F. (1986). Eur. J. Immunol. 16, 1539. 422. Howard, D. R.,Eaves, A. C., and Takei, F. (1986).J. Immunol. 136, 4013. 423. Golde, W.T., Gay, D., Kappler, J., and Marrack, P. (1986). Cell. Immunol. 103, 73.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
99
424. Kearse, K. P., Cassatt, D. R., Kaplan, A. M., and Cohen, D. A. (1988). J. Zmmunol. 140, 1770. 425. Mishra, G. C., Berton, M. T., Oliver, K. G., Krammer, P. H., Uhr, J. W., and Vitetta, E. S. (1986).J.Zmmunol. 137, 1590. 426. Dialynas, D. P., Quan, Z. S.,Wall, K. A., Pierres, A., Quinta, J., Loken, M. R., Pierres, M., and Fitch, F. W. (1983).J.Immunol. 131, 2445. 427. Dialynas, D. P., Wilde, D. B., Marrack, P., Pierres, A., Wall, K. A., Havran, W., Otten, G., Loken, M. R., Pierres, M., Kappler, J., and Fitch, F. W. (1983). Immunol. Rev. 74, 29. 428. Wilde, D. B., Marrack, P., Kappler, J., Dialynas, D. P.,and Fitch, E W. (1983). J. Zmmunol. 131, 2178. 429. Swain, S. L.,Dialynas, D. P.,Fitch, F. W., and English, M. (1984).J.Zmmunol. 132, 1118. 430. Marrack, P., Endres, R., Shimonkevitz, R., Zlotnik, A., Dialynas, D., Fitch, F., and Kappler, J. (1983).J. Exp. Med. 158, 1077. 431. Wassmer, P.J., Chan, C., Logdberg, L.,and Shevach, E. M. (1985).J.Zmmunol. 135, 2237. 432. Tite, J. P., Sloan, A., and Janeway, C. A. (1986).J.Mol. Cell. Zmmunol. 2,179. 433. Owens, T.,and Fazekas de St. Groth, B. (1987).J.Immunol. 138, 2402. 434. Poo, W.J., Conrad, L., and Janeway, C. A., Jr. (1988).Nature (London) 332,378. 435. Taplits, M. S., Henkart, P. A., and Hodes, R. J. (1988).J. Zmmunol. 141, 1. 436. Norcross, M. A., Smith, R. T., and Shimizu, S. (1984).J.Zmmunol. 132, 833. 437. Gunter, K. C., Malek, T. R.,and Shevach, E.M. (1984).J.Ex#. Med. 159, 716. 438. MacDonald, H. R.,Bron, C., Rousseaux, M., Horvath, C., and Cerottini, J. C. (1985).Euc J. Zmmunol. 15, 495. 439. Pont, S.,Regnier-Vigouroux, A., Naquet, P.,Blanc, D., Pierres, A., Marchetto, S . , and Pierres, M. (1985).Euc J. Immunol. 15, 1222. 440. Kroczek, R. A., Gunter, K. C., Germain, R. N., and Shevach, E. M. (1986). Nature (London) 322, 181. 441. Corbel, C., and Melchers, F. (1983).EueJ. Immunol. 13, 528. 442. LeClercq, L.,Bismuth, G., and Theze, J. (1984).Proc. Natl. Acud. Sci. U S A . 81, 6491. 443. LeClercq, L.,Cambier, J. C., Mishal, Z., Julius, M. H., and Theze, J. (1986). J. Zmmunol. 136, 539. 444. Melchers, F., Andersson, J., Lernhardt, W., and Schreier, M. H. (1980). Euc J. Zmmunol. 10, 679. 445. Schreier, M. H.,Andersson, J., Lernhardt, W., and Melchers, E (1980). J. Exp. Med. 151, 194. 446. Singer, A., Morrissey, P. J., Hathcock, K. S., Ahmed, A., Scher, I., andHodes, R.J. (1981).J.Exp. Med. 154, 501. 447. Singer, A., and Hodes, R. J. (1983).Annu. Rev. Immunol. 1, 211. 448. Mosier, D. E., and Feeney, A. J. (1984).J.Exp. Med. 160, 329. 449. Asano, Y.,Singer, A., and Hodes, R.J. (1985).J. Zmmunol. 134, 3682. 450. Krusemeier, M., and Snow, E. C. (1988).J.Zmmunol. 140, 367. 451. Bunjes, D., Hardt, C., Rollinghoff, M., and Wagner, H. (1981).EucJ. Zmmunol. 11, 657. 452. Kronke, M.,Leonard, W. J., Depper, J. M., Arya, S. K., Wong-Staal, F., Gallo, R. C., Waldmann, T. A., and Greene, W. C. (1984).Proc. Natl. Acad. Scz. US.A . 81, 5214.
100
ELLEN S. VITETTA ET A L
453. Yefenof, E.,Sanders, V. M., Uhr, J. W., and Vitetta, E. S. (1986).J.Immunol. 137,85. 454. Springer, T., Dustin, M. L., Kishimoto, T. K., and Marlin, S. D. (1987).Annu. Rev. Immunol. 5, 223. 455. DeFranco, A. L., Ashwell, J. D., Schwartz, R. H., and Paul, W. E. (1984). J. fip. Med. 159, 861. 456. Hirohata, S . , Jelinek, D. F.,and Lipsky, P. E. (1988).J. Immunol. 140, 3736. 457. Cassatt, D. R.,Kaplan, A. M., and Cohen, D. A. (1988).J.Immunol. 141,1437. 458. Ratcliffe, M. J., and Julius, M. H. (1983).J. Immunol. 131, 581. 459. Dutton, R. W., Falkoff, R., Hirst, J. A., Hoffman, M., Kappler, J. W., Kottman, J. R., Lesley, J. F., and Vann, D. (1971). h o g . Immunol. 1, 355. 460. Sporn, M. B., and Roberts, A. B. (1988). Nature (London) 332, 217. 461. O’Garra, A,, Umland, S., DeFrance, T., and Christiansen, J. (1988).Immunol. lbday 9, 45. 462. Abbas, A. K. (1988). Immunol. Today 9, 89. 463. Nicola, N. A. (1987). Immunol. Today 8, 134. 464. Zachary, I., and Rozengurt, E. (1985). Cancer Sum. 4, 729. 465. Fernandez-Botran, R.,Sanders, V. M., and Vitetta, E. S. (1989).J Exp. Med. (in press). 466. Taussig, M. J. (1974).Nature (London) 248, 234. 467. Mozes, E.,Isac, R., and Taussig, M. J. (1975).J. Ex@. Med. 141, 703. 468. Kakiuchi, T., Chesnut, R. W., and Grey, H. M. (1983).J.Immunol. 131, 109. 469. Noelle, R.,Krammer, P. H., Ohara, J., Uhr, J. W., and Vitetta, E. S . (1984). R o c . Natl. Acad. Sci. U.S.A. 81, 6149. 470. Roehm, N. W.,Liebson, H. J., Zlotnik, A., Kappler, J., Marrack, P., and Cambier, J. C. (1984).J. Ex#. Med. 160, 679. 471. Oliver, K., Noelle, R. J., Uhr, J. W., Krammer, P. H., and Vitetta, E. S. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 2465. 472. Rabin, E. M., Ohara. J., and Paul, W. E. (1985).Roc. Natl. Acad. Sci. U.S.A. 82, 2935. 473. Vitetta, E. S., Ohara, J., Myers, C. D., Layton, J. E., Krammer, P. H., and Paul, W.E. (1985).J Ex#. Med. 162, 1726. 474. Sideras, P.,Bergstedt-Lindqvist,S., MacDonald, H. R., andSeverinson, E. (1985). E u r J Immunol. 15, 586. 475. Hamaguchi, Y.,Kanakura, Y., Fujita, J., Takeda, S.-I., Nakano, T., Tarui, S., Honjo, T., and Kitamura, Y. (1987).J. f i p . Med. 165, 268. 476. Hudak, S. A., Gollnick, S . O., Conrad, D. H., and Kehry, M. R. (1987). Proc. Natl. Acad. Sci. U.S.A. 84,4606. 477. Lee, W.T., and Conrad, D. H. (1986).J.Immunol. 136, 4573. 478. Zlotnik, A., Daive. B., Ransom, J., and Zipon, D. (1986).J Leukocyte Biol. 40,314. 479. Crawford, R.M., Finbloom, D. S., Ohara, J., Paul, W. E., and Meltzer, M. S. (1987).J.Immunol. 139, 135. 480. Zlotnik, A., Fisher, M., Roehm, N., andzipon, D. (1987).J Immunol. 138,4275. 481. Peschell, C., Paul, W. E., Ohara, J., and Green, I. (1987). Blood 70, 254. 482. Rennick, D., Yang, G., Muller-Sieburg, C., Smith, C., Arai, N., Takabe, Y., and Gemmell, L. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 6889. 483. DuBois, P., Stein, P., Ennist, D., Greenblatt, D., Mosmann, T., and Howard, M. (1987).J. Immunol. 139, 1927. 484. Stein, P., DuBois, P., Greenblatt, D., and Howard, M. (1986).J.Immunol. 136, 2080.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
101
485. Jones, S., Chen, Y. W., Isakson, P., Layton, J., Pure, E., Word, C., Krammer, P. H., Tucker, P., and Vitetta, E. S. (1983). J. Immunol. 131, 3049. 486. Bergstedt-Lindqvist, S., Moon, H.-B., Persson, U., Moller, G., Heusser, C., and Severinson, E. (1988). Euz J. Immunol. 18, 1073. 487. Finkelman, F. D., Katona, I. M., Urban, J. F., Jr., Snapper, C. M., Ohara, J., and Paul, W. E. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 9675. 488. Hu-Li, J., Shevach, E. M., Mizuguchi, J., Ohara, J., Mosmann, T., and Paul, W. E. (1987).J.Ex#. Med. 165, 157. 489. Ohara, J.. and Paul, W. E. (1987). Nature (London) 325, 537. 490. Park, L. S.,Friend, D., Grabstein, K., andUrdal, D. L. (1987). R o c . Natl. Acad. Sci. U.S.A. 84, 1669. 491. Nakajima, K., Hirano, T., Koyama, K., and Kishimoto, T. (1987).J. Immunol. 139, 774. 492. Lowenthal, J. W., Castle, B. E., Christiansen, J., Schreurs, J., Rennick, D., Arai, N., Hoy, Y., Takebe, Y., and Howard, M. (1988).J. Immunol. 140, 456. 493. Widmer, M. B., and Grabstein, K. H. (1987). Nature (London) 326, 795. 494. Pfeifer, J. D., McKenzie, D. T., Swain, S. L., and Dutton, R. W. (1987). J. Exp. Med. 166, 1464. 495. Trenn, G., Takayama, H., Hu-Li, J., Paul, W. E., andsitkovsky, M. V. (1988). J. Immunol. 140, 1101. 496. Palacios, R.,Sideras, P., and von Boehmer, H. (1987). EMBOJ. 6, 91. 497. Zlotnik, A., Ransom, G., Frank, A., Fisher, M., and Howard, M. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 3856. 498. Lowenthal,J. W., Ransom, J., Howard, M., and Zlotnik, A. (1988).J.Zmmunol. 140, 474. 499. Carding, S . R.,and Bottomly, K. (1988).J. Immunol. 140, 1519. 500. Takatsu, K., Tanaka, K., Tominaga, A., Kumahara, Y., andHamaoka, T. (1980). J. Immunol. 125, 2646. 501. Swain, S. L., and Dutton, R. W. (1982).J. Ex#. Med. 156, 1821. 502. Swain, S. L., Howard, M., Kappler, J., Marrack, P., Watson, J., Booth, R., Wetzel, G. D., and Dutton, R. W. (1983).J. Ex#. Med. 158, 822. 503. Yokota, T., Coffman, R. L., Hagiwara, H., Rennick, D. M., Takebe, Y., Yokota, K., Gemmell, L.. Shrader, B., Yang, G., Meyerson, P., Luh, J. Hoy, P., Pene, J., Briere, E,Banchereau, J., DeVries, J., Lee, F. D., Arai, N., and Arai, K. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 7388. 504. Harada, N., Kikuchi, Y., Tominaga, A., Takaki, S., and Takatsu, K. (1985). J. Immunol. 134, 3944. 505. Kinashi, T., Harada, N., Severinson, E., Tanabe, T., Sideras, P., Kinishi, M., Azuma, C., Tominaga, A., Bergstedt-Lindqvist, S., Takahashi, M., Matsuda, F., Yaoita, Y., Takatsu, K., and Honjo, T. (1986). Nature (London) 324, 7 0 . 506. Sanderson, C. J., O’Garra, A., Warren, D. J., and Klaus, G. G. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 437. 507. O’Garra, A.,Warren, D., Holman, M., Pophan, A. M., Sanderson, C. J., and Klaus, G. G. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 5228. 508. Takatsu, K., Kikuchi, Y., Takahashi, T., Honjo, T., Matsumoto, M., Harada, N., Yamaguchi, N., and Tominaga, A. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 4234. 509. Nakanishi, K., Yoshimoto, T., Katoh, Y., Ono, S., Matsui, K., Hiroishi, K., Noma, T., Honjo, T., Takatsu, K., Higashino, K., and Hamaoka, T. (1988).J. Immunol. 140, 1168.
102
ELLEN S. VITETTA ET AL.
510. Harada, N., Matsumoto, M., Koyama, N., Shimizu, A., Honjo, T., Tominaga, A., and Takatsu, K. (1987). Immunol. Lett. 15, 205. 511. Loughnan, M. S.,Takatsu, K., Harada, N., andNossa1, G. J. (1987).Proc. Natl. Acad. Sci. U.S.A. 84, 5399. 512. Swain, S . L. (1985). J. Immunol. 134, 3934. 513. Layton, J. E., Krammer, P. H., Hamaoka, T., Uhr, J. W., and Vitetta, E. S. (1985).J. Mol. Cell. Immunol. 2, 155. 514. Karasuyama, H., Rolink, A., and Melchers, F. (1988).J. h p . Med. 167, 1377. 515. Takatsu, K., Tominaga, A., Harada, N., Mita, S.,Matsumoto, M., Takahashi, T., Kikuchi, Y.,and Yamaguchi, N. (1988). Immunol. Rev. 102, 107. 516. Nakanishi, K., Malek, T. R., Smith, K. A., Hamaoka, T., Shevach, E. M., and Paul, W.E. (1984).J. Exp. Med. 160, 1605. 517. Purkemn, J. M., Newberg, M., Wise, G., Lynch, K. R., and Isakson, l? C. (1988). J. Ex#. Med. 168, 1175. 518. Diu, A., LeClercq, L., Dautry-Vanat, A., and Theze, J. (1987). Cell. Immunol. 107, 471. 519. Butterworth, A. E., Taylor, D. W., Veith, M. C., Vadas, M. A, Dessein, A., Sturrock, R. F., and Wells, E. (1982). Immunol. Rev. 61, 5. 520. Capron, A., Dessaint, J. P., Capron, M., Joseph, M., and Torpier, G. (1982). Immunol. Rev. 61, 41. 521. Gleich, G. J., and Loegering, D. A. (1984). Annu. Rev. Immunol. 2, 429. 522. Gillis, S.,Mochizuki, Y.,Cordon, l? J., Hefeneider, S. H., Ramthun, C. A., Gillis, A. E., Frank, M. B., Henney, C. S., and Watson, J. D. (1982). Immunol. Rev. 63, 167. 523. Farrar, J. J., Simon, P. L., Farrar, W. L., Koopman, J., and Fuller-Bonar, J. (1980). Ann. N . Y. Acad. Sci. 332, 303. 524. Wagner, H., and Rollinghoff, B. M. (1978). J. Exp. Med. 148, 1523. 525. Okada, M., and Henney, C. S. (1980).J. Immunol. 125, 300. 526. Farrar, W. L., Johnson, H. M., and Farrar, J. J. (198l).J. Immunol. 126, 1120. 527. Liebson, H. J., Marrack. P., and Kappler, J. (1982).J. Immunol. 129, 1398. 528. Pike, B. L., Raubitschek, A., andNosa1, G. J. V. (1984). Proc. Natl. Acad. Sci. U.S.A. 81, 7917. 529. Swain, S. L., Dennert, G., Warner, J. F., and Dutton, R. W. (1981). R o c . Natl. Acud. Sci. U.S.A. 78, 2517. 530. Parker, D. C. (1982).J. Immunol. 129, 469. 531. Farrar, J. J., Howard, M., Fuller-Farrar,J., and Paul, W. E. (1983).J. Immunol. 131, 1838. 532. Osawa, H., and Diamantstein, T. (1984).J. Immunol. 132, 2445. 533. Zubler, R. H., Lowenthal, J. W., Erard, E,Hashimoto, N . , Devos, R., and MacDonald, H. R. (1984).J. Exp. Med. 160, 1170. 534. Malek, T. R.,Robb, R. J., and Shevach, E. M. (1983). Proc. Natl. Acud. Sci. U.S.A. 80, 5694. 535. Prakash, S.,Robb, R.J., Stout, R. D., and Parker, D. C. (1985).J. Immunol. 135, 117. 536. Lowenthal, J. W.. Zubler, R. H., Nabholz, M., andMacDonald, R. H. (1985). Nature (London) 315, 669. 537. Tanalca, T., S a x , O., Doi, S.,Suemura, M., Neguro, S.,and Kishimoto, S. (1988). J. Clin. Invest. 82, 316. 538. Saiki, O.,Tanaka, T., Doi, S., and Kishimoto, S. (1988).J. Immunol. 140,853. 539. Bich-Thuy, L., and Fauci, A. S. (1985). Euz J. Immunol. 15, 1075.
CELLULAR INTERACTIONS I N THE HUMORAL IMMUNE RESPONSE
103
540. Sharon, M., Klausner, R. D., Cullen, B. R., Chizzonite, R., and Leonard, W. J. (1986). Science 234, 859. 541. Budo, M., Kozak, R., Goldman, C. K., and Waldman, T. A. (1986). Proc. Natl. Acad. Sci. US.A. 83, 9694. 542. Teshigawara, K., Wang, H.-M., Kato, K., andsmith, K. A. (1987).J. Erp. Med. 165. 223. 543. Robb, R. J., Rusk, C. M.,Yodoi, J., and Greene, W. C. (1987). Roc. Natl. Acad. Sci. U S A . 84, 2002. 544. Dukovich, M., Wano, Y.,Bich-Thuy, L. T., Katz, P., Cullen, B. R., Kehrl, J. H., and Greene, W. C. (1987). Nature (London) 327, 518. 545. Saragovi, H., and Malek, T. R . (1987).J. Immunol. 139, 1918. 546. Saragovi, H., and Malek, T. R . (1988).J. Immunol. 141, 476. 547. Wang, H.-M., and Smith, K. A. (1987). J. Ex@. Med. 166, 1055. 548. Lowenthal, J. W., and Greene, W. C. (1987).J. fip. Med. 166, 1156. 549. BichThuy, L. T., Dukovich, M., Peffer, N. J., Fauci, A. J.. Kehrl, J. H., and Greene, W. C. (1987).J. Immunol. 139, 1550. 550. Siegel, J. P., Sharon, M., Smith, P. L., a n d h n a r d , W. J. (1987). Science238, 75. 551. Sharon, M., Siegel, J. P., Tosato, G., Yodoi, J., Gerrard, T. L., and Leonard, W. J. (1988). J. Ex#. Med. 167, 1265. 552. Kehrl, J. H., Dukovich, M., Whalen, G.. Katz, P., Fauci, A. S.,and Greene, W. C. (1988). J. Clin. Invest. 81. 200. 553. Robb, R. J., Greene, W. C., and Rusk, C. M. (1984).J. Ex#. Med. 160, 1126. 554. Malkovsky, M., Loveland, B., North, M., Asherson, G. L., Gao, L., Ward, P., and Fiers, W. (1987). Nature (London) 325, 262. 555. Glasebrook, A. L., Kelso, A., and MacDonald, H. R. (1983).J. Immunol. 130, 1545. 556. Handa, K., Suzuki, R., Matsui, H., Shimizu, Y., and Kumagai, K. (1983). J. Immunol. 130, 988. 557. Trinchieri, G., and Perusia, B. (1985). Immunol. Today 6, 131. 558. Schultz, R. M., and Kleinschmidt, W. J. (1983). Nature (London) 305, 239. 559. Steeg, P. S.,Moore, R. N., Johnson, H.M., andoppenheim, J. J. (198Z).J. Exp. Med. 156, 1780. 560. King, D. P., and Jones, P. P. (1983).J. Immunol. 131, 315. 561. Wong, G. H. W., Clark-Lewis, I., McKimm-Breschkin, J. L., Harris, A. W., and Schrader, J. W. (1983).J. Immunol. 131, 788. 562. Warren, M. K., and Vogel, S . N. (1985).J. Immunol. 134, 2462. 563. Pace, J., Russell, S., Torres, B., Johnson, H., and Gray, P. (1983).J. Immunol. 130, 2011. 564. Tite, J. P., Powell, M. B., and Ruddle, N. H. (1985).J. Immunol. 135, 25. 565. Liebson, H. J., Gefter, M.,Zlotnik, A., Marrack, P., and Kappler, J. W. (1984). Nature (London) 309, 799. 566. Sidman, C. L., Marshall, J. D., Schultz, L. D., Gray, P.W., and Johnson, H. M. (1984). Nature (London) 309, 801. 567. Finkelman, F. D., Ohara, J., Goroff, D. K., Smith, J.. Villacreses, N., Mond, J. J., and Paul, W. E. (1986).J. Immunol. 137, 2878. 568. Snapper, C. M., Peschell, C., and Paul, W. E. (1988). J. Immunol. 140, 2121. 569. Finkelman, F. D., Katona, I. M., Mosmann, T. R., andcoffman, R. L. (1988). J. Immunol. 140, 1022. 570. D u r n , S. K., Schmidt, J. A., and Oppenheim, J. J. (1985). Annu. Rev. Immunol. 3 . 263.
104
ELLEN S. VITETTA ET A L .
571. Dinarello, C. A. (1988). FASEBJ. 2, 108. 572. Mizel, S. B. (1982). Immunol. Rev. 63, 51. 573. Beeson, P. B. (1948).J. Clin. Invest. 27, 524. 574. Kampschmidt, R. E,Upchurch, H. F., Eddington, C. L., and Pulliam, L. A. (1973).A m . J . Physiol. 224, 530. 575. Gery, I., Gershon, R. K.,and Waksman, B. H. (1972).J. &p. Med. 136, 128. 576. Lomedico, P. T., Gubler, U., Hellman, C. P., Dukovich, M., Giri, J. G., Pan, Y. E., Collier, K., Semionow, R., Chua, A., and Mizel. S. B. (1984). Nature (London) 312, 458. 577. Gray, P. W.,Glaister, D., Chen, E., Goeddel, D. V., and Pennica, D. (1986). J. Immunol. 137, 3644. 578. Hoffman, M. K., Koenig, S.,Mittler, R. S.,Oettgen, H. F., and Ralph, P. (1979). J. Immunol. 122, 497. 579. Giri, J. G., Kincade, P. W., and Mizel, S. B. (1984).J. Immunol. 132, 223. 580. Howard, M., Mizel, S. B., Lachman, L., Ansel, J., Johnson, B., and Paul, W. E. (1983).J.Exp. Med. 157, 1529. 581. Chiplunkar, S., Langhorne, J., and Kaufmann, S. H. (1986).J. Immunol. 137, 3748. 582. Pike, B. L., and Nossal, G. J. (1985). R o c . Natl Acad Sci. U.S.A. 82, 8153. Iscove, N. N., and Coutinho, A. (1980).Nature (London) 283, 583. Larsson, E.-L., 664. 584. Smith, K. A., Lachman, L. B., Oppenheim, J. J., and Favata, M. F. (1980). J. Ex$. Med. 151, 1551. 585. Kaye, J., Gillis, S.,Mizel, S.B., Shevach, E. M., Malek, T. R., Dinarello, C. A., Lachman, L. B., and Janeway, C. A. (1984).J. Immunol. 133, 1339. 586. Schrader, J. W. (1986). Annu. Rev. Immunol. 4,205. 587. Lee, J. C., Hapel, A. J., and Ihle, J. N. (1982).J. Immunol. 128, 2393. 588. Johnson, G.R., and Metcalf, D. (1977). Proc. Natl. Acad. Sci. U.S.A. 74, 3879. 589. Hara, H., and Ogawa, M. (1978). Am. J. Hematol. 4,23. 590. Burgess, A. W., Metcalf, D., Russell, S. H. M., and Nicola, N. A. (1980). Biochem. J. 185, 301. 591. Fung, M. C., Hapel, A. J., Yuer, S.,Cohen, D. R., Johnson, R. M., Campbell, H. D., and Young, T. G. (1984). Nature (London) 307, 233. 592. Yokota, T.,Lee, F., Rennick, D., Hall, C., Arai, N., Mosmann, T., Nabel, G., Cantor, H., and Arai, K. (1984). Proc. Natl. Acad. Sci. U.S.A. 81, 1070. 593. Schrader, J. W.,and Clark-Lewis, I. (1982).J. Immunol. 129, 30. 594. Palacios, R., Henson, G., Steinmetz, M., and McKearn, J. (1984). Nature (London) 309, 126. 595. Sideras, P., and Palacios, R. (1987). Eur. J. Immunol. 17, 217. 596. Van Snick, J., Vink, A., Cayphas, S., and Uyttenhove, C. (1987).J. f i p . Med. 165, 641. 597. Van Snick, J., Cayphas, S.,Vink, A., Uyttenhove, C.,Coulie, P. G., Rubira, M. R., and Simpson, R. J. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 9679. 598. Nordan, R. P., and Potter, M. (1986). Science 233, 566. 599. Wong, G. G., and Clark, S. C. (1988). Immunol. Today 9, 137. 600. Content, J., DeWit, L., Poupart. P., Opdenakker, G., Van Damme, J., and Billiau, A. (1985). Eur. J. Biochem. 152, 253. 601. Van Damme, J., Opdenakker, G., Simpson, R. J., Rubira, M. R., Cayphas, S., Vink, A., Billiau, A., and Van Snick, J. (1987).J. Exp. Med. 165, 914. 602. Zilbmtein, A., Ruggieri, R., Korn, J. H., and Revel, M. (1986).EMBOJ. 5,2529.
CELLULAR INTERACTIONS IN THE HUMORAL IMMUNE RESPONSE
105
603. Hirano, T., Yasukawa, K., Harada, H., Taga, T., Watanabe, Y.,Matsuda. T., Kashiwamura, S., Nakajima, K., Koyama, K., Iwamatsu, A.. Tsunasawa, S., Sakiyama, F., Matsiu, H., Takahara, Y., Taniguchi, T., and Kishimoto, T. (1986). Nature (London) 324, 73. 604. Muraguchi, A., Hirano, T., Tang, B., Matsuda, T., Horii, Y., Nakajima, K., and Kishimoto, T. (1988).J. Ex@. Med. 167, 332. 605. Takai, Y., Wong, G. G., Clark, S. C., Burakoff, S.J., andHerrmann, S. H. (1988). J. Immunol. 140, 508. 606. Uyttenhove, C., Coulie, P. G., and Van Snick, J. (1988).J.Exp. Med. 167, 1417. 607. Gannan, R. D., Jacobs, K. A., Clark, S. C . , and Raulet, D. H. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 7629. 608. Baroja, M. L., Ceuppens, J. L., Van Damme, J., and Billiau, A. (1988). J. Immunol. 141, 1502. 609. Stanley, E. R.,Bartocci, A., Patinkin, D., Rosendall, M., and Bradley, T. R. (1986). Cell (Cambridge, Mass.) 45, 667. 610. Smith, K. A.,Gibride, K. J., and Favata, M. E (1980). Nature (London) 287, 353. 611. Gillis, S., and Mizel, S. B. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 1133. 612. Farrar, J. J., Mizel, S. B., Fuller-Farrar, J.. Farrar, W. L., and Hilfiker, M. Z. (1980).J. Immunol. 125, 793. 613. Dinarello, C. A.,Marnoy, S. O . , and Rosenwasser, L. J. (1983).J.Immunol. 130, 890. 614. Glasebrook, A. L.,Kelso, A., Zubler, R. H., Ely, J. M., Prystowsky, M. B., and Fitch, E W. (1982). In “Isolation, Characterization, and Utilization of T Lymphocyte Clones”(C. G. Fathman and E W. Fitch, eds.), p. 341. Academic Pres, New York. 615. Nabholz, M., and MacDonald, H. R. (1983). Annu. Rev. Immunol. 1, 273. 616. Jenkinson, E. J., Kingston, R.,and Owen, J. J. T. (1987).Nature (London) 329, 160. 617. Shimonkevitz, R. P.,Hosmann, L. A., Bevan, M. J., and Crispe, I. N. (1987). Nature (London) 329, 157. 618. Brooks, K. H., and Vitetta, E. S. (1986).J. Immunol. 137, 3205. 619. Ohara, J.. and Paul, W. E. (1987). Annu. Rev. Immunol. 5 , 429. 620. Brown, M. A., Pierce, J. H., Watson, C.J., Falco, J., Ihle, J. N., and Paul, W. E. (1987). Cell (Cambridge, Mass.) 50, 809. 621. Thompson, C.B., Schaefer, M. E.,Finkelman, E D., Scher, I., Farrar, J.. and Mond, J. J. (1985).J. Zmmunol. 134, 369. 622. Brown, M., Hu-Li, J., and Paul, W. E. (1988).J. Immunol. 131, 504. 623. Sanderson, C. J., O’Garra, A., Warren, D. J., and Klaus, G. G. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 437. 624. Harriman, G . R.,Kunimoto, D. Y., Elliott, J. F,, Paetkau, V., and Strober, W. (1988).J. Immunol. 140, 3033. 625. Ikebuchi, K., Wong, G. G., Clark, S. C., Ihle, J. N., Hirai, Y.,and Ogawa, M. (1987). Proc. Natl. Acad. Sci. U S A . 84, 9035.
ADVANCES IN IMMUNOLOGY, VOL. 45
MHC-Antigen Interaction: What Does the T Cell Receptor See? PHlLlPPE KOURILSKY AND JEAN-MICHEL CLAVERIE Unite do Biologie Moleculaire du Gene, U277 INSERM and UAC 115 CNRS. and Unite d’informailquo Scientifique, instltut pbrteur, 75724 Fbris Cedex 15, France
1. Introduction
In the last few years, considerable progress has been made in the understanding of the mechanisms by which T cells specifically recognize antigens presented by antigen-presenting cells (APCs). In most instances, a specific receptor carried on T cells (TcR) interacts with the antigen “presented” by molecules of the major histocompatibility complex (MHC). Since the seminal observations that T cells recognized both native and denatured antigens (Gell and Benacerraf, 1959; Franzl, 1962; Campbell and Garvey, 1963), and that APCs were necessary for T cells to recognize antigens (Mosier, 1967), the phenomenon of antigen presentation has been gradually elucidated. It was recognized, after the work of Ziegler and Unanue (1981), that in most cases the antigen is processed by the APC, such that fragments of antigen, rather than antigen per se, are presented by MHC molecules and are recognized by TcRs. The early experimental work on this major finding has been reviewed in detail (Unanue and Allen, 1987; Buus et al., 1987a,b; Long, 1988). There are numerous examples of synthetic peptides corresponding to defined portions of protein antigens that actually substitute for antigen in welldefined immunological reactions involving T cells. In several instances, it was shown directly that such peptides bind to MHC molecules. So far, however, such biochemical data are rather scarce, and the binding of the processed antigen to MHC molecules is often inferred from functional assays involving recognition by TcRs. It is well known that most MHC molecules are highly polymorphic (reviewed in Charron et al., 1988; Kappes and Strominger, 1988; Long et al., 1988; Parham et al., 1988a; Klein, 1986; Nathenson et al., 1986; Lew et al., 1986a). This is indeed how they were initially discovered as major transplantation antigens (Gorer, 1937; Snell, 1948; Dausset, 1958). This is also how their involvement in antigen presentation was initially perceived through the phenomenon known as MHC restriction. The 107 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
108
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
polymorphism of MHC molecules may influence both antigen presentation and recognition by TcRs in many ways. Polymorphic residues from the peptide-binding site may influence the selection of a given set of fragments from the native antigen as well as their conformation and orientation. Other polymorphic residues may positively or negatively interfere with the recognition of the bound peptide with a given TcR. In the whole organism, the MHC-dependent tolerance to self-antigens will also influence the expression of the T cell repertoire (repertoire selection). Thus, much of the experimental evidence pertains to reactions involving a multimolecular complex, rather than solely antigen and MHC. Accordingly, it is often difficult to interpret experimental data relevant to MHC- antigen interaction without making assumptions on what the TcR actually sees, and how the T cell repertoire was generated. Starting from a complex situation, initially described in a global and imaged, albeit sometimes amazingly precise, fashion, the field has rapidly moved toward better defined and more molecular views. These advances have induced shifts of concepts which lead to a reappraisal of several questions. It is the purpose of this review to survey recent data and to analyze critically their interpretation. We show that the data can fit within a consistent, though hypothetical, framework of MHC-antigen-TcR interactions. It is our bias that the underlying hypotheses, which involve the presentation of self-peptides and their variants (Kourilsky and Claverie, 1986, 1989), may provide a unifying view where paradoxes that have intrigued immunologists for years may plausibly be resolved. To facilitate the reading, we have included a few schematic drawings of MHC molecules and TcRs. II. The T Cell Receptor and Accessory Molecules
The TcR is a major partner involved in the recognition of antigen as it is presented by the MHC molecules on an antigen-presenting cell. In addition, so-called accessory molecules may also strengthen the interaction between APCs and T cells. Well-documented reviews on TcRs (Kronenberg et al., 1986; Wilson et al., 1988; Davis and Bjorkman, 1988) and accessory molecules are available (P.Anderson et al., 1988; Dustin et al., 1988). Here, we give a brief outline of features relevant to the present discussion. TYPES OF TcRs A. THEDIFFERENT The polymorphic receptor carried by T cells that reacts with antigen presented by MHC molecule is a heterodimer made of two chains assembled in an antibody-likefashion. In adults, one finds predominantly
MHC-ANTIGEN INTERACTION
109
TcRs made with CY and 0 chains (the a-P TcR). During ontogeny of the immune system, one finds T cells with a y-6 TcR made with y and 6 chains. These T cells seem to be predominantly found in epithelial tissues such as intestine and skin. Still other receptors may exist, as suggested by recent experiments performed in the chicken by Lahti et al. (1988). Matis et al. (1987a) have shown that a mouse 7-6TcR-mediated reaction can be MHC linked, as are ones mediated by the a-/3 TcRs. Diversity of TcRs is generated in much the same way as in immunoglobulins, through the combinatorial joining of DNA segments and the variable addition of nucleotides at junctions (see in Davis and Bjorkman, 1988; Marrack and Kappler, 1988a). No well-documented case of somatic mutation in TcRs has been reported so far. Accordingly, it has been thought that the overall diversity of TcRs might be less than that of antibodies. However, vastly different estimates (from lo9 to 1030) of theoretical repertoire sizes have been given (see in Davis and Bjorkman, 1988; Marrack and Kappler, 1988a). It is striking, however, that a limited number of variable genes have been found to be used in response to several antigens (cf. Section X,A), so that the actual repertoire size in a given individual may not be this high. Alignment of TcR sequences has led to defined hypervariable regions, much in the same way as in antibodies, with one possible major difference: the TcR CY and 0 sequences display only one well-clusteredregion of hypervariable residues (equivalent to CDR3). Outside this region, the variable residues are more evenly distributed than in antibodies and the regions corresponding to CDRl and CDR2 appear less diverse (Bougueleret and Claverie, 1987). In the absence of direct data on the three-dimensional structure of TcRs, models have been built under the assumption that TcRs display an immunoglobulin Fab-like structure (cf. IX,B). B. THECD3 COMPLEX AND ACCESSORY MOLECULES TcR heterodimers (a-p or y-6) are usually found associated with other molecules in the CD3 complex, which includes at least four additional polypeptide chains (reviewed in Clevers et al., 1988). In addition, a number of so-called accessory molecules participate in the interaction between T cells and APCs. Among these are CD4 and CD8, which determine restriction to a given MHC class (CD4, to class 11; CD8, to class I) and, to a certain extent determine helper T cells and CTL functions, respectively. Thus, CD4+ CD8- helper T cells recognize antigens presented by class I1 molecules, and CD8+ CD4CTLs recognize antigens presented by class I molecules. There are also examples of CD4+ CD8- class 11-restricted CTLs (Braakman et al.,
110
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
1987). Other molecules (ICAM-1, LFA-1, etc.) also participate in cellular interactions (reviewed by P. Anderson et al., 1988; Dustin et al., 1988). Transfection experiments have definitively established that MHCrestricted antigen recognition is carried by the a-0 TcR heterodimer (Dembic et al., 1986; Saito et al., 1987; Kuo and Hood, 1987; Malissen et al., 1988). There is evidence that the CD4 and CD8 molecules are essential when the specific recognition of MHC-presented antigens by TcRs is relatively weak (e.g., Gabert et al., 1987). On the contrary, they appear to be dispensable when this interaction is “strong.” For instance, CD8-independent CTLs can be isolated (for a recent example, see Dembic et al., 1986). The physical relationship of CD4 and CD8 to the CDS complex is difficult to assess at this stage. It has been found recently that CD8 can bind to human class I MHC molecules through a disulfide bond (Blue et a l . , 1988; Bushkin et al., 1988). CD8 is disulfide linked to CD1 on immature thymocytes (Snow et al., 1985). CD4 protein binds to class I1 molecules on B cells (Doyle and Strominger, 1987; Gay et al., 1988). Work in progress will rapidly define more precisely which contacts are made by CD4 and CD8. These accessory molecules may bind not only the MHC molecules of the APCs, which present the specific antigen, but may also bind other MHC molecules. Accordingly, one might have to distinguish between MHC-dependent reactivity (not directly dependent upon antigen) and MHC-restricted recognition of antigen. The mode of action of CD4 and CD8 is not yet fully understood. It has been suggested that CD4 might be involved in signal transduction (Rudd et al., 1988), when brought into close physical proximity of the CDS-TcR complex in class II-restricted antigen recognition (Ledbetter et al., 1988). Since CD4, CD8, and other accessory molecules do not contribute to the intrinsic specificity of interactions involving MHC, antigen, and TcR, they will not, in spite of their immunological importance, be discussed further. 111. The MHC Molecules
The major histocompatibility complex is a chromosomal region that has been extensively characterized in several mammals, particularly man and mouse (Flavell et al., 1986) (where it is named the HLA and the H-2 complex, respectively), and a few other organisms, particularly the rat, the rabbit (6.Kulaga et al., 1987), the swine (cf. Ehrlich et al., 1987; Sachs et al., 1988), and the chicken (6.Guillemot et al., 1988; for a general review, see Klein, 1986). Molecules encoded in the MHC (in short, “MHC molecules”) and relevant to antigen presentation so
MHC-ANTIGEN INTERACTION
111
far fall into two classes that differ structurally: class I MHC molecules are made of a polymorphic glycosylated “heavy” chain of about 45 kDa, which spans the cell membrane, associated with a poorly polymorphic, nonglycosylated, 12-kDa “light” chain, called &-microglobulin, not encoded in the MHC. Class I1 molecules are made of two polymorphic chains, a and 0, that both span the cell membrane and associate into a heterodimer. Early biochemical studies involving digestion with proteases suggested that class I heavy chains displayed three external structural domains, while a! and 0 chains of class I1 molecules each exhibited two. This belief was later reinforced by the striking correlation which was found between exons of the cloned genes and these putative domains. Two of these extracellular domains in both class I molecules and class I1 heterodimers display much more variability than the others, which more closely resemble immunoglobulin constant domain, and unambiguously include MHC molecules in the immunoglobulin superfamily (Williams and Barclay, 1988). Thus, it appeared possible that the structure of class I and class I1 molecules was, perhaps, not so different: each could be made, on the extracellular side, of two relatively constant and two more variable domains, albeit connected differently (three domains plus &-microglobulin versus “2 + 2”). Hypothetical molecular models displaying these features were built (Novotny and Auffray, 1984). STRUCTURE OF HLA-A2 A. THECRYSTAL Although the above predictions were grossly correct, the determination at the 3.5-Hi resolution of the crystal structure of the human class I antigen HLA-A2 was illuminating and constitutes a landmark in the field (Bjorkman et al., 1987a,b). The structure revealed two a helixes (correctly predicted by Novotny and Auffray, 1984) laying on a tray of eight antiparallel 0-pleated sheets. The two most variable domains of the class I heavy chain each contribute one a helix, but the 0-pleated sheets are connected in such.a way that these two domains belong to a single structural unit. The two a helixes and portions of P-pleated sheets underneath delineate a probable binding site for the processed antigen, as described below. Schematic drawings of the HLA-A2 structure are shown in Fig. 1. B. MODELSOF OTHERCLASSI MHC MOLECULES Sequence homologies between class I MHC molecules are such that different sequences are usually easily aligned. Variability plots have been made (Parham et al., 1988a,b). So far, the sequences of these other class I molecules do not contain residues that would preclude the folding of
112
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
their sequences into a three-dimensional structure very similar to the HLA-A2 prototype structure. Instead, tridimensional structural models based on the latter have been constructed for several class I HLA and H-2 molecules (Bjorkman et al., 1987a,b; Ajitkumar et al., 1988; Rothbard et al., 1989; Chalufour et al., in preparation). It is thus believed that class I molecules, within a given species as well as in different species, are quite similar in structure. This assumption is also supported by the fact that human HLA class I molecules can serve to present antigen to murine effectors (e.g., Kievits et al., 1987a, 1989). Under this assumption, it is remarkable, as emphasized by Bjorkman et al., (1987b), that many polymorphic residues map within the presumptive peptide-binding pocket. The functional implications will be discussed later. C. CLASSI1 MOLECULES MAYHAVEA SIMILAR STRUCTURE For each of the two polypeptide chains of class I1 molecules, (Y and 0,polymorphic residues are concentrated in the N-terminal domains, a1 and 61. These domains are mostly responsible for binding peptides in what appears to be a single site (Unanue and Allen, 1987; Buus et a l . , 1987a). On this ground, one can align the sequences of the (Y and 0 class I1 chains with the class I heavy chain by matching the al and 01 N-terminal domains of class I1 with the a-1and (YZ domains of class I, respectively (Brown et al., 1988). One can next build a hypothetical model of class I1 molecules, the folding of which closely resemble that of HLA-A2, with two (Y helixes supported by an array of eight /3-pleated sheets (Brown et al., 1988). A modification of this model has been FIG. 1. Schematic representation of the putative peptide-binding site of the HLAA2 molecule (adapted from Bjorkman et al., 1987a,b). A cleft is formed by a floor of fl-pleated sheets on top of which are located two a-helical domains. Various properties of the residues constituting this site have been indicated. (A) Locations in the HLA-A2 structure of the bona fide polymorphic residues in human MHC class I molecules are indicated by stars. (B) Locations of residues the side chains of which are pointing up away from the peptide-binding groove are indicated by arrows. These residues might be available for interaction with the TcR molecule. Among these, only two residues (indicated by asterisks) are polymorphic. (C) Locations of residues the side chains of which are available for direct interaction with peptide antigen bound in the groove are indicated by arrowheads. The correlation with polymorphic locations (A) is striking. B and C are designed according to the indications of Bjorkman et aE. (1987a,b) and are verified on a molecular model (A. Chalufour, personal communication) reconstituted from the C-a coordinates. (D) Locations of hydrophobic (H), noncharged polar (P),and charged (+, - ) residues in the HLA-A2 site. Notice that most of the hydrophobic residues found around the site are not in a position to interact with a bound peptide antigen as shown in C.
113
114
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
D
MHC-ANTIGEN INTERACTION
115
proposed by McKean (1988). The recent results of Perkins et al. (1989) showing that peptides presented by class I molecules can be presented by class 11, and vice versa, support the notion that the structures of the peptide-binding site are similar in class I and class 11. However, although the major traits of this model are likely to be correct, one should remain cautious in extrapolating the details of the HLA-A2 structures (some of which lack a precise definition at 3.5 A resolution) to the various class I and, afortiori, class I1 molecules, the sequences of which are quite divergent. In particular, the orientation of the side chains of the polymorphic residues, which is critical for biological interpretations, cannot be safely extrapolated to molecules other than HLA-A2. D. POLYMORPHIC MHC MOLECULES MAY NOT BE THE ONLY MOLECULES ABLETO PRESENT ANTIGEN Polymorphic MHC molecules were shown to be involved in antigen presentation precisely because of their polymorphic character, but there is no evidence that they are the only molecules involved in antigen presentation. First, the immunoglobulin superfamily is rapidly expanding (reviewed in Williams and Barclay, 1988). It includes many new members of unknown function. For example, the CD1 antigens (Terhorst et a l . , 1981; Martin et al., 1986), not encoded by MHC and recently mapped to human chromosome 1 (Albertson et al., 1988) are associated to &-microglobulin, as are the class I molecules. It is interesting that class I HLA expression is modulated in thymic maturation in an inverse correlation with the expression of CD1 antigens (Gambon et a l . , 1988). Also, mouse CD1 and T L antigens coexist in the same thymus (Bradbury et al., 1988). Second, the inventory of class I and class I1 genes and/or molecules may not yet be complete. Thus, the list of human class I1 molecules continues to grow (Charron et al., 1988; Fernandez et al., 1988). The class I gene family in man and mouse includes many more genes than those coding for HLA-A, -B, and -C and H-2 K, D, and L. Depending on the mouse haplotype, there exist some 20 to 35 genes in the Qa-TLregion, some of which are pseudogenes, while others do encode identified products, such as the Qa-1 and Qa-2 antigens, the Tla antigens, the secreted products of the Q.0 gene, the poorly polymorphic albeit almost ubiquitously expressed product of the 37 gene, and several others (for a selection of recent references, see Gussov and Ploegh, 1987; Lew et al., 1986b; Homer and Murphy, 1987; Robinson, 1987; Stroyonovski et al., 1987; Transy et al., 1987; Sherman et al., 1988; Robinson et al., 1988; Abastado and Kourilsky, 1988). Furthermore, additional class Irelated genes have been found on mouse chromosome 17, outside of the MHC (Singer et al., 1988). and may include determinants coding for
116
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
the maternally transmitted antigen Mta (Fischer-Lindahl et al., 1987). In man, besides HLA-A, -B, and -C, several class I genes, some of them in a broad variety of cell types, are expressed (Geragthy et al., 1987; Shimizu et al., 1988; Mizuno et al., 1988). Apart from a postulated and nonestablished involvement of Qa-2 in embryonic development (Warner et al., 1987a,b), the function of all these molecules is currently unknown, as well as that of minor forms of the polymorphic molecules, translated from alternatively spliced RNAs, some secreted, others not (Kress et al., 1983; Transy et al., 1984; Krangel, 1986; McCluskey et al., 1986a,b; Lew et al., 1986a,b). Although it has been speculated that Qa and Tla gene products may be devoid of any function (Gussov and Ploegh, 1987), computer-aided model building suggests that their structure might resemble that of HLA-A2 (A. Chalufour, personal communication) and comprises a plausible peptide-binding site. Its size may vary though, and the Q l O binding pocket would contain bulky residues which would make it smaller (Mann et al., 1988). Certain of these molecules have a transmembrane region. Others are secreted and/or attached to the cell surface by a phosphoinositol bridge, such as the Qa-2 molecule (Stroyonovski et al., 1987; Soloski et al., 1988). There is evidence that the Q l O molecule, once hooked to a genuine transmembrane region, can serve as a target for alloreactive CTLs (Mann et al., 1988). Cells displaying Qa-2 anchored by a phosphoinositol bridge are also targets for alloreactive anti-Qa-2 CTLs (Mann and Forman, 1988). Lysis is sensitive to a phospholipase that cleaves the phosphoinositol bridge, so that Qa-2 (rather than a Qa-2 peptide) is presumably seen by TcRs of the CTLs. It is not impossible that certain of these molecules, although less polymorphic than the classical ones, might be involved in the presentation of processed (?)antigens (or in the transport of peptides?). The best understood reactions between immune cells involve interactions between APCs (including B cells) and CTLs or helper T cells in response to foreign antigens. For these reactions, a number of exceptions to the “rule” of MHC restriction have been reported (one possible interpretation being that the presenting element is a nonpolymorphic molecule, encoded in MHC or not). Many other cellular immune interactions are poorly understood so far (suppressive reactions, antiidiotypic help, etc.) (we have speculated that Qa and Tla molecules could be involved in the presentation of “idiopeptides,” i.e., processed antibodies and TcRs; Kourilsky et al., 1987). It is conceivable that they utilize self-antigens (rather than foreign antigens) together with other presenting molecules. Thus, our present knowledge in this field is still limited. Conversely, it should be kept in mind that MHC molecules might not
MHC-ANTIGEN INTERACTION
117
exclusively function in antigen presentation. There are reports showing that they may interact with other molecules, such as antibiotics (Horai et al., 1982), the receptor for epidermal growth factor (Schreiber et al., 1984), and the insulin receptor (Fehlmann et al., 1985; Kittur et al., 1987); one may recall here that Allison et al. (1988) have found that transgenic mice overexpressing Kbin the pancreas develop diabetes) and the low-affinity receptor for IgE (CD23) (Bonnefoy et al., 1988; see commentary by Edidin, 1988). Such interactions may or may not be related to antigen presentation. On the other hand, the question of signal transduction by MHC molecules and of the mechanism(s) by which they are modulated on the cell surface (6.IV,F) now receive much attention (Chen et al., 1987; Cambier et a l . , 1987). Recently, two proteins tentatively identified as the cytoplasmic actin A chain and a CAMPdependent protein kinases have been found associated to mouse class I1 molecules (Newell et al., 1988). We do not discuss here the regulation of the expression of class I and class I1 MHC molecules (for a few selected references, see Dorn et al., 1988; Hakem et al., 1989; Kimura et al., 1986; Israel et al., 1987; Korber et al., 1988; Reith et al., 1988; Sakurai and Strominger, 1988; Shirayashi, et al., 1988; Sullivan et a l . , 1987). In the long term, these studies will document important questions, such as the differences of expression of class I and class I1 molecules in various cell types and tumors, the mechanisms of induction by interferons and tumor necrosis factor (and possibly other agents), as well as other phenomena, such as the increased expression of MHC molecules on lymphocytes from aged mice (Sidman et al., 1987).
IV. The Processed Antigen
A. ANTIGEN Is USUALLY PROCESSED INTRACELLULARLY BEFOREITS PRESENTATION BY CLASS I1 MHC MOLECULES When a foreign protein is delivered to the immune system and triggers an MHC class II-mediated reaction, such as a B cell response, part of the antigen has been ingested by class II-bearing APCs. APCs include macrophages and B cells, bone marrow-derived dendritic cells, and a number of nonlymphoid cells, some of which express class I1 constitutively (Langerhans cells, epithelial cells of the gastrointestinal tract, intramesengial dendritic cells of the kidney, etc.), others being induced to express class 11, particularly by interferon-y (astrocytes and microglial cells in the central nervous system, thyroid epithelial cells, keratinocytes, etc.) (Skoskiewicz et al., 1985; Holt et al., 1988; Gaspari and Katz, 1988;
118
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
reviewed in Unanue and Allen, 1987; Bos and Kapsenberg, 1986; Merrill, 1987; Frei et al., 1987; Bland, 1988). Recently, human T cells have been shown to be capable of presenting antigens targeted to their surface molecules (Lanzavecchia et al., 1988a,b). B cells can function as APCs (Rock et al., 1984; Lanzavecchia, 1985), and can be regarded as antigenspecific APCs, inasmuch as antigen is complexed by the specific membrane-bound immunoglobulin, internalized and processed. Dendntic cells appear, by a variety of criteria, to be very efficient in antigen presentation (Boog et al., 1988a,b), a feature which may be related to the abundance and low degree of glycosylation of their class I1 (and class I; see below) MHC molecules (reviewed in Melief et al., 1988). Some of the evidence, however, relies on treatment with neuraminidase, which may have a variety of effects (Taira and Nariuchi, 1988; Krieger et al., 1988). The evidence for intracellular processing includes (1) kinetic data there is a measurable lag between the time native antigen is delivered to an APC and the time it becomes functionally detectable (in contrast, exogenous synthetic peptides become rapidly functional) (for recent studies, see Roosnek et al., 1988; Lakey et a l . , 1988); ( 2 ) sensitivity to inhibitors which block intracellular traffic, such as chloroquine; and (3) experiments using cellular membranes, or class I1 antigens embedded into artificial membranes, unable to activate helper T cells when given native antigen, but able to do so when exposed to processed antigen (often a synthetic peptide). The data have been reviewed by several authors (Unanue and Allen, 1987; Allen, 1987; Buus et al., 1987a,b; Livingstone and Fathman, 1987; Werdelin et al., 1988; see also Berzofsky, 1986) and will not be presented in detail here. We shall also not review studies on the delivery of coupled antigens to APCs such as the uptake of antigen or peptide coupled to antibodies directed against cell surface molecules (for a recent review, see Casten et a l . , 1988).
B. ANTIGENIs OFTENPROCESSED INTRACELLULARLY
BEFORE PRESENTATION BY CLASS I MHC MOLECULES Initially, many functional studies of class I MHC molecules involved the specific lysis of virus-infected cells by cytolytic T cells (CTLs). A number of viral proteins are exposed on the cell surface, or bud out of it. In addition, the status of certain “internal” proteins is not quite clear yet. Thus, viral tumor-associated antigens, such as SV40 large T, are present in small amounts on the cell surface (e.g., Henning and LaugeMutschler, 1983; Sharma et al., 1985) (as is a mouse tumor-specific transplantation antigen related to heat-shock proteins: Ullrich et al., 1986). Accordingly, it was first believed that these surface molecules were seen as such, in association with class I MHC molecules, by CTLs. This
MHC-ANTIGEN INTERACTION
119
assumption was for a time supported by the finding that the adenovirus El9 protein could be immunoprecipitated from the cell surface in association with class I MHC molecules (Kvist et al., 1978). This was, however, an exceptional situation, since other viral antigens could not be convincingly shown to be physically associated with class I molecules. Refinement of the biological analyses, using clones of CTLs and/or target cells with transfected cloned viral genes, then demonstrated that internal proteins, such as influenza DNA polymerase (Bennink et al., 1982) or nucleoprotein (NP) (Townsend and Skehel, 1984; Townsend et al., 1984), could be involved in recognition by CTLs. (The overall CTL response to influenza proteins is surveyed in Gotch et al., 1987a.) The gene coding for influenza hemagglutinin was manipulated in such a way as to produce a cytoplasmic protein, which was as active as the native one in promoting specific CTL recognition (Townsend et al., 1986a). The finding that internal viral proteins are often targets for CTL recognition was soon extended to other systems such as the vesicular stomatitis virus (Yewdell et al., 1986; Puddington et al., 1986), the respiratory syncytial virus (Bangham et al., 1986), and the SV40 T antigen (R. W. Anderson et al., 1988). This was the prelude to the important discovery by Townsend et al. (1985, 1986b) that influenza NP was processed intracellularly. Certain segments of the NP gene, expressed in transfected cells, converted them into specific targets for anti-NP CTL clones. So did certain synthetic peptides added to naive (nontransfected) cells. It could thus be inferred that part of the endogenouslysynthesized NP protein detectable in the nucleus undergoes degradation, leading to restricted presentation of NP fragments at the cell surface. This finding was extended to a variety of systems (see Sections VI,D and E). The case of the adenovirus El9 protein was clarified by the observation that when binding MHC molecules, it inhibits their transport and function (Andelson et al., 1987; Burget and Kvist, 1987; Burget et al., 1987). Most somatic cells express class I molecules and all class I-bearing cells are believed to be able to present antigen. There are, however, vast differences in the levels expressed by different cell types (reviewed in Harris and Gill, 1986; see recent data in Hunt et al., 1988), but a number of agents, particularly interferon-y, stimulate class I gene expression in many different cell types (Momburg et al., 1986). C. EXCEPTIONS TO INTRACELLULAR PROCESSING A few exceptions in which native antigen did not require intracellular processing have been reported. Walden et al. (1985, 1986) challenged the generality of the phenomenon on the basis of experiments in which class I1 molecules were embedded into liposomes together with an enzyme
120
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
(pig lactate dehydrogenase) that could be recovered primarily in active form after a specific stimulation of T helper cells. It was not excluded that proteases released by the T cells used in the assay did not degrade a small part of the enzyme antigen. Also, the homogeneity (possible presence of self-peptideson the class I1 molecules, see Section V) and density of class I1 molecules on the liposomes could be questioned. [In a separate study, Servis et al. (1986) showed that peptide 211-224 of pig lactate dehydrogenase could induce both helper and suppressor T cells in vivo.] Lee et al. (1988) have reported that human fibrinogen can be presented by class I1 molecules to T cells without intracellular processing. The portion of the polypeptide chain that binds to class I1 molecules belongs to a flexible arm. Extracellular proteases in the medium do not seem to play a role. The presentation of Lziteriu monocytogenes antigens does not require processing (Ziegler et al., 1987). The importance of antigen structure is clearly illustrated by experiments on the presentation of apamin, an 18-amino-acid-longpeptide found in bee venom, which does or does not require processing, depending on the presence or absence of an internal disulfide bridge (Regnier-Vigourouxet al., 1988). It thus appears possible that accessible and/or flexible portions of protein antigens are directly presented by cell surface class I1 MHC molecules without processing (Allen, 1987). A similar situation may exist for class I molecules, since Wraith and Vessey (1986) described CTL reactions against purified influenza proteins which did not require processing. D. Is PROCESSED ANTIGENA PEPTIDE? In most cases so far studied where a T cell clone specifically reacts with a given antigen presented by an APC, a synthetic peptide can be added exogenously and can substitute for the native antigen. Although this strongly suggests that processed antigen is a peptide, several notes of caution should be given: (1) the doses of peptides used are often quite high, and presumably well over the physiological ones; (2) in many experiments, the exogenous peptide reported to be functional could have been internalized and processed further (Fox et a l . , 1988); and (3) there are many unknowns in the pathway(s) of processing and presentation. Strikingly, presentation of native antigen by class I1 molecules involves a lipase-sensitive step and is sensitive to cerulinin, an antibiotic which interferes with the posttranslational addition of lipids on proteins (Falo et al., 1986, 1987a,b). (However, possible indirect effects on accessory molecules or other features of the cellular interactions have not been fully ruled out.) Thus, the possibility that presented antigens display some chemical modification should not be discarded. Furthermore, it may happen that nonpeptidic antigens are presented. A recent example is the
MHC-ANTIGEN INTERACTION
121
demonstration that the heme moiety of cytochrome c is a dominant T cell epitope falling under strict MHC restriction (Cooper et al., 1988). In summary, while it is clear that the processed antigens usually include a peptide moiety. their exact structures in vzvo have not been determined. For the sake of simplicity, we shall often use the term peptide to describe processed antigen, but the reservations expressed here should be kept in mind. We shall also use the following terminology to describe two functional features of presented antigens (Schwartz, 1985, 1986): the agretope refers to the subsets of amino acid residues which are involved in the binding to the MHC-restricting element. The epitope refers to the subsets of residues which are seen by (interact with) the TcR. Both types of subsets need not be continuous stretches of residues, and there is ample evidence (cf. Sections VI and VII) that this is usually not the case. E. THELENGTHS OF PRESENTED SYNTHETIC PEPTIDES RANGEBETWEEN 8 AND 14 RESIDUES Overlapping compilations of 57 synthetic peptides (Rothbard and Taylor, 1988) and 63 peptides (Claverie et a l . , 1988) displaying activity versus T cells in various assays suggest that epitopes carried in synthetic peptides lie, on the average, within 9 or 10 residues. Thus, functional epitopes of processed antigens should have about that size (Fig. 2). Optimization of peptide length has been made in a number of systems. In general, it seems that peptides shorter than 8 residues are quite inefficient (with the notable exception of angiotensin 111, which is 7 residues long; Buus and Werdelin, 1986). There seems to be no significant difference so far between the average length of peptides presented by class I or class I1 molecules.
E RECYCLING OF MHC MOLECULES AND ANTIGENPROCESSING Under the assumption that processed antigen is a peptide, MHC molecules can be seen as peptide receptors of loose specificity (Kourilsky and Claverie, 1986). In support of an analogy with several typical cell surface receptors, it has been found that class I and class I1 MHC molecules actively recycle under certain circumstances: thus, class 11, but not class I, MHC molecules rapidly recycle in activated B cells, while class I molecules rapidly recycle in activated T cells. In fibroblasts, no recycling was detected (Aragnol et al., 1986; Tse et al., 1986; reviewed in Pernis, 1988). Class I molecules in T lymphocytes are internalized via coated pits (Machy et al., 1987). It is tempting to speculate that internalized MHC molecules are loaded with processed antigen during recycling, but this has not yet been supported by direct experimental
122
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
evidence. The reason why recycling of class I and class I1 molecules is differentially regulated in B cells and T cells and depends on the activation of the latter has been a matter of speculation (Pernis, 1988; Leserman, 1987; Kourilsky et al., 1987) but is not yet understood. It is conceivable that, in cells where no recycling has been observed, a slow, undetected recycling takes place and plays a role in antigen presentation. It is also possible that de novo synthesized MHC molecules are loaded with peptide during their transit to the cell surface.
G. WHICHFACTORS INFLUENCE ANTIGENPROCESSING? A first factor relates to the structure of the antigen itself. Allen (1987) distinguishes three types of T cell determinants according to their processing requirements. T p e I does not require processing (cf. Section IV,C), type I1 requires unfolding only, and type I11 requires a proteolytic cleavage to allow the epitope to be subsequently formed. It is thus possible that amino acid residues not included in the epitope influence processing. This is actually implied by a variety of experiments. In the lysozyme system, it was found that residues outside the minimal peptide determinant used in H-2bmice could strongly influence antigenicity (Shastri et al., 1986; reviewed in Gammon et al., 1987). In influenza hemagglutinin, residue 17 has been found to be critical for presentation of an epitope located around residues 208-213 (Mills et al., 1988; Mills, 1986). It has also been shown that an acid-induced conformational change of hemagglutinin alters its presentation in various ways (Mills, 1986; Eisenlohr et al., 1988). Conversely, presentation could influence processing in the following sense: in a simple-minded view, one may think that, in many instances, partially denatured antigen binds to MHC molecules, presumably in relatively acidic vesicles, and that proteases next trim off the unprotected parts. If so, the presence of several potential agretopes in the same polypeptide chain might create an internal competition for presentation, FIG. 2. Relative size (van der Waals surfaces) of the putative peptide-binding groove and various peptides in a-helical or extended conformation (as drawn with the program MANOSK). Left: the 16-residue immunogenic peptide K16F (recognized by an HLAA2-restricted anti-HIV CTL line (Claverie et al., 1988) in extended conformation. Top right: the same peptide in canonical helical conformation. Bottom right: the 13-residueimmunogenic peptide (recognized by an HLA-AZ-restricted anti-influenza virus matrix protein (Gotch et al., 1987b) in helical conformation. Center: the a-1 and a - 2 helixes flanking the HLA-A2 putative peptide-binding site as seen from the &pleated sheet floor. The binding of such peptides in a-helical conformation deep in the groove obviously requires a tight tit or, alternatively, a conformational change of the site. More extended and flexible peptide conformations would impose less stringent steric constraints upon the HLA-antigen interaction.
m
123
MHC -ANTIGEN INTERACTION
and certain agretopes might be dominant. This is a plausible explanation of results obtained by Sercarz, Miller, and their colleagues (reviewed by Gammon et al., 1987). Brett et al. (1988)have also shown that in mvo immunization with segments of sperm whale myoglobin reveals new reactivities not previously detected with the native antigen, and Carbone et al. (1988)made similar findings with ovalbumin fragments (see review by Berzofsky, 1988). In antigen processing by specific B cells, a potentially important feature is the site by which the antigen is bound to the membrane attached antibody which mediates its internalization and processing. Data by Ozaki and Berzofsky (1987) and Manca et al. (1988) support the view that antibody binding influences antigen processing. Other factors most certainly pertain to the cellular routes, and probably to the cell types. First, it is important to recall that molecular “addresses” are found in proteins, so far in the form of relatively small stretches of amino acid residues. These addresses specifically direct certain proteins into certain cellular compartments. The first to be identified was the well-known signal peptide, but several others involved in the transport into (or retention within) the endoplasmic reticulum, the nucleus, or the mitochondria have been found. It may be relevant that an address to proteolytic compartments has been postulated to exist in thyroglobulin and in the invariant chain of class I1 molecules (Koch et al., 1987). Second, some cell specificity probably exists. For example, L cells transfected with class I1 genes do not behave like usual APCs inasmuch as they were found to be unable to present lysozyme, although they could present the appropriate lysozyme synthetic peptide (Shastri et al., 1985). Different sets of proteases could operate in different cellular ~~~~~
FIG. 3. A speculative molecular model of the TcR-peptide-HLA-A2 interaction (adapted from Claverie et al., 1988b). Two orthogonal views of the ternary complex. The TcR (top molecule) is modeled after the immunoglobulin Fab fold using the HPBMLT a and @ chain sequences followed by the relaxation of stereochemical constraints and energy minimization. The constant part is not shown. The antigen is represented as a 13-residuepeptide in a-helical conformation (only the C-a skeleton is shown). For clarity, none of the side chains is displayed. The peptide is actually making van der Waals contacts with both HLA-A2 and TcR residues. The HLA-A2 site (a-1 and a-2 domains) model (bottom molecule) is drawn from the available C-a coordinates (Bjorkman et al., 1987a.b). According to the proposed geometry of this complex, the V-a and V-0 CDRS loops (right behind each other) would mainly interact with the bound peptide, while the V-a CDRl/V-@CDR2 loops and V-aCDR2/V-/3 CDRl loops would be available to interact with HLA-A2 residues pointing up (Fig. 1B; see also Table I, p. 150). An additional TcR loop (leftmost) would also be available for interaction with the HLA-A2-antigen complex. The most protruding residue of this loop is at location 74. It is interesting to note that this location (outside the CDR1, CDR2, and CDR3 regions) belongs to a cluster of variable residues in two out of three groups of V-/3 TcR chains (Bougueleret and Claverie, 1987). For discussion in text, see page 160.
124
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
compartments and different cell types. The specificity of proteases in defining the specificity of the immune response is discussed by Werdelin et al. (1988). (For a review on intracellular proteases, see Bond and Butler, 1987.) H. THEPATHWAYS OF ANTIGEN PROCESSING As recently reviewed by several authors (Morrison et al., 1986a; Braciale et al., 1987b; Long, 1988; Jelachich and Biddison, 1988), there exist several pathways of antigen processing. 1. Exogenous Antigens
Exogenous antigens are endocytosed by APCs. A number of compounds block presentation of endocytosed antigens by class I1 MHC molecules. These include acidotropic agents (chloroquine and NH4C1), inhibitors of membrane recycling (such as monensin), several inhibitors of proteases (leucine peptidase and cysteine protease) (reviewed by Grey and Chesnut, 1985; see also Werdelin et al., 1988), and cerulinin (which blocks posttranslational modification of proteins by lipids (Falo et a l . , 1987a; cf. Section IV,D). In contrast, one may note that two compounds that inhibit two cytosolic enzymes that cleave insulin increase antigen presentation of insulin (Naquet et al., 1987). It was elegantly shown by Cresswell (1985) that the pathway followed by endocytosed transferrin intersects that taken by newly synthesized class I1 MHC molecules on their way to the cell surface. Thus, by following the route used by the latter, one may hope to learn something on antigen processing and loading on class I1 molecules. In this regard, Blum and Cresswell (1988)have studied the effects of the protease inhibitor leupeptin on the biosynthesis of class I1 molecules of a human B cell line. They conclude that leupeptin and antipain (which both block cathepsin I) prevent cleavage of the invariant chain and its dissociation from mature class I1 complexes in acidic vesicles probably distinct from lysosomes. Because leupeptin blocks antigen presentation, it appears possible that dissociation of the invariant chain is important in this process. Presentation of measles virus antigens by class I1 molecules was found to occur in the absence of the invariant chain (Sekaly et a l . , 1988). On the other hand, supertransfection by the invariant chain gene of mouse L cells, which spontaneously express little invariant chain, improves their ability to present exogenous antigen in association with class I1 molecules (Koch, personal communication). Because the invariant chain carries an amphipathic a-helical sequence, and because there is some evidence that such sequences bind MHC molecules (cf. Section VII,B), it has been proposed that the invariant chain prevents the loading of endogenous peptides on class I1 molecules (Elliott et al., 1987). Ultrastructural studies of gold-labeled antigens taken up by APCs are compatible with their
MHC-ANTIGEN INTERACTION
125
uptake and intracellular processing in a matter of less than 2-4 h (Lin et a l . , 1988). Watts and Davidson (1988) have documented the rapid recycling of antigen by human B cells. There are cases in which exogenous antigens have been shown to be presented by class I molecules. Yewdell et al. (1988) have shown that cells could process exogenous proteins (in their case, noninfectious influenza virus) for presentation by class I molecules. Staerz et al. (1987) have shown that immunization with soluble ovalbumin could induce class I-restricted CTLs. Also, De Libero et al. (1988) have found that immunization with mycobacteria could elicit a class I restricted CTL response against bacterial proteins. 2 . Endogenous Antigens
Endogenous antigens, a pnori, fall into at least two classes, those which are directed to the cell surface or secreted, and those which remain within the cytosol or the nucleus. The former penetrate the endoplasmic reticulum and are, therefore, in the proper topological configuration to meet, at some stage, de novo synthesized and/or recycling MHC molecules. The latter should somehow cross a membrane to be loaded on MHC molecules by a process which is not understood yet. Both types of endogenous antigens are indeed presented as illustrated by the class I-mediated presentation of endogenously synthesized influenza nucleoprotein and hemagglutinin (Townsend et al., 1986a,b; Braciale et a l . , 1987a). Similarly, an ovalbumin-secretingcell could be recognized by CTLs, implying presentation of ovalbumin peptides by class I molecules (Staerz et al., 1987). Quite interestingly, when the influenza hemagglutinin gene was engineered in such a way as to lack the signal sequence (Townsend et al., 1986b) or the transmembrane anchor sequence (Braciale et al., 1987a), many CTL clones directed against mouse cells expressing the membrane anchored viral protein still recognized cells transfected with the mutant genes. CTL clones directed against the transmembrane region were also found. Antigen processing in the cytoplasm, as judged from CTL recognition, can be highly efficient. For example, Hosaka et al. (1988) studied the presentation of influenza proteins delivered to the target cell by inactivated, albeit fusion-competent, virus. They found that both internal and membrane proteins could be presented, but the former were better presented. In a separate study (Yewdell et al., 1988), they also found that processing and presentation of cytoplasmic proteins could take place in the absence of protein synthesis. Infection by vaccinia virus(es) expressing Kd and/or influenza hemagglutinin or nucleoprotein leads to Kd-restricted presentation of the viral proteins, but with different requirements, which possibly reflect differences in the presentation pathways (Andrew et al., 1987).
126
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
3. Catabolism of Proteins and Processing
of Intracellular Antigens When exogenous antigen is endocytosed, part of it must reach some cellular compartment(s), not yet precisely identified, where it undergoes degradation. But what is the source of endogenously synthesized, processed antigen, with a nuclear or cytoplasmic intracellular localization? Since many cell types can present and therefore process such intracellular antigens (e.g., to CTLs), it is likely that their (partial) degradation uses standard elements of the cell machineries. In other words, the pathway of presentation of such antigens would be plugged on the degradation pathway(s) of the cell. The life-time of cellular proteins can be vastly different, being possibly related to the nature of the N-terminal residue (Bachmair et al., 1986; note that the generality of this N-end rule is not established), to the N-terminal sequence (Yen et al., 1988), or to the presence of PEST sequences (Dice, 1987). Proteins, even when longlived, are not immortal and are catabolized. Furthermore, protein synthesis is not a very accurate process, making numerous errors which produce altered proteins (see review by Buckingham and Grosjean, 1986, and further discussion in Section X,B). Escherichia coli and other bacteria have devices to get rid of these altered products. A similar situation could exist in eukaryotic cells, where ubiquitin could play an important role (reviewed by Rechsteiner, 1987; Townsend, 1987). I. PATHWAYS OF ANTIGEN PROCESSING AND PRESENTATION BY CLASS I OR CLASSI1 MOLECULES It would make biological sense if, as suggested by several authors (Claverie and Kourilsky, 1986; Germain, 1986; Braciale et al., 1987b), endocytosed antigens were preferentially presented by class I1 MHC molecules, while endogenous ones would be preferentially presented by class I MHC molecules. A variant proposal has been that coendocytosis of antigen with class I or class I1 molecules could lead to preferential presentation by the latter (Kourilsky et al., 1987). T helper cells recognize antigen in the context of class I1 molecules, but there exist T helper cells specific for class I molecules (reviewed in Singer et a l . , 1987b). Similarly, CTLs are usually class I restricted, but there exist class I1 restricted CTLs (reviewed in Mills, 1986; Morrison et al., 1986b). Morrison et al. (1986b) have generated CTL clones specific for cells infected by influenza virus and restricted by class I or class 11. Using these clones, they were able to clearly demonstrate the existence of at least two pathways. The endogenous one, involving the processing of molecules synthesized within the cells, led to class I presentation. The exogenous one led to presentation by class I1 molecules. These conclusions have been confirmed by the demonstration that infectious viruses preferentially induce T cell precursors which give rise to class I-restricted
MHC-ANTIGEN INTERACTION
127
CTLs, while noninfectious viruses mostly activate precursors of CD4 class I1 effector T cells (but a small population of class I precursors was also activated) (Morrison et al., 1988). This discrimination, however, cannot be absolute. In the measles system, where the CTL response is largely class I1 restricted, target cells transfected with genes coding for the matrix and nucleocapsid viral proteins are recognized by class 11-restrictedCTLs (Jacobson et al., 1989). Similarly, Bogen and Weiss (1988), have found that a B cell line transfected with the gene coding for a A light chain (A-315)presents a peptide encompassing the third hypervariable region to a proper T helper clone, in association with class I1 molecules (a result which, if generalized, could have important consequences in terms of the functioning of the immune system in relation with the presentation of “idiopeptides”; Jorgensen et al., 1981; Leserman, 1985, 1987; Kourilsky et al., 1987). Conversely, we have mentioned previously (see Section IV,H,l)several examples in which exogenous antigens were presented by class I molecules. Bevan (1987)has discused the question of class I and class I1 discrimination on completely different grounds. He postulated the existence of a category of APCs that would take up large pieces of cellular debris and process them in a pathway which would lead to presentation by class I molecules (endocytosed soluble proteins being shuttled to class 11). +
J. MOLECULAR MECHANISMS Little is presently known about the molecular mechanisms by which antigen is taken up (for the exogenous pathway), processed, and eventually bound by neosynthesized or recycling MHC molecules. Many questions pertinent to immunology overlap the field of basic cellular biology (see commentary by Delovitch et al., 1988). Apart from the matter of antigen-specific receptors (on macrophages, etc.), advances are being made in the identification of cellular structures which handle and process antigen. Thus, Lakey et al. (1987b)have identified a cell surface protein able to bind peptide 81-104 of pigeon cytochrome c, but not the native cytochrome. Antibodies against this nonpolymorphic peptide-binding protein (about 72-74 kDa) block antigen presentation by class I1 molecules. Betancourt et al. (1987) have suggested that a family of peptide-binding molecules exist in the cell membrane, and play a role in the uptake and presentation of peptide antigen. Such data show that binding and competition experiments involving exogenous synthetic peptides and nonpurified MHC molecules should be interpreted with caution. Other membrane proteins, and the membrane, could, for example, attract certain peptides more than others and favor their subsequent uptake by MHC molecules, leading to the false impression that they display a higher affinity for the latter
128
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
(cf. Sections VI,B, C, E, and F and Sections VI1,A and B). Furthermore, under certain conditions, peptides presented by distinct restriction elements may compete with each other for binding to the cell surface at a stage that presumably precedes their selective uptake by class I1 molecules (Lakey et al., 1987b). V. The Questionable Homogeneity of Individual MHC Molecules
A. THEPRESENTATION OF SELF-COMPONENTS There is a long-standing line of immunological thinking according to which self-componentsactively participate in the immune system (e.g., Grabar, 1974). A central concept is that “self’ is learned during ontogeny, such that anything “nonlearned” behaves as ‘‘nonself.”In this scope, the processing and presentation of self-components was postulated by a number of authors. Thus, Benacerraf (1978) suggested that “macrophage Ia molecules could bind macrophage-processedfragments of autologous proteins as well as foreign proteins.” Jorgensen et al. (1981) emphasized that antibodies could be processed and presented by Ia molecules. The presentation by class I1 MHC molecules of self-proteins, circulating in the body fluids or released by dead cells, was relatively easier to figure out than the presentation of self-elements by class I MHC molecules. Several workers trying to decipher alloreactivity (see Section VIII) alluded to the presentation of self-elements. Thus, Matzinger and Bevan (1977) referred to the multiplicity of minor transplantation antigens. Opponents to such views have expressed alternative theories. For example, Lefkovits (1986) postulated the existence of “restriction proteases” which would specifically degrade foreign proteins while sparing self-proteins. SELFMODEL B. THEPEPTIDIC The finding by Townsend et al. (1985, 1986b) that peptides derived from influenza nucleoprotein are presented by class I MHC molecules allowed us, along the same lines, to propose a generalization known as the “peptidic self model” (Kourilsky and Claverie, 1986). We reasoned that cells transfected with and expressing the influenza NP gene could not “know” that the gene is foreign (this would, to a certain extent, suppose the existence of an intracellular immune system). This implied that self-proteins, including nuclear and cytoplasmic ones, were possibly presented. A cell expressing class I MHC molecules (and class I1 molecules?) should thus be coated with a large variety of its own selfpeptides. When infected by a virus, the few additional viral peptides displayed on its surface would signal it to the immune system. This postulate has numerous interesting implications. For example, certain somatic mutations (some of which are possibly associated with a tumoral
MHC-ANTIGEN INTERACTION
129
phenotype) could be seen by the immune system (Kourilsky and Claverie, 1986; Raja,, 1987). Minor transplantation antigens could be peptides of somatic proteins displaying some polymorphism between individuals (Kourilsky et al., 1987). We also speculated that B cells and T cells could present “idiopeptides” derived from their endogenous (rather than exogenous and endocytosed) antibodies or TcRs (Kourilsky and Claverie, 1986; Kourilsky et al., 1987) in association with class I, class 11, or perhaps other presenting molecules (see also Jorgensen et al., 1981; Leserman, 1985, 1987). Implications on alloreactivity (Claverie and Kourilsky, 1986) are discussed later.
c. EXPERIMENTAL EVIDENCE IN FAVOR
PRESENTATION OF SELF-PEPTIDES As logically strong as they may seem to be, these theoretical arguments indeed require experimental support. The following lines of evidence exist: OF THE
1. Experiments by Mitchison’s group involving antibody response against the liver F protein and its variants show that APCs present the self-allelic form (Winchester et al., 1984). 2. The binding of a self-lysozymepeptide to mouse class I1 molecules has been demonstrated (Babbitt et a l . , 1985, 1986). Data with mouse cytochrome c also suggest autologous binding (Lakey et al., 1987a). Adorini et al. (1988a,b) have provided evidence that the mouse lysozyme self-peptide 46-62 would compete in vivo for T cell activation by foreign antigens. 3. Schwann cells present myelin basic protein to T cells upon induction by interferon-y (Wekerle et al., 1986). 4. Recent experiments using hemoglobin variants of the mouse have shown that self-hemoglobin-Ia complexes permanently exist in vivo on a variety of APCs (Lorenz and Allen, 1988). 5 . The analysis of turn- mutants of P815 cells by Boon and his coworkers has led to the important finding that the presentation of a mutant self-protein fragment by class I molecules causes rejection of turnmutants by CTLs (De Plaen et al., 1988). 6. Syngeneic immunoglobulin fragments are presented by class I1 MHC molecules in vitro (Jorgensen and Hannestad, 1982; Weiss and Bogen, 1989; Bikoff et a l . , 1988). 7. When the crystal structure of HLA-A2 was solved, the presumptive peptide-binding site was found to be filled with an uninterpretable cloud of electrons, the overall density of which was comparable to that of the HLA molecule. One interpretation is that HLA molecules in the crystal were loaded with a series of different peptides, possibly selfpeptides (Bjorkman et al., 1987a,b). 8. The product of the Mlsa gene (reviewed in Abe and Hodes, 1988)
130
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
has not been identified yet. Mlsa expression is primarily detectable on B cells. Studies on the reactivity of TcRs containing Vp6 or Vp8.1 can be interpreted by invoking recognition of a self-protein (encoded or regulated by Mls) (an Mls peptide?) in association with class I1 MHC molecules (Kappler et al., 1988; McDonald et al., 1988). 9. TcR-mediated alloreactions specific for given cell types have been described. These results are most easily understood under the assumption that the T cell clones used recognized peptides derived from cell type-specific proteins (Marrack and Kappler, 1988b). 10. Peptides derived from self-classI1 and self-class I MHC molecules have been shown to be presented by MHC molecules of the same individual (Guillet et al., 1986; Song et al., 1988b) (cf. Section VIII on alloreactivity). 11. If MHC molecules are normally associated with peptides, there might exist antibodies specific for MHC-restricted antigen. A few MHCrestricted antibodies have been described (initially an anti-H-Y, HLArestricted antiserum) (van Leeuwen et al., 1979; see also Wylie et a l . , 1982; Froscher and Klinman, 1986) (but others have failed to isolate them, e.g., Kievits et al., 1987b). Recently, we found that monoclonal antibodies seem to react with a small subset of cell surface mouse Dd molecules, perhaps a class of cell peptides bound to Dd (Abastado et al., 1989). (It should indeed be emphasized that antibodies reacting with MHC plus one peptide would experimentally appear as being quite “weak,” since there would be a large variety of peptides on the surface, thus such antibodies may have not been searched properly.) Quite interestingly, a monoclonal antibody reacting with a subpopulation of class I1 MHC molecules has recently been characterized by C. Janeway (personal communication). 12. Finally, it may be recalled that reactions against the minor transplantation antigens are MHC restricted. This is consistent with the notion that at least some of them are presented peptides derived from polymorphic self-proteins. Also, self-tolerance is MHC restricted (Matzinger et al., 1984; Rammensee and Bevan, 1984). In conclusion, a growing body of data suggests that MHC molecules on the cell surface might be coated with a variety of peptides derived mostly from internal self-proteins (for class I), or from internal or internalized self-proteins (for class 11). This association of self-peptides to MHC molecules would be sufficiently strong to be maintained during and after isolation of MHC molecules. A major consequence would then be that, on the cell surface, a given MHC molecule, initially believed to be a homogeneous molecular entity, would, in fact, be coated with a diversity of self-peptides and thereby would be very heterogeneous. There has been and still is some resistance to the concept that self-
MHC-ANTIGEN INTERACTION
131
elements are presented (for example, much of the initial work on pigeon cytochrome c was performed without reference to the mouse homologue). This is partly related to belief that the immune system would be clogged by self and unable to respond to nonself. Indeed, the recent data of -aCobian and Unanue (1988) demonstrate intracellular interference between two antigens upon presentation by murine peritoneal macrophages (discussed in Section V1,E). The price to pay to circumvent this argument is to accept that immune reactions mediated by MHC molecules can be exquisitely sensitive, a small number of peptides (perhaps one; see Section X)being enough to trigger a reaction. This is not clearly established. One should indeed avoid switching from an unproved dominant concept (MHC molecules are homogeneous entities) to a similarly dominant and unproved notion (MHC molecules are permanently coated with a myriad of self-peptides). Nevertheless, in spite of many unknowns, the latter conclusion now appears likely and we take it as being essentially correct, as is discussed below. VI. MHC-Antigen Interactions
In this section we survey what is known about the interactions between MHC molecules and (usually processed) antigen. As mentioned in Section I, much of the available evidence relies on biological assays using T cells. Though experimental designs were gradually refined, particularly with the use of synthetic peptides and T cell clones, direct biochemical data are relatively scarce. General conclusions, therefore, must be carefully evaluated. Central to the discussion is the question of whether, and to what extent, MHC molecules select the peptides which they bind and thus influence the configuration of the bound peptides. This issue is essential for the understanding of MHC restriction. There are more available data for class I1 MHC molecules, which will be discussed first. Comprehensive reviews have been written by Buus et al. (1987b), Unanue and Allen (1987), Allen (1987), McKean (1988), and Jelachich and Biddison (1988) (see also Moller, 1987). A bibliographical analysis can be found in Bjorkman et al. (1987a,b) and Brown et al. (1988) in relation to the interpretation of structural models of class I and class I1 MHC molecules. We focus here on the most recent and main conclusions and on the debatable issues. SYSTEMS INVOLVING A. THEEXPERIMENTAL CLASSI1 MHC MOLECULLS A number of protein antigens have been extensively studied and dissected to the point where synthetic peptides that mimic antigen in specific T cell reactions could be used. A nonexhaustive list includes hen egg white lysozyme, pigeon cytochrome c, ovalbumin, sperm whale
132
PHILIPPE KOURILSKY AND JEAN.MICHEL CLAVERIE
myoglobin; Staphylococcus nuclease, phage X repressor, insulin, and an immunoglobulin light chain (1-315)in the mouse. In humans, antigens from infectious agents have received particular attention, such as proteins from viruses (influenza, hepatitis B, HIV, etc.), parasites (malaria, etc.), and toxins (tetanus toxin). Lists of peptides active with class I1 molecules can be found in Rothbard and Taylor (1988) and Claverie et al. (1988) (see also Livingstone and Fathman, 1987). Recent additional data deal with the T immunization with a synthetic peptide of herpes simplex virus 2 (Watari et al., 1987), the identification of epitopes in horse cytochrome c (Baumhuter et al., 1987a), the localization of epitopes in the a chain of the Torpedo cal~orniuacetylcholine receptor (Yokoi et al., 1987) and on the HIV gp120 envelope protein (Cease et al., 1987), the restriction of an hepatitis B surface antigen peptide by DPw4 (Celis et al., 1988a; Celis and Karr, 1988), the analysis in the mouse of core particles (Milich et al., 1987), the identification of rabies T epitopes (Celis et al., 1988b), the identification of a n HLA-DR-restrictedepitope in the Mycobacterium tuberculosis 19-kDa protein (Lamb et al., 1988), MHC restriction in the mouse of a peptide from tobacco mosaic virus protein (Langton et al., 1988), and further mapping of malaria circumsporozoite epitopes (Sinigaglia et al., 1988). BINDING OF PEPTIDES B. THESELECTIVE BY CLASS I1 MHC MOLECULES In most instances, it was found that exogenous peptides were funtional only when presented by a given class I1 molecule to specific T cells in uilro. When APCs displaying different class I1 molecules were used with the same T cells, the reactivity was lost. This could mean that the peptide presented by MHC-A did not bind to MHC-B, or that it did bind, but was not seen by the T cells. Actually, depending on the experimental stiuations, both explanations were found to be correct. Thus, in many cases, the peptide did not bind to MHC-B, as assessed by competition experiments (using an irrelevant, functional peptide presented by MHC-B) or by direct binding assays. In a number of instances, the peptide was shown to be bound to MHC-B, but was not recognized by the T cells specific for that same peptide presented by MHC-A, imply ing either that the shape of the peptide is influenced by the MHC molecule that presents it, or that the TcR has to recognize polymorphic residues of the presenting MHC molecule, or both. We shall only summarize the critical evidence which supports these very important findings and refer the reader to reviews for more exhaustive surveys (Schwartz, 1986; Unanue and Allen, 1987; Livingstone and Fathman, 1987; Buus et al., 1987a,b; McKean, 1988). Binding of the processed antigen to the MHC molecules can be appreciated in a variety of ways (Ashwell and Schwartz, 1986; Cease et al.,
MHC-ANTIGEN INTERACTION
133
1986a, Fox et al., 1987a,b; Utsunomiya et al., 1988), but direct binding assays between synthetic peptide and class I1 molecules were essential to clarify the interpretation of a large body of experiments using T cell responses as an assay. Babbitt et al. (1985, 1986) first showed by equilibrium dialysis that the lysozyme peptide 46-61 bound to I-Akbut not I-Ad. Then Buus et al. (1986a), also using equilibrium dialysis, found that the ovalbumin peptide 323-339 could bind to I-Ad but not to I-Ed, I-Ak, or I-Ek. Buus et al. (198613) could also separate Ia-peptide complexes by exclusion chromatography. They found that the complexes took a long time to form (on the order of 24 h), but, once formed, were very stable. Other approaches have used biotin labeling of the peptide and physical methods (Watts et al., 1986; Watts and McConnell, 1986; Phillips et al., 1986; reviewed in Watts and McConnell, 1987). Peptides have been cross-linked to purified Ia molecules (Buus et al., 1987a; Luescher et al., 1988). Direct binding assays have been complemented by competition experiments (Rock and Benacerraf, 1983; Lakey et a l . , 1988; Buus and Werdelin, 1986)which turned out to be an efficient approach to extend the binding analyses to a series of variant peptides derived from a core peptide (see below), or to large numbers of distinct peptides. These experiments so far indicate that class I1 molecules bind only one peptide at a time. However, there have been a number of failures in such competition experiments. For example, peptides 48-68 and 118-138 of influenza hemagglutinin did not compete for presentation by I-Ak, as assessed by specific restricted recognition by two T cell clones (Mills et al., 1988). Therefore, although it seems likely that class I1 molecules in most cases bind only one peptide, a generalization might be premature. Competition experiments have also served to extend the correlation between peptide binding to MHC molecules and immunological reactivity (Babbitt et al., 1986; Guillet et al., 1986, 1987). An extensive series of data was generated by Buus et al. (1987a), who have studied the direct binding of 12 peptides to purified I-Ad, I-Ed, I-Ak, and I-Ek MHC molecules. Every peptide could be shown to bind to its functional restriction element. In 9 cases out of 12, binding to the cognate MHC molecule was stronger than to any other MHC tested. However, X repressor peptide 12-26 bound strongly to I-Ed, while its restriction element is I-Ad. This suggests that peptides can be bound without eliciting a T cell response. Such “holes” in the T cell repertoire could be due to tolerance to self under the assumption that the X repressor peptide resembles a self-peptide presented by I-Ed (Guillet et al., 1987; Buus et al., 1987a). These experiments have provided considerable support to the notion of “determinant selection” (Barcinski and Rosenthal, 1978), that is, the hypothesis that class I1 molecules somehow select the part of the antigen
184
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
that will serve as a determinant for T cell recognition. They also support the notion that this selection explains to a certain extent the Ir character of the responding or nonresponding host (see Schwartz, 1986). As a recent example, Adorini et al. (1988b), using lysozyme peptide 107-116 presented by I-Ed, have shown that the semiconservative substitution of arginine for histidine at position 114 has a profound effect on binding, recognition, and immunogenicity of the peptide. On the other hand, it is clear from a number of experiments in which the same peptide binds two distinct restriction elements that it is somehow presented differently (cf. Section V1,C). A recent demonstration has involved a X repressor peptide (12-26) that binds both to I-Ad and I-Ek. A clonal analysis of restricted T cells showed large heterogeneity, and the fine specificities suggest that the configuration of the peptide is altered (Lai et al., 1987). Of course, one should not ignore the body of evidence which indicates that, in addition to determinant selection, repertoire selection also plays a role in T cell responses. For example, in a recent study using pigeon cytochrome c, McElligott et al. (1988) clearly showed that strain differences in T cell repertoires also account for differential responses. Recent results on the T cell response to myoglobin in a high-responder and a low-responder mouse showed equivalent heterogeneity of repertoires in both mice (Gorai st al., 1988). The selectivity of binding, however, may not be as stringent as initially believed. First, the peptide binding studies do correlate with the MHC restriction patterns, but not completely, since several peptides, in the analyses of Buus et al. (1987a), bound immunologically irrelevant MHC molecules. Second, and most importantly, the correlation between these in vitro data and the immunogenicity of peptides in vivo shows major differences. In the lysozyme system, conditions can be found under which, even in the same haplotype, many peptides, in addition to the immunodominant one(s), can be made immunogenic for T helper cells (Gammon et al., 1987; Carbone et al., 1988; Brett et al.,1988). Perkins et al. (1989) have immunized different strains of mice with seven individual “classical” peptides. Depending on the MHC haplotype, proliferative T cell responses in spleens and/or lymph nodes were observed, in agreement with the general notion of MHC restriction. However, all peptides, in one or the other haplotype, could elicit class I- and class 11-restricted T cell responses. Two peptides could stimulate human as well as mouse responses. These results suggest that the selectivity of binding is less stringent than might have been thought. They also emphasize the possible role of the mode of immunization in selecting pathways that lead to presentation by class I or class I1 MHC molecules (6.Section IV,H). In summary, the extent to which binding differences observed in vitro translate into biological differences in vivo remains relatively obscure. It may depend on a variety of parameters, including the mode of
MHC-ANTIGEN INTERACTION
135
immunization (dose of antigen, route of immunization leading to certain types of APCs rather than others, adjuvant, etc.). C. STRUGIVRE-FUNCTION RELATIONSHIPS INVOLVING PEFTIDESAND CLASSI1 MOLECULES 1. Studies with Peptide Analogs
The sequences of a number of well-studied peptides have been systematically varied in such a way as to try and determine which residues are involved in binding to the presenting class I1 molecule (the agretope) and which ones are seen by the TcR (the epitope). Such studies have initially taken advantage of natural variants of the antigen (e.g., insulins, cytochrome c, and lysozymes from different species). Thus, in peptide 81-103 of pigeon cytochrome c, residues 99 and 103 were shown to belong to the agretope (reviewed in Schwartz, 1986). These studies have since been refined by the use of series of analogs (Carbone et a l . , 1987; Fox et al., 1987b). In the lysozyme minimal peptide 52-61, presented by I-Ak, three residues participate in the agretope (52, 58, and 61) and three in the epitope (53, 56, and 57) (Allen et al., 1987). In the ovalbumin peptide 323-339, presented by I-Ad, a core region of seven residues (327-333) is required for peptide I-Ad interaction, through four critical residues (327, 328, 332, and 333). A complex array of residues, including some involved in the agretope, participate in the epitope (Sette et a l . , 1987). Ogasawara et al. (1989) have analyzed peptide 43-58 of pigeon cytochrome c, together with a large series of variants, for presentation by I-Ab and I-Ad and recognition by a degenerate T cell clone which responds to the peptide in association with either I-Ab or I-Ad. They concluded that residues 46 and 56 affected interaction with Ia, residues 50 and 52 were critical for T cell specificity, and residues 47, 48, 49, and 51 were marginally important. Mills et al. (1988) have analyzed in details an influenza hemagglutinin epitope (defined by peptide 54-62) in the H-2khaplotype. Rothbard et al. (1988) have studied the binding of two influenza peptides (18-29 from matrix and 307-318 from hemagglutinin) that are both presented by HLA-DR1. They made the seemingly successful assumption that both peptides would share the same binding site with DR1 and studied a series of hybrid peptides; they were thus able to build a structural model of peptide-DR1 interactions in which bound peptides are in an a-helical conformation. The deduced structure of bound peptides is discussed in Section VII. 2. Studies with Variant Class 11 Molecules Variant class I1 molecules have been obtained from a variety of sources; the available MHC types and subtypes, particularly in man, provide a vast reservoir of variants. In the mouse, where the diversity of
136
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
characterized variants is less, a few mutants have proved useful, in particular the bm12 gene mutation affecting the I-Abmolecule (see below). Other sources include immunoselected mutant cells obtained in vitro, cells transfected with exon-shuffled genes, and cells transfected with genes mutated by site-directed mutagenesis (Buerstedde et a l . , 1988a; Landais et al., 1988). This topic has been reviewed in detail by Glimcher and Griffith (1987) and McKean (1988). Here, we shall leave many questions unanswered, including those pertaining to the choice of the presenting cell lines (discussed in McKean, 1988), the involvement or noninvolvement of the invariant chain (cf. Section IV,F), the preference of certain pairs of a and fl chains over other pairs (Braunstein and Germain, 1987; Lechler, 1988), and the wealth of serologicalstudies performed with these variants. Their allorecognition by alloreactive T cells is surveyed in Section VIII. Variant class I1 molecules have been used in specific assays using defined antigens or synthetic peptides (Allen et al., 1987; Lechler et al., 1986; Ronchese et at., 1987a,b; Rosloniec et al., 1989). The study of mutants immunoselected by antibodies has not been too productive so far, perhaps because the selection is focused upon residues that are accessible to antibodies but, for many of them, irrelevant to the biological specificity of the peptide-binding site. In addition, loss of function is difficult to interpret. Thus, mutations at positions 75 (Griffith et al., 1987) and 69 plus 79 of A-akhad limited effect on the presentation of a lysozyme peptide (Rosloniec et al., 1989), and the mutation at position 49 of E-Pk also had little effect on the function. Cohn et al. (1986) have shown by site-directed mutagenesis that residues 9, 13, and 65-67 of A-pd play some role in antigen presentation. Ronchese et al. (1987b), in a follow-upof previous exon-shufflingand half-exon-shuffling experiments could determine that residue 29 of E-pk is critical to enable the E-pk molecule to present cytochrome c, probably acting at the peptide-binding level. These authors also showed that each of the three substitutions in A-flbm12drastically altered antigen presentation (Ronchese et al., 1987a). A-ak and A - a b genes with single substitutions at many distinct residues were used in conjunction with a lysozyme peptide to identify several critical regions (Rosloniec et al., 1989). Analyses using naturally occurring human class I1 variants have been undertaken in several laboratories (B. Mach, personal communication). These experiments presently provide tentative assignments for the function of several residues of class I1 molecules, primarily on their possible involvement in peptide binding and/or in contacting TcRs. However, in the absence of structural data, their interpretations remain difficult. The use of the model of class I1 structure (Brown et al., 1988) is somewhat hazardous when it comes to pinpointing residues that point “up” or “down” with respect to the presumptive binding cleft. We shall discuss
MHC-ANTIGEN INTERACTION
137
in more detail similar experiments with class I molecules, where a better structural basis is available. D. THEEXPERIMENTAL SYSTEMSINVOLVING CLASSI MHC MOLECULES It is well documented that CTLs kill virus-infected cells in a class I-restricted fashion. Indeed, it must be kept in mind that class 11-restrictedCTLs also exist, and that non-MHC genes also participate in the immunological response to antigens. For example, Plata et al. (1987) have shown that both BALB/c and DBA/2 mice (H-2d)develop CTLs against the env protein of gross munne leukemia virus, but DBA/2, in addition, develops CTLs against the gag protein, while BALB/c does not. Viral proteins have thus been a major and growing source of peptides presented by class I MHC molecules. (For influenza, see reviews by Townsend, 1987; Wraith, 1987.) Another experimental system deals with the presentation of class I MHC molecules by other class I molecules (Maryanski et al., 1986a; Holterman and Engelhard, 1986), which led Maryanski et al. (1986b) to show the restricted presentation of a peptide derived from the human HLA-Cw3 molecules by H-2d mouse cells. Finally, under certain conditions, immunization with soluble antigens has been shown to elicit a CTL response. Thus, antiovalbumin CTL have been described (Staerz et al., 1987). Lists of peptides presented by class I MHC molecules can be found in Rothbard and Taylor (1988) and Claverie et al. (1988). SELECTION OF PEPTIDES E. THE“FUNCTIONAL” BY CLASSI MHC MOLECULES Direct binding of peptides to class I MHC molecules has been difficult to observe. Recently, Song and colleagues detected class I peptides bound by cell surface class I molecules with monoclonal antibodies (E. S. Song, personal communication). Direct biochemical binding assays have just been developed (J.-P. Levy, personal communication). Such assays have not yet been broadly used. Therefore, the majority of data dealing with this question rely on CTL assays and competition experiments. Conclusions similar to that obtained with class I1 molecules tend to prevail: for example, two peptides derived from influenza nucleoprotein (50-63 and 365-380) are restricted by H-2Kk and H-2Db, respectively (Townsend et al., 1986b; Bastin et al., 1987), while peptide 335-349 is quite selectively presented by HLA-37 (McMichael et al.,1986). One peptide derived from influenza matrix protein (55-75) is currently known to be presented by HLA-A2 only (Gotch et al., 1987b). Another epitope from influenza nucleoprotein was defined as peptide 147-161, restricted by Kd (Taylor et al., 1987). It was later refined to 147-158, and it was
138
PHILIPPE KOURILSKY A N D JEAN-MICHEL CLAVERIE
accidently found that the same peptide lacking arginine 156 is 1000-fold more active in the CTL assay (Bodmer et al., 1988). The HLA-Cw3 180-192 peptide is presented only by Kd, not by Dd or other tested mouse molecules (Maryanski et al., 1986b, 1987). A peptide derived from HLA-A2 (56-59) was found to function in certain alloreactive reactions when presented by HLA-Aw69 (Clayberger et al., 1987), and a peptide from Epstein-Barr virus p63 protein can induce specific CTLs (ThorleyLawson and Israelsohn, 1987). Presentation of the Ld peptide 61-85 is H-2k restricted (Song et al., 1988b), and a peptide from the envelope protein of HIV virus is restricted by Dd (Takahashi et al., 1988). A nineamino acid peptide from lymphocyte choriomeningitis virus is restricted by Db(Oldstone et al., 1988). Another peptide from cytomegalovirus also displays restricted presentation (Koszinowski et al., 1987); peptides from HIV gag proteins might be HLA-A2 restricted (Claverie et al., 1988). A peptide from circumsporozoite malaria protein is presented to CTLs by H-2k class I molecules (Kumar et a l . , 1988). Even though the presentation of synthetic peptides by class I MHC molecules nicely correlates with MHC restriction, little is known on the actual selective binding of peptides by MHC molecules. Selectivebinding of peptides to purified HLA and H-2 class I molecules in a direct assay has been observed recently, and binding so far correlates relatively well with restriction patterns, but with exceptions 0. P. Levy, personal communication) (as for class I1 molecules, cf. Section V1,D). Competition experiments have been performed in several systems (Maryanski et al., 1988; Gotch et al., 1988; Bodmer et al., 1989). As for class 11, the results suggest that only a single peptide can bind in the site. They also emphasize the fact that binding, as assessed by competition, is necessary but not sufficient for a peptide to be recognized by CTLs. Presumably tolerance to self is associated with holes in the T cell repertoire for selfpeptides, or peptides closely resembling self. The experiments of Perkins et al. (1989) discussed in Section VI,D must also be kept in mind, as they emphasize possible differences between in uitro experiments and in uiw immunizations. One difference could lie in the generation of class I-peptide complexes. Under the assumptions of the peptidic self-model, it is possible that class I-peptide complex formation in vivo is a highly competitive process, because peptides derived from internal self-proteins would all be candidates for presentation. This internal competition may amplify small affinity differences between peptides and the MHC into large differences at the level of the presented peptides. The situation in that respect could be different from that encountered for presentation by class I1 molecules of endocytosed antigen, because the endocytosis process might favor the exogenous antigen with regard to self-proteins. For example, in B cells carrying a specific antibody, the corresponding antigen is specifically endocytosed. It remains that the uptake of foreign
MHC-ANTIGEN INTERACTION
139
antigens by macrophages and other APCs must take place in competition with self-proteins. This is the “serum albumin paradox” (the concentration of which is so high that it might block all presentation), which has been discussed by us (Kourilsky et al., 1987) and others (Lorenz and Allen, 1988; see also L.evya-Cobian and Unanue, 1988). A solution might be found in the immunization protocols which have often been designed to concentrate the foreign antigen locally. F. STRUCTURE-FUN~ION RELATIONSHIPS INVOLVING PEPTIDES AND CLASS 1 MOLECULES 1. Studies with Peptide Analogs The most extensive study performed so far has involved the peptide 57-68 from the influenza matrix protein that is presented by HLA-A2. A large series of analogs of the minimal peptide (59-68) were analyzed for functional recognition by specific CTLs. The inactive ones were then tested for their binding in a competition assay. The results suggest that residues 60, 64, and 65 participate in the agretope while residues 61, 62, and 63 are recognized by TcRs. Using the crystal structure of HLAA2, Gotch et al. (1988) noted that an a-helical structure of the bound peptide would segregate residues properly. Accordingly, they propose an orientation of the peptide in the HLA-A2 groove. Studies have been performed with the HLA-Cw3 peptide and other peptides derived from HLA or H-2 molecules (Maryanski et a l . , 1988). Recently, Rothbard et al. (1989), following an approach previously used with a human class I1 molecule (Rothbard et al., 1988), made a series of analogs of the nucleoprotein peptide 147-158 (Bodmer et al., 1988). They prepared hybrid peptides between the latter and HLA-Cw3 1670-182, both presented by Kd, and studied interactions with the Kd molecule. They suggest a model of peptide structure, and their results emphasize the possible key role of residue 66 of Kd in interacting with the bound peptide. 2 . Studies with Variant Class I MHC Molecules Variant class I MHC molecules have been obtained in a variety of ways. There exists a large number of naturally occuring variants, particularly among HLA molecules, of which numerous types and subtypes are available (reviewed by Parham et a l . , 1988b). Mouse mutants with mutated class I molecules have been obtained, particularly in the H-2b haplotype, where the Kb molecule was most often affected but mutants in Db and in the H-2d and H-2k haplotypes have also been obtained (reviewed in Nathenson et a l . , 1986). New mutants have been isolated by different procedures. When detected by antibody assays, their frequency of occurrence appears very high (Egoroff and Egoroff, 1988).
140
PHILIPPE KOURILSKY A N D JEAN-MICHEL CLAVERIE
This may provide interesting insights on the generation of mouse class I polymorphism in Vivo, a matter which relates to the question of gene conversion in the mouse class I multigene family (reviewed in Kourilsky, 1986 and Abastado and Kourilsky, 1988). Point mutants of class I molecules have also been isolated by in vilro selection by antibodies or CTLs (Ajitkumar et al., 1988). Several mutants have been made by sitedirected mutagenesis (Cowan et al., 1987; Hogan et d.,1988; McMichael et al., 1988; Salter et al., 1987; Koeller et al., 1987; McLaughlin-Taylor et al., 1988; Jelachich et al., 1988). Finally, chimeric molecules have been constructed using class I-class I1 hybrids (Golding et al., 1985) and chimeric class I molecules produced by exon shuffling (Evans et al., 1982; Murre et al., 1984; Bluestone et al., 1985; Allen et al., 1987; Ozato et al., 1983, 1985; Levy et al., 1985; Arnold et al., 1985; Scholler et al.,1986) and by an intraexon recombination system (known as the snail technique of Abastado et al., 1987a,b; see also Maryanski et al., 1987; Sire et al., 1988). All these variant class I molecules have given rise to a large number of serological studies, which will not be reviewed here (see Darsley et al.,1987; Abastado et al., 1987b, and references therein). Their use in alloreactive reactions is discussed in Section VIII. Exon-shuffling experiments had shown that the a1 and a2 domains of class I molecules carried (most of) the biological specificity, as now visualized in the crystal structure of HLA-A2, where the a3 domain does not directly participate in the peptide-binding site. However, a variety of experiments indicate that the a3 domain and P2-microglobulin(P2m) can exert a certain influence on the peptide-binding site. Thus, mouse P2m cannot functionally replace human &m for the association with certain class I HLA heavy chains, as shown particularly in transgenic mice (Krimpenfort et al., 1987). Similarly, the a3 domain of HLA-A2, when combined with the a1 and a2 domains of Kb, does not induce serologically detectable conformational changes of the latter, but causes decreased allorecognition (Maziarz et al., 1988). Further dissection of class I molecules has attempted to define the function of individual residues, or stretches of residues in binding the peptide and/or in contracting the TcR. The use of the HLA-A2 structure or of its extrapolation to other class I molecules is essential in the interpretation of such experiments. We review below a number of those experiments. Maryanski et al. (1987) have studied the specificity of presentation by the Kd molecule of the HLA-Cw3 and -A24 peptides 180-192 to several Kd-restricted CTL clones. They used a battery of chimeric Kd/Dd molecules, and their study indicated two zones of importance, including the variant Dd/Kd residues 95, 97, 99, 104, 107, and 114 and 152, 155, 156, 173, 174, 177, and 184. The latter region was later refined to 152, 155, and 156 (Maryanski et al., 1989). Healy et al. (1989)
MHC-ANTIGEN INTERACTION
141
similarly generated a series of HLA-A2/-A3 chimeras and tested them with antiinfluenza HLA-A2-restricted CTLs. They concluded that residues 114, 116, and 152 all lying in the groove was important and that these residues are possibly involved in binding the peptide. Koeller et al. (1987) transferred into Ld a stretch of eight Dd residues (63-70) and studied the Ld-restricted recognition of vesicular stomatitis virus (VSV) motifs by CTLs. Recognition was altered. The three Dd/Ld variant residues apparently point into the peptide-binding cleft, Thus, the peptide-binding specificity of the Ld molecule may have been altered. Cowan et al. (1987) and Jelachich et al. (1988) have extensively studied variants of HLA-A3 that differ at positions 152 and 156, for alloreactive recognition (see Section VIII) and recognition for influenza virus CTL clones. By site-directed mutagenesis, they showed that 152 is the important residue, while 156 is less relevant. Both are located in the putative peptide-binding site of HLA-A2. Hogan et al. (1988) have further documented the role of 152 and 156 in HLA-A2 (rather than HLA-AS) by using variants A2-1 and A2-3 that differ by three residues (149, 152, and 156) in the a2domain. By site-directed mutagenesis, they studied the role of each substitution and showed that substitution of 149 had little effect on the recognition of the influenza matrix peptide 57-68 by CTL clones, while substitutions of 152 and 156 induced dramatic changes: some CTL clones completely lost recognition but others did not. Accordingly, it appears likely that 152 and 156 do not dramatically affect peptide binding. Since peptide recognition is altered, and since the crystal structure indicates that they point into the peptidebinding site (Bjorkman et al., 1987a,b), it is possible that changes at positions 152 and 156 affect the conformation of the bound influenza matrix peptide. It is interesting that residue 149 points “up,” and that substitutions at this site are seen by a monoclonal antibody, though they have little effect on recognition by TcRs. McMichael et al. (1988) have performed an extensive analysis of the HLA-A2 peptide-binding site. They made mutants by site-directed mutagenesis at positions 9, 43, 152, and 156 (which are altered in natural A2 variants) and 66, 70, 74, and 107, plus a double 62-63 mutant, and another one with six substitutions between 70 and 80. The mutations represented differences between HLA-A2 and -B7. The variants were tested with influenza matrix peptide 56-68 and a collection of specific CTLs. Residues 43 and 107, which lie outside the putative binding site, had no effect, while 62-63, 66, 152, and 156 strongly reduced or destroyed recognition by all CTL lines tested. Variable results were obtained with 9 and 10. Residue 9 lies in the bottom of the groove, and mutations at this position could affect the orientation or the shape of the peptide,
142
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
which certain TcRs now fail to recognize. Similar results were reached for residue 70, but the substitution at 74 had no effect. In the double mutant 62-63, where 62 points up, mutations at 62 could affect binding to the TcR. In summary, rapid progress is being made in the understanding of peptide-MHC interactions, but more direct binding data and the crystal structure of MHC-peptide complexes will be needed to fully decipher this question. VII. Structural Data on T Cell Epitoper
There is no reason to believe that a peptide derived from a processed antigen displays the same structure as in the native molecule. This indeed limits the use of the atomic coordinates of peptide segments, as they are shaped in native antigens of known crystallographic structure. Nevertheless, a number of hypotheses concerning determinants seen by T cells have been formulated on these grounds (e.g., for insulin; Talmon et al., 1983). Within the average range of sizes (8 to 15 residues) of synthetic T epitopes, free-ended peptides are not believed to have a unique stable conformation in aqueous solutions. Under the assumption that T epitopes might belong to a class of peptides displaying a dominant structure, attempts are being made to characterize their structure by using circular dichroism (Richman and Reese, 1988), NMR, or crystallographic studies. The knowledge of the dominant conformation and of a few interatomic distances in aqueous solution or in a crystal, coupled with the use of computer-assisted molecular graphics, might help building realistic models of MHC class I-peptide interaction. Indeed, the successful cocrystallization of a peptide complexed with an MHC class I or a class I1 molecule (particularly HLA-A2)would allow the determination of the structure of the complex in the bound state. Unfortunately, attempts to obtain sufficient quantities of the homogenous complex involving HLA-A2 and specific peptides have failed so far (T. Jardetzky, personal communication). At present, available data on the structure of T epitopes are thus quite indirect, either inferred from statistical analyses or from the behavior of substituted peptide analogs. Schematic drawings of an antigenic peptide in various conformations as compared to the peptide-binding site (Fig. 2, see page 122) illustrate the significance of the questions raised herein. A. A NUMBER OF PRESENTED PEPTIDES, BUT NOT ALL, MAYDISPLAY A HELICAL CONFORMATION The popular notion that T epitopes tend to adopt an a-helical conformation comes from several lines of evidence, some of them rather
MHC-ANTIGEN INTERACTION
143
weak. The compilation of secondary structure predictions of a number of known T epitopes indicates a propensity toward regular structures rather than disordered ones (Spouge et at., 1987). Their amino acid composition indicates that the proportion of glycines, serines, and prolines (“coil and turn residues”) are statistically less frequent than expected (Spouge et at., 1987; Claverie et at., 1988). This is indeed not sufficient to infer that most T epitopes adopt a canonical a-helical conformation once bound to the presenting class I or class I1 molecule. Experiments involving analogs of peptides and/or MHC molecules (reviewed in Sections VI,C and F) had different outcomes. Initial studies of Pincus et at. (1983) on a pigeon cytochrome c peptide suggested an a-helical conformation. Allen et at. (1987) have engineered a lysozyme peptide by an N-terminal addition that presumably enhances its propensity to adopt an a-helical conformation. The resulting peptide bound the Ia molecule 100- to 1000-foldmore efficiently. Data involving peptides added to APCs or MHC molecules embedded into membranes of living cells should be interpreted with some caution, because the chemical modifications of the peptide may be relevant to the binding of membrane structures other than the MHC molecule. This was, in particular, suggested by Carbone et al. (1987), who have engineered a pigeon cytochrome c peptide (81-104) with an amino-terminal leader sequence to modulate a-helical propensity. Circular dichroism indicated that peptides in solution did have a preferred a-helical conformation, but helixbreaking residues did not affect their recognition by the test T cell clone. Bhayani et at. (1988) have added series of “irrelevant” residues to the 95-104 peptide of pigeon cytochrome c. Certain residues increase antigenic potency in a way that does not correlate with a-helical propensity. Rather, these hydrophobic residues might enhance interaction with the APC membrane. Gotch et at. (1988) have extensively studied variants of an influenza virus matrix peptide presented by HLA-A&,and come to the conclusion that the bound peptide could be in an a-helical configuration. While studying the residues involved in the specific binding to HLA-DR1, Rothbard et at. (1988) concluded in favor of an a-helical bound peptide. A 28-amino acid fragment of the 75K protein of Plasmodiumfalcifirum merozoite displayed a circular dichroism spectra characteristic of an a helix (Richman and Reese, 1988). Conversely, the mapping of residues contacting the MHC and/or TcRs in an ovalbumin peptide (Sette et at., 1987) in two distinct pigeon cytochrome c peptides (Fox et al., 198713; Ogasawara et al., 1989) and a lymphocytic choriomeningitis virus-specific (JTL epitope (Oldstone et al., 1988) has provided evidence for nonhelical structures. D. C. Anderson et al. (1988) systematically studied a variety of analogs to a peptide from Mycobacterium
144
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
leprae 65-kDa protein and found no correlation between reactivity and helical propensity. An HIV gag epitope (K16F: Claverie et al. , 1988) has been studied by circular dichroism, and did not display an a-helical propensity in aqueous solution (C. Abergel, personal communication). In computer-aided modeling of peptides bound to the MHC groove, the epitope is often represented as an a helix (Davis and Bjorkman, 1988; Claverie et al., 1989). This is mostly for convenience, since other irregular conformations could be accommodated. Furthermore, in the HLA-A2 structure (Bjorkman et al., 1987b) the dimension of the groove is such that a-helical peptide sequences containing large side chain residues would not fit (Claverie et al., 1989) (see Fig. 2). Extended or irregular conformation would allow an easy fit and permit interactions with HLAA2 residues pointing from the bottom of the site. It is indeed possible that the MHC molecule displays several conformations and opens up to grasp peptides. PEPTIDES, BUT NOT ALL, B. A NUMBEROF PRESENTED DISPLAY AN AMPHIPATHIC CHARACTER When a sequence folded into an a helix (approximately four residuedturn) exhibits a segregation of the hydrophylic versus hydrophobic residues on opposite “sides” of the helix, it is said to be “amphipathic” (referring in fact to a-helical amphipathy). Because a large fraction of the a-helical segments found in globular proteins are amphipathic, this property is not independent from the a helix propensity (J. Garnier, personal communication). De Lisi and Berzofsky (1985) have proposed that T cell antigenic sites tend to have amphipathic structure. The structure of the HLA-A2 molecule now shows that there is no obvious requirement for a peptide to exhibit a hydrophobic side in order to bind to the presenting MHC molecule, because apolar residues are the minority in and around the class I binding pocket (cf. Fig. 1). Actually, many described T epitopes have no obvious amphipathic character (Claverie et al., 1989). When D. C. Anderson et al. (1988) designed synthetic peptides with an amphipathic structure, some were active, others were not, and there are many examples of active peptides that are not amphipathic. The correlation between the location of amphipathic regions and T epitopes in proteins is difficult to test in a meaningful way without structural data on the peptide (a peptide sequence might look amphipathic albeit not folded as an a helix). The argument that the amphipathic character of a peptide helps binding to the APC membrane (followed by translocation to MHC molecules) rather than direct binding to the MHC can also be made here (as above
MHC-ANTIGEN INTERACTION
145
for the a-helical propensity). Furthermore, there are different ways of computing a numerical amphipathy index. This has generated a certain confusion which we think useful to try and dissipate here. An a-helical amphipathic peptide can be intuitively defined as a succession of highly hydrophilic and highly hydrophobic residues in a way compatible with the four residues-per-turn helical wheel: Ile-Ile-ArgGlu-Val-Leu-Glu-Arg.There are simple computer programs that calculate an index which matches well this biological, intuitive view of amphipathy. This index depends (1) on the proper spacing of hydrophobic versus polar residues and (2) on the hydropathy value associated with each amino acid. According to such a program, the sequence Ala-Ala-Ser-Ser-AlaAla-Ser-Ser, although still exhibiting the proper periodicity, will be counted as much less amphipathic than the previous one because alanine residues are less hydrophobic than isoleucine, and serines are less hydrophilic than arginines. The algorithm used by De Lisi and co-workers (De Lisi and Berzofsky, 1985; Spouge et al., 1987; Margalit et al., 1987; reviewed in Berzofsky et al., 1987) is more mathematically oriented and makes use of a Fourier analysis of the sequence taken as a succession of hydropathy values. If the largest component of this Fourier analysis corresponds to the acceptable a-helical period (100 f 20°), the segment is called amphipathic. By this method, the two above sequences will appear equally amphipathic. Furthermore, the sequence Leu-Thr-Ala-Leu-GlyAla-Ile-Leu, although almost uniformly hydrophobic, will be considered as amphipathic as the two previous ones (Spouge et al., 1987). In fact, totally hydrophobic or hydrophilic peptides displaying the proper oscillation in hydropathy indexes would appear amphipathic in their algorithm. In summary, De Lisi and Berzofsky’sdefinition of amphipathy is broader than its intuitive acceptance by most biochemists and tends to overrepresent “amphipathic” regions in proteins. Discrepancies with more direct, intuitive methods are thus understandable. C. Is
PEPTIDE CONFORMATION MODELED PRESENTING MHC MOLECULE? Is the conformation of the peptide influenced by the presenting molecules? At one extreme, the structure (if any) of the unbound peptide in solution could be irrelevant to that of the bound peptide. An intermediate view, endorsed by a number of protein chemists, is that the peptide probably oscillates between a number of conformations, few of them being suitable for binding to the presenting molecule. The slow binding kinetics (as observed between peptides and Ia molecules by Buus THE
AFTER THE
146
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
et a l . , 1986b) could in part be due to the low probability of productive collisions between MHC molecules and peptides in a rare conformation. Peptide structures might be generated around a central core by the variable conformations of the more flexible extremities. The latter could be more susceptible to shaping by the presenting molecule. Analyses of structure-function relationships between peptides and MHC molecules (surveyed in Section VI) have provided, in several cases, experimental evidence that variant residues of MHC molecules may shape the same peptide differently (e.g., Jelachich et al., 1988; Hogan et al., 1988; McMichael et a l . , 1988).
D. PREDICTION OF THE LOCATION T CELLEPITOPES IN COMPLEX PROTEIN ANTIGENS 1. The Amphipathy Rule The amphipathy rule became popular after De Lisi and Berzofsky (1985), Spouge et al. (1987), and Margalit et al. (1987) analyzed a growing number of peptides with demonstrated T immunogenicity and found that about 70% of them could be called amphipathic. It seems, though, that their broad definition of amphipathy (cf. Section VI1,B) limits its use as an efficient predictive method to locate potential T epitopes in large proteins. When T epitopes are found amphipathic, they rarely correspond to the highest “peaks” of amphipathy, but often have an average index, so that the method leads to rather loose predictions. In addition, there is a growing list of nonamphipathic, nonhelical T epitopes. In summary, even if T immunogenicity for a subclass of peptides correlates with ct amphipathy (as well as with ct-helical propensity), this property appears too common to serve as a precise predictive criterion. It does seem to be valuable on a statistical basis, but it is conceivable that, once the amphipathy rule became popular, more amphipathic peptides were synthesized, some of which were active, thus feeding the list of active amphipathic peptides, even though amphipathy may not be a compulsory character of T cell epitopes. 2 . Search for General Consensus Sequence Patterns Rothbard and Taylor (1988) analyzed a set of 57 T epitopes. Regardless of the MHC restriction elements, they derived a consensus pattern valid for 80% of them. This consensus can be summarized as (charged + Gly) (hydrophobic) (hydrophobic + Pro) (polar + Gly). It is fairly degenerate and allows a considerable number of variations (on a purely statistical ground, this pattern is expected to occur every 30 residues). Even so, a number of T epitopes do not contain it (Claverie et al., 1988). The
MHC-ANTIGEN INTERACTION
147
analysis of several peptides which bind the mouse I-Ad molecule has shown that all (except for one) carry a core sequence compatible with this consensus, but the consensus within the core sequence could be altered without loss of binding (Sette et al., 1988). We feel that more data are needed to decide whether this positive correlation can be turned into a bona fide predictive method. When different peptides are presented by a given MHC molecule, they may share sequence patterns which reflect spatial arrangements of key residues (or charges) required (or compatible with) for their binding to the same site (now assumed to be unique; see Section VI1,C). Such common features can in turn be used as a predictive tool to design new peptides exhibiting the same presentation specificity (Rothbard et a l . , 1988). 3. Search f o r Consensus Sequence Patterns Spec@ for Given MHC Molecules
Results of Guillet et al. (1987) showing a certain degree of sequence homology between peptides presented by I-Ed and the I-Ed molecule itself made it possible that individual MHC molecules would bind peptides sharing some homology with their own sequence. With a growing list of presented peptides, there seems to be only a proportion (a subclass?) that display “significant” homology with MHC sequences. Rothbard and Taylor (1988) have noted subpatterns which emerged when epitopes were segregated according to their restriction element. Such a classification should be limited to the binding properties of peptides, since certain peptides bind but are not recognized by T cells. Sette et al. (1987) identified a core region of six or seven residues in the ovalbumin peptide 323-339. They compared it to other peptides which strongly bind I-Ad and derived a common motif which they showed to be relevant by testing series of analogs (Sette et a l . , 1988). 4 . The “Rarity” Rule
Unlike the previous rules, this rule does not try to translate the physicochemical constraints involved in the binding of peptides to a given MHC molecule. Rather, this “immunological” rule aims at pointing out peptides for which reactive TcRs have the best chance not to have been deleted, by tolerance to self in a given host. Following the principle of the peptidic self-model (Kourilsky and Claverie, 1986), we reasoned that T epitopes recognized as foreign might contain less self-related sequence patterns. For a few organisms for which a sufficient number of sequences have been determined (i.e., presently, man and, to a lesser extent, mouse), a catalog of tetrapeptides found so far in known protein sequences can
148
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
be built. The nonoccurrence of a given tetrapeptide in the catalog is taken as an indication of its relative rarity in this host. The sequence in which one tries to locate T epitopes is then analyzed as a set of overlapping tetrapeptides. An accumulation of nonoccurring tetrapeptides is taken to designate the best candidates. In 63 peptides of proved T immunogenicity in mouse or human, there was a statistically significant bias for sequences that contain the longest stretches of nonoccurring tetrapeptides. The method was applied to the gag protein sequence of HIV-1, and successfully predicted at least two T epitopes (Claverie et al., 1988). This method can be complemented by the use of other physicochemical rules. For instance, given the statistical amino acid composition of known T epitopes, one might choose to discard candidate peptides containing too many glycines, prolines, or serines, and to favor those containing alanine, charged, and aromatic residues. The limitations of this “rarity” rule, in particular those which relate to the size of the sequence data banks, have been discussed elsewhere (Chalufour et al., 1987; Claverie et al., 1988). It may be emphasized that a single residue difference between a foreign and self sequence (e.g., in the lysozyme peptide 46-61) may show as the nonoccurrence of several overlapping tetrapeptides. Therefore, stretches of nonoccurring tetrapeptides do not necessarily signal a long sequence unrelated to self. They may also pinpoint single residues tested several times in the four overlapping tetrapeptides. VIII. Alloreactivity
For quite a while, it was generally accepted that individual MHC molecules were homogeneous entities. It was thus logical to believe that the study of alloreactive reactions mediated by T cells would provide useful and reliable information on the interaction between TcRs and MHC molecules. The high frequency of T cells responding to foreign MHC molecules (e.g., Beretta et al., 1986) was mysterious, and diverse explanations have been proposed (Matzinger and Bevan, 1977; Bevan, 1984). As discussed above (Section II,D), there are good reasons to believe that MHC molecules on the cell surface, being permanently associated with a variety of self-peptides, are in fact very heterogeneous. Accordingly, we earlier reasoned that many experimental data dealing with alloreactivity could reflect recognition, by TcRs, of peptides carried by allo-MHC molecules, rather than of allo-MHC molecules themselves (Claverie and Kourilsky, 1986). Indeed, TcRs could recognize both, weakening the provocative view according to which TcRs might not even
MHC-ANTIGEN INTERACTION
149
recognize haplotype-specific residues of MHC molecules. We discuss below the possibility that the generic notion of alloreactivity encompasses several types of alloreactive reactions.
A. VARIOUSPOSSIBLE TYPES OF ALLOREACTIVE REACTIONS We distinguish three possible nonmutually exclusive types of alloreactive situations. 1. Type I Alloreactiuity
The alloreactive T cell sees self-peptides presented by the allo-MHC molecule (Claverie and Kourilsky, 1986; Werdelin, 1987a). The subset of self-peptides presented by a given MHC is assumed (according to the peptidic self-model)to be haplotype dependent because of peptide selection and/or shaping by MHC molecules. It is therefore understandable that immunization of an organism with cells displaying a distinct MHC yields a strong reaction against the many distinct self-peptides that are presented. This is reminiscent of, but somewhat different from, the proposal by Matzinger and Bevan (1977) that alloreactive T cells see a multiplicity of minor transplantation antigens. The latter are believed to be self-proteins displaying some genetic polymorphism (or presented peptides derived from them). Instead, type I alloreactivity would be due to the presentation of nonpolymorphic molecules by the polymorphic MHC antigens. Type I alloreactivity could explain why T cell clones specific for a foreign peptide presented by a given MHC molecule sometimes react against cells with another haplotype: this might reflect cross-reactionswith self-peptides presented in that haplotype. (A simple calculation shows that this is not unrealistic: assuming that up to 10% of specific T cell clones display alloreactivity against other MHCs, and that these other MHCs each display some lo4 distinct self-peptides, a probability of cross-reaction of lo5 per peptide would be sufficient.) Indeed, in type I alloreactivity, TcRs could react with self-peptides presented by the allo-MHC plus polymorphic or nonpolymorphic residues of the allo-MHC (Fig. 3, p. 123; see also Table I). 2 . Type II Alloreactivity
The alloreactive T cells see allo-MHCpeptides presented by allo-MHC molecules. Here again, the TcRs could react with allo-MHC peptides plus residues of the allo-MHC. Unless there is preferential or exclusive presentation of self-MHC peptides, type I1 alloreactivity may be considered as a particular case of type I MHC molecules yielding self-peptides as any other self-molecule.
150
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
TABLE I RESIDUES OF T W O a!-HELIXES OF THE HLA-2 MOLECULE^ TcR
HLA First a helix Glu 58 Asp 61 Gly 62* Arg 65* Lys 68 Lys Ala Gln Thr
66* 69 72
0 chain CDRZ (around Cly 55)
CDRl (around His 32)
73
Second a helix Ala 150 His 151 Glu 155 Ala 158 Gly 162 Thr 163* Clu 166
a!
chain
CDRZ (around Ser 58)
CDRl (around Thr 31)
“The residues could potentially interact with corresponding variable regions of T cell receptors according to the hypothetical model presented in Fig. 3, p. 123. Asterisk denotes polymorphic locations in human class I MHC molecules.
3 . Type 111 Alloreactivity
The alloreactive T cell sees haplotype-specificmotifs of the allo-MHC molecule alone, irrespective of the content of the peptide-binding groove. For example, it could recognize a conformational determinant formed by polymorphic residues of empty MHC molecules, which might constitute a portion of MHC antigens. It could also recognize MHC molecules loaded with peptides, and thereby distorted.
B. EVIDENCE FOR
TYPES OF ALLQREACTIVITY Experiments involving class I1 molecules are difficult to interpret in the absence of firm structural data. For example, the bm12 mutation in A-pb involves three substitutions (67, 70, and 71), which have all been found important in alloreactive recognition (Donovan et al., 1987; THE VARIOUS
MHC-ANTIGEN INTERACTION
151
Ronchese et a l . , 1987a). A series of mutants made by the introduction, in the A-pb chain, of the k allele residues at positions 9, 13, and 65-67 displayed alterations in allorecognition. Landais et al. (1988) have converted glutamic acid 75 in A-akto 15 alternative amino acids, and found that the alloreactive rewsponse of 10 hybridomas was reduced. In the class I1 model of Brown et al. (1988), this residue is claimed to point up and would be a candidate for contacting TcRs. The series of mutants generated by Buerstedde et al. (1988a) was used with a panel of alloreactive T cell hybridomas. The results emphasize the importance of combination of residues rather than residues alone, and their interpretation is accordingly complex, with two regions possibly involved in the specificity of allorecognition (McKean, 1988). A study of the crossreactivity of T helper clones specific for pigeon cytochrome c has led Matis et al. (1987b) to suggest that MHC-restricted recognition and allorecognition represent differences in the affinity of the MHC-TcR interaction. While such affinity differences may exist, the possiblity that allorecognition involved one or several unknown self-peptides was not taken into account. Not ignoring the importance of this work, we prefer to focus our discussion upon alloreactivity related to class I molecules (reviewed by Forman, 1987), because the structural data on HLA-A2 provide a better framework for less hazardous interpretations. 1. q p e I Alloreactivity
Evidence for type I alloreactive reactions is so far indirect. It is rather a logical inference based on (1) the growing body of evidence suggesting presentation of self-peptides(reviewed in Section V,C) and (2) the evidence indicating peptide selection (and shaping?) by MHC molecules (see Section V). Using a set of chimeric mouse class I molecules made by recombination between Kd and Dd sequences (Abastado et al., 1987a,b), a region important for Kd specificity with respect to presentation of foreign peptides was mapped to residues 152, 155, and 156 (Maryanski et al., 1987, 1989). It is interesting that the same stretch of residues was found critical for recognition of most anti-Kd allo-CTL clones studies (Maryanski et a l . , 1989). This suggests (though does not prove) identical mechanisms for recognition of foreign antigens and allorecognition, which would imply (self?) peptides in the latter, but until further dissection of the residues involved, it is also possible that one or more MHC residues interact with the TcRs carried by the majority of CTL clones. The same three positions (152, 155, and 156) are altered in the Kbml mutant. Site-directed mutagenesis has shown that 152 and 155-156 (either one or both) were important for the recognition by two subsets
152
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
of allo-CTLs (McLaughlin-Taylor et al., 1988). It may be emphasized that the very existence of certain Kbm mutants (reviewed by Nathenson et al., 1986) strongly suggests type I alloreactivity. For example, the bm6 mutation at the bottom of the groove, where it could hardly contact a TcR, has not been revealed by antibodies and yet can be recognized by alloreactive T cells. The bm mutants actually map nicely the presumptive peptide-binding site (Bjorkman et al., 1987b; see Fig. l),and it may be argued that their isolation required that the peptide-binding specificity be modified in order to trigger graft rejection by an alloreactive reaction. Nevertheless, two Kb mutants altered at position 80 and obtained by in vitro selection with antibodies were recently found to display as deep and complex changes in allorecognition as do mutants (Lewis et al., 1988). (Position 80 in HLA-A2 points into the groove.) Along similar lines, Vogel et al. (1988) have performed an elegant analysis of the H-2 Kh2 mutation that encodes three clustered amino acid substitutions at positions 95, 98, and 99, which lie at the bottom of the peptide-binding cleft of the Khz molecule. The binding that these residues strongly influence allorecognition by CTLs is best explained by type I alloreactivity. Salter et al. (1987) have studied substitutions that differentiate HLAAw68-1 from HLA-A2-1. The replacement of glycine 107 in Aw68-1 by tryptophan 107 from A2-1 yielded a molecule which was recognized by several anti-A2-1 alloreactive CTL clones. So did replacements at positions 95 and 97. The latter residues lie in the floor of the peptide-binding site. They are presumably not accessible to TcRs (they have not, so far, been seen by antibodies) and presumably alter the properties of the binding pocket. Residue 107 lies on a loop outside the site, and may be seen directly by TcRs, unless it indirectly affects peptide binding. The analysis of the HLA-A2.4 subtype indicates a role for residues 9 and 99, both on the floor of the groove (Domenech et al., 1988). In a study of allorecognitionof HLA-B27 variants by CTL clones, the critical relevance of residue 152 and residues in the 77-81 region has been noted (Aparicio et al., 1988). In HLA-A2, all these residues point into the peptide-binding groove. In their study of the HLA-A3 molecule, Cowan et al. (1987) and Jelachich et al. (1988) showed that residue 152 plays a critical role both for restricted recognition and for allorecognition (15 out of 16 alloreactive clones failed to recognize the substituted molecule). A substitution at position 156 affected some, but not all, CTL clones. Residues 152 and 156 in the HLA-A2 crystal point into the peptide-binding groove and not “up” to the TcR. This and the correlation between restricted and allorecognition favor recognition of peptides by alloreactive T cells. Healy et al. (1989) have generated a series of HLA-A2/-A3 chimeric
ern
MHC-ANTIGEN INTERACTION
153
molecules and have studied their allorecognition by a battery of CTL lines or clones specific for HLA-A2 or -AS. Again, they found that residues 152 and 114-116, essential for restricted recognition of influenza antigens, are also essential for recognition by most allo-CTL. Residues 114-116 lie in a @pleated sheet at the bottom of the groove, and 152 belongs to an a helix and points into the site. Thus, 114-116 and 152 probably play a role in peptide binding and 114-116 could hardly be seen by TcRs. Their data, therefore, support type I alloreactivity. Interestingly, they also found that the well-conserved residue 161 is important in allorecognition and proposed that it could stabilize the structure of the second a helix. The substitution of the as domain of Kb by that of HLA-A2 leads to complete loss of recognition by a subset of anti-Kb-alloreactiveclones (rather than decreased lysis by all clones). It was concluded that the asdomain influences conformational determinants of the a1and a2 domains (i.e., their ability to bind certain peptides) (Maziarz et al., 1988). Finally, the genes coding for the a and /3 chains of an ovalbumin peptide recognizing TcRs have been isolated and transfected into a Tcr- cell, where they are expressed. It was found that both the ovalbumin specificity and allorecognition of I-Aswere transferred, demonstrating that the latter is most probably due to cross-reaction (Malissen et al., 1988). 2. TyPe 11 Alloreactivity
Guillet et al. (1986) in their studies of presentation by mouse Ia molecules first showed that autologous class I1 peptides were presented by class I1 molecules. Because they noted a certain degree of homology between the amino acid sequence of several foreign presented peptides and the sequence of a segment of class I1 molecules themselves, they proposed a model in which class I1 molecules would bear a selfcomplementary pocket that would open to bind foreign peptides with the appropriate (analogous) structure (Guillet et al., 1986, 1987). Singh et al. (1986) have shown that a peptide corresponding to residues 61-69 of the Kb molecule triggers in Vitro proliferation of H-2k allogenic T cells if presented by syngeneic APCs. The reaction was blocked by anticlass I1 mAbs, suggesting that the class I peptide was presented by a class I1 molecule. Conversely, Shinohara et al. (1986) had shown the recognition of I-Ak by Kb-restricted CTLs. The observation that mouse class I molecules can present peptides derived from an HLA molecule (Maryanski et al., 1986b; Pala et al., 1988) opened the possibility that class I peptides can be presented by self-class I molecules. Maryanski et al. (1988) later showed that mouse class I peptides derived from the same region (170-182) could compete
154
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
with HLA peptides for recognition by anti-HLA CTLs, suggesting (but not proving) that they might be presented. Clayberger et al. (1987) have provided direct evidence for what is called here type I1 alloreactivity by generating a human CTL clone which recognizes a peptide common to HLA-A2 and -B17 (56-69), only when presented by HLA-Aw69. Similarly, Song et al. (1988b) showed that a peptide (61-85) from the Ld molecule could be presented in the H-2k context, and could be recognized in an MHC-restricted fashion by antiLd CTLs generated in H-2k mice. 3 . Type 111 Alloreactivity
The numerous publications which were formerly interpreted in support of type I11 alloreactivity now need reinterpretation taking into account one additional feature, previously ignored, namely, the plausible presentation of self-peptides or MHC peptides. This indeed does not imply that alloreactive reactions of type I11 do not exist. Ajitkumar et al. (1988) have reacted a panel of Kb mutant molecules, many of which were generated by immunoselectionwith antibodies, with a battery of allo-CTL clones. They conclude that TcRs recognize a relatively large surface, including polymorphic residues lying on both a helixes. We feel, however, that this conclusion is debatable, because the authors neglected the possibility that some of their Kb mutants could have lost the ability to bind peptides and were, therefore, not functional while, in others, the effect on the mutations could be indirect, for example, by altering the specificity of peptide binding. Nevertheless, if there were empty MHC molecules, such data could provide support for type I11 alloreactivity- as would all the evidence that suggests recognition of polymorphic MHC residues by TcRs (for example, the competition experiments in which peptides from HLA and H-2 molecules block recognition: Parham et al., 1987; Oldstone et al., 1988; Mazerolles et al., 1988; see Section IX). Mann et al. (1988) have studied CTLs directed against membrane-bound QlO (the mouse class I QlO product is normally secreted in the serum, but a membrane-bound molecule was synthesized from an engineered gene). The cross-reactivityof the CTLs was analyzed versus Kbml (because Q l O is the likely donor of the bml mutation; Nathenson et al., 1986), Kd, and Ld as well as several derivatives thereof. Their study emphasizes the importance of residues in the 63-73 region (in the a-1 helix) as well as of residues 152 and 155-156. The interpretation remains ambiguous: if QlO does not present peptides, their data support type I11 alloreactivity. If it does, the data argue again for the critical role of certain class I residues in binding and/or shaping peptides, in agreement with data reviewed in Section VI,F on the presentation of foreign peptides.
MHC-ANTIGEN INTERACTION
155
Numerous data show that allorecognition is much more focused on certain areas of the MHC molecules than recognition by antibodies. Most of the recent evidence indicates that alloreactive T cells are focused on the peptide-binding site (see above). The question of type I11 alloreactivity then amounts to that of knowing whether there are empty MHC molecules. There is no available evidence on this point. Even if MHC molecules are usually loaded with self-peptides,it is by no means obvious that the population of MHC molecules is fully saturated. C. Do SEVERAL TYPES OF ALLOREACTIVITY COEXIST? The evidence reviewed above indicates that alloreactive reactions of type I1 exist, those of type I probably exist, and those of type I11 may or may not exist. In trying to appreciate the possible importance of the various types, one is faced with the question of how alloreactive T cells were isolated, and of whether the sampling is significant with respect to the in vivo repertoire. In many experiments, the immunization process was such that the alloreactive clones were selected to obtain the desired response. So were, in most cases, the CTL clones directed against MHC peptides, with one exception: recently, Olson et al. (1989) have shown that the Ld peptide 61-85, in association with D/Ldml class I molecule (a recombinant in which the first 114 residues originate from Dd), is recognized by anti-Ld CTLs in bulk cultures. Peptide 61-80 remained active, but other peptides in Ld, Kd, or Dd displayed no activity (Olson et al., 1989). This experiment seems to indicate that type I1 alloreactivity is an important component of alloreactivity in this case (but the precursor frequency of alloreactive cells was not determined). Alternatively, one could argue that the Ld peptide 61-85 reconstitutes, on the L/Ddml molecule (in which the a1 helix is from Dd), an empty Ld-binding site with both helixes (a1being provided by the peptide). This experiment should then be taken in support of type 111 alloreactivity. Given that other experiments of a similar sort (e.g., the addition of the Kd peptide 170-182 to hybrid Dd/Kd molecules; Maryanski et al., 1988a) have failed, we do not think that this scheme will provide a general explanation for alloreactivity. Instead, we consider unlikely that type I1 alloreactivity is a major source of alloreactive reactions. Measurements of the frequency of mouse CTL precursors recognizing HLA-A2 presented by H-2 molecules showed that such precursors were much less frequent than those of alloreactive ones (Holterman and Engelhard, 1986). In experiments devised to isolate mouse CTLs directed against the 61-85 peptide of Ld, Song et al. (1988a) used procedures suggesting that such CTLs were rare (see also in Olson et al., 1989). Recent data involving transgenic mice may be relevant. In transgenic mice expressing human &-microglobulin
156
PHILIPPE KOURILSKY A N D JEAN-MICHEL CLAVERIE
and HLA-B27, it was found, after virus infection, that HLA-B27 could serve as a restricting element for mouse TcRs (Kievits et al., 1987a). Mouse bulk anti-HLA-B27 CTLs were then elicited and were shown to lyse mouse or human HLA-B27-expressingcells in a non-H-2-restricted way (Kievits et al., 1989). Thus, presentation of HLA-B27 by mouse H-2 did not play a significant role in that instance. The authors suggested that HLA peptides are preferentially presented by HLA. They, however, neglected the hypothesis that mouse self-peptideswere presented by HLAB27. Since human cells can also be lysed, this requires that a significant proportion of human self-peptides are identical to mouse selfpeptides. This is a reasonable assumption, given the 80% homology between the bulk of human and mouse amino acid sequences in current data banks. In similar experiments, H-2-nonrestrictedrecognition of PD1, a porcine MHC (SLA) class I antigen, had previously been reported by Bluestone et al. (1987). We propose a similar interpretation for their data, which would also hold for results obtained in HLA-B7 transgenic mice (Chamberlain et al., 1988). This interpretation agrees with analyses of human and mouse CTL clones directed against HLA molecules expressed on mouse or human cells (e.g., Bernhard et al., 1987, 1988). The evidence supporting type I11 alloreactivity appears rather scarce. In a way, this is not too surprising, if one bases interpretations on the crystal structure of HLA-A2. The latter is striking in that only 2-4 (out of 18) polymorphic residues point up from the a1 and a2 helixes in a position to be seen by the TcRS. In contrast, at least 8-11 residues lying in the peptide-binding site have been implied in allorecognition, and are unlikely to be directly seen by TcRS (Fig. 1A and B). This is strong evidence in favor of type I or I1 alloreactivity. The competition experiments involving synthetic peptides which inhibit clonal or bulk CTLs and/or helper T cells (cf. Section IX) could be taken to support type I11 alloreactivity. However, they could possibly mean that TcRs corecognize self-peptides and/or MHC peptides and one or few polymorphic MHC residues-an issue which we deal with in the next section. In contrast, evidence against type I11 alloreactivity exists. For example, Murray et al. (1988) recently produced and analyzed a mutant of the H-2DP molecule in which tyrosine 27 (conserved in all MHC class I DNA sequenced so far) was replaced with aspargine. Assuming the overall class I molecular structure is conserved between mouse and human, residue 27 lies on the floor of the putative peptide-binding site, forming contact with the a-8 domain and &-microglobulin. This mutant is perfectly well recognized by a panel of anti-H-2Dp monoclonal antibodies, thus suggesting that its overall structure is not disrupted. No change was observed in its ability to present antigen to H-2Dp-
MHC-ANTIGEN INTERACTION
157
restricted lymphocytic choriomeningitis virus (LCMV)-specific CTLs. Yet, its recognition by polyclonal DP-allospecific CTLs was totally disrupted. Given the location of the mutated residue, a direct contact with the TcR can hardly be invoked. These results, therefore, strongly argue against type I11 alloreactivity. Our interpretation would be that the mutation modifies the presentation of many, but not all, H-2Dpselected L cell peptides (type I alloreactivity) or, alternatively, the presentation of a mutated H-2Dp (tyrosine 27) peptide (type I1 alloreactivity). In conclusion, many unknowns remain, and many experimental data (for example, the recent analyses performed with hybrid class I-class I1 molecules: McCluskey et a l . , 1988) are not easily interpreted. We feel that the available evidence favors the view that direct recognition of alloMHC molecules without peptide (either empty MHC molecules or molecules loaded with peptide, the latter not being seen by TcRs) is not likely to be a major contribution to the general phenomenon of alloreactivity. Instead, a unifying explanation can be found in type I alloreactivity, type I1 being a particular case of Type I (since MHC molecules, like other self-proteins,would potentially yield peptides that would occasionally be presented by themselves and/or other MHC molecules). These statements are far from being proved right. Nevertheless, immunologists should be aware that alloreactivity cannot be taken as a simple and easily interpretable standard immunological reaction. IX. Do T c B See Polymorphic (Haplotype-Specific) Residues of MHC Molecules?
We have alluded to this question several times. We have chosen to deal with it separately because it cannot be solved without intruding into another field, the ontogeny of the immune system and, more precisely, thymic education, as we do in the next section. Here, we first review the evidence that argues in favor of, or against the recognition by TcRs of polymorphic residues of MHC molecules (in addition to presented antigen). We would like to stress that the term “polymorphic” residue is convenient, but somewhat ambiguous. In depth, it refers to haplotype-specific residues and, therefore, depends on the number of haplotypes known (a nonpolymorphic residue might become polymorphic as more haplotypes are identified). Furthermore, several authors define polymorphic residues as the most diverse and “spread” subset of variant residues (Bjorkman et al., 1987a,b; Bmwn et al., 1988). A note of caution should thus be given, and we would also like to mention that the emphasis given to polymorphic residues should not obscure the possibly essential role of the nonpolymorphic (conserved, selected?) ones, including, perhaps, in MHC-TcR interactions.
158
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
A. REAPPRAISAL OF THE
QUESTION
Had the question been raised in early 1986, the general consensus would have been that TcRs must recognize polymorphic residues of MHC molecules. This appeared as the compulsory correlate of three phenomena: MHC restriction, alloreactivity, and thymic education. We have reviewed above the question of MHC restriction and showed the likely importance of the selection, by the MHC molecule, of a fragment of processed antigen, as well as of its possible shaping by the MHC molecule. We have also discussed alloreactivity, which apparently provided the most direct evidence for MHC-TCR interaction, but can be reinterpreted differently under the assumption that MHC molecules are permanently associated with self-peptides. Along these lines, we performed an intellectual exercise to decide whether the completely opposite proposal, namely that TcRs do not recognize the polymorphic parts of MHC molecules, was tenable (Claverie and Kourilsky, 1986). This provocative view, independently formulated by Werdelin (1987a) and subsequently taken up by Chain (1987) in a different scope, was useful to challenge, along a previously ignored line, the interpretation of a number of experiments. It was, in fact, tenable to account for MHC restriction and alloreactivity. It did, however, raise difficulties with regard to thymic education experiments (Claverie et al., 1987; Werdelin, 1987b). Since 1986, the crystal structure of HLA-A2, solved by Bjorkman et al. (1987a,b), has provided extremely useful information and a framework to interpret a large number of structure-function studies of families of MHC molecules. In addition, further evidence in favor of thymic education has been obtained (see Section X). The crystal structure of HLA-A2 displays only a few polymorphic residues (two or four) that point up from the a! helixes and would thus be likely to be seen by TcRs (Fig. 1). One could argue that, at 3.5 A resolution, the position of the side chains cannot be definitely assessed. It could also be that the structure of the functional molecule is somewhat different, or that the MHC molecule can “breathe” (especially the a1 helix, which is not linked to the &pleated sheets by a disulfide bridge (Fig. 2). However, the prototype structure of HLA-A2 displays polymorphic residues (particularly near the upper kinks of the a!z helixes) that are commonly recognized by antibodies, suggesting that they are easily accessible (Abastado et al., 1987b; Novotny et al., 1986; and J. Novotny, personal communication). Such residues might indeed be seen by TcRs, unless this is prevented by some presently unknown mechanism. There is evidence that polymorphic MHC residues are seen by TcRs. In the above survey (Section VI) of structure-function studies with variant
MHC-ANTIGEN INTERACTION
159
MHC molecules presenting defined antigens, we mentioned a number of experiments which, to some extent, support this conclusion. One should add to this list a number of data gathered with alloreactive T cells, which are more difficult to interpret (Section VIII), such as the recent experiments by Ajitkumar et al. (1988). In addition, competition experiments point in the same direction. Thus, Parham et al. (1987) showed that the HLA-A2 peptide 98-113 could specifically block the recognition of targets by HLA-A2-specificCTLs; the block occurred by binding to the CTL, not to the target, indicating that the peptide might bind to TcRs. Mazerolles et al. (1988) have found that a peptide derived from a human class I1 molecule was able to block, in an isotype- and haplotype-specific fashion, recognition by T helper cells. Recently, Oldstone et al. (1988) found that the Db peptide 37-52 blocked CTLs directed against an LCMV peptide, which displays homology with the Db sequence but also blocked anti-H-2b-specific CTLs. The homologous Kb peptide, which differs only at position 50, did not block. One interpretation is that the Db peptide competes with the TcR for Db recognition. Song et al. (1988b) have synthesized the Ld, Dq,and Kd 61-85 peptides (corresponding to a portion of the a1 helix), detected their binding on appropriate class I molecule with monoclonal antibodies, and showed that peptides could block a bulk population of anti-H-2Ld CTLs. On the other hand, Siciliano et al. (1986) demonstrated the direct binding of nominal antigen to TcR a-p on MHC-restricted human T cell clones specific for fluorescein-S-isothiocyanatein the absence of MHC. Recently, Fraser and Strominger (1988) showed that solubilized human a-p TcR could bind its ligand in the absence of MHC molecules. This might indicate that at least some TcRs can recognize antigen without MHC. One could also argue that all above structure-function studies could be interpreted differently: one might state that all amino acid substitutions in MHC molecules, instead of altering direct recognition by TcRs, affect the shape if not the selective binding of peptides. It should also be kept in mind that certain substitutionswhich cause loss of recognition might similarly prevent recognition, rather than causing the loss of specific recognition. Finally, the conclusions of the competition experiments might not be fully reliable: added peptides could be processed with unexpected consequences, and the specificity of competition experiments is not always easy to assess, because a number of peptides at high concentrations nonspecifically inhibit T cell reactions 0. I? Levy, personal communication). We acknowledge that such arguments are, in principle, tenable, but require a somewhat dogmatically biased interpretation of the existing data.
160
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
In conclusion, given the recent evidence, we now consider it likely (albeit not definitely established) that TcRs often (but perhaps not always) recognize haplotype-specific residues of MHC molecules, although perhaps a small number of those. We discuss in this and the next section the issues associated with this statement.
B. MODELING OF MHC-ANTIGEN-TcR COMPLEXES In the absence of direct structural data on the trimolecular complex, several groups have attempted to build models of it. In such models, where the only known structure is that of HLA-A2, the TcR is assumed to have an antibody-like conformation initially suggested by Novotny et al. (1986). Thus, Poljak (1987) made the guess that, as in an antibodyantigen complex, a TcR could see a surface of about 20 X 10 A that would include peptides and residues from both helixes. This figure agrees well with that proposed by Ajitkumar et al. (1988) (with the reservations formulated above). Bjorkman and Davis (1988) and Claverie et al. (1989), in agreement with Marrack and Kappler (1988a), all agree that the CDR3-like region of TcRs could be of major importance in interacting with the presented peptide. The major argument here is that most TcR chain variability is found at the functional VDJ region of the 0 and y chains, the equivalent of CDR3 in antibodies. This region, centered around residue 100, might thus preferentially interact with the bound variable peptide. The first and second hypervariable regions (which are much less variable in X R s ) would then be in the proper position to interact with MHC a helixes. In support of this notion, Hedrick et al. (1988) found a striking selection for a sequence around residue 100 of several TcR V, chains expressed by several T cell clones specific for pigeon cytochrome c. These models remain very speculative, and even if grossly correct they might encourage oversimplified views. For example, the peptide-binding cleft in the HLA-A2 crystal appears quite narrow (Fig. 2), possibly suggesting that it functions like a jaw in grasping the peptide. Such possible dynamic features of the molecules are ignored at this stage. Experimental evidence on this question is still scarce. Studies of cytochrome c-specific T cell clones (reviewed in Davis and Bjorkman, 1988; Acha-Orbea et al., 1988) tend to support it. However, the extensive studies performed by Acha-Orbea et d . (1988) on TcRs that recognize myelin basic protein peptide 1-9 in the context of I-Ausuggest that all variable components of a and 0 chains of TcRs are involved in recognition. An increasing number of reports show the preferential usage of a V, or a V, TcR chain in an immune response in vivo: V,17A (Kappler et al., 1987), Vp6 (McDonald et al., 1988), Vpll (Hedrick
MHC-ANTIGEN INTERACTION
161
et al., 1988), V,3 (Tan et al., 1988; Lai et al., 1988), and Vg8 (AchaOrbea et al., 1988; see also Hochgeschwender et al., 1987; Urban et al., 1988). It is not clear whether such preference relates to the recognition of the peptide, of the presenting MHC molecule, or both.
c. SYSTEMATIC VERSUS NONSYSTEMATIC RECOGNITION
HAPLQTYPE-SPECIFIC RESIDUES Assuming that TcRs recognize haplotype-specific residues on MHC molecules, which polymorphic residues are involved? It is now quite probable that TcRs can see only a small number of them. This stems from the HLA-A2 structure (with only two to four residues pointing “up”) and from the guessed structure of TcR, which might not be large enough to embrace a large surface of the MHC molecule (Fig. 3). Furthermore, experiments by several groups have shown that a number of polymorphic residues can be modified without altering recognition (e.g., residue 149 of HLA-A2 according to Hogan et al., 1988). MHC residues closer to the bound peptide might be more frequently corecognized with it. For example, the evidence reviewed above and in the previous sections emphasizes recurrently the probable role of residues in the 152-156 region of class I molecules, which precisely map in the kink of the a2 helix. Some of them could interact with the TcR, but pertinent data are still scarce. An important question then arises: do TcRs carried by T cells raised in the same individual recognize systematically a common set of haplotype-specific residues? This would be the molecular translation of the notion that all TcRs are truly “educated’ to recognize antigen only in the context of the self-restrictingelement. This would imply a stringent geometry of MHC-antigen-TcR interaction and a systematic education of TcRs to always see the same polymorphic residues. The alternative is that some TcRs from the same individual recognize one or a given patch of polymorphic residues, while others would recognize another one or another patch of polymorphic residues. The geometry of recognition would then be loose and, importantly, antigen (peptide) dependent. The seemingly homogeneous restriction pattern of T cell recognition would in fact be heteroclite. It is impossible to answer the question rigorously, but we feel that much of the available evidence favors the possibility that recognition is not systematic. Our main argument is that it often happens that T cell clones specific for a given antigen presented by a given MHC molecule frequently display many distinct TcRs (Manca et al., 1984; Babbitt et al., 1985; Cease et al., 1986b; Baumhuter et al., 1987b). This is compatible with the notion that the same peptide can be seen from different OF
162
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
angles, or in different ways, with different portions of its local MHC environment. A similar point can be made for alloreactive T cells. For example, in recent experiments with chimeric class I H-2 molecules, Maryanski et al. (1989) observed that at least one of three residues (152, 155, and 156) of the Kd molecule is essential for Kd-specific recognition by allo-CTLs. This, however, was not true of all allo-CTL clones. Therefore, either recognition of 152, 155, and 156 is irrelevant to TcR contacts (these residues then being essential in the selection and shaping of self-peptides)or their recognition is not systematic. As discussed above (Section IX,A) we believe that the recent data of Ajitkumar et al. (1988) which have been taken as suggesting recognition of residues on both a helixes, do not contradict this view. More data are obviously needed to settle this point, but one may emphasize the conceptual distance that separates MHC recognition by TcRs, as it was usually described several years ago, from the nonsystematic recognition which can be presently conceived. X. Relevance to the Ontogeny of the T Cell Repertoire
As mentiond above, thymic education provides another line of evidence that supports the notion that T cells somehow “learn” to recognize MHC molecules. It has been taken to imply, in molecular terms, the recognition by TcRs of a constant, antigen-independent set of haplotype-specific MHC residues. In particular, this constant set of MHC residues should never be masked or altered upon binding of any peptide antigen. If TcRs did not recognize polymorphic traits of MHC molecules, but only peptides selected and shaped by them, the ontogeny of the system would be easily understood. Deliberately ignoring the interrelationshipsbetween the B cell and the T cell repertoires (Martinez-A et al., 1988), it would be sufficient to postulate the silencing in the thymus of T cells recognizing self-peptides presented by self-MHC molecules (Kourilsky et al., 1987). Along this line of thinking, however, we have not been able to find a satisfactory explanation (Claverie et al. 1987; Werdelin, 1987b) for experiments which imply positive selection in the thymus (e.g., Singer et al.,1981, 1982), recently further supported by experiments using transgenic mice (Kisielow et al., 1988; Bluthmann et al., 1988; Teh et al., 1988). A survey of the literature on thymic selection is outside of our scope. The reader is referred to other reviews (Sprent and Webb, 1987; Sprent et al., 1988; Saito and Germain, 1988; von Boehmer, 1988; von Boehmer et al., 1988) and particularly to the recent one by Marrack and Kappler (1988a). We shall extract here the few elements useful for our discussion.
MHC-ANTIGEN INTERACTION
163
A. THEPROBLEMATIC IMPLICATIONS OF POSITIVE SELECTION IN THE THYMUS The difficulties associated with positive selection in the thymus have been emphasized by many authors (e.g., Singer et al., 1986). The basic problem is that T cells, which are positively selected to grow in the thymus, have to “learn” to corecognize self-MHC and all foreign antigens which the animal might face. The idea gradually emerged, at least for class I1 molecules (Singer et a l . , 1986), that tolerance was acquired by a mechanism which would somehow inactivate cells recognizing MHC molecules and self-antigens.This idea was incorporated within a scheme whereby T cells, which, in the thymus, have high affinity for MHC plus antigen, are somehow inactivated and destroyed, while those with low affinity are stimulated to grow. The difficulties inherent to this model and other ones (reviewed in Marrack and Kappler, 1988a) would be easily solved if there were two receptors, one for MHC and one for antigen. Otherwise, serious constraints are imposed on TcRs, which must retain good recognition of the MHC while keeping enough diversity to react with a vast number of foreign peptides. One way to solve the difficulty is to postulate segmental specialization of TcRs, as appears possible in models built so far (Section IX,B). Thus, the T cell population could be enriched for TcRs in which the first and second variable regions would preferentially match the MHC, and the diversity of the third hypervariable regions would be used for the recognition of foreign peptides. There are hints that certain TcR chains might have preferred interactions with the MHC (Blackman et al., 1986; Morel et a l . , 1987; Kappler et al., 1987; Saito and Germain, 1987; Matis et al., 1987a,b), but these extreme cases might be more the exception than the rule (Goverman et al., 1985; Garman et a l . , 1986; Iwamoto et al., 1987; Sherman et a l . , 1987; Spinella et a l . , 1987; Baumhuter et al., 1987b; Bogen et al., 1986; Lai et a l . , 1988).
B. A NEWINTERPRETATION OF POSITIVE SELECTION IN THE THYMUS Recently, we proposed a new hypothesis to explain positive selection in the thymus (Kourilsky and Claverie, 1989). This hypothesis is not necessarily exclusive of other ones, particularly the low avidity model mentioned above. However, because our new model may provide a consistent view of MHC recognition by TcRs, we summarize it below. 1. Positive Selection in the Thymus Possibly Means That Reactions Against Foreip Antigens Are Secondary Rather than Primary Reactions Positive selection in the thymus and MHC restriction are entirely consistent with the notion that reactions against foreign antigens in the adult
164
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
organism are secondary reactions rather than primary ones. The primary reactions against foreign antigens would take place during ontogeny of T cells, particularly in the thymus; “MHC learning” would be in fact “antigen plus MHC priming” (i.e., priming by processed antigen presented by MHC). Then, it would be no mystery that the T cell repertoire carries the imprint of the MHC of the thymus of the organism in which it developed. This indeed raises the paradox that the thymus should somehow display a broad set of “functionally foreign” epitopes- albeit harbored by the organism. In a way, this is analogous to the antibody repertoire, which, without having been exposed to foreign antigen, is diverse enough to match about any foreign antigen, even molecules not yet conceived of by chemists. The problem then is to identify this internal source of diversity in the thymus. 2 . Possible Sources of Diversity
Several proposals have been made regarding diversity: 1. As mentioned above, presently the most popular model involves cross-reactionswith self-antigens. T cells with weak affinity for self-antigens presented by the MHC would escape inactivation and grow. Thus, the world of foreign epitopes, in this case, is built by cross-reactionswith self (Singer et al., 1986; Sprent and Webb, 1987; Sprent et al., 1988). 2 . We earlier suggested that mutations in host proteins and idiopeptides encompassing the variable regions of TcRs and antibodies could be driving forces in the ontogeny of the T cell repertoire, both creating internally “foreign” motifs (Kourilsky et al., 1987). This hypothesis is compatible with an analysis of antibody sequences as compared to somatic proteins (Chalufour et al., 1987). At this stage, it remains quite speculative. The questions of whether such idiopeptides exist in general and whether or how they might be presented in the thymus (where a small proportion of B cells is found in addition to T cells) are open. As far as mutations are concerned, rough calculation suggests that mutations in thymic cells are probably not frequent enough to generate appropriate diversity. One possibility, though, would be that a hypermutational mechanism operates in the thymus. 5. A growing body of evidence suggests that circulating T cells in the thymus are activated when recognizing self-components (i.e., selfpeptides presented by the self-MHC) on the thymic epithelium. These activated T cells would then be killed when meeting with immigrant cells derived from the bone marrow, probably including macrophages and dendritic cells (reviewed in Sprent and Webb, 1987; Sprent et al., 1988; von Boehmer, 1988; Marrack and Kappler, 1988a). Spleen
MHC-ANTIGEN INTERACTION
165
dendritic cells, which are very efficient APCs, are also very efficient in inducing tolerance (Matzinger and Guerder, 1989). On this basis, Marrack et al. (1988)proposed that the thymic epithelium would synthesize tissue-specific proteins that are not made by dendritic cells and macrophages. These specific proteins would be the source of epitopes able to cross-react with foreign ones. We feel that this hypothesis is rather unlikely to account for the desired diversity of the T cell repertoire, since only a very small fraction of cell putative self-peptides would not be common to both cell types. The T cell repertoire must indeed be rather extensive, as illustrated by the experiments of Ogasawara et al. (1987), who failed to find holes in it. 3. Translation Errors as a Source of Diversity
We proposed that errors in gene expression (as emphasized to us by Dr. M. Radman) might be a major source of variability and generate enough diversity on cells of the thymic epithelium to drive the development of a large T cell repertoire (Kourilsky and Claverie, 1989). The rate of misincorporation of amino acids in proteins has been estimated to be on the order of 5 X per residue (reviewed in Buckingham and Grosjean, 1986). Premature termination of a polypeptide chain, a particular type of translation mistake, may not be that infrequent (it was shown to occur in uivo in 31% of the E. coli 0-galactosidasemolecules, which are about 1000 amino acids long: Manley, 1976). Translational frameshifts (Craigen and Caskey, 1987) and transcriptional restarts (yielding aberrant RNAs and possible frameshift products) have frequently been observed. These estimates must be taken with caution, because data have been gathered in many different systems, in vivo and in ui'tro, and in bacteria more than in eukaryotic cells. Also, there are local effects due to mRNA structure, so that different rates of error are obtained with different sequences. Nevertheless, it is clear that errors in gene expression are much more frequent than errors in DNA synthesis. Thus, expanding on our previous suggestion on mutations (Kourilsky et al., 1987),we propose that, in addition to mutations, errors in gene expression mostly at the translation level can generate a large diversity of variant peptides at the surface of thymic epithelium as well as at the surface of any other cell. Even without invoking a special mechanism that would stimulate errors in gene expression in the thymic epithelium, we make as a conservative guess a minimal average error rate of 5 X per amino acid residue, not including frameshifts and restarts. Any cell displaying on its surface a total of about lo6 peptides in association with its lo6 MHC molecules would then display several hundred erroneous peptides. In our model, these erroneous self-peptides on the thymic epithelium
166
PHILIPPE KOURILSKY A N D JEAN-MICHEL CLAVERIE
activate young T cells, which then have a very low probability of finding again the same variant peptides on immigrant macrophages and dendritic cells before they mature in the periphery and become refractory to tolerance induction (Matzinger and Guerder, 1989). Such activated T cells would, therefore, escape death and be free to expand (thanks to this primary activation) in the periphery. In contrast, T cells recognizing self-peptides would disappear in this double check. Also, we presume that T cells recognizing empty MHC molecules would similarly be killed, because empty MHC molecules (if any) would be found both on the thymic epithelium and the immigrant macrophages and dendritic cells. Elementary calculations show that millions of variants could easily be generated by errors in gene expression as postulated here: assuming that cells of the thymic epithelium express 50,000 different proteins with an average length of 200 to 300 residues, and that presentable peptides are found every 100 residues, the number of different presented self-peptides would be on the order of lo5.If these self-peptides are 10 amino acids long, with 19 possible substitutions at each position, a total of lo5 X 10 X 19, i.e., on the order of lo7 variants could be generated. This is not an absurd number to match the diversity of foreign epitopes that an individual may face: if, in the average peptide, five residues serve in MHC binding (the agretope) and five serve in recognition by TcRs, a total of 205 = 3.2 X lo6 combinations are generated. This number has to be multiplied by the number of agretopes usable by the MHC of the individual, but, due to their size and to MHC restriction, the number of foreign T epitopes may not be astronomical. Furthermore, the repertoire of T cells may be larger if each erroneous self-peptide can be seen at different angles by different TcRs. It may also be fed by other mechanisms (see above) and, in a sufficiently large repertoire, cross-reactions may play a role. In this model, T cells are activated in the thymus by bona fide highaffinity interactions with variant peptides. Variant peptides are selected (possibly shaped by MHC molecules) and recognized by TcRs in exactly the same way as foreign or self-peptides. What is absolutely required by the model is that, in the thymus, T cells can be activated by a single peptide. The sensitivity of T cell recognition is quite high, and, in certain circumstances, it was estimated that a few hundred peptide molecules on the target are sufficient for CTL recognition. We know of no evidence implying that several presented peptides must cooperate to trigger activation of T cells in the thymus. The confined environment of the thymus, where cellular concentrations are high and diffusion is limited around the thymic epithelium, together with accessory molecules, could facilitate recognition of a single peptide. This is indeed a major issue in our model.
MHC-ANTIGEN INTERACTION
167
Conversely, one should note that if a single peptide is able to activate circulating T cells in the thymus, it is almost compulsory that erroneous self-peptides play a role in the ontogeny of the T cell repertoire. BY TcRs C. THEQUESTION OF DIRECTMHC RECOGNITION MIGHT BE OF LIMITED BIOLOGICAL SIGNIFICANCE To conclude, we wish to emphasize that, in the final analysis, the very question of the recognition by the a-p TcRs of polymorphic residues of MHC molecules may appear to have limited biological significance. The three phenomena which led to emphasis on the question, namely, MHC restriction, alloreactivity, and thymic education, may all receive interpretations in which the answer is somewhat irrelevant to the initial question. We take it as established that MHC molecules are peptide receptors endowed with a loose specificity; even if the discrimination is not absolute, they usually bind certain peptides better than others. Furthermore, to an unknown extent, they influence the configuration of the bound peptide. It is the selection of peptides and their possible shaping and orientation by individual MHC molecules that now appear to be of primary importance. From the available information (with the reservations which we underlined; 6.Section IX,A), it seems that numerous polymorphic residues of MHC molecules are involved in those tasks, while only a small number (two to four, according to the HLA-A2 crystal structure; Bjorkman et al., 1987a,b) may be involved in the interaction with TcRs. The reason for our claim that the latter are not of primary importance is that, whether or not such interactions between polymorphic MHC residues and TcRs take place, it does not change the basic interpretations of MHC restriction, alloreactivity, and thymic education that we have proposed. Furthermore, we suspect that this recognition of haplotype-specific residues by TcRs (which is not yet rigorously established) occurs in a nonsystematic way. By this, we mean that several TcRs may recognize the same peptide in different ways, sometimes corecognizing a few MHC polymorphic residues, sometimes others, and perhaps sometimes none. The reason for this has, in our view, to be searched for in the ontogeny of the T cell repertoire. In our model, TcRs with the highest affinity for variant peptides presented by the MHC have the best chance to make their way during ontogeny. Thus, in a purely selective way, in combinations generated at random, haplotype-specific MHC residues may or may not interfere with the recognition of the bound peptide by the TcR. We therefore think that the question of recognition of polymorphic MHC molecules by TcRs has served as an Ariane’s thread to reach other more important problems. The latter deal with the presentation of foreign, self-, and erroneous peptides and are essential in the
168
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
understanding of MHC restriction, alloreactivity, thymic education, and tolerance. Of course, our views on this issue are largely speculative and future experimental evidence may or may not validate them.
ACKNOWLEDGMENTS We are grateful to many colleagueswho have sent us preprints and unpublished information. We would like to acknowledge the fruitful discussions with Dr. M. Radman and A. Singer that gave rise to some of our thoughts on thymic selection. We are grateful to Dr. 0. Acuto, Dr. L. Bougueleret, Dr. G. Chaouat, Dr. E.-L. Larsson-Sciard, and Dr. C. Rabourdin-Combe for critical reading of the manuscript. We thank Dr. L. Bougueleret and A. Chalufour for providing us with unpublished results and pictures of molecular models. We thank Mrs. K. Zorn for her help and acknowledge the excellent editorial assistance of Mn. V. Caput in the preparation of the manuscript.
REFERENCES Abastado,J. P., and Kourilsky, P. (1988). Gene conversion and the generation of polymorphism in class I genes of the mouse major histocompatibilitycomplex. In “Molecular Biology of HLA Class I1 Antigens” Silver, ed.). CRC Press, Boca Raton, Florida (in press). Abastado, J. P., Darche, S., Godeau, F., Cami, B., and Kourilsky, P. (1987a). Intramolecular recombination between partially homologous sequences in E. coli and Xenopus laevis oocytes. Proc. Natl. Acad. Sci. U.S.A. 84, 6496-6500. Abastado, J. P.,Jaulin, C., Schultze, M.-P., Langlade-Demoyen, P., Plata, F., Ozato, K.,and Kourilsky, P. (198713). Fine mapping of epitopes in intradomain Kd/Dd recombinants. J. Exp. Med. 166, 327-340. Abastado, J. P., Darche, S . , and Kourilsky, P. (1989). A monoclonal antibody recognizes a subset of the mouse major transplantation antigen Kd. Submitted for publication. Abe, R.,and Hodes, R. J. (1988). The Mls system: Non-MHC genes that encode strong T-cell stimulatory determinants. Immunol. Today 9, 230-235. Acha-Orbea, H., Mitchell, D. J., Timmermann, L., Wraith, D. C., Tausch, G. S., Waldor, M. K., Samvil, S. S., McDevitt, H. 0.. andsteinman, L. (1988). Limited heterogeneity of T cell receptors from lymphocytes mediating autoimmune encophalomyelites allows specific immune intervention. Cell (Cambridge, Mass.) 54, 263-273. Adorini, L., Muller, S., Cardenaw, F., Lehmann, P. V., Falcioni, F., and Nagy, Z. A. (1988a). In mu0 competition between self peptides and foreign antigens in T-cell activation. Nature (London) 334, 263-265. Adorini, L., Sette, A., Buus, S., Grey, H. M., Darsley, M., Lehmann, P. V., Doria, G., Nagy, Z. A., and Appella, E. (1988b). Interaction of an immunodominant epitope with Ia molecules in T-cell activation. Proc. Natl. Acad. Sci. U.S.A. 85, 5181-5185. Ajitkumar, P., Geier, S. S., Kesari, K. V., Borriello, F,, Nakagawa, M., Bluestone,J. A., Saper, M. A., Wiley, D. C., and Nathenson, S. G. (1988). Evidence that multiple residues on both the alpha-helices of the class I MHC molecule are simultaneously recognized by the T cell receptor. Cell (Cambridge, Mass.) 54, 47-56. Albertson, D. G.. Fishpool, R., Sherrington, P., Nacheva, E., and Milstein, C. (1988). Sensitive and high resolution in situ hybridization to human chromosomes using
u.
MHC-ANTIGEN INTERACTION
169
biotin labeled probes: Assignment of the human thymocyte CD1 antigen genes to chromosome 1. EMBOJ. 7, 2801-2805. Allen, P. M. (1987). Antigen processing at the molecular level. Immunol. Today 8, 270-273. Allen, P. M., Matsueda, G. R., Evans, R. J., Dunbar, J. B., Marshall, G. R., and Unanue, E. R. (1987). Identification of the T-cell and Ia contact residues of a T-cell antigenic epitope. Nature (London) 327, 713-715. Allison, J., Campbell, I. L., Morahan, G., Mandel, T. E., Harrison, L. C., and Miller, J. F. A. P. (1988). Diabetes in transgenic mice resulting from over-expression of class I histocompatibility molecules in pancreatic beta cells. Nature (London) 333, 529-533. Anderson, D. C., Van Schooten, W. C. A., Barry, M., Janson, A. A. M., and de Vries, R. R. P. (1988). Use of multiple synthetic peptide analogs to study structural features of peptides causing T cell activation. Roc. UCLA Symp. Mol. Cell. Biol. (in press). Anderson, M., McMichael, A., and Peterson, P. A. (1987). Reduced allorecognition of adenovirus-2 infected cells. J. Immunol. 138, 3960-3966. Anderson, P., Morimoto, C., Breitmeyer,J. B., and Schlossman, S. F. (1988). Regulatory interactions between members of the immunoglobulin superfamily. Immunol. Toduy 9, 199-203. Anderson, R. W., Tevethia, M. J., Kalderon, D., Smith, A. E., andTevethia, S. S. (1988). Fine mapping of two distinct antigenic sites on SV40 T antigen reactive with SV40 specific cytotoxic T cell clones using SV40 deletion mutants.J. Virol. 62, 285-296. Andrew, LM. E., Coupar, B. E. H., Boyle, D. B., and Blanden, R. V. (1987). Recognition by major histocompatibility complex class I-restricted cytolytic T lymphocytes of cells expressing vaccinia-encoded viral and class I proteins. Eur J. Immunol. 17, 1515-1518. Aparicio, P.,Jaraguemada, D., Rojo, S., and Lopez de Castro, J. A. (1988). Clonal heterogeneity of HLA-B27 cellular allo-recognition. Delineation of immunodominant sites. Euz J. Immunol. 18, 203-209. Aragnol, D., Malissen, B., Schiff, C., Piron, M. A., and Leserman, L. D. (1986). Endocytosis of MHC molecules by B cell-B lymphoma and B cell-T lymphoma hybrids. J. Immunol. 137, 3347-3353. Arnold, B., Horstmann, U., Kuon, W., Burgert, H.-G., Hammerling, G. J., and Kvist, S. (1985). Alloreactive cytolytic T-cell clones preferentially recognize conformational determinants on histocompatibility antigens: Analysis with genetically engineered hybrid antigens. Proc. Natl. Acad. Sci. U.S.A. 82, 7030-7034. Ashwell, J. D., and Schwartz, R. H. (1986). T-cell recognition of antigen and the Ia molecule as a ternary complex. Nature (London) 320, 176-178. Babbitt, B. P., Allen, P. M., Matsueda, G., Haber, E., andunanue, E. R.(1985). Binding of immunogenic peptides to Ia histocompatibility molecules. Nature (London) 317, 359-361. Babbitt, B. P., Matsueda, G., Haber, E., Unanue, E. R., and Allen, P. M. (1986). Antigenic competition at the level of peptide-Ia binding. Proc. Natl. Acad. Sci. U.S.A. 83, 4509-4513. Bachmair, A., Finley, D., and Varshavsky, A. (1986). In vivo half-life of a protein is a function of its amino-terminal residue. Science 294, 179-186. Bangham, C. R. M., Openshaw, P. J. M., Ball, L. A., King, A. M. G., Wertz, G. W., and Askonas, B. A. (1986). Human and murine cytotoxic T cells specific to respiratory syncytial virus recognize the viral nucleoprotein (N), but not the major glycoprotein (G), expressed by vaccinia virus recombinants. J. Immunol. 137, 3973-3977.
170
PHILIPPE KOURILSKY A N D JEAN-MICHEL CLAVERIE
Barcinski, M. A., and Rosenthal, A. S. (1978). Immune response gene control of determinant selection. I. Intramolecular mapping of the immunogenic sites on insulin recognized by guinea pig T and B cells. J. Ex#. Med. 145, 726-742. Bastin, J., Rothbard, J., Davey, J . , Jones, I . , and Townsend, A. (1987). Use of synthetic peptides of influenza nucleoprotein to define epitopes recognized by class I-restricted cytotoxic T lymphocytes.J. Exfi. Med. 165, 1508-1523. Baumhuter, S. C., Wallace, J. A., Proudfoot, A. E. I., Bron, C., and Corradin, G. (1987a). Multiple T cell antigenic determinants identified within a limited region of the horse cytochrome c molecule. Euz J. Immunol. 17, 651-656. Baumhuter, S., F a d , N . , Corradin, G., and Guardiola, J. (1987b). T cell receptor beta chain gene rearrangements and V-beta gene usage in horse cytochrome c-specific T cell hybridomas. Eur. J. Immunol. 17, 657-661. Benacerraf, B. (1978). A hypothesis to relate the specificity of T lymphocytes and the activity of I region-specificIr genes in macrophages and B 1ymphocytes.J. Immunol. 120, 1809-1812. Bennink, J. R., Yewdell, J. W., and Gerhard, W. (1982). A viral polymerase involved in recognition of influenza virus-infected cells by a cytotoxic T-cell clone. Nature (London) 296, 75-76. Beretta, A., Ermonval, M., and Larsson, E.-L. (1986). Degeneracy of H-2 recognition by cytotoxic T lymphocytes: 10% of the total repertoire is “specific” for a given haplotype and up to 1% is self-H-2 reactive. Euz J. Immunol. 16, 605-609. Bemhard, E. J., Le, A.-X. T., Yannelli, R., Holterman, M. J., Hogan, K. T., Parham, P., and Engelhard, V. H. (1987). The ability of cytotoxic T cells to recognize HLA-A2.1 or HLA-B7 antigens expressed on murine cells correlates with their epitope specificity. J. Immunol. 139. 3614-3621. Bernhard, E. J., Le, A.-T. L., Barbosa, J. A., Lacy, E., and Engelhard, V. H. (1988). Cytotoxic T lymphocytes from HLA-A2 transgenic mice specific for HLA-A2 expressed on human cells. J. Exp. Med. 168, 1157-1162. Berzofsky, J. A. (1986). 14th Forum in Immunology. Ann. Immunol. (Paris) 137D, 297-349. Berzofsky, J. A. (1988). Immunodominance in T lymphocyte recognition. Immunol. Lett. 18, 83-92. Berzofsky, J. A., Cease, K. B., Cornette, J. L., Spouge, J. L., Margalit, H., Berkower, I. J., Good, M. F., Miller, L. H., and De Lisi, C. (1987). Protein antigenic structures recognized by T cells: Potential applications to vaccine design. Immunol. Rev. 98, 9-52. Betancourt, S. V., Solvay, M. J., Irani, D. N., Thomas, D. W., and Nairn, R. (1987). Heterogeneity in cellular antigen retention structures.J. Immunol. 189, 3725-3729. Bevan, J. M. (1984). High determinant density may explain the phenomenon of alloreactivity. Immunol. Today 5, 128-130. Bevan, M. J. (1987). Class discrimination in the world of immunology. Nature (London) 325, 192-194. Bhayani, H., Carbone, F. R., and Paterson, Y. (1988). The activation of pigeon cytochrome c-specificT cell hybridomas by antigenic peptides is influenced by nonnative sequences at the amino terminus of the determinant. J. Immunol. 141, 377-382. Bikoff, E. K., Yu, H., and Eckhardt, L. A. (1988). T cell recognition of endogenous IgG2, expressed in B lymphoma cells. Eur. J. Immunol. 18, 841-348. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S.,Strominger, J. L., and Wiley, D. C. (1987a). Structure of the human class I histocompatibility antigen, HLA-A2. Nature (London) 329, 506-512.
MHC-ANTIGEN INTERACTION
171
Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., and Wiley, D. C. (1987b). The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature (London) 329, 512-518. Blackman, M. A., Yagiie, J., Kubo, R., Gay, D., Coleclough, C., Palmer, E., Kappler, J. W., and Marrack, P. (1986). The T cell repertoire may be biased in favor of MHC recognition. Cell (Cambridge, Mass.) 47, 349-357. Blackman, M. A., Kappler, J. W., and Marrack, P. (1988). T-cell specificity and repertoire. Immunol. Rev. 101, 5-19. Bland, P. (1988). MHC class I1 expression by the gut epithelium. Zmmunol. Today 9, 174-178. Blue, M.-L., Craig, K. A., Anderson, P., Branton, K. R., Jr., and Schlossman, S. F. (1988). Evidence for specific association between class I major histocompatibility antigens and the CD8 molecules of human supressor/cytotoxic cells. Cell (Cambnage, Mass.) 54, 413-421. Bluestone, J. A., Foo, M., Allen, H., Segal, D., and Flavell, R. A. (1985). Allospecific cytolytic T lymphocytes recognize conformational determinants on hybrid mouse transplantation antigens. J. Ex$. Med. 162, 268-281. Bluestone, J. A., Pescovitz, M. D., Frels, W. I., Singer, D. S., and Hodes, R. J. (1987). Cytotoxic T lymphocyte recognition of a xenogeneic major histocompatibility complex antigen expressed in transgenic mice. Eul: J. Immunol. 17, 1035-1041. Blum, J. S., and Cresswell, P. (1988). Role for intracellular proteases in the processing and transport of class I1 HLA antigens. Roc. Natl. Acad. Sci. U S A . 85, 3975-3979. Bluthmann, H., Kisielow, F’., Uematsu, U., Malissen, M., Krimpenfort, P., Berns, A., von Boehmer, H., and Steinmetz, M. (1988). T-cell-specificdeletion of T-cell receptor transgenes allows functional rearrangement of endogenous alpha- and beta-genes. Nature (London) 334, 156-159. Bodmer, H. C., Pemberton, R. M., Rothbard, J. B., and Askonas, B. A. (1989). Enhanced recognition of a modified peptide antigen by cytotoxic T cells specific for influenza nucleoprotein. Cell (Cambraage, Mass.) 52, 253-258. Bodmer, H. C., Bastin, J. M., Askonas, B. A., and Townsend, A. R. M. (1988). Competition between unrelated peptides for presentation to influenza specific cytotoxic T lymphocytes. Immunology (in press). 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, New York. Bogen, B., Snodgrass, R., Briand, J. P., and Hannestad, K. (1986). Synthetic peptides and beta-chain gene rearrangements reveal a diversified T cell repertoire for a lambda light chain third hypervariable region. Eul: J. Zmmunol. 16, 1379-1384. Bond, J. S., and Butler, P. E. (1987). Intracellular proteases. Annu. Rev. Biochem. 56, 333-364. Bonnefoy, J. Y., Guillot, O., Spits, H., Blanchard, D., Ishizaka, K., and Banchereau, J. (1988). The low-affinity receptor for IgE (CD23) on B lymphocytes is spatially associated with HLA-DR antigens. J. Exp. Med. 167, 57-72. Boog. C. J. P., Boes, J., and Melief, C. J. M. (1988a). Stimulation with dendritic cells decreases or obviates the CD4+ helper cell requirement in cytotoxic T lymphocyte responses. Euz J. Immunol. 18, 219-223. Boog, C. J. P., Boes, J., and Melief, C. J. M. (1988b). Role of dendritic cells in the regulation of class I restricted cytotoxic T lymphocyte responses.J. Zmmunol. 140, 3331-3337. Bos, J. D., and Kapsenberg, M. L. (1986). The skin immune system. Its cellular constituents and their interactions. Immunol. Today 7 , 235-240.
172
PHILIPPE KOURILSKY AND JEAN.MICHEL CLAVERIE
Bougueleret, L., and Claverie,J. M. (1987). Variability analysis of the human and mouse T-cell receptor beta-chains. Immunogenetics 26, 304-308. Braakman, E., Rotteveel, F. T. M., Van Bleek, G . , Van Seventer, G. A., and Lucas, C. J. (1987). Are MHC class 11-restricted cytotoxic T lymphocytes important? Immunol. Today 8, 265-267. Braciale, T. J . , Braciale, V. L., Winkler, M., Stroynowski, I., Hood, L., Sambrook, J., and Gething, M.-J. (1987a). On the role of the transmembrane anchor sequence of influenza hemagglutinin in target cell recognition by class I MHC-restricted, hemagglutinin-specific cytolytic T lymphocytes.J. Exp. Med. 166, 678-692. Braciale, T. J., Morrison, L. A., Sweetser, M. T., Sambrook, J., Gething, M. J., and Braciale, V. L. (1987b). Antigen presentation pathways to class I and class I1 MHCrestricted T lymphocytes. Immunol. Rev. 88, 95-114. Bradbury, A., Belt, K. T., Neri, T. M., Milstein, C., and Calabi, E. (1988). Mouse CDl is distinct from and co-existswith TL in the same thymus. EMBOJ. 7, 3081-3086. Braunstein, N. S., and Germain, R. N. (1987). Allele-specific control of I molecule surface expression and conformation: Implications for a general model of Ia structure-function relationships. Proc. Natl. Acad. Sci. US.A . 84, 2921-2925. Brett, S. J., Cease, K. B., and Berzofsky, J. A. (1988). Influences of antigen processing on the expression of the T cell repertoire: Evidence for MHC-specific hindering structures of the products of processing. J. Ex@. Med. 168, 357-374. Brown, J. H., Jardetzky, T., Saper, M. A., Samraoui, B., Bjorkman, P.J., and Wiley, D. C. (1988). A hypothetical model of the foreign antigen binding site of class I1 histocompatibility molecules. Nature (London) 332, 845-850. Buckingham, R. H., and Grosjean, H. (1986). The accuracy of messenger RNA: Transfer RNA recognition. In “Accuracy in Molecular Biology” (D. J. Galas, T. B. L. Kirkwood, and R. Rosenberger, eds.), pp. 83-126. Chapman & Hall, London and New York. Buerstedde, J. M., Pease, L. R., Bell, M. P., Nilson, A. E., Buerstedde, G., Murphy, D., and McKean, D. J. (1988a). Identification of an immunodominant region on the I-A beta-chain using site-directedmutagenesis and DNA-mediatedgene transfer. J. Exp. Med. 167, 473-487. Buerstedde, J. M., Pease, L. R., Nilson, A. E., Bell, M. P., Chase, C., Buerstedde, G., and McKean, D. J. (1988b). Regulation of murine MHC class I1 molecule expression. Identification of A-beta rtesidues responsible for allele-specific cell surface expression. J. Exp. Med. 167, 823-837. Burget, H. G., and Kvist, S. (1987). The E3/19K protein of adenovirus type 2 binds to the domains of histocompatibility antigens required for CTL recognition. EMBO J. 6, 2019-2026. Burget, H. G., Maryanski, J. L., and Kvist, S. (1987). “E3/19K’ protein of adenovirus type 2 inhibits lysis of cytolytic T lymphocytes by blocking cell surface expression of histocompatibility class I antigens. Proc. Natl. Acad. Sci. U.S.A. 84, 1356-1360. Bushkin, Y.,Demaria, S., Le,J., and Schwab, R. (1988). Physical association between the CD8 and HLA class I molecules on the surface of activated human T lymphocytes. Proc. Natl. Acad. Sci. U.S.A. 85, 3985-3989. Buus, S., and Werdelin, 0.(1986). Oligopeptide antigens of the angiotensin lineage compete for presentation by paraformaldehyde-treated accessory cells to T cells. J. Immunol. 136, 459-465. Buus, S., Colon, S., Smith, C., Freed, J. H., Miles, C., and Grey, H. M. (1986a). Interaction between a “processed’ ovalbumin peptide and Ia molecules. Proc. Natl. Acad. Sci. U.S.A. 83, 3968-3971.
MHC-ANTIGEN INTERACTION
173
Buus, S.,Sette, A., Colon, S. M., Jenis, D. M., and Grey, H. M. (1986b). Isolation and characterization of antigen-la complexes involved in T cell recognition. Cell (Cambridge, Mass.) 47, 1071-1077. Buus, S., Sette, A., Colon, S. M., Miles, C., and Grey, H. M. (1987a). The relation between major histocompatibility complex (MHC) restriction and the capacity of Ia to bind immunogenic peptides. Science 235, 1353-1358. Buus, S.,Sette, A., and Grey, H. M. (1987b). The interaction between protein-derived immunogenic peptides and I,. Immunol. Rev. 98, 115-141. Cambier, J. C.,Newell, M. K., Justement, L. B., McGuire, J. C., Leach, K. L., and Chen, Z. Z. (1987). Ia binding ligands and cAMPR stimulate nuclear translocation of PKC in B lymphocytes. Nature (London) 327, 629-632. Campbell, D. H., and Garvey, J. S. (1963). Nature of retained antigen and its role in immune mechanisms. Adv. Zmmunol. 3, 261-313. Carbone, F. R., Fox, B. S., Schwartz, R. H., and Paterson, Y. (1987). The use of hydrophobic alpha-helix-defined peptides in delineating the T cell determinant for pigeon cytochrome c. J. Zmmunol. 138, 1838-1844. Carbone, F. R., Moore, M. W., Sheil, J. M., and Bevan, M. J. (1988). Induction of cytotoxic T lymphocytes by primary in Vitro stimulation with peptides.J. Ex$. Med. 167, 1767-1779. Casten, L. A., Kaumaya, P., and Pierce, S. (1988). Enhanced T cell responses to antigenic peptides targeted to B cell surface Ig; la, 'or class I molecu1es.j Ex#. Med.-l68, 171-180. Cease, K. B., Buckenmeyer, G., Berkower, I., York-Jolley,J., and Berzofsky,J. A. (1986a). Immunologically relevant peptide antigen exists on the presenting cell in a manner accessible to macromolecules in solution. J. Ex$. Med. 164, 1440-1454. Cease, K. B., Berkower, I., York-Jolley, J., and Berzofsky, J. A. (1986b). T cell clones specific for an amphipathic alpha-helical region of sperm whale myoglobin show differing fine specificities for synthetic peptides. A multiviewhingle structure interpretation of immunodominance. J. Ex$. Med. 164, 1779-1784. Cease, K. B., Margalit, H., Cornette, J. L., Putney, S. D., Robey, W. G., Ouyang, C., Streicher, H. Z., Fischinger, P.J., Gallo, R. C.,De Lisi, C., and Berzofsky, J. A. (1987). Helper T-cell antigenic site identification in the acquired immunodeficiency syndrome virus gp120 envelope protein and induction of immunity in mice to the native protein using a 16-residuesynthetic peptide. Roc . Natl. Acad. Sci. U.S.A. 84, 4249-4253. Celis, E., and Karr, R. (1988). T-cell recognition of a major epitope of hepatitis B surface antigen is restricted by HLA-DPw4. J. Immunol. (submitted for publication). Celis, E., Ou, D., and Otvos, L., Jr. (1988a). Recognition of hepatitis B surface antigen by human T lymphocytes. Proliferative and cytotoxic responses to a major antigenic determinant defined by synthetic peptides. J. Immunol. 140, 1808-1815. Celis, E., Karr, R. W., Dietzschold, B., Wunner, W. H., and Koprowski, H. (1988b). Genetic restriction and fine specificity of human T cell clones reactive with rabies virus. J. Immunol. 141, 2721-2728. Chain, B. M. (1987). Does the T-cell receptor bind to the MHC? Nature (London) 329, 677-678. Chalufour, A., Bougueleret, L., Claverie, J.-M. and Kourilsky, P. (1987). Rare sequence motifs are common constituents of hypervariable antibody regions. Ann. Inst. Pasteur (Pans) 138, 671-685. Chamberlain, J. W., Nolan, J. A., Gromkowski, S. H., Kelley, K. A., Eisenstadt, J. M.. Henup, K.,Janeway, C. A., Jr., and Weissman, S. M. (1988). Cellsurfaceexpression
174
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
and alloantigenic function of a human class I MHC heavy chain gene (HLA-B2) in transgenic mice. J. Immunol. 140, 1285-1292. Chamn, D. J., Lotteau, V., Hermans, P., Burroughs, D., Turrnel, P., Faille, A., Ju, L. Y., and Teyton, L. (1988). Hybrid HLA class I1 molecules: Expression and regulation. In “Molecular Biology of HLA Class I1 Antigens” 0. Silver, ed.). CRC Press, Boca Raton, Florida (in press). Claverie, J.-M., and Kourilsky, P. (1986). The peptidic self model: A reassessment of the role of the major histocompatibility complex molecules in the restriction of the T-cell response. Ann. Inst. Pasteur (Paris) 137D,425-442. Claverie, J.-M., Chaouat, G., Rabourdin-Combe, C., and Kourilsky, P. (1987). T cells recognize antigen alone and not MHC molecules. Immunol. Today 8 , 202. Claverie, J.-M., Kourilsky, P., Langlade-Demoyden, P., Chalufour-Prochnicka, A., Dadaglio, G., Tekaia, F., Plata, F., and Bougueleret, L. (1988). T-immunogenic peptides are constituted of rare sequence patterns. Use in the identification of T-epitopes in the human immunodeficiency virus gag protein. Eur. J. Immunol.
18, 1547-1553.
Claverie,J. -M., Prochnicka-Chalufour, A., and Bougueleret, L. (1989). Immunological implications of a Fab-like structure for the T-cell receptor. Immunol. Today 10, 10-14.
Clayberger, C., Holmes, N., Wang, P. L., Koller, T. D., Parham, P., and Krensky, A. M. (1985). Determinants recognized by human cytotoxic T cells on a natural hybrid class I HLA molecule. J. Ex$. Med. 162, 1709-1714. Clayberger, C.,Parham, P., Rothbard, J., Ludwig, D. S.,Schoolnik, G. K., and Krensky, A. M. (1987). HLA-A2peptides can regulate cytolysis by human allogeneic T lymphocytes. Nature (London) 330, 763-765. Cleven, H., Alarcon, B., Wileman, T., and Terhont, C. (1988). The T cell receptorKD3 complex: A dynamic protein ensemble. Annu. Rev. Immunol. 6,629-662. Cohn, L. E., Glimcher, L. H., Waldmann, R. A,, Smith, J. A,, Ben-Nun, A., Seidman, J. G., and Choi, E. (1986). Identification of functional regions on the I-Ab molecule by site-directed mutagenesis. Proc. Natl. Acad. Sci. U.S.A. 83, 747-751. Cooper, H. M., Corradin, G., and Paterson, Y.(1988). The heme moiety of cytochrome c is an autoreactive Ir gene-restricted T cell epit0pe.J. Ex$. Med. 168, 1127-1143. Cowan, E. P.,Jelachich, M. L.,Coligan, J. E., and Biddison, W. E. (1987). Site-directed mutagenesis of an HLA-A3 gene identifies amino acid 152 as crucial for major histocompatibility complex-restricted and alloreactive cytotoxic T-lymphocyte recognition. Proc. Natl. Acad. Sci. U.S.A. 84, 5014-5018. Craigen, W. J., and Caskey, C. T. (1987). Translational frameshifting: Where will it stop? Cell (Cambridge, Mass.) 50, 1-2. Cresswell, P. (1985). Intracellular class I1 HLA antigens are accessible to transferrinneuraminidase conjugates internalized by receptor-mediated endocytosis. Proc. Natl. Acad. Scz. U.S.A. 82, 8188-8192. Danley, M. J., Takahashi, H., Macchi, M. J., Frelinger, J. A., Ozato, K., and Appella. E. (1987). New family of exon-shuffled recombinant genes reveals extensive interdomain interactions in class I histocompatibility antigens and identifies residues invo1ved.J. Ex$. Med. 165, 211-222. Dausset, J. (1958). Iso-leuco-anticorps. Actu Haematol. 20, 156-166. Davis, M. M., and Bjorkman, P. A. (1988). T-cell antigen receptor gene and T-cell recognition. Nature (London) 334, 395-401. De Lisi, C., and Berzofsky, J. A. (1985). T-cell antigenic sites tend to be amphipathic structures. Proc. Natl. Acad. Sci. U.S.A. 82, 7048-7052.
MHC-ANTIGEN INTERACTION
175
Delovitch, T. L., Semple, J. W., and Phillips, M. L. (1988). Influence of antigen processing on immune responsiveness. Immunol. Zbday 9, 216-218. Dembic, Z., Haas, W., Weiss, S., McCubrey, J.. Kiefer, H., von Boehmer, H., and Steinmeu, M. (1986). Transfer of specificity by murine alpha and beta T-cell receptor genes. Nature (London) 320, 232-238. De Plaen, 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 of the gene of tum- antigen P91A and identification of the tum- mutation. Proc. Natl. Acad. Sci. U.S.A. 85, 2274-2278. Dice, J. F. (1987). Molecular determinants of protein half-lives in eukaryotic cells. FASEBJ. 1, 349-357. De Libero, G., Flesch, I., and Kaufmann, S. H. E. (1988). Mycobacteria-reactive Lyt-2+ T cell lines. Eur J. Immunol. 18, 59-66. Domenech, N., Ezquerra, A., Castano, R., and Lopez de Castro, J. A. (1988). Structural analysis of HLA-A2.4 functional variant KNE. Implications for the mapping of HLA-A2-specific T-cell epitopes. Immunogenetics 27, 196-202. Donovan, K. A,, Singh, S. K., and David, C. S. (1987). Mutation in the A-beta gene of B6.C-H-Zbm1* generates unique T-cell recognition sites. Immunogenetics 26, 375-377. Dorn, A., Fehling, H. J . , Koch, W., Le Meur, M., Gerliiger, P., Benoist, C., and Mathis, D. (1988). B-cell control region at the 5' end of a major histocompatibility complex class I1 gene: sequence and factors. Mol. Cell. Biol. 8, 3975-3987. Doyle, C., and Strominger, J. L. (1987). Interaction between CD4 and class I1 MHC molecules mediates cell adhesion. Nature (London) 330, 256-259. Dustin, M. L., Staunton, D. E., and Springer, T. A. (1988). Supergene families meet in the immune system. Immunol. Zbday 9, 213-215. Edidin, M. (1988). Function by association? MHC antigens and membrane receptor complexes. Immunol. Today 9, 218-219. Egorov, I. K., and Egorov, 0. S. (1988). Detection of new MHC mutations in mice by skin grafting, tumor transplantation, and monoclonal antibodies: A comparison. Genetics 118, 287-298. Ehrlich, R., Lifshitz, R., Pescovitz, M. D., Rudikoff, S.,andsinger, D. S. (1987). Tissuespecific expression and structure of a divergent member of a class I MHC gene family. J. Immunol. 139, 593-602. Eisenlohr, L. C., Gerhard, W., and Hackett, C. (1988). Acid-induced conformational modification of the hemagglutin molecule alters interaction of influenza virus with antigen presenting cells. J. Immunol. 141, 1870-1876. Elliott, W. L., Stille, C. J., Thomas, L. J., and Humphreys, R. E. (1987). An hypothesis on the binding of an amphipathic alpha helical sequence in Ii to the desetope of class I1 antigens. J. Immunol. 138, 2949-2952. Evans, G. A., Margulies, D. H., Shykind, B., Seidman, J. G., and Ozato, K. (1982). Exon shuffling: Mapping polymorphic determinants on hybrid mouse transplantation antigens. Nature (London) 300, 755-757. Falo, L. D., Jr., Benacerraf, B., and Rock, K. L. (1986). Phospholipase treatment of accessory cells that have been exposed to antigen selectively inhibits antigen-specific Ia-restricted, but not allospecific, stimulatiowof T lymphocytes. Proc. Natl. Acad. S C ~ U.S.A. . 83, 6994-6997. Falo, L. D., Jr., Benacerraf, B., Rothstein, L., and Rock, K. L. (1987a). Cerulenin is a poten; inhibitor of antigen processing by antigen-presenting cells. J. Immunol. 139, 3918-3923.
176
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
Falo, L. D., Jr., Haber, S. I., Herrmann, S., Benacerraf, B., and Rock, K. L. (1987b). Characterization of antigen association with accessory cells: Specific removal of processed antigens from the cell surface by phospholipases. Proc. Natl. Acad. Sci. U.S.A. 84, 522-526. Fehlmann, M., Peyron, J. F.. Samson, M., Van Obberghen, E., Brandenburg, D., and Brossette, N. (1985). Molecular assocation between major histocompatibilitycomplex class I antigens and insulin receptors in mouse liver membranes. Roc. Natl. Acad. Sci. U.S.A.82, 663443637. and Festenstein, H. (1988). A novel HLA class I1 moleFernandez, N., Labeta, M. 0.. cule. Immunogenetics 27, 363-369. Fischer-Lindahl, K., Loveland, B. E., and Richards, C. S. (1987). The end of H-2. In “Major Histocompatibility Genes and their Roles in Immune Function” (C. S. David, ed.), pp. 327-340. Plenum, New York. Flavell, R. A., Allen, H., Burkly, L. C., Sherman, D. H., Waneck, G. L., and Widera, G. (1986). Molecular biology of the H-2 histocompatibility complex. Science 223, 437-441. Forman, J. (1987). Determinants on major histocompatibility complex class I molecules recognized by cytotoxic T lymphocytes. Adu. Immunol. 41, 135-179. Fox, B. S., Quill, H., Carlson, L., and Schwartz, R. H. (1987a). Quantitative analysis of the T cell response to antigen and planar membranes containing purified Ia molecules. J. Immunol. 138, 3367-3374. Fox, B. S.,Chen, C., Fraga, E., French, C. A., Singh, B., andSchwartz, R. H. (1987b). Functionally distinct agretopic and epitopic sites. Analysis of the dominant T cell determinant of moth and pigeon cytochrome c with the use of synthetic peptide antigens. J. Immunol. 139, 1578-1588. Fox, B. S., Carbone, F. R., Germain, R. N., Paterson. Y., andschwartz, R. H. (1988). Processing of a minimal antigenic peptide alters its interaction with MHC molecules. Nature (London) 331, 538-540. Franzel, R. E. (1962). Immunogenic sub-cellular particles obtained from spleens of antigen-injected mice. Nature (London) 195, 457-458. Fraser, J. D., and Strominger, J. L. (1988). A solubilized T-cell receptor from a human leukemia cell line binds to a ligand in the absence of MHC products. Immunogenetics 28, 108-116. Frei, K., Siepl, C., Groscurth, P., Bodmer, S.,Schwerdel, C., and Fontana, A. (1987). Antigen presentation and tumor cytotoxicity by interferon-gamma treated microglial cells. Eur. J. Immunol. 17, 1271-1278. Froscher, B. G., and Klinman, N. R. (1986). Immunization with SV40-transformedcells yields mainly MHC-restricted monoclonal antibodies. J. Exp. Med. 164, 196-210. Gabert, J., Langlet, C., Zamoyska, R., Parnes, J. R.,Schmitt-Verhulst, A. -M., and Malissen, B. (1987). Reconstitution of MHC class I specificity by transfer to the T cell receptor and Lyt-2 genes. Cell (Cambridge, Mass.) 50, 545-554. Gambon, F., Kreisler, M., and Diaz-Espada, F. (1988). Correlated expression of surface antigens in human thymocytes. Evidence of class I HLA modulation in thymic maturation. Eur J. Immunol. 18, 153-159. Gammon, G., Shastri, N., Cogswell, J., Wilbur, S., Sadegh-Nasseri, S., Krzych, U., Miller, A., and Sercarz, E. (1987). The choice of T-cell epitopes utilized on a protein antigen depends on multiple factors distant from, as well as at, the determinant site. Immunol. Rev. 98, 53-73. Garman, R. D., KO, J. L., Vulpe, C. D., and Raulet, D. H. (1986). T-cell receptor variable
MHC-ANTIGEN INTERACTION
177
region gene usage in T-cell populations. Roc. Natl. Acad. Sci. U.S.A. 83, 3987-3991. Gaspari, A. A., and Katz, S. I. (1988). Induction and functional characterization of class I1 MHC (Ia) antigens on murine keratinocytes.J. Immunol. 140, 2956-2963. Gay, D., Buus, S., Pasternak,J., Kappler,J., and Marrack, P. (1988). The T-cell accessory molecule CD4 recognizes a monomorphic determinant on isolated Ia. Proc. Natl. Acad. Sci. U.S.A. 85, 5629-5633. Cell, P. G. H., and Benacerraf, B. (1959). Studies on hypersensitivity. 11. Delayed hypersensitivity to denatured proteins in guinea-pig. Immunology 2, 65-70. Geraghty, D. E.,Koller, B. H., and Orr,H. T. (1987).A human major histocompatibility complex class I gene that encodes a protein with a shortened cytoplasmic segment. Proc. Natl. Acad. Sci. U.S.A. 84, 9145-9149. Germain, R. N. (1986). The ins and outs of antigen processing and presentation. Nature (London) 322, 687-689. Glimcher, L. H., and Griffith, I. J. (1987). Mutations of class I1 MHC molecules. Immunol. Today 8, 274-279. Golding, H., McCluskey, J , , Munitz, T. I., Germain, R. N., Margulies, D. H., and Singer, A. (1985). T-cell recognition of a chimaeric class II/class I MHC molecule and the role of L3T4. Nature (London) 317, 425-427. Gorai, I., Aihara, M., Bixler, G. S., Jr., Atassi, M. Z., Walden, P., and Klein, J. (1988). T cell response to myoglobin: a comparison of T cell clones in high-responder and low-responder mice. Eur. J Immunol. 18, 1329-1335. Gorer, F’. A. (1937). The genetic and antigenic basis of tumor transp1antation.J. Pathol. Bacteriol. 44, 691. Gotch, F., McMichael, A., Smith, G., and Moss, B. (1987a). Identification of viral molecules recognized by influenza-specifichuman cytotoxic T 1ymphocytes.J Ex@. Med. 165, 408-416. Gotch, F., Rothbard, J., Howland, K., Townsend, A., and McMichael, A. (1987b). Cytotoxic T lymphocytes recognize a fragment of influenza virus matrix protein in association with HLA-A2. Nature (London) 326, 881-882. Gotch, F.,McMichael, A., and Rothbard, J. (1988). Recognition of influenza A matrix protein by HLA A2 restricted cytotoxic T lymphocytes. 2. A model for the orientation of the peptide. J. Ex@. Med. 168, 2045-2057. Goverman, J., Minard, K.,Shastri, N., Hunkapiller, T., Hamburg, D., Sercarz, E., and Hood, L. (1985). Rearranged beta T cell receptor genes in a helper T cell clone specific for lysozyme: no correlation between V-beta and MHC restriction. Cell (Cambricige, Mass.) 40, 859-867. Grabar, P. (1974). “Self’ and “Not-self’ in immunology. Lancet 7870, 1320-1322. Grey, H. M., and Chesnut, R. (1985). Antigen processing and presentation to T-cells. Immunol. Today 6 , 101-106. Griffith, I. J.. Choi, E. M., and Glimcher, L. H. (1987). A single base mutation in an I-A-alpha chain gene alters T-cell recognition. Proc. Natl. Acad. Sci. 84, 1090-1093. Guillemot, F., Billault, A., Pourquie, O., Behar, C., Chausse, A.-M., Zoorob, R., Kriebich, G., and Auffray, C. (1988). A molecular map of the chicken major histocompatibility complex: The class I1 B genes are closely linked to the class I genes and the nucleolar organizer. EMBOJ. 7 , 2775-2785. Guillet, J. G., h i , M. Z., Briner, T. J., Smith, J. A., and Gefter, M. L. (1986). Interaction of peptide antigens and class I1 major histocompatibility complex antigens. Nature (London) 324, 260-262.
178
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
Guillet, J. G., h i , M. Z.,Briner, T. J.. Buus, S., Sette, A., Grey, H. M., Smith, J. A., and Gefter, M. L. (1987). Immunological self, nonself discrimination. Science 235, 865-870. Gussov, D., and Ploegh, H. (1987). Soluble class I antigens: A conundrum with no solution? Immunol. Today 8 , 220-222. Guyon-Gruaz, A,, Delmas, A., Pedoussaut, S., Halimi, H., Milhaud. G . , Raulais, D., and Rivaille. P. (1986). Oral immunization with a synthetic peptide of cholera toxin B subunit. Obtention of neutralizing antibodies. Eul: J. Bzochem. 159, 525-528. Hakem, R., Le Bouteiller, P., Barad, M., Trujillo, M., Mercier, P., Wietzerbin, J., and Lemonnier, F. A. (1989). Interferon mediated differential regulation of the expression of HLA-B7 and HLA-A3 class I genes. J. Immunol. (in press). Harris, H. W., & Gill, T. J., 111 (1986). Expression of class I transplantation antigens. Dans#lantation 42. 109-117. Healy, F., Sire, J., Gomard, E., Yssel, H., Jordan, B., and Levy, J. P. (1989). A study of functionally active amino acids involved in the interaction of HLA-A2 or HLA-A3 molecules with cytolytic T lymphocytes. J. Immunol. (in press). Hedrick, S. M., Engel, I., McElligott, D. L., Fink, P. J., Meiling, H. S . U., Hansburg, D., and Matis, L. A. (1988). Selection of amino acid in the beta chain of the T cell antigen receptor. Science 239, 1541-1544. Henning, R., and Lange-Mutschler,J. (1983). Tightly associated lipids may anchor SV40 large T antigen in plasma membrane. Nature (London) 305, 736-738. Hochgeschwender, U., Simon, H. -G., Weltzien, H. U., Bartels, F., Becker, A., and Epplen, J. T. (1987). Dominance of one T-cell receptor in the H-2Kb/TNP response. Nature (London) 326, 307-309. Hogan, K. T., Shimojo, N., Walk, S. F., Engelhard, V. H., Maloy, W. L., Coligan, J. E., and Biddison, W. E. (1988). Mutations in the alpha two helix of HLA-A2 affect presentation but do not inhibit binding of influenza virus matrix peptide. J. Exp. Med. 168, 725-736. Holt, P. G., Schon-Hegrad, M. A., and Oliver, J. (1988). MHC class I1 antigen-bearing dendritic cells in pulmonary tissues of the rat. Regulation of antigen presentation activity by endogenous macrophage populations. J. Ex$. Med. 167, 262-274. Holterman, M. J,, and Engelhard, V. H . (1986). HLA antigens expressed on murine cells are preferentially recognized by murine cytotoxic T cells in the context of the H-2 major histocompatibilitycomplex. R o c . Natl. Acad. S& US.A. 83, 9699-9703. Homer, R. J.. and Murphy, D. B. (1987). Qc-1, a novel Qregion controlled CTL determinant expressed on B lymphocytes.J. Immunol. 138, 4322-4328. Horai, S.,Claas, F. H. J., Poel, J. V. D., and Van Rood, J. J. (1982). The outcome of cell mediated lympholysis is influenced by the antibiotics in the culture medium. Tissue Antigens 19, 347-355. Hosaka, Y., Sasao, F., Yamanaka, K., Bennink, J. R., and Yewdell, J. W. (1988). Recognition of non-infectious influenza virus by class I-restricted murine cytotoxic T lymphocytes. J. Immunol. 140, 606-610. Hunt, J. S., Andrews, G. K., Fishback, J. L., Fees, M.,and Wood, G. W. (1988). Amnion membrane epithelial cells express class I HLA and contain class I HLA mRNA. J. Immunol. 140, 2790-2795. Israel, A., Kimura, A., Fournier, A., Fellous, M., and Kounlsky, P. (1987). The interferon response sequence potentiates the activity of an enhancer in the promoter region of a mouse H-2 gene. Nature (Lmd5n) 322, 743-746. Iwamoto, A., Ohashi, P. S., Pircher, H., Walker, C. L., Michalopoulos, E. E., Rupp, F., Hengartner, H., and Mak, T. W. (1987). T cell receptor variable gene usage in a specific cytotoxic T cell response. Primary structure of the antigen-MHC
MHC-ANTIGEN INTERACTION
179
receptor of four hapten-specific cytotoxic T cell c1ones.J. &p. Med. 165, 591-600. Jacobson, S., Sekaly, R. I?, Jacobson, C., McFarland, H. F., and Long, E. 0. (1989). HLA class-I1 restricted presentation of cytoplasmic measles virus proteins to cytotoxic T cells. J. Virol. (in press). Jelachich, M. L., and Biddison, W. E. (1988). Class I antigen presentation. In “The Year in Immunology,” Vol. 4 (in press). Jelachich, M. L., Cowan, E. P., Turner, R. V., Coligan, J. E., and Biddison, W. E. (1988). Analysis of the molecular basis of HLA-A3 recognition by cytotoxic T cells using defined mutants of the HLA-A3 molecule. J. Immunol. 141, 1108-1113. Jsrgensen, G . , and Hannestad, K. (1982). Helper T cell recognition of the variable domains of a mouse myeloma protein (315). Effect of the major histocompatibility complex and domain conformation. J. Exp. Med. 155, 1587-1596. Jsrgensen, T., Bogen, B., and Hannestad, K. (1981). Recognition of variable (V) domains of myeloma protein 315 by B- and T-lymphocytes. ICN-UCLA Symp. Mol. Cell. Biol. 20, 573-580. Kappes, D., and Strominger, J. L. (1988). Human class I1 major histocompatibility complex genes and proteins. Ann. Rev. Biochem. 57, 991-1028. Kappler, J. W., Wade, T., White, J., Kushnir, E., Blackman, M., Bill, J., Roehm, N., and Marrack, P. (1987). A T cell receptor V-beta segment that imparts reactivity to a class I1 major histocompatibility complex product. 49, 263-271. Kappler, J. W., Staerz, U., White, J., and Marack, I? C. (1988). Self-toleranceeleminates T cells specific for Mls-modified products of the major histocompatibilitycomplex. Nature (London) 332, 35-40. Keivits, F., Ivanyi, P., Krimpenfort, P., Berns, A., and Ploegh, H. L. (1987a). HLArestricted recognition of viral antigens in HLA transgenic mice. Nature (London) 329, 447-449. Kievits, F., Rocca, A.. Opolski, A., Limpens, J., Leuperxs, T., Kloosterman, T., Boerenkamp, W. J., Pla, M., and Ivanyi, P. (1987b). Induction of H-2-specific antibodies by injections of syngeneic Sendai virus-coated cells. Eur. J. Immunol. 17, 27-35. Kievits, F., Wijffels, J., Lnkhorst, W., and Ivanyi, P. (1989). Recognitionof &no-(HLA, SLA)-MHCantigens by mouse cytotoxic T cells is not H-2 restricted. A study with transgenic mice. Proc. Natl. Acad. Sci. U S A . (in press). Kimura, A., Israel, A., Le Bail, O., and Kourilsky, P. (1986). Detailed analysis of the mouse H-2Kbpromoter: Enhancer-like sequences and their role in the regulation of class I gene expression. Cell (Cambridge, Mars) 44, 261-272. Kisielow, I?, Bluthmann, H., Staertz, U. D., Steinmetz, M., and von Boehmer, H. (1988). Tolerance in T-cell-receptor transgenic mice involved deletion of non-mature cD4+8+ thymocytes. Nature (London) 333, 742-746. Kittur, D., Shimizu, Y.,DeMars. R., and Edidin, M. (1987). Insulin binding to human B lymphoblasts is a function of HLA haplotype. Proc. Natl. Acad. Sci. US.A . 84, 1351-1355. Klein, J. (1986). “Natural History of the Major Histocompatibility Complex.” Wiley, New York. Koch, N., Lauer, W., Habicht, J., and Dobberstein, B. (1987). Primary structure of the gene for the murine Ia antigen-associated invariant chains (Ii). An alternatively spliced exon encodes a cysteine-rich domain highly homologous to a repetitive sequence of thyroglobulin. EMBOJ. 6, 1677-1683. Koeller, D., Lieberman, R., Miyazaki,J.-I., Appella, E., Ozato, K., Mann, D. W., and Forman, J. (1987). Introduction of H-2Ld determinants into the H-2Dd antigen by site-directed mutagenesis. J. Exp. Med. 166, 744-760.
180
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
Kojima, M., Cease, K. B., Buckenmeyer, G. K., and Berzofsky, J. A. (1988). Limiting dilution comparison of the repertoires of high and low responder MHC-restricted T cells. J. Exp. Med. 167, 1100-1113. Korber, B., Mermod, N., Hood, L., and Stroynowski, I. (1988). Regulation of gene expression by interferons: control of H-2 promoter responses. Science 239,1302-1306. Koszinowski, U.H., Keil, G. M., Schwarz, H., Schickedanz, J., and Reddehase, M. J. (1987). A non-structural polypeptide encoded by immediate-early transcription unit 1 of murine cytomegalovirus is recognized by cytolytic T 1ymphocytes.J. Ex@. Med. 166, 289-294. Kourilsky, P. (1986). Molecular mechanisms for gene conversion in higher cells. Tr. Genetics 2 , 60-63. Kourilsky, P., and Claverie,J.-M. (1986). The peptidic self-model: A hypothesis on the molecular nature of the immunological self. Ann. Inst. Pdcteur (Pans) 137D,3-21. Kourilsky, F?, and Claverie, J.-M. (1989). MHC restriction, alloreactivity and thymic education: a common link? Cell 56 (in press). 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. Nutl. Acad. Sci. U.S.A. 04, 3400-3404. Krangel, M. S. (1986). Secretion of HLA-A and -B antigens via an alternative RNA splicing pathway. J. Exp. Med. 1173-1190. Kress, M., Glaros, D.,Khoury, G., and Jay, G. (1983). Alternative RNA splicing in expression of the H-2K gene. Nature (London) 306, 602-604. Krieger, J., Jenis, D. M., Chesnut, R. W., and Grey, H. M. (1988). Studies on the capacity of intact cells and purified Ia from different B cell sources to function in antigen presentation to T cells. J. Immunol. 140, 388-394. Krimpenfort, Rudenko, G., Hochstenbach, F., Guessow, D., Berns, A., and Ploegh, H. (1987). Crosses of two independently derived transgenic mice demonstrate functional complementation of the genes encoding heavy (HLA-B27) and light (beta-2-microglobulin) chains of HLA class I antigens. EMBO J. 6, 1673-1676. Kronenberg, M., Siu, G., Hood, L. E., and Shastri, N. (1986). The molecular genetics of the T-cell antigen receptor and T-cell antigen recognition. Annu. Rev. Immunol. 4, 529-591. Kulaga, H., Sogn, J. A., Weissman, J. D., Marche, P. N., Le Guern, C., Long, E. O., and Kindt, T. J. (1987). Expression patterns of MHC class I1 genes in rabbit tissues indicate close homology to human counterparts. J. Immunol. 139, 587-592. Kumar, S.,Miller, L. H., Quakyi, I. A., Keister, D. B., Houghten, R. A., Maloy, W. L., Moss, B., Berzofsky, J. A., and Good, M. E (1988). Induction of circurnsporozoite protein-specific CTL by malaria sporozoites and epitope identification. Nature (London) 334, 258-260. Kuo, C. L., and Hood, L. (1987). Antigedmajor histocompatibility complex-specific activation of murine T cells transfected with functionally rearranged T-cell receptor genes. Proc. Natl. Acad. Sci. U.S.A. 84, 7614-7618. Kvist, S.,Oxtberg, L., Person, H., Philipsson, L., and Peterson, P. A. (1978). Molecular association between transplantation antigens and cell surface antigen in adenovirus transformed cell line. Proc. Natl. Acad. Sci. U.S.A. 75, 5674-5678. Lahti, J. M., Chen, C. L. H., Sowder, J. T., Bucy, R. P., and Cooper, M. D. (1988). Characterization of the avian T cell receptor. Immunol. Res. 7 , 303-317. Lai, M.-Z., Ross, D. T.. Guillet, J.-G., Briner, T. J., Gefter, M. L., and Smith, J. A. (1987). T lymphocyte response to bacteriophage lambda repressor cI protein. Recognition of the same peptide presented by Ia molecules of different haplotypes. f. Immunol. 139, 3973-3980. Lai, M.-Z., Huang, S.-Y.. Briner, T. J., Guillet, J.-G., Smith, J. A., and Gefter, M. L.
MHC-ANTIGEN INTERACTION
187
tion of antigens by the T-cell receptor. Ann. Inst. Pasteur (Paris) 138, 175-180. Puddington, L., Bevan, M. J , , Rose, J. K . , and Lefrancais, L. (1986). N protein is the predominant antigen recognized by vesicular stomatitis virus-specificT ce1ls.J. Vzrol. 60, 708-717. Rajan, T. V. (1987). Is there a role for MHC class I antigens in the elimination of somatic mutants? Immunol. Today 8, 171-172. Rammensee, H.-G., and Bevan, M. J. (1984). Evidence from in Vitro studies that tolerance to self-antigens is MHC-restricted. Nature (London) 308, 741-744. Rechsteiner, M. (1987). Ubiquitin-mediated pathways for intracellular proteolysis. Annu. Rev. Cell. Biol. 3, 1-30, Regnier-Vigouroux, A,, El Ayeb, M., Defendini, M.-L., Granier, C., and Pierres, M. (1988). Processing by accessory cells for presentation of murine T cells of apamin, a disulfide-bonded 18 amino acid peptide. J. Immunol. 140, 1069-1075. Reith, W., Satola, S., Herrero-Sanchez,C., Amaldi, I., Lisowska-Grospierre, B., Gnscelli, C., Hadam, M. R., and Mach, B. (1988). Congenital immunodeficiency with a regulatory defect in MHC class I1 gene expression lacks a specific HLA-DR promoter binding protein, RF-X. Cell (Cambridge, Mass.) 53, 897-906. Richman, S. J., and Reese, R. T. (1988). Immunologic modeling of a 75 kDa material protein with carrier-free synthetic peptides. Proc. Natl. Acad. Sci. U.S.A. 85, 1662-1666. Robinson, P. J. (1987). Two different biosynthetic pathways for the secretion of Qa regionassociated class I antigens by mouse lymphocytes. Proc. Natl. Acad. Sci. U.S.A. 84, 527-531. Robinson, P. J.. Bevec, D., Mellor, A. L., and Weiss, E.H. (1988). Sequence of the mouse class I gene and characterization of the gene product. Immunogenetzcs 27, 79-86. Rock, K. L., and Benacerraf, B. (1983). Inhibition of antigen-specific T lymphocyte activation by structurally related Ir gene-controlled polymers. Evidence of specific competition for accessory cell antigen presentation. J. Exp. Med. 157, 1618-1634. Rock, K. L., Benacerraf, B., and Abbas, A. K. (1984). Antigen presentation by haptenspecific B lymphocytes. J. Exp. Med. 160, 1102-1113. Ronchese, F,, Brown, M., and Germain, R. N. (1987a). Structure-function analysis of the A-alphabmIz mutation using site-directed mutagenesis and DNA-mediated gene transfer. J Immunol. 139, 629-638. Ronchese, F., Schwartz, R. H., and Germain, R. N. (1987b). Functionally distinct subsites on a class I1 major histocompatibility complex molecule. Nature (London) 329, 254-256. Roosnek, E., Demotz, S., Corradin, G., and Lanzavecchia, A. (1988). Kinetics of MHC antigen complex formation on antigen-presentingce1ls.J. Immunol. 140,4079-4082. Rosloniec, E. F., Vitez, L.J., Buerstedde, J.-M., McKean, D. J., Landais, D.,Benoist, C., Mathis, D., and Freed, J. H. (1989). The use of antigen presenting cells bearing chimeric or mutant I-A molecules to present hen egg lysozyme fragments to T-cell hybridomas reveals a functionally dominant region of the I-Ak molecule. J. Immunol. (submitted for publication). Rothbard, J. B., and Taylor, W. R. (1988). A sequence pattern common to T cell epitopes. EMBO J. 7, 93-100. Rothbard, J. B., Lechler, R. I., Howland, K., Vineeta, B. A. L., Eckels, D. D., Sekaly, R., Long, E. O., Taylor, W. R., and Lamb, J. R. (1988a). Structural model of HLADRl restricted T cell antigen recognition. Cell (Cambridge, Mass.) 52, 515-523. Rothbard, J. B., Bodmer, H. C., Howland, K., Pemberton, R. M., Askonas, B. A., and Taylor, W. R. (1989). Molecular model of peptide binding to the MHC class I protein, H-2Kd. EMBO J. (Submitted for publication).
188
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
Rudd, C. E.,Trevillyan, J. M., Dasgupta, J. D., Wong, L. L., and Schlossman, S. F. (1988). The CD4 receptor is complexed in detergent lysates to a protein-tyrosin kinase (pp58) from human T lymphocytes. Proc. Natl. Acad. Scz. U.S.A. 85, 5190-5194. Sachs, D. H., Germana, S., El-Gamil, M., Gustafsson, K., Hinch, F., and Pratt, K. (1988). Class I1 genes of miniature swine. I. Class I1 gene characterization by RFLP and by isolation from a genomic library. Immunogenetics 28, 22-29. Saito, T., and Germain, R. N. (1987). Predictable acquisition of a new MHC recognition specificity following expression of a transfected T cell receptor beta-chain gene. Nature (London) 329, 256-258. Saito, T., and Germain, R. N. (1988). The generation and selection of the T cell repertoire: Insights from studies of the molecular basis of T cell recognition. Immunol. Rev. 101, 81-113. Saito, T., Weiss, A., Miller, J. Norcross, M. A., and Germain, R. N. (1987). Specific antigen-Ia activation of transfected human T cells expressing murine Ti alphabeta-human T3 receptor complexes. Nature (London) 325, 125-130. Sakurai, M., and Strominger,J. L. (1988). B-cell-specificenhancer activity of conserved upstream elements of the class I1 major histocompatibility complex DQB gene. Proc. Natl. Acad. Sci. U.S.A. 85, 6909-6913. Salter, R. D., Calyberger, C., Lomen, C. E.,Krensky, A. M., and Parham, P. (1987). In vitro mutagenesis at a single residue introduces B and T cell epitopes into a class I HLA molecu1e.J. Ex$. Med. 166, 283-288. Schilham, M. W., Lang, R., Benner, R., Zinkernagel, R. M., andHengartner, H. (1986). Characterization of a Lyt-2 alloreactive cytotoxic T cell clone specific for H-2Db that cross-reacts with I-E'. J. Immunol. 137, 2748-2754. Scholler, J., Shimonkevitz, R., McDonald, H. R., and Kvist, S . (1986). Different structural constraints for recognition of mouse H-2Kd and -Kk antigens by alloimmune cytolytic T lymphocytes.J. Ex$. Med. 164, 1823-1834. Schreiber, A. B., Schlessinger, J., and Edidin, M. (1984). Interaction between major histocompatibility complex antigens and epidermal growth factor receptors on human cells. J. Cell Biol. 98, 725-731. Schwartz, R. H. (1985). T-lymphocyte recognition of antigen in association with gene products of the major histocompatibilitycomplex. Annu. Rev. Immunol. 3, 237-261. Schwartz, R. H. (1986). Immune response (Ir) genes of the murine major histocompatibility complex. Adv. Immunol. 38, 31-201. Sekaly, R. P., Jacobson, S . , Richert, J. R., Tonnelle, C., McFarland, H. F., and Long, E. 0. (1988). Antigen presentation to HLA class 11-restrictedmeasles virus-specific T-cell clones can occur in the absence of the invariant chain. Proc. Natl. Acad. Sci. U.S.A. 85, 1209-1212. Servis, C., Seckler, R., Nagy, Z. A , , and Klein, J. (1986). Two adjacent epitopes on a synthetic dodecapeptide induce lactate dehydrogenase B-specifichelper and suppressor T cells. R o c . R . SOC. London 228, 461-470. Sette, A., Buus, S., Colon, S., Smith, J. A., Miles, C., and Grey, H. M. (1987). Structural characteristics of an antigen required for its interaction with Ia and recognition by T cells. Nature (London) 328, 395-399. Sette, A., Buus, S . , Colon, S., Miles, C., and Grey, H. M. (1988). I-Ad-bindingpeptides derived from unrelated protein antigens share a common structural motif. J. Immunol. 141, 45-48. Sharma, S., Rodgers, L., Brandsma, J., Gething, M.;T., and Sambrook, J . (1985). SV40 T antigen and the exocytotic pathway. EMBO J. 4, 1479-1489. Shastri, N., Malissen, B., and Hood, L. (1985). Ia-transfected L-cell fibroblasts present +
MHC-ANTIGEN INTERACTION
181
(1988). T cell receptor gene usage in the response to repressor cI protein. An apparent bias in the usage of V-cr gene element. J. Ex#. Med. 168, 1081-1097. Lakey, E. K., Casten, L. A., Anderson, S.,Smolenski, L. A., Smith, J. A., Margoliash, E., and Pierce, S. K. (1987a). T cell activation by processed antigen is equally blocked by I-E and I-A-restricted immunodominant peptides. Euz J. Immunol. 17, 1605-1609. Lakey, E. K., Margoliash, E., and Pierce, S. K. (198713). Identification of a peptide binding protein that plays a role in antigen presentation. Proc. Natl. Acad. Sci. U.S.A. 84, 1659-1663. Lakey, E. K., Casten, L. A., Niebling, W. L., Margoliash, E., and Pierce, S. K. (1988). Time dependence of B cell processing and presentation of peptide and native protein antigens. J. Immunol. 140, 3309-3314. Lamb, J. R., Eckels, D. D., Lake, P., Woody, J. N., and Green, N. (1982). Human T-cell clones recognize chemically synthesized peptides of influenza haemagglutinin. Nature (London) 300, 66-69. Lamb, J. R., Rees, A. D. M., Bal, V., Ikeda, H., Wilkinson, D., De Vries, R. R. P., and Rothbard, J. B. (1988). Prediction and identification of an HLA-DR-restricted T cell determinant in the 19-kDa protein of Mycobacterium tuberculosis. Euz J. Immunol. 18, 973-976. Landais, D., Marchetto, S.,Waltzinger, C., Pierres, M., Benoist, C., and Mathis, D. (1988). Slot-machine mutagenesis of a polymorphic residue on the Ak-alpha chain (1). J. Immunol. 141, 667-671. Langton, B. C., Mackewicz, C. E., Wan, A. M., Andria, M. L., and Benjamini, E. (1988). Structural features of an antigen required for cellular interactions and for T cell activation in a MHC-restricted response. J. Immunol. 141, 447-456. Lanzavecchia, A. (1985). Antigen-specific interaction between T and B cells. Nature (London) 314, 537-539. Lanzavecchia, A., Roosnek, E., Gregory, T., Berman, P.,and Abrignani, S. (1988a). T. cells can present antigens such as HIV gp120 targeted to their own surface molecules. Nature (London) 334, 530-532. Lanzavecchia, A., Abrignani, S., Scheidegger, D., Obrist, R., Durken, B., and Modenhauer, G. (1988b). Antibodies as antigens. The use of mouse monoclonal antibodies to focus human T cells against selected targets.J. J3P. Med. 167,345-352. Lechler, R. I. (1988). MHC class I1 molecular structure permitted pairs? Zmmunol. Tohy 9, 76-78. Lechler, R. I., Ronchese, E,Braunstein, N. S., andGermain, R. N. (1986). I-A-restricted T cell antigen recognition. Analysis of the roles of A-alpha and A-beta using DNAmediated gene transfer. /. Ex#. Med. 163, 678-696. Ledbetter, J. A., June, C. H., Rabinovitch, P. S., Grossmann, A., Tsu, T. T., and Imboden, J. B. (1988). Signal transduction through CD4 receptors: Stimulatory vs inhibitory activity is regulated by CD4 proximity to the CDWT cell receptor. Eur. J. Immunol. 18, 525-532. Lee, P., Matsueda, G. R., and Allen, P. M. (1988). T cell recognition of fibrinogen. A determinant on the A-alpha-chain does not require pr0cessing.J. Immunol. 140, 1063-1068. Lefkovits, I. (1986). Self and non-self discrimination by “restriction proteases.” Proc. Natl. Acad. Sci. U.S.A. 83, 3437-3438. Leserman, L. (1985). The introversion of the immune response: A hypothesis for T-B interaction. Immunol. Today 6, 352-355. Leserman, L. (1987). All in the superfamily: Presentation of antigen receptors in the context of the major histocompatibility complex. Ann. Znst. Pasteur (Pans) 138, 53-63.
182
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
Levy, R. B., Ozato, K., Richardson, J. C., and Bluestone,J. A. (1985). Molecular localization of allogeneic and self determinants recognized by bulk and clonal populations of cytotoxic T cells. J. Immunol. 134, 677-683. Lew, A. M., Lillehoj, E. P., Cowan, E. P., Maloy, W. L., Van Schravendijk, M. R., and Coligan, J. E . (1986a). Class I genes andmolecules: An update. Immunology 57. 3-18. Lew, A: M., Margulies, D.H., Maloy, W. L., Lillehoj, E. P., McCluskey, J., and Coligan, J. E. (1986b). Alternative protein products with different carboxyl termini from a single class I gene, H-2Kb. Proc. Natl. Acad. Scz. U.S.A. 83, 6084-6088. Lewis,J., Foo, M., Geier, S. S.,Kumar, P. A,, Nathenson, S. G., and Bluestone, J. A. (1988). Cytotoxic T lymphocyte recognition of novel allodeterminant expressed on in wlro selected H-2Kb mutants. J. Immunol. 141, 728-735. Leyva-Cobian, F., and Unanue, E. R. (1988). Intracellular interference with antigen presentation. J. Immunol. 141, 1445-1450. Lin. J. Berzofsky,J. A., and Delovitch, T. L. (1988). Ultrastructural study of internalization and recycling of antigen by antigen presenting cells.J. Mol. Cell. Immunol. 3, 321-343.
Livingstone, A. M., and Fathman, C. G. (1987). The structure of T-cell epitopes. Annu. Rev. Immunol. 5 , 477-501. Long, E. 0. (1988). Processing requirements for presentation of antigens to T lymphocytes. Cum Opinion Immunol. 1, 98-102. Long, E. O., Rosen-Bronson, S.,Jacobson, S.,and Sekaly, R. P. (1988). Isotypic diversity and function of HLA class I1 antigens. In “Molecular Biology of HLA Class 11 Antigens” (J. Silver, ed.). CRC Press, Boca Raton, Florida (in press). Lorenz, R. G., and Allen, P. M. (1988). Direct evidence for functional self protein/Ia complexes in vivo. Proc. Natl. Acad. Sci. U.S.A. 85, 5220-5223. Luescher, I. F., Allen, F! M., and Unanue, E.R. (1988). Binding of photoreactivelysozyme peptides to murine histocompatibility class I1 molecules. Proc. Natl. Acad. Sci. U.S.A. 85, 871-874. McCluskey, J., Bluestone, J. A , , Coligan, J. E., Maloy, W. L., and Margulies, D.H. (1986). Serologic and T cell recognition of truncated transplantation antigens encoded by in mlro deleted class I major histocompatibility genes. J. Immunol. 136, 1472-1481.
McCluskey,J . , Boyd, L. F., Maloy, W. L.,Coligan, J. E., and Margulies, D. H. (1986). Alternative processing of H-2Dd pre-mRNAs results in membrane expression of differentially phosphorylated protein products. EMBO J. 5, 2477-2483. McCluskey, J., Munitz, T., Boyd, L., Germain, R. N., Coligan, J. E., Singer, A., and Margulies, D. H. (1988). Cell surface expression of the amino-terminal domain of A-alphak. Recognition of an isolated MHC antigenic structure by allospecific T cells but not alloantibodies. J. Immunol. 140, 2081-2089. McDonald, H. R., Schneider, R., Lees, R. K., Howe, R. C., Acha-Orbea, H., Festenstein, H.. Zinkernagel, R. M., and Hengartner, H. (1988). T-cell receptor V-beta use predicts reactivity and tolerance to Mlsa-encoded antigens. Nature (London) 332, 40-45.
McElligott, D. L., Sorger, S. B., Matis, L. A., and Hedrick, S. M. (1988). Two distinct mechanisms account for the immune response (Ir) gene control of the T cell response to pigeon cytochrome c’. J. Immunol. 140, 4123-4131. Machy, P., Truneh, A., Gennaro, D., and Hoffstein, S. (1987). The major histocompatibility complex encoded class I molecules are spontaneously internalized via coated pits in T lymphocytes. Nature (London) 328, 724-726.
MHC-ANTIGEN INTERACTION
183
McKean, D. J. (1988). Structural requirements for class I1 molecule recognition by antibodies and T cell antigen receptors. B o g . Allergy (in press). McLaughlin-Taylor,E., Miyada, C. G., McMillan, M., and Wallace, R. B. (1988). Direct demonstration of critical amino acid residues required for cytotoxic Tlymphocyte allorecognition of H-2 class I antigens. Proc. Natl. Acad. Sci. US.A . 85, 3085-3089. McMichael, A. J., Gotch, F. M., and Rothbard, J. (1986). HLA B37 determines an influenza A virus nucleoprotein epitope recognized by cytotoxic T lymphocytes. J. Exp. Med. 164, 1397-1406. McMichael, A. J., Gotch, F. M . , Santos-Aguado, J., and Strominger, J. L. (1988). Recognition of influenza A virus matrix protein by HLA-A2 restricted cytotoxic T lymphocytes. I. Characterization of the peptide binding site in HLA-A2. Proc. Natl. Acad. Sci. LISA. 85, 9194-9198. Malissen, M., Trucy, J., Letourneur, F., Rebai, N., Dunn, D. E., Fitch, F. W., Hood, L., and M a l i , B. (1988). A dual reactive T cell clone expresses two T-cell receptor alpha genes but uses one alpha-beta heterodimer for allo-MHC recognition and self MHC-restricted antigen recognition. Cell (Cambridge, Mass,) 55, 49-59. Manca, F., Clarke, J. A., Miller, A., Sercarz, E. E., and Shastri, N. (1984). A limited region within hen-white lysozyme serves as the focus for a diversity of T cell clones. J. Immunol. 133, 2075-2078. Manca, F., Fenoglio, D., Kunkl, A., Cambiaggi, C., Sasso, M., and Celada, F. (1988). Differential activation of T cell clones stimulated by macrophages exposed to antigen complexed with monoclonal antibodies. A possible influence of paratope specificity on the mode of antigen processing. J. Immunol. 140, 2893-2898. Manley, J. L. (1976). Synthesis and degradation of termination and premature-termination fragments of beta-galactosidase in vitro and in mvo. J. Mol. Biol. 125, 407-432. Mann, D. W., and Forman, J. (1988). Recognition by cytotoxic T lymphocytes of QA-2 antigens. Sensitivity of QA--2molecules to phosphatidylinositol-specificphospholipase C.J.Immunol. 141, 728-735. Mann, D. W., McLaughin-Taylor, E., Wallace, R. B., and Forman, J. (1988). An immunodominant epitope present in multiple class I MHC molecules and recognized by cytotoxic T lymphocytes.J. Exp. Med. 168, 301-324. Margalit, H., Spouge, J. L., Cornette, J. L., Cease, K. B., De Lisi, C., and Berzofsky, J. A. (1987). Prediction of immunodominant helper T cell antigenic sites from the primary sequence. J. Immunol. 138, 2213-2229. Marrack, P., and Kappler, J. (1988a). The T cell repertoire for antigen and MHC. Immunol. Today 9, 308-315. Marrack, P., and Kappler, J. (1988b). T cells can distinguish between allogeneic major histocompatibilitycomplex products on different cell types. Nature (London) 332, 840-843. Marrack, P., Lo, D., Brinster, R., Palmiter, R., Burkly, L., Flavell, R. H., and Kappler, J. (1988). The effect of thymus environment on T cell development and tolerance. Cell (Cambridge, Mass.) 53, 627-634. Martin, L. H., Calabi, F., and Milstein, C. (1986). Isolation of CD1 genes: A family of major histocompatibility complex-related differentiation antigens. Proc. Natl. Acad. Sci. U.S.A. 83, 9154-9158. Martinez-A, C., Pereira, P., Toribio, M. L., Marcos, M. A. R.,Bandeira, A., De La Hera, A., Marquez, C., Cazenave, P.-A., and Coutinho, A. (1988). The participation of B cells and antibodies in the selection and maintenance of T cell repertoires. Immunol. Rev. 101, 191-215.
184
PHILIPPE KOURILSKY A N D JEAN-MICHEL CLAVERIE
Maryanski, J. L., Accolla, R. S., and Jordan, B. (1986a). H2-restricted recognition of cloned HLA class I gene products expressed in mouse cells. J. Immunol. 136, 4340-4347. Maryanski, J. L., Pala, P., Corradin, G., Jordan, B. R., and Cerottini, J.-C. (1986b). H-2 restricted cytolytic T cells specific for HLA can recognize a synthetic HLA peptide. Nature (London) 824, 578-579. Maryanski, J. L., Abastado, J.-P., and Kourilsky, P. (1987). Specificity of peptide presentation by a set of hybrid mouse class I MHC molecules. Nature (London) 380, 660-662. Maryanski, J. L., Pala, P., Cerottini, J.-C., and Corradin, G. (1988). Synthetic peptides as antigens and competitors in recognition by H-2-restricted cytolytic T cells specific for HLA. J. Exp. Med. 167, 1391-1405. Maryanski, J. L., Abastado, J.-P., McDonald, R. H., and Kourilsky, P. (1989). Intradomain KdDd recombinants define the same regions as crucial for recognition by alloreactive or major histocompatibility complex-restricted cytolytic T cells. Eur. J. Immunol. (in press). Matis, L. A., Cron, R., and Bluestone,J. A. (1987a). Major histocompatibility complex linked specificityof gamma-delta receptor-bearing T lymphocytes. Nature (London) 380, 262-264. Matis, L. A., Sorger, S. B., McElligott, D. L., Fink, F! J., and Hedrick, S. M. (198713). The molecular basis of alloreactivity in antigen-specific, major histocompatibility complex-restricted T cell clones. Cell (Cambridge, Muss.) 51, 59-69. Matzinger, F!, and Bevan, M. J. (1977). Why do so many lymphocytes respond to major histocompatibility antigens? Cell. Immunol. 29, 1-5. Matzinger, P.. and Guerder, S. (1989). Is T cell tolerance a case of suicide or murder? Nature (London) (in press). Matzinger, I?, Zamoyska, R., and Waldmann, H. (1984). Self tolerance is H-2-restricted. Nature (London) 308, 738-741. Mazerolles, F., Durandy, A., PiatierTonneau, D., Charron, D., Montagnier, L., Auffray, C., and Fischer, A. (1988). Immunosuppressive properties of synthetic peptides derived from CD4 and HLA-DR antigens. Cell 55, 497-504. Maziarz, R. T., Burakoff, S. J.. and Bluestone, J. A. (1988). The alpha-3 domain of class I MHC proteins influences cytotoxic T lymphocyte recognition of antigenic determinants located within the alpha-1 and alpha-2 domains. J. Immunol. 140, 4372-4377. 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. Merrill, J. E. (1987). Macroglia: Neural cells responsive to lymphokines and growth factors. Immunol. Today 8 , 146-150. Milich, D. R., McLachlan, A., Moriarty, A,, and Thornton, G. B. (1987). Immune response to hepatitis B virus core antigen (HBcAg): Localization of T cell recognition sites within HBcAg/HBeAg. J. Immunol. 139, 1223-1231. Mills, K. H. G. (1986). Processing of viral antigens and presentation to class 11-restricted T cells. Immunol. Today 7 , 260-263. Mills, K. H. G., Skehel, J. J., and Thomas, D. B. (1986). Conformational-dependent recognition of influenza virus hemagglutinin by murine T helper clones. Euz 1. Immunol. 16, 276-280. Mills, K. H.G., Burt, D., Skehel, J. J., and Thomas, D. B. (1988). Fine specificity of
MHC-ANTIGEN INTERACTION
185
murine class 11-restrictedT cell clones for synthetic peptides of influenza virus hemagglutinin. Heterogeneity of antigen interaction with the T cell and the Ia molecule. J. Immunol. 140, 4083-4090. Mizuno, S., Trapani, J. A., Koller, B. H., Dupont, B., and Yang, S . Y. (1988). Isolation and nucleotide sequence of a cDNA clone encoding a novel HLA class I gene. J. Immunol. 140, 4024-4030. Moller, G., ed. (1987). Antigenic requirements for activation of MHC restricted responses. Immunol. Rev. 98, 1-187. Momburg, F., Koch, N., Moller, P., Modenhauer, G., andHammerling, G. (1986). In uzuo induction of H-2K/D antigens by recombinant interferon-gamma. Eu?:J. Immunol. 16, 551-557. Morel, P. A., Livingstone, A. M., and Fathman, C. G. (1987). Correlation of T cell receptor V-beta gene family with MHC restriction. J. Ex#. Med. 166, 583-588. Morrison, L. A., Braciale, V. L., and Braciale, T. J. (1986a). Distinguishable pathways of viral antigen presentation to T lymphocytes. Immunol. Res. 3, 294-304. Morrison, L. A., Lukacher, A. E., Braciale, V. L., Fan, D. l?, and Braciale, T. J. (1986b). Differences in antigen presentation to MHC class I- and class 11-restrictedinfluenza virus-specific cytolytic T lymphocyte clones. J. Ex#. 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 I MHC-restricted cytolytic T lymphocytes. J. Immunol. 141, 363-368. Mosier, D. E. (1967). A requirement for two cell types for antibody formation in uzlro. Science 158, 1573-1575. Murray, R., Hutchinson, C. A., 111, and Frelinger,J. A. (1988). Saturation mutagenesis of a major histocompatibility complex protein domain: Identification of a single conserved amino acid important for allorecognition. Proc. Natl. Acad. Sci. U.S.A. 85, 3535-3539. Murre, C., Choi, E., Weis, J., Seidman, J. G., Ozato, K., Liu, L., Burakoff, S. J., and Reiss, C. S. (1984). Dissection of serological and cytolytic T lymphocyte epitopes on murine major histocompatibility antigens by a recombinant H-2 gene separating the first two external domains. J. Ex$. Med. 160, 167-178. Naquet, P., Ellis, J., Singh, B., Hodges, R. S., and Delovitch, T. L. (1987). Processing and presentation of insulin. I. Analysis of immunogenic peptides and processing requirements for insulin A loop-specific T cells. J. Immunol. 139, 3955-3963. Nathenson, S . G., Geliebter, J., Pfaffenbach, G. M., and Zeff, R. A. (1986). Murine major histocompatibility complex class-I mutants: Molecular analysis and structurefunction implications. Annu. Rev. Immunol. 4, 471-502. Newell, M. K., Justement, L. B., Miles, C. R., and Freed, J. H. (1988). Biochemical characterization of proteins that copurify with class I1 antigens of the murine MHC. J. Immunol. 140, 1930-1938. Novotny, J., and Auffray, C. (1984). A program for prediction of protein secondary structure fron nucleotide sequence data: Application to histocompatibility antigens. Nucleic Acids Res. 12, 243-255. Novotny, J., Tonegawa, S., Saito, H., Kranz, D. M., and Eisen, H. N. (1986). Secondary, tertiary and quaternary structure of T-cell-specific immunoglobulin-likepolypeptide chains. Proc. Natl. Acad. Sci. U S A . 83, 742-746. Ogasawara, K., Maloy, W. L., and Schwartz, R. H. (1987). Failure to find holes in the T-cell repertoire. Nature (London) 325, 450-452. Ogasawara, K., Maloy, W. L., Beverly, B., and Schwartz, R. H. (1989). Functional analysis of the antigenic structure of a minor T cell determinant from pigeon
186
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
cytochrome c: Evidence against an alpha-helical conformation. J. Immunol. (in press). Oldstone, M. B. A., Whitton, J. L., Lewicki, H., and Tishon, A. (1988). Fine dissection of a nine amino acid glycoprotein epitope, a major determinant recognized by lymphocytic choriomeningitis virus specific class I restricted H-2Db cytotoxic T lymphocytes.J. Ex$. Med. 168, 559-570. Olson, C. A., Williams, L. C., Mchughlin-Taylor, E.,and McMillan, M. (1989). Creation of H-2 class I epitope using synthetic peptides: Recognition by alloreactive cytotoxic lymphocytes. Proc. Natl. Acad. Sci. U.S.A. (in press). Ozaki, S . , and Berzofsky,J. A. (1987). Antibody conjugates mimic specific B cell presentation of antigen: Relationship between T and B cell specificity.J. Immunol. 138, 4133-4142. Ozato, K., Evans, G. A., Shykind, B., Margulies, D. H., and Seidman, J. G. (1983). Hybrid H-2 histocompatibility gene products assign domains recognized by alloreactive T cells. Proc. Natl. Acad. Scz. U.S.A. 80, 2040-2043. Ozato, K., Takahashi, H., Appella, E., Sears, D. W., Mume, C., Seidman, J. G., Kimura, S., and Tada, N. (1985). Polymorphism of murine major histocompatibility class I antigen: Assignment of putative allodeterminants to distinct positions of the amino acid sequence within the first external domain of the antigen. J. Immunol. 134, 1749-1758. Pala, P., Corradin, G . , Strachan, T., Sodoyer, R., Jordan, B. R.,Cerottini, J . C . , and Maryanski, J. L. (1988). Mapping of HLA epitopes recognized by H-2-restricted cytotoxic T lymphocytes specific for HLA using recombinant genes and synthetic peptides. J. Immunol. 140, 871-877. Parham, P., Clayberger, C., Zmn, S. L., Ludwig, D. S., Schoolnik, G. K., and Krensky, A. M. (1987). Inhibition of alloreactive cytotoxic T lymphocytes by peptides from the alpha-2 domain of HLA-A2. Nature (London) 325, 625-628. Parham, P., Lawlor, D. A., Saltker, R. D., Lomen, C. E., and Ennis, P. D. (1988a). HLA-A, B, C: Patterns of polymorphism in peptide binding proteins. In “Immunology of HLA” (B. Dupont, ed.), Vol. 2. Springer-Verlag, New York (in press). Parham, P., Lomen, C. E., Lawlor, D. A . , Ways, J. P., Homes, N., Coppin, H. L., Salter, R. D., Wan, A. M., and Ennis, l? D. (1988b). Nature of polymorphism in HLA-A, -B and -C molecules. Proc. Natl. Acad. Sci. U.S.A. 85, 4005-4009. Perkins, D. L., Lai, M.-Z., Smith, J. A., and Gefter, M. L. (1989). Class I and class I1 major histocompatibility restricted T cell responses recognize peptides with similar structural motifs. J. Ex$. Med. (Submitted for publication). Pernis, B. (1988). Recycling of histocompatibility molecules and antigen presentation. In “Antigen Processing and Presentation” (B. Pernis, S.Silverstein, and H. Vogel, eds.), pp. 247-259. Academic Press, Orlando, Florida. Phillips, M. L., Yip, C. C., Shevach, E. M . , and Delovitch, T. L. (1986). Photoaffinity labeling demonstrates binding between Ia molecules and nominal antigen on antigen-presenting cells. Proc. Natl. Acad. Sci. US.A . 83, 5634-5638. Pincus, M. R.,Gerewitz, F., Schwartz, R. H.. and Scheraga, H. A. (1983). Correlation between the conformation of cytochrome c peptides and their stimulatory activity in a T-lymphocyteproliferation assay. Proc. Natl. Acad. Sci. U.S.A. 80, 3297-3300. 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 avo. Cell (Cambridge, Mass.) 48, 231-240. Poljak, R. J. (1987). Hypothesis: A molecular model for the MHC-restricted recogni-
MHC-ANTIGEN INTERACTION
189
a lysozyme peptide but not the native protein to lysozyme-specific T cells. Proc. Natl. Acad. Sci. U.S.A. 82, 5885-5889. Shastri, N., Gammon, G., Horvath, S., Miller, A,, and Sercarz, E. E. (1986). The choice between two distinct T cell determinants within a 23-amino acid region of lysozyme depends on their structural context. J. Immunol. 137, 911-915. Sherman, D. H., Hochman, P. S., Dick, R., Tizard, R., Rachamadran, K. L., Flavell, R. A., and Huber, B. T. (1987). Molecular analysis of antigen recognition by insulinspecific T-cell hybridomas from B6 wild-type and bm12 mutant mice. Mol. Cell. Baol. 7, 1865-1872. Sherman, D. H., Waneck, G. L., and Flavell, R. A. (1988). Qa-2 antigen encoded by Q7b is biochemically indistinguishable from Qa-2 expressed on the surface of CR57B1/10 mouse spleen cells. J. Immunol. 140, 138-142. Shimizu, Y., Geraghty, D. E., Koller, B. H., Orr, H. T., and DeMars, R. (1988). Transfer and expression of three cloned human non-HLA-A, B, C class I major histocompatibility complex genes in mutant lymphoblastoid cells. Proc. Natl. Acad. Sci. U.S.A. 85, 227-231. Shinohara, N., Bluestone,J. A., and Sachs, D. H. (1986). Cloned cytotoxic T lymphocytes that recognize an I-A region product in the context of a class I antigen. J. Exp. Med. 163, 972-980. Shirayoshi, Y., Burke, P. A., Appella, E., and Ozato, K. (1988). Interferon-induced transcription of a major histocompatibility class I gene accompanies binding of inducible nuclear factors to the interferon consensus sequence. Proc. Natl. Acad. Sci. U.S.A. 85, 5884-5888. Siciliano, R. F., Hemesath, T. J.. Pratt, J. C., Dintzis, R. Z., Dintzis, H. M., Acuto, O., Shin, H. S., and Reinherz, E. L. (1986). Direct evidence for the existence of nominal antigen binding sites on T cell surface Ti alpha-beta heterodimers of MHCrestricted T cell clones. Cell (Cambridge, Mass.) 47, 161-171. Sidman, C. L., Luther, E. A., Marshall, J. D. Nguyen, K.-A., Roopenian, D. C., and Worthen, S. M. (1987). Increased expression of major histocompatibility complex antigens on lymphocytes from aged mice. Proc. Natl. Acad. Sci. U.S.A. 84,7624-7628. Singer, A., Hathcock, K. S., and Hodes, R. J. (1981). Self recognition in allogeneic radiation bone marrow chimeras. A radiation-resistant host element dictates the self specificity an immune response gene phenotype of T-helper cel1s.J. Ex#. Med. 153, 1286-1301. Singer, A., Hathcock, K. S., and Hodes, R. J. (1982). Self recognition in allogeneic thymic chimeras. Self recognition by T helper cells from thymus-engrafted nude mice is restricted to the thymic H-2 hap1otype.J. Ex$. Med. 155, 339-344. Singer, A., Mizuochi, T., Munitz, T. I., and Gress, R. E. (1986). Role of self antigens in the selection of the developing T cell repertoire. In “Progress in Immunology VI” (B. Cinader and R. G. Miller, eds), p. 60-66. Academic Press, New York. Singer, A., Munitz, T. I., Golding, H., Rosenberg, A., and Mizuochi, T. (1987). Recognition requirements for the activation differentiation and function of T-helper cells specific for class I MHC alloantigens. Immunol. Rev. 98, 145-170. Singer, D. S., Hare, J. Golding, H., Flaherty, L. and Rudikoff, S. (1988). Characterization of a new sub-family of class I genes in the H-2 complex of the mouse. Immunogenetics 137, 2311-2318. Singh, B.,Waters, C. A., and Swanlund, D. (1986). Alloantigenic sites on class I major histocompatibility complex antigens: 61-69 region in the first domain of the H-2Kb molecule induces specific antibody and T cell responses. J. Immunol. 137, 2311-2318. Sinigaglia, F., Guttinger, M., Gillessen, D., Doran, D. M., Takacs, B., Matile, H.,
190
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
Trzeciak, A , , and Pink, J. R. (1988). Epitopes recognized by human T lymphocytes on malaria cricumsporozoite protein. Euz J. Immunol. 18, 633-636. Sire, J. Chimini, G., Boretto, J., Toubert, A., Kahn-Perles, B., Layet, C., Sodoyer, R., Lemonnier, F., and Jordan, B. (1988). Hybrid genes between HLA-A2 and HLAA3 constructed by in vivo recombination allow mapping of HLA-A2 and HLA-AS polymorphic antigenic determinants. J. Immunol. 140, 2422-2430. Skoskiewicz, M. J., Colvin, R. B., Schneeberger, E. E., and Russell, P. S . (1985). Widespread and selective induction of major histocompatibility complex-determined antigens in wivo by gamma interferon. J. Ex#. Med. 162, 1645-1664. Snell, G. D. (1948). Methods for the study of histocompatibility genes. Genet. 49, 87. Snow, P. M., Van de Rijn, M., and Terhorst, C. (1985). Association between the human thymic differentiation antigens T6 and T8. Euz J. Immunol. 15, 529-532. Soloski, M. J., Lattimore, A., Hereld, D., Krakow, J. L.. Low, M. G., and Einhorn, G. (1988). Further characterization of the membrane anchor found on the tissuespecific class I molecules QaZl. J. Immunol. 140, 3858-3866. Song, E. S . , Linsk, R., Olson, C. A., McMillan, M.. and Goodenow, R. S. (1988). Allospecific cytotoxic T lymphocytes recognize an H-2 peptide in the context of a murine major histocompatibility complex class I molecule. Proc. Natl. Acad. Sci. U.S.A. 85, 1927-1931. Spinella, D. G., Hansen, T. H., Walsh, W. D., Behlke, M. A., Tillinghast, J. P., Chou, H. S., Whitely, P. J., Kapp, J. A., Pierce, C. W., Shevach, E. M.,and Loh, D. Y. (1987). Receptor diversity of insulin-specific T cell lines from C57B1 (H-2b)mice. J. Immunol. 138, 3991-3995. Spouge, J. L., Guy, R. H., Cornette, J. L., Margalit, H., Cease, K., Berzofsky, J. A., De Lisi, C. (1987). Strong conformational propensities enhance T cell antigenicity. J. Immunol. 138, 204-212. Sprent, J., and Webb, S . R. (1987). Function and specificity of T cell subsets in the mouse. Adv. Immunol. 41, 39-133. Sprent, J., Lo, D., Gao, E. K., and Ron, Y. (1988). T cell selection in the thymus. Immunol. Rev. 101, 172-189. Staerz, U. D., Karasuyama, H., and Garner, A. M. (1987). Cytotoxic T lymphocytes against a soluble protein. Nature (London) 329, 449-451. Stroyonovski, I., Soloski, M., Low, M. G., and Hood, L. (1987). A single gene encodes soluble and membrane-bound forms of the major histocompatibility Qa-2 antigen: Anchoring of the product by a phospholipid tail. Cell (Cambridge, Mass.) 50, 759-768. Sullivan, K. E.,Calman, A. F., Nikanishi, M., Tsang, S. Y., Wang, Y.,and Peterlin, B. M. (1987). A model for the transcriptional regulation of MHC class I1 genes. Immunol. Today 8, 289-293. Taira, S., and Nariuchi, H. (1988). Possible role of neuraminidase in activated T cells in the recognition of allogeneic la. J. Immunol. 141, 440-446. Takahashi, H., Cohen, J., Hosmalin, A., Cease, K. B., Houghten, R., Cornette, J. L., De Lisi, C., Moss, B., Germain, R. N., and Berzofsky, J. A. (1988). An immunodominant epitope of the human immunodeficiency virus envelope glycoprotein gp160 recognized by class l major histocompatibility complex moleculerestricted murine cytotoxic T lymphocytes. Proc. Natl. Acad. Sci. U.S.A. 85, 3105-3109. Talmon, J., Ranghino, G., Yonath, A., Cohen, I. R. (1983). Structural analysis of insulin determinants seen by T cells directed by H-2 genes. Immunogenetics 18, 79-89.
MHC-ANTIGEN INTERACTION
191
Tan, K.-N., Datlof, B. M., Gilmore,J. A., Kronman, A. C., Lee, J. H., Maxam, A. M., and Rao, A. (1988). The T cell receptor V-alpha4 gene segment is associated with reactivity to p-azobenzenearsonate. Cell (Cambndge, Mass.) 54, 247-261. Taylor, P. M., Davey, J., Howland, K., Rothbard, J. B. and Askons, B. A. (1987). Class I MHC molecules rather than other mouse genes dictate influenza epitope recognition by cytotoxic T cells. Immunogenetics 26, 267-272. Teh, H. S., Kisielow, P., Scott, B., Kishi, H., Uematsu, Y.,Bluthmann, H., and von Boehmer, H. (1988). Thymic major histocompatibility complex antigens and the or-@ T cell receptor determine the CD4/CD8 phenotype of T cell. Nature (London) 335, 229-233. Terhorst, C., Van Agthoven, A., Le Clair, K., Snow, P., Reinherz, E., and Schlossman, S. (1981). Biochemical studies of the human thymocyte cell-surface antigens T6, T9 and TlO. Cell 23, 771-780. Thorley-Lawson, D. A., and Israelsohn, E. S. (1987). Generation of specific cytotoxic T cells with a fragment of the Epstein-Barr virus-encoded p6Wlatent membrane protein. Proc. Natl. Acad. Sci. U.S.A. 84, 5384-5388. Townsend, A. R. M. (1987). Recognition of influenza virus proteins by cytotoxic T lymphocytes. Immunol. Res. 6, 80-100. Townsend, A. R. M., and Skehel, J. J. (1984). The influenza a virus nucleoprotein gene controls the induction of both subtype specific and cross-reactivecytotoxic T cells. J. Exp. Med. 160, 552-563. Townsend, A. R. M., McMichael, A. J., Carter, N. P., Huddleston, J. A., and Brownlee, G . G . (1984). Cytotoxic T cell recognition of the influenza nucleoprotein and hemagglutinin expressed in transfected mouse L cells. Cell (Cambridge, Mass.) 39, 13-25. Townsend, A. R. M., Gotch, F. M., and Davey, J. (1985). Cytotoxic T cells recognize fragments of the influenza nucleoprotein. Cell (Cambridge, Mass.) 42, 457-467. Townsend, A. R. M., Bastin, J., Gould, K., and Brownlee, G. G. (1986a). Cytotoxic T lymphocytes recognize influenza haemagglutinin that lacks a signal sequence. Nature (London) 324, 575-577. Townsend. A. R. M., Rothbard, J., Gotch, F. M., Bahadur, G., Wraith, D., and McMichael, A. J. (1986b). The epitopes of influenza nucleoprotein recognized by cytotoxic T lymphocytes can be defined with short synthetic peptides. Cell (Cambndge, Mass.) 44, 959-968. Transy, C., Lalanne, J. L.,and Kourilsky, P. (1984). Alternative splicing in the 5’ moiety of the H-2Kd gene transcript. EMBOJ. 3, 2383-2386. Transy, C., Nash, R. S., David-Watine, B., Cochet, M., Hunt, S. W., 111, Hood, L. E., and Kourilsky, P. (1987). A low polymorphic mouse H-2 class I gene from the Tla complex is expressed in a broad variety of cell types. J. Exp. Med. 166, 341-361. Tse, D. B., Cantor, C. R., McDowell, J., and Pernis, B. (1986). Recycling class I MHC antigens: Dynamics of internalization, acidification, and ligand-degradation in murine T lymphoblasts. J. Mol. Cell. 2, 315-329. Ullrich, S. J., Robinson, E. A., Law, L. W, Willingham, M., 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., Kumar, V., Kono, D. H., Gomez, C., Horvath, S. J., Clayton, J., Ando, D. G., Sercarz, E. E., and Hood, L. (1988). Restricted use of T cell receptor V genes in murine autoimmune encephalomyelitis raises possibilities for antibody therapy. Cell (Cambndge, Mass.) 54, 577-592.
192
PHILIPPE KOURILSKY AND JEAN-MICHEL CLAVERIE
Utsunomiya, N., Nakanishi, M., Arata, Y.,Yasuda, T., Saito, S., Koyama, K., and Tadakuma, T. (1988). T cell-mediated recognition of foreign antigen and the la molecule observed by stopped-flow fluorometry. J. Immunol. 141, 1471-1475. van Leuwen, A., Goulmy, E., and Van Rood, J. J. (1979). Major histocompatibility complex-restricted antibody reactivity mainly, but not exclusively, directed against cells from male donors. J. Ex$. Med. 150, 1075-1083. Vogel, J. M., Davis, A. C., McKinney, D. M., McMillan, M., Martin, W. J., and Goodenow, R. S. (1988). Molecular characterization of the CSHfeB/HeN H-2Kbme. J. Ex@. Med. 168, 1781-1800. von Boehmer, H. (1988). The developmental biology of T lymphocytes. Annu. Rev. Immunol. 6, 309-326. von Boehmer, H., Karjalainen, K., Pelkonen, J., Borgulya, P.,and Rammensee, H . G . (1988). The T-cell receptor for antigen in T-cell development and repertoire selection. Immunol. Rev. 101, 21-37. Walden, P., Nagy, Z. A., and Klein, J. (1985). Induction of regulatory T-lymphocyte responses by liposomes carrying major histocompatibility complex molecules and foreign antigens. Nature (London) 315, 327-329. Walden, P., Nagy, Z. A , , and Klein, J. (1986).Major histocompatibilitycomplex-restricted and unrestricted activation of helper T cell lines by liposome-bound antigens.J. Mol. Cell. Immunol. 2 , 191-197. Warner, C. M., Gollnick, S. O., and Goldbard, S. B. (1987a). Linkage of the preimplantation-embryo-development (Ped) gene to the mouse major histocompatibility complex (MHC). Biol. Reprod. 36, 606-610. Warner, C. M., Gollnick, S. O., Flaherty, L., and Goldbard, S. B. (1987b). Analysis of Qa-2 antigen expression by preimplantation mouse embryos: Possible relation(Ped) gene product. Biol. Reprod. ship to the preimplantation-embryo-development 36, 611-616. Watari, E., Dietzschold, B., Szokan, G., and Heber-Katz, E. (1987). A synthetic peptide induces long-term protection from lethal infection with herpes simplex virus 2. J. Exp. Med. 165, 459-470. Watts, C., and Davidson, H. W. (1988). Endocytosis and recycling of specific antigen by human B cell lines. EMBO J. 7 , 1937-1945. Watts, T. H., and McConnell, H. M. (1986). High-affinity fluorescence peptide binding to I-Ad in lipid membranes. Proc. Natl. Acad. Sci. U.S.A. 83, 9660-9664. Watts, T. H., Gaub, H. E., and McConnell, H. M. (1986). T-cell-mediated association of peptide antigen and major histocompatibilitycomplex protein detected by energy transfer in an evanescent wave-field. Nature (London) 320, 179-181. Weiss, S., and Bogen, B. (1989). B lymphoma cells process and present their endogenous immunoglobulin to MHC-restricted T cells. R o c . Natl. Acad. Sci. U.S.A. (in press). Wekerle, H., Schwab, M., Linington, C., and Meyermann, R. (1986). Antigen presentation in the peripheral nervous system: Schwann cells present endogenous myelin autoantigens to lymphocytes. Euz J. Immunol. 16, 1551-1557. Werdelin, 0. (1987a). T cells recognize antigen alone and not MHC molecules. Immunol. Today 8, 80-84. Werdelin, 0. (1987b). Answer to J. M. Claverie. Immunol. Today 8, 202-203. Werdelin, O., Mouritsen, S., Petersen, B. L., Sette, A . , and Buus, S. (1988). Facts on the fragmentation of antigens in presenting cells and on the assocation of antigen fragments with MHC molecules in cell-free systems and speculation on the cell biology of antigen processing. Immunol. Rev. 106, 181-193.
MHC-ANTIGEN INTERACTION
193
Williams, A. F.,and Barclay, A. N. (1988). The immunoglobulin superfamily. Domains for cell recognition. Annu. Rev. Immunol. 6, 381-406. Wilson, R. K., Lai, E., Concannon, P., Barth, R. K., andHood, L. E. (1988). Structure, organization and polymorphism of murine and human T-cell receptor alpha and beta chain gene families. Immunol. Rev. 101, 149-1721. Winchester, G., Sunshine, G. H., Nardi, N., and Mitchison, N. A. (1984). Antigenpresenting cells do not discriminate between self and nonself. Immunogenetics 19, 487-491.
Wraith, D. C. (1987). The recognition of influenza A virus-infected cells by cytotoxic T lymphocytes. Immunol. Today 8 , 239-246. Wraith, D. C.. and Vessey, A. E. (1986). Influenza virus-specificcytotoxic T-cell recognition: Stimulation of nucleoprotein-specificclones with intact antigen. Immunology 59, 173-180.
Wylie, D. E., Sherman, L. A., and Klinman, N. R. (1982). Participation of the major histocompatibility complex in antibody recognition of viral antigens expressed on infected cells. J. Ex#. Med. 155, 403-414. Yen, T. J., Machelin, P. S., and Cleveland, D. W. (1988). Autoregulated instability of beta-tubulin mRNAs by recognition of the nascent amino terminus of beta-tubulin. Nature (London) 334, 580. Yewdell,J. W., Bennink, J. R., Mackett, M., Lefrancois, L., Lyles, D. S., and Moss, B. (1986). Recognition of cloned vesicular stomatitis virus internal and external gene products by cytotoxic T lymphocytes. J. Ex#. Med. 163, 1529-1538. Yewdell, J. W., Bennink, J. R., and Hosaka, Y. (1988). Cells process exogenous proteins for recognition by cytotoxic T lymphocytes. Science 239, 637-640. Yokoi, T., Mulac-Jericevic, B., Kurisaki, J . - I . , and Atassi, M. Z. (1987). T lymphocyte recognition of acetylcholine receptor: Localization of the full T cell recognition profile on the extracellular part of the alpha chain of Torpedo californica acetylcholine receptor. Euz J. Immunol. 17, 1697-1702. Ziegler, H. K., and Unanue, E. R. (1981). Identification of a macrophage antigenprocessing event required for I-region-restricted antigen presentation to T lymphocytes. J. Immunol. 127, 1869-1875. Ziegler, H. K., Orlin, C. A., and Cluff, C. W. (1987). Differential requirements for the processing and presentation of soluble and particulate bacterial antigens by macrophages. Euz J Immunol. 17, 1287-1296.
ADVANCES IN IMMUNOLOGY,VOL. 45
Synthetic T and B Cell Recognition Sites: Implications for Vaccine Development DAVID R. MlLlCH Department of Molecular Biology, Reseamh Institute of Scripps Ciinic, la Joiia, California 92037
1. Introduction
The purpose of this review is twofold: first, to enumerate a representative number of B cell and T cell recognition sites defined by synthetic peptides to provide a cumulative reference for these reagents, and, second, to view this field and specifically synthetic vaccine design from an immunological rather than medical or chemical perspective. In this review all T and B cell epitopes defined by synthetic peptides will be treated equally, and no preference will be given to “medically relevant” antigens. All antigen systems are relevant in terms of their potential to provide basic information applicable to the development of conventional or synthetic vaccines. Historically, synthetic peptide antigens have been utilized since the early 1960s, primarily as linear or branched polymers or copolymers and primarily as research reagents. The availability of synthetic antigens permitted a systematic examination of parameters of antigenicity such as chemical composition, size, shape, charge, and configuration (Sela, 1966, 1969). Additional applications have included studies of cellular immunity, cell-cell interactions, specificity of T and B cell antigen recognition, tolerance, and the phenomenon of genetic restriction as discussed in the following sections. The first observation, that a peptide fragment could induce an antiviral response, was reported as early as 1963 by Anderer. Immunization with a hexapeptide fragment of tobacco mosaic virus (TMV) coupled to an albumin carrier elicited antibodies reactive with the intact virus, which partially neutralized the virus. A model system illustrating viral neutralization by antipeptide antisera was developed using the MS-2 coliphage system in 1976 (Langbeheim et al.,1976). With these notable exceptions, synthetic peptide antigens were used almost exclusively as basic research reagents until the 1980s, when a virtual explosion of interest in synthetic vaccines occurred (Lerner, 1982). The early minimal interest reflected two primary obstacles: the lack of sufficient protein sequence data, and the perception that all antibody recognition sites 195 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
196
DAVID R. MILICH
consist of conformational or discontiguous determinants. Discontiguous determinants are composed of amino acid residues near each other in the native protein structure via tertiary interactions, but distant in relation to the primary amino acid sequence. The first of these obstacles was resolved by recent advances in molecular biology, which have enabled the efficient sequencing of numerous genes. At least two developments have mitigated, if not resolved, the second obstacle: (1) synthetic peptides were found to be capable of efficiently eliciting antibodies that recognize intact proteins, which may or may not bind at the same sites as native-induced antibodies (Amon et a l . , 1971, 1976; Sutcliffe et al., 1980; Walter et al., 1980), and (2) although it is acknowledged that most epitopes on protein antigens are of the discontiguous variety (Benjamin et al., 1984), a number of proteins possess (at least to some degree) sequential or contiguous determinants dependent only on the linear amino acid sequence, which can be mimicked by techniques of peptide chemistry now in common use. This last point is somewhat controversial inasmuch as some investigators maintain that all epitopes are conformational in that the antibody will only interact with those regions of the molecule which present the appropriate “shape”to the antibody combining site. It has been suggested that contiguous determinants are often primarily portions of conformational sites (Amit et al., 1986; Sheriff et al., 1987; Colman et al., 1987; Geysen et al., 1987; Cooper and Paterson, 1987). However, debating the conformational requirements of epitopes which are totally resistent to reduction and denaturation becomes a semantic argument. Antipeptide antibodies to sequential epitopes would be expected to inhibit antibodies elicited to the native protein and, reciprocally, antibodies to the native protein should inhibit the antipeptide antibodies. Examples of such sequential epitopes are observed in a malaria circumsporozoite protein tandem repeat sequence (Nussenzweig and Nussenzweig, 1986a), the pre-S(2) region of the hepatitis B virus (HBV) surface antigen (Neurath et al., 1984b; Milich et al., 1985a), and a sequence within gp41 of the human immunodeficiency virus (HIV-1)(Rosen et a l . , 1987). It stands to reason that these epitopes may be excellent candidates for synthetic vaccine design, whereas conformation-dependent epitopes are not, at least in the context of the present synthetic peptide technology. It is equally clear that a number (possibly the majority) of native reactive antipeptide antibodies bind native protein at sites not “recognized’ by antinative protein antibodies. This category of antibody can only be elicited by immunization with peptide, since immunization with the native protein does not induce it (i.e., it is peptide unique). In terms of synthetic peptide vaccine design, the relevant question is whether this category of antipeptide antibody nonetheless neutralizes infection.
SYNTHETIC T AND B CELL SITES
197
Only relatively recently has the desirability of including antigenrelevant T cell recognition sites into the design of synthetic vaccines been considered. Previously, conjugation of synthetic B cell epitopes to heterologous protein carriers was assumed to be sufficient. The phenomenon of carrier-induced epitopic suppression (Herzenberg and Tokuhisa, 1980; Schutze et al., 1985; Jacob et al., 1985), the lack of well-defind carrier proteins suitable for human use, and the realization that priming a T cell memory response relevant to the pathogen may be beneficial and possibly essential have altered this assumption. Because T cells recognize “processed” fragments of protein antigens within the membrane of an antigen-presenting cell, rather than soluble protein, the technical problem of epitope conformation is largely avoided with respect to building a synthetic T cell site. However, the constraints placed on T cell recognition of antigen due to the dual recognition of antigen plus major histocompatibility complex (MHC) molecules present theoretical problems which may be difficult to resolve. This and other basic immunological phenomena are reviewed in the following section. II. Immunological Considerations
Before considering the use of synthetic peptides as immunogens and as potential vaccine components, it may be useful to review the relatively short history of the cellular immunology of T and B cell antigen recognition. [For a more detailed review of this subject, see Berzofiky (1987).] Although antibodies and their potential protective effects were recognized by the late 1800s, the importance of the thymus and thymus-derived lymphocytes in antibody production was not well recognized until the mid-1960s. Claman et al. (1966) first demonstrated that irradiated mice, given a mixture of thymus and bone marrow cells together with antigen (sheep red blood cells), produced a far higher antibody response than when given either cell source alone. Other investigators reached the same conclusion and lymphocytes were divided into two categories: B cells, which secreted the antibodies, and T cells, which were necessary to “help” B cells make antibody (Miller and Mitchell, 1967; Davies et al., 1967). Although the exact nature of the thymus-marrow interaction was not known, transfer experiments indicated that the thymus-marrow synergism was due to a specific interaction between these two cell types mediated via antigen (Mitchell and Miller, 1968). The discovery of T-B synergism revived interest in the phenomenon known as the “carrier effect.” It had been known for a number of years that the antibody response to a hapten was dependent on the immunogenicity of the carrier, and to obtain a secondary response the hapten must be conjugated to the same carrier as used for the primary immunization (Ovary and Benacerraf, 1963). It was then established that T and B cell interactions were mediated by recognition
198
DAVID R. MILICH
of two distinct determinants on a single antigen moiety (Rajewsky and Rottrander, 1967; Mitchison, 1971). The T cell was shown to be carrier specific, the hapten-specific cell was the B cell, and for effective cooperation, the determinants must be presented on the same hapten-carrier molecule (i.e., intramolecular T cell help) (Raff, 1970; Schirrmacher and Rajewsky, 1970; Katz et al., 1970; Paul et al., 1970). The traditional view of antigen serving as a “bridge” to connect T cells specific for one determinant and B cells specific for another determinant became increasingly complicated and difficult to conceptualize as additional information accumulated. For example, effective cognate T-B cell interaction requires MHC gene identity between T and B cells (Kindred and Shreffler, 1972; Katz et al., 1973a); T cells appear to recognize denatured and cryptic determinants on the antigen, whereas B cells recognize native antigen (Benacerraf and Gell, 1959); unlike B cells, T cells recognize nominal antigen in the context of a self-MHC component (Erb and Feldmann, 1975; Pierce et al., 1976; Swierkosz et al., 1978; Singer et al., 1979; Shih et al., 1980), and macrophages play an essential antigen-processingrole in T cell activation (Ziegler and Unanue, 1981; Allen and Unanue, 1984). Paralleling the studies discriminating the functions of T and B cells was the discovery of immune response (Ir) genes in the guinea pig (Kantor et al., 1963; Levine et al., 1963; Levine and Benacerraf, 1965) and the mouse (McDevitt and Sela, 1965). Ir genes are defined as genes which regulate the ability of an individual to produce an immune response, cellular or humoral, against a specific antigen. In the guinea pig studies, poly(L-lysine) (PLL) was used to demonstrate that inbred strain 13 guinea pigs do not respond to PLL, whereas strain 2 guinea pigs respond to PLL at both the B cell (antibody) and T cell [i.e., delayed-type hypersensitivity (DTH)] levels. Because strain 13 guinea pigs and strain 2 guinea pigs respond equivalently to most other antigens, this was the first observation of antigen-specific genetic control of the immune response. In the murine studies, branched amino acid copolymers (Tyr,Glu)-poly(~ ~ - A l a ) - p o lL-Lys[( y ( Tyr,Glu)-Ala-Lys]) and (His,Glu)poly(~~)-Ala-poly( L-Lys[( His,Glu)-Ala-Lys] ) were used, and similar Ir genes in inbred mice were discovered. Mice of strain C57BL/6 responded to (Tyr,Glu)-Ala-Lys, but not to (His,Gly)-Ala-Lys, whereas the reverse was found for mice of the CBA strain. Shortly thereafter, McDevitt’s group showed that Ir genes were linked to the MHC locus (H-2)in the mouse (McDevitt and Tyan, 1968; McDevitt and Chinitz, 1969). Demonstration of MHC linkage of Ir genes in the guinea pig (Ellman et al., 1970) and the rat (Gunther et al., 1972) indicated the universality of the finding. Use of H-2 congenic and intra-H-2recombinant murine strains facilitated the further mapping of Ir genes to between the serologically defined
SYNTHETIC T AND B CELL SITES
199
H-2 K and H-2 D loci (class I genes) within the MHC complex of the mouse. This new region of H-2 was called the Ir region, or simply the I region, which is divided into subregions (i.e., I-A, I-E; class I1 genes) (McDevitt et al., 1972; Klein et al., 1978). The next major discovery was the finding that the Ir genes encoded cell surface antigens, detectable with alloantisera on subpopulations of spleen cells (Sachs and Cone, 1973; David et al., 1973; Hauptfeld et al., 1973, 1974; Hammerling et al., 1974). These antigens were termed Ia antigens for “Ir associated” (Shreffler et al., 1974). The first evidence that Ir gene effects were mediated at the level of T cells rather than B cells came from hapten-carrier experiments. Using dinitrophenyl (DNP) as hapten attached to a protein carrier such as PLL in guinea pigs (Kantor et al., 1963; Levine et al., 1963; Bluestein et al., 1971) or (Tyr,Glu)-Ala-Lys and (His,Glu)-Ala-Lys in mice (Mozes and McDevitt, 1969), it was demonstrated that Ir gene control of the antiDNP antibody response depended on the carrier used for immunization and not the hapten. These and a number of subsequent studies (Green et al., 1966; McDevitt, 1968; Katz et al., 1971; Mitchell et al., 1972; Schwartz and Paul, 1976; Berzofsky et al., 1979) indicated that Ir gene functions were mediated through regulation of T helper (Th) cell activity. Ir gene influences on T cell interactions with other lymphoid cells soon became apparent. For adoptively transferred T cells to reconstitute an anti-SRBC response in congenitally athymic nude mice, the T cells had to be of the same H-2 genotype as the athymic recipients (Kindred and Shreffler, 1972). Therefore, the T-B cell interaction was genetically restricted, and in subsequent studies in a hapten-carrier system, T-B restriction was mapped to the I region of H-2 (Katz et al., 1975). Similarly, a requirement for histocompatibility between T cells and macrophages [antigen-presenting cells (APCs)] was discovered (Rosenthal and Shevach, 1973). This genetic restriction could also be mapped to the I region of H-2(Yano et al., 1977; Erb and Feldmann, 1975; Miller et al., 1975). Several years later it was shown that cytotoxic T lymphocytes (CTLs) require sharing of MHC antigens with their target as well as recognition of the foreign antigen (Zinkernagel and Doherty, 1974; Koszinowski and Ertl, 1975; Shearer, 1974; Gordon et al., 1975; Bevan, 1975a). The only substantiative difference between T cell-B cell and T cell-macrophage genetic restriction and CTL- target restriction is that CTL restriction maps to the class I, H-2 K, H-2 D, and H-2 L MHC antigens rather than to Ia (Shearer et al., 1975; Blanden et al., 1975; Gordon et al., 1975; Bevan, 1975b). Although the mechanisms were unclear, it was apparent that helper, proliferating, and cytotoxic T cells recognized antigen plus an MHC molecule and not antigen alone. Studies of Ir gene phenomenon (i.e., high to nonresponse) and genetic
200
DAVID R. MILICH
restriction of T cell interactions appeared to converge. For example, carrier-primed Th cells from (high responder X low responder) F1 mice could help hapten-specific B cells plus macrophage populations of the high-responder parent, but not of the low-responder parent (Katz et al., 1973b). This indicated that histocompatible T cells recognized antigens differently in combination with different MHC molecules. Two major theories emerged to explain nonresponse (i.e., Ir gene defects): (1) a gap in the T cell repertoire for recognition of a particular antigen in association with a particular Ia haplotype, termed the repertoire selection model (Jerne, 1971; Langman, 1978; Schwartz, 1978; von Boehmer et al., 1978), and (2) the inability of certain Ia molecules to effectively present certain antigens, termed the antigen presentation hypothesis or determinant selection model (Rosenthal et al., 1977; Benacerraf, 1978). This second model presumed some type of chemical interaction between antigen and Ia, and the Ir gene defect reflected constraints on this interaction. Although a tremendous amount of experimental data support both concepts, which are not mutually exclusive, recent evidence suggests that the bulk of Ir gene defects may be explained by the inability of antigenic fragments to bind MHC efficiently. This hypothesis is strengthened by the direct demonstration of a physical interaction between antigen and purified Ia molecules (Babbitt et a l . , 1985, 1986; Buus et al., 1986a,b, 1987; Guillet et al., 1987). For example, an I-Ak-restricted lysozyme peptide (residues 46-61) was shown to bind to purified I-Akbut not I-Ad in equilibrium dialysis experiments (Babbitt et al. , 1985). Reciprocally, an ovalbumin peptide (residues 323-339) restricted by I-Ad bound purified I-Ad but not I-Ak(Buus et al., 1986a). Buus et al. (1987) then analyzed 12 peptides and showed that binding correlated well with the genetic restriction data. The relevance of these peptide-Ia complexes was demonstrated by their ability to stimulate T cell hybridomas after insertion into artificial membranes (Buus et al., 198613). Watts and McConnell(l986) also demonstrated binding of peptide to Ia in planar membranes, and the binding was significantly enhanced by the addition of T cells specific for the peptide (Watts et al., 1986). This experiment suggests that in uivo the involvement of the T cell receptor (TCR) in the trimolecular complex of antigen-Ia-TCR may stabilize the complex. Peptide-Ia binding is consistent with the ability of related synthetic amino acid polymers (Werdelin, 1982; Rock and Benacerraf, 1983) or synthetic peptides (Buus and Werdelin, 1986) to compete for presentation to T cells. These earlier results have been extended by the demonstration that synthetic peptides from unrelated antigens, manifesting the same genetic restriction, compete with each other for presentation (Guillet et al., 1986, 1987; Buus et al., 1987). These studies
201
SYNTHETIC T AND B CELL SITES
implied a single binding site for antigen on any given Ia molecule. This interpretation is fortified by the recent determination of the threedimensional structure of a human class I protein (HLA-A2) (Bjorkman et al., 1987). This molecule contained a single deep groove 25 A long and 10 wide between the two long a-helices of the a1 and a2 domains, which is the likely candidate for the binding site for foreign antigen. In general, T h cells do not recognize native protein antigen, but only antigen that has been processed or physically altered (i.e., denatured, partially degraded, unfolded) and subsequently displayed in association with class I1 molecules by APCs (Ellner et al., 1977; Ziegler and Unanue, 1981, 1982; Chesnut et al., 1982; Unanue, 1984; Shimonkevitz et al., 1984). It has been suggested that intracellular processing represents events necessary to produce or select for that portion of the molecule or epitopes that has affinity for the Ia molecule, and that the epitope associated with Ia molecules creates the determinant recognized by Th cells (Unanue and Allen, 1987). The structural characteristics of the antigen required for its interaction with Ia has been the subject of numerous studies. Heber-Katz et al. (1983) suggested that since antigen interacts with Ia as well as the T cell receptor, the two sites on the antigen, one which is bound by the T cell receptor (epitope) and the other which interacts with Ia (agretope), should be distinguishable. The earliest study which suggested distinct epitope-agretope functions on the same antigen compared T cell responses between a pigeon cytochrome c fragment and the acetimidylated fragment (Hamburg et al., 1981). Murine T cells primed to a fragment consisting of residues 81-104 cross-reacted with a panel of cytochromes c from different species with a definite hierarchy of response, but failed to react with the acetimidylated fragments. Reciprocally, T cells primed to an acetimidylated fragment consisting of residues 81-104 did not cross-react with the underivatized fragments, but reacted with the acetimidylated species variants with exactly the same hierarchy of response as shown for the native fragment. These results were interpreted to suggest that the T cell receptor was binding a site which was acetimidylated, whereas, the Ir gene control was determined by a second site not affected by acetimidylation. This same group subsequently refined these studies using synthetic peptides to identify an epitope around residue 99 interacting with the T cell receptor, and an agretope at position 103 interacting with Ia (Hamburg et al., 1983a,b). Several other groups described similar findings in the myoglobin (Berkower et al., 1982) and hen egg lysozyme (Katz et al., 1982; Sercarz and Shastri, 1984; Manca et al., 1984) systems. In the hepatitis B virus system we observed that although d and y subtype determinants of the major envelope protein (HBsAg) were not cross-reactive at the T and
-
202
DAVID R. MILICH
B cell levels, the hierarchy of responsiveness between H-2 congenic strains was identical regardless of which subtype was used for immunization. These results suggested that an important agretope(s) is conserved between subtypes (Milich et al., 1984a). More recently, a number of investigators using synthetic peptide antigens have identified residues within several T cell recognition sites which bind the T cell receptor and residues which bind Ia (Sette et al., 1987; Allen et al., 1987; Fox et al., 1987). Returning to the issue of T cell-B cell interaction, T and B cells can recognize distinct entities on an antigen, as was definitively shown by Senyck et al. (1971) in the glucagon system. This study showed that the antibody response was directed against the N-terminal portion of the molecule (residues 1-17), and the T cell reactivity was specific for the C-terminal sequence (residues 18-29). Subsequently a number of investigators made similar observations in the insulin (D. W. Thomas et al., 1981), 0-galactosidase (Krzych et al., 1982), lysozyme (Maizels et al., 1980), myoglobin (Berkower et al., 1982), myelin basic protein (Chou et al., 1979), and hepatitis B virus (Milich et al., 1986a) systems. Retrospectively, this was seen in the context of T cell-B cell interaction through an antigen “bridge.” As previously noted, this traditional view became difficult to reconcile with the fact that the T cell must corecognize a processed form of antigen and MHC molecules on an antigenpresenting cell, while B cells recognize native antigenic determinants in solution. Therefore, it became clear that the antigen “bridge” could not be as simple as previously viewed. This paradox was elegantly resolved by the demonstration that antigen-specific B cells were able to process and present antigen to T cells in an MHC-restricted manner indistinguishable from that characteristic of conventional antigenpresenting cells (i.e., macrophages) (Lanzavecchia, 1985). Antigen processing by specific B cells easily explains the linked cooperation between two cells of different specificities inasmuch as the T cells selected by macrophage-processedantigen will recognize the same specificity on the antigen-specific B cell independently of the original hapten epitope recognized by the B cell. These experiments demonstrated that the “antigen bridge” is due to sequential, rather than simultaneous, recognition of antigen by B and T cells (Lanzavecchia, 1985). 111. Enumeration and Functional Characteristics of T Cell Recognition Sites Defined by Synthetic Peptides
In terms of vaccine design, it is imperative that a synthetic T cell determinant primes a memory T cell response which can be recalled by a determinant on the pathogen. Our experience in the HBV system has
SYNTHETIC T AND B CELL SITES
203
indicated that not all peptides capable of eliciting peptide-specific T cell activation (i.e., T cell proliferation, IL-2 production) are relevant to the native protein. For example, the HBsAg pre-S(2) sequence (residues 120-132) is a potent T cell antigen in the H-2s and H-2khaplotypes in terms of priming T cell (Tp) proliferation and T h cell activity specific for the peptide, but it does not prime a native-specific T cell response or represent a site recognized by native protein-primed T cells (Milich et al., 1986a). In fact, Sercarz and colleagues have categorized peptide T cell determinants as immunodominant, subdominant, and cryptic, depending on the ability to recall a proliferative response from native protein-primed T cells (Gammon et al., 1987). In this regard, we have observed that very slight modifications in a peptide (i.e., one to two residue deletions or additions), which do not affect peptide-specific T cell activation, can dramatically affect the ability to generate a native protein-specific response (D. Milich, unpublished observation). This phenomenon may reflect processing events within the native antigen which do not occur on a peptide antigen, for example, the location of a proteolytic cleavage site or an influence of “distant” residues on processing or presentation. Consistent with this hypothesis, recent studies in the myoglobin system have indicated that the fragments produced by natural processing may bind Ia molecules differently from the corresponding synthetic peptides (Brett et al., 1988). This author has not attempted to categorize T cell determinants cited herein with regard to this characteristic and refers the reader to the actual reference for this information. Table I represents a compilation of published T cell recognition sites defined with synthetic peptides. The accompanying text is an attempt to review the role these reagents have played in furthering our understanding of T cell antigen recognition. A. GLOBULAR PROTEIN ANTIGENS The T cell response to pigeon cytochrome c was one of the first to be studied in detail, and it remains one of the best characterized systems to date. One advantage of this antigen system is that cytochrome c is a small protein that has been highly conserved throughout evolution, and species variants have been quite useful in determining the importance of specific residues in the T cell response. As noted previously, studies in this system provided early evidence that antigens possessed separate sites for interaction with the T cell and interaction with Ia molecules (Hamburg et al., 1981). The immunodominant region of cytochrome c lies within the sequence 81-104 and is restricted to I-E molecules rather than I-A molecules (Table I). The minimal peptide required for a T cell response was shown to be p97-103 on moth cytochrome c (Schwartz et a l . , 1985). Studies using synthetic peptides
TABLE I ENUMERATION OF T CELL SITE System/antigen
Globular protein antigens Cytochrome dpigeon
Cytochrome dhorse Cytochrome dbovine Hen egg lysozyme NJ 0 rp
Hen ovalbumin
Site 81-104 97-103 (moth) 39-65 1-38 45-58 1-38 47-53, 48-53 13-25 1-12 1-17-S-S-120-129 1-17-S-S-120-129 46-61 74-96 52-61 74-90, 81-96 74-86 85-96 34-45 1-18 1-18 (TF’) 1-18 (Phe3) 30-53 13-35 108-120 112-129 105-120 323-339 326-336
T subseta Th Th Th Th Th Th Th Th Ts Ts Th Th Th Th Th Th Th Th DTH Th Ts Th Th Th Th Th Th Th
MHC restriction EL, Es/Ek Ek Ak Ak
Ab
Ab Ab Ad H-Zb H-2b H-2‘ Ak
Ak, Ab Ak
Ab Ak Ek Ak H-2b/d H-2b/d H-2b Ab, Ak Ab, Ak, Ad Ed
Ak
Ed Ad Ad
Reference Hedrick et al. (1982) Schwartz et al. (1985) Suzuki and Schwartz (1986) Suzuki and Schwartz (1986) Suzuki and Schwartz (1986) Baumhuter et al. (1987) Baumhuter et al. (1987) Corradin et al. (1983) Yowell et al. (1979) Adorini et al. (1979a) Adorini et al. (1979a) Allen et al. (1984) Manca et al. (1984) Allen et al. (1985a) Shastri et al. (1985) Shastri et al. (1985) Shastri et al. (1985) Allen et al. (1985b) Colizzi et al. (1985) Sette et al. (1986) Sette et al. (1986) Gammon et al. (1987) Gammon et al. (1987) Gammon et al. (1987) Adorini et al. (1988) Adorini et al. (1988) Shimonkevitz et al. (1984) Watts et al. (1985)
Myoglobin
Viral antigens Bacteriophage A repressor CI protein EBV/LMP Flu/ hemagglutinin (A/Texas/l/77) (A/Teuas/l/77) (PR8) (PR8) (A/X31) (A/JAP) (A/JAP) (A/X31) (A/X31) (A/Teuas/l/77) FWnucleoprotein (34/68) (34/68) (1934)
136-146 106-118 111-118 70-78 15-22, 56-62 145-151 113-119 10-22, 46-59, 51-63 69-80, 87-100, 107-120, 137-151 10-22, 46-59, 71-82 111-124, 138-152
Th Th Th Th Th Th Th Th
Ed Ad Ed
Ab.k H-2d
Berkower et al. (1986) Cease et al. (1986) Livingstone and Fathman (1987) Livingstone and Fathman (1987) Okuda et al. (1979) Okuda et al. (1979) Okuda et al. (1979) Bixler and Atassi (1984)
Th
H-2s
Bixler and Atassi (1984)
-
A/Cd.f.s Ad.s
12-26, 15-26 15-26 43-53
Th Th CTL
Ad Ad, Ek A1
Guillet et al. (1986) Lai et al. (1987) Thorley-Lawson and Israelsohn (1987)
306-329 105-140 111-120 302-313 48-68 103-123 181-204 118-138 54-62 307-318
Th Th Th Th Th Cl-L CTL Th Th Th
DRl/DRwl Human H-Zd H-2d Ak H-Zd H-2d Ak Ak DRl
Lamb et al. (1982) Lamb and Green (1983) Hackett et al. (1983) Hunvitz et al. (1984) Mills et al. (1986) Wabuke-Bunoti et al. (1984) Wabuke-Bunoti et al. (1984) Mills et al. (1988) Mills et al. (1988) Rathbard et al. (1988)
147-161 147-158 365-380
CTL CTL CTL
Kd
Taylor et al. (1987) Bodmer et at. (1988) Townsend et al. (1986)
Kd Db
(continued)
TABLE I (Continued) ENUMERATION OF T CELL SITES System/antigen (1968) (34/68) (34/68) (34/68) Flu/matrix FMDV/VP-1 (01)
HBV/HBsAg Pre-S(l) (ad) (ad) (ad) (ad) HBV/HBsAg Pre-S(2) (ad/ay) HBV/HBsAg (ad) (ad) (ad) (ad) (ad) HBV/HBcAg
Site
T subseta
365-379 335-349 339-347 50-63 17-31 56-68
CTL CTL CTL CTL Th Cl-L
141-160 150-160
Th Th
12-21 94-117 53-73 21-28
Th Th Th Cl-L
MHC restriction Db B37 B37
Reference
DRl A2
Townsend et al. (1986) Townsend et al. (1986) McMichael et al. (1986) Bastin et al. (1987) Rothbard et al. (1988) Gotch et al. (1987)
Guinea pig H-2k.r
Franciset al. (1985) Francis et al. (1987a)
A= H-2s.f.k.b All
Milich et Milich et Milich et Jin et al.
Kk
AS
al. (1986b) al. (1987a) al. (1987a) (1988)
120-132 126-140
Th Th
Aq.s,k Human
Milich et al. (1986a) Steward et al. (1988)
212-226 269-283 314-328 284-311 193-202 85-100 100-120 120-131 129-140
Th Th Th Th Th Th Th Th Th
AS AS H-2' H-2'
Milich et al. (1985~) Milich et al. (1985~) Milich et al. (1985~) Milich et al. (1985~) Celis et al. (1988) Milich et al. (1987b) Milich et al. (1987b) Milich et al. (1988) Milich et al. (1988)
DPw4
H-2* Af4 AS Ab
(HSDCG) HIV/gpl60 HIV/gp4l HSV/gD
H-2k.s.d H-2k.s.d H-2k.s.d Chimpanzee DR4 Gibbon Dd Human H-2' Ak
8-23
Th
Ek
1-23
Th
Human
Cease et al. (1987) Cease et al. (1987) Cease et al. (1987) Krohn et al. (1987) Siliciano et al. (1988) L w o et al. (1988) Takahashi et al. (1988) Schrier et al. (1988) Heber-Katz et al. (1985) Heber-Katz and Dietzschold (1986) Heber-Katz and Dietzschold (1986) DeFreitas et al. (1985)
428-443 112-124 386-400 426-450 410-429 426-450 308-322 584-609 1-23 1-16
Th Th Th Th Th Th CTL Th Th Th
272-293 18-44 93-112
m L Th Th
Db H-2a H-2a
Whitton et al. (1988) Madarlan et al. (1984) Wan et al. (1986)
326-343 330-343 (NANPI, 300-310 361-380 103-122 378-398 368-390
Th Th Th Th Th Th Th CTL
A', human Ad Ab Human Human Human DR2/DRw9 H-2k
Good et al. (1987) Good et al. (1988a) Good et al. (1986) Good et al. (1988a) Good et al. (1988a) Sinigaglia et al. (1988) Sinigaglia et al. (1988) Kumar et al. (1988)
Human Human
Kabilan et al. (1988) Crisanti et al. (1988)
C-terminal repeats 225-237
Th Th
(continued)
TABLE 1 (Contmued) ENUMERATION OF T CELLSITES
System/antigen
Bacterial antigens M . tuberculoslj/65-kDa protein M . bovis BCG (adj. Arth.) Staphylococcal nuclease Streptococcus pyogenes Type 5 M protein Type 24 M protein Tetanus toxin
m
Other Acetylcholine receptor a-subunit
Angiotensin I1 Bee venom/apamin Fibrinogen/fibrinopeptide B Fibrinogen/A,-chain Hemoglobin/b-chain
Site
T subset"
112-132, 437-459 180-188 91-110 81-100, 51-70 61-80
Th Th Th Th Th
1-10. M6 1-11 1-12 830-843 1273-1284 52-70 73-90 100-116 152-167 172-205 125-147 1-8 1-8 (va15) 1-18 87-14 fl-13
fl -14
571-578 12-24, 60-69, 73-8 112-124, 133-146 13-24, 73-83, 133-146
MHC restriction
Reference
Human Rat Ab Ek H-Zd
Lamb et al. (1987) van Eden et al. (1988) Finnegan et al. (1986) Finnegan et al. (1986) Finnegan et al. (1986)
Th Th Th Th
DR5 DR52a
Beachey et al. (1987) Beachey et al. (1987) Demotz et al. (1988) Demotz et al. (1988)
Th Th Th Th Th Th Th Th Th Th Th Th Th Th Th Th
Rat (Buffalo) Rat (Lewis) Rat (Lewis) Rat (W/F) Rat (B/N) Rat (Lewis) GP strain 2 GP strain 13 Ab, Ad GP strain 2 GP strain 13 H-2k EL H-2' H-2s H-2d
Fuji and Lindstrom (1988) Fuji and Lindstrom (1988) Fuji and Lindstrom (1988) Fuji and Lindstrom (1988) Fuji and Lindstrom (1988) Lennon et al. (1985) Thomas et al. (1981) Thomas et al. (1981) Regnier-Vigouroux et al. (1988) Thomas et al. (1978) Thomas et al. (1978) Peterson et al. (1983) Lee et al. (1988) Yoshioka et al. (1986) Yoshioka et al. (1986) Yoshioka et al. (1986)
Insulin/A chain (Pork) Insulin/B chain (Pork) MHC (HLA-A2) (HLA-CW3) (HLA-A24) (H-2Ld) Myelin basic proteidguinea pig
Myelin basic protein/mouse Myelin basic protein/bovine Myelin basic proteinhat Myelin basic proteinhabbit Myelin proteolipid protein Ragweed allergen/Ra3
A8-10 A1-14-S-S-B7-15
Th Th
GP strain 2 Ad
Barcinski and Rosenthal (1977) Naquet et al. (1987)
B5-16
Th
GP strain 13
Thomas et al. (1981)
Aw69
Clayberger et al. (1987) Maryanski et al. (1986) Maryanski et al. (1988) Song et al. (1988) Chou et al. (1977) Vandenbark et al. (1985a) Pettinelli et al. (1982) Pettinelli et al. (1982) Fritz et al. (1985) Fritz et al. (1985) Martenson et al. (1975) Zamvil et al. (1986) Kono et al. (1988) Kono et al. (1988) Kira et al. (1986) Tuohy et al. (1988) Kurisaki et al. (1986) Kurisaki et al. (1986) Kurisaki et al. (1986) Kurisaki et al. (1986) Kurisaki et al. (1986) Kurisaki et al. (1986) Kurisaki et al. (1986) Rothbard et al. (1988)
56-69, 98-113 171-186 170-182 61-85 68-88 75-80 1-37 44-48 1-37 89-169 37-42, 108-113 1-9 NAc 87-98 91-104 15-31 103-116 7-16, 52-64 73-83, 87-99 10-19, 52-64 8-18, 52-64 92-101 10-19, 52-64 87-99 54-61
Cl-L
CTL CTL CTL Th Th Th Th Th Th Th Th Th Th Th Th Th Th Th Th Th Th Th Th
Kd
Kd Dk Rat Rat A" AS
A", Ak AS, AS Rat A" AS AS Rabbit AS H-2d H-2d H-2' H-2s H-2' H-2s H-2s DR1
"Th, T helper cell; Ts, T suppressor cell; DTH, delayed-type hypersensitivity; Cl-L, cytotoxic T lymphocyte.
210
DAVID R. MILICH
with three different residues at position 99 indicated that T cells immune to a peptide reacted only with peptides bearing the same residue at position 99, and did not cross-react with peptides bearing another residue at that site. Therefore, T cell cross-reactivity, an indicator of receptor specificity (i-e., epitope), depended on the residue at position 99. However, Ir gene control measured as immunogenicity depended on residue 103 in the B1O.A and BlO.A(SR) strains. This data suggested that an epitope existed around residue 99 which interacted with the T cell receptor, and an agretope existed at position 103 which interacted with Ia (Hamburg et al., 1988a,b). Another approach using synthetic peptides involved making analogs with several amino acid substitutions at position 99 or 103 and assaying the peptides for the ability to stimulate T cell clones specific for the parent peptide. It was found that only very few conservativesubstitutions were tolerated at position 99 (i.e., ornithine and arginine for lysine), whereas a greater number of substitutions were tolerated at position 103, the agretope. Therefore, the Ia molecule appears to have more flexibility in its interaction with antigen as compared to the T cell receptor (Fox et al., 1987). This subject has also been addressed in detail in the ovalbumin system (Sette et al., 1987). Recent studies in the cytochrome c system have addressed the issue of determinant selection versus repertoire selection models to explain the mechanism of Ir gene effects. The BlO.A(5R) strain is a nonresponder to pigeon cytochrome c, but does respond to moth cytochrome c, and these T cells will respond to pigeon or moth cytochrome c on BIO.A APCs (Heber-Katz et al., 1983). Molecular analysis of the rearranged T cell receptors of BIO.A and BlO.A(5R) pigeon cytochrome c-specific T cells showed that these strains utilize identical T cell receptors (McElligott et al., 1988). The authors suggest that this data, together with the observed low affinity of pigeon cytochrome c peptides for I-Eb molecules (Buus et al., 1987), argue that the Ir defect in the BlO.A(5R) strain is due to a defect in antigen presentation (i.e., determinant selection model). In this same study a comparison of T cells from the highresponding strains BIO.A and BlO.S(9R), which are capable of responding to cytochrome c presented by either BIO.A or BlO.S(SR) APCs, revealed that the T cell receptors expressed in these two strains were totally unrelated. The authors conlcuded that both determinant selection and repertoire selection can be mechanisms of Ir gene control of an immune response (McElligott et al., 1988). Additional T cell recognition sites have been localized to the N-terminal fragment of cytochrome c, fragment 1-65. However, these T cell sites are recognized in association with I-A molecules in contrast to T cell recognition of the C-terminal end of pigeon cytochrome c (Suzuki
SYNTHETIC T AND B CELL SITES
211
and Schwartz, 1986). T cell sites have been defined within peptides 1-38, 39-65, and 45-58, and similar sites were reported for horse cytochrome c (Baumhuter et a l . , 1987) (Table I). Examination of clones specific for the 39-65 site revealed that each clone had a different fine specificity. Since the Ia molecule is constant in these experiments, the differences in fine specificity among clones most likely stem from differences in the structure of the T cell antigen receptor (Suzuki and Schwartz, 1986; Fink et al., 1986). This is a reoccurring theme, since similar observations of T cell clonal diversity for recognition of the same peptide have also been made in the lysozyme (Shastri et al., 1984; Allen and Unanue, 1984) and myoglobin (Berkower et al., 1984; Cease et al., 1986) systems. Corradin et al. (1983) identified a T cell recognition site at residues 13-25 with bovine apocytochrome c. This T cell site is of interest in that it appears similar to a site bound by a monoclonal antibody specific for this antigen and generated in the same BALB/c strain. Typically, T and B cell recognition sites on a molecule appear to be distinct. Synthetic peptides representing T cell recognition sites on cytochrome c have been used to investigate another important immunologic phenomenon, namely, neonatal tolerance induction. Three versions of the 93-103 sequence of moth cytochrome c, consisting of the parent peptide and two peptides with substitutions for lysine at position 99 (epitope), were used to induce neonatal tolerance. It was found that tolerance was specific for the peptide used for its induction, and the response to the other peptides was unaffected (Gammon et al., 1986a). Thus, the specificity of tolerance was dependent on substitutions at position 99 within the T cell epitope matching the specificity of T cell proliferation. The authors suggest that these data support the hypothesis of clonal deletion, and exclude a mechanism of suppression in induction of neonatal tolerance since a suppressive epitope distinct from the proliferative epitope would have resulted in cross-reactive tolerance (Gammon et al., 198613). Hen egg lysozyme (HEL) and constituent peptides have served as important antigens to study the phenomenon of T cell “immunodominance,” to analyze the influence of Ir genes, to examine antigen-Ia interaction, and to identify epitopes able to induce selected T cell subsets [i.e., suppressor T cells (Ts), T h cells, and T cells mediating DTH] . Mice of the H-2bhaplotype are responders to four species of bird lysozymes, but are nonresponders to six others, including hen lysozyme. The immunogenic species variants possess a tyrosine at position 3, whereas, the nonimmunogenic lysozymes have a phenylalanine at this position (Hill and Sercarz, 1975). It was observed that immunization with a cleavage fragment of HEL comprising residues 13-105 elicited a T cell
212
DAVID R. MILICH
response in nonresponder H-2bmice (Sercarz et al., 1978; Adorini et al., 1979a). Furthermore, preimmunization with a fragment of HEL called N-C, consisting of residues 1-17 from the N-terminus linked by a disulfide bond to the C-terminal 120-129 residues, induced suppression in H-2b mice similar to preimmunization with intact HEL (Adorini et al., 1979a,b).These data indicated that Th cells recognized an epitope within residues 13-105, whereas, T suppressor cells recognized an epitope within the N-C peptide. Subsequent studies have localized the suppressor epitope to the N-terminal residues (Yowell et al., 1979; Sette et al., 1986) (Table I). Induction of Ts cells by specific epitopes has been observed in other systems as well, including: 0-galactosidase(Eardley and Sercarz, 1976; Turkin and Sercarz, 1977; Krzych et al., 1983, 1985), myelin basic protein (Swanborg, 1975; Swierkosz and Swanborg, 1975; Hashim et al., 1976), and bovine serum albumin (BSA) (Muckerheide et al., 1977; Ferguson et al., 1983). Therefore, the outcome of an immune response may depend on the balance between T h cell and Ts cell activation by distinct sites within the antigen. The implications of these observations for synthetic vaccine design are obvious. A synthetic antigen could be constructed to lack suppressor-inducing sites and, therefore, possibly would represent a superior immunogen as compared to the native protein. However, a theoretical objection may be that a suppressor determinant for one individual may be a helper determinant for another due to MHC restriction of T cell antigen recognition. Examination of T cell recognition in a number of protein systems has indicated that the T cell response appears highly focused on a few determinants, e.g., cytochrome c (Solinger et al., 1979), staphylococcal nuclease (Finnegan et al., 1986), sperm whale myoglobin (Berkower et al., 1984), 0-galactosidase (Krzych et al., 1982), and HEL (Maizels et al., 1980; Katz et al., 1982; Bixler et al., 1985). However, as these systems and partiularly HEL were examined in greater detail, more T cell determinants were identified. It has become apparent that the number of potential T cell sites is not the limiting factor, but rather distinct sites are “selected’ by each H-2 haplotype and even within a single haplotype immunodominant epitopes further focus the T cell response (Gammon et a l . , 1987). In the HBV system, the specificity of the T cell response to the nucleocapsid (HBcAg) is quite diverse among strains; however, each of five different haplotypes predominantly recognized a distinct determinant (Milich et a l . , 1987b), and similar results were obtained upon examination of the T cell response to the pre-S(Z) region of the envelope of HBV (D. Milich, unpublished observation). Sercarz and colleagues have suggested that in addition to Ia-antigen interactions and the T cell receptor repertoire, other factors also affect the choice of
SYNTHETIC T AND B CELL SITES
213
determinants recognized in a protein molecule. For example, immunization with three molecular forms of HEL (i.e., HEL, peptide 74-96, and cleavage fragment 13-105) elicited three distinct patterns of fine specificity (Shastri et al., 1985). It was suggested that the three molecules were processed differently. Whatever the mechanism, this observation is relevant to synthetic vaccine design since it implies that qualitative changes in T cell recognition may result from the use of peptide rather than native immunogen (Gammon et al., 1987). As is evident in Table I, a number of T cell sites have been identified within HEL, and a number of these sites are I-Ak-restricted.However, priming with HEL does not elicit equivalent peptide-specific responses, and some peptide responses are not recalled by HEL at all (Gammon et al., 1987). It was suggested that multiple determinants (i.e., 13-35,30-53,46-61,74-86) may compete in mvo for the same MHCbinding site, thereby resulting in a hierarchial response. The selection of “dominant” T cell recognition sites is important in the design of a synthetic vaccine. However, again MHC restriction of the T cell recognition process may mitigate against the ability to define “dominant” epitopes relevant for an outbred population. In a detailed study of an HEL T cell determinant p46-61, recognized by T cells of H-2k strains, Allen et al. (1985a) demonstrated that the minimal stimulatory peptide consisted of residues 52-61. Only a single residue, Leu56,is different from mouse lysozyme, and this residue is responsible for the immunogenicity of this peptide (Allen et al., 1985b). These investigators determined three Ia molecule contact residues (agretope) and three T cell receptor contact residues (epitope) within the peptide by testing the ability of alanine-substituted analogs to stimulate a T cell hybridoma and to compete for the presentation of the stimulatory HEL peptide, p46-61 (Allen et al., 1987). The side chains of the T cell contact residues T ~ TLeu56, ~ ~and , Glu5’ were all oriented in one direction, whereas the side chains of the Ia contact residues Asp52, Ile5*,and A r e 1 were oriented in an opposite direction as determined by macromolecular modeling. This orientation would allow one face of the peptide to bind Ia and the opposite face to contact the T cell receptor in an a-helical conformation. A number of T cell epitopes have been shown to correlate with a-helical conformation (Pincus et al., 1983; Spouge et al., 1987; Carbone et al., 1987), and a predictive model has been suggested based on amphipathic a-helical propensity (De Lisi and Berzofsky, 1985; Berkower et al., 1986; Berzofsky et al., 1987). In addition to HEL (46-61), three other I-Ak-restnctedT cell peptides were used in competition experiments, and although these three peptides possessed no sequence homology, they all competed for binding to Ia
214
DAVID R. MILICH
molecules with HEL (46-61) (Allen et al., 1987). The authors concluded that many unrelated peptides bind to a single Ia site and that the Ia molecule has a broad binding specificity. Similar conclusions were reached in analysis of the cytochrome c and ovalbumin systems (Fox et al., 1987; Sette et al., 1987). These in vitro Ia competition experiments have recently been extended to self-peptides in an in vivo system. Adorini et al. (1988) have shown that the mouse lysozyme peptide 46-62 (i.e., selfpeptide) can inhibit induction of a T cell response to HEL 46-61 if injected together as a mixture, depending on the competing peptide/antigen molar ratio. Although the self-peptide is not immunogenic due to tolerance, it was still capable of competing with a foreign peptide at the level of Ia interaction. The murine antibody and T cell responses to sperm whale myoglobin (Mb) were shown to be under the control of two H-2-linked Ir genes, one mapped in the I-A subregion (Ir-Mb-1)and the other in the I-E/I-C subregion (Ir-Mb-2)(Berzofskyet al., 1979). Mice carrying only Ir-Mb-1 produced a T cell proliferative response to both N- and C-terminal cyanogen bromide fragments, whereas, T cells of mice carrying only Ir-Mb-2 proliferated only to the N-terminal fragment. This study demonstrated that distinct Ir genes controlled responses to different antigenic determinants on the same protein, and indicated that Ir genes were epitope specific and not molecule specific. Therefore, the response to a multideterminant antigen would be expected to be controlled by numerous Ir genes. Similar conclusions were suggested by studies in the staphylococcal nuclease (Berzofsky et al., 1977), insulin (Barcinski and Rosenthal, 1977), and hepatitis B (Milich et al., 1984a) systems. Using synthetic peptides, Berzofsky and colleagues have identified two “dominant” T cell recognition sites in sperm whale myoglobin within residues 106-118 and 136-146 (Berkower et al., 1982, 1984, 1985a, 1986; Cease et al., 1986). It was shown that myoglobin can be presented in association with either I-Ad or I-Ed molecules. However, all 106-118-specific clones were restricted to I-Ad (Berkower et al., 1985b). This finding indicated determinant selection at the clonal level, and suggested that each class I1 molecule interacts differently with the antigen and this results in the presentation of a different epitope. More recently a T cell site was localized to residues 111-118; however, this epitope is presented in association with I-Ed, in contrast to p106-118, which is I-Ad restricted (Livingstone and Fathman, 1987). As noted in other antigen systems, a heterogeneity of fine specificity was found among T cell clones which are specific for the same peptide sequence and the same MHC molecules (Cease et al., 1986; Livingstone and Fathman, 1987). The authors suggest that this, taken together with Ia-antigen binding studies,
SYNTHETIC T AND B CELL SITES
215
reflects heterogeneity in the T cell repertoire for a single peptide bound to a single site on the Ia molecule. Atassi and colleagues have arrived at a somewhat different view of the T cell response to myoglobin. A large number of T cell recognition sites were identified in a given H-2haplotype (see Table I; Okuda et al., 1979; Bixler and Atassi, 1984). For example, H-Zd haplotype-bearing strains recognized synthetic peptides 10-22, 46-59, 51-63, 69-80, 87-100, 107-120, and 137-151 (Bixler and Atassi, 1984). It was reported earlier by this group that a number of myoglobin-derived peptide T cell sites were not recalled by native myoglobin, and this led to the hypothesis that T cell recognition may require a high level of conformational dependency (Young and Atassi, 1983a,b). Berkower et al. (1984) reported that monoclonal or polyclonal myoglobin-specific T cells could not distinguish between native and denatured forms of myoglobin. The T cell sites reported by Bixler and Atassi (1984) and Okuda et al. (1979) differ from those identified by other investigators in another respect as well, in that many also represent antibody-binding sites defined by this group (Atassi, 1975). In contrast, no similarities between the antigen-binding sites of T cell receptors and antibodies were reported by Berkower et al. (1984) in the myoglobin system. The T cell response to chicken ovalbumin and constituent peptides have served as a model for the analysis of the structural requirements for peptide-Ia interaction. Shimonkevitz et al. (1984) first described a tryptic fragment of ovalbumin, 323-339, which could be presented by fixed APCs to I-Ad-restricted, ovalbumin-specific T cell hybridomas. Subsequently, synthetic peptides were used to more narrowly define the T cell determinant to residues 326-336 (Watts et a l . , 1985). Kinetic analysis of the interaction between ovalbumin peptide 323-339 and purified I-Ad molecules indicated that the rate of complex formation was very slow (k, = 1 M-' s-l), however, once formed, the complexes are very stable ( k d = 3 X s-l) (Buus et al., 1986b). The finding that the association rate is 10-foldslower at low pH as compared to pH 7.2 suggests that lysosomes are not likely cadidates for the organelle where association between Ia and processed antigen takes place. It was suggested that antigen may associate with Ia on the plasma membrane or in some undefined post-Golgi storage compartment (Buus et al., 1986b). These same authors showed that peptide-I-Ad complexes incorporated into planar membranes were 20,000-fold more efficient in stimulating a T cell response than peptide added to planar membranes containing uncomplexed I-Ad. These results suggest that the peptide-Ia interaction represents a physiologically relevant event in T cell activation. It was also found in glutaraldehyde cross-linking experiments that only the
216
DAVID R. MILICH
I-Ad a-chain was efficiently cross-linked to the ovalbumin peptide; however, this may have been due to the cross-linker used (Buus et al., 1986b). In a detailed analysis of the structural requirements for T cell recognition of ovalbumin peptide 323-339, a series of analogs and truncated versions were synthesized and examined for their capacity to bind to I-Ad and to stimulate ovalbumin-specificT cell hybridomas (Sette et al., 1987). The results indicated that the requirements for peptide-I-Ad binding were highly permissive, since a large number of substituted analogs allowed binding. It was noted that this type of specificity is predicted by the determinant selection model, which requires that relatively few Ia molecules have the capacity to bind a large universe of foreign antigens. In contrast, it was found that T cell recognition was far more sensitive to single amino acid substitutions than I-Adpeptide interaction, and the authors equated this type of fine specificity with that of antibody recognition (Sette et al., 1987). It was also found that Ia and T cell receptor interacting residues were intermingled on the peptide antigen, which suggested that some residues interacted with both Ia and the T cell receptor. A model of the trimolecular complex was proposed in which the peptide is “sandwiched” in a @-sheetor other extended conformation between Ia and the T cell receptor (Sette et a l . , 1987). The data using this ovalbumin peptide were not consistent with the model that predicts that immunogenic peptides form a-helical structures, with the caveat that some of the substitutions may affect conformation of the peptides. B. VIRALANTIGENS Protective immunity to virus infection requires the development of helper T cells specific for viral antigens. A T h function is required for the production of neutralizing antibodies, the induction of a DTH response, and possibly the development of cytotoxic precursor cells into cytotoxic effector cells. The study of T cell recognition in a viral system offers several advantages. For example, human as well as murine systems can be explored, an infectious system allows one to examine the induction of all aspects of the cellular response, and the existence of viral subtypes serves as a convenient source of primary sequence variants similar to the species variants of the small globular proteins discussed in the preceding section. Additionally, there is great interest in vaccine development for many viral diseases. The antigens of the influenza virus have been extensively studied in terms of T cell immunity. Lamb et al. (1982) first identified a human T h recognition site, residues 306-329, within the hemagglutinin (HA)
SYNTHETIC T AND B CELL SITES
217
molecule using synthetic peptide technology (Table I). The minimal residues required for T cell activation were subsequently defined as 307-318 (Rothbard et al., 1988). In the murine system, a number of T cell recognition sites have been identified for several H-2 haplotypes. The fine specificity of a panel of hybridomas obtained from BALB/c mice immunized with a synthetic peptide representing residues 111-120 of the HA1 chain of the HA molecule was shown to be indistinguishable from the fine specificities of T cells generated by priming with intact wild-type PR8 virus when tested against a panel of mutant and variant influenza viruses (Hackett et al., 1985). Influenza is a recurrent disease because of the rapid mutation rate, particularly in the gene coding for HA. It is well appreciated that alterations in antibody recognition sites on HA are a mechanism by which the virus eludes the immune response; however, escape from Th cell recognition may be another important factor in recurrent disease. In this regard, it was demonstrated that antigenic drift in HA occurring between the years 1934 and 1957 and between 1977 and 1980 altered two of three T cell determinants studied (Hurwitz et al., 1984). Therefore, knowledge of human T cell determinants will be a necessary requirement for the design of synthetic peptide vaccines, since it may permit modification and optimization of T cell recognition. In response to influenza A virus infection, mice and humans produce a vigorous cytotoxic T lymphocyte (CTL) response. The CTLs generally express the CD8+ phenotype and are MHC class I restricted. It is believed that CTLs play an important role in limiting the spread of virus in vivo. However, until recently little was known about the nature of the viral epitopes recognized by CTLs on the surface of an infected target cell. It was observed that the majority of CTLs cross-reacted on target cells infected with all strains of A viruses but not B viruses (Askonas et al., 1982; Townsend and McMichael, 1985). This was later explained by the finding that the highly conserved nucleoprotein (NP) was the major antigen recognized in murine strains (Townsend et al., 1984; Yewdell et al., 1985) and in the majority of humans (Gotch et al., 1986). Furthermore, the CTL response to NP appeared to be under strict MHC gene control (Townsend et al., 1985; Pala and Askonas, 1986; McMichael et al., 1986). The question then became how a viral component that was not a transmembrane protein and did not possess a recognizable leader sequence was transported to the plasma membrane of the target cell, where CTL recognition is assumed to occur. It was observed that CTLs could recognize short overlapping regions of nucleoprotein fragments expressed in transfected murine fibroblasts (Townsend et al., 1985). The authors suggested that CTLs may recognize segmental
218
DAVID R. MILICH
epitopes of denatured or degraded intracellular proteins that are exported to the cell surface. Such degraded viral protein may then become available for recognition by CTLs in association with class I MHC molecules in a way similar to T h cell recognition of protein fragments in association with class I1 molecules. Synthetic peptides representing sequences within the influenza nucleoprotein were utilized to sensitize target cells, and the results of these experiments have substantiated the view that somatic cells bearing class I molecules are capable of degrading and presenting newly synthesized viral proteins to CTLs (Townsend et al., 1986). An epitope was mapped to a short region of NP that has undergone genetic change since 1934, represented by residues 365-380 (Townsend et al., 1986). Two mutually exclusive epitopes were defined using CTL clones that distinguished between the 1934 and 1968 isolates, which differ by two amino acids in this region. The murine CTL clones were both restricted to the H-2Db molecule. In this same report, the authors also defined a human, HLA-B37-restricted, CTL recognition site on the NP at residues 335-349 (Townsend et al., 1986). The minimal peptide required for stimulation was subsequently defined as residues 339-347 (McMichael et a l . , 1986). It was also observed that CTLs from six HLA-B37-positive donors recognized amino acids 335-349 of the NP sequence, whereas CTLs from 12 donors of other HLA types failed to recognize this determinant. Similarly, mice of the H-2k haplotype do not recognize the 365-380 sequence of the NP as do H-2bhaplotype mice, but recognize amino acids 50-63 in the context of H-2 Kk(Bastin et al., 1987). Furthermore, mice of the H-2dhaplotype recognize a third site on the NP, residues 147-161, in association with H-2 Kd (Taylor et al., 1987). Cumulatively, these data argue very strongly that a direct interaction between a unique viral epitope and a given class I MHC molecule “selects” the appropriate determinant recognized by CTLs. The finding that ClXs recognize small synthetic peptides in uilro raises the possibility of using peptide vaccines to prime or restimulate CTLs in vivo. Successful studies using synthetic peptides to induce CTLs in uivo have not been reported to date. However, a recent study using NP 365-380 showed that peptide fragments can efficiently generate strong primary in vitro CTL responses under conditions in which conventional virusinfected cells cannot (Carbone et al., 1988). Although the in m2ro peptide-primed CTLs were inefficient in lysing target cells expressing endogenous antigen (i.e., influenza virus-infected cells), further study and optimization of this system may yield results directly applicable to the design of synthetic antiviral vaccines. However, similar to T h cell sites, any single peptide representing a CTL site would not be expected
SYNTHETIC T AND B CELL SITES
219
to be sufficient in an outbred human population due to the genetic restriction imposed on CTL antigen recognition. It will, therefore, be important to define a variety of peptides with which given HLA molecules can interact. Influenza-immune CTLs have also been shown to recognize another internal component of the virus, the matrix protein. Two T cell recognition sites within the influenza matrix protein have been defined with synthetic peptides. A T h cell site corresponding to residues 17-31 is recognized in association with HLA-DR1 (Rothbard et al., 1988), and a CTL site restricted by HLA-A2 corresponds to residues 56-68 (Gotch et al., 1987). Comparison of the matrix T h site (17-31) sequence with another HLA-DR1-restrictedT h site (307-318) from the hemagglutinin molecule revealed a similar pattern that could account for their binding to DR1. If the peptides were to adopt a helical conformation, the two DR1-restricted epitopes could be aligned in a similar manner (Rothbard et al., 1988). It was postulated that the similar amino acids in the two determinants may constitute the residues interacting with DR1 (agretope), while the other amino acids on the opposite face of the a-helix would contact the T cell receptor (epitope). To test this hypothesis, the investigators synthesized complementary sets of peptides; in one set, amino acids from the HA epitope were substituted into the matrix sequence, and in the other set, amino acids from the matrix epitope were substituted into the HA determinant. It was demonstrated that the hybrid peptides stimulated the appropriate T cell clone corresponding to its epitope specificity (Rothbard et al., 1988). Induction of T cell immunity may contribute to protection against HIV for several reasons, including the fact that neutralizing antibodies coexist with progressive infection and disease. This may be explained by the fact that a major mode of viral transmission within the infected patient is cell to cell ( L i h n et al., 1986; Sodroski et al., 1986). Therefore, a vaccine that primes Th cells for production of lymphokines that augment natural killer cell activities as well as virus-specific cytotoxic T lymphocyte immunity may be essential for an effective AIDS vaccine. In the murine system, two T h cell sites have been defined within the envelope protein gp120 of HIV, using synthetic peptides, namely, residues 428-443 and residues 112-124 (Cease et al., 1987). These peptides were stimulatory in mice of the H-Zksssd haplotypes but not of the H-Zb haplotype. Surprisingly, these two gp120 peptides, defined in a murine system, were recently shown to elicit T cell proliferative responses from a majority of gpl60-immunized (recombinant vaccinia virus) human subjects (Berzofsky et al., 1988). Of 14 vaccine recipients, 8 responded to residues 428-443 and 4 responded to residues 112-124. The reason so
220
DAVID R. MILICH
many individuals responded to the same sequence is unknown, and no correlation between HLA type and responsivenesswas apparent. However, it is possible that each individual’s T cells actually recognized unique sites within the 428-443 sequence. A similar peptide, p426-450, was shown to be highly reactive in immunized nonhuman primates (Krohn et al., 1988) and gibbon apes infected with HIV (Lusso et al., 1988). A CTL recognition site was also defined in a murine model, and consisted of residues 315-329 within gp160 and was recognized in association with H-2 Dd molecules (Takahashi et al., 1988). This CTL site occurs in a region of great interisolate variation, similar to the sites recognized by neutralizing antibodies, and would present similar problems to the design of a synthetic vaccine. Using a quite different approach, Siliciano et al. (1988) cloned HIV gpl2O-specificT cells from seronegative donors using a soft-agar cloning technique. Two HLA-DR4-restricted T h cell clones were derived which recognized residues 410 -429 within the gp120 sequence. Although this determinant partially overlaps the murine epitope 428-443, the sites were determined to be distinct. Analysis of a series of variant isolates, which possessed from 1 to 10 amino acid changes within segment 410-429, showed that the majority of variants were sigdicantly less active relative to the reference sequence (Siliciano et al., 1988). A very interesting observation was that one of the CD4+ T h cell clones specific for gp120 residues 410-429 was cytotoxic toward class 11-positiveAPCs pulsed with gp120. Furthermore, this T cell clone was able to lyse Ia-positive, uninfected CD4+ T cells in the presence of gp120, but not CD8+ T cells. Presumably the ability of the CD4+ T cell subset to present gp120 is due to the presence of a high-affinity receptor for HIV gp120 on CD4+ T cells; this receptor is lacking on the CD8+ subset of T cells. This was further indicated by the ability of the gpl2O-specific T cell clone to lyse both C D 4 + - and CD8+ -activated T cells pulsed with the synthetic peptide 410-429, which can apparently interact with HLA-DR4 molecules on both T cell subsets quite efficiently. The authors suggest that this cytotoxic mechanism may explain, at least partially, the depletion of CD4+ T cells during HIV infection (Siliciano et al., 1988). This important study also suggests that immunization with intact a 1 2 0 may generate gpl20-specific T cells with cytotoxic potential, which may be important for protection against HIV infection. However, the down side of priming gpl2O-specific CD4+ T cells through immunization is the possibility that these cells may mediate the depletion of uninfected CD4+ T cells if an infection occurs. Similar to HIV, there is no inbred animal model for hepatitis B virus infection. Therefore, the studies performed in animal systems have
SYNTHETIC T AND B CELL SITES
221
examined HBV proteins as immunogens rather than as an infectious agent. The hepatitis B surface envelope (HBsAg) is composed of a major polypeptide, P25, and larger polypeptides which share the 226 amino acids of P25 (S region; 175-400) at the C-terminus and possess additional N-terminal residues (pre-S region; 1-174). The GP33 consists of the P25 sequence plus an N-terminal 55 residues [pre-S(2) region; 120-1741, and P39 consists of the GP33 sequence plus an N-terminal 119 residues [pre-S(1) region; 1-1191 (Heermann et al., 1984; Machida et al., 1984). The nucleocapsid of HBV is a 27-nm particle composed of multiple copies of a single polypeptide (P21), which exhibits hepatitis B core antigenicity (HBcAg). It was demonstrated that the murine immune response to the S region of HBsAg is regulated by at least two H-2-linked immune response genes, and S region nonresponder phenotypes (H-2f9were identified (Milich and Chisari, 1982). This observation has been extended to the human immune response to HBsAg by the reports of an association between HLA phenotype and nonresponsiveness to HBsAg vaccination (Walker et al., 1981; Craven et al., 1986). More recent studies have shown that the T cell response to the pre-S(2) region is regulated by H-2-linked Ir genes independently of the S region response, such that immunization of S region nonresponder, pre-S(2) region responder mice (H-29 with HBsAg/GP33 circumvents nonresponse to the S region (Milich et al., 1985a,b). Furthermore, the immune response to the pre-S(l) region is influenced by H-2-linked Ir genes independently of the S and pre-S(2) region responses (Milich et al., 198613). Therefore, HBsAg/P39 immunization in S region and pre-S(2) region nonresponder mice (H-29 elicits an anti-pre-S(1)-specificresponse as well as anti4 and anti-preS(2) antibody production. The mechanism responsible for circumvention of nonresponsivenessis mediated at the T cell level, as T cell recognition of a single region is sufficient to provide T cell helper function for multiple B cell specificities present on S , pre-S(2), and pre-S(1) regions of HBsAg (Milich et al., 1985b, 1986b). These studies indicate that the envelope polypeptides of the HBV present an array of T cell determinants to the host’s immune response and that the specificity of the recognition process is influenced by major histocompatibility complex-linked genes. Additional studies have been aimed at identifying T cell recognition sites within the S and pre-S regions of HBsAg. In the murine system, two sites restricted by the I-AS molecule were defined with synthetic peptides at residues 212-226 and 269-283 within the S region. Two additional sites relevant to the H-Zk haplotype were mapped to residues 284-311 and 314-328 (Milich et al., 1985~).In the human system, a
222
DAVID R. MILICH
dominant T cell site for DPw4-positive individuals was localized to residues 193-202 within the S region using T cell clones derived from vaccine recipients (Celis et al., 1988). Human peripheral T lymphocytes from two vaccine recipients were shown to recognize either 212-226 and 284-311 or 222-255, which indicated that variability in T cell recognition exists between individuals and that human and murine T cell fine specificities could overlap (Milich et al., 1985~). A dominant antibody-binding region within the pre-S(2) region of HBsAg has been localized to residues 133-143 (Milich et al., 1986~); however, this site did not activate T cells and the unconjugated peptide is nonimmunogenic. The N-terminal sequence of the pre-S(2) region, 120-132, was shown to possess an efficient T cell recognition site for several H-2 haplotypes (Milich et al., 1986a). In fact, the composite peptide containing both a T cell and a B cell site proved to be quite immunogenic, and elicited high-titer antibody specific for p133-143 which was highly cross-reactivefor native HBsAg/GP33 (Milich et al., 1986~).This demonstrated that synthetic T and B cell sites could be combined to yield an efficient immunogen. However, this synthetic immunogen lacked an important characteristic, namely, that the synthetic T cell site did not prime T cells relevant to native HBsAg/GP33 in any murine strain tested to date. In contrast, the pre-S(2) sequence 126-140 was reported to elicit T cell proliferation in a number of HBsAg vaccine recipients (Steward et al., 1988). The caveat to this finding was that no pre-S(2) antigen could be detected on the HBsAg preparation used for immunization. Recently, several murine T cell recognition sites within the pre-S(1) region were identified (i.e., residues 12-21, 53-73, and 94-117) which did prime native HBsAg/PJg-specific T h cells (Milich et al., 1987a). Linkage of the pre-S(l) T cell site (12-21) with the pre-S(2) B cell site (133-143) did yield a functional synthetic immunogen relevant to native antigen-specific T and B cells (Milich et al., 1987a). Predictably, this composite peptide is immunogenic only in the strains bearing the H-2 haplotypes (H-2spf)which recognize the synthetic T cell site. A similar construction of a synthetic immunogen containing both T and B cell recognition sites from the homologous antigen has been reported in the malaria circumsporozoite system (Good et al., 1987). Synthetic immunogens containing T and B cell sites from heterologous systems have also been reported (Francis et al., 1987b; Leclerc et al., 1987; BorrasCuesta et al., 1987). However, the heterologous systems lack the advantage of priming both T and B cell memory to the pathogen of interest. More recently, a synthetic T cell site from the nucleocapsid of HBV was used as a carrier moiety in combination with a synthetic pre-S(2)specific B cell epitope. Immunization with this synthetic construction
SYNTHETIC T AND B CELL SITES
223
yielded envelope-reactive antibody production, and primed memory T cells specific for the nucleocapsid protein. This result demonstrated the feasibility of constructing complex synthetic immunogens that represent multiple proteins of a pathogen and are capable of engaging both T and B cells relevant to the native antigens (Milich et al., 1988). A pre-s(1)-specificT cell site recognized in association with HLA-A11 was recently mapped to residues 21-28 using synthetic peptides and a human T cell clone derived from a vaccine recipient (Jin et al., 1988). Although this was a class I-restricted clone, it was shown to respond to both endogenous and exogenous antigen and was not sensitive to leupeptin or chloroquine inhibition of antigen presentation. This suggested that a nonendosomal processing pathway is accessible to both exogenous and endogenous antigen. In a reciprocal experiment, a class II-restricted HBsAg-specific T cell clone was totally inhibited by leupeptin whether the antigen source was endogenous or exogenous. This surprising result suggests that the endosomal pathway is also accessible to exogenous and endogenous antigen. These results have implications for vaccine design, because a noninfectious viral component vaccine was demonstrated to elicit a class I-restricted T cell response capable of recognizing endogenously produced (i.e., viral infection) antigen (Jin et al., 1988). A number of pre-S(1)-specific antibody-binding sites have also been recently elucidated (Milich et al., 1987a). Identification of distinct T cell and B cell recognition sites within the pre-S region has allowed examination of the ability of a T cell population specific for a single peptide to provide functional T cell help for a series of B cell specificities on HBsAg/P39. For this purpose, mice were primed with the pre-S(l) T cell determinants p12-21 or p94-117 and were challenged 4 weeks later with a suboptimal dose of HBsAg/P39 particles: the fine specificities of the antibodies produced were then measured. Priming with p12-21 elicited T cell helper function, resulting in in vivo antibody production to a set of antibody specificities in the pre-S region, and group- and subtypespecific determinants in the S region. Similarly, priming with p94-117 elicited antibody production specific for a set of antibody specificities in the pre-S region but did not prime antibody production to the S region determinants. These results indicated that T cells primed to a single determinant are sufficient to provide functional help to multiple B cell clones which recognize unique epitopes on a complex HBsAg particle. It was noteworthy that p12-21 and p94-117 primed antibody production to unique as well as common B cell determinants. These data provide strong evidence that the fine specificity of the T helper cell can influence the fine specificity of the antibody produced (Milich et al., 1987a). The molecular mechanism whereby T cells can provide differential help for
224
DAVID R. MILICH
B cell clones with differing specificities on the same polypeptide is difficult to explain in the context of T cell-B cell interaction models; however, similar phenomena have been previously observed in other antigen systems (Checka et al., 1976; Berzofsky et al., 1977; CamposNet0 et a l . , 1982; Celada et al., 1984). Berzofsky (1983) has proposed a T cell-B cell reciprocity circuit in which B cell receptor-antigen-Ia interactions may limit T cell specificity, which in turn limits B cell specificity. Whatever the mechanism, the availability of defined T and B cell sites represented by synthetic peptides has allowed a thorough examination of a heretofore controversial issue. Examination of the fine specificity of T cell recognition of the nucleocapsid of HBV in a murine system revealed multiple but distinct T cell sites among a panel of H-2 congenic strains. Each strain recognized a predominant T cell determinant dependent on the H-2 haplotype of the responding strain. For example, H-2S strains recognize residues 120-131 predominantly, H-2b strains recognize 129-140, H-2f strains recognize 100-110, and H-2s strains recognize residues within 100-120 (but distinct from H-2f strains), and H-2d strains predominantly recognize residues 85-100 (Milich and McLachlan, 1986; Milich et a l . , 1987b, 1988). Using synthetic peptides it was shown that these T cell recognition sites could prime functional T h cell activity and induce antiHBc antibody production in wvo after challenge with suboptimal doses of native HBcAg. This was a direct demonstration that synthetic T cell sites, which elicit no antinative antibody production themselves, can nonetheless prime Th cells which can be recalled by native antigen. This may explain a number of observations from the earlier synthetic peptide literature in which the term “priming effect” was used to explain the absence of antibody production after peptide immunization but enhanced antibody production after the challenge with native antigen (Emini et al., 1983;Jacob et al., 1986; Shapira, 1987). Possibly, the peptides used contained T cell as well as B cell determinants. Another finding which emerged fmm the functional studies of HBcAgderived peptides was that HBcAg-specific T h cells, elicited by peptide immunization, can help B cells produce antibody specific for envelope antigens as well as for HBcAg, even though these antigens are found on separate molecules within the virion (Milich et al., 1987~).Mice primed with a synthetic T cell site and challenged with a mixture of HBcAg and HBsAg produced no anti-HBs. However, peptide-primed mice challenged with intact virions produced antibody to the S, preS(2), and pre-S(l) regions of the envelope. This was a demonstration of intermolecular/intrastructural T cell helper function because the antigens were on separate molecules but within the same particle. This
SYNTHETIC T AND B CELL SITES
225
phenomenon has important implications for synthetic vaccine design because multiple T cell sites can be combined within a natural or artificial particle and serve as T cell carrier moieties for any B cell epitope(s) on the surface of the paricle. Due to the limits imposed on T cell antigen recognition by MHC restriction, it is becoming obvious that a number of T cell sites will be required to ensure that a synthetic vaccine will engage the majority of an outbred population. Presentation of a number of short synthetic T cell sites within a particulate structure may be an economical and feasible way to accomplish this task. Herpes simplex virus (HSV) is one of the most common infectious agents of man. One important feature of this virus is its ability to produce a latent infection. The reactivation of this infection, however, generally does not result in a viermia, but rather infection of additional cells. Thus, it might be expected that a cellular mechanism of protection is required. Heber-Katz et al. (1985) have defined residues 1-23 of the glycoprotein D (gD) as a T cell site. This group then defined two T cell sites on the glycoprotein at residues 1-16 (I-Ak restricted) and 8-23 (I-Ekrestricted) (Heber-Katz and Dietzschold, 1986). The T cell responses appear to have the same fine specificity as B cell responses to this protein (Cohen et al., 1984; Dietzschold et al., 1984). As noted previously, this is an unusual finding because T and B cell sites generally do not overlap (Lamb and Green, 1983; D. W. Thomas et al., 1981; Maizels et al., 1980; Corradin and Chiller, 1979; Berkower et al., 1984; Hurwitz et al., 1984; Milich et al., 1984b, 1985c, 1986~).Finally, it was shown that the I-Ak-restricted peptide, consisting of residues 1-16, is helical, but residues 8-23, which are I-Ek restricted, are nonhelical. The I-Ek response was found to be highly MHC degenerate in that T cell hybridomas specific for the 8-23 peptide responded to antigen on APCs derived from BlO.A, BlO.A(5R), and BlO.A(SR) mice, and also showed differences in antigenic fine specificity with APCs of different haplotypes (Heber-Katz et al., 1988). Watari et al. (1987) showed that residues 1-23 could induce a strong T cell response when injected into mice, but did not confer protection. When this peptide was covalently coupled to palmitic acid and incorporated into liposomes, it induced a virus-specific T cell response that conferred protection against a lethal challenge of HSV in the absence of antibody, and this protection by T cells could be adoptively transferred. It is also noteworthy that the N-terminal peptide, residues 1-23 of HSV gD, activated a panel of human peripheral blood T cells from HSV-seropositiveand -seronegativeadult donors, but not a panel of T cell lines from seronegative children. It was suggested that all adults have been exposed to HSV and retain T cell memory. It was also observed that each adult donor demonstrated a unique pattern of T cell reactivity
226
DAVID R. MILICH
against a panel of truncated and substituted analogs of residues 1-23, which suggested that the human T cell response was variable, as predicted by the MHC restriction of T cell recognition (DeFreitas et al., 1985). Epstein-Barr virus (EBV) is a human herpes virus that infects the majority of the human population and is associated with disorders such as infectious mononucleosis, Burkitt’s lymphoma, and nasopharyngeal carcinoma. One of the sites for maintenance of persistent infection is within peripheral blood B cells. It has been shown that T cells from seropositive individuals suppress EBV-infected B cell outgrowth in vitro (Thorley-Lawson et al., 1977; Moss et al., 1978). Furthermore, this phenomenon is mediated by MHC -restricted, EBV-specific CTLs (Sugamura and Hinuma, 1980; Misko et al., 1980; Wallace et al., 1982; Slovin et al., 1983). Recently ThorleyLawson and Israelsohn (1987) have defined a CTL recognition site located within residues 43-53 of the latent membrane protein (LMP) encoded by EBV and present in the plasma membrane of transformed B cells (Hennessy et al., 1984). This represents the first definition of a target structure for a human EBVspecific CTL. The possible role of cell-mediatedimmunity in rabies infection is poorly understood. Rabies, because of its prolonged incubation period, is the only viral disease in which postexposure vaccination is routinely practiced. However, the mechanisms underlying its mode of action have remained obscure. Analysis of the cellular immune response in rabies may elucidate some of the critical events occurring in the rabies virus-host cell interactions and may help in establishing the most effective protocol for postexposure treatment. Toward this end, MacFarlan et al. (1984) have identified residues 18-44 of the rabies virus glycoprotein as a T h site in the A/J murine strain (H-2a). Lymphocytic choriomeningitis virus (LCMV) can establish a persistent infection when acquired in utero or neonatally, or when the host cellular immune response is compromised. The virus then persists throughout the lifetime of the host; the severity of the resultant virus-related pathology varies depending on the mouse strain used. In contrast, infection of an immunocompetent adult mouse results in the rapid induction of an efficient MHC class I-restricted cytotoxic T lymphocyte response, with consequent virus clearance. Whitton et al. (1988) have defined a LCMVspecific CTL site which is H-2b, class I-restricted at residues 272-293 within glycoprotein 2 (GP-2). Another interesting finding which emerged from this study was that mice of H-2dor H-29 haplotype exhibited a different gross pattern of CTL reactivity as compared to the H-2b haplotype. These two haplotypes demonstrated CTL activity toward the nucleoprotein and not to either glycoprotein (GP-1 or GP-2). Therefore,
SYNTHETIC T AND B CELL SITES
227
the MHC phenotype determined not only the T cell site within a protein, but in this case, the viral protein recognized by CTLs. This implies that the viral proteins as well as the specific T cell determinants chosen for a subunit vaccine must be considered carefully and with some understanding of the specific system. Foot and mouth disease virus (FMDV) is the causative agent of the most economically important infectious disease of farm animals. A synthetic peptide corresponding to the amino acid sequence 141-160 of protein VP1 from FMDV serotype 0 1 coupled to keyhole limpet hemocyanin (KLH) was shown to elicit neutralizing antibodies that could protect guinea pigs against experimental infection (Bittle et al., 1982). Subsequently, Francis et al. (1985) observed that immunization with 141-160/KLH primed guinea pigs for an FMDV serotype-specific neutralizing antibody response to a second subimmunizing dose of the same peptide. The secondary response was not dependent on the carrier, since uncoupled 141-160 or 141-160/tetanus toxoid could also boost the response. It was therefore suggested that residues 141-160 contained a Th cell determinant as well as a neutralizing antibody site. This observation was confirmed in a subsequent study and the minima1 T cell site was defined as residues 150-160 within VPl (Francis et ad., 1987a). This T cell site was later shown to be restricted to the H-2k and H-2' haplotypes in a murine system (Francis et al., 1987b). As mentioned previously, tobacco mosaic virus protein (TMVP) was the first system in which small peptide fragments were used to induce neutralizing antisera (Anderer, 1963). Immunization with TMVP in many species leads to antibody production to several areas of the protein. One such area is a tryptic fragment representing residues 93-112 of TMVP, and specifically antibodies bind to the C-terminal decapeptide (Benjamini, 1988). Subsequently, it was demonstrated that residues 93-112 could also stimulate a T cell proliferative response in H-2a mice (Wan et al., 1986). It was further noted that immunization of some strains with residues 93-112 induced T cells capable of responding to 93-112 or TMVP in vitm, and these strains were referred to as crossreactive (CR). Other strains did not exhibit cross-reactivity and were referred to as noncross-reactive (NCR). Crosses between CR and NCR strains yielded CR offspring. However, T cells from CR F1 animals were only cross-reactivein the presence of macrophages from CR or F1 mice (Benjamini et al., 1986). It was then found that T cells of CR but not NCR mice recognized the C-terminal portion of the peptide, residues 97-112. These investigators also showed that nonstimulatory fragments of 93-112 could be converted to an antigenic peptide by conjugation to various protein carriers. It was suggested that this mechanism may be
228
DAVID R. MILICH
of benefit to vaccine design in the circumstance in which one wished to induce T cells to an area of a protein that is normally not recognized by T cells (Wan et al., 1986). The N-terminal domain (Pl-102) of bacteriophage X CI repressor (cI) and many mutant forms of the domain are readily available, and certain mutants with single amino acid substitutions can be used for defining antigenic determinants. Additionally, CIis not likely to contain only nonself antigenic determinants. Guillet et al. (1986) defined an immunodominant T cell site within residues 12-26 in association with the I-Ad molecule, and residues 15-26 were also stimulatory. It was found that the truncated analog p12-24 could inhibit the activation of one T cell hybridoma but not another following stimulation with 15-26. This suggested that the T cell was involved in the competition of peptides for binding to I-A, most likely by enhancing the avidity of the active peptide (15-26) for I-A (Guillet et al., 1986). It was next observed that two other I-Ad-restrictedT cell sites [staphylococcal nuclease (Nase) 61-80; ovalbumin (Ova) 824-3361 unrelated to CI could also inhibit activation of cI-specificT cell hybridomas by p15-26. The Nase peptide sequence showed remarkable homology with p15-26, whereas the Ova peptide had little similarity. These findings suggested that all peptides restricted by I-Ad may bind to the same site on I-A. These investigators then showed that cI-specific T cell hybridomas from two strains (BALB/c and A/J) recognized determinants within the same 12-amino acid sequence, 15-26 (Lai et al., 1987). However, I-Ad (BALB/c)-restricted and I-Ek (A/J)-restricted T cell responses illustrated different patterns of fine specificity, suggesting that the same peptide is presented in different configurations by different Ia molecules. Furthermore, cI-specific T cells of these two strains were not cross-reactive. C. PARASITE ANTIGENS Much effort is currently being devoted to vaccine development as a possible means of preventing malaria. The malaria parasite (Plasmodium species) is carried by mosquitoes that inoculate the mammalian host with sporozoites. The sporozoites migrate to the liver. If this migration could be blocked before the sporozoites entered hepatocytes, the liberation of the blood stage merozoites from the liver and thus disease could be prevented. Vaccines directed against sporozoites are based on surface circumsporozoite (CS) proteins, which are found on all malarial sporozoites. The CS protein contains an immunodominant repeating epitope, and antibodies to this repeating epitope block infection (Potocnjak et aZ., 1980; Nardin et al., 1982; Ballou et al., 1985; Mazier et a l . , 1986).
SYNTHETIC T AND B CELL SITES
229
A murine T cell site was identified within the repeating epitope; however, only mice carrying the I-Ab gene could recognize this site (Good et al., 1986). In a human vaccine trial using the repeat epitope linked to tetanus toxoid, a similarly weak repeat sequence-specificT cell proliferative response was noted (Etlinger et al., 1988). Because a malaria-specific T cell memory response, and possibly antibodyindependent T cell immunity may be required for protection, various groups have searched for additional T cell recognition sites on malarial proteins. In the murine system, T cell sites were localized to residues 326-343, restricted to I-Ak, and residues 330-343, restricted to I-Ad, within the CS protein (Good et al., 1987, 1988a). Site 326-343 was shown to represent a T h cell site as it primed antibody production to the repeat B cell epitope. Furthermore, the synthetic T cell site was linked to the synthetic B cell epitope to construct a totally synthetic immunogen (vaccine) (Good et al., 1987). Mice (I-Ak) unable to respond to the B cell epitope alone responded to the composite immunogen, with antibody production to the repeat epitope. It was shown that CD8+ T cells may play a role in immunity to malaria infection, because depletion of CD8 T cells from a sporozoite-immunized animal completely abrogated immunity (Schofield et al., 1987; Weiss et al., 1988). Subsequently, Kurnar et al. (1988) demonstrated the existence of CD8+ CTLs that recognized the CS protein, specifically residues 368-390, in BlO.BR (H-2k) mice immunized with sporozoites. Human T cell recognition sites have also been defined on the CS protein using synthetic peptides and peripheral blood lymphocytes or cloned T cell lines from malaria-immune donors. Three immunodominant T cell determinants were localized outside the repetitive region to residues 300-310, 326-345 (similar to the murine site), and 361-380 (Good et al., 1988a). It was noted that these domains occurred in polymorphic regions of the CS molecule, suggesting that parasite mutation and selection may have occurred in response to pressure from T cell immunity. Such polymorphism may present a problem for vaccine development (Good et al., 198813). Another group defined residues 103-122 and 378-398 within the CS protein as additional human T cell recognition sites (Sinigaglia et al., 1988). In studies aimed at developing a merozoite-based vaccine, two human T cell recognition sites were defined within a nonpolymorphic region of gp190 of Plasmodium falciparum merozoites to be residues 225-237 (Crisanti et al., 1988). These authors examined peptide-specific T cell clones for the ability to respond to 11 different P falciparum isolates in culture, and the identified T cell determinants or cross-reacting sequences were present in a11 isolates tested. Identification of invariant T cell recognition sites of +
230
DAVID R. MILICH
the gp190 protein has implications for the design of a synthetic merozoitebased vaccine. Three human T cell recognition sites were also identified on another merozoite antigen, PFl55/RESA, localized to the C-terminal repeat region of this protein (Kabilan et al., 1988). This region of the PFl55/RESA antigen is also known to contain immunodominant linear B cell epitopes (Berzins et al., 1986; Perlmann et al., 1987). These characteristics suggest that this region of the protein may serve as a very attractive candidate for synthetic vaccine development.
D. BACTERIAL ANTIGENS Staphylococcal nuclease has proved to be a useful antigen for the study of the genetics of the immune response. It is simple in composition (149 residues), has no disulfide bridges, and its three-dimensional structure at high resolution is known. In previously studied systems (i.e., cytochrome c, myoglobin, lysozyme, ovalbumin) it appeared that only a small number of antigenic determinants on the overall protein were recognized by T cells. Also a strong preferential association appeared to exist between the antigenic determinant recognized and the self-Ia molecule. Each of the previously studied antigens differed from the homologues expressed in the responding experimental mice by only a few amino acids. It was possible that the apparent limitations in the number of T cell sites on a protein may reflect the influence of selftolerance to most of the determinants on closely related molecules. Because of its evolutionary distance from mammalian proteins, Nase is ideal for evaluating the diversity expressed within the T cell repertoire. Toward this end, Finnegan et al. (1986) have identified four T cell recognition sites on Nase using synthetic peptide analogs of this protein. Residues 81-100 and 51-70 were recognized by T cell clones in association with I-Ak; p91-110 was I-Ab-restricted and H-2d mice recognized peptide 61-80. The fine specificity of T cell recognition of the nucleoprotein of HBV is reminiscent of this pattern (Milich et al., 1987b). In both systems, which represent very foreign proteins, fine specificity of T cell recognition is dependent on the H-2 haplotype of the responding strain, and although the T cell sites are distinct, they are concentrated within a region of the molecule. These findings suggested that T cell recognition of a complex and highly foreign protein antigen was limited to a small number of peptide epitopes, which are preferentially recognized by T cells in association with a given Ia molecule. This indicated that tolerance is not a universally important component in the limited repertoire of T cells for sites on foreign proteins. Pathogenic mycobacteria are the cause of widespread chronic diseases, particularly in developing countries, with an estimated 30 million
SYNTHETIC T AND B CELL SITES
231
individuals suffering with tuberculosis and a further 10-15 million with leprosy (Bloom and Godal, 1983). The T cell response to mycobacterial antigens is thought to determine the outcome of mycobacterial infection. The gene for a major mycobacterial antigen, the 65-kDa protein, has been cloned, and T cell clones from patients with tuberculosis or leprosy have been shown to proliferate in response to the 65-kDa protein, which is highly conserved in the two mycobacteria (Emmrich et al., 1986; Lamb et a l . , 1986; Oftung et al., 1987). Using a combination of synthetic peptide technology and a recombinant DNA expression system, Lamb et al. (1987) have mapped several human T cell determinants within the 65-kDa protein. Residues 112-132 and 437-459 were identified as human T cell determinants. These two regions are identical in Mycobacterium leprae and Mycobacterium tuberculosis and are distinct from the known B cell epitopes of the 65-kDa protein (Mehra et al., 1986). Adjuvant arthritis is a chronic disease inducible in rats by immunization with M . tuberculosis (Pearson, 1964). Rat T cell clones specific for an epitope on M . tuberculosis also recognized a fraction of cartilage proteoglycan (van Eden et al., 1985). Furthermore, a fraction of M . tuberculosis containing the epitope that is cross-reactivewith cartilage was also recognized by T cells from patients with rheumatoid arthritis (Holoshitz et al., 1986). Van Eden et al. (1988) recently demonstrated that the epitope recognized was present on a Myco bacterium boms BCG antigen of 65-kDa. Use of synthetic peptides enabled this group to define the critical T cell determinant as residues 180-188. An amino acid homology between the 180-188 sequence and the link protein of rat proteoglycan was found. Furthermore, immunization with the 65-kDa antigen induced resistance to adjuvant arthritis induction. The possible therapeutic suppression of the disease with the 180-188 determinant and a relationship between bacterial infections and susceptibility to autoimmune arthritis were suggested (van Eden et al., 1988). The serotype-specific M protein emanating as a-helical coiled-coil fibrils from the surface of Streptococcus pyogenes is the major virulence determinant of these organisms. The fibrinogen-binding and antiopsonic properties of these fibrils is inhibited by type-specific antibodies directed against epitopes of M protein (Whitnack and Beachy, 1982; Lancefield, 1962). In addition to protective epitopes, certain M proteins contain epitopes cross-reactive with host tissues, especially cardiac tissue (Dale and Beachy, 1982), which explains the association of streptococcal infections with rheumatic heart disease. Therefore, a synthetic peptide vaccine containing protective epitopes and excluding those epitopes cross-reactive with host tissues would be desirable (Beachey et al., 1981). These
232
DAVID R. MILICH
investigators have shown that the N-terminal region of types 5, 6, and 24 M proteins contain protective, but not tissue cross-reactive, epitopes (Beachey et al., 1987). Studies to define T h sites which may be effective in vaccine development have focused on the M protein. A trivalent peptide containing the N-terminal sequences of types 5, 6, and 24 M proteins was synthesized -M5(l-lO)/M6(l-ll)/M24(1-12). This trivalent peptide elicited type-specificprotective immunity. Furthermore, this composite peptide was shown to sensitize rabbit T lymphocytes to respond to each of the natural M proteins as well as to the synthetic, trivalent peptide. These findings have implications for the construction of safe and effective synthetic vaccines against a variety of different serotypes of S. pyogenes. A limited number of human T cell epitopes have been defined, mostly due to the inability to immunize humans for experimental purposes. The recent cloning and expression of tetanus toxin fragments and the fact that most humans have been immunized against tetanus toxoid make this antigen an ideal model for studies of human T cell recognition, in addition to its application to synthetic peptide vaccine design. Recently, Demotz et al. (1988) defined two human T cell recognition sites on tetanus toxin, residues 1273-1284, in association with HLA-DR52a, and residues 830-843, most likely in association with HLA-DR5.
E. T CELLSITESWITHIN OTHERANTIGENS Small peptides for which three-dimensional and structural analogs are available represent appropriate systems to investigate processing requirements and Ia- antigen interaction for presentation to T cells. Apamin, an 18-amino acid natural peptide of bee venom, is such a molecule owing to its compact conformation with an a-helix core surrounded by two 6-turns stabilized by two disulfide bonds (Bystrov et al., 1980; Freeman et al., 1986). Regnier-Vigouroux et al. (1988) established that both antibody and lymph node T cell proliferative responses to this peptide were controlled by H-2-linked Ir genes. Responder mice use Ad or Ab class I1 molecules as restriction elements. They further demonstrated that apamin-specific T cell hybridomas required antigen processing of this small peptide antigen by APCs for stimulation. Additionally, the simple unfolding of apamin was sufficient to eliminate the need for antigen processing. Therefore. this natural peptide, which is an appropriate size for T cell triggering, acquires its antigenic conformation after processing by APCs, which primarily involves an alteration of disulfide bond-dependent peptide folding. In another system using a small, well-definednatural peptide antigen, the octapeptide hormone angiotensin I1 (AII) has been employed to
SYNTHETIC T AND B CELL SITES
233
investigate T and B cell recognition. Early studies focused on the specificity and spatial constraints of antibody binding to A11 (Vallotton, 1974). It was also shown that unconjugated A11 elicited immediate and delayed-type hypersensitivity reactions in guinea pigs (Dietrich, 1966). These findings were extended by D. W. Thomas et al. (1981), who examined Ir gene control and the specificity of T cell responses to a variety of synthetic homologues and analogs. The genetic control of T cell responses to these peptides was found to be highly specific and capable of distinguishingsubtle differences in the antigens. For example, strain 2 guinea pigs responded to A11 but were low responders of A11 (Val5), whereas, strain 13 guinea pigs responded to A11 (Val5) but not A11 (D. W. Thomas et al., 1981). The T cell responses of these two strains were differentiating between one methyl group between Val5 and Ile5. Another well-defined natural peptide antigen used in a guinea pig system to examine T cell recognition and Ia-antigen interaction is the 14-residuepeptide antigen, human fibrinopeptide B (hFPB). Fibrinopeptide B is cleaved from the 0-chain of fibrinogen during coagulation. It was found that strain 2 guinea pigs responded to hFPB (Bpl-14), and the minimum immunogenic fragment was B07-14. However, strain 2 animals failed to respond to the des-ArghFPB homologue (Bm-13). In contrast, strain 13 animals produced T cell responses to Bpl-1’3, but not to Bpl-14. Thus, the genetic control of T cell responsiveness was determined by a single residue, Arg14 (Thomas et al., 1979). Subsequent studies were performed using analogs containing substitutions at residues within the B05-14 peptide. These elegant studies showed that substitutions at positions 10 and 14 remained immunogenic but T cell responses elicited were not cross-reactive with the parent peptide, BP5-14. Reciprocally, replacing residues 7, 8, or 11 abrogated or dramatically reduced immunogenicity (Thomas et al., 1980). The authors proposed that residues 10 and 14 contact the T cell receptor, whereas Ia interacts with residues 7, 8, and 11. These investigators then examined the effects of inverting the amino acid sequence of BP7-14. The inverted peptide was nonantigenic and nonimmunogenic, suggesting that T cell recognition involves more than simple interactions with amino acid side chains and that the ordering of the amino acids within the peptide fragment is critical (Thomas et al. 1982). Studies of this antigen in a murine system revealed that H-2kmice are responders to Bpl-14 in terms of T cell proliferation, whereas H-2bmice are not (Peterson et d., 1983). However, both H-2k and H-2b mice demonstrated a helper T cell response to Bpl-14, and the fine specificity of the Th cell response differed from the T cell proliferative response. The authors concluded that Th and T proliferative cells may be distinct subpopulations that express
234
DAVID R. MILICH
differences in genetic control and fine specificity. However, another explanation may be related to quantitative differences between the sensitivity of assays for T cell proliferation versus T h cell function. More recent studies of the T cell response to fibrinogen revealed a T cell recognition site within the A, chain of the molecule at residues 571-578, which was I-ELrestricted (Lee et al., 1988). The majority of studies of T cell recognition have revealed that some degree of antigen processing is required. The processing requirement ranges from simply unfolding the polypeptide chain to proteolytic cleavage. Although 340,000), it did not fibrinogen is a very large nonglobular protein (M, require processing to activate two I-Ek-restricted, fibrinogen-specific murine T cell hybridomas. The explanation for this observation appears to be the fact that the determinant recognized (571-578) was located to the carboxy-terminal portion of the A, chain, which has no defined secondary structure and may possess the necessary conformational flexibility to directly associate with an 1-Ekmolecule (Lee et al., 1988). It was previously shown that class I-restricted CTLs can recognize processed antigenic fragments of the nucleoprotein (NP) of the influenza virus (Townsend et al., 1985, 1986). To determine if CTLs recognize peptide fragments on antigens normally expressed as integral membrane proteins at the cell surface (for example, allogenic MHC molecules), Maryanski et al. (1986) have isolated class I-restricted mouse CTL clones that recognize class I gene products of the human MHC (HLA-CW3) as antigens in mouse cell HLA transfectants. They showed that these anti-CW3 CTLs can lyse HLA-negative syngeneic mouse cells in the presence of a synthetic HLA-CWS peptide (residues 171-186) very efficiently. They then showed that antigens recognized by Kd-restricted CTLs specific for HLA-A24 can also be mimicked by synthetic peptides (Maryanski et al., 1988). These results are consistent with the viral model systems and indicate that MHC class I molecules bind protein fragments or peptides, which results in specific recognition by CTLs. In another study of the molecular nature and involvement of MHCbound peptides as antigens recognized by alloreactive T cells, Clayberger et al. (1987) examined the effects of peptides derived from the human class I gene product HLA-A2 on cytolysis of HLA-A2-specific CTLs. It was demonstrated that two peptide sequences, A2.56-59 and A2.98-113, could inhibit the killing of HLA-A2-expressingtarget cells. To determine if lysis was inhibited by the peptide interacting on the CTLs or the target cells, either the CTLs or the target cells were pretreated with the peptides. Lysis was inhibited only when the CTLs were pretreated, presumably indicating competition of the peptide with membrane HLA-A2 for the T cell receptor. It was also observed that the
SYNTHETIC T AND B CELL SITES
235
A2.56-69 peptide could sensitize syngeneic target cells for lysis by HLAA2-specific CTLs, which suggested that the peptide bound to the HLAAw69 molecule expressed by the syngeneic target cell, and that HLAAw69 plus peptide mimicked the native structure of HLA-A2 recognized by the CTLs (Clayberger et al., 1987). Using CTLs specific for the a1 domain of the H-2Ld molecule, it was demonstrated that a synthetic a1 peptide (61-85) corresponding to one of the helices of the H-2Ld molecule can be presented by a class I restriction element (Dk) to reconstitute a CTL determinant present on the intact H-2Ld molecule (Song et al., 1988). It has been suggested that the in vivo relevance of these findings relates to the possibility for peptides derived from one HLA molecule to bind to another. The number of T cell determinants which could be generated is potentially large and may in part account for the intensity of the alloresponse (Clayberger et al., 1987). Myasthenia gravis (MG) and experimental autoimmune MG (EAMG) are autoimmune diseases characterized by the production of autoantibodies directed against the nicotinic acetylcholine receptor (AChR) of the neuromuscular junction. The autoantibodies are known to be T cell dependent, which has prompted interest in identifying T cell recognition sites on this autoantigen. McCormick and Atassi (1984) reported that an antibody to a synthetic peptide, residues 125-147, derived from the cr-subunit of the Torpedo electric organ AChR, bound native AChR provided the Cys128and Cys142were disulfide linked. It was then demonstrated that this peptide also stimulated a T cell response in Lewis rats (Lennon et al., 1985). More recently, Fuji and Lindstrom (1988) identified five T cell recognition sites on the a-subunit of the AChR, which were strain dependent. Residues 100-116 and 73-90 stimulated AChR-primed T cells from Lewis rats, fragment 52-70 was active in the Buffalo strain, residues 152-167 stimulated T cells from Wistar-Furth rats, and fragment 172-205 was active in Brown-Norway rats. These findings are consistent with the genetically restricted T cell recognition observed in other systems, and suggests that MG patients with different MHC haplotypes may recognize different AChR determinants. Experimental allergic autoimmune encephalomyelitis (EAE) is a paralytic disease induced by immunization with myelin basic protein (MBP) in suitable adjuvant and has been used as an animal model for demyelinating autoimmune disorders of the central nervous system, such as multiple sclerosis. Basic proteins of guinea pig and rat are highly encephalitogenic, whereas other mammalian basic proteins of nonrodent origin and basic proteins of chicken, turtle, and frog are only weakly active (Martenson et al., 1972). EAE has also been shown to be mediated by class 11-restrictedT cells and is thus a prototype model for the study
236
DAVID R. MILICH
of antigen-specific T helper cell-mediated autoimmune disease (Zamvil et al., 1985a;Pettinelli and McFarlin, 1981;Brostoff and Mason, 1984; Waldor et al., 1985; Pettinelli et al., 1982). Many groups have tested proteolytic cleavage fragments and synthetic peptides of myelin basic proteins for their ability to induce EAE, and these studies have generated much information about the T cell epitopes involved in this disease. Martenson et al. (1975)found encephalitogenic determinants at residues 37-42 and 108-113 in bovine MBP. Chou et al. (1977)identified residues 68-88 of GP MBP as an active site for EAE, which was more narrowly defined as residues 75-80 by Vandenbark et al. (1985a). Vandenbark et al. (1985b)went on to show that the antigen-presenting cell from a given rat strain plays the key role in determining which part of the MBP molecule is recognized preferentially, which implies the association of MBP epitopes with different MHC products. Kibler et al. (1977),determined that a serine at residue 79 of the GP peptide residues 68-88,which differentiates GP from rat MBP, results in the stimulation of greater numbers of T cells involved in (1) the induction of EAE, (2) the in vitm proliferative response, and (3) the helper function in antibody production. Fritz et al. (1985)have mapped the induction of EAE to the I-A subregion of the MHC, and showed that I-As-and I-As-bearingstrains recognize residues 87-169 of mouse MBP, whereas I-Ak- and I-AUbearing strains recognize residues 1-37. Similarly, Zamvil et al. (1986) have identified residues 1-9 and 1-11 on rat BP as being I-Au restricted. Kono et al. (1988)more precisely mapped the encephalitogenic determinants recognized by H-28mice to be residues 87-98 and 91-104.Since both peptides are I-ASrestricted and have a large region of overlapping amino acids (residues 91-98),it is likely that they share sequences which bind to the I-As molecule but have at least two different sets of residues recognized by T cells. This finding of a single encephalitogenic peptide region with multiple T cell epitopes, and the fact that encephalitogenic T cell epitopes may be subdominant, have implications for the design of treatments directed at the T cell receptor-MHC-peptide epitope complex in autoimmune disease. Therefore, different strains of mice respond to different encephalitogenic determinants dependent on H-2 haplotype. Sites showing clonal activation with no disease have been defined at residues 43-88 and 1-37 of GP MBP (Pettinelli et al., 1982). McCarron and McFarlin (1988)studied the three T h sites, 1-37,43-88, and 89-169 of GP MBP in I-AS,I-A', and F1 murine strains. They showed that F1 mice are passively susceptible to EAE induced by adoptive transfer of cells reactive to either the N-terminal or C-terminal fragment, and that the encephalitogenic determinant of GP MBP is related to the MHC
SYNTHETIC T AND B CELL SITES
237
of the macrophage present in vitro. Finally, Kira et al. (1986) defined a T cell recognition site at residues 15-31 of rabbit MBP and Tuohy et al. (1988) have identified an encephalitogenic determinant within the proteolipid protein (PLP) molecule, 103-116, in SWR mice. This is the first report of a synthetic encephalitogenic peptide from myelin PLF? It has been demonstrated that strain 2 guinea pigs respond to the loop region of the bovine insulin A chain (residues A8-10) and are @chain nonresponders (Barcinski and Rosenthal, 1977), whereas strain 13 guinea pigs recognize a determinant on the 0-chain (B5-16) and are nonresponders to the A chain loop (D. W. Thomas et al., 1981). These results indicated that the nature of the antigenic determinant recognized by T cells was dependent on the associative interaction between antigen and Ia. A recent study of T cell recognition of an A chain loop determinant of insulin revealed that this T cell site is actually conformational and requires portions of both the A and B chains, and that residues A1-14 disulfide linked to B7-15 represented the minimal antigen (Naquet et al., 1987). However, antigen processing was still required since fixed APCs were poor presenters of insulin to a panel of I-Ad-restricted T cell hybridomas. The authors suggested that a partial proteolytic cleavage and reduction of insulin, which permits an unfolding of the molecule, give rise to the immunogenic, a-helical amphipathic A1-14-SS-B7-15 peptide. The T cell antigenic site on this peptide most likely involves the glutamic acid residue at A4, since this is the only residue that differs between pork insulin peptide and its mouse homologue (Naquet et al., 1987). The T cell response to ragweed allergen Ra3 has been studied in four murine strains, and a number of T cell sites were identified (Kurisaki et al., 1986). Uniquely, fragment 52-64 was reported to contain a dominant T cell site for all four strains. This site was also reported to activate T cells from several human Ra3-sensitive, DR1-positive donors (Rothbard et al., 1988). Similarly, a large number of T cell recognition sites have been mapped on the 0-chain of hemoglobin A in two murine strains (Yoshioka et al., 1986). IV. Enumeration of B Cell Recognition Sites Defined by Synthetic Peptides
The classical approach to “searching” for B cell recognition sites on a protein using synthetic peptides has been to synthesize a panel of candidate peptides, immunize, and test the antipeptide antisera for reactivity with the immunizing peptide and the native protein. Reciprocally, antibody to the native protein can be screened for reactivity
238
DAVID R. MILICH
on a panel of synthetic peptides. It is important at this point to distinguish between peptide antigenicity and immunogenicity. The term B cell antigenicity refers to the ability of a protein or peptide surface region to bind specifically to an antibody in a test tube, whereas immunogenicity refers to the ability of any antigenic site to elicit antibody production. Antigenicity is primarily an intrinsic characteristic of a peptide region while immunogenicity is dependent in large part on the quality of T h cell function, the immunization protocol, the nature of T cell-B cell interactions, the genetics of the immunized animal, and other extrinsic characteristics. The majority of B cell sites referred to in this review were defined using synthetic peptide antigens conjugated to heterologous carrier proteins, and, therefore, represent studies of B cell antigenicity and not immunogenicity. Antibody-binding sites on proteins have long been categorized into two types, originally designated as sequential and conformational (Sela et al., 1967). Sequential antigenic sites represent residues located on a continguous region of the polypeptide chain, whereas conformational determinants are composed of regions of the protein that are discontiguous in sequence but topographically adjacent due to protein folding. Although, as noted previously, these distinctions remain somewhat controversial. It was originally thought that antibodies reactive with native protein would only be produced if the fragment or peptide used for immunization closely reproduced the tertiary conformation of the protein. However, numerous studies showed that linear synthetic peptides corresponding to parts of a protein are able to elicit antibodies that react with the intact antigen (Arnon et al., 1971, 1976; Atassi, 1975; Sutcliffe et al., 1980; Walter et al., 1980; Sutcliffe et al., 1983; Lerner, 1982; Shinnick et al., 1983). It has also been noted that monoclonal antibodies raised against a native protein are usually conformation specific and do not bind to peptide fragments, while the majority of monoclonal antibodies raised against peptide fragments also recognize the intact molecule (Todd et al., 1982; Niman et al., 1983; Van Regenmortel, 1984). One can, at least operationally, define three categories of antipeptide antibodies: (1) an antipeptide that reacts with the immunizing peptide, but not with the native protein (i.e., peptide specific); (2) an antipeptide that reacts with the immunizing peptide and with the native protein, but does not compete with antibodies produced by immunization with the native protein (i.e., peptide unique); and (3) an antipeptide that reacts with the immunizing peptide and with native protein, and inhibits reactivity of antibodies elicited by immunization with the native protein (i.e., sequential determinant specific). Reciprocally, the peptides used to elicit peptide-specific
SYNTHETIC T AND B CELL SITES
239
antibodies and peptide-unique antibodies would not be expected to bind antibody elicited by the native protein, whereas peptides representing a sequential determinant would be expected to bind antibodies raised to the native protein. In other words, antibodies to a sequential determinant can be elicited by immunization with either a peptide or the native protein, and these antibody specificities should be mutually competitive. We prefer to examine the cross-reactivity of a peptide antisera with the native protein antigen by antipeptide/antiprotein antibody competition experiments rather than attempting to inhibit antiproteidprotein interactions with soluble peptide, which is fraught with problems of antigen valency, solubility, affinity, etc. In terms of synthetic vaccine design, only antipeptide antibodies capable of neutralizing infection are primarily of interest. Both peptideunique and sequential determinant-specific antipeptide antibodies may be capable of neutralization, since both categories of antibodies bind native protein, whereas peptide-specificantibodies do not. The emphasis placed on using peptides which represent sequential determinants within the native protein is based on the desire to elicit antibody specificities which are elicited by the native protein during infection, since a proportion of these antibodies are sure to be neutralizing. In the absence of additional information, this is a reasonable approach; however, it does not exclude the possibility that peptide-unique antibodies are also capable of possessing neutralizing function. For example, antibodies raised against residues 284-311 of the S region of HBsAg bind the immunizing peptide, bind native HBsAg in a subtype-specific manner, and do not compete with subtype-specific antinative HBsAg monoclonal antibodies (D. Milich, unpublished observation), but nevertheless, this peptide protected chimpanzees from infection with HBV (Gerin et d.,1983). Therefore, in this case it appears that antipeptide antibodies which bound native HBsAg at “unique” sites other than those bound by antibodies to the native protein were capable of viral neutralization in wvo. The generality of this observation in other viral systems and with other peptide antisera needs to be determined. Table I1 represents a compilation of representative antibody-binding sites from numerous protein systems, which have been defined with synthetic peptides. The only qualification for definiton as a B cell site was that the peptide elicited antibodies that reacted with the immunizing peptide. The reader is referred to the cited references for information relating to the cross-reactivity of antipeptide antisera with native protein or its neutralizing ability (when relevant). In addition, there are numerous nonlinear antibody sites that have been characterized using monoclonal antibody and/or species variant proteins which have not
240
DAVID R. MILICH
TABLE I1 ENUMERATION OF B CELL SITES System/antigen
Globular prnteidantigem Cytochrome dbovine Cytochrome drabbit Cytochrome dhorse *zyme/hen egg white
Myoglobin/sperm whale
Myoglobin/ bovine Myohemerythrin
Viral antigens BaEV/p2PV Bovine Rotavim/VP6 Dengue type 4/ polypmein
Dengue type 4/E(gp) EBV/EBNA-1 (P62)
Site
Reference
42-50 13-30 N-acetyl-(1-4)-Gly( 87-100) 56-73 64-83 38-54 1-18 74-96 16-21 56-62 94-99 113-119 146-151 72-89 25-55 140-153 3-16 7-16 22-35 26-35 37-46 57-66 63-72 79-72 69-82 73-82 96-109 100-109 107-116
Atassi (1981) Corradin et al. (1983) Jemmemn et al. (1985) Jemmemn and F’atemn (1986) Arnon and Sela (1969) Takagaki et al. (1980) Sette et al. (1986) Genving and Thompson (1968) Atassi (1975) Atassi (1975) Atassi (1975) Atassi (1975) Atassi (1975) Leach (1983) Leach (1983) Cooper et al. (1986) Tainer et al. (1984) Tainer et al. (1984) Tainer et al. (1984) Tainer et al. (1984) Tainer et al. (1984) Tainer et al. (1984) Tkiner et al. (1984) Tainer et al. (1984) Tainer et al. (1984) Tainer et al. (1984) Tainer et al. (1984) Tainer et al. (1984) Gepn et al. (1987)
61-73 103-115 133-147
Copeland et al. (1985) Bornas-Cuesta et al. (1987) Born-Cuesta et al. (1987)
538-551 577-590 477-490 498-613 30-38 67-69
Markoff Markoff Markoff Markoff Markoff Markoff
150-168 315-327 369-383
D h e r et al. (1984) Ddner et al. (1984) Ddner et al. (1984)
et al. (1988) et al. (1988) et al. (1988) et al. (1988) et al. (1988) et al. (1988)
241
SYNTHETIC T AND B CELL SITES
BIBLE I1
(Continued) ~~
System/antigen
EBV/gpllO FELV-B/=70
Flu/hemagglutinin (A/Texas/l/77) (A/X31) (H3X:47) (A/X31) (A/Eng/42/72) (A/Eng/42/72) (A/X47) (A/X47) (A/X47) (A/X47) (A/X47) (A/X47) (A/X47) (A/X47) (A/X47) (A/X47) (A/X47) (A/X47) (AIX47) (A/ X47) (A/X47) (AIX47) (AIX47) (A/X47) (A/X47) (AIX47) FMDV/Vpl
Site
Reference
1-14 57-64 83-102 116-131 215-227 228-248 807-816 45-64 203-220 221-237 259-276 348-368 44-65 107-120 121-143
Rhoda et al. (1985) Rhoda et al. (1985) Rhoda et al. (1985) Rhoda et al. (1985) Rhoda et al. (1985) Rhoda et al. (1985) Roudier et al. (1988) Elder et al. (1987) Elder et al. (1987) Elder et al. (1987) Elder et al. (1987) Elder et al. (1987) Elder et al. (1987) Elder et al. (1987) Elder et al. (1987)
91-108 123-151 75-110 305-328 181-200 138-164 1-15 1-36 15-53 39-65 53-60 53-65 53-87 73-83 76-90 76-111 105-140 130-151 132-146 140-147 140-156 140-175 174-196 184-196 201-227 306-329 9-24 17-32 25-41
Muller et al. (1982) Jackson et al. (1982) Wilson et al. (1984) Natorowicz et al. (1985) Shapira et al. (1984) Shapira et al. (1984) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Green et al. (1982) Bittle et al. (1982) Bittle et al. (1982) Bittle et al. (1982) (continued)
DAVID R. MILICH TABLE I1 (Continued) System/antigen
HAV/Vpl HBV/HBsAg
HBV/HBsAg pR-S(l)
Site 1-41 141-160 151-160 200-213 144-159 167-181 205-213 146-152 143-149 144150 13-24 176-190 186-190 196-209 212-226 221-226 222-255 269-283 278-283 281-311 314-322 323-337 312-323 296-311 291-311 302-308 308-320 313-321 313-332 314-332 319-332 324-332 309-329 284-311 299-313 175-194 195-221 222-255 330-359 273-295 218-239 243-253 12-32 32-53 41-53 94-105
Reference Bittle et al. (1982) Bittle et al. (1982) Bittle et al. (1982) Bittle et al. (1982) Pfaff et al. (1982) Pfaff et al. (1982) Pfaff et al. (1982) Geysen et al. (1985) Geysen et al. (1985) Geysen et al. (1985) Emini et al. (1985) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Lemer et al. (1981) Hopp and Woods (1981) Dreesman et al. (1982) Dreesman et al. (1982) Bhatnagar et al. (1982) Bhatnagar et al. (1982) Bhatnagar et al. (1982) Bhatnagar et al. (1982) Bhatnagar et al. (1982) Bhatnagar et al. (1982) Bhatnagar et al. (1982) Neurath et al. (1982) G r i n et al. (1983) G r i n el al. (1983) Shih et al. (1983) Shih et al. (1983) Shih et al. (1983) Shih et al. (1983) Audibert et al. (1982) Neurath et al. (1984a) Neurath et al. (1984a) Neurath et al. (1985) Milich et al. (1987a) Milich et al. (1987a) Milich et al. (1987a)
243
SYNTHETIC T AND B CELL SITES TABLE I1 (Continued) System/antigen
HBV/HBsAg p ~ - S ( 2 ) (ad)
HBV/pol HBV/X HIV/@20-gp41 HIV2/gp41 HIVgp41(HIV-l) HIV/gp41 HIV/gp41(HIV-1/2) HIV/gpl20 HIV/gpl20 HIV/pl7 HSV/gDl HSV/gDI HSV/gDl HSV/gDl HSV/gD, and gD, HSV/gD, and gD, HSV/gCl HSV/IE12 1 HSV/IE751 Mo-MULV MS-2 coliphage/P Polio/l/Vpl
Site 106-117 16-27 21-47 1-21 120-145 133-151 133-139 135-143 137-145 126-140 822-838 781-795 29-38 100-115 144-154 115-131 503-532 735-752 589-610 598-609 586-620 254-274 303-321 86-115 8-23 340-356 288-297 314-323 268-287 1-23 128-139 11 residues, C-terminus 65-76 527-562 89-108 93-104 70-75 97-103 70-80 11-17 141-147 24-40 61-80 86-103
Reference Milich et al. (1987a) Milich et al. (1987a) Neurath et al. (1986) Neurath et al. (1986) Neurath et al. (1984b) Okamoto et al. (1985) Milich et al. (1986~) Milich et al. (1986~) Milich et al. (1986~) Steward et al. (1988) Feitelson et al. (1988) Feitelson et al. (1988) Feitelson et al. (1988) Moriarty et al. (1985) Moriarty et al. (1985) Feitelson et al. (1986) Chanh et al. (1986) Kennedy et al. (1986) Rosen et al. (1987) Gnann et al. (1987) Norrby et al. (1987) Ho et al. (1988) Paker et al. (1988) Naylor et al. (1987) Cohen et al. (1984) E m b e r g et al. (1985b) Strynadka et al. (1988) Strynadka et al. (1988) Eisenberg et al. (1985b) Eisenberg et al. (1985a) Zweig et al. (1984) Palfreyman et al. (1984) Palfreyman et al. (1984) Sutcliffe et al. (1980) Langbeheim et al. (1976) Wychowski et al. (1983) Emmi et al. (1983) Emini et al. (1983) Emini et al. (1983) Emini et al. (1983) Bonin et al. (1985) Chow et al. (1985) Chow et al. (1985) Chow et al. (1985) (continued)
244
DAVID R. MILICH TABLE I1 (Continued) System/antigen
RabiedG
RSV/Fl&LS Rhinovirus 14/VP1 Rhinovirus 14/VP3 SV40/T
TMVP
Site
Reference
121-141 161-181 202-221 244-264 270-287 286-302 222-241 100-109 91-109 182-201 222-241 244-264 192-203 162-173 71-81 75-81 93-103 97-103 93-103 97-103 89-100 113-121 165-172 1-44 57-102 103-179 188-236 244-291 292-323 324-376 386-452 216-236 147-162 126-141 1-8 Last 11 residues, C-terminus 1-10 34-39 62-68 93-112 142-158 55-61 80-90
Chow et al. (1985) Chow et al. (1985) Chow et al. (1985) Chow et al. (1985) Chow et al. (1985) Chow et al. (1985) Chow et al. (1985) Chow et al. (1985) Chow et al. (1985) Diamond et al. (1987) Diamond et al. (1987) Diamond et al. (1987) Emini et al. (1984) Bonin et al. (1985) Bonin et al. (1985) Bonin et al. (1985) Bonin et al. (1985) Bonin et al. (1985) Bonin et al. (1985) Bonin et al. (1985) Ferguson et al. (1985) Jameson et al. (1985) Jamemn et al. (1985) Dietzschold et al. (1982) Dietzschold et al. (1982) Dietzschold et al. (1982) Dietachold et al. (1982) Dietzschold et al. (1982) Dietzschold et al. (1982) Dietzschold et al. (1982) Dietachold et al. (1982) Tmdel et al. (1987) McCray and Werner (1987) McCray and Werner (1987) Walter et al. (1980) Walter et al. (1980) Milton and Regenmortel (1979) Milton and Regenmortel (1979) Milton and Regenmortel (1979) Milton and Regenmortel (1979) Milton and Regenmortel (1979) Altschuh et al. (1983) Altschuh et al. (1983)
245
SYNTHETIC T AND B CELL SITES
TABLE I1 (Continued) System/antigen Pamitic antigens l? falcipammlCS l? falaparum/meromite/ Pfl55 I? falci$anm/merozoite/ Pfl55 l? falciparum/merozoite and late schizont/35K I! falciparundmerozoite and late schizont/55K I? falciprumlmerozoite and late schkontA95K
l? bergheilCS Baaerial antigens Cholera/ toxin/ B
Diphtheriahoxin E. colilLT
Site
Reference
13-28
Dore et al. (1987)
(NAN%
Zavala et al. (1985)
EENVEHDA
Benins et al. (1986)
EENVEENV
Perlmann et al. (1987)
N-terminus
Patamyo et al. (1987)
N-terminus
Patarroyo et al. (1987)
43-53 27-38 40-52 540-552 575-587 731-742 614-624 78-92 106-117 595-606 694-707 277-287 (DPPPPNPN)PD
Patarroyo et al. (1987) Patarroyo et al. (1987) Patamyo et al. (1987) Patarroyo et al. (1987) Patarroyo et al. (1987) Patamyo et al. (1987) Patarroyo et al. (1987) Patarroyo et al. (1987) Patarroyo et al. (1987) Patarroyo et al. (1987) Patarroyo et al. (1987) Patarroyo et al. (1987) Zavala et al. (1987)
8-20 30-42 50-64 69-85 75-85 83-97 30-50 50-75 57-69 47-60 45-64 27-36 21-36 188-201 50-64 8-20
Jacob et al. (1983) Jacob et al. (1983) Jacob et al. (1983) Jacob et al. (1983) Jacob et al. (1983) Jacob et al. (1983) Delmas et al. (1985a) Delmas et al. (1985a) Ghose and Karush (1988) Ghose and Karush (1988) Ghose and Karush (1988) Ghose and Kamh (1988) Ghose and Karush (1988) Audibert et al. (1981) Jacob et al. (1984) Jacob et al. (1984) (continued)
246
DAVID R. MILICH
TABLE I1 (Continued) Systemlantigen E. colzlpilinlGa1-Gal
M. tuberculosisll8kDa M. tuberculosisl30kDa M. tuberculosisl23kDa M. tuberculosirll3kDa M. bovislBCG N. gonorrhoeaelpilinl MSll
Streptococcw/M24
StreptococcuslMS StreptococcurlM6 Hormones Erythropoietinlhuman
hCGIP
Site
Reference
5-12 65-75 93-104 103-116 131-143 11-16 17-27 1-16 1-9 1-13
Schmidt et al. (1984) Schmidt et al. (1984) Schmidt et al. (1984) Schmidt et al. (1984) Schmidt et al. (1984) Patarroyo et al. (1986) Patamyo et al. (1986) Patarroyo et al. (1986) Patamyo et al. (1986) Minden et al. (1986)
21-35 41-50 48-60 69-84 107-121 121-134 135-151 1-35 13-35 18-35 23-35 1-29 1-20 20-40 164-179 1-20
Rothbard et al. (1984) Rothbard et al. (1984) Rothbard et al. (1984) Rothbard et al. (1984) Rothbard et al. (1984) Rothbard et al. (1984) Rothbard et al. (1984) Beachey et al. (1981) Seyer et al. (1986) Seyer et al. (1986) Seyer et al. (1986) Beachey and Dale (1988) Dale et al. (1983) Dale et al. (1983) Beachey and Dale (1988) Beachey and Dale (1988)
8-15 40-59 80-99 99-118 111-129 131-150 109-145 125-137 116-145 109-118 118-136 136-145 40-52 38-57 82-101 92-101 110-116
Sytkowski and Fisher (1985) Sytkowski and Donahue (1987) Sytkowski and Donahue (1987) Sythwski and Donahue (1987) Sytkowski and Donahue (1987) Sytkowski and Donahue (1987) Stevens (1976) Stevens (1976) Talwar et al. (1976) Bidart et al. (1985) Bidart et al. (1985) Bidart et al. (1985) Stevens et al. (1986) Stevens et al. (1986) Iyer et al. (1986) Iyer et al. (1986) Bidart et al. (1987)
247
SYNTHETIC T AND B CELL SITES
TABLE I1 (Continued) System/antigen
hCS and hGH hPTH LH-RH Somatostatin/28 Thyroglobulin Autoantibodiea AChR/Torpedo a
AChR/Torpedo 6 AChR/human a MBP/bovine
MBP/guinea pig
MBP/rat MBP/human
Site
Reference
134-139 139-145 121-145 165-174 44-68 1-10 1-12 1-19
Bidart et al. (1987) Bidart et al. (1987) Bidart et al. (1987) Neri et al. (1984) Delmas et al. (1985b) Carelli et al. (1985) Lipkin et al. (1988) De Boer et al. (1987)
426-437 432-437 159-169 125-147 1-20 126-143 169-181 185-196 193-210 262-277 330-340 351-368 394-409 354-367 367-374 373-387 185-196 351-368 160-167 68-83 65-83 43-88 74-83 69-83 69-84 79-88 89-115 68-88 69-84 72-84 75-84 43-88 43-88 51-67 67-80 74-84 79-88
Barkas et al. (1984) Barkas et al. (1984) McCormick et al. (1985) McCormick and Atassi (1984) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Fuchs et al. (1987) Dyrberg and Oldstone (1986) Day et al. (1981) Day et al. (1981) Fritz and Chou (1983) Fritz and Chou (1983) Day and Hashim (1988) Day and Hashim (1988) Price et al. (1986) Sheng et al. (1988) Fritz and Chou (1983) Hashim et al. (1987) Hashim et al. (1987) Hashim et al. (1987) Fritz and Chou (1983) Whitaker (1982) Whitaker (1982) Whitaker (1982) Whitaker (1982) Whitaker (1982) (continued)
248
DAVID R. MILICH TABLE I1 (Continued) Systemlantigen
MBPlrabbit
Ribosomal protein P2 Lymphokinen IL-l/hul@
IL-3
IFNlylhu IFN/y
Oncogenea v-fes/FeSV v-fes/ST-FeSV
v-fes/85-kDa fusion protein c.Ha-(EJ)ras/NRK-H c-Ha-ras v-Ha-ras
Site
Reference
83-95 80-89 82-91 88-109 1-21 1-44 1-87 15-44 22-44 45-87 96-128 101-111
Whitaker (1982) Price et al. (1986) Price et al. (1986) Sheng et al. (1988) Sheng et al. (1988) Sheng et al. (1988) Sheng et al. (1988) Sheng et al. (1988) Sheng et al. (1988) Sheng et al. (1988) Sheng et al. (1988) Ellcon et al. (1988)
1-18 45-65 71-90 1-140 1-6 7-140 1-29 30-43 44-75 76-89 64-82 91-112 123-140 130-135 131-146 1-39 95-133 1-20 3-20 5-20 10-30
Bomford et a!. (1987) Bomford et al. (1987) Bomford et al. (1987) Clark-Lewis et al. (1986) Ziltener et al. (1987) Ziltener et al. (1987) Ziltener et al. (1988) Ziltener et a!. (1988) Ziltener et al. (1988) Ziltener et al. (1988) Ziltener et al. (1988) Ziltener et al. (1988) Ziltener et al. (1988) Ziltener et al. (1988) Ichimori et al. (1985) Russell et a!. (1986) Russell et al. (1986) Magazine et al. (1988) Magazine et al. (1988) Magazine et al. (1988) Magazine et al. (1988)
12 raidues, C-terminus 541-555 584-593 782-796 893-905 901-913 754-769 1-18 1-18 1-18
Sen et al. (1983) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Niman et al. 0985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985)
249
SYNTHETIC T AND B CELL SITES
TABLE I1 (Continued) Sytem/antigen
H-ras hu EGF receptor/AEV v-Ki-ras/NRK-K
v-mos/M-MuSV v-myc/MCZI)
v-myb/BM2
pp60"rc/RSV
sk/PDGE-2 v-Si/SSV p28"/SSV C-SK/~T~'~
Other CD8 ( T cell)
Cob box gene/ S. cerevisiae/ intron bi2
Site
Reference
91-108 126-136 146-155 160-179 37-59 984-996 1-18 119-135 161-175 167-185 4981-4999 4963-4999 323-334 340-350 363-371 389-403 395-405 11-24 33-47 146-162 168-186 170-185 247-260 247-265 415-424 103-108 155-160 315-321 326-333 409-415 419-427 458-463 500-506 1-18 1-35 139-155 409-425 468-482 499-508
Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Niman et al. (1985) Kris et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Papkoff et al. (1981) Papkoff et al. (1981) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985) Wong and Goldberg (1981) Tamura et al. (1983) Tmura et al. (1983) Tamura et al. (1983) Tamura et al. (1983) Tamura et al. (1983) Tamura et al. (1983) Tmura et al. (1983) Tamura et al. (1983) Nman et al. (1985) Robbins et al. (1982) Niman (1984) Tanaka et al. (1985) Tanaka et al. (1985) Tanaka et al. (1985)
1-13
Brunati et al. (1987)
709-738
Gulso et al. (1984) Gulso et al. (1984)
844-876
(continued)
250
DAVID R. MILICH
TABLE I1 (Continued) System/antigen
LDH/C4
Site 5-15 97-110 211-220 99-111
Wheat et al. (1985) Wheat et al. (1985) Wheat et al. (1985) Chen et al. (1984)
1-16 167-180 207-218 237-251 291-305 338-352 372-385 55-66 50-60
Schluter et al. (1987) Robertson and Liu (1988) Robertson and Liu (1988) Robertson and Liu (1988) Robertson and Liu (1988) Robertson and Liu (1988) Robertson and Liu (1988) Atlas et al. (1985) Galen et al. (1987); Bouhnik et al. (1987) Galen et al. (1987); Bouhnik et al. (1987) Galen et al. (1987); Bouhnik et al. (1987) Galen et al. (1987); Bouhnik et al. (1987) Galen et al. (1987); Bouhnik et al. (1987) Galen et al. (1987); Bouhnik et al. (1987) Galen el al. (1987); Bouhnik et al. (1987) Galen et al. (1987); Bouhnik et al. (1987) Galen el al. (1987); Bouhnik et al. (1987) Van Rietschoten et al. (1975) Bahraoui et al. (1986) Bahraoui et al. (1987)
63-71 61-90 118-126 133-144 180-188 215-224 300-310 304-312 Venoms/ bee/apamin Venoms/scorpion/AaHII Venoms/scorpion/AaHII
Reference
1-18 50-59 19-28
been included in the table. For example, three nonlinear epitopes have been defined for myoglobin involving 4-79, 83-144-145, and residues centered around 140 (Berzofsky et al., 1982). V. Methods of Predicting T and
B Cell Recognition Sites
A. B CELLRECOGNITION SITES For purposes of vaccine development, it would be useful to predict the regions of a protein that will be recognized by the immune system.
SYNTHETIC T AND B CELL SITES
251
This is complicated by the fact that in addition to intrinsic factors related to structural features of the antigen, many extrinsic factors related to the regulation and repertoire of the immune system determine what sites on a protein are actually recognized by T and B cells (Berzofsky, 1985). Therefore, biological realities may severely limit the ability to predict T and B cell recognition sites based on chemical and structural features alone. With regard to intrinsic factors of B cell antigenicity, there have been a number of attempts to predict the antigenic sites of proteins from either their primary sequence structure or from properties of the protein that are related to polypeptide folding. All models assume that antigenicity is a surface property, and accessibility on the surface of the native structure is a minimum requirement for an antigenic B cell site. The question is whether other additional structural properties can be used to predict which regions of the accessible surface are potential antigenic sites. Structural properties that have been considered are regions of local hydrophilicity in the sequence (Hopp and Woods, 1981; Kyte and Doolittle, 1982), the occurrence of 0-turns (Rose et al., 1985; Nestor et a l . , 1985), protruding regions of proteins (Thornton et al., 1986; Fanning et al., 1986; Novotny et al., 1986), segmental mobility (Westhof et al., 1984; Tainer et al., 1984, 1985), and the occurrence of specific amino acids in the sequence (Welling et al., 1985). These intrinsic properties are generally highly correlated with one another. For instance, clusters of hydrophilic residues are more likely to occur on the aqueous surface of a water-soluble protein. Similarly, surface residues will have fewer packing constraints and contacts and, therefore, tend on the average to be more mobile than internal residues. Proponents of each of these intrinsic features argue that they are more than mere indicators of surface accessiblity. These methods have been partially successful in locating possible surface regions of proteins that can be used to derive antipeptide antibodies that show cross-reactivity with the native protein. SITES B. T CELLRECOGNITION The recent interest in development of synthetic peptide vaccines and the increased understanding of the mechanisms of T cell antigen recognition have prompted much interest in the prediction of peptide sequences in protein antigens that can elicit a T cell response. Berzofsky and colleagues (De Lisi and Berzofsky, 1985; Margalit et al., 1987; Berzofsky et a l . , 1987) have reasoned that since all regions of a protein are not capable of eliciting a T cell response in a given individual, there may be a general property of T cell-active peptide sequences that enhance their ability to associate with a binding site on Ia molecules, which is a prerequisite for T h cell activation. They have proposed that T cell
252
DAVID R. MILICH
determinants correlate with regions of a protein which have a propensity to adopt amphipathic a-helical structures after processing events have occurred in the APC. The hypothesis states that one face of the a-helix, such as the hydrophobic face, may interact with the Ia molecule and the other face, such as the hydrophilic face, may bind the T cell receptor. Another proposal has suggested that a common motif is found in the majority of helper and cytotoxic T cell determinants represented by a four- or five-residue peptide sequence in which a charged residue or glycine is followed by two or three hydrophobic residues and ends with a polar residue (Rothbard, 1986; Rothbard and Taylor, 1988). Although differing in emphasis, the two hypotheses are not mutually exclusive, and with the exception of glycine in the first position, such a sequence could occur as the first turn of an amphipathic a-helix, but may also occur in other conformations in proteins (Paterson, 1988). Both of these proposals have been substantiated by statistical analysis of known T cell sites which show a strong correlation with the property being studied (De Lisi and Berzofsky, 1985; Margalit et al., 1987; Rothbard and Taylor, 1988). Furthermore, previously undefined T cell determinants have been predicted by these algorithms (Cease et al., 1987; Lamb et al., 1987; Rothbard and Taylor, 1988; Rothbard et al., 1988; Good et al., 1987; Kabilan et al., 1988; Hohlfeld et al., 1988). Rothbard and Taylor (1988) reported that 81% of the T cell determinants used in their data set possessed the four- or five-sequence motif. However, a limitation of this algorithm as a predictor of T cell sites is the high frequency of this motif in proteins in general (Gotch et al., 1987; Paterson, 1988). A theoretical limitation of both algorithms is that neither take into account the influence of the MHC on the T cell site recognized by a given murine strain or individual, although allele-specific subpatterns of the motif hypothesis have been suggested (Guillet et al., 1986; Buus et al., 1987; Rothbard and Taylor, 1988; Rothbard et al., 1988). As the data base increases, these and possibly other predictive algorithms will surely evolve, and the potential benefits of an ability to predict accurately T cell recognition sites within a protein will be enormous in terms of synthetic peptide vaccine design. VI. Status of Candidate Synthetic Peptide Vaccines
The majority of this review has focused on the use of synthetic peptides to study basic immunological mechanisms of T and B cell recognition. It is also important to review the application of synthetic peptide technology to vaccine design. Ideally, synthetic peptide vaccines should (1) represent contiguous and/or noncontiguous epitopes that are
SYNTHETIC T AND B CELL SITES
253
important for protective antibody production: (2) contain T cell recognition sites that will serve to induce antibody production and cellular immunity and to prime a memory T cell response relevant to the pathogen: (3) be specific for a particular pathogen and not induce immunity cross-reactivewith self-antigens: (4) be readily administered; and (5) provide long-lasting immunity not requiring frequent booster doses. Although no synthetic peptide vaccines have yet been licensed for human or veterinary use, there are a number of systems which have progressed to the level of in uivo protection studies. The status of several candidate synthetic peptide vaccines is summarized in Table 111. Foot and mouth disease (FMD) affects both domestic and wild ruminants and swine. It is a worldwide problem, as the disease is present in every continent except Australia and North America. The extent of the problem is reflected in the over 1000 million doses of vaccine which are used each year, and, nonetheless, the disease is still widespread. Current vaccines are produced by inactivation of virus cultured in baby hamster kidney cells or in bovine tongue epithelium fragments. Although this vaccine is generally effective, a number of disadvantages to its use have prompted continued studies to develop an alternative vaccine (Bittle et al., 1982). Of the four structural polypeptides of FMD virus, the 213-amino acid VPl polypeptide bears the determinants for cellular attachment and induction of neutralizing antibody. Therefore, VP1 has been the focus of attempts to localize B cell recognition sites using synthetic peptides. Two groups used similar peptide sequences (Table 111) coupled to heterologous protein carriers to elicit neutralizing antibodies (Bittle et al., 1982; Pfaff et al., 1982). Bittle et al. (1982) demonstrated that a single inoculation of peptide conjugate protected guinea pigs against subsequent challenge with the virulent virus. Subsequently, a chemically synthesized peptide consisting of two regions of VPl, 141-158 and 200-213, was used as an immunogen free of any carrier (DiMarchi et al., 1986). This composite peptide elicited neutralizing antibody and protected cattle against intradermolingual challenge with infectious virus. The authors claimed that inclusion of the C-terminal sequence enhanced the protective response. During influenza virus infection, hemagglutinin is the major antigenic protein against which neutralizing antibodies are directed, and it is also responsible for attachment of the virus to the cell. Vaccine programs against influenza make use of killed or attenuated virus, as well as its subunit, yet none of the vaccines is capable of providing full and satisfactory protection against viral infection. Moreover, there are side effects ranging from fever and mild disease symptoms to neurological disorders
STATUS OF
System/antigen FMDV/VPl FMDV/VPl FMDV/VPl Flu/HA Flu/HA HBV/S HBV/pre-S(2) HBV/pre-S(2) HSV/gpD Chlora toxin/B subunit Diptheria toxin S. pyogenes/M protein
E. colz/Gal-Gal Pilin Malaria/CS Malaria/CS Malaria/CS Malaria/late schizont/merozoite Malaria/late schizont/merozoite HCG
TABLE 111 CANDIDATE SYNTHETIC PEPTIDE VACCINFS
Immunogen 141-160 - KLH 144-159-KLH, BSA 141-158-S-S-200-213 91-108-TT 138-166-TT 284-311 /y- KLH 133-151/y-BSA 120-145/d-KLH 1-23-liposome 50-64-IT 188-201-BSA CB7(1-34)-PLL M5(1-20)-S-CB7(23-35) M5(1-10)-M6(1-11) -M24(1-l2)-KLH 5-12-thyraglobin 6 5- 75- thyroglobin (DPPPPNPN)z-TT (NANP)S-TT (NANP)S-TT 83K(43-53)-55K(1-13) - 35K(l-l1)-BSA 83K(43-53)-55K(1-13) -35K(1-11) 109-145-DT 109-145-DT
Comments
Reference
Guinea pigs protected Guinea pigs protected Cattle protected Mice partially protected Mice partially protected Chimps partially protected Chimps protected Chimps protected No antibody/mice protected Neutralized toxin Neutralized toxin Mice protected Mice protected
Bittle et al. (1982) Pfaff et al. (1982) DiMarchi et al. (1986) Miiller et al. (1982) Sharpira et al. (1984) Gerin et al. (1983) Itoh et al. (1986) Thornton et al. (1987) Watari et al. (1987) Jacob et al. (1983) Audibert et al. (1981) Beachey et al. (1981) Beachey et al. (1986)
Mice protected Mice protected Mice protected Mice protected Humans partially protected Human seroconversion/low T cell Monkeys protected
Beachey et al. (1987) Schmidt et al. (1988) Schmidt et al. (1988) Zavala et al. (1987) Hemngton et al. (1987) Etlinger et al. (1988) Patarroyo et al. (1987)
Humans partially protected
Patarroyo et al. (1988)
Baboon fertility reduced Human antibody
Stevens (1981) Jones et al. (1988)
SYNTHETIC T AND B CELL SITES
255
such as the Guillain Barre syndrome. The inability to control influenza by vaccination is due mainly to the capacity of the two surface glycoproteins of the virus, hemagglutinin and neuraminidase, to change antigenically; thus, new vaccines need to be prepared every few years. Analysis of the three-dimensional structure of influenza virus hemagglutinin has proceeded rapidly since X-ray crystallography defined precisely the tertiary structure of the two polypeptide contituents (Wilson et al., 1981). Four antigenic sites on the three-dimensional structure of the molecule were proposed (Wiley et al., 1981; Laver et al., 1980). During natural infection, antibodies are generated against all four sites. ’ However, immunization with synthetic peptides representing over 75% of the linear amino acid sequence resulted in antibodies which reacted with HA1 or the intact virus, which demonstrated that the sites in proteins accessible to peptide antibodies are more numerous than are the few sites recognized in the course of a natural immune response (Green et al., 1982). These results also indicated that peptide antigens can induce antibodies not normally elicited, since the antibodies raised during the natural infection did not react with the peptide antigens. Nonetheless, these peptide-unique antibodies were capable of protection, as several groups have reported that immunization with synthetic peptides (residues 91-108 and 138-164) linked to tetanus toxoid partially protected mice from a challenge with infectious virus (Miiller et al., 1982; Sutcliffe et al., 1983; Shapira et al., 1984). Furthermore, many of the antipeptide antisera recognized amino acids common to all hemagglutinins, regardless of the serotype specificity of the native molecule. It was postulated that by using peptide antigens it may be possible to construct a vaccine with a broad range of specificity not attainable with the intact protein (Sutcliffe et al., 1983). Similar to the influenza system, studies in the hepatitis B virus system demonstrated that synthetic peptides can elicit antibodies that are reactive with the native S region of the hepatitis B surface antigen (Lerner et al., 1981; Bhatnagar et al., 1982; Dreesman et al., 1982). This was true, notwithstanding the fact that antigenicity of the native S region is highly dependent on the presence of intact disulfide bonds (Vyas et al., 1972; Dreesman et al., 1973). Using a synthetic peptide from a hydrophilic section of the S region of HBsAg (residues 284-311) linked to KLH, Gerin et al. (1983) demonstrated that this region represented a protective epitope when evaluated in a chimpanzee model of HBV infection. The pre-S(2) region of HBsAg has been shown to possess sequential, as opposed to conformationally dependent, antibody-binding sites (Neurath et al., 1984b; Milich et al., 1985a). Two groups defined peptide antigens representing sequences within the pre-S(Z) region, p120-145
256
DAVID R. MILICH
(Neurath et al., 1984b) and p133-151 (Okamoto et al., 1985). Antipeptide antibodies were shown to compete with antibodies raised to the protein, and, reciprocally, antinative protein antibodies competed with antipeptide antibodies for binding to HBsAg-pre-S(2)-containing particles (Milich et d., 1985a). Further studies indicated the presence of two dominant overlapping antibody-binding sites at residues 133-139 and 137-143 (Milich et al., 1986~).Two subsequent studies examined the protective efficacy of pre-S(2) region synthetic peptides p120-140 and p133-151 in chimpanzee infection models. Itoh et al. (1986) showed that the pre-S(2) sequence 133-151 linked to BSA induced protection from HBV challenge; the same result was achieved by Thornton et al. (1987) using the pre-S(2) sequence 120-145 coupled to KLH. These results demonstrated that an epitope(s) within the pre-S(2) region was capable of eliciting protection in the absence of an immune response to S region antigens. The current HBV vaccines consist of the S region alone. The inclusion of pre-S(2) antigens would increase the number of neutralizing epitopes, and could be accomplished by synthetic as well as DNA recombinant technology, since the pre-S(2) region appears to be composed of sequential determinants at the level of B cell and T cell immunity. Protection against infection with the herpes simplex virus by immunization with a synthetic peptide has also been reported. In contrast to the other viral systems discussed, the mechanism of protection appeared to be at the T cell level, since no antibody was produced. Watari et al. (1987) constructed an immunogen consisting of a &%aminoacid synthetic peptide corresponding to the N-terminus of gD from either HSV-1 or HSV-2 coupled to palmitic acid side chains, and then inserted the acylated peptides into a liposome structure. This particular construct was used to enhance the T cell response. These investigators found that immunization with this form of the antigen produced long-term protective immunity with a single dose in the absence of neutralizing antibody production. Furthermore, T cells, but not serum, from such immune mice could adoptively transfer this protection (Watari et al., 1987). The induction of T cell immunity is believed to play an important role in neutralizing HSV infection because reactivation of a latent infection generally does not result in a viremia, but rather viral infection of other cells. In fact, high anti-HSV virus-neutralizing antibody (VNA) does not necessarily protect from reinfection and reactivation in man (Pass et al., 1979), and may even play a negative role (L. A. Wilson et al., 1984). The synthetic approach to the induction of immune responses to bacterial toxins has been studied using peptides derived from the cholera and diphtheria toxins. The toxin of Vibrio cholerae is composed of two
SYNTHETIC T AND B CELL SITES
257
subunits, A and B. Subunit B is responsible for binding to cell receptors, and antibodies to this subunit are capable of neutralizing the biological activity of the intact toxin. Synthetic peptides representing various regions of the B subunit were shown to induce antibody, which cross-reacted with the native molecule (Jacob et al., 1983, 1984). One peptide (residue 50-64), which reacted with antisera to the native cholera toxin, induced antipeptide antibodies that significantly neutralized the biological activity of cholera toxin (Jacob et al., 1983). Similarly, guinea pigs immunized with residues 188-201 of the diphtheria toxin linked covalently to BSA produced antibodies which bound specifically to the toxin and neutralized its dermonecrotic and lethal effects in vivo (Audibert et al., 1981). The surface M proteins of group A streptococci prevent phagocytosis by the nonimmune host. In the immune host anti-M antibodies opsonize the organism to allow phagocytosis and killing. Because group A streptococci contain several different antigens that are cross-reactivewith host tissues, the concerns of vaccine development have been that an M protein vaccine could cause rather than prevent rheumatic fever. Therefore, Beachey and colleagues have synthesized selected regions of various M proteins in an attempt to construct protective vaccines free of autoimmune epitopes. They have shown that the N-termini of types 5 and 6 M proteins, as well as a 34-residue peptide of type 24 M protein from the C-terminus of an internal fragment (CB7), contain protective, but not tissue-cross-reactive,epitopes (Dale et al., 1983; Beachey and Seyer, 1986; Beachey et al., 1981). This approach revealed the problem of a high degree of type specificity of the highly variable N-terminal sequences of the M proteins. To overcome this problem, these investigators began synthesizing short peptide sequences of different serotypes of M protein in tandem, hoping to produce broadly protective vaccines against many serotypes. Recently, bivalent peptide and trivalent peptide immunogens were reported to induce high titers of antibodies against each of the included M protein serotypes, which also opsonized the respective serotypes of streptococci and lacked tissue cross-reactivity(Beachey et al. 1987). Uropathogenic Eschen'chia coli adhere to and colonize urogenital mucosa by means of adhesion proteins that bind epithelial cell surface receptor glycoconjugates (Leffler and Svanborg-Eden, 1980). Digalactoside (Gal-Gal)-binding adhesins are associated with Gal-Gal pili. A recombinant Gal-Gal pilus vaccine was reported to prevent renal colonization and invasion by the homologous pyelonephritis strain (O'Hanley et al., 1985). Subsequently, linear antigenic epitopes were identified with synthetic peptides corresponding to regions of the Gal-Gal pilin sequence.
258
DAVID R. MILICH
(Schmidt et al., 1984). A number of these peptides have been studied in the murine pyelonephritis model to determine their efficacy as vaccines and to map protective, linear epitopes. Synthetic peptides corresponding to residues 5-12 and 65-75 were found to induce antibodies that bound the homologous pilin and prevented disease in most vaccine recipients. Furthermore, anti-65-75, which represents an unconserved sequence, bound only the homologous pilin, whereas, anti-5-12, which corresponds to a conserved region, bound Gal-Gal pilins from seven of eight pyelonephritis strains (Schmidt et al., 1988). These results indicated that an eight-residue peptide corresponding to a highly conserved linear domain, which is not normally immunogenic within the folded protein, can confer protection on a mucosal surface equal to that obtained with the intact protein (Schmidt et al., 1988). Malaria is an important worldwide health problem infecting 800 million people annually around the world. Plasmodium falczparum is transmitted through Anopheles mosquitoes. The mosquito bite transmits the sporozoites into the circulation of the host. Sporozoites give rise to merozoites as they multiply in the host’s hepatocytes. As they rupture from the hepatocytes they either mature into the sexual gametocyte stage or invade the red blood cells, multiplying into the asexual stage. The three targets for an immune-mediated vaccine correspond to the three extracellular stages, i .e., sporozoites, merozoites, and the gametocytes. The sporozoite antigen has been the main focus of vaccine development because, if successful, it would interrupt all parasite development and prevent disease (Nussenzweig and Nussenzweig, 1986b). Zavala et al. (1985) synthesized a peptide known to correspond to the dominant epitope on the circumsporozoite protein of I? falci@rmm. This epitope on the CS molecule is formed by tandem repeated sequences of amino acids, NANP. (NANP)S was found to be the best peptide to induce antibody reactivity with itself and the suface of the parasite, and had the capacity to neutralize the infectivity of the parasite. (NANP)s conjugated to tetanus toxoid plus adjuvant was used by two independent groups in human clinical trials. The volunteers seroconverted to a-NANP in 53 % of those receiving a 100-pgdose and in 71% of those receiving a 160-pgdose. Upon subsequent challenge with parasite there was delay or suppression of disease (Herrington et al., 1987). In immunogenicity trials of (NANP),-TT in human subjects, Etlinger et al. (1988) reported seroconversion in 8 of 11 volunteers, and a relatively weak T cell proliferative response. Patarroyo et al. (1987) synthesized 18 peptides representing sequences on the 83K, 55K, and 35K proteins specific for the late schizont and merozoite stages of I? falciparum. None of the peptides was completely
SYNTHETIC T AND B CELL SITES
259
protective after subsequent challenge. However, by combining three peptides that exhibited partial protection individually [83K(43-53)55K(1-13)- 35K(1-11)], a completely protective response was elicited in Aotus monkeys. Encouraged by these results, the same construct was tested in human volunteers. All vaccine recipients seroconverted, and the PBMCs of most recipients demonstrated a proliferative response to parasite antigens in vitro. However, antibody levels and the magnitude of the proliferative responses did not correlate with degree of protection. In general, the development of parasitemia after infection was delayed or suppressed in virtually all vaccine recipients (Patarroyo et al., 1988). A vaccine specific for human chorionic gonadotropin (hCG), which is essential for gestation, is being considered as a potential contraceptive. Over 45 years ago the observation was made by Osterguard (1942) that antibodies against pituitary gonadotropins affect gonadal function. Active immunization of rats and rabbits with bovine luteinizing hormone (LH) or passive immunization with anti-LH antisera produced marked reduction in gonadal function and fertility (Laurence and Ichikawa, 1968; Wakabayaski and Tamaoki, 1966). The synthetic approach to vaccine development was applied to alleviate problems of vaccine expense and cross-reactivity with other endogenous hormones. Stevens et al. (1981) immunized female baboons with residues 109-145 of the &subunit of hCG conjugated to diphtheria toxoid. Antipeptide antibodies reacted well with intact hCG but not with human LH. This conjugated peptide also produced marked reductions in fertility in the female baboons. Levels of antibody reactive with hCG in animals injected with this fragment reached only 5-10% of that observed following injections with native hCG. Recently, a phase 1 clinical trial of 109-145-DTwas completed, which demonstrated that 30 female volunteers produced potentially contraceptive levels of antibodies to hCG (Jones et al., 1988). VII. Conclusions
Although the advantages and disadvantages surrounding the development of synthetic peptide vaccines have been discussed in the course of this review, it may be helpful to list them (Table IV). There are obvious practical advantages to synthetic vaccines as compared to conventional killed, attenuated, or other subunit vaccines. Conventional antiviral vaccines require large cultures to provide sufficient protein for mass immunization, and not all viruses can even be propagated in culture (i.e., HBV). Large cultures require stringent containment procedures and extensive purification protocols to ensure the absence of contaminating
260
DAVID R. MILICH TABLE IV SYNTHETIC VACCINES: THEORETICAL CONSIDERATIONS Advantages
Chemical purity/safety Unlimited source of material costs Stability/storage/delivery Contiguous B cell epitopes on many pathogens Nature of T cell recognition sites Defined immunogen Exclusion of adverse epitopes Adaptability (i.e., mutations) Multivalent (intradinterpathogen) Immunomodulation (i.e., target antibody) Predictive algorithms
Disadvantages Requires amino acid sequence Requires identification of T and B cell sites Noncontiguous B cell epitopes Genetic restriction of T cell recognition Immunogenicity Complexity of protective immune response
substances. Whole virus vaccines must be properly inactivated or contain only attenuated virus. Many virus preparations are unstable and require special handling, which may prevent their worldwide distribution in underdeveloped areas. In contrast, the chemical synthesis of peptide vaccines implies relative purity and safety. Chemical synthesis guarantees an unlimited source of protein for mass vaccination. Synthetic vaccines would be chemically stable and may be dehydrated for storage or delivery to remote areas. Presumably, synthetic vaccines on a large scale would be cost effective assuming the costs of development were not exorbitant. There are also a number of more subtle immunological benefits to the use of synthetic vaccines. First, the synthetic vaccine strategy relies on the occurrence of contiguous B cell epitopes on protein antigens from a number of pathogens and/or the ability of peptide-unique antibodies to bind the intact protein and neutralize infectivity. These characteristics have been demonstrated in a number of synthetic antigen systems. Second, synthetic vaccine design can capitalize on the fact that T cell recognition sites are, for the most part, represented by small peptide fragments. It is clear from many of the studies reviewed herein that synthetic peptides can adequately substitute for the native protein in a number of functionally important situations. Therefore both T and B cell recognitions sites relevant to the native protein can be mimicked by chemical synthesis of small peptides. A chemically synthesized immunogen represents a defined entity, which can be thoroughly characterized with respect to its antigenicity and immunogenicity. For
SYNTHETIC T AND B CELL SITES
261
example, antigenic regions of a protein which activate suppressor mechanisms or are otherwise harmful can be exluded from a synthetic immunogen. A particular T h cell site, which may elicit an allergic or autoimmune response in some vaccine recipients, could be deleted from the immunogen. Furthermore, a defined synthetic immunogen can be easily altered to accommodate viral mutations that may occur over time (i.e., influenza and HIV). Another advantage to defined synthetic epitopes is that they can be used to construct a synthetic vaccine against multiple pathogens. An example of such an approach was a combined diphtheria toxin, streptococcus type 24 M protein, and HBsAg synthetic immunogen conjugated to a single carrier (ChCdid et al., 1983). The synthetic immunogen elicited high-titered antibodies to each of the three antigens. Although pathogen-related T cell recognition sites were not included in this experiment, inclusion of such T cell sites would be expected to prime T h cell memory responses to each pathogen as well as antibody production. Combination of multiple proteins from the same pathogen may also be desirable, as was shown recently in the HBV system (Milich et al., 1988). Nucleoprotein-specificT h cells were shown to elicit antienvelope antibody production as well as antinucleoprotein antibody. This finding and the proposed importance of nucleoprotein-specific T cell responses in HBV infection prompted us to construct a synthetic immunogen composed of nucleoprotein T cell sites and an envelope B cell epitope. The synthetic composite HBV immunogen primed nucleoprotein-specific T h cells and induced envelope-specificantibody production relevant to the native HBV proteins. Another example of intrapathogen multivalency would be to combine B cell epitopes from multiple serotypes of a virus to obtain a broadly cross-reactivevaccine. Perhaps the greatest advantage of a synthetic peptide approach will be its potential to modulate the immune response. For example, regions of a protein which are not normally immunogenic, and therefore have not been subjected to selective pressure by the host’s immune system, may be targeted for antibody production by the use of synthetic immunogens. Similarly, regions that are not T cell recognition sites in the native protein, by virtue of conditions of antigen processing, may activate T cell responses when presented as small peptide fragments. This would allow priming of a beneficial T h cell clone which may be normally silent. However, in order for these strategies to be successful, the memory T and B cell clones elicited in this manner should be recalled by the native protein on the pathogen should humoral immunity wane over time. This may be a general problem when using peptide antigens which elicit peptide-unique antibodies, which cross-react with native proteins but which are not induced by native protein immunization.
262
DAVID R. MILICH
It is possible that once these peptide-specific clones are efficiently activated, there would be some degree of recall by the native protein. Alternatively, neutralizing antibody titers must be maintained. Another application of synthetic peptide strategy to modulate the immune response may be to selectively combine T and B cell recognition sites for use as immunogens. As demonstrated in a number of antigen systems, Th cell fine specificity can effect the fine specificity of the antibody produced. In other words, not all T h cell clones are equal in their ability to provide help to a given B cell clone. Altering the T h cell component of native proteins may qualitatively or quantitatively alter the antibody response. Synthetic immunogens may also provide a means of selectively activating a specific immune cell or T cell subset. In a circumstance in which a T cell response to a particular determinant was desired but an antibody response was not (i.e., autoantibody or immunoglobulin Fc receptor-mediated virus-cell interaction), a vaccine consisting of a synthetic T cell site, which did not induce antibody cross-reactive with the pathogen, would be ideal. Furthermore, the recent findings that peptide fragments can sensitize target cells for lysis by CTLs- and indeed induce CTLs in vitro whereas intact proteins cannot-suggest the possibility that peptide antigens may be used to prime CTLs in vivo if an effective delivery system can be devised. The ability to immunize CTLs in vivo without the use of infectious virus or vector systems would be a tremendous advance in the development of antiviral and possibly antitumor vaccines. Continued efforts to predict T and B cell recognition sites from primary sequence data will be required for the generalized development of synthetic vaccines. Highly predictive algorithms will greatly enhance the feasibility of the synthetic peptide vaccine strategy. There are also a number of obstacles to the development of synthetic vaccines. Obvious requirements are the amino acid sequence of the protein, and identification of the T and B cell recognition sites of interest. This is time consuming and labor intensive, and possibly prohibitive in the absence of a method to predict T and B cell recognition sites. However, one could argue that a detailed characterization of the immunogen, such as is required to identify antibody and T cell recognition sites, should be performed as a matter of course during the development of any vaccine. Another, at least perceived, limitation is the fact that most antibody recognition sites are of the conformational type and cannot be easily duplicated by the present synthetic peptide technology. This limitation may be resolved in time by advances in peptide chemistry. Until such advances, the question remains whether sequential epitopes, whether immunogenic within the native protein or not, can induce sufficient levels of neutralizing antibody to provide adequate protection from infection. Evidence from a number of systems suggests that antibody
SYNTHETIC T AND B CELL SITES
263
produced to predominantly sequential B cell epitopes is sufficient to protect against infection. It appears that the limitation is more a problem of optimization rather than theory. In terms of T cell immunity, the limitation of the peptide strategy is clearly not a chemical and structural one in that T cell antigen recognition and activation can be generally induced by synthetic peptide antigens. The fact that this recognition process involves the dual recognition of MHC gene products as well as the peptide fragment, and that the fine specificity of the T cell antigenic site is determined by the MHC phenotype of the individual, present a problem for synthetic peptide vaccine design. Not all individuals will respond to the same T cell recognition site. For example, in the Th cell response to the nucleoprotein of HBV, H-2 congenic mice, differing only in the MHC, focus on different antigenic determinants; H-2Smice recognize residues 120-131 predominantly, H-2b mice recognize fragment 129-140, H-2fJ4mice respond to residues 100-120, and H-2d strains predominantly respond to residues 85-100 (Milich et al., 1987b, 1988). T cells of a percentage of outbred mice “see” one or the other of these defined determinants or other T cell sites (D. Milich, unpublished observation). Therefore, a composite synthetic vaccine containing numerous T cell sites may be required to ensure T cell responsiveness in an entire outbred population. If one considers that the human MHC is more complex (i.e., more loci) than the murine equivalent, and that there are estimated to be from 50 to 100 different allelic forms of class I1 molecules in outbred murine populations (Klein and Figueroa, 1981), it will be no small task to identify and include sufficient T cell sites into a synthetic vaccine. However, the complexity of the human MHC and the fact that most humans are heterozygous in the MHC may actually increase the potential for any given peptide to interact with MHC gene products. In this regard, it was quite surprising that a high percentage of HIV gpl60-immune donors demonstrated a T cell proliferative response to a single peptide (Berzofsky et al., 1988). Similar findings in a number of additional antigen systems would cause a reevaluation of this theoretical limitation. The finding that synthetic T cell sites need not be present on the same molecule as B cell epitopes (intermolecular Th), as long as they are within the same particulate structure (intrastructural Th) (Milich et al., 1987c), may lessen the burden of delivering multiple T cell sites per B cell epitope inasmuch as individual synthetic T cell sites may be combined in a particulate structure, as discussed previously. Furthermore, a synthetic vaccine could also include B cell sites conjugated to a protein carrier (i.e., tetanus toxoid) in addition to free synthetic peptides consisting of B and T cell sites to ensure antibody production. A number of investigators have reported that synthetic peptides are
264
DAVID R. MILICH
poor immunogens. However, this is a very subjective characterization, depending on the adjuvant systems, the species or strain immunized, the form of the antigen (i.e., monomeric or polymeric), the inclusion or exclusion of relevant T cell epitopes, etc. It is possible that this limitation is more a measure of our lack of understanding of what comprises a strong immunogen, rather than an inherent characteristic of peptide immunogens. Another possible limitation of a synthetic peptide vaccine may lie in the characteristic which is most appealing, its intrinsic simplicity. As host-pathogen interactions have become better understood, it is clear that the host immune response can be anything but simple. Multiple antigens on both structural and nonstructural proteins may serve as inducers of T h cell and/or Ts cell function, targets for CTL, and sites for neutralizing antibody. Viral clearance may require a myriad network of cellular interactions with differing specificities. A disease such as AIDS emphasizes that inducing protective immunity may be no less complex than viral clearance mechanisms. This is not an argument against synthetic vaccines, merely a caution against the assumption that provision of one to several antibody sites on a T cell carrier will necessarily represent an effective vaccine. More than with any other approach to vaccine development, the application of synthetic antigens to vaccine production requires a thorough knowledge of the mechanisms of protective immunity precisely because a synthetic vaccine must be “engineered” to suit the purpose. A number of the immunological advantages and disadvantages discussed may apply equally to both synthetic and recombinant protein antigens. The ultimate choice of a technology for development of a subunit vaccine will be specific to each pathogen given the special circumstances that may pertain. ACKNOWLEDGMENTS The author wishes to thank Drs. J. Bittle, J, A. Benofsky, G. B. Thornton, A. Moriartp and A. McLachlan for helpful discussions and Kathy Carpenter for preparation of the manuscript. I am especially indebted to Janice Hughes and Joyce Jones for their efforts in the research library, which made this review possible. Original work from the author’s laboratory summarized in this paper was supported by Grants A120720 and A100585 from the National Institutes of Health, and a grant from the Johnson and Johnson Company. This is Publication No. 5562-MBfrom the Research Institute of Scripps Clinic.
REFERENCES Adorini, L . , Harvey, A., Miller, A., and Sercarz, E. E. (1979a).J. f i f i . Med. 150, 293. Adorini, L., Miller, A., and Sercarz, E. E. (1979b).J. Immunol. 122, 861. Adorini,L., Muller, S.,Cardinaux, E,Lehmann, I? V., Falcioni, E, and Nagy, Z. (1988). Nature (London) 334, 623.
SYNTHETIC T AND B CELL SITES
265
Allen, P. M., and Unanue, E. R. (1984). J. Immunol. 132, 1077. Allen, P. M., Strydom, D. J., and Unanue, E. R. (1984).Proc. Natl. Acad. Sci. U.S.A. 81, 2489. Allen, P. M., Matsueda, G. R., Haber, E., andunanue, E. R. (1985a).J. Immunol. 135, 368. Allen, P. M., McKean, D. J., Beck, B. N., Sheffield, J., and Glimcher, L. H. (198513). J. Exp. Med. 162, 1264. Allen, P. M., Matsueda, G. R., Evans, R. J., Dunbar, J. B., Jr., Marshall, G. R., and Unanue, E. R. (1987). Nature (London) 327, 713. Altschuh, D., Hartman, D., Reinbolt, J., and Van Regenmortel, M. H. V. (1983).Mol. Immunol. 20. 271. Amit, A. G., Mariuzza, R. A., Phillips, S. E. V., and Poljak, R. J. (1986). Science 233, 747. Anderer, F. A. (1963). Biochim. Biophys. Acta 71, 246. Amon, R., and Sela, M. (1969). Proc. Natl. Acad. Sci. U.S.A. 62, 163. Amon, R., Maron, E., Sela, M., and Anfinsen, C. B. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 1450. Amon, R., Bustin, M., Calef, E.,Chaitchik, S., Haimovich, J., Novik, N., and Sela, M. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 2123. Askonas, B. A., McMichael, A. J., and Webster, R. G. (1982). In “Basic and Applied Influenza Research” (A. S. Beare, ed.), p. 159. CRC Press, Boca Raton, Florida. Atassi, M. Z. (1975). Immunochemistry 12, 423. Atassi, M. Z. (1981). Mol. Immunol. 18, 1021. Atlas, S.A., Christofalo, P., Hesson, T., Sealey, J. E.,and Fritz, L. C. (1985).Biochem. Biophys. Res. Commun. 132, 1038. Audibert, F. M., Jolivet, M. E., Chkdid. L. A., Alouf, J. E., Boquet, P., Rivaille, P., and Siffert, 0. (1981). Nature (London) 289, 593. Audibert, F. M., Jolivet, M. E., Chkdid, L.A., Amon, R.. and Sela, M. (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 5042. Babbitt, B. l?, Allen, P. M., Matsueda, G., Haber, E., andunanue, E. R. (1985). Nature (London) 317, 359. Babbitt, B. P., Matsueda, G., Haber, E., Unanue, E. R., and Allen, P. M. (1986). Proc. Natl. Acad. Scz. U.S.A. 83, 4509. Bahraoui, E. M., Granier, C., Van Rietschoten, J., Rochat, H., and El Ayeb, M. (1986). J. Immunol. 136, 3371. Bahraoui, E. M., El Ayeb, M., Granier, C., and Rochat, H. (1987). Euz J. Biochem. 167,371. Ballou, W. R., Rothbard, J., Wirtz, R. A., Gordon, D. M., Williams, J. S., Gore, R. W., Schneider, I., Hillingdale, M. R.,Beaudoin, R. L., Maloy, W. L., Miller, L. H., and Hockmeyer, W. T. (1985). Science 228, 996. Barcinski, M. A., and Rosenthal, A. S. (1977).J. Ex$. Med. 145, 726. Barkas, T., Juillerat, M., Kistler, J., Schwendimann, B., and Moody, J. (1984). Euz J. Biochem. 143, 309. Bastin, J., Rothbard, J., Davey, J., Jones, I., and Townsend, A. (1987).J. Exp. Med. 165, 1508. Baumhuter, S.,Wallace, C. J. A., Proudfoot, A. E. I., Bron, C., and Corradin, G. (1987). E uzJ . Immunol. 17, 651. Beachey, E. H . , and Dale, J. B. (1988). In “Vaccines 88” (H. Ginsberg, E Brown, R. A. Lemer, and R. M. Chanock, eds.), p. 95. Cold Spring Harbor Lab., Cold Spring Harbor, New York.
266
DAVID R. MILICH
Beachey, E. H.,and Seyer, J. M. (1986).J. Immunol. 136, 2287. Beachey, E. H., Seyer, J. M., Dale, J. B., Simpson, W. A., and Kang, A. H. (1981). Nature (London) 292, 457. Beachey, E. H., Gras-Masse, H. S., Tarter, A., Jolivet, M. E., Audibert, F. M., Chedid, L. A., and Seyer, J. M. (1986). J. Ex$. Med. 163, 1451. Beachey, E. H., Seyer, J. M., and Dale, J. B. (1987).J. Exp. Med. 166, 647. Benacerraf, B. (1978).J. Immunol. 120, 1809. Benacerraf, B., and Cell, P. G. H. (1959). Immunology 2, 53. Benjamin, D. C., Berzofsky, J. A., East, I. J., Gurd, F. R. N., Hannum, C., Leach, S. J., Margoliash, E., Michael, J. G., Miller, A., Prager, E. M., Reichlin, M., Sercarz, E. E., Smith-Gill, S. J., Todd, P. E., and Wilson, A. C. (1984).Annu. Rev. Immunol. 2 , 67. Benjamini, E. (1988).In “Immunochemistry of Proteins” (M. Z. Atassi, ed.), Vol. 2, p. 265. Plenum, New York. Benjamini, E., Langton, B. C., Wan, A. M., Andria, M. L., and h u n g . C. Y. (1986). In “Progress Towards Better Vaccines” (R. Bell and Tomgiani, eds.), p. 1986.Oxford Univ. Press, New York. Berkower, I. J., Buckenmeyer, G. K., Gurd, F.R. N., and Berzofsky, J. A. (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 4723. Berkower, I. J., Matis, L. A., Buckenmeyer, G. K., Gurd, F. R. N., Longo, D. L., and Berzofsky, J. A. (1984)..J.Immunol. 132, 1370. Berkower, I. J., -Kawamura,-H., Matis, L. A., and Berzofsky,J. A. (1985a).J.Immunol. 135, 2628. Berkower, I. J., Buckenmeyer, G. K., and Berzofsky, J. A. (19858).Fed. Proc., Fed. A m . SOC. Ex#. Biol. 44, 1312. Berkower, 1. J., Buckenmeyer, G. K., and Berzofsky, J. A. (1986).J. Immunol. 136, 2498. Bekrzins, K.,Perlmann, H., Carlsson, J., Wahlgren, M., Udomsangpetch, R., Bjorkman, A., Patarroyo, M. E., and Perlmann, P. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 1065. Berzofsky, J. A. (1983). Sum Immunol. Res. 2, 223. Berzofsky, J. A. (1985). Science 229, 932. Berzofsky, J. A. (1987). In “The Antigens” (M. Sala, ed.), Vol. 7, Chapter 1, p. 1. Academic Press, Orlando, Florida. Berzofsky, J. A., Schechter, A. N., Shearer, G. M., and Sachs, D. H. (1977).J. Exp. Med. 145, 123. Berzofsky, J. A., Richman, L. K., and Killion, D. J. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 4046. Berzofsky, J. A,, Buckenmeyer, G. B., Hicks, G., Gurd, F. R. N., Feldmann, R. J., and Minna, J. (1982).J. Biol. Chem. 257, 3189. Berzofsky,J. A., Cease, K. B., Cornette, J. L., Spouge, J. L., Margalit, H., Berkower, I. J., Good, M. F.,Miller, L. H., and De Lisi, C. (1987). Immunol. Rev. 98, 9. Berzofsky, J. A., Bensussan, A., Cease, K. B., Bourge, J. F., Cheynier, R., Lurhuma, Z., Salaun, J. J., Gallo, R. C., Shearer, G. M., and Zagury, D. (1988). Nature (London) 334, 706. Bevan, M. J. (1975a).J.Exp. Med. 142, 1349. Bevan, M. J. (1975b). Nature (London) 256, 419. Bhatnagar, P. K.,Papas, E., Blum, H. E., Milich, D. R., Nitecki, D., Karels, M. J., and Vyas, G. N. (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 4400. Bidart, J.-M., Oxturk, M.. Bellet, D. H., Jolivet, M. E., Gras-Masse, H. S., Troalen,
SYNTHETIC T AND B CELL SITES
267
F., Bohoun, C. J., and Wands, J. R. (1985).J. Immunol. 134, 457. Bidart, J.-M., Bellet, D. H., Alberici, G. F., Van Besien, F., and Bohuon, C. (1987). Mol. Immunol. 24, 339. Bittle, J. L., Houghten, R. A., Alexander, H., Shinnick, T. M., Sutcliffe, J. G., and Lerner, R. A. (1982). Nature (London) 298, 30. Bixler, G. S., Jr., and Atassi, M. Z. (1984).J. Immunogenet. 11, 339. Bixler, G. S., Jr., Yoshida, T., and Atassi, M. Z. (1985). Immunology 56, 103. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S.,Strominger, J. L., and Wiley, D. C. (1987). Nature (London) 329, 506. Blanden, R. V., Doherty, P. C., Dunlop, M. B. C., Gardner, I. D., Zinkernagel, R. M., and David, C. S. (1975). Nature (London) 254, 269. Bloom, B. R.,and Godal, T. (1983). Rev. Inject. Dis. 5, 765. Bluestein, H. G., Green, I., and Benacerraf, B. (1971).J. Exp. Med. 134, 458. Bodmer, H. C., Pemberton, R. M., Rothbard, J. B., and Askonas, B. A. (1988). Cell (Cambridge, Mass.) 52, 253. Bomford, R.,Abdulla, E., Hughes-Jenkins, C., Simpkin, D., and Schmidt, J. (1987). Immunology 62, 543. Bonin, J., Jameson, B. A., Emini, E. A., Diamond, D. C., and Wimmer, E. (1985). Microbiology (Washington, D.C.) p. 255. Borras-Cuesta, E,Petit-Camurdan, A., and Fedon, Y. (1987). EuzJ. Immunol. 17, 1213. Bouhnik, J., Galen, F.-X., Menard, J., Corvol, P., Seyer, R., Fehrentz, J.-A., Le Nguyen, D., Fulcrand, P., and Castro, B. (1987).J. Biol. Chem. 262, 2913. Brett, S. J., Cease, K. B., and Berzofsky, J. A. (1988).J. Ex@. Med. 168, 357. Brostoff, S. W., and Mason, D. W. (1984).J. Immunol. 133, 1938. Brunati, S., Corradin, G., and Bron, C. (1987).J. Immunol. Methods 97, 105. Buus, S., and Werdelin, 0. (1986).J. Immunol. 136, 459. Buus, S.,Colon, S.,Smith, C., Freed, J. H., Miles, C., and Grey, H. M. (1986a). PTOC. Natl. Acad. Sci. U.S.A. 83, 3968. Buus, S., Sette, A., Colon, S. M., Jenis, D. M., and Grey, H. M. (1986b). Cell (Cambridge, Mass.) 47, 1071. Buus, S . , Sette, A., Colon, S. M., Miles, C., and Grey, H. M. (1987). Science 235, 1353. Bystrov, V. F., Okhanov, V. V., Miroshnikov, A. I., and Ovchinnikov, Y.A. (1980). FEBS Lett. 119, 113. Campos-Neto, A,, Levine, H., and Schlossman, S. F. (1982). Cell. Immunol. 69,128. Carbone, F. R., Fox, B. S., Schwartz, R. H., and Patterson, Y. (1987).J. Immunol. 138, 1838. Carbone, F. R., Moore, M. W., Sheil, J. M., and Bevan, M. J. (1988).J. Exp. Med. 167, 1767. Carelli, C., Ralamboranto, L., Audibert, F. M., Gaillard, J.. Briquelet, N., Dray, E, Fafeur, V., Haour, F., andChedid, L. A. (1985). Int..J Immunopharmacol. 7,215. Cease, K. B., Berkower, I., York-Jolley, J., and Berzofsky, J. A. (1986).J. Ex#. Med. 164,1779. Cease, K. B.. Margalit, H., Cornette, J., Putney, S.,Robey, W., Wuyang, C., Streicher, H., Fischinger, P., Gallo, R.,De Lisi, C., and Berzofsky, J. A. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 4249. Celada, E, and Strom, R. (1972). Q Rev. Biophys. 5, 395. Celada, F., Kunkl, A., Manca, F., Fenoglio, D., Fowler, A., Krzych, U., and Sercarz, E. E. (1984). In “Regulation of the Immune System” (H. Cantor, L. Chess, and E. Sercarz, eds.), p. 637. Alan R. Liss, New York.
268
DAVID R. MILICH
Celis, E., Ou, D., and Otvos, L., Jr. (1988).J. Immunol. 140, 1808. Chanh, T. C., Kennedy, R. C., Alderete, B. E., Kanda, P., Eichberg, J. W., and Dreesman, G. R. (1986). Eur. J. Immunol. 16, 1465. Checka,J. M., Stratton, J. A., Miller, A., and Sercan, E. (1976).Eu7:J.Immunol. 6,639. Chcdid, L. A., Jolivet, M., Audibert, F. M., Przewlocki, G., Beachey, E. H., GrasMasse, H. s., and Tartar, A. (1983). Biochem. Biophys. Res. Commun. 117,908. Chen, P. P., Houghten, R. A., Fong, S., Rhodes, G. H., Gilbertson, T. A., Vaughan, J. H., Lerner, R. A., and Carson, D. A. (1984). Proc. Natl. Acad. Sci. U.S. A. 81, 1784. Chestnut, R., Colon, S., and Grey, H. (1982).J. Immunol. 129, 2382. Chou, C.-H. J., Chou, F. C.-H., Kowalski, T. J., Shapira, R., and Kibler, R. F. (1977). J. Neurochem. 28, 115. Chou, C.-H. J., Fritz, R. B., Chou, F. C. H., and Kibler, R. F. (1979).J. Immunol. 123, 1540. Chow, M., Yabrov, R., Bittle, J., Hogle, J., and Baltimore, D. (1985).Proc. Natl. Acad. Sci. U.S.A. 82, 910. Claman, H. N., Chaperon, E. A., andTriplet, R. E (1966). Proc. SOC.E%p. Biol. Med. 122, 1167. Clark-Lewis, I.. Aebersold, R., Ziltener, H., Schrader, J. W., Hood, L. E., and Kent, S. B. H. (1986). Science 231, 134. Clayberger, C., Parham, P., Rothbard, J., Ludwig, D. S.,Schoolnik, G. K., and Krensky, A. M. (1987). Nature (London) 330, 763. Cohen, G. H., Dietzschold, B., Ponce de Leon, M., Long, D., Golub, E., Varrichio, A., Periera, L.,and Eiaenberg, R. J. (1984).J. Virol. 49, 102. Colizzi, V., Palmieri, G., Sette, A., Appella, E., Dona, G., and Adorini, L. (1985). Folia Biol. (Prague) 31, 396. Colman, P. M., Laver, W. G., Varghese,J. N., Baker, A. T., Tulloch, P. A,, Air, G. M., and Webster, R. G. (1987). Nature (London) 326, 358. Cooper, H. M., and Paterson, Y. (1987). Protides Biol. Fluids 35, 523. Cooper, H. M., Todd, P. E. E., and Leach, S. J. (1986). Mol. Immunol. 12, 1289. Copeland, T. D., Xai, W.-P., and Oroszlan, S. (1985). Biochem. Biophys. Res. Commun. 126, 672. Corradin, G., and Chiller, J. M. (1979).J. Ex$. Med. 149, 436. Corradin, G., Juillerat, M. A., Vita, C.. and Engers, H. D. (1983). Mol. Immunol. 20, 763. Craven, D. E., Awdeh, Z. L., Kunches, L. M., Yunis, E. J., Dienstag, J. L., Werner, B. G., Polk, B. F,, Snydman, D. R., Crumpacker, C. S., Grady, G. F., and Alper, C. A. (1986). Ann. Intern. Med. 105, 356. Crisanti, A., Muller, H.-M., Hilbich, C., Sinigaglia, E , Matile, H., McKay, M., Scaife, J., Beyreuther, K., and Bujard, H. (1988). Science 240, 1324. Dale, J. B., and Beachey, E. H. (1982). J. Ex#. Med. 156, 1165. Dale, J. B., Seyer, J. M., and Beachey, E. H. (1983). J. Exp. Med. 158, 1727. David, C. S., Shreffler, D. C., and Frelinger, J. A. (1973).Proc. Natl. Acad. Sci. U.S.A. 70,2509. Davies, A. J. S., Leuchers, E., Wallis, V., Marchant, R., and Elliot, E. B. (1967). Transplantation 5, 222. Day, E. D., and Hashim, G. A. (1984). Neurochem. Res. 9, 1453. Day, E. D., Hashim, G. A,, Varitek, V. A., Jr., and Patenon, P. Y. (1981).J. Neuroimmunol. 1, 205.
SYNTHETIC T AND B CELL SITES
269
De Boer, M., Admiraal, P., Kok, K., Ossendorp, F. A., De Vijlder, J. M., and Tager, J. M. (1987). Hybridoma 6, 655. DeFreitas, E. C., Dietzschold, B., and Koprowski, H. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 3425. De Lisi, C., and Berzofsky, J. A. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 7048. Delmas, A., Guyon-Gruaz, A., Halimi, H., Raulais, D., and Rivaille, P. (1985a). Int. J. Pept. Protein Res. 25, 421. Delmas, A., Milhaud, G., Raulais, D., and Rivaille, P. (1985b). Mol. Immunol. 22, 675. Demotz, S., Lanzavecchia, A., Eisel, U.,Niemann, H., Widmann, C., andcorradin, G. (1988).J. Immunol. (Submitted for publication). Diamond, D. C., Wimmer, E., and Emini, E. A. (1987). In ‘Synthetic Vaccines” (R. Arnon, ed.), Vol. 2, p. 80. CRC Press, Boca Raton, Florida. Dietrich, F, M. (1966). Int. Arch. Allergy Appl. Immunol. 30, 497. Dietzschold, B., Wiktor, T.J., Madarlan, R.,and Varrichio, A. (1982).J. Virol. 44,595. Dietzschold, B., Eisenberg, R. J., Ponce de Leon, M., Golub, E., Hudecz, F., Varrichio, A., and Cohen, G. H. (1984).J. Virol. 52, 431. Dillner, J., Sternas, L., Kallin, B., Alexander, H., Ehlin-Henriksson, B., Jornvall, H., Klein, G., and Lerner, R. (1984). R o c . Natl. Acad. Sci. U.S.A. 81, 4652. DiMarchi, R., Brooke, G., Gale, C., Cracknell, V., Doel, T., and Mowat, N. (1986). Science 232, 639. Dore, I., Altschuh, D., Moudallal, Z. A., and Van Regenmortel, M. H. V. (1987). Mol. Immunol. 24, 1351. Dreesman, G. R., Hollinger, E B., McCombs, R. M., and Melnick, J. L. (1973).J. Gen. Virol. 19, 129. Dreesman, G. R., Sanchez, Y.,Ionescu-Matiu, I., Sparrow,J. T., Six, H. R., Peterson, D. L., Hollinger, E B.,and Melnick, J. L. (1982). Nature (London) 295, 158. Dyrberg, T., and Oldstone, M. B. A. (1986).J. Ex$. Med. 164, 1344. Eardley, D. D., and Sercarz, E. E. (1976).J. Immunol. 116, 600. Eisenberg, R. J., Long, D., Ponce de Leon, M., Matthavs, J. T., Spear, P. G., Gibson, M. G., Lasky, L. A., Berman, P., Golub, E.,and Cohen, G. H. (1985a).J Virol. 53, 634. Eisenberg, R. J., Cerini, C. P., Heilman, C. J., Joseph, A. D., Dietzschold, B., Golub, E., Long, D., Ponce de Leon, M., and Cohen, G. H. (1985b).J. Virol. 56, 1014. Elder, J. H., McGee, J. S., Munson, M., Houghton, R. A., Kloetzer, W., Bittle, J. L., and Grant, C. K. (1987).J. Virol. 61 8. Elkon, K., Bonfa, E., Llovet, R., Danho, W., Weissbach, H., andBrot, N. (1988).Proc. Natl. Acad. Sci. U.S.A. 85, 5186. Ellman, L., Green, I., Martin, W. J., and Benacerraf, B. (1970). Proc. Natl. Acad. Sci. U.S.A. 66, 322. Ellner, J., Lipsky, P., and Rosenthal, A. (1977).J Immunol. 118, 2053. Emini, E. A., Jameson, B. A., and Wimmer, E. (1983). Nature (London) 304, 699. Emini, E. A., Jameson, B. A., and Wimmer, E. (1984). J. Virol. 52, 719. Emini, E. A., Hughes, J. V., Perlow, D. S., and Boger, J. (1985). J. Virol. 55, 836. Emmrich, F., Thole, J., van Embden, J., and Kaugmann, S. H. E. (1986).J. Exp. Med. 163, 1024. Erb, P., and Feldmann, M. (1975).J. Exp. Med. 142, 460. Etlinger, H., Felix, A. M., Gillessen, D., Heimer, E. P., Just, M., Pink, J. R. L., Sinigaglia, F., Sturchler, D., Takacs, B., Trzeciak, A , , and Matile, H. (1988). J. Immunol. 140, 626.
270
DAVID R. MILICH
Fanning, W. E., Smith, J. A., and Rose, G. (1986). Biopolymers 25, 863. Feitelson, M. A. (1986). Hepatology 6, 191. Feitelson, M. A., Millman, I., Duncan, G. D.,and Blumberg, B. S. (1988).J. Med. Virol. 24, 121. Ferguson, M., Evans, D. M. A., Magrath, D. I., Minor, P. D., Almond, J. W., and Schild, G. C. (1985). Virology 143 505. Ferguson, T.A,, Peters, J., Jr., Reed, R., Pesce, A., and Michael, J. G. (1983). Cell. Immunol. 78, 1. Fink, P. J., Matis, L. A., McElligott, D. L., Bookman, M., andHedrick, S. M. (1986). Nature (London) 321, 219. Finnegan, A., Smith, M. A., Smith, J. A., Berzofsky, J. Sachs, D. H., and Hodes, R. J. (1986).J. Exp. Med. 164, 897. Fox, B. S.,Chen, C., Fraga, E., French, C. A., Singh, B., andschwartz, R. H. (1987). J Zmmunol. 139, 1578. Francis, M. J., Fry, C. M., Rowlands, D.J., Brown, R.E,Bittle, J. L., Houghten, R. A., and Lerner, R. A. (1985). J. Gen. Virol. 66, 2347. Francis, M. J., Fry, C. M., Clarke, B. E., Rowlands, D. J., Brown, F., Bittle, J. L., Houghten, R. A., and Lerner, R. A. (1987a). Zn "Vaccines 87" (R.M. Chanock, R. A. Lerner, F. Brown, and H. Ginsberg, eds.), p. 60. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Francis, M. J., Hastings, G. Z.,Syred, A. D., McGinn, B., Brown, F., and Rowlands, D. J. (1987b). Nature (London) 330, 168. Freeman, C. M., Catlow, C. R. A,, Hemmings, A. M., and Hider, R. C. (1986).FEBS Lett. 197, 289. Fritz, R. B., and Chou, C.-H. J. (1983).J. Zmmunol. 130, 2180. Fritz, R. B., Skeen, M.J., Chou, C.-H. J., Garcia, M., and Egorov, I. K. (1985). J. Immunol. 134, 2328. Fuchs, S.,Neumann, D., Safran, A., Souroujon, M., Barchan, D., Fridkin, M., Gershoni, J. M., Mantegazza, R., and Pizzighella, S. (1987). Ann. N.Y. Acad. Sci. 505, 256. Fujii, Y.,and Lindstrom, J. (1988).J. Immunol. 140, 1830. Galen, F.-X., Evin, G., Carelli, C., Bouhnik, J., Michel, J.-B., Ferhentz, J.-A., Le N'Guyen, D.,Seyer, R., Carlson, W. D., Castro, B., and Corvol, P. (1987). Nucl. Med. Biol. 4, 305. Gammon, G. M., Dunn, K., Shastri, N., Oki, A., Wilbur, S., andsercarz, E. E. (1986a). Nature (London) 319, 413. Gammon, G. M., Oki, A., Shastri, N., and Sercan, E. E. (1986b).J.Ex$. Med. 164,667. Gammon, G . , Shastri, N., Cogswell, J., Wilbur, S., Sadegh-Nasseri, S., Krzych, U., Miller, A., and Sercarz, E. (1987). Zmmunol. Rev. 98, 5 3 . Gnann, J. W., Jr., McCormick, J. B., Mitchell, S., Nelson, J. A., and Oldstone, M. B. A. (1987). Science 237, 1346. Gerin, J. L., Alexander, H., Shih, J. W.-K., Purcell, R. H., Dapolito, G., Engle, R., Green, N., Sutcliffe, J. G., Shinnick, T. M., and Lerner, R. A. (1983).Proc. Natl. Acad. Sci. U S A . 80, 2365. Genving, J., and Thompson, K. (1968). Biochemistry 7, 3888. Geysen, H. M., Barteling, S. J., and Meloen, R. H. (1985).R o c . Natl. Acad. Sci. U.S.A. 82, 178. w e n , H. M., Tainer, J. A., Rodda, S. J., Mason, T. J., Alexander, H., Getzoff, E. D., and Lerner, R. A. (1987). Science 235, 1184.
SYNTHETIC T AND B CELL SITES
271
Chose, A. C., and Karush, F. (1988). Mol. Immunol. 25, 223. Good, M. F., Berzofsky, J. A., Maloy, W. L., Hayashi, Y.,Fujii, N., Hockmeyer, W. T., and Miller, L. H. (1986).J. Ex$. Med. 164, 655. Good, M. F,, Maloy, W. L., Lunde, M. N., Margalit, H., Cornette, J. L., Smith, G. L., Moss, B., Miller, L. H., and Berzofsky, J. A. (1987). Science 235, 1059. Good, M. E , Pombo, D., Maloy, W.L., de la Cruz, V. F., Miller, L. H., and Berzofsky, J. A. (1988a).J. Immunol. 140, 1645. Good, M. F., Pombo, D., Quakyi, I. A., Riley, E. M., Houghten, R. A., Menon, A., Alling, D. W., Berzofsky, J. A., and Miller, L. H.(1988b). Proc. Natl. Acad. Sci. U.S.A. 85, 1199. Gordon, R. D., Simpson, E., and Samelson, L. E. (1975).J. Ex#. Med. 142, 1108. Gotch, F. M., McMichael, A. J., Smith, G. L.,and Moss, B. (1986).J.Exfi. Med. 165,
408.
Gotch, E M., Rothbard, J., Howland, K., Townsend, A., and McMichael, A. J. (1987). Nature (London) 326, 881. Green, I., Paul, W. E., and Benacerraf, B. (1966).J.Exp. Med. 123, 859. Green, N., Alexander, H., Olson, A., Alexander, S., Shinnick, T. M., Sutcliffe, J. G., and Lerner, R. A. (1982). Cell (Cambricige, Mass.) 28, 477. Guillet, J.G., Lai, M.-Z., Briner, T. J., Smith, J. A., and Gefter, M. L. (1986). Nature (London) 324, 260. Guillet, J.-G., Lai, M.-Z., Briner, T. J., Buus, S., Sette, A., Grey, H. M., Smith, J. A., and Gefter, M. L. (1987). Science 235, 865. Gulso, N., Dreyfus, M., Siffert, O., Danchin, A., Spyridakis, A., Gargouri, A., Claisse, M., and Slonimski, P. P. (1984). EMBO J. 3, 1769. Gunther, E., Rude, E., and Stark, 0. (1972). Eur. J. Immunol. 2, 151. Hackett, C. J., Dietzschold, B., Gernard, W., Christ, B., Knorr, R., Gillessen, D., and Melchers, F. (1983).J. Ex#. Med. 158, 294. Hackett, C. J.. Hurwitz, J. L., Dietzschold, B., and Gerhard, W. (1985).J. Immunol.
135, 1391.
Hammerling, G. J., Deak, B. D., Mauve, G., Hammerline, U., and McDevitt, H. 0. (1974). Immunogenetics 1, 68. Hamburg, D., Hannum, C., Inman, J. K., Appella, E., Margoliash, E., and Schwartz, R. H. (1981).J. Immunol. 127, 1844. Hamburg, D., Fairwell, T., Schwartz, R. H., and Appella, E. (1983a).J. Immunol. 131, 319. Hansburg, D., Heber-Katz, E., Fairwell, T., and Appella, E. (1983b).J.Ex#. Med. 158,
25.
Hashim, G. A.,Sharpe, R. D., Carvalho, E. F., and Stevens, L. E. (1976).J. Immunol.
116,126.
Hashim, G. A., Day, E. D., Carvalho, E., and Abdelaal, A. (1987).J. Neurosci. Res.
17, 375.
Hauptfeld, V., Klein, D., and Klein, J. (1973). Science 181, 167. Hauptfeld, V., Hauptfeld, M., and Klein, J. (1974).J. Immunol. 113, 181. Heber-Katz, E.,and Dietzschold, B. (1986). Cum Top. Microbiol. Zmmunol. 130, 51. Heber-Kaa, E.,Hamburg, D., andschwartz, R. H. (1983).J. Mol. Cell. Immunol. 1, 3 . Heber-Katz, E., Hollosi, M., Dietzschold, B., Hudecz, F., and Fasman, G. D. (1985). J. Immunol. 135, 1385. Heber-Katz, E., Valentine, S., Dietzschold, B., and Burns-Purzycki, C. (1988).J. Ex#. Med. 167,275.
272
DAVID R. MILICH
Hedrick, S. M., Matis, L. A., Hecht, T. T., Samelson, L. E., Longo, D. L., HeberKatz, E., and Schwartz, R. H. (1982). Cell (Cambridge, Mass.) 30, 141. Heermann, K. H., Goldmann, U., Schwartz, W., Seyffarth, T., Baumgarten, H., and Gerlich, W. H. (1984).J. Virol. 52, 396. Hennessy, K., Fennewald, S.,Hummel, M., Cole, T., and Kieff, E. (1984). R o c . Natl. Acad. Sci. U.S.A. 81, 7207. Herrington, D. A., Clyde, D. F., Losonsky, G., Cortesia, M., Davis, J., Murphy, J. R., Baqar, S.,Felix, A. M., Heimer, E. P., Gillessen, D., Nardin, E., Nussenzweig, R. S.,Nussenzweig, V., Hollingdale, M. R., and Levine, M. M. (1987). Nature (London) 528, 257. Herzenberg, L. A., and Tokuhisa, T. (1980). Nature (London) 285, 664. Hill, S. W., and Sercarz, E. E. (1975). Eur. J. Immunol. 5 , 317. Ho, D. D., Kaplan, J. C., Rackauskas, I. E., andGumey, M. E. (1988). Science259, 1021. Hohlfeld, R., Conti-Tronconi, B., Kalies, I., Bertrams, J., and Toyka, K. V. (1985). J. Immunol. 135, 2393. Hohlfeld, R., Toyka, K. V., Miner, L. L., Walgrae, S. L., and Conti-Tronconi, B. M. (1988).J. Clin. Invest. 81, 657. Holoshitz, J., Drucker, I., Yaretzky, A., van Eden, W., Klajman, A., Lapidot, Z.,Frenkel, A., and Cohen, I. R. (1986). Lancet 2, 305. Hopp, T. F!, and Woods, K. R. (1981). Proc. Nutl. Acud. Sci. U.S.A. 78, 3824. Hurwitz, J. L., Heber-Katz, E., Hackett, C. J., and Gerhard, W. (1984).J. Immunol. 153, 3371. Ichimori, Y.,Kurokawa, T., Honda, S.,Suzuki, N., Wakimasu, M., and Tsukamoto, K. (1985).J. Immunol. Methods 80, 55. Itoh, Y., Takai, E., Ohnuma, H., Kitajima, K., Tsuda, F., Machida, A., Mishiro, S., Nakamura, T., Miyakawa, Y., andMayumi, M. (1986).Proc. Nutl. Acud. Sci. US.A. 83, 9174. Iyer, K. S. N., Sahal, D., and Talwar, G. P. (1986). Int.J. Pept. Protein Res. 27, 604. Jackson, D. C., Murray, J. M., White, D. O..,Fagan, C. N., andTregear, G. W. (1982). Virology 120, 273. Jacob, C. O., Sela, M., and Amon, R. (1983). Proc. Nutl. Acud. Sci. U.S.A. 80, 7611. Jacob, C. O.,Pines, M., and Amon, R. (1984). EMBOJ. 3, 2889. Jacob, C. O.,Arnon, R.,and Sela, M. (1985). Mol. Immunol. 22, 1333. Jacob, C. O.,Grossfeld, S., Sela, M., and Arnon, R. (1986).EuzJ Immunol. 16,1057. Jameson, B. A., Bonin, J., Murray, M., Wimmer, E., and Kew, 0. M. (1985).In “Vaccines 85” (R. Lerner, R. M. Chanock, and F. Brown, eds.), p. 191. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Jemmerson, R., and Paterson, Y. (1986). BzoTechniques 4, 18. Jemmerson, R., Morrow, P. R., Klinman, N. R., and Paterson, Y. (1985). Proc. Nutl. Acud. Scz. U.S.A. 82, 1508. Jerne, N. K. (1971). Eur. J. Immunol. I, 1. Jin, Y.,Shih, J. W.-K., and Berkower, I. (1988).J. Ex@. Med. 168, 293. Jones, W. R., Judd, S.J., Ing, R. Y.,Powell, J.. Bradley, J., Denholm, E. H., Mueller. U. W., Griffin, P. D., and Stevens, V. C. (1988). Lancet I, 1295. Kabilan, L., Troye-Blomerg, M., Perlmann, H., Andersson, G . , Hogh. B., Petersen, E., Bjorkman, A., and Perlmann, F! (1988). Proc. Nutl. Acud. Sci. U.S.A.85, 5659. Kantor, F. S., Ojeda, A., and Benacerraf, B. (1963).J. Exp. Med. 117, 55. Katz, D. H., Paul, W. E.,Goidl, E. A, and Benacerraf, B. (1970).J. f i p . Med. 132, 261. Katz, D.H., Paul, W. E., Goidl, E. A., and Benacerraf, B. (1971).J. f i p . Med. 153, 169.
SYNTHETIC T AND B CELL SITES
273
Katz, D. H., Hamaoka, T., and Benacerraf, B. (1973a).J. Exp. Med. 137, 1405. Katz, D. H., Hamaoka, T., Dorf, M. E., Maurer, P. H., and Benacerraf, B. (1973b). J. Exp. Med. 138, 734. Katz, D. H., Graves, M., Dorf,M. E., Dimuzio, H., and Benacerraf, B. (1975).J. Exp. Med. 141, 263. Katz, M. E., Maizels, R. M., Wicker, L., Miller, A., and Sercarz, E. E. (1982). Euz J. Immunol. 12, 535. Kennedy, R. C., Henkel, R. D.,Pauletti, D., Allan, J. S., Lee, T. H., h e x , M., and Dreesman, G. R. (1986). Science 231, 1556. Kibler, R. F., Fritz, R. B., Chou, F. C.-H., Chou, C.-H., J.. Peacocke, N. Y., Brown, N. M., and McFarlin, D. E. (1977).J. Ex$. Med. 146, 1323. Kindred, B., and Shreffler, D. C. (1972).J. Immunol. 109,940. Kira, J., Bacon, M. L., Martenson, R. F.,Deibler, G. E., Kies, M. W., and Alvord, E. C., Jr. (1986).J. Neuroimmunol. 12, 183. Klein, J., and Figueroa, E (1981). Immunol. Rev. 60, 23. Klein, J., Flaherty, L., Van de Berg, J. L., and Shreffler, D. C. (1978). Immunogenetics 6, 489. Kono, D. H., Urban, J. L., Horvath, S. J., Ando, D. G., Saavedra, R. A., andHood, L. (1988).J. Exp. Med. 168,213. Koszinowski, U., and Ertl, H. (1975). Nature (London) 255, 552. Kris, R. M., Lax, I., Gullick, W., Waterfield. M. D., Ullrich, A., Fridkin, M., and Schlessinger, J. (1985). Cell (Cambridge, Mass.) 40, 619. Krohn, K. J. E., Cease, K. B., Arthur, L. O., Fisher, R., Ranki, A., Putney, S., Lusso, P., Eischinger, P., Gallo, R. C., and Berzofsky, J. A. (1987).In “Vaccines 87” (R. M. Chanock, R. A. Lerner, F. Brown, and H. Ginsberg, eds.), p. 143. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Krohn, K. J. E., Lusso, P.,Gallo, R. C., Ranki, A., Arthur, L. O., Moss, B., and Putney, S. (1988). In “Vaccines 88” (H. Ginsberg, F. Brown, R. A. Lerner, and R. M. Chanock, eds.), p. 357. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Krzych, U., Fowler, A. V., Miller, A., Sercarz, E. E. (1982).J. Immunol. 128, 1529. Krzych, U., Fowler, A. V., and Sercarz, E. E. (1983). In “Protein Conformation as an Immunological Signal” (F. Celeda, V. Schumaker, and E. E. Sercarz, eds.), p. 395. Plenum, New York. Krzych, U., Fowler, A. V., and Sercarz, E. E. (1985).J. Ex$. Med. 162, 311. Kumar, S.,Miller, L. H., Quakyi, I. A., Keister, D. B., Houghten, R. A., Maloy, W. L., Moss, B., Berzofsky, J. A., and Good, M. F. (1988). Nature (hndon) 224, 258. Kurisaki, J,, Atassi, H., and Atassi, M. Z. (1986). Eur J. Immunol. 16, 236. Kyte, J., and Doolittle, R. F. (1982).J. Mol. Biol. 157, 105. Lai, M.-Z., Ross, D. T., Guillet, J.-G., Briner, T. J., Gefter, M. L., and Smith, J. A. (1987).J. Immunol. 139, 3973. Lamb,J. R., and Green, N. (1983). Immunology 50, 659. Lamb, J. R., Eckels, D. D., Lake, P., Woody, J. N., and Green, N. (1982). Nature (Lcndon) 300, 66. Lamb, J. R., Ivanyi, J., Rees, A., Young, R. A., and Young, D. B. (1986). Lepr. Rev. 57(2), 131. Lamb, J. R., Ivanyi, J., Rees, A., Rothbard, J., Howland, K., Young, R.,and Young, D. (1987). EMBOJ. 6,1245. Lancefield, R. C. (1962).J. Immunol. 89,307.
274
DAVID R. MILICH
Langbeheim, H., Arnon, R.. and Sela, M. (1976).Roc. Natl. Acad. Sci. U.S.A.73,4636. Langman, R. E. (1978). Rev. Physiol. Biochem. Phannacol. 81, 1. Lanzavecchia, A. (1985). Nature (London) 314, 537. Laurence, K. A,, and Ichikawa, S. (1968). Endocnbology (Baltimore) 82, 1190. Laver, W. G., Air, G. M., Dopheide, T. A., and Ward, C. W. (1980). Nature (London) 283, 454. Leach, S. J. (1983). Biopolymers 22, 425. Leclerc, C., Przewlocki, G., Schutze, M.-P., and Chedid, L. A. (1987). EuzJ. Immunol. 17, 269. Lee, P., Matsueda, G. R., and Allen, P. M. (1988).J. Immunol. 140, 1063. Leffler, H., and SvanborgEden, C. (1980). Microbiol. Lett. 8, 127. Lennon, V. A.,McCormick, D. J., Lambert, E. H., Griesmann, G. E., and Atassi, M. Z. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 8805. Lerner, R. A. (1982). Nature (London) 299, 592. Lerner, R. A.. Green, N., Alexander, H., Liu, F.-T., Sutcliffe,J. G., andshinnick, T. M. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 3403. Levine, B. B., and Benacerraf, B. (1965). Science 147, 517. Levine, B. B., Ojeda, A,, and Benacerraf, B. (1963). Nature (London) 200, 544. Lifson, J. D., Feinberg, M. B., R e p , G. R., Rabin, L., Banapour, B., Chakrabarti, S., Moss, B., Wong-Staal, E,Steimer, K. S., and Engelman, E. G. (1986).Nature (London) 323,725. Lipkin, W. I., Schwimmbeck, P. L., and Oldstone, M. B. A. (1988).J. Histochem. Cytochem. 36, 447. Livingstone, A. M., and Fathman, C. G. (1987). Annu. Rev. Immunol. 5 , 477. Lusso, P., Ranki, A., Gallo, R. C., Krohn, K. J. E., Markham, P. D., and Kueberuwa, S. S. (1988). In “Vaccines 88” (H. Ginsberg, E Brown, R. A. Lerner, and R. M. Chanock, eds.), p. 361.Cold Spring Harbor Lab., Cold Spring Harbor, New York. McCarron, R., and McFarlin, D. E. (1988). J. Immunol. 141, 1143. McCormick, D. J., and Atassi, M. Z. (1984). Bi0chem.J. 224, 995. McCormick, D.J., Lennon, V. A., and Atassi, M. Z. (1985). Biochem. J. 226, 193. McCray, J., and Werner, G. (1987). Nature (London) 329, 736. McDevitt, H. 0. (1968).J. Immunol. 100, 485. McDevitt, H. O.,and Chinitz, A. (1969). Science 163, 1207. McDevitt, H.0.. and Sela, M. (1965). J. Exp. Med. 122, 517. McDevitt, H. O.,and Tyan, M. L. (1968).J. Exp. Med. 128, 1. McDevitt, H. O.,Deak, B. D., Shreffler, D. C., Klein, J., Stimpfline, J. H., and Snell, G. D. (1972). J. Ex#. Med. 135, 1259. McElligott, D. L., Sorger, S. B., Matis, L. A., and Hedrick, S. M. (1988).J. Immunol. 140, 4123. McMichael, A. J., Gotch, F. M., and Rothbard, J. (1986). J. Exp. Med. 164,1397. Madarlan, R. I., Dietzschold, B., Wiktor, T. J., %el, M., Houghten, R., Lerner, R. A., Sutcliffe, J. G., and Koprowski, H. (1984).J. Immunol. 133, 2748. Machida, A., Kishimoto, S., Ohnuma, H., Baba, K., Ito, Y.,Miyamoto, H., Funatsu, G., Oda. S.,Usuda, S., Togami, S., Nakamura, T.. Miyakawa, Y.. and Mayumi, M. (1984). Gastroenterology 86,910. Magazine, H. I., Carter, J. M., Russell, J. K., Torres, B. A., Dunn, B. M., and Johnson, H. M. (1988). Proc. Natl. Acad. Sci. U.S.A. 85, 1237. Maizels, R. M., Clarke, J. A., Harvey, M. A,, andSercarz, E. E. (1980).EuzJ. Immunol. 10, 509. Manca, F., Clarke, J. A., Miller, A., Sercarz, E. E., andShastri, N. (1984).J. Immunol. 133, 2075.
SYNTHETIC T AND B CELL SITES
275
Margalit, H., Spouge, J. L., Cornette, J. L., Cease, K. B., De Lisi, C., and Berzofsky, J. A. (1987). J. Immunol. 138, 2213. Markoff, L. J., Bray, M., Lai, C.-J., Chanock, R. M., Eckels, K., Summers, P., Gentry, M. K., Houghten, R. A., and Lerner, R. A. (1988). In “Vaccines 88” (H. Ginsberg, F. Brown, R. A. Lerner, and R. M. Chanock, eds.), p. 161. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Martenson, R. E., Deibler, G. E., Kies, M. W., Levine, S.,and Alvord, E. C., Jr. (1972). J. Immunol. 109, 262. Martenson, R. E., Lwine, S., and Sowinski, R. (1975). J. Immunol. 114, 592. Maryanski, J. L., Pala, P., Corradin, G., Jordan, B. R., and Cerottini, J.-C. (1986). Nature (London) 324, 578. Maryanski, J. L., Pala, P., Cerottini, J.-C., and Corradin, G. (1988).J. Exp. Med. 167, 1391. Mazier, D., Mellouk, S.,Beaudoin, R. L., Texier, B., Druihle, l?, Hockmeyer, W., Tmper, J., Paul, C., Charoenvit, Y.,Young, J.. Miltgen, F., Chedid, L., Chigot, J. P., Galley, B., Brandicourt, O., and Gentilini, M. (1986). Science 231, 156. Mehra, V., Sweetser, D., and Young, R. A. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 7013. Milich, D. R., and Chisari, E V. (1982). J. Immunol. 129, 320. Milich, D. R., and McLachlan, A. (1986). Science 234, 1398. Milich, D. R., Leroux-Roels, G. G., Louie, R. E., and Chisari, F. V. (1984a).J. Exp. Med. 159, 41. Milich, D. R., Peterson, D. L., and Chisari, E V. (198413). In “Viral Hepatitis and Liver Disease” (G. N. Vyes, J. L., Dienstag, and J. H. Hoofnagle, eds.), p. 573. Grune & Stratton, New York. Milich, D. R., Thornton, G. B.. Neurath, A. R., Kent, S. B., Michel, M.-L., Tiollais, P., and Chisari, F. V. (1985a). Science 228, 1195. Milich, D. R., McNamara, M. K., McLachlan, A., Thornton, G. B., and Chisari, F. V. (1985b). Proc. Natl. Acad. Sci. U.S.A. 82, 8168. Milich, D. R., Peterson, D. L., Leroux-Roels, G. G., Lerner, R. A., andchisari, F. V. (1985~).J. Immunol. 134, 4203. Milich, D. R., McLachlan, A., Chisari, F. V., and Thornton, G. B. (1986a).J. Exp. Med. 164, 532. Milich, D. R., McLachlan, A., Chisari, F. V., Kent, S. B. H., and Thornton, G. B. (1986b). J. Immunol. 137, 315. Milich, D. R., McLachlan, A., Chisari, F. V., Nakamura, T., and Thornton, G. B. (1986~).J Immunol. 137, 2703. Milich, D. R., McLachlan, A., Moriarty, A., and Thornton, G. B. (1987a).J. Immunol. 138, 4457. Milich, D. R., McLachlan, A., Moriarty, A., and Thornton, G. B. (1987b).J. Immunol. 139, 1223. Milich, D. R . , McLachlan, A., Thornton, G. B., and Hughes, J. L. (1987~).Nature (London) 329, 547. Milich, D. R., Hughes, J. L., McLachlan, A., Thornton, G. B., and Moriarty, A. (1988). Proc. Natl. Acad. Sci. U.S.A. 85, 1610. Miller, J. F. A. P., and Mitchell, G. F. (1967). Nature (London) 216, 659. Miller, J. F. A. l?, Vadas, M. A., Whitelaw, A., and Gamble, J. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 5095. Mills, K. H. G., Skehel, J. J., and Thomas, D. B. (1986). J. Exfi. Med. 163, 1477. Mills, K. H. G., Burt, D. S.,Skehel, J. J., and Thomas, D. B. (1988).J. Immunol. 140, 4083.
276
DAVID R. MILICH
Milton, R. C. de L.. and Regenmortel, M. H. V. (1979). Mol. Immunol. 16, 179. Minden, P., Houghten, R. A., Spear, J. R., and Shinnick, T. M. (1986). In “Vaccines 86”(E Brown, R. M. Chanock, and R. A. Lerner, eds.), p. 225. ColdSpring Harbor Lab., Cold Spring Harbor, New York. Misko, I. S.,Moss, D. J., and Pope, J. H. (1980). Roc. Natl. Acad. Sci. U.S.A. 77, 4247. Mitchell, G. E, and Miller, J. F. A. P. (1968). J. Exp. Med. 128, 821. Mitchell, G. F., Grumet, F. C., and McDevitt, H. 0. (1972). J. Ex#. Med. 135, 126. Mitchison, N. A. (1971). Eur. J. Immunol. 1, 18. Moriarty, A. M., Alexander, H., and Lerner, R. A. (1985). Science 227, 429. Moss, D. J., Rickinson, A. B., and Pope, J. H. (1978). Int. J. Cancer 22, 662. Mozes, E., and McDevitt, H. 0. (1969). Immunochemistry 6, 760. Muckerheide, A,, Pesce, A. J., and Michael, J. G. (1977).J. Immunol. 119, 1340. Miiller, G. M., Shapira, M., and Arnon, R. (1982). Pmc. Natl. Acad. S c i U.S.A.79, 569. Naquet, P., Ellis, J., Singh, B., Hodges, R. S., and Delovitch, T. L. (1987).J. Immunol. 139, 3955. Nardin, E. H., Nussenzweig, V., Nussenzweig, R. S., Collins, W. E., Harinasuta, K. T., Tapchaisri, P., and Chomcharn, Y. (1982). J. Exp. Med. 156, 20. Naylor, P. H., Naylor, C. W., Badamchian, M., Wada, S., Goldstein, A. L., Wang, S.-S., Sun, D. K., Thornton, A. H . , and Sarin, P. S . (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 2951. Neri, P., Antoni, G., Barbarulli, G., Casagli, C., Mariani, M., Nencioni, L., Presentini, R., and Tagliabue, A. (1984). Mol. Immunol. 21, 151. Nestor, J. J., Newman, S. R., DeLustro, B. M., Todaro, G. J., and Schreiber, A. B. (1985). Bzochem. Biophys. Res. Commun. 129, 226. Nestorowicz, A., Tregear, G. W., Southwell, C. N., Martyn, J., Murray, J. M., White, D. O., and Jackson, D. C. (1985). Mol. Immunol. 22, 145. Neurath, A. R., Kent, S. B. H., and Strick, N. (1982). R o c . Natl. Acad. Scz. U.S.A. 79, 7871. Neurath, A. R., Kent, S. B., Strick, N., and Lindsley, F. (1984a). V i m Res. 1, 321. Neurath, A. R., Kent, S. B. H., and Strick, N. (198413). Science 224, 392. Neurath, A. R., Kent, S. B. H., Strick, N., Taylor, P., andStevens, C. E. (1985). Nature (London) 315, 154. Neurath, A. R., Kent, S. B. H., Strick, N., and Parker, K. (1986). Cell (Cumbridge, Mass.) 46, 429. Niman, H. L. (1984). Nature (London) 307, 180. Niman, H. L., Thompson, A. M. H., Yu, A., Markman, M., Willems, J. J., Herwig, K. R., Habib, N. A., Wood, C. B., Houghton, R. A., and Lerner, R. A. (1983). Proc. Natl. Acad. Sci. U.S.A. 80, 4949. Norrby, E., Biberfeld, G., Chiodi, F., von Gegerfeldt, A., Naucler, A., Parks, E., and Lerner, R. (1987). Nature (London) 329, 248. Novotny, J., Handschumacher, M., Haer, E., et al. (1986). R o c . Natl. Acad. Sci. U.S.A. 83, 226. Nussenzweig, V., and Nussenzweig, R. S . (1986a). A m . J. no#. Med. Hyg. 35, 678. Nussenzweig, V., and Nussenzweig, R. S . (1986b). Czba Found. Symp. 119, 150. Oftung, E,Mustafa, A. S.,Husson, R., Young, R. A., andGodal, T. (1987).J. Immunol. 138, 927. O’Hanley, P., Lark, D., Falkow, S., and Schoolnik, G. K. (1985).J. Clin. Invest. 75, 347. Okamoto, H., Imai, M., Usuda, S.,Tanaka, E., Tachibana, K., Mishiro, S.,Machida, A., Nakamura, T., Miyakawa, Y.,and Mayumi, M. (1985).J. Immunol. 134, 1212. Okuda, K., ’Iivining, S. S.,David, C. S.,and Atassi, M. Z. (1979).J. Immunol. 123, 182.
SYNTHETIC T AND B CELL SITES
277
Osterguard, E. (1942). “Antigonadotropic Substances.” Munksguard, Copenhagen. Ovary, Z., and Benacerraf, B. (1963). Proc. SOC. Ex$. Biol. Med. 114, 72. Pala, P., and Askonas, B. A. (1986). Immunogenetics 23, 379. Palfreyman, J. W., Maclean, J. B., Messeder, E., and Sheppard, R. C. (1984).J. Gen. Virol. 65, 865. Palker, T. J., Clark, M. E., Langlois, A. J., Matthews, T. J., Weinhold, K. J., Randall, R. R., Bolognesi, D. P., and Haynes, B. E (1988). R o c . Natl. Acad. Sci. U.S.A. 85, 1932. Papkoff, J., h i , M. H.T., Hunter, T., and Verma, I. M. (1981). Cell (Cambdge, Muss.) 27, 109. Pass, R. F., Whitley, R. J., Whelchel, J. D., Diethelm, A. G., Reynolds, D. W., and Alford, C. A. (1979).J. Inject. Dis. 140,487. Patarroyo, M. E., Parra, C., Pinella, C., del Portillo, P., Lozada, D., Oramas, M., Torres, M., Clavijo, P., Ramirez, C., Fajardo, N., Cruz, N., and Jimenez, C. (1986). In “Vaccines 86” (E Brown, R. M., Chanock, and R. A. Lerner, eds.), p. 219. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Patarroyo, M. E., Romero, P., Torres, M. L.,Clavijo, P., Moreno, A., Martinez, A., Rodriguez, R., Guzman, F., and Cabezas, E. (1987). Nature (London) 328, 629. Patarroyo, M. E.,Amador, R., Clavijo, P., Moreno, A., Guzman, E,Romero, P., Tascon, R., Franco, A., Murillo, L. A., Ponton, G., andTrujillo, G. (1988). Nature (London) 332, 158. Paterson, Y. (1988). Int. Rev. Immunol. (in press). Paul, W. E.,Katz, D. H., Goidl, E. A., and Benacerraf, B. (1970).J. Exp. Med. 132,283. Pearson, C. M. (1964). Arthritis Rheum. 7, 80. Perlmann, P., Berzins, K., Kabilan, L., Perlmann, H., Blomberg, T., Wahlin, M., Bjorkman, A., Hogh, B., Petersen, E., and Patarroyo, E. M. (1987). In “Vaccines 1987” (R. M. Chanock, R. A. Lerner, F. Brown, and H. Ginsberg, eds.), p. 86. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Peterson, L. B., Wilner, G. D.,and Thomas, D. W. (1983).J. Immunol. 130, 2542. Pettinelli, C. B., and McFarlin, D. E. (1981).J. Immunol. 127, 1420. Pettinelli, C. B., Fritz, R.B., Chou, C.-H. J., and McFarlin, D. E. (1982).J.Immunol. 129, 1209. Pfaff, E.,Mussgay, M., Bohm, H. O., Schultz, G. E., and Schaller, H. (1982).EMBOJ. 1, 869. Pierce, C. W., Kapp, J. A., and Benacerraf, B. (1976).J. Ex#. Med. 144,371. Pincus, M., Gerewitz, F., Schwartz, R. H., and Scheraga, H. A. (1983). R o c . Natl. Acad. Sci. U.S.A. 80, 3297. Potocnjak, P., Yoshida, N., Nussenzweig, R. S., and Nussenzweig, V. J. (1980).J. Exp. Med. 151, 1504. Price, J. O., Whitaker, J. N., Vasu, R. I., and Metzger, D. W. (1986).J. Immunol. 136, 2426. Raff, M. C. (1970). Nature (London) 226, 1257. Rajewky, K., and Rottrander, E. (1967). Cold Spring Harbor Symp. Qwmt. Baol. 32,547. Regnier-Vigouroux, A., Ayeb, M. E.,Defendini, M.-L., Granier, C., and Pierres, M. (1988).J. Immunol. 140, 1069. Rhodes, G.,Carson, D. A., Valbracht, J., Houghten, R., andvaughan, J. H. (1985). J. Immunol. 134,211. Robbins, K. C., Devare, S.G., Reddy, E. P., and Aaronson, S. A. (1982). Science 218,1131. Robertson, M. W., and Liu, F.-T. (1988). Mol. Immunol. 25, 103.
278
DAVID R. MILICH
Rock, K. L., and Benacerraf, B. (1983). J. Ex$. Med. 157, 1618. Rose, G. D., Gierasch, L. M., and Smith, J. A. (1985). Adv. Protein Chem. 37, 1. Rosen, J , , Hom, Y.-L., Whalley, A., Smith, R., and Naso, R. B. (1987). In “Vaccines 87”(R. M. Chanock, R. A. Lerner, F. Brown, a n d H . Ginsberg, eds.), p. 188. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Rosenthal, A. S., and Shevach, E. M. (1973).J. Exp. Med. 138, 1194. Rosenthal, A. S., Barcinski, M. A., and Blake, J. T. (1977). Nature (London) 267, 156. Rothbard, J. B. (1986). Ann. Inst. Rzsteur, Paris 137E, 518. Rothbard, J. B., Fernandez, R.,and Schoolnik, G. K. (1984).J. Exp. Med. 160,208. Rothbard, J. B., and Taylor, W. R. (1988). EMBOJ. 7, 93. Rothbard, J. B., Lechler, R. I., Howland, K., Bal, V., Eckels, D. D., Sekaly, R., Long, E. 0..Taylor, W. R., and Lamb, J. R. (1988). Cell (Cambridge, Mass.) 52, 515. Roudier, J., Petersen, J., Rhoda, G. H., Luka, J., andcarson, D. A. (1988). Unpublished. Russell, J. K., Hayes, M. P., Carter, J. M., Torres, B. A., Dunn, B. M., Russell, S. W.. and Johnson, H. M. (1986).J. Immunol. 136, 3324. Sachs, D. H., and Cone, J. L. (1973).J. Ex#. Med. 138, 1289. Schirrmacher, V., and Rajewsky, K. (1970).J. Ex$. Med. 132, 1019. Schluter, S. F., Rosenshein, I. L., Hubbard, R. A., and Marchalonis, J. J. (1987). Biochem. Biophys. Res. Commun. 145, 699. Schmidt, M. A., O’Hanley, P., and Schoolnik, G. K. (1984). J. Ex$. Med. 161, 705. Schmidt, M. A., O’Hanley, P.,Lark, D., andSchoolnik, G. K. (1988). Proc. Natl. Acad. Sci. U.S,A. 85, 1247. Schofield, L., Villaquiran, J., Ferrei, A., Schellekens, H., Nussenzweig, R., and Nussenzweig, V. (1987). Nature (London) 330, 664. Schrier, R. D., Gnann, J. W., Jr., Langlois, A. J., Shriver, K., Nelson, J. A., and Oldstone, M. B. A. (1988).J. Virol. 62, 2531. Schutze, M.-P., Leclerc, C., Jolivet, M., Audibert, E,andchedid, L. (1985).J. Immunol. 135,2319. Schwartz, R. H. (1978). Scand.J. Immunol. 7, 3 . Schwartz, R. H., and Paul, W. E. (1976).J. Ex$. Med. 143, 529. Schwartz, R. H., Fox, B. S., Fraga, E., Chen, C., and Singh, B. (1985).J. Immunol. 135, 2598. Sela, M. (1966). Adv. Immunol. 5, 29. Sela, M. (1969). Science 166, 1365. Sela, M., Schechter, B., Schechter, I., and Borek, F. (1967). Cold SpnngHarbor Symp. Quant. Biol. 32, 537. Sen, S., Houghten, R. A., Sherr, C. J., and Sen, A. (1983). Proc. Natl. Acad. Sci. U.S.A. 80, 1246. Senyck, G . , Williams, E. B., Nitecki, D. E., and Goodman, J. W. (1971).J. Ex$. Med. 133, 1294. Sercarz, E. E., and Shastri, N. (1984). Ann. Immunol. (Paris) 135, 392. Sercarz, E. E., Yowell, R. L., Turkin, D., Miller, A., Araneo, B. A., and Adorini, L. (1978). Immunol. Rev. 39, 108. Sette, A., Colizzi, V., Appella, E., Doria, G., and Adorini, L. (1986). Eur.J. Immunol. 16, 1. Sette, A., Buus, S., Colon, S., Smith, J. A., Miles, C., and Grey, H. M. (1987). Nature (London) 328, 395. Seyer, J. M., Dale, J. B., and Beachey, E. H . (1986). Dew. Bzol. Stand. 63, 101. Shapka, M. (1987). In “Synthetic Vaccines” (R. Arnon, ed.), Vol. 2, p. 51. CRC Press, Boca Raton. Florida.
SYNTHETIC T AND B CELL SITES
279
Shapira. M., Jibson, M., Muller, G., and Arnon, R. (1984). Proc. Natl. Acad. Sci. US.A. 81,2461. Shastri, N., Miller, A., and Sercarz, E. E. (1984).In “Regulation of the Immune System” (H. Cantor, L. Chess, and E. E. Sercarz, eds.), p. 153. Alan R. Liss, New York. Shastri, N., Oki, A., Miller, A., and Sercarz, E. E. (1985).J.Exp. Med. 162, 332. Shearer, G. M. (1974). Eur. J. Immunol. 4, 527. Shearer, G. M., Rehn, T. G., and Garbarino, C. A. (1975).J. Ex$. Med. 141, 1348. Sheng, H. Z.,Martenson, R. E., Grgacic, E. V., Dowse, C. A., Carnegie, R. L., and Bernard, C. C. A. (1988).J. Neuroimmunol. 17, 183. Sheriff, S., Silverton, E. W., Padlan, E. A., Cohen, G. H., Smith-Gill, S. J., Finzel, B. C., and Davies, D. R. (1987). Proc. Natl. Acad. Sci. U S A . 84, 8075. Shih, J. W.-K., Gerety, R. J., Liu, D. T., Yajima, H., Fuji, N., Nomizu, M., Hayashi, Y., and Katakura, S. (1983). In “Modem Approaches to Vaccines” (R. M. Chanock and R. A. Lerner, eds.), p. 127. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Shih, W. H., Matzinger, P. C., Swain, S. L.,and Dutton, R. W. (1980).J. Ex$. Med. 152, 1311. Shimonkevitz, R., Colon, S., Kappler, J. W., Marrack, P., and Grey, H. M. (1984). J. Immunol. 133, 2067. Shinnick, T. M., Sutcliffe, J. G., Green, N., and Lerner, R. A. (1983). Annu. Rev. Microbiol. 37, 415. Shreffler, D., David, C., Gotze, D., Klein, J., McDevitt, H., and Sachs, D. (1974). Immunogenetics 1, 189. Siliciano, R. F., Lawton, T., Knall, C., Karr, R. W., Berman, P., Gregory, T., and Reinherz, E. L. (1988). Cell (Cambridge, Mass.) 54, 561. Singer, A. S., Hathcock, K. S., and Hodes, R. J. (1979).J. Ex#. Med. 149, 1208. Sinigaglia, F., Guttinger, M., Gillessen, D., Doran, D. M., Takacs, B., Matile, H., Trzeciak, A., and Pink, J. R. (1988). Eul: J. Immunol. 18, 633. Slovin, S. F., Schooley, R. T., andThorley-Lawson, D. A. (1983).J.Imrnunol. 130,2127. Sodroski,J., Goh, W. C., Rosen, C., Campbell, K., and Haseltine, W. A. (1986). Nature (London) 322, 470. Solinger, A. M., Ultee, M. E., Margoliash, E., andSchwartz, R. H. (1979).J.Exfi. Med. 150, 830. Song, E. S., Linsk, R., Olson, C. A., McMillan, M., andGoodenow, R. S. (1988). R o c . Natl. Acad. Sci. U S A . 85, 1927. Spouge, J. L.,Guy, H. R., Cornette, J. L., Margalit, H., Cease, K., and Berzofsky, J. A. (1987).J. Immunol. 138, 204. Stevens, V. C. (1976).In “Development of Vaccines for Fertility Regulation,” W.H.O. Symposium, Varna, Scriptor, Copenhagen. Stevens, V. C. (1986). Immunol. 7bday 7, 369. Stevens, V. C., Powell, J. E., Lee, A. C., and Griffin, P. D. (1981).Fertil. Steril. 36,98. Stevens, V. C., Chou, W.-S., Powell, J. E., Lee, A. C., and Smoot, J. (1986).Immunol. Lett. 12, 11. Steward, M. W., Sisley, B. M., Stanley, C., Brown, S. E.,and Howard, C. R. (1988). Clin. Exp. Immunol. 71, 19. Strynadka, N. C. J., Redmond, M. J., Parker, J. M. R., Scraba, D. G., and Hodges, R. S. (1988).J. Virol. 62, 3474. Sugamura, K . , and Hinuma, Y. (1980).J. Immunol. 124, 1045. Sutcliffe, J. G., Shinnick, T. M., Green, N., Liu, F.-T., Niman, H. L., and Lerner, R. A. (1980). Nature (London) 187, 80.
280
DAVID R. MILICH
Sutcliffe,J. G., Shinnick, T. M., Green, N., and Lerner, R. A. (1983).Science 219,660. Suzuki, G., and Schwartz, R. H. (1986).J. Immunol. 136, 230. Swanborg, R. H. (1975).J.Immunol. 114, 191. Swierkosz, J. E., and Swanborg, R. H. (1975).J. Immunol. 115, 631. Swierkosz, J. E., Rock, K., Marrack, P., andKappler, J. W. (1978).J. Exp. Med. 147,554. Sytkowski, A. J., and Donahue, K. A. (1987).J. Biol. Chem. 262, 1161. Sytkowski, A. J., and Fisher, J. W. (1985).J. Biol. Chem. 260, 14727. Tainer, J. A., Getzoff, E. D., Alexander, H., Houghten, R.A,, Olson, A. J., and Lerner, R. A. (1984). Nature (London) 312, 127. Tainer, J. A., Getzoff, E.D., Paterson, Y. A. et al. (1985). Annu. Rev. Immunol. 3,423. Takagaki, Y., Hirayama, A., Fujio, H., and Amano, T. (1980). Biochemistry 19,2498. Takahashi, H., Cohen, J., Hosmalin, A., Cease, K. B., Houghten, R., Cornette, J. L., De Lisi, C., Moss, B., Germain, R. N., and Berzofsky, J. A. (1988). Proc. Natl. Acad. Sci. U.S.A. 85, 3105. Talwar, G. P.,Sharma, N. C., Dubey, S. K., Salahuddin, M., Das, C., Ramakrishnan, S., Kumar, S., and Hingorani, V. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 218. Tamura, T., Bauer, H., Birr, C., and Pipkorn, R. (1983).Cell (Cambridge, Mass.) 19,587. Tanaka, T., Slamon, D. J., and Cline, M. J. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 3400. Taylor, P. M., Davey, J., Howland, K., Rothbard, J. B., and Askonas, B. A. (1987). Immunogenetics 26, 267. Thomas, D. W.,Clement, L., and Shevach, E. M. (1978). Immunol. Rev. 40, 182. Thomas, D. W., Meltz, S. K., and Wilner, G. D. (1979). J. Zmmunol. 123, 759. Thomas, D. W., Hsieh, K.-H., Schauster,J. L., Mudd, M. S.,and Wilner, G. D. (1980). J. Exp. Med. 152, 620. Thomas, D. W., Hsieh, K.-H., Schauster, J. L., and Wilner, G. D. (1981).J. Ex#. Med. 153, 585. Thomas, D. W., Hoffman, M. D., and Wilner, G.D. (1982).J. Ex$. Med. 156, 289. Thomas, J. W., Danho, W., Bullesbach, E., Fohles, J., and Rosenthal, A. S. (1981). J. Immunol. 126, 1095. Thorley-Lawson, D. A.,and Israelsohn, E. S. (1987).Proc. Natl. Acad. Sci. U.S.A. 84, 5384. Thorley-Lawson, D. A.,Chess, L., and Strominger, J. L. (1977).J. Exp. Med. 146,495. Thornton, G. B., Milich, D., Chisari, F., Mitamura, K., Kent, S. B., Neurath, R., Purcell, R., and Gerin, J. (1987).In "Vaccines 87" (R. M. Chanock, R. A. Lerner, F. Brown, and H. Ginsberg, eds.), p. 77. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Thornton, J. M., Edwards, M. S., Taylor, W. R., and Barlow, D. J. (1986). EMBOJ. 5, 409. Todd, P. E. E., East, 1. J., and Leach, S. J. (1982). nends Biochem. Sci. 7 , 212. Townsend, A. R. M.. and McMichael, A. J. (1985). Prog. Allergy 36, 10. Townsend, A. R.M, McMichael, A. J., Carter, N. P., Huddleston, J. A., and Brownlee, G. G. (1984). Cell (Cambridge, Mass.) 39, 13. Townsend, A. R. M., Gotch, F. M., and Davq, J. (1985). Cell (Cambridge, Mass.) 42,457. Townsend, A. R. M., Rothbard, J., Gotch, F. M., Behadur, G., Wraith, D., and McMichael, A. J. (1986). Cell (Cambridge, Mass.) 44,959. Trudel, M., Nadon, F.,Seguin, C., Dionne, G., and Lacroix, M.(1987). J. Gen. Virol. 68 2273. Tuohy, V. K., Lu,Z.,Sobel, R. A., Laursen, R. A., and Lees, M. B. (1988).J.Immunol. 141, 1126.
SYNTHETIC T AND B CELL SITES
281
Turkin, D., and Sercarz, E. E. (1977). Proc. Natl. Acad. Sci. U.S.A. 74, 3984. Unanue, E. R. (1984). Annu. Rev. Zmmunol. 2, 395. Unanue, E. R., and Allen, P. M. (1987). Science 236, 551. Vallotton, M. B. (1974). Handb. Ex#. Pharmakol. p. 185. Vandenbark, A. A., Gill, T., and Offner, H. (1985a).J. Zmmunol. 135, 223. Vandenbark, A. A., Offner, H., Reshef, T., Fritz, R., Chou, C.-H. J., and Cohen, I. R. (1985b).J. Ex#. Med. 135, 229. van Eden, W., Holoshitz, J., Nevo, Z., Frenkel, A., Klajman, A., and Cohen, I. R. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 5117. van Eden, W., Thole, J. E. R., van der Zee, R., Noordzij, A,, van Embden, J. D. A., Hensen, E. J., and Cohen, I. R. (1988). Nature (London) 331, 171. Van Regenmortel, M. H. V. (1984). In “Hybridoma Technology in Agricultural and Veterinary Research” (N. Stem and H. R. Gamble, eds.), p. 43. R o m a n h Allenheld, Totowa, New Jersey. Van Rietschoten, J., Granier, C., Rochat, H., Lissitzky, S.,and Miranda, F. (1975).Eul: J. Biochem. 56, 35. von Boehmer, H., Haas, W., and Jerne, N. K. (1978). Proc. Natl. Acad. Sci. U.S.A. 75, 2439. Vyas, G. N., Rao, K. R., and Ibrahim, A. B. (1972). Science 178, 1300. Wabuke-Bunoti, M. A., Taku, A., Fan, D. P., Kent, S., and Webster, R. G. (1984). J. Immunol. 133, 2194. Wakabayaski, D., and Tamaoki, B. (1966). Endocrhology (Baltimore) 79, 477. Waldor, M. S., Sriram, S.,Hardy, R., Herzenberg, L. A., Lanier, L., Lim, M., and Steinman, L. (1985). Science 227, 415. Walker, M. E., Szmuness, W., andStevens, C. (1981). Proc. A m . Assoc. Blood Bank4, 4. Wallace, L. E., Rickinson, A. B., Rowe, M., and Epstein, M. A. (1982). Nature (London) 297, 413. Walter, G., Scheidtmann, K.-H., Carbone, A., Laudano, A. P., and Doolittle, R. F. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 5197. Wan, A, M., Langton, B. C., Andria, M. L., andBenjamin, E. (1986).Mol. Zmmunol. 23, 467. Watari, E . , Dietzschold, B., Szokan, G., and Heber-Katz, E. (1987).J.EX#.Med. 165, 459. Watts, T. H., and McConnell, H. M. (1986). R o c . Natl. Acad. Sci. U.S.A. 83, 9660. Watts, T. H., Gariepy, J., Schoolnik, G. K., and McConnell, H. M. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 5480. Watts, T. H., Gaub, H. E., and McConnell, H. M. (1986). Nature (London) 320, 179. Weiss, W. R., Sedegah, M., Beaudoin, R. L., Miller, L. H., and Good, M. E (1988). Proc. Natl. Acad. Sci. U.S.A. 85, 573. Welling, G. W., Weijer, W. J., van der Zee, R. ,and WellingWester, S. (1985).FEBS Lett. 188, 215. Werdelin, 0. (1982). J. Zmmunol. 129, 1883. Westhof, E.. Altschuh, D., Moras, D., Bloomer, A. C., Mondragon, A., Klug, A., and Van Regenmortel, M. H. V. (1984). Nature (London) 311, 1123. Wheat, T. E., Shelton, J. A., Gonzales-Prevatt, V., and Goldberg, E. (1985). Mol. Immunol. 22, 1195. Whitaker, J. N. (1982).J. Immunol. 129, 2729. Whitnack, E., and Beachey, E. H. (1982).J. Clin. Invest. 69, 1042. Whitton, J. L., Lewicki, H., Gebherd, J. R., Tishon, A., Southern, P. J., and Oldstone, M. B. A. (1988). In “Vaccines 88’(H. Ginsberg, E Brown, R. A. Lerner, and R. M.
282
DAVID R. MILICH
Chanock, eds.), p. 215. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Wiley, D. C., Wilson, I. A., and Skehel, J. J. (1981). Nature (London) 298, 373. Wilson, I . A., Skehel, J. J., and Wiley, D. C. (1981). Nature (London) 289, 366. Wilson, I. A, , Niman, H. L., Houghten, R. A,, Cherenson, A. R., Connolly, M. L., and Lerner, R. A. (1984). Cell (Cambridge, Mass.) 37, 767. Wilson, L. A., Karabin, J. M., Smith, J. W., Dawson, D., and Scott, D. W. (1984). J. Immunol. 132, 1522. Wong, T. W., and Goldberg, A. R. (1981). Proc. Natl. Acad. Sci. U S A . 78, 7412. Wychowski, C., van der Werf, S., Siffert, O., Crainic, R., Bruneau, P., and Girard, M. (1983). EMBOJ. 2, 2019. Yano, A., Schwartz, R. H., and Paul, W. E. (1977). J. Exp. Med. 146, 828. Yewdell, J. W., Bennink, J. R., Smith, G. L., and Moss, B. (1985). Proc. Natl. Acad. Sci. U.S.A. 82, 1785. Yoshioka, M., Yoshioka, N., and Atassi, M. Z. (1986). Biochem. J. 234, 449. Young, C. R . , and Atassi, M. Z. (1983a). J. Immunogenet. 10, 139. Young, C. R . , and Atassi, M. Z. (1983b). J. Immunogenet. 10, 151. Yowell, R. L., Araneo, B. A., Miller, A. and Sercarz, E. E. (1979). Nature (London) 279, 70. Zamvil, S., Nelson, P. Trotter, J., Mitchell, D., Knobler, R., Fritz, R., and Steinman, I. (1985a). Nature (London) 317, 355. Zamvil, S . , Nelson, P. A., Mitchell, D. J., Knobler, R. L., Fritz, R. B., and Steinman, L . (1985b).J. Exp. med. 162, 2107. Zamvil, S. S., Mitchell, D. J.. Moore, A. C., Kitamura, K., Steinman, L, and Rothbard, J. B. (1986). Nature (London) 324, 258. Zavala, F.,Tam, J. P., Hollingdale, M. R., Cochrane, A. H., Quakyi, I., Nussenzweig, R . S., and Nussenzweig, V. (1985). Science 228, 1436. Zavala, E,Tam, J. P.. Barr, P.J., Romero, P.J., Ley, V., Nussenzweig, R. S., and Nussenzweig, V. (1987).J. Ex@. Med. 166, 1591. Ziegler, H. K., and Unanue, E. R. (1981). J. Immunol. 127, 1869. Ziegler, H. K., and Unanue, E. R. (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 175. Ziltener, H. J., Clark-Lewis, I., Fazekas d e St. Groth, B., Hood, L. E., Kent, S. B. H . , and Schrader, J. W. (1987). J. Immunol. 138, 1105. Ziltener, H. J., Clark-Lewis, I., Fazekas d e St. Groth, B., Organ, P. C., Hood, L. E., Kent, S. B. H., and Schrader, J. W. (1988). J. Immunol. 140, 1181. Zinkernagel, R. M., and Doherty, P. C. (1974). Nature (London) 248, 701. Zweig, M., Showalter, S. D., Simms, D. J., and Hampar, B. (1984). J. Virol. 51, 430.
ADVANCES IN IMMUNOLOGY. VOL. 45
Rationale for the Development of an Engineered Sporozoite Malaria Vaccine VICTOR NUSSENZWEIG" and RUTH S. NUSSENZWElGt *Department of Pathology and Kaplan Cancer Center, and tDepartment of Modlcal and Molecular Pamsitology, New York Unlvenity Medical Center, New York, New York 10016
I. Introduction
Humans are infected by four species of malaria parasites. The greatest attention, however, has been given to Plasmodium falciparum, because it affects millions of people, causes a potentially lethal disease, and is becoming increasingly resistant to chemoprophylaxis and chemotherapy. Infection by Plasmodium vivax is also widespread and causes an acute or subacute disease, chronic anemia and splenomegaly, but is not usually lethal or drug resistant. At the other end of the spectrum, Plasmodium malariue is well adapted to the human host and produces chronic illness which can persist for many years. Infection with Plasmodium ovule is rare and occurs mainly in West Africa. The symptoms are similar to those of I? uimx infection, but relapses are less frequent. Attempts to control the spread of the infection by combating the Anopheles mosquito vectors have failed. Some have become resistant to insecticides, and many species or subspecies are not strictly domiciliary and are therefore very difficult to eliminate. For these reasons vaccine development has become an important research priority. Unfortunately, however, this task is complicated by several problems-some are associated with unique features of this parasite and others, of a more fundamental nature, are associated with the obstacles facing the design of rational, efficient, engineered subunit vaccines that contain multiple B and T cell epitopes. One of the problems in developing a malaria vaccine is that this parasite has a very complex life cycle, and each stage bears different protective antigens. Therefore, protective immunity is stage specific, although, as discussed below, there are exceptions to this rule. Another difficulty is that sufficient amounts of antigen can only be obtained by genetic engineering or chemical synthesis. This requires the identification of the relevant protective antigens, which is no simple task considering the complexity and variability of the parasite. In spite of these 283 Copyright Q 1989 by Academic Press, lnc. All rights of reproduction in any form reserved.
284
VICTOR NUSSENZWEIG AND RUTH S . NUSSENZWEIG
formidable obstacles, considerable progress has been made in the last few years, and a few candidate malaria vaccines have been tried in humans with partial success. In this review, we will only discuss the rationale for developing sporozoite vaccines. Sporozoites are the infective stages of the parasite found in the salivary glands of female Anopheles mosquitoes. They are introduced into the host's circulation by the bite of the mosquito, and in a few minutes they enter hepatocytes. There they multiply by schizogony inside a vacuole and develop into exoerythrocytic forms (EEFs). After a few days, each EEF contains thousands of newly formed parasites, the merozoites. These exit the hepatocyte and enter the circulation to start the red blood cell cycle, which is associated with the clinical symptoms of malaria. Some merozoites develop inside red cells into the sexual stages, or gametocytes. which are taken up by the mosquitoes during their blood meal to complete the cycle (Fig. 1). Recent reviews of studies dealing with blood-stage and gametocyte vaccines are to be found elsewhere (1-3).
~~
Anopheles SPOROZOITES ,Salivary Gland)
~~
Mammalian Host
-
L
SPOROZOITES
c
Blood Stream
E XOERY THROCY TIC STAGES ( Hepotocyte) I
OOCYSTS Widgut Wall)
)OKY NETES ( Midgut)
TROPHOZOITES
RING FORMS GAMETES
a"?
MEROZOITES (Blood Streom)
GAMETOCYTES
FIG.1. Schematic representation of the life cycle of the malaria parasite. The main targets of a vaccine are the sporozoites, merozoites, and exoerythrocytic stages. Gametocytes leave the red cell in the gut of mosquitoes, where they fuse to form the gametes. Both gametocytes and gametes can also be recognized by antibodies ingested by mosquitoes.
SPOROZOITE MALARIA VACCINES
285
II. Sporozolte-Induced Immunity to Malaria
Although protection against avian malaria by immunization with ultraviolet-irradiated sporozoites was demonstrated in the 1940s (4,5), this line of investigation was not followed for more than 20 years. For a long time the concept predominated that protective immunity to mammalian sporozoites was impossible to achieve because of the very short time during which they are accessible in the ciculation. There was also uncertainty whether the findings in avian malaria were applicable to mammalian malaria, because of the profound differences in the life cycles, particularly in the initial phases of infection: for example, avian malaria sprozoites do not invade hepatocytes, but rather macrophages, where they develop into EEFs. The initial observations of sporozoite-induced protection in mammalian malaria were made in mice injected intravenously with y-irradiated Plasmodium berghei (6-8). The degree of protection was dependent on the dose of sporozoites used for immunization, but some of the animals survived challenge with a lethal dose of sporozoites following a single immunization with a few thousand attenuated parasites. Increasing the number of immunizing doses to three or more resulted in protection of more than 90% of the mice, even when the challenge dose contained thousands of viable sporozoites. This is a remarkable degree of protection considering that the number of sporozoites introducted by an infected mosquito is very small, most likely only a few hundred, or even fewer. The total number of sporozoites in the salivary glands of mosquitoes from endemic areas is usually less than lo4, and a very small proportion is injected. An estimate of the number of sporozoites injected by a feeding mosquito was recently obtained by allowing A. stephensi infected with €? faleiparum sporozoites to salivate for up to 10 minutes into oil droplets. The median number of ejected sporozoites was only 38, or 0.5% of gland infection (8a). Although the protection lasted only a few months, resistance to infection could be maintained in mice for more than 1 yr by subjecting them periodically to the bite of I? berghez-infected, nonirradiated mosquitoes (9). This suggested that in endemic areas the periodic injection of sporozoites by mosquito bite would have the same effect, that is, would boost resistance to infection. Also of interest was the observation that sporozoite-immunized, adult female mice transferred their immunity to their offspring through their milk (10). Of mice born to nonimmunized mothers, and nursed by immune foster mothers, 55 % resisted an infective challenge, while age-matched controls foster-nursed by nonimmunized mothers were uniformly susceptible. The log2 of antisporozoite serum
286
VICTOR NUSSENZWEIC AND RUTH S. NUSSENZWEIC
titers of the mice nursed in the immunized mothers was usually only twofold lower than the titer in the serum of the mother, and the passively transferred antibody was of IgG class. These experiments do not rule out, however, the concomitant transfer of immune cells, which could contribute to the immunity of the offspring (11). Immunization with viable P berghei sporozoites, administered by intravenous injection or by mosquito bite, was also effective. In these experiments, chloroquine was added to the drinking water of the immunized animals to suppress development of erythrocyticforms without affecting the EEFs (12). Using the same or similar methods, sporozoite-inducedprotection has been obtained in monkey and human malarias. Rhesus monkeys were protected against challenge with Plasmodium cynomolgi and Plasmodium knowlesi sporozoites by vaccination with y-irradiated parasites (13-15). However, in contrast to the findings in the P berghei rodent model, the monkeys had to be injected with higher doses and multiple times in order to achieve optimal results. Clyde and other investigators vaccinated humans with sporozoites (16-21). The method utilized was unique (22), that is, the volunteers were subjected to the bite of hundreds of infected mosquitoes which had been y-irradiated. (Sporozoites obtained by dissection of salivary glands of mosquitoes are heavily contaminated with mosquito tissue and cannot be injected into humans. When the radiation-attenuated sporozoites are delivered by mosquito bite, the amount of mosquito-contaminating material is kept to a minimum.) A total of 13 volunteers were vaccinated against falciparum. Only after multiple exposures to the infected and irradiated mosquitoes, over periods of time ranging from 3 to 10 months, were six volunteers protected against challenge with the same or different strains of sporozoites of P falciparum. Two out of five volunteers were protected against I? vivax infection when immunized by the bite of P miax-infected mosquitoes (23). A. IMPORTANCEOF VIABILITY AND ROUTEOF IMMUNIZATION
Sporozoites of €? berghei had to be viable in order to mediate a strong degree of protective immunity. Various attempts have also been made to achieve protection using formalin-treated or lyophilized sporozoites, or crude extracts of the parasite, but they were not successful (24). On the other hand, if the viability of the irradiated sporozoites was increased by adding mouse serum to the incubation medium, the doses and number of injections required to establish a solid immunity were reduced (25). As an explanation for these findings, it has been suggested that for effective immunity, the I? berghei sporozoites have to enter hepatocytes and that antigens expressed following invasion may be required (26). This seems unlikely for several reasons. Primary immune responses usually
SPOROZOITE MALARIA VACCINES
287
do not occur on the periphery but in central lymphoid organs, and hepatocytes do not express class I1 MHC antigens and should not function as antigen-presenting cells. In addition, mice are not the natural host of I? berghei, and only about 1% of the few thousand sporozoites used for immunization enter hepatocytes (27). Most sporozoites are cleared rapidly from the mouse circulation and are taken up by macrophages in the spleen (27-29). Removal of the spleen prior to a single administration of I? berghei sporozoites markedly diminishes protective immunity and the production of specific antibodies (28). We believe that to elicit a vigorous immune response to the irradiated sporozoites, it is essential that products from the viable parasites enter macrophages and other antigen-presenting cells, mainly of the spleen or liver. In the I? berghei model, protective immunity is relatively easy to achieve probably because about 99% of the irradiated sporozoites are taken up by the macrophages. When sporozoites are introduced in a susceptible or natural host, a large proportion enter hepatocytes (30-32). This is most likely one of the reasons why, in endemic areas, it may take many years before children develop immunity to sporozoites (33), and may also be the explanation for the greater difficulty in successfully vaccinating rhesus monkeys with irradiated sporozoites of I? cynomolgi or I? knowlesi (34,35), and humans with irradiated sporozoites of I? falcipurum (16-21) or I? Vivax (23). The superiority of attenuated versus killed sporozoitesprobably reflects the need for adequate antigen processing. It could be argued that only when sporozoites are metabolically active do the antigens involved in protective immunity enter the intrinsic route of processing to generate class I MHC-associated peptides and cytotoxic cells. (This is futher discussed in Section V.) Another unusual characteristic of protective immunity to sporozoites in adult mice is that immunization by the intravenous route is required. Repeated administration of irradiated sporozoites of I? berghei by the intraperitoneal, intracutaneous, and intramuscular routes or per 0 s failed to induce good degrees of protection reproducibly (24). This is consistent with the idea that in order to induce immunity, the irradiated sporozoites have to enter macrophages of the spleen or the liver while they are alive. Sporozoites lose their infectivity very quickly, and this will not happen unless they are injected directly into the bloodstream. AND SPECIES SPECIFICITY OF SPOROZOITE-INDUCED B. STAGE PROTECTION In rodent malaria, sporozoite-inducedprotection is stage specific (36). Mice immunized with and protected against sporozoites of I? berghei were susceptible to infection with red blood cells infected by the parasite,
288
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
and developed parasitemias similar to those of control animals (37). The species specificity of protection has not been extensively studied in rodent malaria, but some examples of cross-protection have been noted. Mice immunized with €? berghei sporozoites resisted challenge with sporozoites of Plasmodium vinckei and Plasmodium chabaudi (8). This crossprotection was accompanied by serological cross-reactivity (7). In human malaria, the protection was also stage specific: a volunteer who resisted two challenges with sporozoites of 19 falciparum developed malaria when injected with blood stages of €? falciparum. Immunity was also species specific. No resistance to challenge with sporozoites of €? vivax was observed in a I! falciparum-immunized host and vice versa. Notably, however, extensive protection against the various geographical isolates and strains of the homologous species was observed. For example, immunization with a strain of €? falcijmrum from Burma induced protection against sporozoite challenge by strains from Malaysia, Panama, and the Philippines. 1. Mechanism of Protective Immunity: Earlier Studies
When the sera of vaccinated animals were incubated with viable sporozoites, characteristic morphological changes were seen by phase microscopy: a threadlike precipitate appeared at the posterior end of the parasite (38,39) (Fig. 2). This reaction, designated the circumsporozoite, or CSP, reaction, was temperature dependent and was later shown to represent the shedding of the parasite surface coat when it is cross-linked by antibodies (40). It is unlikely that the CSP reaction represents the parasite’s attempt to escape the immune response of the host, as some have suggested (41), since the parasite loses infectivity as the intensity of the reaction increases. Several studies were performed to determine the correlation between the CSP reaction and protective immunity (28,42), because in the avian malaria model there was a n association between agglutination titers of sporozoites and protection. Some of the results of these early studies cannot be interpreted in a straightforward fashion because of the relative insensitivity of the CSP reaction for detecting the presence of serum antibodies. While a good correlation between levels of antibodies and protection was observed in sporozoite-immunized rhesus monkeys (16), in many experiments in the rodent malaria model protection could not be clearly correlated with the intensity of the CSP reaction. An example of the dissociation of the two phenomena occurred during the earlier phases of immunization of normal mice with irradiated sporozoites when protection was achieved in the absence of CSP reaction. Also, the reverse
SPOROZOITE MALARIA VACCINES
289
FIG. 2. Scanning electron micrographs showing the alterationsof sporozoites following incubation with serum from animals vaccinated with X-irradiatedsporozoites. (Top) Sporozoites incubated with normal serum. (Bottom) Sporozoites incubated in immune serum. (Photograph by M . Aikawa.)
was observed, that is, immunization with sonicated sporozoites led to the production of antibodies but not protection (928,42). In favor of a role for antibodies, however, were the findings that in vitro incubation of sporozoiteswith immune serum inhibited their infectivity, and that partial protection could be transferred to naive mice with large volumes of immune serum. Passive transfer of immune serum also led to a rapid clearance of injected sporozoites and to a significant decrease of EEFs in the livers of the recipients (29). The results of the challenge of the human volunteers who had been
2 90
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
vaccinated with irradiated sporozoites also led Clyde (17) to propose that protection was antibody mediated. Not only were there higher levels of antisporozoite antibodies in all successfully immunized volunteers, but the stage and species specificities of protection and of the serum antibodies were closely correlated. The results of early epidemiological studies also suggested that antibodies to sporozoites played a role in naturally acquired immunity to malaria (see Section VIII). In The Gambia, West Africa, the sera of villagers living in endemic areas contained antibodies to the surface membrane of sporozoites of I? fakiparum. The level of antibodies increased with age, in parallel with acquired immunity, and in several adults high levels of antibodies were detected in the absence of parasitemia and of a serologically detectable immune response to the blood stage of the parasite (33). Furthermore, antibodies to sporozoites were transferred congenitally to infants (43). The prevalence of malaria is very low in children below 1 yr of age and it is likely that the maternally transferred antibodies to both the sporozoite and the blood-stage parasite help to prevent malaria infection. Additional experiments were designed to investigate the role of T cells in protection. Thymectomized, bone marrow-recontitutedmice and nude mice which had been immunized with irradiated sporozoites failed to develop either protective immunity or the CSP reaction. Both were fully restored when thymocytes were injected into the mice prior to immunization (44,45). The only firm conclusion of these experiments was that T cells helped in the primary response for the production of the CSP antibodies. Left unresolved, however, was the question of the participation of T cells in the effector arm of the immune response. This question was approached by immunizing with irradiated sporozoites mice that had been depleted of B cells by injecting them at birth with antibodies to the p chain of immunoglobulins. When these mice were challenged with I? berghei sporozoites, they were partially protected from malaria infection. These findings showed that a weaker degree of protection could be achieved in the absence of serum antibodies. However, the mechanisms remained obscure, because the idea that sporozoites could be attacked by specifically sensitized T cells was very difficult to reconcile with the observation that the parasites remain free in the circulation for only a few minutes before entering the liver cells (29,46). This paradox was later resolved when it was found that the targets of the T cells were not sporozoites, but were the next stage of development of the sporozoite inside hepatocytes, that is, the EEFs (see Section V).
SPOROZOITE MALARIA VACCINES
291
2. Identqication of the Protective Antigen Several observations in the rodent, monkey, and human malaria models indicated that the protective immunity mediated by immunization with irradiated sporozoites was, at least in part, antibody mediated and that the protective antigen was found on the surface of sporozoites. That this hypothesis was correct was first shown in the rodent malaria system. A monoclonal antibody (SDll), raised against the surface of €? bmghei sporozoites, neutralized their infectivity in vitro and in vivo (40,47). Passive transfer of as little as 10,ug of 3Dll protected most mice against sporozoite challenge (40). In order to quantitate sporozoite neutralization by the 3Dll antibody, a DNA probe was used to compare the number of EEFs which developed in the livers of rats challenged with SD11-treated versus nontreated sporozoites. A substantial degree of neutralization was observed with antibody concentrations of less than 1 pg/ml(48). More important, sporozoites could be neutralized by Fab fragments (Fig. 3). This implied that secondary events triggered by antigen-antibody reactions, such as agglutination or complement fixation, were not required to inhibit parasite infectivity, and suggested that the surface antigen recognized by 3Dll (CS protein) was involved in sporozoite penetration. In fact, it was later found that sporozoites do not enter target cells in vitro when Fab fragments of antibodies to the CS protein are added to the incubation medium (49-51). The inhibitory effect of
Fob Frogments
9:
m
10 0-
M of Antibody
X10-'
FIG.3. Inhibition of I? bergheisporozoite infectivity by purified IgG and Fab of 3Dll monoclonal antibody. Norway Brown rats were injected with 1.5 X 105 sporozoites and preincubated with different concentrations of antibodies or of Fab fragments. Shown are the amounts of parasite DNA in the livers of the rats sacrificed 44 h after injection of the sporozoites. Each value represents the mean and standard deviations obtained from four independent measurements in four livers. (Reproduced from Ref. 48.)
292
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
antibodies to the CS protein on the infectivity of rodent malaria sporozoites was confirmed in more recent studies (52), and was also demonstrated in the monkey and human malaria models (53,54). Incubation of sporozoites of 19 falciparum and l? mvax with Fab fragments of monoclonal antibodies to the corresponding CS proteins decreased their infectivity for chimpanzees. The involvement of the CS protein in inducing immunity was also suggested by the observation that irradiated oocyst sporozoites, found in the stomach of the mosquitoes, afforded minimal protection (55). The immature oocyst sporozoites contain much smaller amounts of CS protein (56), but must share many other antigens with the mature salivary gland sporozoites. 111. Properties of
CS Proteins
These early findings clearly implicated antibodies to the CS protein in protection and suggested a yet undefined role for cell-mediated immunity. Further studies revealed many structural, biosynthetic, and immunological similarities among the CS proteins of various species of malaria sporozoites and indicated that they belonged to a family of proteins (57). CS proteins were strictly stage specific, and were found mainly in mature salivary gland sporozoites, uniformly distributed on their surface membrane (Fig. 4). A large proportion of the protein synthesized by the mature salivary gland sporozoites consisted of CS protein (58), a finding that argued that the CS protein was necessary for the develoment of the parasite in the mammalian host. All CS proteins had an M,between 40,000 and 60,000 by SDS-PAGE under reducing or nonreducing conditions, but as was shown later, when the genes were cloned, their molecular weights were considerably lower. The anomalous migration on SDS-PAGE is probably due to their unusual structure and to the fact that they contain a high proportion of polar amino acids. Two intracellular precursors of the CS protein were detected. They had a somewhat higher M , and higher isoelectric points (58). Neither the N-terminal sequences of the mature CS nor the biosynthetic pathways involved in the processing of the precursors have been determined. CS proteins were shown to have immunodominant repetitive epitopes (59). A large majority of monoclonal and polyclonal antibodies obtained in many laboratories against sporozoites, as well as the antisporozoite antibodies found in sera from endemic areas or humans vaccinated with irradiated sporozoites, were directed against these repetitive epitopes (60). Although some cross-reactivities between a few monoclonal antibodies to different CS proteins were observed, most immune sera were strictly
SPOROZOITE MALARIA VACCINES
293
FIG.4. Electron micrograph showing the distribution of the CS protein on the surface of a I! brasilianum sporozoite. The parasites were treated in succession with a monoclonal antibody to the CS protein and with 15 nm gold particles bearing rabbit antibodies to mouse Ig. Magnification, X 39,000. (Courtesy of Dr. A. Cochrane and Dr. M. Aikawa.)
species specific (7,37,61,62) and were directed against the repeat domain. The species specificity of the monoclonal antibodies was exploited to develop assays to detect and identify sporozoites in extracts of whole mosquitoes captured in endemic areas (63). One striking exception to this species specificity was the finding that the monoclonal antibodies to the CS protein of E! malanue (a human malaria parasite) not only reacted with the CS protein of Plasmodium brasilianum (a parasite of monkeys), but also neutralized their infectivity (64,65). Because these two parasites are remarkably similar morphologically and share several
2 94
VICTOR NUSSENZWEIG AND RUTH
S. NUSSENZWEIG
biological properties, it has been proposed that I? brasilianum is a New World derivative of I? malunae and that monkeys can serve as a reservoir for infection of humans (66). More recently, the CS protein sequences of both parasites were elucidated by gene cloning and were found to be nearly identical (67,68). A. CLONING THE cs GENEAND DETERMINING THE PROTEIN STRUCTURE Following the identification of the CS protein as an attractive target for the development of a vaccine, a practical impediment was that its only source was the mature sporozoite. This difficulty was overcome when the CS genes were cloned and the amino acid sequences were deduced from the nucleotide sequences. The first CS gene to be cloned was that of €? knowlesi (69,70), and this was quickly followed by the cloning of CS genes from human (68,71-74), monkey (75-77), and rodent (78-81) malaria parasites. The CS gene was found to be present as a single copy in the genome and it contained no introns. In vitro translation and immunoprecipitation from CS mRNA yielded polypeptides of slightly smaller M, values than those of the larger intracellular precursors synthesized by the parasite (69,79). The nature of the post translational modification which accounts for this difference is not known. There are no sites for N-linked glycosylation or evidence for the presence of O-linked oligosaccharides. However, biosynthetic studies indicate that [3H]-labeled fatty acids are incorporated into the CS polypeptides when sporozoites are incubated with [sH] palmitate (N. Andrews, unpublished observations). Palmitylation of transmembrane receptors almost invariably occurs on a cysteine residue near the transmembrane/cytoplasmic interface, with basic residues nearby (82). A pair of very conserved cysteine residues is found close to the C-terminal end of all CS molecules, and they could be acylated (Fig. 5). Another possibility is that CS proteins are anchored to the outer leaflet of the membrane by means of glycosylphosphatidylinositol (GPI). Several parasite proteins have been shown to be attached to the surface membrane by GPI (83), and the amino acid sequences of the predicted C-terminal portions of CS proteins resemble those of other GPI-anchored proteins; that is, they all contain about 15-20 hydrophobic residues and, in contrast to most other integral membrane proteins, there is no typical hydrophilic cytoplasmic domain (Fig. 5). Perhaps the translated hydrophobic C-terminal amino acids are removed from the nascent polypeptide chain and are substituted by GPI. Anchorage of the CS protein by a fatty acid would facilitate lateral movement and the shedding observed in the CSP reaction. This class of anchor would provide a structural basis for the intriguing
SPOROZOITE MALARIA VACCINES Parasiter
dlrufflb bond?
295
anchor sequence
P cynomolgi I? v/vox
00
FIG. 5. C-terminal sequence of CS proteins. Notable are the conserved pair of cysteine residues and the absence of a predicted hydrophilic cytoplasmic domain. The conserved residues are boxed.
observations that the shed material in the CSP reaction is not amorphous: it resembles a sheath and it resists the shearing forces which develop during centrifugation. Perhaps the shed material consists of a bilayer of CS molecules formed by the juxtaposition of the terminal fatty acids. The gene sequences of the various CS proteins code for 300-400 amino acids. On the basis of sequence, charge densities, and secondary structure predictions, the CS polypeptides can be divided into several domains (Fig. 6). The central area consists of tandem repeats of amino acids and is rich in amino acids such as asparagine and proline, which are commonly found in reverse turns. Conformation energy minimization analysis, used to predict the conformation of an oligopeptide representing the repeats of the CS protein of I? fakiparum [(NANP)g], led to the conclusion that it has only two stable helical conformations, one of them likely to be adopted in polar and the other in nonpolar media (84). Others have predicted a similar helical structure of 12 residues per turn, but with a narrower pitch and stabilizing hydrogen bonds (85). Two regions flanking the repeat domain have charged residues and may contain a-helical structures.
FIG.6. Schematic representation of the structure of the CS protein of P falciparum. The dots indicate the position of NVDP repeats. (Adapted from Ref. 72.)
296
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
Immediately before the 5 ’ end of the repeat domain there is a conserved sequence of amino acids, LKQP (region 1) (72,86). The presence of several synonymoussubstitutions in the corresponding codons in various CS genes supports the idea that region I is important for parasite survival. The suggestion has been made that region I is involved in parasite infectivity on the basis of the observation that synthetic peptides representing region I bind to hepatocytes in vitro (87). In addition, antibodies to a larger peptide (Nl) that included the region I sequence (86,88) inhibited sporozoite invasion of a hepatoma cell line. However, because the repeat region and region I are immediately adjacent, it is possible that antibodies against one region have steric effects on the other. The limits of the repeats at the 3’ end are less precise, and the repeat units appear to degenerate. As mentioned, closer to the C-terminal there are conserved pairs of cysteines. They do not form intermolecular disulfide bonds, but might form intramolecular disulfide bonds. The amino acid sequences between both pairs of cysteines and surrounding the first pair of cysteines (region 11) (72,86) are well conserved between CSs of different species. Sequences similar to region I1 have been found in a blood-stage protein of I! falciparum and in repeated motifs, each of 60 amino acids, found in thrombospondin and in the complement protein properdin (89,90). These intriguing findings could be of functional significance. Thrombospondin is thought to mediate cell-cell and cell-matrix interactions and the common motif could have a similar function in the parasite proteins. The homology to properdin is more enigmatic. Properdin has specific binding affinity for complement fragment C3b and stabilizes the enzyme CSbBb, the C3 convertase of the alternative pathway of complement activation. There is no evidence that the complement system and, in particular, C3 fragments are necessary or that they enhance invasion of the host cells by malaria parasites in vitro or in uivo.
B. EVOLUTION OF REPEAT^ Because of the importance of the repeats in the formulation of the malaria vaccine, it is important to understand their genetic variability and evolution. As shown in Table I, the amino acid sequences of the repeats differ markedly among species of Plasmodium. However, they share a limited repertoire of amino acids: Pro, Asn, Gln, Gly, Asp, Glu, and, more rarely, Arg and Val. This must reflect the influence of strong selective pressures to maintain some shared functional or structural feature of this domain.
SPOROZOITE MALARIA VACCINES
297
TABLE I AMINOACIDSEQUENCES OF SUBUNIT REPEATS OF Plasmodium SPECIES Species (strain)
Sequence4
I? falciparum
NXEP
I? vivax
~GQPAGDRA
I? malartae
NXAG
I? bmsilianum
N~AG
I? cynomolgi (NIH)
NAEG
I? knowlesi (H)
AGQPQAQGDGAN
I? yoelii
QGPGAP(QQPP)
I? berghei
P$PPN$ND(PQ)
"Amino acids used: P, N , 4, A, G, D, (E), (V), and (R); minor tandem repeat sequences found in the N-terminal region of the repeat domain are shown in parentheses.
Evidence for the importance of preserving repeat sequences within a species of Plasmodium stems from the analysis of the I? falciparum repeats within and between different isolates of the parasite. It is remarkable that in all isolates from widely different geographical regions (72,91-93), most of the CS protein repeats are NANP, with a few NVDP repeats interspersed. In addition, within each CS sequence, the DNA coding for the NANP repeats displays several silent nucleotide substitutions. Perhaps the most striking example of conservation of the repeats is the finding that they are identical (NAAG, and a few NDAG) in the CS proteins of I? brasilianum and I? malariae (67,68),which are closely related parasites of monkeys and humans, respectively. In this instance, even change of host did not lead to a change in the sequence of the subunit repeats. An important factor to be taken into consideration when studying the evolution of the CS protein is that the analysis of the DNA composition and genome arrangement of malaria parasites permitted their separation into two groups. One group includes the human parasite I? fakiparum and the rodent malarias, and the other includes the human parasite I? vivax and monkey parasites. This is in agreement with the
298
VICTOR NUSSENZWEIG AND RUTH S . NUSSENZWEIG
finding that there is a close homology between the CS gene of €? m'mx and those of the simian malarias and much less homology between the CS genes of the two human malarias l? vivax and I?fulciparum (94). In fact, a short insertion, found recently in the repeats of a North Korean strain of €? uivax, is part of the repeats of some strains of €? cynomolgi, a malaria parasite of Asian monkeys. Perhaps nonreciprocal genetic exchanges occurred between the tandem repeats of these genes, and I! vivax is evolutionarily related to I? cynomolgi. Analysis of the repeats of the CS genes of l? knowlesi (77) and €? cynomolgicomplex (Table 11)showed that this domain evolved more rapidly than the other domains (75). The CS genes of two strains of I! knowlesi, a simian malaria parasite, were virtually identical except for the repeats. The sequences flanking the repeats in six strains (or subspecies) of €? cynomolgz'were 95% conserved, while the repeats were markedly diverse. More recently, sequences from repeat regions of CS genes of two €? Vivax isolates from Latin America and North Korea were compared (95,96). They were found to be very similar, even in the patterns of silent mutations. However, although the number and arrangement of repeat units were similar, the patterns of substitution at the gene level appeared to shift. These shifted patterns might result from gene conversion or from homologous but unequal crossing-over. Such processes can lead to the
TABLE I1 AMINOACIDSEQUENCES OF SUBUNIT REPEATS OF STRAINS OR SUBSPECIES OF Plasmodium cynomolgia Subspecies (?) or strain (?)
Sequence
Mulligan/NIH
NA~G
Gombak
&A A AAGGGGN
London
XDGARA
GN&GQAGAG
Ceylon
AGNNA A ~ G ~ AGNNAA AGEAGAGGAGR
Berok
PDGDGAPA~
NRAGGQPA AGGNQAG~ 'The different parasites have been isolated from various monkey species in parts of Asia. (Adapted from Ref. 75.)
SPOROZOITE MALARIA VACCINES
299
rapid emergence of new repeat domains or the replacement of an existing domain with another one. The finding that some repeat domains contain two separate repeat sequences supports this idea. An unusual feature of the repeat domain of the CS genes is the nonrandom use of synonymous codons (95-97). The remarkable codon bias which has been noted may result from the spreading of local changes from one repeat unit to the neighboring ones (75,98,99). It should also be pointed out that the presence of polymorphic repeats is a common feature in most of the other cloned antigens from malaria parasites (100). Whatever the genetic mechanisms involved, variation in the repeats is a potential problem for the immunoprophylaxis of malaria. If, under selective pressure of the immune system, populations of sporozoitescan rapidly change the CS protein repeats, the effectiveness of repeat-containing vaccines will be limited. The available evidence, however, does not justify a pessimistic outlook. In endemic areas there are relatively high levels of antibodies to this epitope in the sera of individuals, who after prolonged exposure have developed resistance to malaria infection (33,101-104). However, although marked differences have been noted in the number of subunit repeats in CS proteins of €?falciparum and €? Vivax isolated from several areas, the sequences of the subunits are the same (91,92,95,105,106). Single monoclonal antibodies recognize CS proteins from all isolates of I? fulciparum and I! n ' w obtained from Latin America, Asia, and Africa (107). Perhaps the genetic changes leading to the entire substitution of the repeats of the human malaria CS proteins may compromise the viability or infectivity of sporozoites. Alternatively, the human parasites may have arisen relatively recently from another species of Plasmodium. C. FUNCTION OF REPEATS The function of the repeats and the nature of the constraints limiting their variation within a CS protein have been the subject of much speculation. An attractive possibility is that the repeats from neighboring CS molecules may interlock to form a sheath surrounding the parasite (70). The amino acid sequence of a subunit could then vary between different CS proteins, provided that the molecules could still form the putative sheath. Drastic variations in the structure of the subunit repeat within a CS molecule might interfere with the assembly of this quaternary structure and diminish parasite viability. The repeats could also serve as ligands for the postulated sporozoite receptor on hepatocytes. The regular arrangement of the repeats, and their high local concentration on the parasite surface membrane, would enhance the probability of encounter with multiple low-affinity
300
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
hepatocyte receptors. Internal duplications in a gene encoding a ligand for a host cell receptor may be advantageous for parasite survival. The redundant expression of the ligand would enhance the probability of cooperativity in the interaction with a mobile or multimeric receptor. It would also render a key stretch of the CS polypeptide impervious to single amino acid substitutions or to small deletions and insertions, since mutations in one subunit would be of little consequence unless the others were equally affected (108). One apparent contradiction to the idea that the CS protein repeats bind to liver cell receptors is the observation that sporozoites from different species (or strains) of malaria parasites that have distinct CS protein repeats can invade hepatocytes of the same host species. For example, strains (or subspecies?)of I? cynomolgi have CS proteins with quite different subunit repeats, but experimentally they all infect rhesus monkeys. It is possible, however, that different hepatocyte receptors are recognized by the various sporozoites. It is becoming apparent that the blood stages of human malaria parasites enter red cells via different receptors. Alternatively, the repeats with different amino acid sequences can fold to form similar three-dimensional ligands. In support of the latter hypothesis, unexpected cross-reactions have been observed between monoclonal antibodies against CS protein repeats that appear distantly related or unrelated judging from the primary amino acid sequences (109). Because antibodies (and perhaps liver receptors) are multivalent, cooperativity will greatly increase the energy of interaction and lead to enhanced binding. IV. Role of Antibodies in Protection
The question of whether a vaccine that only generates antibodies against the repetitive domain of the CS protein can mediate effective protection against malaria infection was first approached experimentally when the CS g n e s of the rodent malaria parasites €? berghei and l? yoelii were cloned, and the corresponding amino acid sequences of the CS proteins were elucidated. The answer to the question was predictable on the basis of prior findings; that is, the effectiveness of such a vaccine increased as the level (and probably the binding affinity) of antibodies increased, and it became less effective when the dose of sporozoites used for challenge increased. In the li berghei rodent model of malaria, one particular CS-derived immunogen was effective in the production of high levels of antibodies: it consisted of a synthetic peptide representing a variant subunit of the repeat domain of the CS protein (PPPPNPND)z, conjugated to tetanus
SPOROZOITE MALARIA VACCINES
301
toxoid by means of bis(diaz0benzidine). The effects of this vaccine on sporozoite infectivity were evaluated using a specific DNA probe to measure the number of EEFs in the livers of vaccinated rats following challenge with sporozoites. The titer of antisporozoite antibodies in the serum of the rats was about 1:10,000. Following challenge with sporozoites, their livers contained 97% less parasite DNA than those of the controls vaccinated with tetanus toxoid alone. The vaccine also afforded protection to 7 5 4 7 % of mice against challenge with 1000 sporozoites. The animals had serum titers of antisporozoite antibodies between 16,000 and 64,000 at the time of challenge (110). No evidence for cell-mediated immunity was obtained in cell proliferation assays in which lymph node cells were incubated with the synthetic peptide. When the challenge dose was 5000 sporozoites, however, or when the antibody titers were below 1:10,000, the mice were only partially protected. This is in contrast to the solid protection against challenge with 20,000 or even higher doses of parasites obtained following immunization with irradiated sporozoites. Using the same experimental model but a different vaccine formulation, others reported a lower degree of protection against challenge with only 500 sporozoites (52). The lower efficacy may have been due to the use of a different variant subunit repeat in the vaccine formulation or to the method used to prepare the conjugate. Another group of investigators immunized mice with a repeat peptide of the I? yoelii CS protein conjugated to keyhole limpet hemocyanin, and obtained little protection against challenge with 500 sporozoites (111). It was not clear, however, in that report whether most of the antipeptide antibodies detected in the serum of the vaccinated mice reacted with the sporozoites of I! yoelii. V. Interferon-y Affects the Liver Stages
After a few minutes in circulation, sporozoites invade hepatocytes and each sporozoite generates progeny consisting of several thousand merozoites inside a parasitophorous vacuole. The EEFs are therefore protected from attack by antibodies. It is also difficult to envision a direct effect of cytotoxic T cells on the infected hepatocytes. Blood flows in the sinusoids lined with endothelium and Kupffer’s cells. According to the current concepts of liver structure, the sinusoid wall contains gaps and small fenestrae, which permit an exchange between the fluid phase of the blood and the parenchymal cells and are sufficiently large for the passage of sporozoites. However, they do not permit the exit of blood elements. In fact, red blood cells and leukocytes are not normally found
302
VICXOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
in the perisinusoidal space of Disse and it seems unlikely that few sensitized T cells would encounter the very rare infected hepatocytes even if they expressed parasite-processedantigens on their surface membrane. Products of the immune system, however, can inhibit the EEFs. It has been known for some time that interferon-a and interferon-P inducers, such as Newcastle disease viruses and double-stranded RNA, affect the EEFs (112,113). More recently, it was shown that interferon-?, a product of T cells, has a profound effect on the development of EEFs (114,115). In one study a DNA probe was used to measure the number of EEFs in the liver. It was found that interferon-? at very small doses prevented the development of EEFs in mice and rats infected with sporozoites of I? berghei. A significant reduction in the EEFs in rat liver was achieved when 15 ng of rat interferon-? was injected a few hours prior to sporozoite challenge (Fig. 7). Human interferon-? also diminished the parasitemia in chimpanzees infected with sporozoites of the human malaria parasite €? vivux (114) and prevented the malaria infection of rhesus monkeys
bc
W LL W
B 40-
20-
L
0
a0015 Units IFN-8 injected ( x
FIG. 7. Inhibition of the development of I! berghei EEFs by recombinant rat interferon-y(adapted from Ref. 114). Groups of rats were injected with the lymphokine 5 h before intravenous injection of 105 I? berghei sporozoites. After 44 h the livers were removed and the DNA was purified and probed for the presence of I? berghei DNA. Ordinate values represent the percentage inhibition of EEF development as compared to control groups.
SPOROZOITE MALARIA VACCINES
303
injected with the simian malaria parasite I! cynomolgi (115). The lymphokine had no effect on sporozoites or on the blood stages. The target of interferon-y activity seemed to be the infected hepatocytes. The degree of inhibitory activity was independent of the number of sporozoites used for challenge. Furthermore, the lymphokine was fully active in vitro and inhibited EEFs from developing in hepatoma cells (116). How the lymphokine inhibits EEF development is not known. The alterations that interferon-y induces in the metabolism of tryptophan, which inhibit the intracellular development of Toxoplusma gondii (117), do not seem to be involved in killing the EEFs (116). Although very small doses of interferon-? were inhibitory, some EEFs escaped the administration in vivo of very high doses of the lymphokine and developed normally (118). The reasons for this are not clear. We favor the idea that interferon-? binds to the receptors on the liver cells and renders them refractory to parasite development. Perhaps some sporozoites escape because they enter liver cells in a random fashion, and some enter cells with no or few interferon-y receptors. Hepatocytes within an acinus are functionally quite heterogeneous (119) and it is therefore conceivable that numbers of interferon-? receptors vary among liver cells. The importance of these findings is that they raised the possibility that interferon-y might play a role in the protective immunity observed following immunization with irradiated sporozoites and in natural immunity to malaria infection. In fact, spleen cells of mice immunized with irradiated sporozoites released high levels of interferon-y when challenged in mlro with antigen (120), and interferons have been detected in the serum of malaria patients (121,122). Perhaps in endemic areas lymphokines from the sensitized T cells of immune individuals help to prevent or limit superinfection. AND INTERFERON-y IN SPOROZOITE IMMUNITY ROLEOF CD8+ T CELLS In the past, several investigators presented suggestive evidence that T cell-mediated effector mechanisms are involved in the solid protection achieved by vaccination of mice with irradiated sporozoites. Only recently, however, could this question be approached more directly. In one study, the administration of neutralizing antibodies against interferon-y to immune hosts reversed sterile immunity to sporozoite challenge by allowing the growth of EEFs and the development of parasitemia (123) (Fig. 8). In this and in another study (124), it was found that the immune animals also developed infections when depleted in vivo of their CD8+ cytotoxic cells. Depletion of CD4+ T cells had no detectable effect on EEF development. The reasons for this are not clear. Evidence
304
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
FIG.8. Endogenous interferon-y inhibits development of EEFs in immune hosts. Rats immunized with four doses of 105 X-irradiated sporozoites of I! berghei were challenged with various doses of sporozoites. Immediately, or 2-1/2 h afterward, animals received an intravenous injection of a monoclonal antibody, DB-1, which inhibits interferon-y, or received control antibodies. EEF-DNA levels were measured 44 h later. Data are mean values from five animals f SD; clear areas, naive animals plus antibody DB-1; crosshatched areas, naive animals plus control antibody; diagonally striped areas, immunized animals plus antibody DB-1; solid areas, immunized animals plus control antibody.
is accumulating that the mouse CD4+ cells can be divided into subsets, Thl and Th2, on the basis of the ability to produce lymphokines. The Thl subset secretes preferentially interferon-y, while the Th2 subset secretes the factors involved in B cell differentiation and provides help for antibody synthesis. Perhaps in the sporozoite-immunizedmice, mainly the Th2 subset and the CD8+ T cells are activated. CD8+ cells recognize antigenic peptides associated with the class I MHC. Accumulated experimental evidence has led to the suggestion that there are two pathways for antigen processing, each associated with one class of MHC (125). Exogenous soluble antigens that can be pinocytosed are degraded in the pinocytic vesicles and enter the class I1 MHC pathway.
SPOROZOITE MALARIA VACCINES
305
In contrast, the endogenously synthesized antigens enter only the class I MHC pathway. If these ideas are correct, how could parasite antigens have entered the endogenous pathway? One group of investigators has proposed that the antigens do not derive from sporozoites but from the EEF, and that parasite materials leak into the cytoplasm of the hepatocytes and enter the endogenous pathway of processing. They also proposed that CD8+ T cells might enter directly in contact with the membrane of infected liver cells and recognize EEF-derivedpeptides in association with the class I MHC (26). In addition to the anatomical considerations which would make the encounter between specific lymphocytes and the few (less than 100) infected hepatocytes a very unlikely event, this explanation would be inconsistent with the accumulated evidence that the induction phase of immune responses does not occur in the periphery, but rather centrally, in lymph nodes and spleen. It is also improbable that sporozoite peptides are generated in the infected liver cells and are subsequently leaked into the circulation to be presented elsewhere and preferentially in association with the class I MHC. We favor the hypothesis that exogenous antigens can enter the endogenous pathway provided that they are particulate and are phagocytosed by specialized antigen-presenting cells (125). As mentioned in Section II,A, the majority of the injected irradiated sporozoites do not enter hepatocytes, but are ingested and destroyed by macrophages, mainly in the spleen. In vitto studies showed that, in the absence of specific antibodies, the sporozoites which had been ingested by macrophages remained for at least 1 h inside the phagosome without ultrastructural evidence for degeneration. An active role of sporozoites in their interiorization by macrophages was suggested by experiments showing that cytochalasin B, a powerful inhibitor of phagocytosis, does not inhibit the uptake of the parasites (126). It is conceivable that the sporozoites remain metabolically active for some time in the macrophages, and that either some of them destroy the phagocytic vacuole and reach the cytoplasm, or they remain in the vacuole but some proteins leak into the cytoplasm and enter the endop o u s pathway of antigen processing. A similar suggestion has been made to explain the participation of CD8+, class I-restricted T cells in resistance to murine listeriosis and several other intracellular microbes (126a). A second question regards the nature of the effector mechanisms used by CD8+ cells to destroy the EEFs. One possibility is that, in the immunized mice, CD8 cells recognize parasite antigens on the surface of macrophages and subsequently produce interferon-y , which will then act in a hormonal fashion against EEFs. In support of this interpretation +
306
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
is the observation that the protection afforded by adoptive transfer of CD8+ splenic T cells was abolished by monoclonal antibodies to interferon-y (123). Alternatively, CD8 + cells might be directly cytotoxic for infected hepatocytes which present processed antigens of sporozoite or EEF origin (26). This would require direct contact and, as discussed above, seems implausible for anatomical reasons. Furthermore, many workers have noted a total absence of cellular response to the EEF parasite. Cellular infiltration has been noted only around ruptured mature EEF schizonts (127). Whatever the molecular mechanisms involved, with the proviso that the findings in the rodent malaria model are equally applicable to human malaria, the effectiveness of a sporozoite vaccine will be greatly enhanced if it contains epitopes capable of expanding the population of specific CD8+ (or subsets of CD4+ cells) capable of producing interferon-y. In addition, it is essential that the vaccine-elicited T cells are triggered to produce the lymphokine when the vaccinated individuals are injected with sporozoites. Which sporozoite antigens elicited the protective T cells in the I! berghei rodent malaria model? Although it is not possible to exclude a role for internal sporozoite antigens, some observations suggest the participation of epitopes from the CS protein in the T cell-mediated immunity. For example, a recombinant Salmonella typhimurium oral live vaccine of cells expressing the CS protein of I? berghei protected mice against malaria infection in the absence of serum antibodies (128). The recombinant live bacteria invade and survive for some time within macrophages and, similar to the living X-irradiated sporozoites, elicit efficiently the T cell protective responses. Another Salmonella transformant which had a C-terminal deletion in the CS gene and did not express the repeats also induced protection. These results indicate that there are protective T cell epitopes in the N-terminal region of the CS protein. Others have immunized mice with vaccinia recombinant virus expressing the CS protein of €! fulcipurum and have located an epitope capable of inducing cytotoxic cells in the C-terminal region of the CS protein (129a).
VI. The Need for CS-Specific T Cells In Sporozolte Vaccines
To be effective in endemic areas, a sporozoite CS protein vaccine should not only induce high levels of antibodies to the repeats, but should also maintain these levels for prolonged periods of time. Because of the difficulties in delivering booster injections, it is desirable that the sporozoites
SPOROZOITE MALARIA VACCINES
307
injected by the mosquitoes serve as a natural booster. This can only occur if the vaccine also contains parasite-derived T helper epitopes. The question arises, however, whether it is essential that the vaccine contain a T epitope from the CS protein or whether the help for producing antibodies to the CS protein could derive from an epitope present on another parasite antigen. In other words, is it possible to bypass the need for T and B cell epitopes to be tightly linked for boosting to occur? The desirable objective is to trigger the expanded population of the B memory cells in the vaccinated individuals, causing them to differentiate and produce antibodies to the CS repeats. The most efficient way for this to occur is if the B memory cells themselves function also as antigen-presenting cells. That is, these B cells should recognize and internalize the CS molecule via their Ig receptors to the repeats and, following “processing,”present parasite-derived peptides associated with the class I1 MHC to T cells. The T cells will, in turn, provide the necessary signals for triggering antibody production in the B cells. In the case of the influenza virus, the T cells sensitized to an internal viral antigen can help very effectively the antibody response to a surface antigen, the hemagglutinin. The probable explanation is that the entire virus particle can be recognized and endocytosed by a hemagglutininspecific B cell. Following processing, T cells recognize peptides originating from internal nonrelated antigens (130) and provide help to the B cell for the production of antihemagglutinin antibodies. Similar findings have been reported for hepatitis B virus (13Oa). Sporozoites, however, are very few in number, and the probability that they will encounter a CS-specific B cell during the few minutes in which they remain in circulation is virtually nil. Even if they did meet, the sporozoites would not enter the B cells. We have observed, however, that during incubation in vitro at 37OC, large amounts of CS protein are shed rapidly by the parasite, and this probably also occurs in vivo. The CS protein released from the parasite following mosquito bite may function as a boosting stimulus to the B memory cells. If this view is correct, T help for antibody production should be mainly provided by lymphocytes recognizing CS protein peptides, and the vaccine should contain the corresponding T helper epitopes. This does not mean, however, that it will be useless to incorporate in the vaccine internal sporozoite antigens, which do not stimulate the production of antibodies. On the contrary, they could have protective value if they expand the population of T cells which produce interferon-y. Clearly, these views are speculative, because we still have little understanding of the cellular and molecular mechanisms
308
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
involved in the processing of antigens, particularly those derived from infectious agents. A. Ir GENECONTROL OF THE ANTIBODY RESPONSE TO CS PROTEIN REPEATS Most adults living in endemic areas have serum antibodies to the repeats of the CS protein of l? fulcipurum-in many instances in relatively high levels, considering the minute doses of antigen (nanograms) which are contained in the few sporozoites injected by mosquito bite. Even Europeans who presumably had very little exposure to the parasite in endemic areas have detectable levels of antibodies to (NANP), (131). When mice are injected with sporozoites, they produce mainly antibodies to the repeats of the CS protein, and the immune response to the CS is not restricted in inbred strains of mice immunized with sporozoites of l? berghei (Table 111) (132) and l? wvax (133). The considerations in the previous sections strongly suggest that the cognate help for the production of antibodies to the repeats is provided by T cell epitopes from the CS molecule and not from another internal sporozoite antigen. Studies aiming at the elucidation of the structure of the T cell epitopes of the CS protein have been performed in experimental models in mice and in cells from the peripheral blood of humans. One important observation was that the antibody response to the NANP repeats of l? falciparum was genetically restricted and under Ir gene control (134,135). NANP polymers elicited T cell proliferative responses only in mice bearing I-Ab in the H-2 region, and only these mice made antibodies to the peptide. Other repeats were also poorly immunogenic. Among 15 strains of mice with different H-2 regions and different backgrounds, none responded to the PPPPNPND repeats of r! berghei (132), and only mice bearing I-Ak responded to the l? w%ux repeats (133). The implication of these findings was that if the same situation occurred in humans, a vaccine containing only the repeats might be immunogenic in only a small proportion of individuals. This may be the explanation for the weak immunogenicity for humans of a recombinant l? falciparum vaccine consisting mostly of NANP repeats (136). Also, in a few studies, the T cells from some individuals naturally exposed to l? falcipurum malaria infection did not proliferate in the presence of synthetic peptides or recombinant proteins representing the NANP repeats (137,138). However, other regions of the CS protein contain T cell epitopes recognized by several inbred strains of mice. In animals vaccinated with recombinant CS proteins containing the repeat and portions of the nonrepeat domains of l? fakiparum (139) or l? berghei (132), the
309
SPOROZOITE MALARIA VACCINES
TABLE 111 OF 12 STRAINS OF INBRED MICETO IMMUNIZATION ANTIBODY RESPONSE WITH IRRADIATEDE! berghei SPOROZOITES' IRMA titersCwith
Strain BIO.A C57B1/10 DBA/2 BIO.M C3H.JK B1O.BR B1O.AKM C3H.NB DBAA ASW B10.PL NZW
H-2 haplotype
IFA titersb
repeat peptide
recombinant CS protein
a b d f j k k P q
2048 1024 4096 1024 4096 2048 1024 8192 8192 2048 2048 2048
8192 4096 32,768 4096 8192 8192 8192 8192 16,384 4096 8192 8192
8192 4096 32,768 4096 16,384 16,384 8192 16,384 32,768 8192 8192 16,384
S
V
z
'Groups of four mice of each strain received two i.v. injections of lo5 irradiated I! bergheisporozoites at two-week intervals. Serum samples were collected 10 days after the last injection. Indirect immunofluorescence (IFA) titers using glutaraldehyde-fixed l? berghez sporozoites as antigen. Pooled Sera from each group were titrated by means of an immunoradiometnc assay (IRMA), using as antigen either the repeat peptide or a recombinant CS protein containing the repeats and other regions of the CS protein. Titers are expressed as the reciprocal of the highest positive serum dilution. (Adapted from Ref. 132.)
production of antibodies to the repeats was not restricted. While only mice bearing I-Ak responded to a yeast-derived recombinant I! wvux CS protein (133), no additional restriction was observed when the recombinant antigen included an additional portion of the C-terminal end of the CS protein (140). In other studies, several strains of mice responded poorly to immunization with a vaccinia recombinant virus containing the CS gene of I! fulci#urum (141) or I! berghei (142). The antigen dose may have been a contributing factor for the poor immunogenicity. Low levels of CS protein are produced in vaccinia-infected cells and the protein is not secreted (143,144). When the recombinant vaccinia is injected into animals, the recombinant virus may not replicate sufficiently due to a very effective immune response to the vaccinia antigens. Another explanation for the poor antibody response could be that, in the vaccinia
310
VICTOR NUSSENZWEIG AND RUTH S . NUSSENZWEIG
virus-infected cells, the CS protein is processed by the endogenous pathway and is presented in the context of the class I MHC. Perhaps the most important question stemming from these studies is whether the restriction observed in inbred strains of mice foreshadows a poor immune response in an outbred population. That this may not be the case is shown by the observation that the recombinant P vivax vaccine, which was restricted to mice bearing I-Ak, induced the production of high levels of antibodies to sporozoites when injected into outbred mice (145) or Suimzri sciureus monkeys (146).
B. T CELLEPITOPES OF THE CS PROTEIN RECOGNIZED BY HUMANS Several T cell epitopes of the CS protein of €? fulczpurum have been identified by performing proliferation assays in the peripheral blood lymphocytes of humans in The Gambia, West Africa (137). The antigens were synthetic peptides spanning the entire sequence of the CS protein, but only 60% of individuals responded. Since most of the nonresponders had antibodies to the NANP repeats in their serum, the most likely explanation for the lack of proliferative response is that the universe of all possible T cell epitopes of the CS protein was not represented among the tested peptides. Furthermore, testing was performed in a single sample of blood, and this may be a significant problem, particularly when low numbers of specific T cells are present. More recently, the same group of investigators reported that two-thirds of individuals from an endemic area responded to at least one of the three variant peptides representing two of the immunodominant domains of the CS protein (147). In another study involving individuals living in a malaria endemic area in the Ivory Coast, two methods were used to predict the location of T cell epitopes in the CS protein of €? fulcipurmm (148,149); the corresponding synthetic peptides were used to perform proliferation assays (150). Although only three peptides were tested, the peripheral blood lymphocytes from all individuals responded to at least one of them. The majority of individuals (9 out of 16) responded to one of the peptides, located in the C-terminal portion of the CS (residues 378-398, DIEKKIAKMEKASSVFNVVNS, except that the two cysteines at positions 384 and 390 were substituted for alanine). The observation that the peptide corresponding to residues 378-398 was recognized by cells from most individuals suggests that it can associate with many different MHC class I1 molecules. More recently, mononuclear cells from the peripheral blood of 20 nonimmune donors, with no antibodies to sporozoites, were challenged in Vitro with the peptide. From the stimulated cells close to 300 peptide-specific clones were derived from eight donors. Experiments with truncated peptides
SPOROZOITE MALARIA VACCINES
311
located the epitopes in the segment containing amino acids 380-395 (EKKIAKMEKASSVFNV). Other studies showed that the peptide was recognized by most mouse and human MHC class I1 molecules and that it can function as a helper epitope. Many strains of mice which were genetically unresponsive to the NANP repeats produced antibodies to the repeats if the antigen consisted of (NANP)S coupled to the 378-398 peptide. The 378-398 peptide was also recognized by at least seven DR alleles, found in 80-90% of African or Caucasian individuals (151,152). A I! fakiparum sporozoite recombinant or synthetic vaccine should elicit T helper cells in the majority of the population from endemic areas and serve as a carrier for the (NANP)S B cell epitope. On the basis of the above-mentionedfindings, the peptide 378-398 satisfies both of these requirements. Virtually nothing is known about the physiological pathways for “processing” a native protein antigen, or about the specificity of the proteolytic enzymes which generate the fragments destined to bind to the MHC class I1 molecules. Therefore, the synthetic peptides, which by trial and error are found to stimulate T cells from immune individuals in Vitro, may not be generated in vivo. In this case the T cells elicited by a vaccine containing such “minimal” peptides might recognize poorly or not at all the physiological products, and might not be boosted by them. However, the T cell clones recognizing 378-398 in association with several DR antigens also recognized the processed, parasite-derived CS protein, as well as the cysteine-substituted 378-398 peptide. In theory, therefore, a I? falci@rum vaccine containing the 379-398 epitope linked covalently to a NANP polymer should be able to induce parasite-specific humoral and cellular immunity in the genetically diverse human population and, in addition, this population’s T cells should be boosted by exposure to sporozoites. C. POLYMORPHISM OF T CELLEPITOPES One possible difficulty for the development of a I? fakiparum malaria vaccine based on the CS protein is the variability of candidate T cell epitopes in different strains of the parasite (26,95,137,153). Inspection of the sequences of the CS proteins of three isolates of I? fakiparum revealed the presence of variation in only nine amino acids, all of them in the regions flanking the repeats. Because there are no silent mutations at the nucleotide level in these regions (there are several in the repeat domain), it has been suggested that selective pressure at the protein level dictates the variation (95). The variation was seen in the C-terminal region containing human T cell epitopes (137), which led to the hypothesis that
312
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
the pressure for variation was from immune T cells (95). Some support for this idea comes from the observation that a human T cell clone recognized only one of the variants at P0sit;on 339 of the CS protein (151). On the other hand, it could be argued that the observed polymorphisms in the C-terminal region reflect the absence of constraints for amino acid variation. The remarkable low frequency of silent substitutions at the nucleotide level is most likely an unrelated phenomenon, which has been noted in other E! fulciparum genes and may be a feature associated with plasmodia1 DNA (100). Because variations in T helper epitopes of the CS would not save sporozoites from destruction by antibodies directed to the NANP repeats, it was further suggested that the variant, C-terminal sequences could function as T cytotoxic epitopes. The rationale was that hepatocytes expressing on their surface a CS variant epitope would escape destruction by the CD8+ cells, and the intracellular parasite would develop normally (26). There are several difficulties with this scenario, which assumes the existence of selectively maintained polymorphisms in the C-terminal region of the CS protein. The hypothesis would require that infected hepatocytes express on their surface only the polymorphic T cell epitopes, since if others were equally expressed, recognition by CD8 + cells could still take place. This proposal also hinges on the idea that CD8+ cells destroy the liver stages directly, and, as mentioned above, anatomical and statistical considerations render this unlikely. Most important, however, is the experimental evidence that interferon-y inhibits EEF development, and that the protective activity of the CD8+ lymphocytes in mice vaccinated with irradiated sporozoites is mediated by the hormonal action of interferon-y. This was demonstrated in adoptive transfer experiments in which naive mice received splenic lymphocytes from naive or immune donors and were challenged with viable sporozoites. Those mice receiving immune lymphocytes were partially protected against infection. The transferred immunity was, however, reversed by the administration to the recipients of a monoclonal antibody which neutralized interferon-y activity. These results strongly suggest that, even if a mosquito injects into the host a population of sporozoites containing a rare mutant that expresses a variant T cell epitope, the mutant would have no selective advantage over the other sporozoites. This is because although the production of interferon-y by T cells is triggered specifically, the lymphokine acts nonspecifically on any EEF-infected liver cells bearing interferon-y receptors. That is, the interferon-y produced by T cells against one form of the CS protein should act against any EEF, variant or nonvariant, provided that the mosquitoes inject mixed populations of parasites. In short, while there has been a lot of speculation about the possible
SPOROZOITE MALARIA VACCINES
313
problems for malaria vaccine development which might result from the observed variations in the amino acid sequence of the C-terminal end of the CS protein, direct experimental evidence in support of these ideas is nonexistent. In fact, the results referred to earlier of immunization of the human volunteers with irradiated sporozoites would support the opposite conclusion, since they showed that the protection mediated by vaccination was not strain specific. It will be necessary to perform experiments with cloned parasites bearing different CS alleles to answer this question in a definitive fashion. It should be pointed out that variation in the CS protein from different isolates or subspecies of the monkey malaria parasites R cynomolgi and R Knourlesi occurs predominantly or exclusively in the repeat domain containing the immunodominant B epitopes, rather than in the flanking regions containing the various T epitopes. This is also true in the few isolates of the human malaria parasite I! vivax which have been examined. It seems more reasonable to argue that the immune pressure for selection derives from antibodies and not from T cells, similar to what has been observed in other infectious diseases, and that, at least in part, this is the explanation for the more rapid evolution of the repeat domain of the CS protein. VII. Human Trials of Sporozoite Vacclnes
Two sporozoite vaccines have been given to volunteers, both aiming only at raising antibodies to the repeat domain of the E? falciparum CS protein (141,154). The rationale for trying such a vaccine was that if the levels of the induced antibodies were sufficiently high, they would block invasion of hepatocytes and prevent disease. Other attractive considerations were that the NANP repeats were conserved among different geographical isolates of I? falciparum, and that the sporozoites had a very large number of NANP-containingepitopes on their surface, which made them excellent targets for the neutralizing activity of antibodies. One of the tested vaccines was a synthetic vaccine consisting of cysteine-(NANP)sconjugated to tetanus toxoid (155). Three injections of different doses of this vaccine and a control vaccine containing only tetanus toxoid were given to groups of volunteers. The adjuvant was aluminum hydroxide. The frequency and magnitude of the antibody response correlated with the dose of vaccine administered, but there was no clear boost with succeeding injections. Of the volunteers who received the highest dose of vaccine (160 pg). 70% showed significant rises in serum antibodies to NANP repeats, but the titers were in general lower than 1/500. There was an excellent correlation between the titers of
314
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
antipeptide and antisporozoite antibodies (TS = 0.858, P < 0.001) in the sera of the vaccinees. Three individuals with the highest titers of serum antibodies and four controls were challenged by the bite of five infected mosquitoes. The four controls developed parasitemia 7, 8, 9, and 10 days following challenge (mean, 8.5 days). The minimum time for EEF development with €? fakiparum is 7 days. One peptide-vaccinated individual did not develop a blood infection, and the other two had detectable parasitemias only 11 days after challenge. The significantly increased prepatent period in the two vaccinees most likely is due to the neutralization of a large proportion of the invading sporozoites. Studies in sporozoite-inducedmalaria in man and in animal models have shown an inverse relationship between the size of the sporozoite inoculum and the duration of the prepatent period, and the dose-response curve is extremely flat. For example, a 100-folddecrease in the dose of I? knowlesi sporozoites injected into rhesus monkeys led to an increase of less than 1 day in the prepatent period (156). An important factor to be taken into consideration when interpreting these results is that the efficacy of the sporozoite vaccine depends on the size of the inoculum, and that five infective bites are a high challenge dose. Except in highly endemic areas, the proportion of infected mosquitoes is less than 1/100 and in many areas it is less than 1/1000. Although in endemic areas individuals can be subjected to many infective bites in a single day, it seems very unlikely that the bites will be simultaneous. In other words, a malaria vaccine has to protect individuals in most instances only against single infective bites even in areas of high endemicity. One disappointing finding was that the serum titers of antibodies in these and in other volunteers injected with (NANP)3-tetanus toxoid (154) were much lower than those of mice or rabbits immunized with the same antigen (60,157). This may in part be due to the fact that the doses injected into the 25- to 30-g mice were much higher relative to their body weight (100 pg/mouse). In fact, when the vaccine dose was increased to 320 pg per volunteer, twice as many recipients developed high serum titers of antisporozoite antibodies ( > 512) as compared to the recipients of 160 pg of antigen. However, the dose of this vaccine cannot be increased because of the toxic effects of the tetanus toxoid carrier. It is also possible that the immune response in humans was lower than expected because all volunteers had been previously vaccinated with tetanus toxoid. Several studies have shown that the immune response to a hapten or to a peptide is inhibited if the animals were previously
SPOROZOITE MALARIA VACCINES
315
sensitized to the carrier (epitopic suppression) (158-160).In one study (157), mice were immunized with tetanus toxoid prior to receiving the (NANP)s-tetanus toxoid vaccine. This resulted in a specific inhibition of the anti-(NANP)s antibody response, but this effect disappeared following the second booster injection with the vaccine. Another important disadvantage of this particular peptide conjugate vaccine is that it lacks CS protein T cell epitopes that will prime parasitespecific T cells in most individuals. Therefore, the immune response to the vaccinated individuals will not be boosted in endemic areas following exposure to sporozoites. The other vaccine used in humans (136)was produced in Eschen’chiu coli and consisted of a polypeptide containing 32 repeats and 32 nonrelevant amino acids corresponding to part of a tetracycline resistance gene read out of frame (161). Of 15 individuals inoculated, 12 developed antibodies to NANP repeats, but only in a single person was the titer between 1/500 and l/lOOO as assessed by immunoassay. Six immunized volunteers and two controls were challenged. Parasitemia did not develop in the volunteer with the highest titer of antibodies and was delayed in two others. The rather poor antibody response could have been due to lack of T cell recognition or inadequate processing of the subunit vaccine, as suggested by studies showing that the immune response to NANP polymers in mice is severely restricted. Others have presented evidence suggesting that the 32 tetracycline-derivedamino acids could have had a suppressive effect on the immune response to the NANP repeats (162). VIII. Relationship between Immunity to Malaria in Endemic Areas and the Presence in Serum of Antibodies to NANP,
Several investigators reported that antibodies to the CS protein of I! fulcipurum were present in the serum of individuals from endemic areas and noted that the antibody titers and frequency increased with age, in parallel with resistance to infection. As the structure of the I! fulcipurum CS gene was elucidated, it became clear that most or all antibodies in the serum from endemic areas were directed to the NANPS epitope. Several studies were then performed in the endemic areas to investigate the role of these antibodies in the natural history of the disease. These studies confirmed the finding that antibodies to sporozoites increased in amount with the age of the subjects (33) and showed that, if the titers were above 1/1000,they inhibited invasion of hepatoma cells (102). These observations do not prove that these antibodies can protect
316
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
against the natural infection or diminish significantly the parasite load, but they support other studies pointing in the same direction. For example, long-term drug prophylaxis, which leads to the diminution or disappearance of antibodies to the blood stage, did not appear to increase susceptibility to infection when drug administration was terminated (168). Also, a negative correlation was observed between the levels of antibodies to (NANP), and both spleen enlargement and the presence of parasites in thick smears of individuals living in an endemic area in rural Tanzania (103). In investigations performed in The Gambia, West Africa, Marsh et al. (164) observed a close temporal relationship between a fall in parasite rates and the rise in serum levels of antibodies to (NANP)S. In addition, children who were seropositive at the beginning of the malaria season had fewer episodes of malaria than those who were seronegative, but the results did not achieve conventional levels of significance. However, in the same studies it was pointed out that even adults with the highest levels of antibodies to the repeats often had parasitemia . In studies performed in Kenya, in areas where malaria transmission is very intense, 83 adults were treated for malaria and then were monitored for new malaria infection for 98 days. The antibody levels were indistinguishable among individuals who did and did not develop parasitemia (165). In short, antibodies to (NANP), do not appear to play a predominant role in natural immunity to infection with I! fakiparum, and their effect may or may not be detected, depending on the parameters used to measure immunity and probably also on the rates of transmission of the disease in the area under study. A related problem is the reason for the prolonged delay required in endemic areas for achieving high levels of serum antibodies to (NANP),. Probably the most important factor is the very low numbers of sporozoites injected by the mosquitoes (8a). Furthermore, as pointed out before, in the natural host a large proportion of parasites enter hepatocytes rather than macrophages. Hepatocytes do not normally express class I1 molecules, are secluded from the circulation, and do not function at antigen-presenting cells. Epidemiological studies lend strong support to the idea that the levels of antibodies to the CS protein of l? falcifiarum in children depend on the number of infective mosquito bites and the cumulative number of injected sporozoites. In two African villages, which differed greatly in the level at which mosquitoes transmit the disease (between 1 and 100 infective bites per year and per individual), both the prevalence by age group and the levels of antisporozoite antibodies differed markedly. In areas of low transmission, these antibodies were not detected between ages 2 and 10 yr, and increased
SPOROZOITE MALARIA VACCINES
317
thereafter with age. In contrast, in areas of high transmission, all subjects were positive, including very young children. The average titers of antibodies to sporozoites were 100 times higher in the village with high transmission (166). Another possible reason for the delay in mounting an antibody response to the CS protein is that the blood-stage infections suppress the immune response to a variety of antigens (167),including sporozoites (168). Compatible with this idea is the observation that following a period of heightened malaria transmission, when parasitemias are higher, levels of antibodies to (NANP)3 in children tend to be lower (164). IX. Vaccination with P, vivux CS Protein
Plasmodium mvux is the agent of one of the tertian malarias. It is widespread in Asia and Central and South American, and is less frequent in Africa. Although infection with I! mvux is rarely fatal, the course of disease is characterized by frequent relapses, slow development of acquired immunity, and the presence of dormant forms in the liver, which can persist for several years. As mentioned in Section 11, protective immunity against l? vivax infection has been observed in a few humans immunized with irradiated sporozoites, and the development of protective immunity was accompanied by the presence in the serum of antibodies giving the circumsporozoite reaction. The I! vimx CS protein resembles the other members of the family but the structure of the repeat domain is more complex than that of l? fulcipurum (Fig. 9). The repeat subunits consist of nine rather than four amino acids and there are frequent (but conserved) substitutions in two positions of the tandem repeats (73,74). As in the case of the l? fulciparum CS protein, the repeats are conserved among strains of parasites from different areas of the world. However, the greater complexity of the repeats was also associated with a greater complexity in the serological response observed in individuals residing in endemic areas (169). That is, in different sera, the antibodies reacted predominantly with one or the other variant repeat subunits. These findings, taken together with the appealing idea that malaria vaccines should also contain T cell epitopes, have encouraged the development of recombinant CS antigens for vaccination purposes. DNA coding for 234 amino acids of the CS protein of l? mvax was incorporated into yeast expression vectors (Fig. 9). The DNA encoded all the repeat domain, the codons for the highly conserved region I sequence KLKQP, and part of the sequence which follows the repeat domain (170). The transformed yeast cells expressed high levels of the
318
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
FIG.9. Schematic representation of the CS protein of F! Vivax and of the fragment expressed in yeast. The assigned positions correspond to those of Arnot et al. (73). As shown, the repeats bearing alanine in the fourth position ( 0 ) predominate at one end of the repeat domain.
CS polypeptide upon appropriate induction. The malaria antigen was
purified in good yields from the yeast extracts and was injected into mice and Suimiri monkeys. Although several inbred strains of mice were low responders, both outbred mice (170) and the monkeys (146) made high levels of antibodies to the CS antigen using alum as adjuvant. The antibodies recognized the authentic CS protein and at high dilutions inhibited the invasion of hepatocytes by sporozoites in vitro (170). Upon challenge of the immunized Suimzn’monkeys,a few did not develop the infection, whereas some others developed parasitemia only after a prolonged prepatent period. However, a definitive conclusion about the protective value of the vaccine could not be reached because there was a considerable degree of variation in the unvaccinated control group of animals (146). In mice, the yeast-derived CS antigen elicited helper T cells which recognized epitopes from the native CS antigen. The T-dependent secondary responses in vitro were observed when lymphoid cells from immunized mice were challenged with I? vivax sporozoite extracts, and the antibody responses in these mice were vigorously boosted by injection of the sporozoite extracts (171). The latter results suggest that the immune response of individuals receiving the vaccine might also be boosted in endemic areas by the bite of infected mosquitoes. Human trials of the immunogenicity and efficacy of this recombinant product should start in the near future.
SPOROZOITE MALARIA VACCINES
319
X, Perspectives
For maximum efficacy, a sporozoite malaria vaccine should incorporate epitopes from the CS protein capable of inducing high titers of neutralizing antibodies and of sensitizing the CD8+ and CD4+ T cell subsets that respond with the production of interferon-y to parasite antigens introduced during challenge. Genetically engineered purified CS proteins are probably the most promising candidates for vaccine development, since they contain all B and T epitopes. It is not clear, however, whether the incorporation of the antigen in alum will sensitize the T cell subsets involved in protective immunity. Live recombinant animal viruses or bacteria, containing the CS gene, are advantageous with respect to vaccine delivery (172), and probably also in their ability to induce cytotoxic cells and cell-mediated immunity (173). Additional studies are necessary to improve the expression of CS protein and to estimate their safety, particularly in populations where there are high rates of seropositivity for the HIV-1 virus. With regard to synthetic vaccines, they have the advantage that they can be engineered to contain only the relevant T epitopes. Before certain T epitopes are selected for inclusion in such a vaccine, it is essential to establish whether they vary among different strains of parasites, and to determine the significance of this variation for the development of protective immunity. Another important consideration is that a malaria vaccine for use in endemic areas should include not only the CS antigen, but also bloodstage antigens. On the basis of our current knowledge, this ideal malaria vaccine is not likely to become available for widespread use in the near future. Therefore prevention of morbidity and mortality from this disease in endemic areas should continue to rely on integrated primary measures of health care and on chemotherapy, chemoprophylaxis, and mosquito control. Significant progress has been made in recent years in the identification and characterization of relevant malaria antigens, particularly from sporozoites, and in the understanding of mechanisms involved in protective immunity. An important conceptual advance was the realization that EEFs are very susceptible to interferon-? and that a sporozoite vaccine can therefore attack two successive stages of parasite development, the sporozoites and the liver stages. If some sporozoites escape destruction by antibodies to the CS protein repeats and enter hepatocytes, their further development can be prevented by the interferon-y released by sensitized T cells upon encounter with antigens from the invading parasites. Recent evidence strongly suggests that the relevant T cell
320
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
epitopes for eliciting protective immunity are contained in the CS protein (128,129,174). Further studies are necessary to better characterize these epitopes and to determine the most effective way to incorporate them in a vaccine which will sensitize the T cells from humans. However, protection against challenge with a few hundred sporozoites can be achieved in experimental models in the absence of cell-mediated immunity when the serum titers of antibodies to the repeats of the CS protein are high. This dose of sporozoites corresponds to that injected through one bite of an infected mosquito in an endemic area. Therefore, malaria vaccines containing only the CS protein repeats may protect a certain proportion of travelers or migrant workers who do not reside for prolonged periods of time in endemic areas. The probability that sporozoites will escape the neutralizing effect of antibodies will increase with the number of infective bites, even in individuals with high antibody titers, but a large proportion of sporozoites will have been destroyed. The effects which this may have in the course of the disease are difficult to predict. One favorable outcome could be that with a smaller dose of sporozoites the malaria attacks will be less severe and the mortality decreased. This scenario is more likely to occur in individuals who have already developed some degree of immunity to blood stages of the parasite, and who would succumb to challenge with a high dose of parasites. It is also possible that the reduction in the infective doses will affect malaria transmission, if it leads to a decrease in the pool of circulating gametocytes, and a consequent diminution in the rates of infected mosquitoes. Before such a vaccine is introduced in endemic areas, the possible beneficial effects of serum conversion will have to be carefully balanced against the possibilities of vaccine-related injuries. Effective vaccines could also eventually lead to the selection of parasites bearing CS protein repeat variants. Significant enhancement in the humoral and cellular responses to a sporozoite vaccine may be achieved with better adjuvants than those currently used, aluminum hydroxide and aluminum phosphate. More effective adjuvants, such as incomplete Freund's, ISCOMs, liposomes, and muramyl peptide derivatives, have not yet been approved for human use. Particularly encouraging is the demonstration of enhancement of the antibody response to CS antigen by its incorporation into proteosomes (175). The addition of recombinant IL-2 to the vaccines overcame the genetic unresponsiveness of mice to l? fulcipurum sporozoite (NANP), repeats (176). If the safety of some of these immunostimulants is established, and they can be incorporated into a CS vaccine, the efficacy of the vaccine may improve greatly. The efficacy of antibody responses to peptide vaccines may be also
SPOROZOITE MALARIA VACCINES
321
improved by utilizing the multiple-antigen peptide (MAP) system (177). The MAP consists of a small peptidyl core matrix of lysine residues bearing dendritic arms of synthetic peptides (Fig. 10). The MAP, synthesized by the solid-phasemethod, can have molecular weights above 10,000 and can be used directly as antigen. Larger doses of the relevant epitopes can be administered and the need for a carrier bypassed. This is important because the immune response to the carrier may interfere with the immune response to the attached peptide. In addition, MAPs are chemically defined, while peptide conjugates to proteins are not. One of the most attractive features of the MAPs is that they permit the attachment of mulitple different peptides in a single engineered molecule. A single MAP may contain B and T epitopes not only from sporozoites but also from blood stages. MAPs bearing multiple branches of the B epitope of the l? fulcipurum CS protein, (NANP)*, have already been shown to be highly immunogenic in outbred mice and in rabbits, and the elicited antibodies did recognize the native CS protein. It remains to be determined whether CS T epitopes will function jointly with the B epitopes to sensitize parasite-specific T cells.
Multiple- Antigen €@tide
0
A
B
FIG. 10. Schematic representation of the core matrix of the MAP. (A)The lysil core (seven residues) of the octabranching MAP and (B) MAP with peptide antigens attached.
322
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
Since short peptides are disordered structures in water, perhaps “better” antipeptide antibodies, that is, antibodies with higher affinity, can be generated by restricting the conformation of the peptides and by confining them to a shape which resembles that found in the natural immunogen. Although the three-dimensional structure of the NANP repeats is not known, by Chou-Fasman analysis there is a preference for a series of reverse turns, with PN at the corners of the turn. Gibson and Sheraga (84) predicted that the repeats could form a helix containing hydrogen-bonded asparagines around a proline (NPNA). On this basis, a conformationally restricted NANP3 peptide was synthesized by replacing the putative hydrogen bonds between the asparagines with an ethylene bridge (178). Antibodies to this shaped peptide reacted strongly with f?Julcifiarum sporozoites. This is clearly still the first stage of a new strategy for developing peptide vaccines, but it opens the way for synthesizing other shaped peptides and for determining the effect of these defined changes in the immunogenicity of the CS protein B cell epitope. The development of effective malaria vaccines, which only a few years ago faced insurmountable problems, seems now a likely possibility. Although there are still scientific challenges ahead, the rapid progress made in recent years should accelerate even further by the application of innovative and rational approaches to vaccine design and by the increasing participation of basic scientists and the pharmaceutical industry. ACKNOWLEDGMENTS We wish to acknowledge the helpful comments of Dr. F. Sinigaglia, Dr. D. Arnot, Dr. F. Zavala, Dr. E. Nardin, and Dr. R. Pink, and the editorial help of R. Rose. Supported by grants from the Agency for International Development (DEP 0453-A-00-5012-00), the MacArthur Foundation, The UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases, and the National Institutes of Health (United States).
REFERENCES 1. Miller, L. H., Howard, R. J., Carter, R., Good, M. F., Nussenzweig, V., and Nussenzweig, R. S. (1986). Research toward malaria vaccines. Science 234, 1349. 2 . Anden, R. F., Coppel, R. L., Brown, G. V., and Kemp, D. J. (1988). Antigens with repeated amino acid sequences from the asexual blood stages of Plasmodium falciparum. Prog. Allergy 41, 148. 3. Carter, R., Kumar, N., Quakyi, I., Good, M.,Mendis, K., Graves, F!, andMiller, L. (1988). Immunity to sexual stages of malaria parasites. Prog. Allergy 41, 193. 4. Mulligan. H. W., Russell, F! F.,and Mohan, B. N. (1941). Active immunization of fowls against Plasmodium gallinaceum by injections of killed homologous sporozoites. J. Malaz Inst. India 4, 25. 5. Russell, P. F., and Mohan, B. N. (1942). The immunization of fowls against
SPOROZOITE MALARIA VACCINES
6.
7. 8. 8a.
9. 10. 11. 12. 13. 14. 15.
16. 17.
18. 19. 20. 21.
323
mosquito-borne Plasmodium gallinaceum by injections of serum and of inactivated homologous sporozoites. J, Ex#. Med. 76, 477. Nussenzweig, R. S.,Vanderberg, J,, Most, H., and Orton, C. (1967). Protective immunity produced by the injection of X-irradiated sporozoites of Plasmodium berghei. Nature (London) 216, 160. Nussenzweig, R., Vanderberg, J., and Most, H. (1969). Protective immunity produced by the injection of X-irradiated sporozoites of Plasmodium berghei. IV. Dose response, specificity and humoral immunity. Mil. Med. 134, 1176. Nussenzweig, R. S.,Vanderberg, J., Spitalny, G. L., Rivera, C. I., Orton, C., and Most, H. (1972). Sporozoite-induced immunity in mammalian malaria. A review. A m . J. Trop. Med. Hyg. 21, 722. Rosenberg, R., Wirtz, R., and Schneider, I. (1987). An estimation of sporozoite number ejected by a feeding mosquito. Paper presented at the meeting of the American Society of Tropical Medicine and Hygiene, Los Angeles, California. Orjih, A. U., Cochrane, A. H., and Nussenzweig, R. S. (1982). Comparative studies on the immunogenicity of infective and attenuated sporozoites of Plasmodium berghei. Trans. R . SOC. Dop. Med. Hyg. 76, 57. Orjih, A. U., Cochrane, A. H., and Nussenzweig, R. S. (1981). Active immunization and passive transfer of resistance against sporozoite-induced malaria in infant mice. Nature (London) 291, 331. Beer, A. E., Billingham. R. E., and Head, J. (1974). The immunologic siflicance of the mammary gland. J. Invest. Dermatol. 63, 65. Verhave, J. P. (1975). Immunization with sporozoites: An experimental study of Plasmodium berghei malaria. Ph.D. Thesis, Katholieke Universiteit te Nijmegen. Collins, W. E., and Contacos, P. G. (1972). Immunization of monkeys against Plasmodium cynomolgi by X-irradiated sporozoites. Nature (London), New Biol. 236, 176. Nussenzweig, R. S.,and Chen, D. (1974). Various characteristics of the immune response to sporozoites of simian and human malaria. Bol. Of.Sanit. Panam. 77, 469. Gwadz, R. W., Cochrane, A. H., Nussenzweig, V., and Nussenzweig, R. S. (1979). Preliminary studies on vaccination of rhesus monkeys with irradiated sporozoites of Plasmodium knowlesi and characterization of surface antigens of these parasites. Bull. W H . 0 . 57, 165. Clyde, D. F. (1975). Immunization of man against falciparum and vivax malaria by use of attenuated sporozoites. Am. J. Dop. Med. Hyg. 24, 397. McCarthy, V. C., and Clyde, D. F. (1977). Plasmodium wimx: Correlation of circumsporozoite precipitation (CSP) reaction with sporozoite-inducedprotective immunity in man. Exp. Parasitol. 41, 167. Rieckmann, K. H., Carson, €? E., Beaudoin, R. L., Cassells,J. S.,and Sell, K. W. (1974). Letter: Sporozoiteinduced immunity in man against an Ethiopian strain of Plasmodium falciparum. nuns. R . SOC. no$. Med. Hyg. 68, 258. Clyde, D. E, McCarthy, V. C., Miller, R. M.,and Homick, R. B. (1973). Specificity of protection of man immunized against sporozoite-inducedfalciparum malaria. Am. J. Med. Sci. 266, 398. Rieckmann, K. H., Beaudoin, R. L., Cassells, J., and Sell, K. (1979). Clinical studies with a sporozoite vaccine against falciparum malaria. Bull. W H . 0 . 57, Suppl. 1, 261. Clyde, D. E , Most, H., McCarthy, V., and Vanderberg, J. P. (1973). Immunization of man against sporozoite induced falciparum malaria. Am.J. Med. Sci. 266, 169.
324
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
22. Vanderberg, J. P., Nussenzweig, R., and Most, H. (1969). Protective immunity 23. 24.
25. 26. 27. 28. 29.
30.
31. 32. 33.
34.
35.
produced by injections of X-irradiated sporozoites of Plasmodium berghei. V. In vitro effects of immune serum on sporozoites. Mil. Med. 134, Suppl., 1183. Cochrane, A. H., Nussenzweig, R. S., and Nardin, E. H. (1980). Immunization against sporozoites. In “Malaria” (J. F! Kreier, ed.), Vol. 3, p. 163. Academic Press, New York. Spitalny, G. L., and Nussenzweig, R. S. (1972). Effect of various routes of immunization and methods of parasite attenuation on the development of protection against sporozoite-induced rodent malaria. Proc. Helminthol. SOC. Wash. 39, 506. Orjih, A. U., and Nussenzweig, R. S. (1980). Immunization against rodent malaria with cryopreserved irradiated sporozoites of Plasmodium berghei. Am. J. ’Pop. Med. Hyg. 29, 343. Good, M. F., Berzofsky, J. A., and Miller, L. H. (1988). The T cell response to the malaria circumsporozoite protein: An immunological approach to vaccine development. Annu. Rev. Immunol. 6, 663. Ferreira, A., Enea, V., Morimoto. T., and Nussenzweig, V. (1986). Infectivity of Plasmodium bergheisporozoites measured with a DNA probe. Mol. Biochem. Parasitol. 19, 103. Spitalny, G. L., Rivera Ortiz, C. I., and Nussenzweig, R. S. (1976). Plasmodium berghei: The spleen in sporozoite-induced immunity to mouse malaria. Ex#. Parasitol. 40, 179. Nussenzweig, R. S., Vanderberg, J. P., Sanabria, Y., and Most, H. (1972). Plasmodium berghei: Accelerated clearance of sporozoites from blood as part of immune-mechanism in mice. Exp. Parasitol. 31, 88. Nussenzweig, R., Herman, R., Vanderberg, J., Yoeli, M., and Most, H. (1966). Studies on sporozoite-induced infections of rodent malaria. 3. The coume of sporozoite-induced Plasmodium berghei in different hosts. A m . J. T7op. Med. Hyg. 15, 684. Yoeli, M., Upmanis, R. S., Vanderberg, J., and Most, H. (1966). Life cycle and patterns of development of Plasmodium berghei in normal and experimental hosts. Mil. Med. 131, Suppl., 900. Desser, S. S., Weller, I., and Yoeli, M. (1972). An ultrastructural study of the pre-erythrocyticdevelopment of Plasmodium berghei in the tree rat, Thumnomys surdastez Can. J. 2001. 50, 821. Nardin, E. H., Nussenzweig, R. S., McGregor, I. A., and Bryan, J. H. (1979). Antibodies to sporozoites: Their frequent occurrence in individuals living in an area of hyperendemic malaria. Science 206, 597. Gwadz, R. W., Cochrane, A. H., Nussenzweig, V., andNussenzweig, R. S. (1979). Preliminary studies on vaccination of rhesus monkeys with irradiated sporozoites of Plasmodum knowlesi and characterization of surface antigens of these parasites. Bull. W.H.O. 57, 165. Collins, W. E., and Contacos, P. G. (1972). Immunization of monkeys against Plasmodium cynomolgi by X-irradiated sporozoites. Nature (London), New Biol.
236, 176. 36. Nardin, E. H., and Nussenzweig, R. S. (1978). Stage-specific antigens on the surface membrane of sporozoites of malaria parasites. Nature (London) 274, 55. 37. Nussenzweig, R. S., Vanderberg,J. I?, Most, H., and Orton, C. (1969). Specificity
of protective immunity produced by X-irradiated Plasmodium berghei sporozoites. Nature (London) 222, 488.
SPOROZOITE MALARIA VACCINES
325
38. Vanderberg, J., Nussenzweig, R., and Most, H. (1969). Protective immunity produced by the injection of X-irradiated sporozoitesof Plasmodium berghei. V. In Vitro effects of immune serum on sporozoites. Mil. Med. 134, 1183. 39. Cochrane, A. H., Aikawa, M., Jeng, M., and Nussenzweig, R. S. (1976). Antibodyinduced ultrastructural changes of malarial sporozoites.J. Immunol. 116, 859. 40. Potocnjak, P., Yoshida, N., Nussenzweig, R. S., and Nussenzweig, V. (1980). Monovalent fragments (Fab) of monoclonal antibodies to a sporozoite surface antigen (Pb44) protect mice against malarial infecti0n.J. Ex#. Med. 151, 1504. 41. Godson, G. N. (1985). Molecular approaches to malaria vaccines. Sci. Am. 252, 52. 42. Spitalny, G. L., and Nussenzweig, R. S. (1973). Plasmodium berghei: Relationship between protective immunity and anti-sporozoite (CSP) antibody in mice. Exp. Parasitol. 33, 168. 43. Nardin, E. H., Nussenzweig, R. S., Bryan, J. H., and McGregor, I. A. (1981). Congenital transfer of antibodies against malarial sporozoites detected in Gambian infants. Am. J. no$. Med. Hyg. 30, 1159. 44. Chen, D. H., Tigelaar, R. E., and Weinbaum, E I. (1977). Immunity to sporozoiteinduced malaria infection in mice. I. The effect of immunization of T and B cell-deficient mice. J. Immunol. 118, 1322. 45. Spitalny, G. L., Verhave, J. P., Meuwissen,J. H., andNussenzweig, R. S. (1977). Plasmodium berghei: T cell dependence of sporozoite-induced immunity in rodents. Exp. Parasitol. 42, 73. 46. Shin, S. C., Vanderberg, J. P., and Terzakis, J. A. (1982). Direct infection of hepatocytes by sporozoites of Plasmodium berghei. J. Protoxool. 29, 448. 47. Yoshida, N., Nussenzweig, R. S., Potocnjak, P., Nussenzweig, V.,and Aikawa, M. (1980). Hybridoma produces protective antibodies directed against the sporozoite stage of malaria parasite. Science 207, 71. 48. Ferreira, A., Morimoto, T., Altszuler, R., and Nussenzweig, V. (1987). Use of a DNA probe to measure the neutralization of Plasmodium berghei sporozoites by a monoclonal antibody. J. Immunol. 138, 1256. 49. Hollingdale, M. R., Nardin, E. H., Tharavanij, S., Schwartz, A. L., and Nussenzweig, R. S. (1984). Inhibition of entry of Plasmodium falciparum and f! vivax sporozoites into cultured cells; an in vitro assay of protective antibodies. J. Immunol. 132, 909. 50. Hollingdale, M. R., Zavala, F., Nussenzweig, R. S., and Nussenzweig, V. (1982). Antibodies to the protective antigen of Plasmodium berghei sporozoites prevent entry into cultured cells. J. Immunol. 128, 1929. 51. Zavala, E,Hollingdale, M. R., Schwartz, A. L., Nussenzweig, R. S., andNussenzweig, V. (1985). Immunoradiometric assay to measure the in Vitro penetration of sporozoites of malaria parasites into hepatoma cells. J. Immunol. 134, 1202. 52. Egan, J. E., Weber, J. L., Ballou, W. R., Hollingdale, M. R., Majarian, W. R., Gordon, D. M., Maloy, W. L., Hoffman, S. L., Wirtz, R. A., Schneider, I., Woollett, G. R., Young, J. F., and Hockmeyer, W. T.(1987). Efficacy of murine malaria sporozoitevaccines: Implicationsfor human vaccine development. Science 236, 453. 53. Cochrane, A. H., Santoro, E, Nussenzweig, V., Gwadz, R. W., and Nussenzweig, R. S. (1982). Monoclonal antibodies identlfy the protective antigens of sporozoites of Plasmodium knowlesi. Proc. Natl. Acad. Sci. U.S.A. 79, 5651. 54. Nardin, E. H., Nussenzweig, V., Nussenzweig, R. S., Collins, W. E., Harinasuta, K. T., Tapchaisri, P., and Chomcharn, Y. (1982). Circumsporozoite
326
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIC
proteins of human malaria parasites Plasmodiumfalciparum and Plasmodium viuax. J. &p. Med. 156, 20. 55. Vanderberg, J. P. (1975). Development of infectivity of the Plasmodium berghei sporozoite.J. Parasitol. 61, 43. 56. Aikawa, M., Yoshida, N., Nussenzweig, R. S., and Nussenzweig, V. (1981). The protective antigen of malarial sporozoites (Phmodium berghez) is a differentiation antigen. J. Immunol. 126, 2494. 57. Santoro, F,, Cochrane, A. H., Nussenzweig, V., Nardin, E. H., Nussenzweig, R. S., Gwadz, R. W., and Ferreira, A. (1983). Structural similaritiesamong the protective antigens of sporozoites from different species of malaria parasites. J. Biol. Chem. 258, 3341. 58. Yoshida, N., Potocnjak. P., Nussenzweig, V., and Nussenzweig, R. S. (1981). Biosynthesis of Pb44, the protective antigen of sporozoites of Phmodium berghei. J. &p. Med. 154, 1225. 59. Zavala, F., Cochrane. A. H., Nardin, E. H., Nussenzweig, R. S., and Nussenzweig, V. (1983). Circumsporozoite proteins of malaria parasites contain a single immunodominant region with two or more identical epit0pes.J. Ex#. Med. 157,
1947. 60. Zavala, F., Tam, J. l?, Hollingdale, M. R., Cochrane, A. H., Quakyi, I., Nussenzweig, R. S., and Nussenzweig, V. (1985). Rationale for development of a synthetic vaccine against Plasmodiumfalciparum malaria. Science 228, 1436. 61. Nussenzweig, R. S., and Chen, D. (1974). The antibody response to sporozoites
of simian and human malaria parasites: Its stage and species specificity and strain cross-reactivity. Bull. W H.O.50, 293. 62. Chen, D. H., Nussenzweig, R. S., and Collins, W. E. (1976). Specificity of the circum-sporozoite precipitation antigen(s) of human and simian malarias. J. Pamsitol. 62, 636. 63. Zavala, E,Gwadz, R. W., Collins, E H., Nmnzweig, R. S., and Nussenzweig, V. (1982). Monoclonal antibodies to circumsporozoite proteins identify the species of malaria parasite in infected mosquitoes. Nature (London) 299, 737. 64. Cochrane, A. H., Collins, W. E., and Nussenzweig, R. S. (1984). Monoclonal antibody identifies circumsporozoiteprotein of Plasmodium malanae and detects a common epitope on Plasmodium brasilianum sporozoites. Infect. Immun. 45, 592. 65. Cochrane, A. H., Barnwell, J. W., Collins, W. E., and Nussenzweig, R. S. (1985). Monoclonal antibodies produced against sporozoites of the human parasite Plasmodium malanae abolish infectivity of sporozoites of the simian parasite Plasmodium brasilianum. Infect. Zmmun. 50, 58. 66. Coatney, G. R., Collins, W. E., Warren, M., and Contacos, P. G. (1971). “Primate Malarias.” U.S.Dept. of Health, Education and Welfare, Washington, D.C. 67. Lal, A. A., de la Cruz, V. F., Collins, W. E., Campbell, G. H., Procell, P. M., and McCutchan, T. F. (1988). Circumsporozoite protein gene from Plasmodium brasilianum. Animal reservoirs for human malaria parasites.J. Biol Chem. 263,
291. 68. Lal, A. A., de la Cruz, V. F., Campbell, G. H., Procell, I? M., Collins, W. E., and McCutchan, T. F. (1988). Structure of the circumsporozoite gene of Plasmodium malanae. Mol. Biochem. Parasitol. 30, 291. 69. Ellis, J . , Ozaki, L. S., Gwadz, R. W., Cochrane, A. H., Nussenzweig, V., Nussenzweig, R. S.,and Godson, G. N. (1983). Cloning and expression in E. coli
of the malarial sporozoitesurface antigen gene from Plasmodium knowlesi. Nature (London) 302, 536.
SPOROZOITE MALARIA VACCINES
327
70. Godson, G. N., Ellis, J., Svec, P., Schlesinger, D. H., and Nussenzweig, V. (1983).
Identification and chemical synthesis of a tandemly repeated immunogenic region of Plasmodium knowlesi circumsporozoite protein. Nature (London) 305, 29. 71. Enea, V., Ellis, J., Zavala, E,Amot, D. E., Quakyi, I., and Nussenzweig, R. S. (1984).DNA cloning of Plasmodium falci’arum circumsporozoite gene: Amino acid sequence of repetitive epitope. Science 225, 628. 72. Dame, J. B., Williams, J. L.,McCutchan, T. F., Weber, J. L., Wirtz, R. A., Hockmeyer, W. T., Maloy, W. L., Haynes, J. D., Schneider, I., Roberts, D., Sanders, G. S., Reddp E. P., Diggs, C. L., and Miller, L. H. (1984). Structure of the gene encoding the immunodominant surface antigen on the sporozoite of the human malaria parasite Plasmodium falciparum. Science 225, 593. 73. Amot, D. E., Barnwell, J. W., Tam, J. P., Nussenzweig, V., Nussenzweig, R. S., and Enea, V. (1985). Circumsporozoiteprotein of Plasmodium wvax: Gene cloning and characterization of the immunodominant epitope. Science 230, 815. 74. McCutchan, T. F., Lal, A. A., de la Cruz, V. F., Miller, L. H., Maloy, W. L., Charoenvit, Y.,Beaudoin, R. L., Guerry, P., Wistar, R., Jr., Hoffman, S. L., Hockmeyer, W. T., Collins, W. E., and Wirth, D. (1985). Sequence of the immunodominant epitope for the surface protein on sporozoites of Plasmodium viuax. Science 230, 1381. 75. Galinski, M. R., Amot, D. E., Cochrane, A. H., Barnwell, J. W., Nussenzweig, R. S., and Enea, V. (1987).The circumsporozoitegene of the Plasmodium cynomolgi complex. Cell (Cambridge, Mass.) 48(2), 311. 76. Ozaki, L.S., Svec, P., Nussenzweig, R. S., Nussenzweig, V., and Godson, G. N. (1983). Structure of the Plasmodium knowlesigene coding for the circumsporozoite protein. Cell (Cambridge, Mass.) 34, 815. 77. Sharma, S., Svec, P., Mitchell, G. H., and Godson, G. N. (1985). Diversity of circumsporozoite antigen genes from two strains of the malarial parasite Plasmodium knowlesi. Science 229, 779. 78. Weber, J. L.,Egan, J. E., Lyon, J. A., Wirtz, R. A., Charoenvit, Y., Maloy, W. L., and Hockmeyer, W. T. (1987). Plasmodium berghei: Cloning of the circumsporozoite protein gene. Ex’. Parasitol. 63, 295. 79. Eichinger, D. J., Amot, D. E., Tam, J. P., Nussenzweig, V., and Enea, V. (1986). Circumsporozoiteprotein of Plasmodium berghei: Gene cloning and identification of the immunodominant epitopes. Mol. Cell. Biol. 6, 3965. 80. Lal, A. A., de la Cruz, V. F., Welsh, J . A., Charoenvit, Y.,Maloy, W. L., and McCutchan, T. F. (1987). Structure of the gene encoding the circumsporozoite protein of Plasmodium yoeliz. A rodent model for examining antimalarial sporozoite vaccines. J. Biol. Chem. 262, 2937. 81. de la Cruz, V. E , Lal, A. A., and McCutchan, T. E (1988). Variation among circumsporozoiteprotein genes from rodent malarias. Mol. Biochem. Pamntol. 28,31. 82. Magee, T., and Hanley, M. (1988). Sticky fingers and CAAX boxes. Nature (London) 335, 114. 83. Low, M. G. (1987). Biochemistry of the glycosylphosphatidylinositolmembrane protein anchors. Biochem. J. 244, 1. 84. Gibson, K. D., and Scheraga, H. A. (1986). Predicted conformations for the immunodominant region of the circumsporozoite protein of the human malaria parasite Plasmodium falciparum. Proc. Natl. Acad. Sci. US.A. 83, 5649. 85. Brooks, B. R.,Pastor, R. W., andcanon, E W. (1987).Theoretically determined three-dimensional structure for the repeating tetrapeptide unit of the circumsporozoite coat protein of the malaria parasite Plasmodium fakiparum. Proc. Natl. Acad. Sci. U.S.A. 84, 4470.
328
VICTOR NUSSENZWEIC AND RUTH S. NUSSENZWEIG
86. Vergara, U., Ruiz, A., Ferreira, A., Nussenzweig, R. S., and Nussenzweig, V. (1985). Conserved pup-specific epitopes of the circumsporozoite proteins revealed by antibodies to synthetic peptides. J. Immunol. 134, 3445. 87. Aley, S. B., Bates, M. D., Tam, J. P., and Hollingdale, M. R. (1986). Synthetic
peptides from the circumsporozoite proteins of Plasmodium falciparum and Plasmodium knowlesi recognize the human hepatoma cell line HepG2-A16 in v i 2 m J. Exp. Med. 164, 1915. 88. Vergara, U., Gwadz, R., Schlesinger, D., Nussenzweig, V., andFerreira, A. (1985). Multiple non-repeated epitopes on the circumsporozoiteprotein of Plasmodium knowlesi. Mol. Biochem. Parasitol. 14, 283. 89. Robson, K. J. H., Hall, J. R. S.,Jennings, M. W., Harris, T. J. R., Marsh, K., Newbold, C. I., Tate, V. E., and Weatherall, D. J. (1988). A highly conserved amino-acid sequence in thrombospondin, properdin and in proteins from sporozoites and blood stages of a human malaria parasite. Nature (London) 335, 79. 90. Goundis, D., and Reid, K. B. M. (1988). Properdin, the terminal complement components, thrombospondin and the circumsporozoite protein of malaria parasites contain similar sequence motifs. Nature (London) 335, 82. 91. del Portillo, H. A., Nussenzweig, R. S.,and Enea, V. (1987). Circumsporozoite gene of a Plasmodiumfalciparum strain from Thailand. Mol. Biochem. Pamitol. 24, 289. 92. k k y e r , M. J., and Schwarz, R. T. (1987). Strain variation in the circumsporozoite protein gene of Plasmodium fakiparum. Mol. Biochem. Parasitol. 22, 101. 93. McCutchan, T. F,, de la Cruz, V. F., Good, M. F., and Wellens, T. E. (1988). Antigenic diversity in Plasmodium falciparum. Prog. Allergy 41, 173. 94. McCutchan, T. F., Dame, J. B., Miller, L. H.,and Barnwell, J. (1984). Evolutionary relatedness of Plasmodium species as determined by the structure of DNA. Science 225, 808. 95. de la Cruz, V. E,Lal,A. A., Welsh, J. A., and McCutchan, T. E (1987). Evolution of the immunodominant domain of the circumsporozoite protein gene from Plasmodium vi’mx. Implications for vaccines. J. Biol. Chem. 262, 6464. 96. Amot, D. E., Barnwell, J. W., and Stewart, M. J. (1988). Does biased gene conversion influence polymorphism in the CS gene of Plasmodium viuax? Proc. Natl. Acad. Sci. U.S.A. 85, 8102. 97. de la Cruz, V. F., Lal, A. A., and McCutchan, T. F. (1987). Sequence variation in putative functional domains of the circumsporozoite protein of Plasmodium fulciparum. Implications for vaccine development. J. Biol. Chem. 262, 11935. 98. Enea, V., Galinski, M., Schmidt, E., Gwadz, R., and Nussenzweig, R. S. (1986). Evolutionary profile of the circumsporozoitegene of the Plasmodium cynomolgi complex. J. Mol. Biol. 188, 721. 99. McCutchan, T. F., de la Cruz, V. E , Good, M. F., and Wellems, T. F, (1988). Antigenic diversity in Plasmodium falciparum. Prog. Allergy 41, 173. 100. Amot, D. (1989). Malaria and the major histocompatibility complex. Parasitol. 7bday (in press). 101. Tapchaisri, P., Dekumyoy, P., Asavanich, A., Tharavanij, S.,and Harinasuta, T. (1986). Occurrence of a common epitope in circumsporozoite proteins of P h o d i u m faln$arum isolated from different areas in Thailand. Southeart Asian J. nap. Med. Public Health 17, 13. 102. Hofman, S. L., Wistar, R., Jr., Ballou, W. R., Hollingdale, M. R., Wirtz, R. A.,
SPOROZOITE MALARIA VACCINES
329
Schneider, I., Marwoto, H. A., and Hockmeyer, W. T. (1986). Immunity to malaria and naturally acquired antibodies to the circumsporozoite protein of Plasmodium falciparum. N. Engl. J. Med. 315, 601. 103. Del Giudice, G., Engers, H. D., Tougne, C., Biro, S. S., Weiss, N., Verdini, A. S., Pessi, A., Degremont, A. A., Freyvogel, T. A., Lambert, P. H., and Tanner, M. (1987). Antibodies to the repetitive epitope of Plasmodium falciparum circumsporozoite protein in a rural Tanzanian community: A longitudinal study of 132 children. A m . J . Trop. Med. Hyg. 36, 203. 104. Esposito, F., Lombardi, S., Modiano, D., Zavala, F., Fkeme, J., Diallo, M., Coluzzi, M., and Nussenzweig, R. S. (1989). Prevalence and level of antibodies to the circumsporozoite protein of Plasmodium falciparum in an endemic area. Its relationship to resistance against malaria infection. Truns R . SOC. no$.Med. Hyg. (in press). 105. Weber, J. L., and Hockmeyer, W. T. (1985). Structure of the circumsporozoite protein gene in 18 strains of Plasmodium falciparum. Mol. Biochem. Purasitol. 15, 305. 106. Yoshida, N., del Portillo, H. A., di Santi, S. M., Nussenzweig, R. S., and Enea, V. (1987). Plasmodiumfalciparum: Epidemiological studies on the circumsporozoite gene. Exp. Rzrasztol. 64, 510. 107. Zavala, E,Masuda, A., Graves, F? M.,Nussenzweig, V., andNussenzweig, R. S. (1985). Ubiquity of the repetitive epitope of the CS protein in different isolates of human malaria parasites. J Immunol. 135, 2790. 108. Nussenzweig, V., and Nussenzweig, R. S. (1985). Circumsporozoite proteins of malaria parasites. Cell (Cambridge, Mass.) 42, 401. 109. Nussenzweig, V., and Nussenzweig, R. S. (1988). Sporozoite malaria vaccines. In “The Biology of Parasitism” (P. T. Englund and A. Sher, eds.), p. 183. Alan R. Liss, New York. 110. Zavala, F., Tam, J. P., Barr, P. J., Romero, P. J., Ley, V., Nussenzweig, R. S., and Nussenzweig, V. (1987). Synthetic peptide vaccine confers protection against murine ma1aria.J. Exp. Med. 166, 1591. 111. Lal, A. A., de la Cruz, V. E, Good, M. F., Weiss, W. R., Lunde, M., Maloy, W. L., Welsh, J. A. and McCutchan, T. E (1987). In Viva testing of subunit vaccines against malaria sporozoites using a rodent system. Proc. Natl. Acad. Sci. U.S.A. 84, 8647. 112. Jahiel, R. I., Nussenzweig, R. S., Vanderberg, J., and Vilcek, J. (1968). Anti-malarial effect of interferon inducers at different stages of development of Plasmodium berghei in the mouse. Nature (London) 220, 710. 113. Jahiel, R. I., Vilcek, J., Nussenzweig, R., and Vanderberg,J. (1968). Interferon inducers protect mice against Phmodium berghei malaria. Science 161, 802. 114. Ferreira, A., Schofield, L., Enea, V., Schellekens, H., van der Meide, P., Collins, W. E., Nussenzweig, R. S., and Nussenzweig, V. (1986). Inhibition of development of exoerythrocyticforms of malaria parasites by gamma-interferon. Science 232, 881. 115. Maheshwari, R. K., Czarniecki, C. W., Dutta, G. P., Pun, S. K.,Dhawan, B. N., and Friedman, R. M. (1986). Recombinant human gamma interferon inhibits simian malaria. Infect. Immun. 53, 628. 116. Schofield, L., Ferreira, A., Altszuler, R., Nussenzweig, V., and Nussenzweig, R. S. (1987). Interferon-gamma inhibits the intrahepatocytic development of malaria parasites in vitro. J. Immunol. 139, 2020.
330
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
117. Pfefferkorn, E. R. (1984). Interferon-gamma blocks the growth of Toxoplasma gondii in human fibroblasts by inducing the host cell to degrade tryptophan. Proc. Natl. Acad. Sci. U.S.A. 81, 908. 118. Vergara, U.,Ferreira, A,, Schellekens, H., and Nussenzweig, V. (1987). Mechanism of escape of exoerythrocytic forms (EEF) of malaria parasites from the inhibitory effects of interferon-gamma. J. Immunol. 138, 4447. 119. Wojcik, E., Dvorak, C., Chianale, J., Traber, P. G., Keren, D., and Gumucio, J. J. (1988). Demonstration by in situ hybridization of the zonal modulation of rat liver cytochrome P-450b and P-450e gene expression after phenobarbita1.J. Clin. Invest. 82. 658. 120. Ojo-Amaize, E. A., Vilcek, J., Cochrane, A. H., and Nussenzweig, R. S. (1984). Plasmodium berghei sporozoites are mitogenic for murine T cells, induce interferon, and activate natural killer cells. J. Immunol. 133, 1005. 121. Ojo-Amaize, E., Salimonu, L. S., Williams, A., Akinwolere, O., Shabo, R., Alm, G. V., and Wigzell, H. (1981). Positive correlation between the degree of parasitemia, interferon titres and natural killer cell activity in Plasmodium falciparum infected children. J. Immunol. 127, 2296. 122. Rhodes-Feuillete,A., Druilhe, P., Canivet, M., Gentilini, M., and Penes, J. (1981). Presence dinterferon circulant dans le serum des malades infect& par Plasmodium falciparum. C.R. Hebd. Seances Acad. Sci. 293, 635. 123. Schofield, L., Villaquiran, J., Ferreira, A . , Schellekens, H., Nussenzweig, R., and Nussenzweig, V. (1987). Gamma interferon, CDBt T cells and antibodies required for immunity to malaria sporozoites. Nature (London) 330, 664. 124. Weiss, W. R., Sedegah, M., Beaudoin, R. L., Miller, L. H., and Good, M. F. (1988). CD8+ T cells (cytotoxic/suppressors)are required for protection in mice immunized with malaria sporozoites. Proc. Natl. Acad. Sci. U.S.A. 85, 573. 125. Bevan, M. J. (1987). Class discrimination in the world of immunology. Nature (London) 325, 192. 126. Danforth, H. D., Aikawa, M., Cochrane, A. H., andNussenzweig, R. S . (1980). Sporozoites of mammalian malaria: Attachment to, interiorization and fate within macrophages. J. Protozool. 27, 193. 126a. Kaufmann, S . H. E. (1988). CD8+ T lymphocytes in intracellular microbial infections. Immunol. Today 9, 168. 127. Shortt, H. E., and Garnham, P. C. C. (1948). The pre-erythrocytic development of Plasmodium cynomolgi and Plasmodium Vivax. 33ans. R . SOC. Trop. Med. Hyg. 41, 785. 128. Sadoff, J. C., Ballou, W. R., Baron, L. S., Majarian, W. R., Brey, R. N., Hockmeyer, W. T., Young, J. F., Cryz, S . J., Ou, J., Lowell, G. H., and Chulay, J. D. (1988). Oral Salmonella typhimunum vaccine expressing circumsporozoite protein protects against malaria. Science 240, 336. 129. Sadoff, J. C., Ballou, W. R., Barron, L. S . , Ou, J., and Young, J. F. (1988). Oral Salmonella typhimunum circumsporozoite recombinant vaccines protect against malaria. In “Modern Approaches to New Vaccines, Including Prevention of AIDS” (P.Chernock, H. S. Ginsberg, R. A. Lerner, and F. Brown, eds.), p. 35. Cold Spring Harbor Lab., Cold Spring Harbor, New York. 129a. Kumar, S., Miller, L. H., Quakyi, I. A , , Keister, D. B., Houghton, R. A., Maloy, W. L., Moss, B., Berzofsky, J. A., and Good, M. F. (1988). Cytotoxic T cells specific for the circumsporozoite protein of Plasmodium falciparum. Nature (London) 334, 258.
SPOROZOITE MALARIA VACCINES
33 1
130. Scherle, P.A., and Gerhard, W. (1986).Functional analysis of influenza-specific helper T cell clones in Vivo. T cells specific for internal viral proteins provide cognate help for B cell response to hemagglutinin. J. Ex$. Med. 164, 1114. 13Oa. Milich, D. R., McLachlan, A., Thornton, G. B., and Hughes, J. L. (1987). Antibody production to the nucleocapsid and envelope of the hepatitis B virus primed by a single synthetic T cell site. Nature (London) 329, 547. 131. DelGiudice, G., Verdini, A. S.,Pinori, M., Pessi, A , , Verhave, J. P., Tougne, C., Ivanoff, B., Lambert, P. H., and Engers, H. D. (1987). Detection of human antibodies against Plasmodiumfalciparum sporozoites using synthetic peptides. J. Clin. Microbiol. 25, 91. 132. Romero, P. J., Tam, J. P., Schlesinger, P., Calvijo, P., Barr, P. J., Nussenzweig, R. S., Nussenzweig, V., and Zavala, F. (1989). Multiple T-helper cell epitopes of the circumsporozoite protein of Plasmodium berghei. Euz J. Zmmunol. (in press). 133. Nardin, E. H., Ban; P. J., Heimer, E., and Ettlinger, H. M. (1988). Genetic restriction of the murine humoral response to a recombinant Plasmodium vimx circumsporozoite protein. Euz J. Zmmunol. 18, 1119. 134. Good, M. F., Berzofsky, J. A., Maloy, W. L., Hayashi, Y., Fujii, N., Hockmeyer, W. T., and Miller, L. H. (1986). Genetic control of the immune response in mice to a Platmodium falciparum sporozoite vaccine. Widespread nonresponsiveness to single malaria T epitope in highly repetitive vaccine. J. Exp. Med. 164, 655. 135. Del Giudice, G., Cooper, J. A., Merino, J., Verdini, A. S.,Pessi, A., Togna, A. R., Engers, H. D., Corradin, G., and Lambert, P. H. (1986). The antibody response in mice to carrier-free synthetic polymers of Plasmodium falciparum circumsporozoite repetitive epitope is I -Ab-restricted: Possible implications for malaria vaccines. J. Immunol. 137, 2952. 136. Ballou, W. R.,Hoffman, S. L., Sherwood,J. A., Hollingdale, M. R., Neva, E A.. Hockmeyer, W. T., Gordon, D. M., Schneider, I., Wirtz, R. A., Young, J. F., Wasserman, G. F., Reeve, P.,Diggs, C. L., and Chulay, J. D. (1987).Safety and efficacy of a recombinant DNA Plasmodium falciparum sporozoite vaccine. Lancet 1, 1277. 137. Good, M. F., Pombo, D., Quakyi, 1. A,, Riley, E. M., Houghten, R. A., Menon, A., Alling, D. W., Berzofsky, J. A., and Miller, L. H. (1988). Human T-cell recognition of the circumsporozoite protein of Plasmodium falciparum: Immunodominant T-cell domains map to the polymorphic regions of the molecule. Proc. Natl. Acad. Sci. U.S.A. 85, 1199. 138. Campbell, J. R., Paleologo, F. D., Franke, E. D., Ratiwayanto, S., Hadiputranto, H., Kurniawan, L., Wistar, R., Jr., Hoffman, S. L., Annis, B. A., Wasserman, G., and Laughlin, L. W. (1988). Immune response of humans to the circumsporozoite protein of Plasmodium falciparum: Limited T-cell response to the immunodominant central repeat region. A m . J. Trop. Med. Hyg. 39,232. 139. Zavala, F. et al. (1989). In preparation. 140. Nardin, E. et al. (1989). In preparation. 141. Good, M. F., Maloy, W. L., Lunde, M. N., Margalit, H., Cornette, J. L., Smith, G. L., Moss, B., Miller, L. H., and Berzofsky, J. A. (1987). Construction of synthetic immunogen: Use of new T-helper epitope on malaria circumsporozoite protein. Science 235, 1059. 142. Satchidanandam, V. et al. (1989). In preparation.
332
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
143. Smith, G. L., Godson, G. N., Nussenzweig,V., Nussenzweig, R. S..Bamwell, J., and Moss, B. (1984). Plasmodium Rnowlesi sporozoite antigen: Expression by infectious recombinant vaccinia virus. Science 224, 397. 144. Cheng, K.-C., Smith, G. L., Moss, B., Zavala, F., Nussenzweig, R. S., and Nussenzweig, V. (1986). Expression of malaria circumsporozoite protein and hepatitis B virus surface antigen by infectious vaccinia virus. In “Vaccines 86” (F. Brown, R. M. Chanock, and R. A. Lerner, eds). p. 165. Cold Spring Harbor Lab., Cold Spring Harbor, New York. 145. Ban; P. J., Gibson, H. L., Enea, V., Amot, D. E., Hollingdale, M. R., and Nussenzweig, V. (1987). Expression in yeast of a Plasmodium Vivax antigen of potential use in a human malaria vaccine.]. ESP. Med. 165, 1160. 146. Collins, W. E., Nussenzweig, R. S., Chulay, J. D., Nussenzweig, V., Ruebush, R. K.,11, Nardin, E. H., Ballou, W. R., Wasserman, S. S., Brodenon, J. R., Skinner, J. C., Filipsky, V. K., Wilson, C. L., Young, J. F., Wasserman, G. W., and Bathurst, I. (1989). Immunization of Saimin’sciurew boliuiensis monkeys with recombinant vaccines based on the circumsporozoite protein of Plasmodium w‘vax. Submitted for publication. 147. De Groot, A. S., Maloy, W. L.,Johnson, A., Berzofsky, J. A., and Good, M. E (1988). Human immune response to polymorphic malaria circumsporozoiteT-cell epitopes. In “Modem Approaches to New Vaccines, Including Prevention of AIDS’ (R. Chanock, H. S. Ginsberg, R. A. Lemer, and F. Brown, eds.), p. 66. Cold Spring Harbor Lab., Cold Spring Harbor, New York. 148. Margalit, H., Spouge, J. L., Cornette, J. L., Cease, K. B., DeLisi, C., and Berzofsky, J. A. (1987). Prediction of immunodominant helper T cell antigenic sites from the primary sequence. J. Immunol. 138, 2213. 149. Rothbard, J. B.,and Taylor, W. R. (1988). A sequence pattern common to T cell epitopes. EMBOJ. 7 , 93. 150. Sinigaglia, F.,Guttinger, M., Gillessen, D., Doran, D. M., Takacs, B., Matile, H., ’Iheciak, A,, and Pink, J. R. (1988). Epitopes recognized by human T-lymphocytes on malaria circumsporozoite protein. Eur. J. Immunol. 18, 633. 151. Guttinger, M., Caspen, P., Takacs, B., Trzeciak, A., Gillesson, D., Pink, J. R., and Sinigaglia, F. (1988). Human T cells recognize polymorphic and non-polymorphicregions of the Plasmodiumfala$arum circumsporozoiteprotein. EMBOJ. 7, 2555. 152. Sinigaglia, F.,Guttinger, M., Kilgus, J.. Doran, D. M., Matile, H., Etlinger, H., Trzeciak, A., Gillessen, D., and Pink, J. R. L. (1988). A malaria T cell epitope recognized in association with most mouse and human MHC class I1 molecules. Nature (London) 336, 778. 153. Good, M. F.,Pombo, D., Maloy, W. L., de la Cruz, V. F., Miller, L. H., and Berzofsky, J. A. (1988). Parasite polymorphism present within minimal T cell epitopes of Plasmodium falci’parum circumsporozoite pr0tein.J. Immunol. 140, 1645. 154. Etlinger, H. M., Felix, A. M., Gillessen, D., Heimer, E. P.,Just, M., Pink, J. R., Sinigaglia, F., Sturchler, D., Takacs, B., Trzeciak, A., and Matile, H. (1988). Assessment in humans of a synthetic peptide-based vaccine against the sporozoite stage of the human malaria parasite,
[email protected]. Immuml. 140,626. 155. Herrington, D. A., Clyde, D. F., Losonsky, G., Cortesia, M., Murphy, J. R., Davis, J., Baqar, S., Felix, A. M., Heimer, E. P., Gillessen, D., Nardin, E., Nussenzweig, R. S., Nussenzweig, V., Hollingdale, M. R., and Levine, M. M. (1987). Safety and immunogenicity in man of a synthetic peptide malaria vaccine
SPOROZOITE MALARIA VACCINES
333
against Plasmodium falciparum sporozoites. Nature (London) 328, 257. 156. Schmidt, L. H., Fradkin, R., Genther, C. S., Rossan, R. N., and Squires, W. (1982). I. The characteristics of untreated sporozoite-induced and trophozoiteinduced infections. Am. J. ‘prop. Med. Hyg. 31, Suppl., 612. 157. Etlinger, H. M., Heimer, E. P., Trzeciak, A., Felix, A. M., and Gillessen, D. (1988). Assessment in mice of a synthetic peptide-based vaccine against the sporozoite stage of the human malaria parasite, I! falciparum. Immunology 64, 551. 158. Herzenberg, L. A., and Tokuhisa, T. (1982). Epitope-specific regulation. I. Carrier-specific induction of suppression for IgG anti-hapten antibody responses. J. Ex#. Med. 155, 1730. 159. Jacob, C. O.,Amon, R., and Sela, M. (1985). Effect of carrier on the immunogenic capacity of synthetic cholera vaccine. Mol. Immunol. 22, 1333. 160. Schutze. M.-P., Leclerc, C., Vogel, F. R., and Chcdid, L. (1987). Epitopic suppression in synthetic vaccine models: Analysis of the effector mechanisms. Cell. Immunol. 104, 79. 161. Young, J. F., Hockmeyer. W. T., Gross, M., Ballou, W. R., Wirtz, R. A., Trosper, J. H., Beaudoin, R. L., Hollingdale, M. R., Miller, L. H., Diggs, C. L., and Rosenberg, M. (1985). Expression of Plasmodium falciparum circumsporozoite proteins in Escherichia colifor potential use in a human malaria vaccine. Science 228, 958. 162. RUSSO,D. Sundy, J., and Weidanz, W. (1988). Identification and characterization of T-epitopes residing within recombinant and synthetic peptides derived from the circumsporozoiteprotein of I!falciparum. In “Modem Approaches to New Vaccines” (R. M. Chanock and R. A. Lerner, eds), p. 37. Cold Spring Harbor Lab., Cold Spring Harbor, New York. 163. Archibald, H. M., and Bruce-Chwatt, L. J. (1956).Suppression of malaria with pyrimethamine in Nigerian schoolchildren. Bull. W.H.O. 15, 755. 164. Marsh, K.,Hayes, R. H., Carson, D. C., Otoo, L., Shenton, F., Byass, P., Zavala, F., and Greenwood, B. M. (1988). Anti-sporozoite antibodies and immunity to malaria in a rural Gambian population. Tmns R . SOC. Dop. Med. Hyg. 82, 532. 165. Hoffman, S. L., Oster, C. N., Plowe, C. V., Woollett, G. R., Chulay, J. D., Wirtz, R. A,, Hollingdale, M. R., and Mugambi, M. (1987). Naturally acquired antibodies to sporozoites do not prevent malaria: Vaccine development implications. Science 237, 639. 166. Druilhe, P., Pradier, O., Marc, J. P., Miltgen, F., Mazier, D., and Parent, G. (1986). Levels of antibodies to Plasmodium falciparum surface antigens reflect malaria transmission rates and are persistent in the absence of re-infections. Infect. Immun. 53, 393. 167. Bomford, R.,and Wedderbum, N. (1973). Depression of immune response to Moloney leukaemia by malarial infection. Nature (London) 242, 471. 168. Orjih, A. U.,and Nussenzweig, R. S. (1979). Plasmodium berghei: Suppression of antibody response to sporozoite stage by acute blood stage infection. Clin. Exp. Immunol. 38, 1. 169. Romero, P.,Heimer, E. P., Herrera, S., Felix, A. M., Nussenzweig, R. S., and Zavala, E (1987). Antigenic analysis of the repeat domain of the circumsporozoite protein of Plasmodium viwx. J. Immunol. 139, 1679. 170. Ban; P. I., Gibson, H. L., Enea, V., Amot, D. E., Hollingdale, M. R., and Nussenzweig, V. (1987). Expression in yeast of a Plasmodium wiuax antigen of
334
VICTOR NUSSENZWEIG AND RUTH S. NUSSENZWEIG
potential use in a human malaria vaccine. J. Ex$. Med. 165, 1160. 171. Nardin, E. H., Barr, P. J . , Gibson, H . L . , Collins, W. E . , Nussenzweig, R. S.. and Nussenzweig, V. (1987). Induction of sporozoite-specificmemory cells in mice
immunized with a recombinant Plasmodium vivax circumsporozoite protein. Euz J. Immunol. 17, 1763. 172. Jacobs, W. R., Jr., Tuckman, M., and Bloom, B. R. (1987). Introduction of foreign DNA into mycobacteria using a shuttle plasmid. Nature (London) 327. 532. 173. Moss, B., and Flexner, C. (1987). Vaccinia virus expression vectors. Annu. Rev. Immunol. 5 , 305. 174. Kumar, S., Miller, L. H., Quakyi, I. A., Keister, D. B., Houghten, R. A . , Maloy, W. L . , Moss, B., Berzofsky, J. A., and Good, M. E (1988). CytotoxicT-cells specific for the circumsporozoite protein of Plasmodium falcipawm. Nature (London) 334, 258. 175. Lowell, G. H., Ballou, W. R., Smith, L. F., Wirtz, R . A . , Zollinger, W. D., and Hockmeyer, W. T.(1988). Proteosome-lipopeptide vaccines: Enhancement of immunogenicity for malaria CS peptides. Science 240, 800. 176. Good, M. F., Pombo, D., Lunde, M. N., Maloy, W. L., Halenbeck, R., Koths, K., Miller, L. H., and Berzofsky, J. A. (1988). Recombinant human IL-2 overcomes genetic non-responsiveness to malaria sporozoites.J. Immunol. 141, 972. 177. Tam, J. P. (1988). Synthetic peptide vaccine design: Synthesis and properties of a high-density multiple antigenic peptide system. PTOC.Natl. Acad. Sci. U S A . 85, 5409. 178. Satterthwaite,A. C., Arrhenius, T., Hagopian, R. A., Zavala, F., Nussenzweig, V., and Lerner, R. A. (1988). Conformational restriction of peptidyl immunogens with covalent placements for the hydrogen bond. Vaccine 6 , 99.
ADVANCES IN IMMUNOLQGY, VOL. 45
Virus-Induced Immunosuppression: Infections with Measles Virus and Human Immunodeficiency Virus MICHAEL B. McCHESNEY A N D MICHAEL B. A. OLDSTONE Department of Immunology, kripps Clinic and Research Foundatlon, b Jolla, California 92037
1. Introduction
The literature on virus-induced immunosuppression has grown and expanded in scope greatly in the last 45 years (reviewed by Mims, 1986; Rouse and Horohov, 1986; McChesney and Oldstone, 1987). But the current epidemic of a newly described disease, the acquired immunodeficiency syndrome (AIDS) (Gottleib et al., 1981), has accelerated research in this subject in an inevitable and unprecedented way. Historically, measles virus, and now human immunodeficiency virus (HIV), can be regarded as infectious agents that perturb immunologic functions of the host in a critical way, permitting opportunistic infections to confer the distinction of fatality to these relatively noncytopathic viruses. The phenomenon of virus-induced immunosuppression was initially demonstrated by von Pirquet (1908), with his study of tuberculosis immunity during measles virus infection. Prior to this era, a concept of “opportunistic infection” can be identified in the speculation of an eighteenth century physician (William Heberden, 1802), in the conclusion of his chapter on measles in “Commentaries on the History and Cure of Diseases.” The lungs are sometimes so injured by this distemper, that a lasting cough succeeds; and sometimes a pulmonary consumption.Weak eyes, inflamed eyelids, glandular tumors, and many other scrofulous appearanceshave followed the measles; whether they were formed by them, or, the seeds being before in the constitution, were only excited by this distemper; or possibly the appearance of scrofulous symptoms was wholly owing to other causes, and would have come on at this time though there had been no measles [p. 3131.
This temporal association of two well-defined diseases, measles and tuberculosis, could allow nothing more than an empirical speculation in Heberden’s era, when the germ theory of disease had not yet supplanted popular ideas of spontaneous generation and the host immune response had not become an object of scientific inquiry (Silverstein and Bialasiewicz, 1980). A similar temporal association has been observed 335 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
336
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
recently for AIDS and tuberculosis (Stoneburner et al., 1987). The term, opportunistic infection, was generated with the recognition that a variety of ubiquitous infectious agents of low virulence could cause fatal disease in an immunocompromised host. This concept has been useful in the context of the AIDS literature, and can be expanded to include virusassociated neoplasms such as Kaposi’s sarcoma and certain B cell lymp hom as. Do infections with these viruses, which are from different taxonomic classes, have anything in common? The epidemiology of both virus infections is by now similar. Infection with measles virus (Assaad, 1983) and HIV (Piot et al., 1988) is endemic worldwide, and sporadic epidemics associated with radical sociologic change are common to both (Aaby, 1988; Nzilambi et al., 1988). The high morbidity and mortality due to measles virus infection has not changed significantly since the time of Rhazes’ discourse (850) except in the developed countries. Soon HIV will likely compete with measles virus as a major cause of childhood mortality (Piot et al., 1988). Vaccination strategy is effective in reducing the mortality due to measles virus infection (Halsey et al., 1985; Aaby et al., 1988), and will have to contend with HIV. Biologically, although the route and mode of infection are different, both viruses are tropic for cells of the immune system, and some aspects of the virus-lymphocyte interaction may be similar (Fig. 1). The immunologic dysfunction due to measles virus develops soon after infection and is transient as long as there is no prior T lymphocyte dysfunction (Burnet, 1968). The host immune response effectively limits virus replication. In contrast, immunosuppression or immunodeficiency during HIV infection develops insidiously and most often does not manifest clinically for several years, during which a persistent infection successfully evades the host immune response. Knowledge about the biology and molecular biology of both viruses has increased greatly in recent years. In this review, we focus on aspects that elucidate possible mechanisms of immunosuppression. Further, with HIV infection, we concentrate on the biology of primary and persistent infection before the progression to AIDS. II. Measles Virus
A. VIROLOGY An enveloped RNA virus, measles virus is a member of the Morbillivirw genus (family Paramyxoviridae). Man is the only known natural host, although the virus can infect and produce disease in some species
337
VIRUS-INDUCED IMMUNOSUPPRESSION
Natural Cytoxity
Proliferation
Ig Production
NIL 120
E
Measles .c e Virus
601
HIV
l 60 Z01 0 24
48
Hours
72
48
72
Hours
96
Days
FIG. 1. Lymphocyte functions in vitro are altered by infection of lymphocytes with measles virus or HIV. (me.), Spontaneous Ig secretion by lymphocytes cultured with inactivated virus preparations.
of Old World primates (Blake and Trask, 1921a,b,c). Taxonomically related viruses, canine distemper virus and rinderpest virus, are immunosuppressive in their hosts. Measles virus was originally isolated and characterized by Enders and Peebles (1954), who described cytopathic effects in human and monkey kidney cells consisting of intranuclear inclusions and multinucleated giant cell formation, similar to the cellular syncytia seen in lymphoid and epithelial tissues during measles virus infection in uivo (Warthin, 1931). Viral nucleocapsids are present in both the nucleus and cytoplasm of infected cells. Viral replication occurs in the cytoplasm and mature virions bud from the plasma membrane. As yet uncharacterized host cell nuclear and cytoplasmic factors are required for viral replication (Fraser and Martin, 1978). Measles virus infects a wide range of epithelial and lymphoblastoid cell lines in vitro. The cell membrane receptor for measles virus is a nonsialated oligosaccharide present on many cells (Krah and Choppin, 1988). Productive infection of cells by measles virus is not lytic and constitutive cellular protein synthesis is not altered (Haspel et al., 1977). The major cytopathic effect, syncytial cell formation, is a consequence of viral replication and insertion of the viral fusion protein into the cellular plasma membrane during maturation and budding of virions. But syncytium formation is not required for viral replication: a synthetic peptide that blocks fusion
338
MICHAEL B. MCCHESNEY AND MICHAEL B. A . OLDSTONE
(Norrby, 1971;Richardson and Choppin, 1983)and antibody to the fusion protein of the virus (Graves et a l . , 1978) inhibit syncytium formation without inhibiting efficient virus replication. B. IMMUNOSUPPRESSION 1. Measles V i m Infection and Tuberculosis Immunity Active tuberculosis infection is a common sequela of measles virus infection (Heberden, 1802; Osler, 1892;Holt, 1897). Following the discovery of the tubercle bacillus, the first scientific report of virus-induced immunosuppression was von Pirquet’s (1908)observation that the tuberculin skin test response of immune individuals is depressed during the time course of acute measles virus infection: from a day or two prior to onset of rash and for 7 to 20 days following the appearance of rash. This transient depression of the tuberculin skin test response occurred in every patient examined, although clinically active tuberculosis would not always follow. Von Pirquet also noted that clinical cases of nephritis would go into remission following measles infection. This observation was extended with the therapeutic use of measles infection for the treatment of nephrotic syndrome (Blumberg and Cassady, 1947)prior to the discovery of steroids. The von Pirquet phenomenon, transient depression of tuberculosis immunity by acute measles virus infection, was extended in an in vitro system by Smithwick and Berkovich (1969).They showed that the blastogenic response of lymphocytes from the peripheral blood of tuberculosis-immune individuals to tuberculin antigen is significantly reduced during acute measles virus infection. Further, they demonstrated that the phenomenon was due directly to the virus, rather than being a nonspecific effect of acute infection: the blastogenic response to tuberculin by immune lymphocytes is reduced by infection in vitro with measles virus. Subsequently it was shown that the lymphocyte proliferative response to the mitogen phytohemagglutinin (PHA), measured by [3H]thymidine uptake, is suppressed by infection in vitro of peripheral blood mononuclear cells (PBMCs) by measles virus. The suppression by measles virus of the proliferative response to mitogen occurs with lymphocytes from both nonimmune and measles-immune donors, but serum containing measles antibody abrogates the suppression (Zweiman, 1971).
2. Measles Virus Is Lymphotropic
A corollary of the suppression of lymphocyte functions by measles virus is that the virus can productively infect lymphocytes. During acute measles infection, followingprimary replication in the respiratory tract, the virus
339
VIRUS-INDUCED IMMUNOSUPPRESSION
disseminates throughout the body and secondary replication occurs in lymphoid tissue (Yamanouchi et al., 1973). Close to the onset of rash, there is a viremia during which infectious virus can be recovered from the leukocytes but not from the plasma of peripheral blood (Gresser and Chany, 1963). Measles virus antigen was found in mitogen-stimulated T and B lymphocytes from the peripheral blood of patients with measles (Osunkoya et al., 1974; Whittle et al., 1978). Human and monkey leukocytes support measles virus replication in Vitro (Berg and Rosenthal, 1961). Monocytes and T helper, T suppressor, and B lymphocytes can be infected in Vitro, but polymorphonuclear cells cannot (Joseph et al., 1975; Sullivan et al., 1975; Huddlestone et al. , 1980). Viral replication is restricted (i.e., nonreplicative) in unstimulated cells, but lymphocytes stimulated by a mitogen either before or after infection replicate virus, and nearly all cells contain viral antigens (Joseph et al., 1975; Lucas et al., 1978; Huddlestone et al., 1980). Infected and stimulated CD4+ T lymphocytes support higher titer viral replication than do CD8 cells, and this was inversely correlated with the amount of interferon-a secreted by infected CD4+ and CD8+ cells (Jacobson and McFarland, 1984). Both T and B lymphoblastoid cell lines can be infected by measles virus and persistent infection is easily established (Barry et al., 1976). The establishment of persistent infection in lymphoblastoid cell lines is related to the ability of cells to cleave the fusion glycoprotein of measles virus from its precursor Fo, to the active F1 protein, as is required for viruscell fusion (Fujinami and Oldstone, 1981). +
3. Suppression of Lymphocyte Functions by Measles Virus
Both cellular and humoral immune responses are altered during acute measles virus infection. The time course of suppression of the tuberculin skin response and lymphocyte blastogenesis to tuberculin in tuberculosis-immune donors during natural measles infection or after live virus vaccination is seen in Table I. Suppression of the responses is not seen until 7 days following immunization and recovery occurs within 2-3 weeks (Smithwick and Berkovich, 1969). The delayed skin hypersensitivity reaction to dinitrochlorobenzene and to recall antigens is impaired for 4-6 weeks following measles infection or immunization with live virus (Fireman et al., 1969; Wesley et al., 1978). Immunization with an inactivated measles virus preparation does not alter immune responses to other antigens (Fireman et al., 1969). Although attenuated live measles virus vaccine suppresses immune responses in vivo, vaccination does not result in clinical immunosuppression even in malnourished infants (Halsey et al., 1985). This is probably due to the fact that attenuated strains of measles virus do not replicate in lymphoid tissue in vivo as extensively
340
MICHAEL B. MCCHESNEY AND MICHAEL B. A . OLDSTONE
SKIN AND
TABLE I LYMPHOCYTE RESPONSETO PPD IN T W O INFECTED WITH MEASLESVIRUS"
CHILDREN
Response to PPD Event
Postvaccination (days)
Skin (mm)
4 7 9 11 16 24
11 0 15 15 15 8 ?+ 0 0 8
Natural measles 2 weeks before Second day of rash 10 days later Prior to vaccination for measles
Lymphocytes
(% mitotic cells) 10.1 0.6 6.2 7.6 8.2 3.0 0.4 1 .o 0.6 10.4
"From Smithwick and Berkovich (1969).
as do wild-type strains (Yamanouchi et al., 1970). During measles virus infection antibody responses to typhoid vaccination are depressed, but established antibody titers are not markedly altered (Whittle et al., 1973; Coovadia et al., 1974). Polyclonal B cell activation, a component of humoral immunosuppression during many acute viral infections, is also seen during measles infection (Arneborn et al., 1983). Pokeweed mitogen (PWM)-driven immunoglobulin (Ig) secretion in Vitro is impaired during measles virus infection (Joffe and Rabson, 1981; Arneborn et al., 1983). We recently had the opportunity to extend this observation during a natural epidemic of measles virus infection at a primate research center (McChesney et a l . , 1989). During this outbreak there was a high case/fatality ratio associated with opportunistic infections. Measles virus was detected in the PBMCs of acutely infected rhesus monkeys, and as shown in Fig. 2, Ig secretion in PWM-stimulated cultures of B and T lymphocytes from infected monkeys was markedly reduced compared with uninfected controls. PWM-stimulated Ig secretion was still reduced at 6 months, but was in the normal range at 1 y with PBMCs from two convalescent monkeys. Alteration of a variety of lymphocyte functions is seen in measles virus infection of lymphocytes in Vitro (reviewed in McChesney and Oldstone, 1987). In general, measles virus suppresses lymphocyte functions that
VIRUS-INDUCED IMMUNOSUPPRESSION
341
T Normal
; Measles-
Normal
-Measles-
FIG. 2. Ig secretion by B and T lymphocytes from monkeys infected with measles virus. Open (unshaded) histograms, unstimulated cultures; shaded histograms, PWMstimulated cultures. Reprinted from McChesney et d.(1988).
require terminal differentiation, but already differentiated functions are not altered (Galama et al., 1980). For example, proliferation and generation of cytotoxic T lymphocytes (CTLs) in an allogeneic mixed lymphocyte reaction (MLR) are suppressed by measles virus, but infection of 3- or 5-day-oldMLR cultures does not affect the lytic activity of allogeneic CTLs. Lymphocytes infected with measles virus in vitro cannot secrete Ig in PWM culture, but if the virus is added to PWM cultures after day 3, less suppression of Ig secretion is observed (Galama et al., 1980; Casali et al., 1984) (Fig. 3). This temporal kinetics of suppression of Ig secretion by measles virus in vitro is reminiscent of the dichotomy of radiosensitive and radioresistant phases of the antibody response in vivo (Dixon et al., 1952; Taliaferro et al., 1952). Both phenomena suggest a requirement for cell proliferation, during which the cell is vulnerable to radiation or measles virus suppression, prior to differentiation into Ig-secreting cells. PWM cultures are a complex multicellular system with requirements for helper factors from monocytes and T cells to drive B cell differentiation, as well as for down regulation of Ig secretion by suppressor T cells and natural killer (NK) cells. By culturing isolated B lymphocytes in T cell/monocyte-conditionedmedia stimulated with PWM (Hirano et a l . , 1977), the effect of virus infection of T or B lymphocytes in this system can be seen (Fig. 4; McChesney et al., 1986). Measles virus-infected B lymphocytes are unable to proliferate or secrete Ig when cultured in
342
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
- ’ 0
0 1 2 3 4 5 6 7 Measles Virus Added at Day
FIG. 3. Suppression of PWM-stimulated Ig secretion by measles virus infection of B and T lymphocytes in vitro. All cultures were assayed for Ig on day 7. Hatched area indicates the mean f SD Ig production by uninfected cultures. Reprinted from Casali et al. (1984).
conditioned media containing sufficient growth and differentiation factors. However, measles virus-infected T cells and monocytes are able to produce B cell growth and differentiation factors. Is there a final common pathway for the multiple effects of measles virus on lymphocyte functions in vitro? Extending the observations described above that measles virus infection suppresses proliferation of PBMCs, when relatively pure populations of T lymphocytes or B lymphocytes are infected with measles virus, mitogen-stimulated proliferation of either T or B lymphocytes also is suppressed (McChesney et a l . , 1987, 1988). The suppression of T cell proliferation by measles virus is not reversed with exogenous interleukin (IL-2). There is no discrepancy between the two observations that (1) measles virus-infected T lymphocytes can secrete B cell growth and differentiation factors, and (2) infected T lymphocytes cannot proliferate in response to mitogens. The helper T cell function in PWM culture is resistant to y-radiation (Keightley et al., 1976), i.e., the capacity of T lymphocytes to secrete growth factors after mitogenic stimulation is not dependent on the capacity to proliferate. Thus the proximate effect of measles virus infection on T and B lymphocytes is the same, and the final effect on differentiated lymphocyte functions is determined by the requirement for proliferation prior to differentiation. The mechanisms(s) by which measles v i r u s suppresses lymphocyte proliferation is not known. Viral replication in lymphocytes is required:
343
VIRUS-INDUCED IMMUNOSUPPRESSION T LYMPHOCYTE CONDITIONED MEDIA
IgG nglml ~ 6 0
760
IgM ng/ml
ooo r
Medium Control
T + MO, no PWM
T + MO+ PWM
T-MV+MB+PWM
P P P
260
,
600
,
150
,
1.000 1 2 6 0 1.500
,
,
,
P P P
F F
FIG. 4. Ig secretion by B lymphocytes cultured in T cell/monocyte-conditioned media. The light, unshaded areas represent uninfected B cells; the darkened areas represent B cells infected with measles virus; TMV, T cells infected with measles virus; PWM, pokeweed mitogen; M@, macrophages. Reprinted from McChesney et al. (1986).
inactivation of virus by heat, UV light, or neutralizing antibody (Lucas et al., 1977; Casali et al., 1984), and culture at a temperature nonpermissive for viral replication (Lucas et a l . , 1977), abrogate the suppression. Yet productive infection of lymphocytes is not lytic (Lucas et al., 1977, 1978; Casali et al., 1984; McChesney et al., 1986). The formation of cellular syncytia in infected lymphocyte suspension cultures is not widespread as it is in measles virus infection of adherent monolayer cultures (Enders and Peebles, 1954) or in some lymphoblastoid cell lines (Barry et al., 1976; Fujinami and Oldstone, 1981). Attempts to demonstrate suppressor T cells or monocytes both in vivo (Arneborn et al., 1983) and in Vitro (Sanchez-Lanier, 1988)have been unsuccessful. Infected lymphocyte culture supernatants are variably suppressive of the proliferation of uninfected cells. As yet no soluble suppressive factor has been identified (Minagawa et a l . , 1974), and interferon (IFN-a) and prostaglandin E have been excluded as mediators of suppression in this cell-virus system (Lucas et al., 1978; Sanchez-Lanier, 1988). These results together with the data mentioned above (that measles virus suppresses the proliferation of relatively pure populations of T and B lymphocytes) may suggest
344
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
that an autocrine mechanism in the infected cell is aborted (or activated) during viral replication, thus inhibiting lymphocyte proliferation. 4 . Effect of Measles Virus on T and B Lymphocyte Activation The suppression of mitogen-stimulated cell proliferation by measles virus suggests a paradox, because the virus cannot replicate in lymphocytes unless stimulated by a mitogen. Neither a DNA virus nor a retrovirus, measles virus does not require host cell DNA synthesis for its replication, but it does require cellular factors that are not present in the metabolically quiescent, unstimulated, or Go lymphocyte. The other side of this paradox poses a question: what effect does measles virus have on the events of cell activation in the infected lymphocyte or at what stage in the cell cycle does virus-induced suppression occur? Measurable events of B cell activation were examined in B cells infected with measles virus (McChesney et al., 1987). Measles virus-infected small, Go B lymphocytes, when stimulated by mitogen, undergo cell volume increase at 24 and 48 h after stimulation, as do uninfected B cells. The expression of cell surface antigens 4F2, HLA-DS, and transferrin receptor are the same in infected and uninfected cells at 24 to 72 h after stimulation. Total RNA synthesis as measured by [3H]uridine uptake by infected and uninfected cells is increased similarly at 24 h after mitogenic stimulation, but is decreased by 20-50% relative to uninfected cells at 48 h. This was correlated with a fivefold reduction in the steady-state level of mRNA for histone 2B in infected cells at 48 h after stimulation. Transcription of the histone 2B group of genes is normally upregulated during the S phase of the cell cycle (Schumperli, 1986). Efficient measles virus replication occurs in mitogen-stimulated, infected cells during the time course of cellular activation. Likewise, events of T cell activation were measured in measles virusinfected T lymphocytes (McChesney et al., 1988). There is no significant difference in expression of cell surface antigen 4F2, Tac (IL-2 receptor), transferrin receptor, or HLA-DR after stimulation of infected and uninfected T lymphocytes. The capacity of infected cells to secrete IL-2 and IFN-y after stimulation is not reduced. But total RNA synthesis in infected T lymphocytes is reduced by about 50% compared with uninfected cells at 72 h after stimulation. Again, production of infectious virus occurs in infected, stimulated T lymphocytes during the time course of cellular activation. Thus both T and B lymphocytes infected with measles virus can enter the cell cycle when stimulated with mitogens, but the cells are arrested in late GI(Fig. 5 ) . Measles virus infection of unstimulated lymphocytes results in secretion of IFN-(r (Lucas et a l . , 1978; Sanchez-Lanier, 1988),
345
VIRUS-INDUCED IMMUNOSUPPRESSION ACTIVATION
..ly-Gll~,~-S+G7
Cell Cycle G,-GI
B Cell
PROLIFERATION
@
-@-
DIFFERENTIATION
LJ
BCGF
1
- - -- - - - - +
T I
t
b
a@; 0 ,
Ig-Secreting Celts
MEASLES VIRUS REPLICATION
T Cell
CTL
@-@---------+ 11.2
FIG.5. Measles virus infection of B or T lymphocytes results in cell cycle arrest following mitogenic stimulation. This arrest cannot be overcome by exogenous growth factors; BCGF, B cell growth factor; IL-2, interleukin-2;CTL, cytotoxic T lymphocyte.
but does not produce any of the changes of lymphocyte activation described above. There is a restriction point in the normal lymphocyte cell cycle in late GI that is growth factor dependent (Howard and Paul, 1983; Melchers and Lernhardt, 1985). In contrast, measles virus infection results in growth factor-independent cell cycle arrest in late G , . 5. The Immune Response to Measles Virus Immunity to measles virus after natural infection is long-lived and complete; second infections resulting in clinical disease have not been reported. Antibodies specific for measles virus are detectable during acute infection by the time of onset of rash (Graves et al., 1984), and neutralizing antibody titers remain elevated for life in most individuals (Norrby, 1986). Effective lysis of measles virus-infected cells in vitro occurs both by antibody plus complement (Sissons and Oldstone, 1980) and by antibody-dependent cytotoxicity (ADCC) (Kreth and Wiegand, 1977; Perrin et al., 1977). During acute measles virus infection, low levels of a virus-specific CTL response could be detected in some individuals, although there was a high background of nonspecific lysis in assays of primary viral CTL activity (Lucas et al., 1982; Sissons et al., 1985). An MHC class II-restricted CTL response, first reported from in vitro studies
346
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
using T cell clones generated from a patient with multiple sclerosis (Jacobson et al., 1984, 1985a), can be detected in measles-immune subjects (Jacobson et al., 1985b). The relative roles in protection in vivo of virusspecific CTL, delayed hypersensitivity,or other T lymphocyte-dependent mechanisms have not been determined. 6. Immunosuppression during Persistent Measles Virus Infection
The onset of lymphocyte dysfunction in mvo following measles infection or immunization (Table I) demonstrates an “incubation period” coincident with the interval after primary viral replication at the site of infection and dissemination of measles virus to lymphoid tissues (Yamanouchi et al., 1970, 1973). If test antigens are administered at the same time as measles infection, antibody responses to both T-dependent and T-indepenent antigens are not impaired (Hicks et al., 1977). But if antibody production in PWM culture is measured at the time of rash and viremia, marked suppression is seen (Fig. 2). In rare cases following acute infection, measles virus can establish a slowly progressive infection in the brain leading to fatal neurological disease, subacute sclerosing panencephalitis (SSPE) (reviewed by ter Meulen et al., 1983). Neurons and glial cells contain large accumulations of viral nucleocapsids. There is a perivascular accumulation of mononuclear cells, abundant plasma cells, but not marked infiltration of parenchyma with inflammatory cells. A large percentage of PBMCs from patients with SSPE contain measles virus RNA detected by in situ hybridization (Fournier et al., 1985), although it has proved difficult to isolate measles virus from brain or lymphocytes (reviewed by ter Meulen and Carter, 1984). Clinical immunosuppression is not a feature of persistent measles virus infection, although subtle lymphocyte dysfunction has been noted (ter Meulen and Carter, 1984). Patients with SSPE have elevated titers of antibody to measles virus antigens in serum and cerebrospinal fluid, and ADCC function of PBMCs is normal (Kreth and Wiegand, 1977). It was recently reported that the measles virusspecific, MHC class II-restricted CTL response of patients with SSPE is diminished (Dhib-Jalbut et al., 1988). This would suggest a pattern of viral antigen-specificimmunosuppression in SSPE restricted to a component of the host immune response that is required for viral clearance, as in the pathogenesis of persistent murine lymphocytic choriomeningitis virus infection (reviewed by McChesney and Oldstone, 1987). Interestingly, measles immune sera and monoclonal antibodies to the hemagglutinin protein of measles virus not only modulate the external membrane glycoproteins of the virus from infected cells, but also down regulate the expression of internal structural proteins of the virus inside the cell
VIRUS-INDUCED IMMUNOSUPPRESSION
347
(Fujinami et al., 1984), a process termed antibody-induced antigenic modulation (reviewed by Oldstone and Fujinami, 1982). This process protects infected cells from immune recognition by removal of cell surface viral antigens, and promotes the selection of a replication-defective variant of the virus. Animal experiments demonstrate that antibodyinduced modulation of the measles virus hemagglutination occurs in Vivo and is associated with establishment and maintenance of persistence. 111. Human Immunodeficiency Virus and Immunosuppression
A. VIROLOGY First isolated in 1983 (Barre-Sinoussi et al., 1983), HIV and related viruses of man and other animals (Table 11) have been recently cloned and their genomes sequenced. These viruses comprise a recently described taxonomic group of the family Retroviridae, the lentivirus group in the subfamily Lentivirinae (reviewed by Haase, 1986; Cheevers and McGuire, 1988; Weber and Weiss, 1988). The prototypic virus of TABLE I1
TAXONOMY OF LENTIVIRUSES~ Retroviridae Oncovirinae HTLV I and I1 Spumavirinae Foamy virus Lentivirinae Visna-maedi Caprine arthritis encephalitis virus Equine infectious anemia virUS Human immunodeficiency virus HIV-1 HIV-2 Simian immunodeficiency virus (also called STLV 111) Feline T lymphotropic virus Bovine immunodeficiencv virus
Host Man Man Sheep Goat Horse Man Man Old World monkey Cat cow
‘Referenced in publications: Sonigo et al. (1985); Clements et al. (1988); Wain-Hobson et al. (1985); Guyader et al. (1987); Hirsch et al. (1987); Fukasawa et al. (1988); Pedersen et al. (1987); Gonda et al. (1987).
348
MICHAEL B. MCCHESNEY AND MICHAEL B. A . OLDSTONE
this family, visna, produces a chronic disease in sheep after a relatively long incubation period, and is thus designated a slow virus infection by Sigurdsson (1954). Lentiviruses are nononcogenic retroviruses that produce cytopathic effects (multinucleated giant cells or syncytia) in tissue culture. The viral life cycle in tissue culture has two stages: (1) a chronic, nonproductive infection of quiescent, growth-arrested cells maintained by reverse transcription of the viral genome into double-stranded DNA, and (2) a productive infection of permissive cells with complete transcription of the provirus genome and replication of infectious virus. Unlike the oncogenic retroviruses, lentivirus replication may not require integration of the proviral genome into host cell DNA (reviewed by Haase, 1986; Cheevers and McGuire, 1988). During chronic HIV infection in mvo and in vitro, proviral DNA exists in both integrated and nonintegrated forms (Shaw et al., 1984, 1985). Also different from the oncogenic retroviruses, the lentivirus genome encodes a number of regulatory proteins (Fig. 6) that act on viral RNA transcription, splicing, and translation in ways that are not yet well understood (reviewed by Chen, 1986; Weber and Weiss, 1988). In marked contrast to the efficient replication of virus by a large percentage of infected cells in vitro, lentivirus replication in m'vo is characterized by a slowly progressive accumulation of viral genomes with very little replicating virus detected at any time in a small number of cells (reviewed by Haase, 1986), even in immunosuppressed animals
0
1
2
3
I
4
I
5
I
6 1
7
I
8
1
9
I
kb
FIG. 6. Genomic structure and proteins of HIV. The long terminal repeats (LTRs), gag,pol, and enu genes are typical of oncogenic retroviruses, but the shaded bars indicate viral regulatory genes not found in oncogenic retroviruses: RT, reverse transcriptase. Modified from Weber and Weiss. 1988.
VIRUS-INDUCED IMMUNOSUPPRESSION
349
(Nathanson et al., 1976; Narayan et al., 1977). The lentivirus infections of ungulates, e.g., visna in sheep, caprine arthritis encephalitis virus (CAEV) (Narayan and Cork, 1985; Nathanson et al., 1985), and equine infectious anemia virus (EAIV) (Cheevers and McGuire, 1985), are characterized by restricted cell tropism for monocyte-macrophages, asymptomatic infection or immunopathologic disease of lung, brain, joints, and immune complex disease. In contrast, infections with HIV-1, HIV-2 (Clavel et al., 1987), and simian immunodeficiency virus (SIV) (Daniel et al., 1987; Benveniste et al., 1988) are characterized by tropism for CD4+ lymphocytes and monocytes, and by asymptomatic infection or disease due to immunosuppression and encephalopathy (Table 11). Unlike the other lentiviruses and most retroviruses, a cellular receptor for HIV has been well characterized (reviewed by Sattentau and Weiss, 1988). Shortly after the initial isolation of HIV it was first reported that the virus replicates selectively in CD4+ T lymphocytes both in vivo and in Vitru (Klatzmann et al., 1984a). Infection of lymphocytes in vitro as detected by syncytium formation (Dalgleish et al., 1984) or viral replication (Klatzmann et al., 1984b) can be blocked specifically by some monoclonal antibodies to the CD4 cell surface molecule. A secreted, soluble form of CD4 blocks HIV infection in vitro, and the binding of CD4 to the gp120 envelope protein of HIV in solution has a dissociation constant of l o p 9M (Smith et al., 1987). Concentrations of soluble CD4 protein that efficiently block infection of susceptible T lymphocytes by HIV do not inhibit the MHC class II-restricted functions of CD4+ T cell clones (Hussey et al. 1988). This suggests that the binding affinity of CD4 to the gp120 envelope protein is greater than the CD4-MHC class I1 interaction. After binding to the CD4 molecule on cells, HIV is probably not internalized by a process of receptor-mediated endocytosis (Maddon et al., 1986; Hoxie et al. 1988). Following virus-receptor binding, enveloped viruses gain entry to cells by a process of fusion of the viral envelope and the cell membrane (reviewed by Kohn, 1985). In this regard, viruscell fusion and cell-cell fusion leading to multinucleated giant cell formation by HIV (Lifson et al., 1986a) may occur in a manner similar to that of the classical syncytium-forming paramyxoviruses. The viral determinant(s) for fusion by HIV is encoded in the env gene (Lifson et al., 198613). The env gene product of HIV is a glycosylated protein, gp160, which is endoproteolytically cleaved to yield a transmembrane protein, gp41, and the other spike protein gp120 (Montagnier et al., 1985). Regions within gp120 and gp41 involved in CD4 binding, cell fusion, and the noncovalent interaction of gp120 and gp41 have been defined by insertion and deletion mutants of an infectious cDNA clone
350
MICHAEL B. MCCHESNEY AND MICHAEL B. A . OLDSTONE
(Kowalski et al., 1987). The amino-terminal end of gp41 contains a hydrophobic amino acid sequence that is similar to the fusion peptide sequences of the paramyxoviruses, respiratory syncytial virus, and measles virus (Gallagher, 1987). Site-directed mutagenesis of a cloned HIV provirus, replacing the normal trypsin-sensitive cleavage site in gp160 that yields gp41 and gp120 with a chymotrypsin-sensitivesite, resulted in production of noninfectious virus that contained noncleaved gp160 (McCune et al., 1988). The defective virions, morphologically similar to infectious virions, became infectious and induced syncytia after chymotrypsin treatment. Thus cleavage of the glycoprotein precursor during virus assembly is required to activate a fusion site in the amino-terminal end of gp41 for virus infectivity and fusion, a process similar to that described for measles virus (Fujinami and Oldstone, 1981). Cleavage activation of precursor viral glycoproteins by host cell enzymes is a major determinant of pathogenesis for paramyxoviruses and myxoviruses (reviewed by Scheid and Choppin, 1984). Multinucleated giant cells of probable lymphocyte origin can be found in lymph nodes (reviewed by Racz et al., 1986); those of monocyte origin can be found in brain (Wiley et al., 1986) and lung (Salahuddin et a l . , 1986) from patients infected with HIV. Syncytium formation is a possible consequence of HIV replication in cells, but it is not required for efficient viral replication (Tersmette et al., 1988). Persistent, productive infection of human T cell lines, in spite of syncytium formation, allowed high-efficiency propagation of HIV and adaptation to tissue culture (Popovic et al., 1984). Both viral and cellular factors are required for syncytium formation during HIV infection. Different isolates of HIV vary widely in syncytium production in the same cell line (Dahl et al., 1987; Evans et al., 1987; Somasundaran and Robinson, 1987; Tersmette et a l . , 1988), and different cell lines, even of the same lineage, vary in susceptibility to syncytium formation by a particular strain of HIV (Kikukawa et al., 1986; Montefiori and Mitchell, 1987; Somasundaran and Robinson, 1987). Two experimental approaches suggest that the tropism of HIV for cells is restricted primarily at the level of receptor binding or virus fusion and entry into cells. 1. HIV is able to infect and replicate in cells that express the CD4 molecule; these cells are T lymphocytes and monocytes (discussed below), some B lymphoblastoid and promyelomonocytic cell lines (Levy et al., 1985), monkey PBMCs (Levy et al., 1985), HeLa cells transfected with a cDNA encoding CD4 (Maddon et al., 1986), central nervous system glial cells (Funke et al., 1987; Dewhurst et al., 1987), and human colon carcinoma cells (Adachi et a l . , 1987).
VIRUS-INDUCED IMMUNOSUPPRESSION
351
2. Transfection of infectious molecular clones of HIV into CD4cell lines demonstrate that both human and some nonhuman CD4cells are permissive (Adachi et al., 1986; Levy et al., 1986).
Murine cell lines transfected with a cDNA for human CD4 can express CD4 at the cell surface and bind HIV, but virus entry is blocked (Maddon et al., 1986). HIV infection of cells not currently known to express CD4 has been reported, including infection in Vitro of some Epstein-Barr virus (EBV)-transformedB cell lines (Montagnier et al., 1984; Salahuddin et al., 1987), infection of brain endothelial cells in vivo (Wiley et a l . , 1986), and infection of bone marrow endothelial cells by SIV in macaques (Benveniste et ul., 1988). A high degree of genetic variation between different isolates of HIV has been observed, due not only to point mutations but also to insertions and deletions, especially in the enu gene (Alizon et al., 1986). Similarly, a high degree of genetic variation occurs in the enu genes of the ungulate lentiviruses, but there are different patterns of genetic variation in visna virus and in EIAV (reviewed by Clements et al., 1988). Nucleotide sequencing of sequential isolates of HIV from one infected individual over a 2-y period reveals a significant degree of genetic variation, as do isolates from different infected individuals or from different geographic areas (Hahn et al., 1986). Variants of HIV with different cell tropism in vilro were isolated simultaneously from different tissues of AIDS patients (Gartner et al., 1986; Koyanagi et al., 1987). In this case, the pattern of variation suggests a parallel evolution of mutant viruses from a common progenitor rather than sequential evolution (Hahn et al., 1986). This pattern has been seen with visna virus infection of sheep (Narayan and Cork, 1985) and CAEV infection of goats (Ellis et al. , 1987), in which antibody neutralization variants may coexist in the animal with the parental type (Narayan and Cork, 1985). Antigenic variation may be a mechanism for the maintenance of persistent infection by evasion of the host immune response, but it does not play a significant role in disease pathogenesis of visna and CAEV infections (Narayan and Cork, 1985; Nathanson et al., 1985); it may be a factor, however, in equine infectious anemia (Cheevers and McGuire, 1985; Payne et al., 1987). Overall, there are regions showing a remarkable pattern of variability as well as regions having conservation in the enu gene of HIV (reviewed by Coffin, 1986). bearing a resemblance to the immunoglobulin supergene family (Maddon et al., 1986; Lasky et al., 1987). The CD4 binding domain of gp120 is conserved between HIV-1 and HIV-2 and SIV (Fig. 7) (Lasky et al., 1987). The deduced fusion peptide sequence in the amino-terminal region of gp41 is also highly conserved (McCune et al., 1988). An immunodominant region responsible for antibody
352
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
N
200
100
0
'
r
&
k
300 "
400aa C
/\
CRITICAL DOMAIN FOR BlNOlNG TO C04
FIG. 7. Amino acid sequence diversity in the a120 molecule of HIV, determined by comparing the protein sequences of many independent isolates of the virus. Light (unshaded) regions indicate highly conserved sequences: dark areas indicate nonconsexved and shaded areas indicate moderately conserved sequences. Modified from Lasky et al. (1987).
production in HIV infections, located in gp41, is also highly conserved (Gnann et al., 1987b,c; Gnann and Oldstone, 1989; Table 111, p. 364). B. NATURAL HISTORY OF HIV INFECTION Acute infection is often subclinical, but it may be associated with an infectious mononucleosis syndrome, similar to primary infection with the Epstein-Barr virus or human cytomegalovirus(Cooper et al., 1985; Ho et al., 1985). An initial phase of lymphopenia, with a proportionate decrease in both CD4+ and CD8+ lymphocytes, is followed by an increase in CD8+ cells and a return toward normal of CD4+ cells. This pattern of peripheral blood lymphocyte phenotypic alteration is seen in acute infections by herpes viruses (Reinherz et al., 1980). Months following seroconversion, lymphocyte counts stabilize, with a reduction of CD4 + lymphocytes (15-40% below seronegative controls) and an elevation of CD8+ lymphocytes (35-55% above controls) (Fahey et al., 1987; Cooper et al., 1988). HIV can be isolated from PBMCs (Ho et al., 1985) and plasma (Albert et al., 1987), and the viral p24 antigen is detectable in serum (Cooper et al., 1988; von Sydow et al., 1988) for up to 3 months following infection. However, detection of viremia is difficult at later times during the asymptomatic period of infection (see discussion below). An HIV-specific antibody can be detected 3-12 weeks following infection. Prior to the initiation of seroepidemiologic studies of HIV infection, it was realized that AIDS is an extreme in a spectrum of epidemic disease that includes chronic unexplained lymphadenopathy in high-risk patient groups (homosexual males, hemophiliacs, and parenteral drug abusers) (Gold et al., 1985). Following the discovery of HIV and the detection of specific antibody, an asymptomatic seropositive carrier state was identified (Gallo et al., 1984; Laurence et al., 1984). The incubation period of HIV infection prior to the diagnosis of AIDS can be long. Based on three prospective studies of HIV
VIRUS-INDUCED IMMUNOSUPPRESSION
353
seropositivity and AIDS in populations of homosexual males, it is estimated that from the date of seroconversion, 50% ’ of individuals will progress to AIDS in about 9 yr (Jaffe et al., 1985; Polk et a l . , 1987; Moss et al., 1988). The rate of progression to AIDS in a seropositive African heterosexual population may be similar (Mann et al., 1986). To date, no subpopulations of HIV-infected homosexual males with rapid progression or no progression to AIDS have been identified. It has been suggested that immunologic abnormalities associated with life-style may predispose high-risk groups to primary infection or progression to AIDS, but, at present, no significant immunologic abnormalities have been detected in HIV-seronegative homosexuals compared to seronegative heterosexuals (see discussion below) (Nicholson et al., 1985a; Dobozin et a l . , 1986). The incubation period of HIV infection in other patient populations is shorter, e.g., 3-4 months from birth in cases of perinatal infection (reviewed by Bernstein and Rubenstein, 1986) and 2 y or less in transfusion-associated AIDS (Jaffe et al., 1984). This suggests that among other factors, the route of infection affects the rate of disease progression. The altered immunoregulation of pregnancy in HIVseropositive women does not appear to accelerate disease progression (Scott et al., 1985; Johnstone et al., 1988). The immunologic hallmark of AIDS is a functional and phenotypic depletion of CD4+ lymphocytes (reviewed by Gluckman and Klatzmann, 1986). In patients with AIDS or at high risk for developing AIDS, the mean absolute number of CD4+ lymphocytes in peripheral blood is 250 cells/pl or less (Dobozin et al., 1986; Kaslow et al., 1987; Moss et al., 1988). There is a strong inverse correlation between the number of peripheral blood CD4+ cells and the duration of HIV infection or the risk of progression to AIDS (Goedert et al., 1987; Polk et a l . , 1987; Schwartz et al., 1985; Moss et al., 1988). As shown in Fig. 8, the decrease in peripheral blood CD4 + cell concentration during the incubation period of HIV infection in a population is progressive and almost linear with time. However, individual healthy HIV seropositives have variable rates of decline in CD4+ cells, with plateau intervals (Fahey et al., 1987). A relative depletion of CD4+ lymphocytes and an increase in CD8 cells is also seen in lymph nodes with follicular hyperplasia from HIV-seropositive individuals with lymphadenopathy (reviewed by Racz et al., 1986). Other immunologic factors associated with duration of HIV infection or progression to AIDS include increased peripheral blood CD8 cells, decreasing titer of HIV antibodies, elevated serum IgG or IgA (Polk et al., 1987), elevated serum &-microglobulin, and elevated serum level of viral p24 antigen (Moss et al., 1988). In the absence of any illness, lymphoadenopathy is not associated with disease progression +
+
354
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
FIG. 8. Median peripheral blood T4 lymphocyte counts at successive follow-up examinations in HIV-seropositive individuals who were not progressing to AIDS. Modified from Moss et al. (1988).
or declining CD4+ lymphocyte counts (Kaslow et al., 1987; Polk et al., 1987; Moss et al., 1988). Lymphadenopathy may resolve, but low CD4 cell counts do not return to normal (Gold et al., 1985; reviewed by Racz et al., 1986). Animals infected with visna or CAEV have persistent viremia and virus can be isolated from blood monocytes (Narayan and Cork, 1985). The investigation of viremia during HIV infection has been methodologically difficult, but virus or viral antigen is frequently detected in patients with AIDS, indicating that a persistent infection with HIV is required for disease progression. However, CD8 + cells from HIV-seropositive donors inhibit the replication of HIV (Walker et al., 1986; Tersmette et al., 1988; discussed below). In patients with AIDS, viral RNA can be detected in peripheral blood lymphocytes by in situ hybridization in 50% of PBMC samples and 86% of lymph nodes. Yet, very few cells (1 in 10+4-10+5) contain sufficient viral RNA for detection by this method (Harper et al., 1986). Southern blot analysis of cellular DNA from PBMCs, spleen, and lymph node of AIDS patients suggests that there are very low copy numbers of viral DNA in these tissues (Shaw et al., 1984). A higher rate of detection of viremia may be possible by hybridization on solid supports +
VIRUS-INDUCED IMMUNOSUPPRESSION
355
of total cellular RNA with HIV-specificprobes (Richman et al., 1987), or by hybridization to DNA after amplification of HIV proviral DNA in a polymerase chain reaction (Ou et al., 1988; Ameisen et al., 1989). The latter method is able to detect viral DNA in PBMCs from seropositive (but culture-negative) donors. In summary, infection with HIV can be characterized as persistent, or as latent with frequent episodes of reactivation, during which lymphocyte populations undergo radical and progressive alterations from the normal (Fig. 9). The interval of asymptomatic infection indicated in Fig. 9 may be regarded as an incubation period of disease. Mechanisms of CD4+ lymphocyte depletion could involve direct or indirect effects of HIV infection of these cells. Conversely, virus-specific immune effector mechanisms could destroy HIV-infected cells by an immunopathologic process, as in the ungulate lentiviral diseases. C. IMMUNOSUPPRESSION The immunosuppression in AIDS patients has been characterized by global dysfunction of cellular immunity and polyclonal B cell activation (reviewed by Seligmann et al., 1984; Lane and Fauci, 1985). In this review, we focus on immune dysfunction during HIV infection prior to the onset of AIDS and on the effects of HIV infection of lymphocytes and monocytes in mvo. 1. Immunologic Abnormalities in Healthy HZV-Seropositive Individuals a. Functional Abnormalities. Lymphocyte proliferation in response to mitogens and antigens is significantly decreased in patients with AIDS, but if CD4+ cells are adjusted to normal levels, only the antigen-specific response is lost (Lane et al., 1985; Giorgi et al., 1987a). Immunologic evaluation of healthy homosexual men in a cross-sectional (Nicholson et al., 1985a) or prospective analysis (Dobozin et al., 1986) demonstrates the absence of significant defects of cellular immunity in HIV-seronegative subjects and the presence of subtle defects in healthy HIV-seropositive subjects. In general, lymphocyte proliferative responses to mitogens are either in the normal range or are moderately decreased. The percentage of healthy HIV-seropositive subjects with positive proliferative responses to purified protein derivative (PPD), tetanus toxoid, and Candida antigens is similar to controls, but the percentage of subjects with positive skin test responses to PPD and tetanus toxoid is decreased (Murray et al., 1985; Dobozin et al., 1986). The proliferative response to Candida or tetanus toxoid are in the normal range when corrected for the number of CD4+ cells, but with duration of infection there is
356
MICHAEL B. MCCHESNEY AND MICHAEL B. A . OLDSTONE
an increasing number of nonresponders (Giorgi et al., 1987a). PHAstimulated IL-2 secretion is either normal in healthy homosexuals or chronic lymphadenopathy patients (Kirkpatrick et al., 1985; Dobozin et al., 1986), or decreased (Gluckman et al., 1985). In healthy homosexuals or lymphadenopathy patients with reduced proliferative responses to PHA, there was a significant correlation with decreased mitogenstimulated IL-2 production and/or IL-2 receptor expression, but there was no correlation with reduced CD4 cell numbers. Deficient mitogen responses could be due to activation in vivo, as there may be increased numbers of IL-2 receptor-positive and transferrin receptor-positive cells in the peripheral blood (Gupta, 1986). The proliferative response to PPD is decreased, but not the response to PHA (Antonen and Krohn. 1986). Antigen-induced IFN-7 secretion is impaired late in the course of HIV infection (Murray et al., 1985). Proliferative or CTL responses to the Epstein-Barr virus (Blumberg et al. , 1987), cytomegalovirus (Epstein et al., 1985), and influenza A virus (Shearer et al., 1986) are either normal or mildly depressed and reversible with IL-2 in healthy HIV seropositives. Defects in humoral immunity during HIV infection prior to AIDS include polyclonal B cell activation (PBA) and impaired specific antibody responses. Serum immunoglobulin levels are moderately increased in healthy HIV seropositives (Nicholson et al., 1986; Yarchoan et al., 1986), although a marked elevation of serum IgD has been found in healthy seropositives by Mizuma et al. (1987). The number of B lymphocytes in peripheral blood is in the normal range (Nicholson et al., 1984; Martinez-Mata et al., 1987), but B cell activation in mvo is evident by an increased percentage of B cells expressing the transferrin receptor and a decreased percentage of Leu-8+ B cells (Martinez-Mata et al., 1987). PWM-stimulated Ig secretion is depressed. A component of this depression is due to CD8+ lymphocytes, but there is a B lymphocyte defect as well (Nicholson et al., 1984). In cultures of purified peripheral blood B lymphocytes there is increased spontaneous IgG and IgM secretion (Martinez-Mata et al., 1987) and increased numbers of spontaneous immunoglobulin-secreting cells (Yarchoan et al., 1986). Unstimulated PBMCs (Teeuwsen et al., 1987) and purified B cells (Yarchoan et al., 1986) produce HIV-specificantibody, indicating a chronic active humoral response during persistent HIV infection. As in PBA associated with other diseases, antigenic or mitogenic stimulation either suppresses or does not enhance spontaneous antibody production (Yarchoan et al., 1986; Teeuwsen et al., 1987). The report of Yarchoan et al. (1986) presents a paradox: the serum antibody titer to a ubiquitous antigen, influenza A virus, is decreased relative to HIV-seronegativeconrols, but there are spontaneous influenza A virus-specific immunoglobulin-secreting cells +
VIRUS-INDUCED IMMUNOSUPPRESSION
357
in B cell cultures from HIV seropositives and not from controls. This suggests that there are restrictions controlling PBA in uivo during HIV infection. The antibody response to immunization with either T-dependent or T-independent antigens is impaired in HIV-seropositivesubjects (Ballet et al., 1987; Teeuwsen et al., 1987; Collier et al., 1988). Following booster immunization with tetanus toxoid, there is little or no specific antibody rise in vivo, and in PBMC cultures there is no spontaneous or PWMstimulated antibody to tetanus toxoid. The defect is at the level of antigen presentation or T cell help, since matrix-bound antigen rather than soluble antigen stimulates tetanus antibody production in PBMC cultures from these subjects (Teeuwson et al., 1987). Monocyte chemotaxis is reduced during HIV infection, the greatest reduction being in AIDS (Smith et al., 1984). However, microbicidal activity against intracellular parasites in response to IFN--yor stimulated lymphocyte supernatants is normal (Eales et al., 1987; Murray et al., 1987a). Lipopolysaccharide (LPS)-stimulated IL-1 secretion is not impaired, and accessory function for T lymphocyte proliferation in response to concanavalin A (Con A) and PHA is normal (Murray et al., 1987b). Accessory function for anti-CD3 monoclonal antibody-stimulated proliferation may be impaired (Prince et al., 1985). Natural killer cell numbers, defined by the Leu-11 (CD16) phenotype, are in the normal range and NK activity is normal in healthy HIV seropositives (PlaegerMarshall et al., 1987). b. Phenotypic Alterations in Pen'pheral Blood T Lymphocyte Subsets. Several monoclonal antibodies distinguish subsets within the CD4+ and CD8+ lymphocyte populations and may correlate with lymphocyte activation or with discrete effector functions (reviewed by Beverley, 1987). In the CD4+ lymphocyte population of individuals in an AIDS high-risk group, the percentage of Leu-8+ cells is decreased (Nicholson et al., 1984, 1985b) or normal (Giorgi et al., 1987b). The numbers of CD4 lymphocytes doubly fluorescent with monoclonal antibodies 4B4, 2H4, and HB-11 are distributed normally, indicating that the decrease in CD4+ lymphocytes during HIV infection is not selective for a particular phenotype of these cells (Giorgi et al., 1987a; Gupta, 1987). The percentage of CD4+ cells that are HLA-DR+ may be increased (Nicholson et al., 1985b). The increased numbers of CD8+ lymphocytes during HIV infection can be attributed to a relative and absolute increase in the Leu-8- subset (Nicholson et al., 1985b; Giorgi et al., 198713). An increase in the CD8 /Leu-8- subset is not seen in HIV-seronegativehigh-risk individuals, and thus may be a point of interest for HIV infection. The functional implications of this are not clear. +
+
358
MICHAEL B. MCCHESNEY AND MICHAEL B. A . OLDSTONE
However, the antigen recognized by Leu-8 is a marker on mature, resting lymphocytes and its expression is lost during T cell activation (Kanof and James, 1988). Thus, the Leu-8- phenotype on CD8+ lymphocytes during HIV infection may be due to activation in uivo. Another phenotypic subset of CD8+ cells, CD8+ /Leu-’/+, is significantly elevated early and throughout the course of HIV infection (Lewis et al., 1985; Stites et al., 1986; Plaeger-Marshall et al., 1987). This subset is also detected in HIV-seronegativehomosexual males and during infection with herpes viruses (Plaeger-Marshall et al., 1987). Leu-8 antigen expression overlaps with expression of Leu-7 on PBMCs (Lanier and Loken, 1984), so the CD8+/Leu8- and the CD8+/Leu-7+ subsets during HIV infection are probably overlapping and not distinct. CD8 /Leu-7 cells are not NK cells, but have LGL morphology and non-MHC-restricted CTL activity (Phillips and Lanier, 1986; reviewed by Lanier and Phillips, 1986). It is not yet clear whether there is a phenotype that distinguishes T suppressor cells from CTLs. This issue is relevant to the analysis of CD8+ HIV-specific effector T lymphocytes (see discussion below). +
+
2 . Infection of Lymphocytes and Monocytes HIV can be isolated from mitogen-stimulated CD4 + lymphocytes but not CD8 lymphocytes from the peripheral blood of HIV-seropositive donors, and normal CD4+ cells but not CD8+ cells can be infected with HIV in vitro (Klatzmann et al., 1984a). Viral replication in cells, detected by reverse transcriptase (RT) enzyme activity, is associated with loss of expression of CD4 on the cells, syncytium formation, and shortened survival in long-term culture with exogenous IL-2. Unlike infection of lymphocytes with Epstein-Barr virus or HTLV-1, productive infection by HIV requires mitogenic stimulation (Klatzmann et al., 1984a; McDougal et al., 1985b),but does not require cell division (McDougal et al., 1985b). In this requirement for cellular activation, productive infection of lymphocytes with HIV is similar to infection of T lymphocytes by measles virus (McChesney et al., 1987) and many other lymphotropic viruses (reviewed by Wheelock and Toy, 1973). Specific antigenic stimulation of primed PBMCs in the absence of exogenous IL-2 is sufficient for viral replication when HIV-seronegativenormal PBMCs are infected with HIV in Vitro (Margolick et al., 1987). In cultures of normal PBMCs or enriched CD4+ lymphocytes that are stimulated with PHA for 3 days and then infected with tissue culture-adapted strains of HIV, the number of cells containing viral antigen close to the time of peak RT activity is less than 10% in PBMC cultures (Klatzmann et al., 1984a) or is 16-40% in CD4+ cell cultures (McDougal et al., 1985b; Hoxie et al., 1985). CD4+ +
359
VIRUS-INDUCED IMMUNOSUPPRESSION
cell death occurs rapidly thereafter (McDougal et al., 1985b), or, with a different laboratory strain of HIV, a 50% reduction in viable cells is followed by maintenance of a long-term, noncytopathic, productive infection (Hoxie et al., 1985). When lymphocytes from an HIV-seropositive individual were maintained in long-term culture, HIV replication occurred transiently without significant loss of proliferating, HIV-infected cells (Fig. 9) (Gallo et al., 1984). Thus, the productive infection by HIV of lymphocytes has been characterized as lytic, in the presence (McDougal et al., 198513) or absence of syncytium formation (Somasundaran and Robinson, 1987), or persistent and noncytopathic (Gallo et al., 1984; Hoxie et al., 1985). The initial, independent isolates of HIV were characterized by syncytial cytopathic effect in mitogen-stimulatedPBMCs (Barre-Sinoussi et al., 1983; Gallo et a l . , 1984; Levy et al., 1984). However, tissue culture-adapted strains of HIV were found to vary in their ability to induce syncytia in PBMCs or in cell lines. The characterization of viral cytopathic effects in productively infected cells is a useful
1
1 VlAEMlA
-.
1
2
3
Acute Infection (months)
1
2
3
4
5
6
Asymptomatic Infection (years)
7
8
9
’-
1 0 1 1
AIDS (years)
FIG. 9. A schematic illustration of the natural history of HIV infection. The concentrations of peripheral blood (and lymph node) CD4+ and CD8+ lymphocytes are plotted over duration of the infection. Viremia, indicated by the shaded area under the dotted-line curve, is based on the relative frequency of detection of HIV in peripheral blood during infection.
360
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
virologic technique, but it has limited relevance to the fate of virusinfected cells in vivo (reviewed by Enders, 1954). HIV expression has been identified in tissue macrophages, in epidermal Langerhans cells (Tschachler et al., 1987), and in lymph node follicular dendritic cells (reviewed by Racz et al., 1986). HIV was isolated from primary cultures of brain and lung tissue of AIDS patients, in which macrophage cells are known to contain virus (Gartner et al., 1986; Wiley et al., 1986; reviewed by Wiley and Nelson, 1989), from alveolar macrophages (Salahuddin et al., 1986), and from blood monocytes by cocultivation on PBMCs (Ho et al., 1986). In Vitro, normal blood monocytes (Gartner et a l . , 1986; Ho et al., 1986) and alveolar macrophages (Salahuddin et al., 1986) are susceptible and permissive for HIV. Long-term cultures of HIV-infected monocyte/macrophages have sustained RT activity, release infectious virus in the absence of cell division, and undergo minimal syncytial cytopathic effect (Gartner et al., 1986; Nicholson et al., 1986; Salahuddin et al., 1986). Only a small percentage of cells contain viral antigen at times of optimal RT activity (Gartner et al., 1986; Ho et al., 1986). Comparing the rate of HIV replication in T lymphocytes and macrophages, RT activity appears lower in macrophages than in T cells using laboratory strains of virus adapted to growth in T cell lines (Gartner et a l . , 1986; Nicholson et al., 1986). However, in contrast, HIV isolates of tissue origin replicate equally well or better in macrophages than in T cells (Gartner et al., 1986). The cell surface receptor on mononuclear phagocytes for HIV is believed to be the CD4 molecule, which is present at low levels on some of these cells and is expressed at higher levels on the histiocytic cell line U937 (Wood et al., 1983; Moscicki et al., 1983). Blocking the expression of CD4 on monocytes (Nicholson et al., 1986) or on U937 cells (Clapham et al., 1987; Firestein et al., 1988) inhibits infection by HIV. The replication of HIV in monocytes and U937 cells can be modulated by the cellular differentiation factors IFN-y, GM-CSF, phorbol ester, and vitamin D in Vitro (Hammer et al., 1986; Gendelman et al., 1988; Pauza et al., 1988). It is not known whether primary infection with HIV is cytopathic in mononuclear phagocytes in uivo. Prior to the onset of AIDS, peripheral blood monocyte counts are in the normal range (Nicholson et al., 1984; Smith et al, 1984). However, dendritic cells in lymph nodes with reactive hyperplasia may contain HIV antigen and are hyperplastic or motheaten in appearance (reviewed by Racz et al., 1986; Piris et al, 1987). Evidence suggests that peripheral blood dendritic cells, the precursors of lymph node paracortical dendritic cells, can be infected with HIV in vitro (Patterson and Knight, 1987). PBMCs enriched for NK cells and stimulated with IL-2 can be infected with a T lymphotropic laboratory
VIRUS-INDUCED IMMUNOSUPPRESSION
361
strain of HIV and are permissive for viral replication (Ruscetti et al., 1986). Persistent infection of CD4+ cells by HIV is associated with the altered expression of CD4 and MHC class I1 molecules on the surface of infected cells. Infection of PBMC or CD4+ lymphocytes by HIV in Vitro is associated with the loss of CD4 molecules from the cell surface coincident with the onset of viral replication detectable by RT activity, from 5 to 10 days postinfection (Dalgleish et al., 1984; Klatzmann et al., 1984a; Hoxie et al., 1985; McDougal et al., 1985b). During productive infection of T cell lines, cell surface CD4 expression is diminished coincident with the onset of efficient viral replication and remains diminished after long-term culture (Hoxie et al., 1986; Stevenson et al., 1987). This phenomenon is characterized by intracellular accumulation of CD4 molecules complexed with the HIV glycoproteins gp160 and gp120, an absolte reduction of immunoprecipitable CD4 compared with uninfected cells, and a reduction of CD4 mRNA steady-state level. Similar results were reported using a retroviral vector expressing only the gp120 of HIV (Stevenson et al., 1988). Persistent productive infection of T cell lines by both HIV and SIV resulted in diminished cell surface CD4 and increased expression of MHC class I1 molecules (Kannagi et al., 1987). The increased expression of MHC molecules was not due to a soluble factor produced by infected cells, or specifically, IFN-y. In comparison, sheep monocytes infected with visna virus express increased levels of Ia antigen due to a virus-induced interferon (Kennedy et al., 1985). MHC class I1 molecules are incorporated into the HIV virion envelope (Kannagi et al., 1987), and comprise a significant amount of protein in virus preparations purified from extraneous cellular proteins by sucrose density centrifugation (Henderson et al., 1987; Hoxie et al., 1987). 3. Effects of HIV on the Functions of Normal
Lymphocytes/Monocytes in Vitro Antigen- or mitogen-stimulated proliferation of normal PBMCs exposed to HIV in Vitro is suppressed. When PBMCs from normal donors recently immunized to tetanus toxoid or keyhole limpet hemocyanin (KLH) were cultured with antigen for 3 days, then infected with a laboratory strain of HIV and cultured for 14 to 18 days, proliferation in response to antigens or PHA was significantly suppressed (Margolick et al., 1987). This was not due to loss of viable cells in long-term culture. Very low dilutions of virus stock were used for infection. Suppression required live virus because heat-inactivated virus had no effect, but it did not require a high rate of viral replication as assayed by RT activity. The mechanism for this virus-induced suppression of
362
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
lymphocyte proliferation is not clear. Inducible transcription of the IL-2 gene in some T cell lines is not altered by infection with HIV (Arya and Gallo, 1985). Exposure of normal PBMCs to inactivated HIV preparations has variable effects. Purified HIV virions (detergent solubilized) moderately inhibited lymphocyte proliferation to several mitogens and antigens (Pahwa et al., 1985). In another system, purified, freeze-thawed virions or purified viral gp120 inhibited proliferation to PHA, but not to Con A, PWM, or allogeneic stimulation (Mann et al., 1987). A recombinant molecular form of gp120 moderately suppressed lymphocyte proliferation to soluble tetanus toxoid (Shalaby et al., 1987). In the latter two systems, exposure of PBMCs to gp120 blocked the binding of monoclonal antibody OKT4a but not OKT4 to cells. The OKT4a but not the OKT4 monoclonal antibody competes with HIV binding to the CD4 molecule (reviewed by Sattentau and Weiss, 1988). The data presented by Shalaby et al. (1987) suggest that soluble gp120 binding to CD4 on cells does not result in shedding or endocytosis of the bound CD4 complex, as does the binding of some antibodies to CD4. Finally, two short synthetic peptides derived from the deduced amino acid sequence of HIV gp41, when conjugated to KLH or bovine serum albumin (BSA), moderately inhibited human and murine lymphocyte proliferation (Chanh et al., 1988). In all of the systems described above using inactivated virion or protein preparations, significant inhibition of lymphocyte proliferation required protein concentrations in the range of 1 j&ml or higher. The detergent-solubilized HIV virion preparation did not have mitogenic activity (Pahwa et al., 1985), but an intrinsic mitogenic effect of the other virus preparations was not reported. The effect of HIV on B lymphocyte functions is variable, depending on the experimental system, but HIV, like other viruses (reviewed by Ahmed and Oldstone, 1984), can be a polyclonal B cell activator. B lymphocytes are stimulated to Ig secretion in the presence of lymphocytes and a concentrated HIV preparation (Yarchoan et al., 1986). In another system using different strains of HIV, culturing purified B cells with virus at very low multiplicity (one virion per lo3 cells) resulted in proliferation and Ig secretion (Schnittman et al., 1986). In neither system was infection of cells or viral replication reported. A concentrated, detergent-solubilized HIV virion preparation acts as a T-dependent polyclonal B cell activator (Pahwa et al., 1986). The effect of HIV infection of monocytes in vitro on antigen-presenting function has not been reported. However, infection of the U937 cell line with HIV resulted in decreased accessory cell function for T lymphocyte proliferation stimulated by antLCD3 monoclonal antibody or Con A
VIRUS-INDUCED IMMUNOSUPPRESSION
363
(Petit et al., 1988). Normal peripheral blood NK cells, stimulated with IL-2, are able to lyse HIV-infected CD4+ T cell blasts. IL-2-stimulated T cells and monocytes activated with IFN-y do not have the lytic capability. However, cocultivation of normal NK cells with PHA-activated T lymphocytes and an HIV-producing T cell line did not prevent infection and replication of HIV (Ruscetti et a l . , 1986). RESPONSE TO HIV D. THEIMMUNE 1. The Humoral Immune Response Despite the immunologic defects during asymptomatic HIV infection as described above, a vigorous humoral immune response to HIV develops that has kinetics similar to that associated with other viral infections and is sustained throughout the infection until the progression to AIDS. Only a small number of HIV viremic individuals without detectable antibody by enzyme-linkedimmunosorbent assay (ELISA), Western blot, or immunoprecipitation have been identified (Mayer et al., 1986). Loss of detectable antibody to HIV with continued viremia has also been observed in 0.4% of asymptomatic carriers in a prospective study (Farzadegan et al., 1988). This failure to detect viral antibodies in serum may be due to immune complex formation in antigen excess (reviewed by Casali and Oldstone, 1983; discussed below). The primary antibody response during HIV infection includes an early and transient IgM response and a later sustained IgG response (Cooper et al., 1987). Antibodies in HIV-seropositivedonors are directed to the structural and nonstructural viral proteins, with the major response to enu and gag gene products (Pan et a l . , 1987). Antibody reactive with the nonstructural viral protein nef can be detected during early infection prior to the detection of antibodies to the other viral proteins (Ameisen et al., 1989). Within a highly conserved region of the amino-terminal end of the transmembrane protein gp41 of HIV-1, an immunodominant B cell epitope has been defined (Banapour et al., 1987; Gnann et a l . , 1987a,b; Gnann and Oldstone, 1989). Nearly 100% of American HIV-seropositive donors, including AIDS patients, have antibody reactive with this epitope in a synthetic peptide ELISA. The homologous region of HIV-2 also contains an immunodominant epitope for HIV-2-infectedWest Africans (Table 111) (Gnann et a l . , 1987c; Gnann and Oldstone, 1989). Whether the humoral immune response to HIV has a role in protection or immunopathology is yet to be determined. Viremia during HIV infection is persistent or intermittent. Significant levels of circulating immune complexes have been detected during asymptomatic infection (McDougal et al. , 1985a). Complexes precipitated in polyethylene glycol
TABLE 111 AMINOACID CONSERVATION IN THE TRANSMEMBRANE GLYCOPROTEINS OF HIV-1, HIV-2, SIV CONTAINING AN IMMUNODOMINANT B CELL EPITOPE~ Isolate name
Source
HIV-1 (LAV-lB,; HTLV-111,; HTLV-111,; WMJ-1; ARV-2) HIV-1 (CDC-451) HIV-1 (Z3) HIV-1 (LAV-lELI) HIV-1 (LAV-1MAL)
France, United States, Zaire, Haiti United States Zaire Zaire Zaire
HIV-2 ( H 1 v - 2 ~ 0 ~ ) Cape Verde Islands Kenya SIV (SIVAGM) -
AND
Amino acid sequenceb 598 609 Leu-Gly-Ile -Trp-Gly-Cvs-Ser-Glv-Lvs-Leu-Ile-Cvs
I
* p h *
t
*Leu*
t
t
t
* M e t
t
t
t
t
t
t
t t
t
t
t
*
t
t
t
t
Asn Ser * * * Asn Ser * *
* *
592
*
t
*
t
t
t
I
t
:
* & * *
* H i s t * 603 Ala Phe Arg Gln Val * Ala Trp * Gin Val * 1
t
"Modified from Gnann et al. (1987b). b T h e amino acid position numbers for HIV-1 are based on the sequence of LAV-lBRU (WainHobson et al., 1985). T h e underlined sequence is the minimal B cell epitope.
VIRUS-INDUCED IMMUNOSUPPRESSION
365
contained immunoglobulins, complement proteins, and HIV antigens or infectious virus (Morrow et al., 1986). A decline of titer or loss of antibody to the core antigen p24 is strongly associated with progression to AIDS (Kalyanaraman et al., 1984; Weber et al., 1987; discussed above). It is possible that the disapperance of p24 antigenemia during acute infection and the later disappearance of antibody to p24 are both due to immune complex formation and clearance (von Sydow et al., 1988). An unusual feature of HIV infection is that serum complement does not inactivate the virus (Banapour et a l . , 1986) and that antibody plus complement does not lyse virus-infected cells (Weiss et al., 1986; Lyerly et a l . , 1987). We predicted these events over a decade ago, prior to the discovery of human retroviruses, when the interactions of human complement with a variety of animal retroviruses were studied (Welsh et a l . , 1975, 1976; Cooper et al., 1976). Low-titer neutralizing antibody is present during HIV infection and may not diminish with disease progression (Clavel et al., 1985; Robert-Guroff et al., 1985; Weiss et d., 1985; Prince et a l . , 1987). Individual human HIV-positive sera are able to neutralize antigenically diverse isolates of the virus (Weiss et al., 1986; Prince et al., 1987), but different virus strain isolates vary in susceptibility to neutralization (Weiss et al., 1986). Both variable and conserved neutralization epitopes within the envelope glycoprotein have been defined (Ho et a l . , 1988; Matsushita et al., 1988; Palker et al., 1988), but the mechanisms and role of neutralization are not yet clear. The neutralizing antibody response to HIV appears similar to that to CAEV, in which a low or absent titer of antibody is associated with persistent infection without significant evolution of neutralization-resistantvariants (Narayan and Cork, 1985). Neutralization epitopes on the envelope glycoprotein of CAEV are masked by sialic acid (Huso et al., 1988). Antibodies able to mediate ADCC are present in the serum during HIV infection (Ljunggren et al., 1987; Ojo-Amaize et al., 1987; Rook et al., 1987). The gp120 of HIV expressed on target cells is sufficient for lysis by ADCC with HIV-positive donor sera (Lyerly et a l . , 1987; Shepp et al., 1988). 2 . The Cellular Immune Response Recent evidence suggests that there is a complex and vigorous T lymphocyte response during HIV infection. HIV-specific, MHC-restricted cytotoxic T lymphocytes are present in the peripheral blood (Walker et al., 1987, 1988; Riviere et al., 1989), in inflammatory lesions of lung (Plata et al., 1987), and in brain (Sethi et al., 1988). Viral antigens recognized by primary CTLs from peripheral blood of HIV-seropositive donors include e m , gag, and pol gene products as well as the nef and vy gene viral regulatory proteins (Walker et al., 1988; Riviere et al.,
366
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
1989). CTL recognition of viral proteins other than the envelope glycoproteins (Townsend and Skehel, 1984), and including viral polymerases (Bennink et al., 1987), has been demonstrated in other systems (reviewed by Whitton and Oldstone, 1989). Primary immunization in Vitro of lymphocytes to soluble gp120 of HIV presented by autologous antigen-presentingcells resulted in the outgrowth of CD4+, gpl20-specific T lymphocytes (Lanzavecchia et al., 1988; Siliciano et al. , 1988). T cell clones established by this method were restricted by defined MHC class I1 gene products, and two of four clones effectively lysed autologous class 11-expressingmonocytes and B and T lymphocytes (Siliciano et al., 1988). Thus CD4+ CTLs can be generated to HIV antigens in vitro. Although human antiviral CD4+ CTLs have been generated by secondary in zrilro culture, CTLs with this phenotype have not yet been detected in vivo. The route of antigen presentation of viral proteins, whether by an endogenous pathway in virus-infected cells or by the uptake and processing of viral proteins in antigenpresenting cells, is a major determinant of the phenotype and MHC restriction of antiviral CTLs (reviewed by Braciale et al., 1987). There may be discrete virus-specific CD8+ lymphocyte populations that mediate either CTL activity or suppression of virus replication. In another assay of HIV-immune T lymphocyte function, peripheral blood CD8+ lymphocytes suppress the replication of HIV (or SIV) in autologous CD4+ lymphocytes (Walker et al., 1986; Kannagi et al., 1988; Tersmette et al., 1988). Depletion and reconstitution experiments demonstrated that CD8 cells in HIV-infected lymphocyte cultures reduced RT activity (see Fig. 10) and suppressed ongoing HIV replication when added back to CD8+ lymphocyte-depleted cultures, in an MHC-restricted manner. This occurred without destroying autologous HIV-infected cells (Walker et al., 1986; Tersmette et al, 1988). A similar phenomenon was described in SIV-infected rhesus monkeys. In this system, a CD8+ T lymphocyte cell line from an SIV-infected monkey suppressed the replication of SIV in autologous cells, but did not lyse virus-infected cells in a chromium release assay (Kannagi et al., 1988). Lymphocytes from healthy HIV-seropositive donors can secrete IFN-y in response to inactivated HIV plus exogenous IL-2 (Rinaldo et al., 1988). IFN-y, per se, can suppress HIV replication (Hammer et al., 1986), but the combination of IFN-y plus TNF-(r can protect cells from HIV infection, suppress production of viral RNA, and kill HIV-infectedcells (Wong et al., 1988). Lymphocyte proliferation in response to protein antigens, including inactivated virus preparations and peptides, is thought to be a consequence of antigen processing and presentation to CD4+ lymphocytes +
VIRUS-INDUCED IMMUNOSUPPRESSION
20
40
367
60
Time in Culture (days)
FIG. 10. Cell viability and replication of HIV in culture of lymphocytes from an HIVseropositive individual. Viral replication was measured by reverse transcriptase activity. A sudden vertical drop in viable cell count indicates the time of subculturing of cells. The arrow indicates the time of addition of rabbit antiserum to interferon-a. Reprinted from Gallo et d.(1984).
(Unanue and Allen, 1987). Lymphocyte proliferation in response to inactivated HIV antigen is weak or absent in HIV seropositives (Reddy et al., 1987; Wahren et al., 1987). In marked contrast, the proliferative responses of the same donors to inactivated cytomegalovirus and herpes simplex virus are similar to normal controls. Proliferation in response to the single HIV protein, p24, was generally better than response to whole virus (Wahren et al., 1987). These results suggest that there is antigen-specific anergy during infection with HIV. However, the PBMCs from 90% of HIV-seropositive individuals had a proliferative response to at least 1 of 21 short synthetic peptides deduced from the HIV gag, pol, and env genes (Table IV) (Schrier et al., 1988). T cell activation, as measured by IL-2 production, correlated with proliferation in this assay. No single peptide elicited a response in all donors, but 48% of donors responded to a peptide from gp41 or from the polymerase. Most donors responded to more than one peptide. The heterogeneity of responses to the HIV peptides in this outbred population was correlated to peptide size and secondary structure, and to some alleles of the MHC DR locus. An immunodominant CD4+ T cell epitope for mouse strains of several MHC haplotypes has been described in the envelope protein of HIV (Cease et al., 1987). An epitope for human CD4+ T lymphocytes partially overlaps but is not continuous with the murine T cell epitope (Siliciano et al., 1988).
368
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
TABLE IV NUMBERS OF HIV-SEROPOSITIVE DONORS WITH ANTIBODY AND T CELL PROLIFERATIVE RESPONSES TO HIV P E ~ I D E S ~ Peptide gag aa 22-2gC
Antibodyb
(n = 53)
T cell proliferationb (n = 29)
7.5 7.5 0 3.8 3.8
13.8 13.8 24.1 13.8 13.8
aa 720-730
5.7
899-913 923-937 942-954 env gp120 aa 74-85 108- 119 115-126 179-186 233-244 301-31 2 368-377 env gp41 aa 584-609 603-614 609-620 655-667 737-749
3.8 1.9 3.8
34.5 27.6 37.9 48.3
0 5.7 3.8 0 4.0 0 1.9
6.9 31.0 37.9 0 6.9 13.8 20.7
100.0 60.0 35.0
24.1 41.3 42.3 37.9 20.7
228-235 282-301 439-446 466-473 PO1
3.8 3.8
"Modified from Schrier et al. (1988). Expressed as percentage of positive responses. CAmino acid position numbers are based on the HIV nucleotide sequence published by Wain-Hobson et al. (1985).
IV. Conclusion
Immunosuppression associated with HIV infection currently commands significant attention in biomedical research. Yet the basic phenomenology of disordered lymphocyte function due to infection by a lymphotropic virus is neither novel nor unexpected. The problem of measles virus-induced immunosuppression is contemporaneous with modern immunology and virology. Nearly two decades ago, animal
VIRUS-INDUCED IMMUNOSUPPRESSION
369
retroviruses were found to be lymphotropic and to profoundly alter the host immune system, resulting in immunosuppression or autoimmunity (reviewed by Hirsch, 1977). Thus any unexplained disorder of immune response, either heightened or suppressed, might well be investigated with an infectious etiology in mind. Viral probes should be useful reagents in dissecting the molecular basis of immune regulation. ACKNOWLEDGMENTS This is Publication No. 5581-IMM from the Department of Immunology, Scripps Clinic and Research Foundation, La Jolla, California 92037. This work was supported in part by USPHS Grants AI-07007, NS-12428, and United States Army Research and Development Command Contract No. DAMD17-88-C-8109.Opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the United States Army. M.B.Mc. was the recipient of an Arthritis Foundation Postdoctoral Fellowship. We thank G.L. Schilling and R. Landes for assistance in preparing this manuscript.
REFERENCES Aaby, P. (1988). Rev. Infect. Dis. 10, 451. Aaby, I?, Bukh, J., Lisse, I. M., and da Silva, M. C. (1988). Br. Med. J. 296, 1225. Adachi, A., Gendelman, H. E., Koenig, S., Folks, T., Willey, R., Rabson, A., and Martin, M. A. (1986).J. Virol. 59, 284. Adachi, A., Koenig, S., Gendelman, H. E., Daugherty, D., Gattoni-Celli, S., Fauci, A. S.,and Martin, M. A. (1987).J. Virol. 61, 209. Ahmed, R., and Oldstone, M. B. A. (1984). In "Concepts in Viral Pathogenesis" (A. L. Notkins and M. B. A. Oldstone, eds.), p. 231. Springer-Verlag, New York. Albert, J., Gaines, H., Sonnerborg, A., Nystrom, G., Pehrson, P. O., Chiodi, F., von Sydow, M., Moberg, L., Lidman, K., Christenson. B., Asjo, B., andFenyo, E. M. (1987).J. Med. Virol. 23, 67. Alizon, M., Wain-Hobson, S., Montagnier, L., and Sonigo, P. (1986). Cell (Cambridge, Mass.) 46, 63. Ameisen,J. C., Guy, B., Loche, M., Chamaret, S.,Neyrinck, J. L., Khalife,J., Boutillou, C., Gras-Masse,H., Kieny, M. P., Mouton, Y.,Beaucaire, G., Maniez. M., Laustriat, C., Berthier, A., Mach, B., Montagnier, L., Lecocq, J. P., andcaprou, A. (1989). AIDS Res. (in press). Antonen, J., and Krohn, K. (1986). Clin. Ex#. Immunol. 65, 489. Ameborn, P., Biberfeld, G., Forsgren, M., andvonstedingk, L. V . (1983). Clin.Exfi. Immunol. 51, 165. Arya, S. K., and Gallo, R. C. (1985). R o c . Natl. Acad. Sci. U S A . 82, 8691. Assaad, F. (1983). Rev. Infect. Dis. 5 , 452. Ballet, J. J., Sulcebe, G., Couderc, L. J., Danon, F., Rabian, C., Lathrop, M., Clauvel, J. P., and Seligmann, M. (1987). Clin. Ex#. Immunol. 68, 479. Banapour, B., Sernatinger, J., and Lekvy, J. A. (1986). Virology 152, 268. Banapour, B., Rosenthal, K., Rabin, L., Sharma, V., Young, L., Fernandez, J., Engelman, E., McGrath, M., Reyes, G., and Lifson, J. (1987).J. Immunol. 139, 4027. Barre-Sinoussi, F., Chermann, J. C., Rey, F., Nugeyre, M. T., Chamaret, S., Gruest,
370
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
J., Dauget, C., Axler-Blin, C., Vezinet-Brun, F., Rouzioux, C., Rozennbaum, W., and Montagnier, L. (1983). Science 220, 868. Barry, D. W., Sullivan, J. L., Lucas, S. J., Dunlap, R. C., and Albrecht, P. (1976). J. Immunol. 116, 89. Bennink, J. R., Yewdell, J. W., Smith, G . L., and Moss, B. (1987).J. Virol. 61, 1098. Benveniste, R. E., Morton, W. R., Clark, E. A., Tsai, C. C., Ochs, H . D., Ward, J. M., Kuller, L., Knott, W. R., Hill, R. W., Gale, M. J., and Thouless, M. E. (1988). J. Virol. 62, 2091. Berg, R. B., and Rosenthal, M. S. (1961). Proc. SOC. Exp. Biol. Med. 106, 581. Bernstein, L., and Rubenstein, A. (1986). Prog. Allergy 37, 194. Beverley, P. C. L. (1987). Immunol. Lett. 14, 263. Blake, F. G., and Trask, J. D. (1921a).J. Ex$. Med. 33, 385. Blake, F. G.,and Trask, J. D. (1921b).J. Ex$. Med. 33, 413. Blake, F. G., and Trask, J. D. (1921~).J. Ex#. Med. 33, 621. Blumberg, R. S., Paradis, T., Byington, R., Henle, W., Hirsch, M. S., and Schooley, R. T. (1987).J. Infect. Dis. 155, 877. Blumberg, R. W., and Cassady, H. A. (1947). Am. J. Dis. Child. 63, 151. Braciale, T. J., Morrison, L. A., Sweetser, M. T., Sambrook, J., Gething, M. J., and Braciale, V. L. (1987). Zmmunol. Rev. 98, 95. Burnet, F. M. (1968). Lancet 2, 610. Casali, P., and Oldstone, M. B. A. (1983). Cum Top. Microbiol. Immunol. 104, 7. Casali, P., Rice, G. P. A., and Oldstone, M. B. A. (1984).J. Exp. Med. 159, 1322. Cease, K. B., Margalit, H., Cornette, J. L., Putney, S. D., Robey, W. G., Ouyang, C., Streicher, H. Z.,Fischinger, P. J., Gallo, R. C., DeLisi, C., and Berzofsky, J. A. (1987). Proc. Natl. Acad. Sci. U.S,A. 84,4249. Chanh, T. C., Kennedy, R. C., and Kanda, P. (1988). Cell. Immunol. 111, 77. Cheevers, W. P., and McGuire, T. C. (1985). Rev. Infect. Dis. 7, 83. Cheevers, W. P., and McGuire, T. C. (1988). Adv. V i m Res. 34, 189. Chen, I. S. Y. (1986). Cell (Cambridge, Mass.) 47, 1. Clapham, P. R., Weiss, R. A,, Dalgleish, A. S.,Exley, M., Whitby, D., and Hogg, N. (1987). Virology 158, 44. Clavel, F., Klatzmann, D., and Montagnier, L. (1985). Lancet 1, 879. Clavel, E , Mansinho, K., Chamaret, S.,Guetard, D., Favier, V., Nina, J., Santos-Ferreira, M. 0..Champalimaud, J. L., and Montagnier, L. (1987). N . Eng1.J. Med. 316, 1180. Clements, J. E.,Godvin, S. L., Montelaro, R. C., andNarayan, 0. (1988). Annu. Rev. Immunol. 6, 139. Coffin, J. M. (1986). Cell (Cambridge, Mass.) 46, 1. Collier, A. C., Corey, L., Murphy, V. L., and Handsfield, H. H. (1988). Ann. Intern. Med. 109, 101. Cooper, D. A., Gold, J., Maclean, P., Donovan, B., Finlayson, R., Barnes, T. G., Michelmore, H. M., Brooke, P., and Penny, R. (1985). Lancet 1, 537. Cooper, D. A., lmrie, A. A., and Penny, R. (1987).J. Infect. Dis. 155, 1113. Cooper, D. A., Tindall, B.. Wilson, E. J., Imrie, A. A , , and Penny, R. (1988).J. Infect. Dis. 157, 889. Cooper, N., Jensen, E C., Welsh, R. M., and Oldstone, M. B. A. (1976).J. Ex#. Med.
144, 970. Coovadia, H. M., Parent, A., Loening, W. E. K., Wesley, A., Burgess, B., Hallett, F., Brain, P., Grace, J., Naidoo, J., Smythe, P. M., and Vos,G. H. (1974).Am.J. Clin. Nutr. 27, 665. Dahl, K., Martin, K., and Miller, G . (1987).J. Virol. 61, 1602.
VIRUS-INDUCED IMMUNOSUPPRESSION
371
Dalgleish, A. C., Beverley, P. C. L., Calpham, P. R., Crawford, D. H.,Graves, M. F., and Weiss, R. A. (1984). Nature (London) 312, 763. Daniel, M. D., Letvin, N. L., Sehgal, P. K., Hunsmann, G . , Schmidt, D. K., King, N. W., and Desrosiers, R. C. (1987).J. Gen. Virol. 68, 3183. Dewhurst, S., Sakai, K., Bresser, J., Stevenson, M., Evinger-hodges, W. J., and Volsky, D. J. (1987).J. Virol. 61, 3774. Dhib-Jalbut, S., Jacobson, S., and McFarland, H. F. (1988).J. Inject. Dis. (submitted for publication). Dixon, F. J., Talmadge, D. W., and Maurer, P. H. (1952).J. Zmmunol. 68, 693. Dobozin, B. S., Judson, F. N., Cohn, D. L., Penley, K. A , , Rickmann, P. E., Blasser, M. J., Sarin, P. S., Weiss, S. H., and Kirkpatrick, C. H. (1986). Cell. Immunol. 98, 156. Eales, L. J., Moshtael, 0.. and Pinching, A. J. (1987). Clin. Ex#. Immunol. 67, 227. Ellis, T. M., Wilcox, G . E., and Robinson, W. F. (1987).J. Gen. Virol. 68, 3145. Enders, J. F. (1954). Annu. Rev. Microbiol. 8, 473. Enders, J. F., and Peebles, T. C. (1954). R o c . SOC. Exp. Biol. Med. 86, 277. Epstein, J. S., Frederick, W. R., Rook, A. H., Jackson, L., Manischewitz,J. F., Mayner, R. E., Mausr, H., Enterline, J. C., Djeu, J. Y., and Quinnan, G. V. (1985).J. Znject. Dis. 152, 727. Evans, L. A., McHugh, T. M., Stites, D. P., and Levy, J. A. (1987). J. Immunol. 138, 3415. Fahey, J. L.,Giorgi, J.. Martinez-Maza, O., Detels, R., Mitsuyasu, R., and Taylor, J. (1987). Ann. Immunol. (Pans) 138, 245. Farzadegan, H., Polis, M. A., Wolinski, S. M., Rinaldo, C. R., Sninsky, J. J., Kwok, S . , Griffith, R. L., Kaslow, R. A., Phair, J. P., Polk, F., and Saah, A. J. (1988). Ann. Intern. Med. 108, 785. Fireman, P., Friday, G., and Kumate, J. (1969). Pediatrics 43, 264. Firestein, G . S., Reifler, D.. Richman, D., and Gruber, H. E. (1988). Cell. Zmmunol. 113, 63. Fournier, J. G . , Tardieu, M., Lebon, P., Robain, O., Ponsot, G., Rozenblatt, S., and Bouteille, M. (1985). N. Engl. J. Med. 313, 910. Fraser, K. B., and Martin, S. J. (1978). “Measles Virus and Its Biology.” Academic Press, New York. Fujinami, R. S., and Oldstone, M. B. A. (1981).J. Exp. Med. 154, 1489. Fujinami, R. S., Norrby, E., and Oldstone, M. B. A. (1984).J.Immunol. 132, 2618. Fukasawa, M., Muira, T., Hasegawa, A., Moikawa, S., Tsujmoto, H., Keizaburo, M., Kitamura, T., and Hayami, M. (1988). Nature (London) 333, 457. Funke, I., Hahn, A , , Rieber, E. P., Weiss, E.,and Reithmuller, G . (1987).J.Exp. Med. 165, 1230. Galama, J. M. D., Ubels-Postma, J., Vos, A., and Lucas, C. J. (1980). Cell. Zmmunol. 50,405. Gallagher, W. R. (1987). Cell (Cambridge, Mass.) 50, 327. Gallo. R. C., Salahuddin, S . Z., Popovic, M., Shearer, G . M., Kaplan, M., Haynes, B. F., Palker, T. J.. Redfield, R., Oleske, J., Safai, B., White, G., Foster, P., and Markham, P. D. (1984). Science 224, 500. Gartner, S., Markovitz, P.Markovitz, D. M . , Kaplan. M. H., Gallo, R. C., and Popovic, M. (1986). Science 223, 21. Gendelman, H . E., Orenstein, J. M., Martin, M. A., Ferrua, C., Mitra, R., Phipps, T., Wahl, L. A., Lane, H. C., Fauci, A. S., Burke, D. S., Skillman, D., and Meltzer, M. E. (1988).J. Exp. Med. 167, 1428.
372
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
Giorgi, J. V., Fahey, J. L., Smith, D. C., Hultin, L. E., Cheng, H.. Mitsuyasu, R. T., and Detels, R. (1987a). J. Immunol. 138, 3725. Giorgi, J. V., Nishanian, P. G., Schmid, I., Hultin, L. E.,Cheng, H., and Detels, R. (1987b). J. Clin. Immunol. 7, 140. Gluckman, J. C., and Klatzmann, D. (1986). Annu. Rev. Immunol. 4, 97. Gluckman, J. C., Klatzmann, D., Cavaille-Coll, M., Brisson, E.,Messiah, A., Lachiver, D., and Rozenbaum, W. (1985). Clin. Ex#. Immunol. 60, 8 . Gnann, J. W., and Oldstone, M. B. A. (1989). “Concepts in Viral Pathogenesis” (A. L. Notkins and M. B. A. Oldstone, eds.), Vol. 3. Springer-Verlag, New York. Gnann, J. W., Schwimmbeck, €? L., Nelson, J. A., Truax, A. B., and Oldstone, M. B. A. (1987a). J. Infect. Dis. 156, 261. Gnann, J. W., Nelson, J. A., and Oldstone, M. B. A. (1987b). J. Virol. 61, 2639. Gnann, J. W., McCormick, J. B., Mitchell, S . , Nelson, J. A ., and Oldstone, M. B. A. (1987~).Science 237, 1346. Goedert, J. J , , Biggar, R. J., Melbye, M., Mann, D. L., Wilson, S . , Gail, M. H., Grossman, R. J., DiGiola, R. A., Sanchez, W. C., Weiss, S. H., andBlattner, W. A. (1987).JAMA, J. Am. Med. Assoc. 257, 331. Gold, J. W. M., Weikel, C. S . , Godbold, J., Garcia, C., Urmacher, C., CunninghamRundles, S., Koziner, B., Pollack, M., Gallo, R. C., Sarngadharan, M . G., and Armstrong, D. (1985). Medicine (Baltimore) 64, 203. Gonda, M. A., Braun, M. J., Carter, S . G., Kost, T. A., Bess, J. W., Arthur, L. O., and Van Der Maaten, M. J. (1987). Nature (London) 330, 388. Gottlieb, M. S . , Schroff, R., Schanker, H. M., Weisman, J. D., Fan, P. T., Wolf, R. A., and Saxon, A. (1981). N . Engl. J. Med. 305, 1425. Graves, M. C., Griffin, D. E., Johnson, R. T., Hirsch, R. L., d e Soriano, I. L., Roedenbeck, S . , and Vaisberg, A. (1984). J. Virol. 49, 409. Graves, M. C., Silver, S. M., and Choppin, P. W. (1978). Virology 86, 254. Gresser, I., and Chany, C. (1963). Proc. SOC. Exp. Biol. Med. 113, 695. Gupta, S . (1986). Clin. Immunol. Zmmunopathol. 38, 93. Gupta, S. (1987). Clin. Exp. Immunol. 68, 1. Guyader, M., Emerman, M., Sonigo, P., Clavel, F., Montagnier, L., and Alizon, M. (1987). Nature (London) 326, 662. Haase, A. T. (1986). Nature (London) 322, 130. Hahn, B. H., Shaw, G. M., Taylor, M. E., Redfield, R. R., Markham, P. D., Salahuddin, S . Z., Wong-Staal, F., Gallo, R. C., Parks, E. S., and Parks, W. P. (1986). Science 232, 1548. Halsey, N. A., Boulos, R., Mode, F., Andrew, J., Bowman, L., Yaeger, R. G., Toureau, S., Rohde, J., and Boulos, C. (1985). N. Engl. J. Med. 313, 544. Hammer, S . , Gillis, J. M., Groopman, J. E., and Rose, R. M. (1986). Proc. Nutl. Acud. Sci. U.S.A. 83, 8734. Harper, M. E., Manelle, L. M., Gallo, R. C., and Wong-Staal, F. (1986). Proc. Natl. Acad. Sci. U.S.A. 83, 772. Haspel, M. V., Pelegrino, M. A., Lampert, P. W., andoldstone, M. B. A. (1977).J Exp. Med. 146, 146. Heberden, W. (1802). “Commentaries o n the History and Cure of Diseases,” p. 313. T. Payne, London. Henderson, L. E.. Sowder, R., Copeland, T. D., Oroszlan, S . , Arthur, L. O., Robet, W. G., and Fischinger, P. J. (1987). J. Virol. 61, 629. Hicks, J. T., Sullivan, J. L., and Albrecht, P. (1977).J. Immunol. 119, 1452. Hirano, T., Kuritani, T.,Kishimoto, T., and Yamamura, Y. (1977).J.Immunol. 119, 1235.
VIRUS-INDUCED IMMUNOSUPPRESSION
373
Hirsch, M. S. (1977).In “Mechanisms of Tumor Immunity” (I. Green, S . Cohen, and R. T. McCluskey, eds.), p. 287. Wiley, New York. Hirsch, V., Riedel, N., and Mullins, J. I. (1987). Cell (Cambn’dge, Mass.) 49, 307. Ho, D. D., Sarngadharan, M. G., Resnick, L., Dimarzo-Veronese,F., Rota, T. R., and Hirsch, M. S . (1985). Ann. Intern. Med. 103, 880. Ho, D. D., Rota, T. R., and Hirsch, M. S . (1986).J. Clin. Invest. 77, 1712. Ho, D.D., Kaplan, J. C., Rackauskas, I. E., and Gurney, M. (1988). Science 239, 1021. Holt, L. E. (1897). “The Diseases of Infancy and Childhood.” Appleton, New York. Howard, M., and Paul, W. E. (1983). Annu. Rev. Immunol. 1, 307. Hoxie, J. A., Haggarty, B. S., Rackowski, J. L., Pillsbury, N., and Levy, J. A. (1985). Science 229, 1400. Hoxie, J. A., Alpers, J. D., Rackowski, J. L., Huebner, K., Haggarty, B. S., Cedarbaum, A. J., and Reed, J. C. (1986). Science 234, 1123. Hoxie, J. A., Fitzharris, T. P., Youngbar, P.R., Mathews, D. M., Rackowski,J. L., and Radka, S. F. (1987). Hum. Immunol. 18, 39. Hoxie, J. A., Rackowski,J. L., Haggarty, B. S . , and Gaulton, G. N. (1988).J.Immunol. 140, 786. Huddlestone, J. R., Lampert, P. W., and Oldstone, M. B. A. (1980). Clin. Zmmunol. Immunopathol. 15, 502. Huso, D. L., Narayan, O., and Hart, G. W. (1988).J. vil-01. 62, 1974. Hussey, R. E., Richardson, N. E.,Kowalski, M., Brown, N. A., Chang, H. C., Siliciano, R. F., Dorfman, T., Walker, B.,Sodroski, J., and Reinherz, E. L. (1988). Nature (London) 331, 78. Jacobson, S., and McFarland, H. F. (1984).In “Nonsegmented Negative Strand Viruses: Paramyxoviruses and Rhabdoviruses” (D. H. L. Bishop and R. W. Compans, eds.), p. 435. Academic Press, New York. Jacobson, S., Richert, J. R., Biddison, W. E., Satinsky, A , , Hartzman, R. J., and McFarland, H. F. (1984).J. Immunol. 133, 754. Jacobson, S., Nepom, G. T., Richert, J. R., Biddison, W. E., and McFarland, H. F. (1985a).J. Exp. Med. 161, 263. Jacobson, S., Flerlage, M. L., and McFarland, H. F. (1985b).J.Exp. Med. 162, 839. Jaffe, H. W.,Francis, D. P., McLane, M. F., Cabradilla, C., Curran, J. W., Kilbourne, B. W., Lawrence, D. N., Haverkos, H. W., Spira, T. J . , Dodd R. W., Gold, J., Armstrong, D., Ley, A., Groopman, J., Mullins, J., Lee, T. H., and Essex, M. (1984). Science 223, 1309. Jaffe, H. W., Darrow, W. M., Echenberg, D. E, O’Malley, P. M.. Getchell, J. P., Kalyanaraman, V. S., Byers, R. H., Drennan, D. P., Braff, E. H., Curran, J. W., and Francis, D. P. (1985). Ann. Intern. Med. 103, 210. Joffe, M. I., and Rabson, A. R. (1981). Clin. Immunol. Immunopathol. 20, 215. Johnstone, F. D., MacCallum, L., Brettle, R., Inglis, J. M., and Peutherer, J. F. (1988). Br Med. J. 296, 467. Joseph, B. S., Lampert, P. W., and Oldstone, M. B. A. (1975).J. Virol. 16, 1638. Kalyanaraman, V. S . , Cabradilla, C. D., Getchell, J. P., Narayanan, R., Braff, E. H., Chermann, J. C., Barre-Sinoussi, F., Montagnier, L., Spira, T., Kaplan, J., Fishbein, D., Jaffe, H. W., Curran, J. W., and Francis, D. P.(1984).Science 225, 321. Kannagi, M . , Kiyotaki, M., King, N. W., Lord, C. I., and Letvin, N. L. (1987).J. Virol. 61, 1421. Kannagi, M., Chalifoux, L. V., Lord, C. I., and Letvin, N. L. (1988).J. Immunol. 140, 2237. Kanof, M. E.,and James, S. P. (1988).J. Immunol. 140, 3701.
374
MICHAEL B. MCCHESNEY AND MICHAEL B. A . OLDSTONE
Kaslow, R. A., Phair, J. P., Freidman, H. B., Lyter, D., Solomon, R. E., Dudley, J., Polk, F., and Blakwelder, W. (1987). Ann. Intern. Med. 107, 474. Keightley, R. G . , Cooper, M. D. and Lawton, A. R. (1976).J. Immunol. 117, 1538. Kennedy, P. G. E., Narayan, O., Ghotbi, Z.,Hopkins, J., Gendelman. H. E., and Clements, J. E. (1985).J. Exp. Med. 162, 1970. Kikukawa, R., Koyanagi, Y., Harado, S . , Kobayashi, N., Hatanaka, M., and Yamamoto, N. (1986).J. Virol. 57, 1159. Kirkpatrick, C. H., Davis, K. C., Horsburgh, C. R., Cohn, D. L., Penley, K., andJudson, F. N. (1985).J. Clin. Immunol. 5 , 31. Klatzmann, D., Barre-Sinoussi, F., Nugeyre, M. T., Dauget, C., Vilmer, E., Griscelli, C., Brun-Vezinet, E,Rouzioux, C.. Gluckman, J. C., Chermann, J. C., and Montagnier, L. (1984a). Science 225, 59. Klatzmann, D., Champagne, E., Chamaret, S., Gruest, J., Guetard, D., Hercend, T., Gluckman, J. C., and Montagnier, L. (1984b). Nature (London) 312, 767. Kohn, A. (1985). B o g . Med. Virol. 31, 109. Kowalski, M., Potz, J., Basiripour, L., Dorfman, T., Goh, W. C., Terwilliger, E., Dayton, A., Rosen, C., Haseltine, W., and Sodroski, J. (1987). Science 237, 1351. Koyanagi, Y., Miles, S . , Mitsuyasu, R. T., Merrill, J. E., Vinters, H . V., and Chen, I. S. Y. (1987). Science 236, 819. Krah, D. L., and Choppin, P. W. (1988).J. Virol. 62, 1565. Kreth, H. W., and Wiegand, G. (1977).J. Immunol. 118, 296. Lane, H. C., and Fauci, A. S . (1985). Annu. Rev. Immunol. 3 , 477. Lane, H. C., Depper, J. M., Whalen, G., Waldmann, T. A., and Fauci, A. S . (1985). N . Engl. J. Med. 313, 79. Lanier, L. L., and Loken, M. R. (1984).J. Immunol. 132, 151. Lanier, L. L., and Phillips, J. H. (1986). Immunol. Today 7, 132. Lanzavecchia, A., Roosnek, E., Gregory, T., Berman, P., and Abrignami, S . (1988). Nature (London) 334, 530. Lasky, L. A., Nakamura, G., Smith, D. H., Fennie, C., Shimasaki, C., Patzer, E., Berman, F?, Gregory, T., and Capon, D. J. (1987). Cell (Cambridge, Mass.) 50, 975. Laurence, J., Brun-Vezinet, F., Schutzer, S . E., Rouzioux, C., Klatrmann, D., BarreSinoussi, E,Chermann, J. C., and Montagnier, L. (1984).N . Engl.J. Med. 311, 1269. Levy, J. A., Hoffman, A. D., Kramer, S . M.,Landis, J. A., Shimabukuro, J., and Oshiro, L. S . (1984). Science 225, 840. Levy,J. A., Shimabukuro, J., McHugh, T., Casavant, C., Stites, D., and Oshiro, L. (1985). Virology 147, 441. Levy,J. A.. ChengMayer, C., Dina, D., and Luciw, P. A. (1986). Science 232, 998. Lewis, D. E., Puck, J. M., Babcock, G. F., and Rich, R. R. (1985).J. Inject. Dk. 151, 555. Lifson, J. D., Reyes, G . R., McGrath, M. S . , Stein, B. S . , and Engelman, E. G. (1986a). Science 232, 1123. Lifson, J. D., Feinberg, M. B., Reyes, G. R., Rabin, L., Banapour, B., Chakrabarti, S . , Moss, B., WongStaal, F., Steiner, K. S., andEngelman, E. C. (1986b).Nature (London) 323, 725. Ljunggren, K., Bottiger, B., Biberfeld, G., Karlson, A., Fenyo, E. M., and Jondal, M. (1987).J. Immunol. 139, 2263. Lucas, C. J., Galama, J. M. D., and Ubels-Postma, J. C. (1977).Cell. Immunol. 32, 70. Lucas, C. J., Ubels-Postma. J. C., Rezee, A., and Galama, J. M. D. (1978).J.Ex$. Med. 148, 940.
VIRUS-INDUCED IMMUNOSUPPRESSION
375
Lucas, C. J., Biddison, W. E., Nelson, D. L., and Shaw, S. (1982). Infect. Immun. 38, 226. Lyerly, H. K., Matthew, T. J., Langlois, A. J., Bolognesi, D. P., and Weinhold, K. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 4601. McChesney, M. B., and Oldstone, M. B. A. (1987). Annu. Rev. Immunol. 5, 279. McChesney, M. B., Fujinami, R. S., Lampert, P. W., and Oldstone, M. B. A. (1986). J. fip. Med. 163, 1331. McChesney, M. B., Kehrl, J. H., Valsamakis, A., Fauci, A . S., and Oldstone, M. B. A. (1987).J. vi70l. 61, 3441. McChesney, M . B., Altman, A., and Oldstone, M. B. A. (1988).J.Immunol. 140, 1269. McChesnq! M. B., Fujinami, R. S.,Lerche, N. W., M a n , I? A., and Oldstone, M. B. A. (1989).J. Infect. DiS. (in press). McCune, J. M., Rabin, L. B., Feinberg, M. B., Lieberman, M., Kosek, J. C., Reyes, G. R., and Weissman, I. L. (1988). Cell (Cambridge, Mass.) 53, 55. McDougal, J. S., Hubbard, M., Nicholson, J. K. A., Jones, B. M., Holman, R. C., Roberts, J., Fishbein, D. B., Jaffe, H. W., Kaplan, J. E., Spira, T. J., and Evatt, B. L. (1985a).J. Clin. Immunol. 5, 130. McDougal, J. S., Mawle, A., Cort, S. P., Nicholson, J. K. A., Cross, G. D., SchepplerCampbell, J. A., Hicks, D., and Sligh, J. (1985b).J. Immunol. 135, 3151. Maddon, P. J., Dalgleish, A. G., McDougal, J. S., Clapham, P. R., Weiss, R. A., and Axel, R. (1986). Cell (Cambridge, Mass.) 47, 333. Mann, D. L., Lasane, F., Popovic, M., Arthur, L. O., Robey, W. G., Blattner, W. A., and Newman, W. J. (1987). J. Immunol. 138, 2640. Mann, J. M., Colebunders, R. L., Khonde, N., Nzilambi, N., Jansegers, L., McCormick, J. B., Quinn, T. C., Bila, K., Kalemba, K., Bosenge, N., Malonga, M., Francis, H., Piot, P. and Curran, J. W. (1986). Lancet 2, 707. Margolick, J. B., Volkman, D. J., Folks, T. M., and Fauci, A. S. (1987).J. Immunol. 138, 1719. Martinez-Mata, O . , Crabb, E., Mitsuyasu, R. T., Fahey, J. L., and Giorgi, J. V. (1987). J. Immunol. 138, 3720. Matsushita, S . , Robert-Guroff, M., Rusche, J., Koito, A., Hattori, T., Hoshino, H., Javaherian, K., Takatsuki, K., and Putney, S. (1988).J. firol. 62, 2107. Mayer, K. H., Stoddard, A. M., McCusker, J., Ayotte, D., Ferriani, R., and Groopman, J. E. (1986). Ann. Intern. Med. 104, 194. Melchers, F., and Lernhardt, W. (1985). Proc. Nutl. Acad. Scz. U.S.A. 82, 7681. Mims, C. A. (1986). Clin. Ex#. Immunol. 66, 1. Minagawa, T., Nakaya, C., and Iida, H. (1974).J. virol. 13, 1118. Mizuma, H., Zolla-Pazner, S., Litwin, S., El-Sadr, W., Sharpe, S . , Zehr, B., Weiss, S., Saxinger, W. C., and Marmor, M. (1987). Clin. Ex@. Immunol. 68, 5. Montagnier, L., Gruest, J., Chamaret, S., Dauguet, C., Axler, C., Guetard, D., Nugeyre, M. T., Barre-Sinoussi, F., Chermann, J. C., Brunet, J. B., Klatzmann, D., and Gluckman, J. C. (1984). Science 225, 63. Montagnier, L., Clavel, F., Krust, B., Chamaret, S., Rey, F., Barre-Sinoussi, F.,and Chermann, J. C. (1985). Virology 144, 283. Montefiori, D. C., and Mitchell, W. M. (1987). Virology 160, 372. Morrow, W. J. M., Wharton, M., Strickler R.B., and Levy, J. A. (1986). Clin. Immunol. Immunopathol. 40, 515. Moscicki, R. A., Amento, E. P., Krane, S. M., Kurnick, J. T., and Colvin, R. B. (1983). J. Immunol. 131, 743.
376
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
Moss, A. R.. Bacchetti, P., Osmond, D., Krampf, W., Chaisson, R. E., Stites, D., Wilber, J., Allain, J. P., and Carlson, J. (1988). Br. Med. J. 296, 745. Murray, H. W., Hillman, J. K., Rubin, B. Y., Kelly, C. D., Jacobs, J. L., Tyler, L. W., Donelly, D. M., Carriero, S. M., Godbold, J. H., and Roberts, R. B. (1985). N . Engl. J. Med. 313, 1504. Murray, H. W., Scavuzzo, D., Jacobs, J. L.. Kaplan, M. H., Libby, D. M., Schindler, J., and Roberts, R. B. (1987a). J. Immunol. 138, 2457. Murray, H. W., Jacobs, J. L., and Bovbjerg, D. H. (1987b).J. Infect. Dis. 156, 696. Narayan, 0.. and Cork, L. C. (1985). Rev. Infect. Dis. 7, 89. Narayan, 0.. Griffin, D. E., and Silverstein, A. M. (1977).J. Infect. Dis. 135, 800. Nathanson, N., Panitch, H., Palsson, P. A,, Petursson, G . , and Georgsson, G. (1976). Lab. Invest. 35, 444. Nathanson, N., Georgsson, G., Palsson, P. A., Najjar, J. A., Lutley, R., and Petursson, G. (1985). Rev. Infect. DZj. 7, 75. Nicholson, J. K. A., McDougal, J. S.,Spira, T. J., Cross, G. D., Jones, B. M., and Reinherz, E. L. (1984).J. Clin. Invest. 73, 191. Nicholson, J. K. A., McDougal, J. S.,Jaffe, H. W., Spira, T.J.. Kennedy, S.,Jones, B. M., Darrow, W. W., Morgan, M., andHubbard, M. (1985a).Ann. Intern. Med. 103, 37. Nicolson, J. K. A , , McDougal, J. S.,and Spira, T. J. (1985b).J. Clin. Immunol. 5, 269. Nicholson,J. K. A,, Cross, G. D., Callawy, C. S.,and McDougal,J. S. (1986).J. Immunol. 137, 323. Norrby, E. (1971). Virology 44, 599. Norrby, E. (1986). In “Virology” (B. N. Fields, ed.), p. 1305. Raven Press, New York. Nzilambi, N., De Cock, K., Forthal, D. N., Francis, H., Ryder, R. W., Malebe, I., Getchell, J., Laga, M., Piot, P., and McCormick, J. B. (1988). N . Engl. J. Med. 318, 276. Ojo-Amaize,E. A., Nishanian, P., Keith, D. E., Houghten, R. L., Heitjan, D. F., Fahey, J. L., and Giorgi, J. V. (1987).J. Immunol. 139, 2458. Oldstone, M. B. A., and Fujinami, R. S. (1982). SOC. Gen. Microbzol. 33, 185. Osler, W. (1892).“The Principles and Practice of Medicine,” p. 77.Appleton, New York. Osunkoya, B. O., Adeleye, G. I., Adejuno, T. A., and Salimonu, L. S. (1974).Arch. Gesamte Virusforsch. 44,523. Ou, C. Y., Kwok, S.,Mitchell, S. W., Mack, D. H., Sninsky,J. J., Krebs, J. W., Feorino, P., Warfield, D., and Schochetman, G. (1988). Science 239, 295. Pahwa, S., Pahwa, R., Saxinger, C., Gallo, R. C., and Good, R. A. (1985).Proc. Natl. Acad. Sci. U S A . 82, 8198. Pahwa, S., Pawha, R., Good, R. A., Gallo, R. C., andSaxinger, C.(1986).Proc. Natl. Acad. Sci. U S A . 83, 9124. Palker, T. J., Clark, M.E., Langlois, A. J., Mathews, T. J.. Weinhold, K. J., Randall, R. R., Bolognesi, D. P., and Haynes, B. F. (1988). Proc. Natl. Acad. Sci. U S A . 85, 1932. Pan, L., ChengMayer, C., and Levy, J. A. (1987).J. Infect. Dis. 155, 626. Patterson, S., and Knight, S. C. (1987).J. Gen. Virol. 68, 1177. Pauza, C. D., Galindo, J., and Richman, D. D. (1988).J. Virol. 62, 3558. Payne, S. L., Fang, F. D., Liu, C. P., Dhurva, B. R., Rwambo, P., Issel, C. J., and Montelaro, R. C. (1987). Virology 161, 321. Pedemn, N. C., Ho, E. W., Brown, M.L., and Yamamoto, J. K. (1987). Science 235,790. Perrin, L. H.,Tishon, A., and Oldstone, M. B. A. (1977).J Immunol. 118, 282.
VIRUS-INDUCED IMMUNOSUPPRESSION
377
Petit, A. J. C., Tersmette. M., Terpstra, F. G., de Goede, R. E. Y., van Lier, R. A. W., and Miedema, F. (1988). J. Immunol. 140, 1485. Phillips, J. H.. and Lanier, L. L. (1986).J. Immunol. 136, 1579. Piot, P., Plummer, F. A., Mhalu, F. S., Lamboray, J. L., Chin, J., and Mann, J. M. (1988). Science 239, 573. Piris, M. A., Rivas, C., Morente, M., Rubio, C., Martin, C., and Olivia, H. (1987). A m . J . Clin. Pathol. 87, 716. Plaeger-Marshall, S., Spina, C. A., Giorgi, J. V., Mitsuyasu, R. T., Wolfe, P., Gottlieb, M., and Beall, G. (1987).J. Clin. Immunol. 7, 16. Plata, E,Autran, B., Martins, L. €?,Wain-Hobson,S., Raphael, M., Mayaud, C., Denis, M., Guillon, J. M., and Debre, P. (1987). Nature (London) 328, 348. Polk, B. F., Brookmeyer, R. , Kanchanaraksa, S., Kaslow, R., Visscher, B., Rinaldo, C., and Phair, J. (1987). N . Engl. J, Med. 316, 61. Popovic, M., Sargadharan, M. G., Read, E., andGallo, R. C. (1984). Science224, 497. Prince, A. M., Pascual, D., Kosolapov, L. B., Kurokawa, D., Baker, L.,and Rubinstein, P. (1987)..J Infect. Dk. 156, 268. Prince, H. E.,-Moodi, D.J., Shubin, B. I., and Fahey, J. L. (1985).J. Clin. Immunol. 5, 21. Racz, P., Tenner-Racz, K., Kahl, C., Feller, A. C., Kern, P., and Dietrich, M. (1986). h o g . Allergy 37, 81. Reddy, M. M., Englard, A., Brown, D., Buimovici-Klein, E., and Greico, M. H. (1987). J. Infect. Dis. 156, 374. Reinherz, E. L., O’Brien, C., Rosenthal, P., and Schlossman, S. F, (1980).J.Immunol. 125, 1269. Rhazes (850). “A Treatise on the Small Pox and Measles” (W. A. Greenhill, transl.) [Med. Classics 4, 22 (1939)l. Richardson, C., and Choppin, P. W. (1983). Virology 131, 518. Richman, D. D., McCutchan, A., and Spector, S. (1987).J. Infect. Dk. 156, 823. Rinaldo, C., Piazza, P., Wang, Y., Armstrong, J., Gupta, P.,Ho, M., Petteway, S., Reed, D., Lyter, D., and Kingsley, L. (1988).J. Immunol. 140, 3389. Riviere, Y., Tanneau-Salvadori, E, Regault, A., Lopez, O., Sansonetti, €?, Guy, B., Kieny, M. P., Fournel, J. J., and Montagnier, L. (1989).J. Ex$. Med. (in press). Robert-Guroff, M., Brown, M., and Gallo, R. C. (1985). Nature (London) 316, 72. Rook, A. H., Lane, H. C., Folks, T., McCoy, S., Alter, H., and Fauci, A. S. (1987). J. Immunol. 138, 1064. Rouse, B. T., and Horohov, D. W. (1986). Rev. Infect. Dis. 8 , 850. Ruscetti, F. W., Mikovits, J. A., Kalyanaram, V. S., Overton, R., Stevenson, H., Stromberg, K., Herbexman, R. B., Farrar, W. L., and Ortaldo, J. R. (1986). J. Immunol. 136, 3619. Salahuddin, S. Z., Rose, R. M., Groopman, J. E., Markham, P. D., and Gallo, R. C. (1986). Blood 68, 281. Salahuddin, S. Z.,Ablashi, D. V., Hunter, E. A., Gonda, M. A., Sturzenegger, S., Markham, P. D., and Gallo, R. C. (1987). Int. J. Cancer 39, 198. Sanchez-Lanier, M., Guerin, P., McClaren, L., and Bankhurst, A. D. (1988). Cell. Immunol. 116, 367. Sattentau, Q. J., and Weiss, R. A. (1988). Cell (Cambridge, Mars.) 52, 631. Scheid, A., and Choppin, P. W. (1984). In “Concepts in Viral Pathogenesis” (A. L. Notkins and M. B. A. Oldstone, eds.), p. 26. Springer-Verlag, New York. Schnittman, S. M., Lane, H. C., Higgins, S. E., Folks, T., and Fauci, A. S. (1986). Science 233, 1084.
378
MICHAEL B. MCCHESNEY AND MICHAEL B. A. OLDSTONE
Schrier, R. D., Gnann, J. W., Landes, R., Lockshin, C., Richman, D., McCutchan, A., Kennedy, C., Nelson, J. A., and Oldstone, M. B. A. (1989).J.Immunol. (in press). Schumperli, D. (1986). Cell (Cambridge, Mass.) 45, 473. Schwartz, K., Visscher, B. R., Detels, R., Taylor,J., Nisanian, P., andFahey, J. L. (1987). Lancet 2, 831. Scott, G. B., Fischl, M. A., Klimas, N., Fletcher, M. A., Dickinson, G. M., Levine, R. S . , and Parks, W. P. (1985).JAMA,J. Am. Med. Assoc. 253, 363. Seligmann, M.,Chess, L., Fahey, J. L., Fauci, A. S., Lachmann, P. J., L’Age-Stehr, J., Ngu, J., Pinching, A. J., Rosen, F. S., Spira, T. J., and Wybran, J. (1984). N. Eng1.J. Med. 311, 1286. Shalaby, M. R., Krowka, J. F., Gregory, T. J., Hirabayashi, S. D. E., McCabe, S. M., Kaufman, D. S., Stites, D. P., and Ammann, A. J. (1987).Cell. Immunol. 110, 140. Shaw, G. M., Hahn, B. H., Arya, S. K., Groopman, J. E., Gallo, R. C. and Wong Staal, F. (1984). Science 226, 1165. Shaw, G. M., Tarper, M. E., Hahn, B. H., Epstein, L. G. Gadjusek, D. C., Price, R. W., Navia, B. A,, Petito, C. K., Groopman, J. E.,Cho, E. S., Oleske, J. M., WongStaal, F., and Gallo, R. C. (1985). Science 227, 177. Shearer, G. M., Bernstein, D. C., Tung, K. S., Via, C. S., Redfield, R., Salahuddin, S. Z., and Gallo, R. C. (1986).J. Immunol. 137, 2514. Shepp, D. H., Chakrabarti, S., Moss, B., and Quinnan, G. V. (1988).J.Infect. Dis. 157, 1260. Sigurdsson, B. (1954). Br. Vet.J. 110, 341. Siliciano, R. F., Lawton, T., Knall C., Karr, R. W., Berman, P., Gregory, T., and Reinherz, E. L. (1988). Cell (Cambridge, Mass.) 54, 561. Silverstein, A. M., and Bialasiewicz, A. A. (1980). Cell. Immunol. 51, 151. Sissons, J. G. P., and Oldstone, M. B. A. (1980). Adu. Immunol. 29, 209. Sissons, J. G. P.. Colby, S . D., Harrison, W. O., and Oldstone, M. B. A. (1985). Clin. Immunol. Immunopathol. 34, 60. Smith, D. H., Bym, R. A., Marsters, S. A., Gregory, T., Groopman, J. E., a n d c a p o n , D. J. (1987). Science 238, 1704. Smith, P. D., Ohura, K., Masur, H., Lane, H. C., Fauci, A. S., and Wahl, S. M. (1984). J. Clin. Invest. 74, 2121. Smithwick, E. M., and Berkovich, S. (1969). In “Cellular Recognition” (R. T. Smith and R. A. Good, eds.), p. 131. Appleton, New York. Somasundaran, M., and Robinson, H. L. (1987). J. Virol. 61, 3114. Sonigo, P., Alizon, M., Staskus, K., Klatzmann, D., Cole, S . , Danos, O., Retzel, E., Tiollais, P., Haase, A., and Wain-Hobson, S. (1985). Cell (Cambridge, Mass.) 42, 369. Stevenson, M., Zhang, X.,and Volsky, D. J. (1987).J. Virol. 61, 3741. Stevenson, M., Meier, C., Mann, A. M., Chapman, N., and Wasiak, A. (1988). Cell (Cambridge, Marc.) 53, 483. Stites, D. P., Casavant, C. H., McHugh, T. M., Moss, A. R., Beaql, S . L., Ziegler, J. H., Saunders, A. M., and Warner, N. L. (1986). Clin. Immunol. Immunopathol. 38, 161. Stoneburner, R. L., Ruiz, M. M., Milberg, J. A., Schultz, S., Vennema, A., and Morse, D. L. (1987). Morbid. Mortal. Wkly. Rep. 36, 785. Sullivan, J. L., Barry, D. W., Lucas, S . J., and Albrecht, P. (1975).J.Erp. Med. 142,773. Taliaferro, W. H., Taliaferro, L. G., and Janssen, E. F. (1952).J. Infect. Dis. 91, 105.
VIRUS-INDUCED IMMUNOSUPPRESSION
379
Teeuwson, V. J. P., Logtenberg, T., Siebelink, K. H. J., Lange, J. M., Goudsmit, J., Uytedhaag, E G. C. M., and Osterhaus, A. D. M. A. (1987).J.Immunol. 139,2929. ter Meulen, V., and Carter, M. J. (1984). h o g . Med. Virol. 30, 44. ter Meulen, V., Stephenson, J. R. and Kreth, H. W. (1983). Compz Virol. 18, 105. Tersmette, M., de Goede, R. E. Y., Al, B. J. M., Winkel, I. N., Gruters, R. A., Cuypers, H. T., Huisman, H. G., and Miedema, F. (1988).J. Virol. 62, 2026. Townsend, A. R., and Skehel, J. J. (1984).J. Ex$. Med. 160, 552. Tschachler, E.,Groh, V.,Popovic, M., Mann, D. L., Konrad, K., Safai, B., Eron, L., Veronesse, F., Wolff, K., and Sting], G. (1987).J. Invest. Dermatol. 88, 233. Unanue, E. R., and Allen, P. M. (1987). Science 236, 551. von Pirquet, C. (1908). Dtsch. Med. Wochenschr. 30, 1297. von Sydow, M., Gaines, H., Sonnerborg, A., Forsgren, M., Pehrson, P. 0.. and Strannegard, 0. (1988). Bz M e d . J . 296, 238. Wahren, B., Morfeldt-Mansson, L.,Biberfeld, G., Moberg, L., Sonnerborg, A., Ljungman, I?, Werner, A., Kurth, R., Gallo, R., and Bolognesi, D. (1987).J. Virol.
61,2017.
Wain-Hobson, S., Sonigo, P., Danos, O., Cole, S., and Alizon, M. (1985). Cell (Cambridge, Mass.) 40. 9. Walker, B. D., Chakrabarti, S., Moss, B., Paradis, T. J., Flynn, T., Durno, A. G., Blumberg, R. S., Kaplan, J. C., Hirsch, M. S . , and Schooley, R. T. (1987). Nature (London) 328, 345. Walker, B. D., Flexner, C., Paradis, T. J., Hirsch, M. S . , Schooley, R. T., and Moss, B. (1988). Science 240, 64. Walker, C. M., Moody, D. J., Stites, D. P., and Levy, J. A. (1986). Science 234, 1563. Warthin, A. S. (1931). Arch. Pathol. 11, 864. Weber, J. N., and Weiss, R. A. (1988). Bz Med. Bull. 44, 20. Weber, J. N.,Clapham, P. R., Weiss, R. A., Parker, D., Roberts, C., Duncan, J., Weller, I., Came, C., Tedder, R. S., Pinching, A,, and Cheingsong-Popov, R. (1987). Lancet 1, 119. Weiss, R. A., Clapham, P. R., Cheingsong-Popov, R., Dalgleish, A. G., Came, C. A., Weller, I. V. D., and Tedder, R. S. (1985). Nature (London) 316, 69. Weiss, R.A.,Clapham, P. R., Weber, J. N., Dalgleish, A. G., Lasky, L. A., and Beman, P. W. (1986). Nature (London) 324 572. Welsh, R. M., Cooper, N., Hensen, F. C., and Oldstone, M. B. A. (1975). Nature (London) 257, 612. Welsh, R. M., Jensen, F. C., Cooper, N. R., and Oldstone, M. B. A. (1976). Virology
74, 432.
Wesley, A., Coovadia, H. M., andHenderson, L. (1978). Clin. Ex#. Zmmunol. 32,540. Wheelock, E. E , and Toy, S . (1973). Adv. Immunol. 16, 123. Whittle, H. C., Bradley-Moore, A., Fleming, A., and Greenwood, B. M. (1973).Arch. Dis. Child. 48, 753. Whittle, H. C., Dosseter, J., Oduloju, A., Bryceson, A. D. M., and Greenwood, B. M. (1978).J. Clin. Invest. 62,678. Whitton, J. L., and Oldstone, M. B. A. (1989). Zn “Concepts in Viral Pathogenesis” (A. L. Notkins and M. B. A. Oldstone, eds.), Vol. 3. Springer-Verlag, New York (in press). Wiley, C. A., and Nelson, J. A. (1989). Curz Top. Microbiol. Zmmunol. (in press). Wiley, C. A., Schrier, R. D., Nelson, J. A., Lampert, P. W., and Oldstone, M. B. A. (1986). R o c . Natl. Acad. Sci. U S A . 83, 7089.
380
MICHAEL B. MCCHESNEY AND MICHAEL B. A . OLDSTONE
Wong, G . H. W., Krowka, J. F., Stites, D. P.,and Goeddel, D. V. (1988). J. Immunol. 140 120. Wood, G . S . , Warner, N. L . , and Warnke, R. A . (1983). J. Immunol. 131, 212. Yamanouchi, K . , Egashira, Y., Uchida, N . , Kodama, H . , Kobune, F., Hayami, M . , Fukuda, A . , and Shishido, A. (1970). J@. J. Med. Sci. Biol. 23, 131. Yamanouchi, K., Chino, F., Kobune, F.,Kodoma, H., andTsuruhara, T. (1973).J InJect. Dis. 128, 795. Yarchoan, R.,Redfield, R . , and Broder, S. (1986). J. Clin. Invest. 78, 439. Zweiman, B. (1971). J. Immunol. 106, 1154.
ADVANCES IN IMMUNOLOGY. VOL. 45
The Regulators of Complement Activation (RCA) Gene Cluster DENNIS HOURCADE, V. MICHAEL HOLERS, AND JOHN P. ATKINSON The Howard Hughes Medical Institute laboratories and Department of Mediclne, DIvlslon of Rheumatoiogy, Washington Univenlty School of Medicine, St. Louis, Missouri 63110
I. Introduction
The complement proteins collectively play a leading role in the immune system, both in the identification of and in the removal of foreign substances and immune complexes (Reid, 1986; Ross, 1986; Atkinson and Eisen, 1988). Central to the complement system are the C3 and C4 proteins, which, when activated, covalently attach to nearby targets, marking them for clearance. In order to help control this process, a remarkable family of receptor and regulatory proteins has evolved, with each member interacting with activated C3 and/or C4 derivatives. This group of proteins constitutes the regulators of complement activation (RCA) gene family (de Cordoba et al., 1985) and is encoded at a single chromosomal location, the RCA cluster (Campbell et al., 1988). The study of the RCA proteins has been accelerated by the availability of molecular techniques. Recently it has been established that each of these proteins is composed primarily of a repeated 60- to 70-amino acid motif (reviewed in Reid et al., 1986). This discovery has placed the biochemistry and evolution of the RCA proteins in a new light. In this review we will discuss the physiological roles of the RCA family, stress the common structural and evolutionary relationships among the RCA proteins and their genes, and present the case that from a common structural motif there has evolved an elaborate regulatory network involving numerous proteins and cell types and featuring interactions in plasma, interactions at the cell surface, and interactions within cells. II. The Complement Pathways
The complement system can be divided into two main pathways, the classical pathway and the alternative pathway (Reid, 1986; Pangburn and Muller-Eberhard, 1984). As can be seen in Fig. 1, when antibody/ antigen recognition occurs (the classical pathway), the interaction of C1 381 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
382
DENNIS HOURCADE ET AL.
Classical Pathway Antibody-Antigen ComDlex
c3
Alternative Pathway C3a Microbes, Polysaccharldes. etc.
C3b-Target
Ba
B
A
’
FIG. 1. The complement pathways. Emphasis is placed on the formation of the enzymes that cleave C3.
with the Fc regions of IgG or IgM in immune complexes activates a Cl protease that can cleave the plasma protein C4, resulting in the C4a and C4b fragments. C4b, which may covalently attach to the target immune complex, can bind to another plasma protein, C2. The resulting species, C4bC2, is cleaved by the C1 protease to form the classical pathway C3 convertase, C4bC2a. The convertase, in turn, carries its own proteolytic activity, which can activate CS to form C3b, a central molecule in the complement pathway. C3b can covalently attach to any nearby structure, marking it for endocytosis, or for transport to the liver or spleen for further processing. The CS convertase is also required for the activation of C5, the first step in the assembly of a membrane attack complex which leads to cell destruction (Mayer et al., 1981; Muller-Eberhard, 1984). While the initial interaction which triggers the alternative pathway is not understood as well as that for the classical pathway, the alternative pathway is accelerated by a wide variety of surfaces, including those of many microbes. These foreign substances provide a focused site for the efficient assembly of convertases. Initial complement deposition on these surfaces is thought to arise from spontaneously activated C3. This differentiates it from the classical pathway, which requires interaction of
REGULATORS OF COMPLEMENT ACTIVATION
383
Ab and complement for initiation. In a cascade that parallels that of the classical pathway, a resulting C3 convertase proteolytically modifies CS, producing the C3b fragment, which again may covalently attach to the target. C3b can bind plasma protein B, which is then cleaved by protease D to form the alternative pathway convertase (C3bBb). Properdin stabilizes this bimolecular enzyme complex. The alternative pathway convertase, like the classical pathway convertase, cleaves more CS, producing large numbers of C3b and terminal component deposition (Fig. 1). 111. Biochemical Interactions of the RCA Proteins with C3b/C4b
The classical and alternative pathway convertases (the C3 convertases) are important control points in complement activation: since newly activated C3b can bind covalently to normal cells and molecules as well as microbes and immune complexes, the generation and activity of C3b must be carefully directed. Two distinct biochemical mechanisms are instrumental in this process: decay acceleration and factor I-mediated degradation (Holers et al., 1985; Reid, 1986; Ross and Medof, 1985). Activities of the C3 convertases are dependent on the association of two components, C4b plus C2a in the classical pathway convertase and C3b plus Bb in the alternative pathway convertase. Although the C3 convertases are inherently unstable, their dissociation is hastened by protein-mediated decay acceleration (Fig. 2A) in which the protease, C2 or Bb, is dissociated from the cell-bound C4b or C3b. In addition, the proteins which exhibit decay-accelerating activity may associate with C3b or C4b alone, inhibiting their interactions with B or C2, respectively, and thus prohibit the generation of new C3 convertases. In one respect, decay acceleration is reversible because dissociated C4b and C3b retain the capacity to bind C2 or B. Decay acceleration of the classical convertase is mediated by RCA proteins-C4bp in plasma and decayacceleration factor (DAF) and CR1 at the cell surface-while decay acceleration of the alternative pathway is mediated by factor H (H) in plasma and DAF and CR1 on cell surfaces (Table I). The C3 convertases can also be irreversibly inactivated (Fig. 2B). In that case, factor I, a highly specific serine protease, inactivates the convertase component, CSb or C4b, by proteolytic cleavage, but only in the presence of a cofactor. The cofactor function is carried out by RCA proteins-H and C4 binding protein (C4bp) in plasma and CR1, membrane cofactor protein (MCP), and possibly CR2 on cell surfaces. The first four proteins are sufficient to allow the first factor I-mediated cleavage, which produces soluble C3f and attached C3bi. Only CRl can
384
DENNIS HOURCADE ET AL.
A
FIG. 2. Biochemical interactions of the RCA proteins with C3b. (A) Decayaccelerating activity; (B) cofactor activity. Figure indicates membrane-bound CSb, but similar reactions occur with fluid-phase C3b. Parallel reactions occur with C4b and its binding to C2. In that case, C4b interacts with C4bp in the same way that C3b interacts with H (A). Binding of H, DAF, MCP, or CR1 to C3b is not specifically shown.
efficiently mediate the first and second cleavage, resulting in a soluble CSc fragment and the attached CSd,g fragment. In addition to decay-accelerating and cofactor activities, two surface RCA proteins exhibit receptor activity: CR1 binds C3b and C4b while CR2 primarily binds CSd,g. In some cell types, receptors mediate the transport of immune complexes to the liver and spleen for clearance, while in other cells, receptors are involved in endocytosis, phagocytosis, or in signal transduction across cell surface membranes (see discussions below). IV. The Roles of the RCA Proteins
A. HISTORICAL DEVELOPMENT In the late 1970s the control of the complement system appeared to be relatively straightforward: an immune complex triggered the classical pathway and a foreign membrane triggered the alternative pathway. For the classical pathway a C1 inhibitor provided additional control of the
REGULATORS OF COMPLEMENT ACTIVATION
385
TABLE I RCA PROTEINS: STRUCTURAL FEATURES AND DISTRIBUTION M~
Protein
x 10-3
Nonreduced
Reduced
Distribution
H
160
170
C4bp CR1"
590 190
70 220
CR2
130
140
DAF
70
75
MCP MCP
58 63
63 68
Plasma; synthesized in liver, fibroblasts, and U937 cells Plasma; synthesized in liver Membranes of erythrocytes, monocytes, granulocytes, some T cells, B cells (pre-B, mature, and plasma), follicular- dendritic cells, glomerular podocytes Membranes of B cells (mature), follicular-dendritic cells, epithelial cells (stage specific), T cell lines (absent from peripheral T cells), platelets(?) Membranes of erythrocytes, T cells, B cells, monocytes, granulocytes, platelets, endothelial cells, epithelial cells, fibroblasts Membranes of T cells, B cells, monocytes, granulocytes, platelets, endothelial cells, epithelial cells, fibroblasts
"Most common phenotypic variant.
activated C1 complex, limiting inappropriate activation in the fluid phase. Two other plasma proteins, H and C4bp, provided control at the next critical step in the system, the formation of the C3 convertases. Through their decay-accelerating activity and their factor I-mediated cofactor activity, H regulated the alternative pathway convertase as well as fluid-phase and cell-bound C3b, while C4bp regulated the classical pathway convertase as well as fluid-phase and cell-bound C4b. While there were problems with the above model (reviewed in Atkinson and Farries, 1987), it was the identification and characterization of three membrane proteins, the C3b/C4b receptor (CRl), decay-accelerating factor, and membrane cofactor protein, each with the capacity to modulate convertase activity, that showed that some other form of regulation must be occurring. This led not only to a reassessment of the role of H and C4bp but also to a critical look at the control of the amplification stage of the complement system and to the elucidation of a role for the complement system in the processing of immune complexes.
386
DENNIS HOURCADE ET AL.
B. PROTECTION OF AUTOLOGOUS TISSUE The initiating principle of the alternative pathway is the instability of C3: C3 contains an internal thiol ester bond that links two of its amino acid residues (Tack, 1985; Campbell et al., 1988). The bond is unstable and spontaneously breaks, mediating covalent attachment of C3 to nearby molecules, forming amide linkages with amines and ester linkages with hydroxyl groups. Thus, C3 derivatives may attach to nearby autologous tissue just as well as foreign structures. Activated C4b is also nonspecific in its reactions with nearby molecules due to a similar thiol ester in its structure. In fact, during complement activation by either pathway, a majority of the C3bK4b that is formed does not become covalently bound to the target. Although a cluster of C3b and C4b molecules is deposited at the site of complement activation, as much as 70 to 90% does not effectively interact with a target and thus remains in the fluid phase (see below). Moreover, of the 10 to 30% that does become bound, especially at an inflammatory site, some binds to self-tissue. Since C3b molecules must not be allowed to cluster on self-tissue, two RCA proteins, DAF and MCP, have evolved which protect cells by patrolling their surfaces and inactivating C3b and C4b (Fig. 3).
A
FIG.3. Regulation of membrane-bound C3b by (A) DAF and (B) MCP. Similar reactions occur to membrane-bound C4b. Note the complementing inhibitory profile of DAF and MCP. It is not known if DAF binds to the same or a different site than Bb. However, the binding of DAF does not prevent the action of MCP (T. Seya, T. Kinoshita, and J. P. Atkinson, unpublished). Cleavage of C3b (B) requires the factor I protease (not shown).
387
REGULATORS OF COMPLEMENT ACTIVATION
-
DAF is an 70-kDa glycolipid-anchored membrane protein with a wide tissue distribution that possesses regulatory activity for the C3 convertases (reviewed in Medof, 1988; Lublin and Atkinson, 1988). It was described as an inhibitor of complement action by Hoffman (1969a,b) and was purified by Nicholson-Weller et al. (1982). MCP is composed of two distinct but highly homologous proteins with mean molecular weights of 68,000 and 63,000. Like DAF, it has a broad tissue distribution. MCP was discovered by Cole et al. (1985) based on its ability to bind to C3b and was purified and shown to possess cofactor activity by Seya et al. (1986). DAF and MCP are ideally suited to provide protection for autologous tissue (reviewed in Atkinson and Farries, 1987; Lublin and Atkinson, 1988). First, they are widely distributed, being found not only on most peripheral blood cells but also on endothelial and epithelial cells and on fibroblasts. Second, they have complementary functions: DAF can prevent the formation of convertases and dissociate preformed convertases. MCP can permanently inactivate C3b or C4b through its cofactor activity. The importance of DAF for the protection of autologous tissue is seen in paroxysmal nocturnal hemoglobinuria (PNH), an acquired hemolytic anemia in which affected blood cells show increased sensitivity to complement-mediated lysis, partly or entirely due to an increased uptake of C3b (Medof, 1988). These cells lack DAF, but the reincorporation of DAF into their membrane normalizes their C3b uptake and partially corrects their complement sensitivity. Factor H in plasma does not efficiently protect these cells. A membrane protein on the same cell as the deposited C3b appears necessary. C. CONTROL OF PLASMA COMPLEMENT ACTIVITY Factor H is a relatively abundant plasma protein of 150 kDa that was first described as a complement control protein by Whaley and Ruddy (1976). It binds to C3b, preventing the formation of the alternative pathway convertase, and it dissociates the preformed C3bBb complex (decay acceleration). Further, H serves as a cofactor for the proteolytic degradation of C3b. C4bp is a relatively abundant plasma regulatory protein of 550 kDa, formed from seven identical 70-kDa subunits. It binds to C4b, preventing the formation of the classical pathway convertase, and it dissociates the preformed C4bC2a complex. C4bp serves as a cofactor for the proteolytic degradation of C4b. It was characterized and purified by Ferreira et al. (1977) and Scharfstein et al. (1978). C4bp can also interact with C3b in a manner analogous to that of H.
-
388
DENNIS HOURCADE ET AL
The major physiological role of H and C4bp appears to be the inhibition of convertase activity in the fluid phase, although these proteins undoubtedly contribute in the inhibition of C3b and C4b attached to cell surfaces. Thus, the small quantities of activated C3 and C4 formed during the low-grade turnover of these components and the larger amounts that occur during complement activation can be inhibited. The newly formed C3b or C4b that does not become bound to the target but which escapes into the surrounding milieu becomes bound by H or C4bp. In addition, a convertase already formed can be rapidly disassociated by H or C4bp. Consequently, the resulting C3b/C4b can be proteolytically degraded by factor I, with H and C4bp serving as cofacton, The importance of these proteins is best illustrated in individuals with H or factor I deficiency (reviewed in Lachman and Walport, 1986). In these cases, C3 levels are low or undetectable, due to increased activity of C3 convertase. Infusion of normal plasma corrects the defect.
D. TRANSPORT AND CLEARANCE OF IMMUNE COMPLEXES CR1 is a polymorphic (190-280 kDa) receptor whose primary ligands are C3b and C4b (reviewed in Ross and Medof, 1985). It has a limited tissue distribution, being found primarily on mature hematopoietic cells in the peripheral blood. It was known by its function (binding of CSbK4b-coated ligands) in the 1940s and 1950s, but it was not until 1979 that it was purified by Fearon (1979). CRl has been shown to possess decay-accelerating,cofamr, and C3b/C4b receptor activities (Iida and Nussenzweig, 1981; Medof et al., 1982; Fearon, 1980). Its major biological function is to bind, process, and transport CSbK4b-coated immune complexes (IC) and particles. The adherence of serum-treated microbes to mammalian peripheral blood cells was first shown around the turn of the century. This phenomenon was rediscovered and termed the immune adherence (IA) phenomenon by Nelson (1953), who demonstrated that immunoglobulin-and complement-coatedbacteria and spirochetes would adhere to primate erythrocytes (E) and subprimate platelets. The binding was complement dependent and was subsequently shown to be mediated by C3b and to a lesser extent by C4b. The physiological role of this activity was not understood until recently. If ICs are placed in whole blood, they become bound within seconds, not to monocytes, B lymphocytes, or granulocytes, but to erythrocytes (Ross and Medof, 1985). Erythrocytes in turn transport the IC to the liver and spleen where the IC (possibly including CRl) is stripped from the E. The E remains viable and returns to the circulation. This process for the metabolism of ICs is called the primate-E-IC-clearance mechanism (Hebert and Cosio, 1987). The erythrocytes serve as a sump (bind),
REGULATORS OF COMPLEMENT ACTIVATION
389
processing station (inhibit further complement activation), and taxi (transfer to liver and spleen) for complement-fixing IC. ICs form in association with most infectious illnesses. In diseases such as malaria, where there is intravascular release of large amounts of antigenic and infectious material, the rarity of synovitis, dermatitis, and nephritis attests to the efficiency of this system. While the mechanism for the removal of ICs from the erythrocytes in the liver and spleen is poorly understood, other features of the process are clear. The host must make an antibody response and activate complement. Clusters of C3b must be deposited and, in turn, serve as a ligand for CR1 on erythrocytes. In an experimental system employing baboons, inefficient complement activation actually led to a more rapid clearance of the IC but the site of deposition was inappropriate, namely, the kidneys and lung (reviewed in Hebert and Cosio, 1987). A similar outcome would be expected if an ineffective antibody response were mounted or if the antibodies did not fix complement or if the receptors on erythrocytes or in the liver were deficient or blocked.
E. TRANSMEMBRANE AND INTRACYTOPLASMIC FUNCTIONS The roles played by the RCA family of integral membrane proteins (CRl, CR2, DAF, and MCP) in transmembrane signaling, phagocytosis, and intracellular trafficking in nucleated cells are areas of growing interest. The majority of experimental data has been accumulated for CR1. It is clear that CRI mediates the binding of C3b-coated targets to both polymorphonuclear cells and monocytes (Fearon et al., 1981; Abrahamson and Fearon, 1983). A second signal, however, is required to induce phagocytosis (O’Shea et al., 1985b; Wright and Meyer, 1985a). Phorbol esters stimulate this phagocytosis in both polymorphonuclear leukocytes and monocytes. In addition, interaction of monocytes with extracellular matrix proteins, such as fibronectin or peptides containing the sequence Arg-Gly-Asp(Wright et al., 1983, 1986; Pommier et al., 1984; Wright and Meyer, 1985b; Bohnsack et al., 1986), and interaction of peritoneal macrophages with certain lymphokines(Griffin and Mullinax, 1981) induce CR1-mediated phagocytosis. An intracytoplasmic pool of CRI in polymorphonuclear cells has been demonstrated (Fearon and Collins, 1983; O’Shea et al., 1985a; Berger and Cross, 1984). This pool can be rapidly translocated to the cell surface and then be reinternalized by phorbol dibutyrate, Net-Leu-Phe, C5a, and other modulators. CR1 can be phosphorylated in polymorphonuclear (PMN) cells and monocytes (but not in B cells) upon treatment with phorbol esters (Changelian and Fearon, 1986). However, polymorphonuclear CR1 has not as yet been shown to be phosphorylated upon interaction with physiological ligands or anti-CR1
390
DENNIS HOURCADE ET A L .
antibodies alone. A critical but as yet unanswered question is whether these experimentally induced CR1 modulations accurately reflect in vivo physiology. Although an early report suggested that ligand binding to CRl caused PMN histaminase release, this has not been confirmed (Melamed et al., 1982). A second report suggests that C3b can lead to translocation of an intracellular pool of CR1 in PMN cells to the plasma membrane (Porteu et al., 1987). However, most laboratories have been unable to demonstrate that ligand binding to CRl generates any change in cell responses or CRl expression. CR2 is a 145-kDa receptor for the -40-kDa degradation fragment of C3b, now termed C3d,g (Fig. 2). It also serves, in a pathologic sense, as a receptor for the Epstein-Barr virus (Fingeroth et al., 1984; Nemerow et al., 1985b). CR2 was discovered as a functional entity by Ross et al. (1973) and by Eden et al. (1973) and was identified as the C3d receptor by Iida et al. (1983). The role of CR2 in transmembrane signaling has gained recent interest because of observations which strongly indicate that CR2 plays a role in normal B cell physiology. Several lines of evidence support this: (1) CR2 is found predominantly on mature B cells, which can be stimulated to proliferate and differentiate (Tedder et al., 1984), (2) CR2 is linked functionally to membrane IgM, indicated by reciprocal co-capping of each when B cells are treated with an antibody to the other protein (Tanner et al., 1987), (3) intracytoplasmic Ca2+ can be increased by the interaction of CR2 with anti-CR2 antibodies in the presence of limiting amounts of anti-IgM (Carter et al., 1988), (4) some polyclonal and monoclonal anti-CR2 antibodies stimulate T cell-dependent proliferation of B cells (Carter et al., 1988; Nemerow et al., 1985a; Frade et al., 1985; Wilson et al., 1985), and ( 5 ) fragments of C3 stimulate growth of B lymphoblastoid cell lines in serum-free media (Schulz et al., 1987). As opposed to B cell CR1, B cell CR2 can be phosphorylated using phorbol esters (Changelian and Fearon, 1986), although the physiological relevance of this is also unknown. No specific role has been ascribed to DAF or MCP in transmembrane signaling. However, a polyclonal anti-DAF antibody was found to specifically stimulate proliferation of peripheral T cells (Davis et al., 1988). This effect was abrogated by removal of DAF with phosphatidyl inositol-specific phospholipase C. Whether this effect is specific to DAF, or due to a general phenomenon associated with a glycolipid anchor, is unknown.
391
REGULATORS OF COMPLEMENT ACTIVATION
V. The Short Consensus Repeat: Structure and Function
A. A COMMON STRUCTURAL MOTIF The use of molecular techniques has led to the cloning and sequencing of cDNAs corresponding to CR1 (Klickstein et al., 1987a,b; Hourcade et al., 1988), CR2 (Moore et al., 1987; Weis et a l . , 1988), DAF (Caras et al., 1987; Medof et al., 1987b), MCP (Lublin et al., 1988), C4bp (Chung et al., 1985b; Kristensen et al., 1987), and H (Kristensen and Tack, 1986; Ripoche et al., 1988). From this work we have learned that the RCA proteins are composed mainly of a tandemly repeated motif (short consensus repeat; SCR) of about 60-70 amino acids in length (Reid et al., 1986) (Fig. 4). Each SCR features 5 residues that occur in greater A
B SCRs
non-SCR Domains
FIG.4. (A) Short consensus repeat (SCR). Those amino acids shown are conserved in at least 44% of 102 SCRs analyzed (Perkins et al., 1988). The cysteines at positions 2, 31,45, and 58 and the tryptophan at position 51 are found in 95-100% of the SCRs analyzed. Positively or negatively charged residues (not shown) occur in more than 40% of the SCRs at positions 9, 12, 26, 28, 30, 32, 59, and 60.(B) Organization of the RCA proteins into SCR-containing and non-SCR-containing regions (NHp-terminalends at the left). TM, Transmembrane domain: C, cytoplasmic domain; 0, 0-linked glycosylation domain: G, glycolipid anchor: U, domain with unknown functional significance; D, disulfide bridge-containing domain.
392
DENNIS HOURCADE ET A L
than 95% of the cases, 17 residues that are conserved in at least 44% of the cases, and 8 additional positions that are occupied by charged amino acids in at least 40% of the cases (Fig. 4; see also Perkins et a l . , 1988). In addition, SCRs have been found in a number of other complement proteins that also interact with C3b or C4b, and in several proteins with no known relationship to complement (see discussionsbelow). Most of the available secondary structure information for the RCA proteins is derived from studies of C4bp and of H, but it is generally assumed that all multiple SCR arrays will have much in common, given the highly conserved primary structure. Electrophoresis of C4bp (Dahlback, 1983) taken together with amino acid sequence studies (Reid and Gagnon, 1982; Dahlback and Stenflo, 1981; Chung et al., 1985a) show it to be composed of six to eight identical disulfide-linked chains of 70 kDa. Each subunit is composed of eight SCRs and a carboxyterminal non-SCR domain (Chung et al., 1985a). Subunits are attached to a central core through disulfide linkages in the non-SCR domains. From electron microscopic (EM) observations (Dahlback et al., 1983), C4bp is a seven-tentacled species, with each flexible a m 33 nm long and 3 nm wide. From X-ray scattering and hydrodynamic analysis (Perkins et al., 1986) and the EM data (Dahlback et al., 1983), the SCRs are arranged about 4.5 nm from center to center. The SCR structure features four invariant cysteine residues, and reduction of any RCA protein will result in a dramatic change in electrophoretic mobility (Dykman et al., 1983a; Van Dyne et al., 1987). From structural studies of C4bp (Janatova and Reid, 1987), as well as 02-glycoprotein 1 (Lozier et al. 1984), the cysteine pairs at positions 2 and 45 and at positions 31 and 58 are cross-linked. More recently, Fourier transform infrared spectroscopic analysis of H (Perkins et al., 1988), a monomeric protein that is composed entirely of 20 SCRs (Ripoche et a l . , 1988), has resolved this picture further. In that study it was seen that SCRs are predominantly in 0-sheet conformation. This was supported by secondary structure averaging of 101 SCR sequences which predicted that residues 21-51 in the 61-residue repeat constitute four strands of structure and four p turns. Carrying this further, Perkins et al. (1988) proposed a model for the secondary structure of the SCR. By this model, the SCR is a typical globular polypeptide form, much of it being an antiparallel 0-sheet structure that is folded into a sandwich and held in place by the disulfide linkages in the hydrophilic center.
B. LOCALIZATION OF ACHVESITES A full understanding of the function and evolution of the RCA proteins requires a detailed biochemical picture of their active sites. This
REGULATORS OF COMPLEMENT ACTIVATION
393
picture must account for different interactions (cofactor activity, decayaccelerating activity, and receptor activity) and different ligands (CSb, C4b, CBbi, C3d,g, etc.). The discovery that the RCA proteins share a common structural motif strongly suggests that the interactions of the RCA proteins with C3b/C4b and their derivatives share common structural requirements and that differences in ligand specificity and activity may be due to subtle variations in that common structural theme. The observation that all of the major in vitro activities of the RCA proteins are substantially enhanced by low salt concentrations may indicate the importance of ionic bonding in these interactions. The fact that all of the RCA proteins are composed of multiple SCRs (Fig. 4) leads to several simple questions involving the possible role of multiple SCRs in the determination of specificity: Is each active site on an RCA protein mediated by a single SCR, or are several SCRs directly involved? Do the interactions between an RCA protein and a CSb/C4b derivative occur at a single active site on the RCA protein, or do multiple active sites occur? Does one active site mediate more than one activity (i.e., decay acceleration and cofactor activity in CR1) or ligand (i.e., C4b and C3b in DAF), or are there different active sites for each activity and each ligand? The characterization of the RCA active sites is currently the subject of much interest. The available evidence suggests in some cases the presence of active sites near the amino-terminal ends of SCR arrays: in the EM study of C4bp, it was shown that each C4bp arm may attach to C4b by its peripheral (amino-terminal) half (Dahlback et al., 1983; Dahlback and Muller-Eberhard, 1984). Similar conclusions can be drawn from a study of the proteolytic fragments of H. In that case the CSb-binding activity and the coenzyme activity can be found in a proteolytic fragment (Alsenz et al., 1985) which is derived from the amino-terminal 5.5 SCRs of the protein (Alsenz et al., 1985; Schulz et al., 1986; Ripoche et al., 1988). Of interest is a comparison between human and mouse C4bp (Kristensen et al., 1987). In that study it was shown that the mouse protein is 51% homologous to the human counterpart except that it lacks equivalents to the human SCRs 5 and 6. Since the mouse protein binds to human C4b as well as it does to mouse C4b (Ferreira et al., 1977, 1978; Kaidoh et al., 1981; Kai et al., 1980) and it is an active cofactor in the cleavage of human C4bp mediated by mouse I (Kai et al., 1980), it is likely that human SCRs 5 and 6 do not play essential roles in these activities. These conclusions are also notable in light of the tentative assignment by Chung and Reid (1985), based on analysis of proteolytic fragments,
394
DENNIS HOURCADE ET AL.
of the cofactor active site to the human C4bp residues 177-332. Since most of this region has no corresponding region in the mouse C4bp (where homologues to human residues 248-375 are lacking), the active site could be limited to residues 177-247, which is little more than SCR 4. In addition, Chung and Reid (1985) identified a site in regions 332-395 of importance to C4b binding in an in Vitro competition assay. Since this region is nearly deleted from the mouse protein, it is not likely to play a major role in mvo (Kristensen et al., 1987). In view of the gene duplication that must have occurred during the evolution of these genes (see later), it would not be surprising if there were multiple sites in the larger proteins which could manifest biochemical activities in mlro, although there may still be one site per molecule that is responsible for most biological activity. In addition to the proteins of the RCA cluster, several other complement proteins that interact with C3b or C4b contain SCRs. The homologous proteins C2 (Bentley, 1986) and B (Morley and Campbell, 1984; Campbell et al., 1984) both contain three SCRs at their aminoterminal ends. C2b and Ba, derived from C2 and B, each carry the SCRs (along with about 10 kDa of non-SCR polypeptide) and have been shown to bind C4b (Oglesby et al., 1988) and C3b (Ueda et al., 1987), respectively. Clr and Cls, homologous proteins that are components of the C1 complex, both contain two tandem internal SCRs (Journet and Tosi, 1986; Tosi et al., 1987). It has also been proposed that factor I carries a single highly divergent SCR (Goldberger et al., 1987). Several noncomplement proteins also contain SCRs. T h e /3 subunit of clotting factor XI11 is composed of 10 contiguous SCRs beginning with the amino-terminus and followed by a non-SCR segment (Ichinose et al., 1986). &-glycoprotein I consists of five SCRs followed by a carboxy terminal domain (Lozier et al., 1984), while the interleukin 2 receptor contains two WRs separated by 37 amino acids and followed by an extracellular domain, an intermembrane segment, and a cytoplasmic domain (Shimuzu et al., 1985). Human haptoglobin also contains an SCR (Kurovsky et al., 1980). Thus far, little is known about how any of these SCRs function. The binding site(s) for RCA proteins within C3b/C4b are known only in a general way. Primarily through the work of Lambris and colleagues (Becherer and Lambris, 1987), peptides of C3 that interact with CRl and CR2 have been defined but the actual binding sites are unknown. VI. Variations in Structure and Expression of the RCA Proteins
A. TISSUE-SPECIFIC EXPRESSION Targeting of regulatory and other functions of the members of the RCA cluster is due in large part to cell and tissue-specific regulation
395
REGULATORS OF COMPLEMENT ACTIVATION
of gene expression. Although no detailed analysis is yet available of the nucleotide structure of the promoter/enhancer elements of the genes, immunochemical means of detection of the gene products allow certain generalizationsto be made. The six members can be subdivided into three primary categories: (1) cell surface autoregulatory proteins (DAF and MCP), (2) membrane receptors (CR1 and CR2), and (3) serum regulatory proteins (H and C4-binding protein). In general, these three categories also parallel the specific tissue expression. The most divergence occurs, though, in CR2 expression, as it is more restricted than CR1 in its distribution on peripheral blood cells. In this section we discuss the regulation of expression in these three categories. First, DAF and MCP are widely distributed on circulating cells (reviewed in Medof, 1988; Lublin and Atkinson, 1989; see summary in Table 11). With the exception of natural killer cells, DAF is expressed in detectable levels on all circulating cells (Nicholson-Welleret al., 1986), including erythrocytes and platelets. Little is known of the specific levels of expression during ontogeny of these cells; however, DAF has been detected on erythroid and mononuclear progenitors (Moore et al., 1985). In addition to circulating and bone marrow cells, DAF is also found on a wide variety of epithelial cells and cell lines as well as cultured endothelial and fibroblast cells (Medof et al., 1987a; Asch et al., 1985). The recent finding that DAF has a protective effect on survival of tumor cells may also indicate that DAF levels may be modulated by tumorpromoting viral or oncogene products (Cheung et al., 1987). To support this concept, surface expression of DAF can be increased in the monomyelocytic cell line HL-60 by treatment with phorbol esters or vitamin D (Lublin et al., 1986). Several molecular-weight forms of cell membrane and soluble DAF have been described (Medof et al., 1987a; Seya et al., 1987) and are likely due to posttranslational modifications, enzymatic degradation,
TABLE I1 RCA PROTEINS:BIOCHEMICAL ACTIVITIES Protein
Primary ligand(s)
Receptor activity
Decay acceleration
H C4bp CR1 CR2 DAF MCP
C3b C4b C3b, C4b C3d,g C3b, C4b C3b, C4b
-
+ +
+ + -
+ -
+
-
Cofactor activity
+
+ + ?
-
+
396
DENNIS HOURCADE ET A L
or alternative RNA splicing (Seya et al., 1987; Medof et al., 1987a; Caras et al., 1987; Kinoshita et al., 1987). MCP also exhibits a wide tissue distribution (reviewed in Lublin and Atkinson, 1989; see also Table 11). Expression of MCP is not as well characterized as DAF; however, with the exception of erythrocytes (it is absent) and natural killer cells (it is present), MCP expression appears to parallel that for DAF (Cole et al., 1985; Yu et al., 1986; Seya et al., 1988; McNearney et a l . , 1988). MCP also exhibits an interesting polymorphism in the expression of its larger and smaller forms (Ballard et al., 1987). In a population survey of 74 unrelated individuals, three phenotypes were defined: upper band-predominant expression (65%), equal expression (29%), and lower band-predominant expression (6%). Family studies support a two-allelic system, as none of the 29 offspring from 10 matings had a phenotype that deviated from expected. The observation that both protein forms are made to some extent in all individuals suggests that the polymorphism occurs in a linked cis-acting regulatory element which modulates the relative expression of the upper and lower proteins (Ballard et al., 1987). Recent identification of MCP cDNA (Lublin et a l . , 1988) and genomic clones (Lublin et al., 1988) should allow an analysis of this expression of polymorphism at the molecular level. CR1 and CR2 are primarily expressed on circulating cell populations, though each also has been detected on other cell types (reviewed in Fearon, 1985; Ross and Medof, 1985; see also Table 11). CR1 is expressed on human erythrocytes, granulocytes, monocytes, a small subset of T cells (both CD4 and CD8 positive), glomerular podocytes, follicular dendritic cells, and B cells at various levels during ontogeny. Much of the analysis of the regulation of CR1 expression has been performed using erythrocytes. These investigations utilized anti-CR1 monoclonal and polyclonal antibodies (Medof and Nussenzweig, 1984; Holers et al., 1986; Wong et a l . , 1985) and took advantage of the CR1 size polymorphisms as defined by SDS-PAGE analysis (Dykman et al., 1983a, 1984, 1985; Wong et al., 1983). Based on this work, a number of points can be made. Levels of expression vary widely in the population, ranging from < 50 to > 1000 molecules/erythrocyte with a mean level of 500 (Iida et a l . , 1982; Wilson et al., 1982; Moldenhauer et al., 1987). The level of erythrocyte CR1 expression is regulated by a cis-acting regulatory element (de Cordoba and Rubinstein, 1986; Wilson et al., 1986a; Van Dyne et al., 1987). Restriction fragment-length polymorphism analysis using a partial CR1 cDNA correlated a 6.9-kb genomic Hind111 fragment to a low level of expression and a 7.4-kb fragment to a high level of expression (Wilson et al., 1986a), although considerable overlap occurs for CR1 numbederythrocyte among heterozygotes and those individuals
-
REGULATORS OF COMPLEMENT ACTIVATION
397
homozygous for the 7.4-kb fragment (Wilson et al., 1986a; Moldenhauer et al., 1987). Expression of CR1 in nonerythrocyte populations also varies considerably (Wilson et al., 1986b; Yoshida et al., 1987), probably due to cell-specificposttranscriptional and posttranslational mechanisms but also due to differential activity of cis-actingregulatory elements (Dykman et al., 1983a,b; Van Dyne et al., 1987). CRl expression can be experimentally modulated in HL-60 cells. Treatment with phorbol esters, vitamin D, dimethyl sulfoxide, and other compounds causes cell differentiation and concomitant surface and mRNA expression of CR1 (Lublin et al., 1986; Holers et al., 1987). Analysis of CR1 expression during B cell ontogeny has indicated that the majority of pre-B and almost all mature B cells express CR1 (Wilson et al., 1982). In addition, tissue monocytes and glomerular podocytes (Fischer et al., 1986) are recognized by anti-CR1 monoclonal antibodies. The glomerular product has been purified and found to have a molecular weight identical to the erythrocyte form (Fischer et al., 1986). CR2 expression is very limited as compared to CR1, DAF, or MCP (Cooper et al., 1988). On circulating cells, CR2 is expressed on mature B cells (Tedder et al., 1984). Stem cells, early pre-B cells, and plasma cells are negative for CR2 by fluorescence-activated cell sorting (FACS) analysis; however, the exact switch points are not well defined, and some late pre-B cells and pre-B cell EBV-transformedcell lines are CR2 positive. Some T cell tumor-derived cell lines are also CR2 positive; however, by FACS analysis, peripheral T cells are negative for CR2 expression (Tedder et al. , 1984). Utilizing polyclonal and monoclonal anti-CR2 antibodies and purified platelets, low-level expression on platelets has been reported (Nunez et al., 1987). However, another study utilizing lZ5Isurface labeling and iC3-Sepharose affinity chromatography and also highly purified platelets did not confirm these findings (Yu et al., 1986). CR2 is also expressed on follicular dendritic cells (Reynes et al., 1985) as well as in a stage-specificmanner on oropharyngeal and cervical epithelial cells. H and C4bp are both serum regulatory proteins. The predominant site of synthesis is the liver; however, a form of H has also been reported to be synthesized and cell membrane associated in the U937 monocytic cell line (Malhotra and Sim, 1985). Both molecular-weight forms of H are also synthesized and secreted by human skin fibroblasts (Katz and Strunk, 1988). Human monocytes and the human hepatoma cell line HepG2 did not synthesize detectable H. Treatment of fibroblasts with IFN-..I addition led to a marked increase of H mRNA and biosynthesis. Tumor necrosis factor (TNF) a and IFN-/3 also slightly increased H synthesis, as opposed to LPS, IFN-a, and IL-1, which had no effects (R.W. Strunk, personal communication).
398
DENNIS HOURCADE ET AL
Overall, it is tempting to speculate, given the clear evidence of genomic duplication during the evolution of the RCA family (see below) and several similarities in tissue-specific expression, that homologies in the transcriptional regulatory regions will also be found. One likely scenario involving DAF and MCP, for example, would be the finding of identical nucleotide sequences which bind ubiquitous transcriptional factors in conjunction with the specific transcriptional promoters, enhancers, or repressors which mediate stage-specific or lineage-specificexpression. One of the most interesting genes in this regard is CR2, whose expression is turned on, then off, as B cells progress through an ordered ontogeny. It will be of great interest to analyze expression of CR2 relative to other stage-specific markers of B cells. In addition, as H and C4-binding protein are both serum proteins with complementary functions for the classical and alternative pathway, similar regulation of expression would be expected. B. ALTERNATIVE RNA PROCESSING Alternative RNA processing (alternative splicing and/or polyadenylation) plays a role in expression of many genes and gene families. One of the best characterized is the switch from membrane to secreted IgM, which occurs by alternative splicing and exclusion of the sequence encoding the membrane-spanning domain of IgM from the mRNA (Mather et al., 1984). Examples of alternative RNA processing have been described in the RCA family for CR1, CR2, DAF, and H. We will discuss what is known in these genes and the potential or postulated role this mechanism plays in the RCA family. The majority of the evidence for alternative processing has come from analysis of cDNA clones, and only a partial understanding exists of the in vivo relevance of these observations. Alternative polyadenylation in the CR1 gene has been detected by analysis of the structure of CRl cDNA clones (Holers et al., 1987; Hourcade et al., 1988). One class of clones encodes the amino-terminal 8.5 SCRs followed by a stop signal, 3’ untranslated sequence, and a typical polyadenylation signal and tail. Sequence analysis of genomic subclones has determined that the SCR and 3’ untranslated cDNA sequences are contiguous; therefore, this class arises from alternate use of an intronic polyadenylation signal. A subset of poly-A+ mRNAs, which is smaller than the full-length 10.0- and 11.6-kb CR1 mRNAs, is specifically recognized by a probe derived from the 3 ’ untranslated portion of the cDNA. Lacking a transmembrane domain, this class of cDNAs could direct the synthesis of a secreted form of CRI. A soluble form of CR1 has been described (Yoon and Fearon, 1985); however, this product
REGULATORS OF COMPLEMENT ACTIVATION
399
comigrated with the erythrocyte 200-kDa receptor. The putative secreted CR1 would be -60 kDa. Experiments are in progress to assess the biological relevance of this form of CR1. Alternative processing in CR2 gene expression can be inferred, though not yet ripmusly proved, by results from different laboratories in obtaining cDNA clones. Full-length clones with both 15 and 16 SCRs have been identified, some from the same mRNA source (Weis et al., 1988; Moore et al., 1987; Fujkaku et al., 1989). Whether this reflects alternative splicing or allelic differences is currently unknown. Cosmid clones have been identified which encode exons containing the 16 SCRs (Fujisaku et al., 1989), each of which contains appropriate splice donor and acceptor sites. The best evidence for alternative RNA processing playing a physiological role has been determined by analysis of cDNA clones for H. One class of cDNA clones isolated encodes the amino-terminal seven SCRs of H, followed by a stop signal, 3’ untranslated sequence, and a polyadenylation signal and tail (Schulz et al., 1986; Kristensen et al., 1986; Ripoche et al., 1988). Two major bands of 4.3 and 1.8 kb are recognized on Northern blots of poly-A+ liver mRNA. The 1.8-kb band appears to be specific for this class of smaller cDNAs (Schwaeble et al., 1987). In addition, partial tryptic analysis and epitope mapping of H has localized CSb-binding and cofactor activity to this amino-terminal portion of H (Alsenz et al., 1985). Antisera to H have detected smaller serum forms (Schwaeble et al., 1987) and biosynthetically labeled fibroblast-derived forms (Katz and Strunk, 1988) that are of the appropriate size to be encoded by this smaller transcript. Thus, at least two lines of evidence support alternative RNA splicing as playing a physiological role in H expression in humans. C. DIFFERENT ANCHORING SYSTEMS CR1, CR2, and MCP are typical type I membrane glycoproteins in that they each possess a hydrophobic peptide, a charged stop transfer signal, and a cytoplasmic tail at their COOH terminus (Klickstein et al., 198713; Weis et al., 1988; Moore et al., 1987; Lublin et al., 1988). The exception among RCA proteins is DAF, which is anchored by a glycolipid (Caras et al., 1987; Medof et al., 1987b). This type of an anchor accounts for the remarkable observation that purified membrane DAF, when added to a cell suspension, reincorporates and displays functional activity (Medof and Nussenzweig, 1984), as well as the fact that DAF is released from erythrocytes and leukocytes by phosphatidyl inositol-specific phospholipase C (PIPLC) (Davitz et al., 1986). In the latter case, PIPLCreleased DAF loses its ability to reinsert. The deduced amino acid sequence of DAF ends in a 24-amino acid
400
DENNIS HOURCADE ET A L
segment with markedly hydrophobic character (Caras et al., 1987; Medof et al., 1987b). It does not possess a series of basic residues to serve as an anchor such as those found in most membrane proteins, including CR1, CR2, and MCP. By analogy with several other similarly anchored proteins, part or all of this hydrophobic segment is removed and the glycolipid tail is covalently attached to the resulting COOH-terminal amino acid (reviewed in Low and Saltiel, 1988). This appears to occur very rapidly following translation and before the molecule is transported to the Golgi apparatus. The hydrophobic extension peptide may provide a temporary membrane anchor in the endoplasmic reticulum. Preliminary evidence suggests that this type of anchor provides greater lateral mobility in the plane of the membrane (Thomas et al., 1987; Ishihara et al., 1987; Noda et al., 1987). This increased mobility could be an advantage to a protein such as DAF, in which several hundred molecules must patrol the large surface of a cell, for example, an erythrocyte. Other possible roles for such an anchor could be as a means to release the protein from the cell membrane or the transduction of an intracellular signal (reviewed in Low and Saltiel, 1988). In the acquired hemolytic disorder PNH, the deficiency in DAF has been correlated to a defect in the modification of the DAF COOH terminus. The hemolytic defect can be corrected in some forms of PNH by the addition of normal DAF (Medof and Nussenzweig, 1984). The structural gene for DAF appears normal in PNH, and DAF mRNA is produced in normal to increased amounts (Stafford et a l . , 1988). Since several other glycophospholipid anchor proteins are also missing from the cells of these patients (Medof et al., 1987a; Medof, 1988), the defect is likely to be in the biosynthetic pathway for the synthesis or attachment of the glycophospholipid. D. DIFFERENTIAL GLYCOSYLATION All members of the RCA gene family are glycoproteins. CR1 (Lublin et al., 1986), CR2 (Wek and Fearon, 1985), C4bp (Campbell et al., 1988), and H (Sim and DiScipio, 1982; Ripoche et al., 1988) possess only N-linked CHO moieties. DAF (Lublin et al., 1986; Medof et al., 1987a) and MCP (Yu et al., 1986; Ballard et al., 1987) possess both N- and 0-linked sugars. These data have been derived by chemical, enzymatic, and biosynthetic analyses (see Table 111). The biosynthetic pathway for glycosylation of these proteins is similar to that described for other mammalian glycoproteins (reviewed in Kornfeld and Kornfeld, 1980). The precursors of these N-linked sugars are attached cotranslationally en bloc as a high-mannose (simple sugar) unit to asparagine residues. They are then processed in the Golgi apparatus
REGULATORS OF COMPLEMENT ACTIVATION
401
TABLE 111 GLYCOSYLATION OF RCA PROTEINS Protein
N-linked4
0-linked
Other features
H C4bp CR1
+ + +
-
Only three conserved in mouse factor H 3 per 70,000 monomer CHO accounts for cell-specific variation in molecular weight. Nonglycosylated receptor has reduced affinity for CSb, half-life on plasma membrane, and intracellular survival; lactosaminoglycans Half-life of aglycosylated receptor reduced on the plasma membrane Ser/Thr-rich region Ser/Thr-rich region
CR2
+ (8-11)
-
DAF MCP
+ (1) + (3)
+ +
(6 or 7) (3) (7-10)
ONumber of units is given in parentheses.
by trimming some of the mannoses and by enzymatic attachment of N-acetylgalactosaminesand terminal sialation (complex sugar). A core fucose is often added and the degree of branching of the complex units is variable. The 0-linked units are attached in the Golgi apparatus to serine and threonine residues (see Fig. 5 ) . Of interest is that DAF and MCP each possess serine/ threonine-rich regions near their membrane anchors (Caras et al., 1987; Medof et al.,
N-LINKED
SP
SP
G,aI
G,"I
GlcYAc M a c
GlcYAc Man
M Y n l GlcTJAc GlcrJAc-Fuc Asn
0-LINKED
SP G,aI SA-GalvAc SerlThr
FIG. 5 . Structure of N-linked and 0-linked oligosaccharides. The precursor to N-linked sugars (high-mannose units) consists of nine mannoses linked to the N-acetylglucosaminecore. After processing of the high-mannose precursor, the degree of branching at the two mannoses varies from two (as shown in the diagram) to four. The fucose unit is also variably present. Lactosaminoglycans are N-linked oligosaccharides whose outer sequences contain a repeating lactose rather than sialic acid. SA. Sialic acid; Gal, galactose; GlcNAc, N-acetylglucosamine; Man, mannose; Fuc, fucose. 0-Linked units may be attached to isolated or clustered Ser/Thr.
402
DENNIS HOURCADE ET AL.
1987b; Lublin et al., 1988). These segments consist of 70 (DAF) and 27 (MCP) amino acids and 45% of the residues are serines or threonines. Such regions are typically rich in O-linked glycosylation. A role for O-linked glycosylation within DAF and MCP is yet to be determined. Of interest, several other membrane proteins, including the IL-2 receptor (Greene and Leonard, 1986; Waldmann, 1986) and low-density lipoprotein (LDL) receptor, have 0 -rich regions near their transmembrane-spanning domains. This region (Davis et al., 1986)was deleted from the LDL receptor, but no functional defect was noted. In another study, all of the O-linkedglycosylation was blocked and the receptor was unstable (Kingsley et a l . , 1986). The authors suggested that this type of glycosylation may protect the LDL receptor from proteolytic digestion. It is not known what percentage of the O-linked sugars in DAF or MCP are in the serine/threonine-rich region compared to those distributed in other parts of the molecule. The biosynthesis and glycosylation of the RCA proteins have been investigated in detail for CR1 and, to a lesser degree, for CR2, DAF, and MCP. Several interesting features related to the glycosylation of CR1 (Lublin et al., 1986) are as follows: (1) The molecular-weight differences of -5 kDa by SDS-PAGE analysis between CR1 of erythrocytes and peripheral blood granulocytes is accounted for by variation in glycosylation. Also, the HL-60 human promyelocytic tumor cell line can be induced to pursue granulocyte or monocyte developmental pathways. In these cases the mature CR1 produced is 10 or 15 kDa larger, respectively, than erythrocytic CR1, even though their aglycosylated precursors (as well as those of human leukocytes) have identical molecular weights. Variations in the structure of complex oligosaccharideunits correlates with the differences in molecular weights. (2) Erythrocytic CRl and probably CRl of most other cells are fucosylated and also possess relatively large amounts of lactosoaminoglycans. For CR1 and CR2, functions have been assigned to the CHO moieties. CR1 was synthesized without oligosaccharide (in the presence of tunicamycin, an inhibitor of N-linked glycosylation; Lublin et al., 1986). This nonglycosylated form of CR1 had twice the turnover rate (3.6 versus 7.9 h) of normal CR1. The nonglycosylated protein did reach the membrane, but its efficiency in this regard was much less than the glycosylated molecule. Finally, nonglycosylated CR1 had 41% of the binding efficiency to CSb-Sepharose of glycosylated CR1. Of interest, the precursor of CR1 (containing high-mannose units) had 98% of the binding efficiency of the mature receptor (complex units). These data suggest the carbohydrate of CR1 stabilizes CR1 as it moves intracellularly and while it is in the membrane. Furthermore, although CHO is important
REGULATORS OF COMPLEMENT ACTIVATION
40 3
for ligand binding, the chemical identity of the unit is not necessarily of great importance. In the case of CR2 (Weis and Fearon, 1985), CHO residues do not appear to be necessary for ligand binding or plasma membrane expression, but the half-life of the nonglycosylated CR2 on the plasma membrane is considerably reduced. These results can be explained by an effect of oligosaccharide on the conformation of CR1 and CR2, especially in their resistance to proteolytic degradation. VII. Organization and Evolution of the RCA Genes
A. ORGANIZATION OF THE RCA CLUSTER It was not apparent that RCA proteins were closely related when they were first being isolated and characterized. Factor H and C4bp were very different in molecular weight and each had its own ligand specificity, while the receptor, CR1, had not yet appeared structurally similar to either plasma protein (Tables I and 11). The earliest evidence of their common evolutionary origins came through a collaboration between our group in St. Louis and that of Santiago Rodriguez de Cordoba and Pablo Rubinstein and their collaborators at the New York Blood Center. In that investigation, classical genetic methods utilizing polymorphic protein variants established that the genes encoding H, C4bp, and CRl are tightly clustered in the human genome (de Cordoba et al., 1985). This finding prompted us to write a review on the functional, structural, and genetic aspects of these proteins and to propose that they were members of a new multigene family derived from a common ancestral C3b-binding protein (Holers et al., 1985). We also suggested, based largely on functional similarities, that DAF and MCP would be members of this family. Soon afterward, the derived primary structures of all of these proteins showed their common structural motif, thus supporting our original proposal and carrying it further. The C3bK4b-binding proteins are now seen largely as the evolutionary result of duplication and divergence of an ancestral SCR-encoding genetic element (Reid et al., 1986). Genetic mapping studies were extended once cDNA probes were available for in situ hybridization: It was thus shown that CR1, CR2, DAF, and MCP probes all hybridize to band q32 on the long arm of chromosome 1 (Weis et al., 1987; Lublin et al., 1987, 1988). The use of pulsed-field gel electrophoresis has recently given the first detailed picture of the RCA cluster (Fig. 6). Presently, a physical map can be drawn with the rough placement of the genes encoding C4bp, CR1, CR2, DAF (Rey-Campos et al., 1988; Carroll et al., 1988). and
404
DENNIS HOURCADE ET AL.
s
MNSf
s s
Sf MSf
I
I )I
I
f
I
I
M
Sf I
I
__ MCP ____ CR1___- CR2 - DAF _----_C4Bp________
I
0
I
I
I
I
I
500
I
I
I
1
1
1000
I
I
I
kb
FIG.6 . Physical map of RCA cluster. Data from Rey-Campos et a l . , 1988; Carroll et al., 1988; Bora et al., 1988.
MCP (N. Bora and J. P. Atkinson, unpublished), all within 900 kb. While genetic analysis indicates that the H gene must also be in this general vicinity, it does not appear within 1500 kb of these genes. Several of the other genes that encode SCRs have been mapped in the genome. B and C2 are located at the class I11 region of the major histocompatibility complex on the short arm of human chromosome 6 (Weitkamp and Lamm, 1982), while Clr and Cls lie on chromosome 12 (Tosi et al., 1987). Both pairs are likely to be the result of gene duplication. The IL-2 receptor gene is found on chromosome 10 (Green and Leonard, 1986) and the factor I gene is found on chromosome 4 (Goldberger et al., 1987).
B. EXONSTRUCTURE Many of the genes within the RCA cluster (as well as those encoding other SCR-containingproteins) have been cloned or partially cloned and these sequences are now being characterized. In general it is found that each SCR is encoded by a single separate exon (Kristensen et al., 1987). There are a number of exceptions to this rule, most following a single "split" pattern. Split exons are found in genes encoding CRl (Hourcade et al., 1988), CR2 (Holers, 1989), mouse H (Vik et al., 1987), mouse C4bp (Barnum et al., 1987), and a CR1-likegenomic region (Hourcade et al., 1988), and probably will be found in other RCA genes. In all of these cases, an intervening sequence interrupts the usual exon organization in the triplet encoding Gly 34 (Fig. 4) and, therefore, it is likely that all these exon pairs arose through a common ancester. One other split exon has been found in the human haptoglobin gene (Maeda, 1985). In that case the split is found at a different site in the SCR motif, and therefore probably arose independently from the rest of the split exons. Two models can account for the split exons. A primordial exon encoding an entire SCR could have been interrupted through transposition by an intervening sequenc:. Alternatively, the primordial SCR may have been encoded by two separate but adjacent exons with intron
REGULATORS OF COMPLEMENT ACTIVATION
405
deletion leading to the predominant composite (unsplit) exons. Since two different split patterns have been described, the simplest model would be the first one, with the primordial SCR encoded by a single exon and the two split patterns arising from two different insertions. The second model cannot be precluded because, in that case, once a composite exon was formed, a second split exon could be produced by insertion. In either case, it appears that insertion of an intervening sequence into an exon occurred at least once. It is of interest that all of the characterized RCA genes contain at least one exon “split” at Gly 34, while no other SCRcontaining gene has been found to carry such an exon. It suggests that the RCA cluster may have evolved from a C3b/C4bp gene that contained an ancestral split exon. It is also possible that there is an RCA-specific activity encoded by the split exon. In addition, exons encoding two full SCRs have been shown. They occur in CR2 (Holers, 1989) and the CR1-like sequence (D. Hourcade and J. P. Atkinson, unpublished results) and may be expected in CR1, given the close relationship between CR1 and the CR1-likegenomic region (see below). C. GENEDUPLICATION It is evident that gene duplication has played an important part in the evolutionary development of the RCA cluster. This can be deduced by comparison of amino acid sequences of the different proteins (see, for example, Weis et al., 1988). A relatively recent case appears to be that of CR1 and a CR1-like partial replica (Wong et al., 1987; Hourcade et al., 1988): A CR1-like genomic region of at least 40-50 kb in size has been isolated using CR1 cDNA probes at high stringency. Sequence analysis has shown potential exons which could encode a polypeptide very similar to the amino-terminal region of CR1, faithful to the SCR motif and homologous to CRl cDNA at the 95 % level (D. Hourcade and J. P. Atkinson, unpublished results). Divergence from the CR1 sequence occurs mainly in the nonconserved SCR residues. Some corresponding intron sequence is also known and it, too, exhibits 95% homology. Since this sequence hybridizes to the probes used in the pulsedfield gel electrophoresis mapping (Fig. 6), it must be located near CR1 at the RCA cluster. Although the products of the CR1-like genomic region have yet to be characterized, the conservation of the SCR motif in potential coding regions strongly suggests that selection at the protein level has guided its divergence from CR1. In any case, here is a clear indication of gene duplication in the RCA cluster.
406
DENNIS HOURCADE ET AL.
D. INTRAGENIC DUPLICATION It is clear that intragenic duplications have also occurred in the RCA cluster. In the case of CR2, containing 16 SCRs, the protein can be organized into four groups of repeats by analysis of the derived amino acid sequence. Thus, SCRs 1-4 are homologous to SCRs 5-8, 9-12, and 13-15, and possibly SCR 16, suggesting that duplication of a four-SCR element occurred in the evolution of CR2 (Weis et al., 1988; Moore et al., 1987). This model is strongly supported by the exon structure of CR2 (Fujisaku et al., 1989). The first four SCRs are encoded by an exon carrying two SCRs followed by an unsplit exon and then a split exon, and this pattern repeats through SCR 15. Intragenic duplication has also been part of the evolution of CRl. In that case the first 28 of 30 SCRs of the primary allotype (190 kDa; Fig. 7A) can be organized into four long homologous repeats (LHRs A, B, C, and D (Fig. 7B)-that are 70-99% homologous at the amino acid level (Klickstein et al., 1987a,b;Hourcade et al., 1988). In addition, the SCRs of CR1 can be organized in a different form (Fig. 7C), which utilizes degree of intramolecular homology as a physical basis for assigning separate regions. Thus, region 11, composed of 16 SCRs, exhibits internal homology of the seven-SCR pattern at the 99% level, while region 111, composed of 10 SCRs, exhibits internal homology at the 91% level. Region I is composed of two SCRs and is not long enough to carry a
A
B
S
TM
30
A
B
I
C
IC
D
4
C FIG. 7. Schematic representation of CR1. S, Signal peptide; T M , transmembrane domain; IC, intracytoplasmic domain. Numbers refer to SCRs (A) in the different LHRs (B) and homology regions (C). (Reproduced from Hourcade et al., 1988.)
REGULATORS OF COMPLEMENT ACTIVATION
407
LHR, but it is 61 and 59% homologous to regions I1 and 111, respectively. Region I1 is 67% homologous to region 111. The complex homologies between the SCRs suggest specific evolutionary relationships. While it is clear that duplication of a seven-SCR ancestral element was instrumental in the evolution of CRl, the strict repetitions in regions I1 and I11 are suggestive of further unequal crossingover and/or gene conversion. In addition, several CR1 allotypes have been found that differ in apparent molecular weight from the predominant form by a multiple of 30 kDa (Table I) (Dykman et al., 1983a,b, 1984, 1985; Wong et al., 1983). Tryptic fragment analysis (Nickells et al., 1986) and mRNA analysis (Holers et al., 1987; Wong et a l . , 1986) of four different allotypes and genomic analysis (Wong et al., 1986) of the two most prevalent allotypes (190 and 220 kDa) indicate that these fonns could differ by multiples of an internal repeat. These allotypes may have evolved by addition or deletion of multiples of seven SCRs in homology region I1 or 111. It is interesting to note the contrast between the repeating structure of CR2, in which the repeats appear to have diverged once duplication occurred, and CRl, in which parts of the repeating pattern evolved in a rigorously concerted fashion. The extensive homology and multiple polymorphisms in CR1 may simply be the consequences of very recent duplications. Further evolution of CR1 may eventually lead to one major polymorphic form and a level of internal repetition more consistent with that seen in CR2. On the other hand, the contrast between contemporary CR1 and CR2 may indicate different functional constraints. For example, continued unequal crossing-over and/or gene conversion in CR1 could allow multiple binding sites to evolve together in homology region I1 or in homology region 111. Alternatively, the product of concerted evolution by unequal crossing-overwould by necessity be the continuous generation of allotypes that differ in the length of their extracellular regions. It is possible that there may be an advantage for a cell to carry CR1 molecules of two different extracellular lengths, perhaps allowing more efficient function as a receptor. E. ALTERNATIVE POLYADENYLATION As described above, evidence has been obtained for the presence of an alternative polyadenylation site in the CR1 transcriptional unit, located in the intron which splits SCR 9. The utilization of this site leads to the synthesis of a truncated RNA which could encode a secreted form of CR1, 8.5 SCRs in length. It has been suggested that alternative polyadenylation may have played a role in the generation of secreted
408
DENNIS HOURCADE ET AL.
proteins from receptor genes (Early et al., 1980; Hourcade et al., 1988). A receptor gene such as CR1, bearing an alternative polyadenylation site, could have been duplicated and one of the products could have diverged to produce a secreted form alone (Fig. 8). This process could have been instrumental in the evolution of RCA plasma proteins from primordial C3b/C4b receptors. F. THEEVOLUTION OF DIVERGENT FUNCTION
The evolution of function at the RCA cluster clearly involved shuffling of non-SCR domains (Gilbert, 1978, 1979). Non-SCR domains mediate the cytoplasmic functions of the receptors, the quaternary structure of C4bp, serine/threonine-rich regions in DAF and MCP, secretion, and membrane anchoring (Fig. 4). Of special interest is the evolution of functions mediated by the SCRs. The evolution of the RCA cluster by necessity involved modification of SCRs to account for varying molecular specificity, but beyond that little is known. Moreover, little is known of the function of SCRs in the noncomplement proteins. The results of structure/function analyses, underway in a number of laboratories, will lead not only to the molecular basis for different interactions, but will also provide a scenario of their evolution from a common structural/functional domain. Finally, it is interesting that the RCA genes have remained in a cluster,
S
4
C
lntron Insertion
AATAAA
S
C
4
Primitive Receptor Gene
\
Alternative Polyadenylation Both Secreted & Receptor Forms Gene Duplication
AATAAA
AACAA
Y
C
S
1
C
S
1
Deletion
FIG. 8. Alternative polyadenylation in the evolution of secreted proteins from receptor genes. S, Signal peptide; C, transmembrane region and cytoplasmic domain.
REGULATORS OF COMPLEMENT ACTIVATION
409
because, in contrast, many gene families are composed of members scattered throughout the genome. At this time, however, there is no evidence for a common control mechanism that would require the close physical proximity of the RCA genes. In any case, as a cluster, the RCA genes appear to have retained the capacity for rapid evolution through unequal crossing-over and in this manner further evolution of the complement regulatory network may continue. By this reasoning, substantial differences in the organization of the RCA cluster may be expected in other species. VIII. Utilization of RCA-Like Proteins by Human Pathogens
The complement system stands as one of our most important defense mechanisms against biological pathogens. The formation of convertases on microbial surfaces by the alternative and classical pathways, and the resulting accelerated deposition of C3b, usually leads to the phagocytosis and destruction of the intruder (Fig. 1). In principle, there should be substantial selective advantage for a pathogen to interfere with this complement activity, and it might be surprising if the RCA genes, involved with the protection of host tissues from complement, were not utilized by some pathogens for their own protection. There are few well-documented examples where bacteria, fungi, or viruses have been shown to possess the means of repressing the normal activities of the complement system (reviewed for bacteria in Frank et al., 1984; Joiner et al., 1984). Two recent examples of this type of activity are of special note because they suggest a direct evolutionary relationship between the RCA cluster and foreign (viral) genes. In the case of vaccinia virus, the viral genome encodes a protein that could easily be mistaken for a member of the RCA family (Kotwal and Moss,1988). A secreted 35K protein has been purified from the medium of vaccinia-infected cells and its complete amino acid sequence has been determined indirectly by correlation with its genomic coding region. By this method it is seen that the 35K vaccinia protein, consisting of a signal sequence followed by four SCRs, is 38% homologous at the amino acid level to the amino-terminal half of C4bp. It is likely that active sites for C4bp expression are located in this half of the protein (see Section V,B). Thus, this vaccinia protein appears to have evolved directly from the host C4bp gene and could have retained some of its biochemical activity. In the case of the herpes viruses, the structural relationship to the RCA family is not as clear as that of vaccinia, but the functional studies more directly illustrate the potential of such viral proteins to promote pathogenicity. Herpes simplex virus type I (HSV-1) encodes a CSb-binding
410
DENNIS HOURCADE ET A L .
protein, glycoprotein C-1, that is an envelope protein of the virus and also appears in large quantities on the surface of the infected cell (Cines et al., 1982; Friedman et al., 1984; Smiley and Friedman, 1985; Smiley et al., 1985; Kubota et al., 1987; McNearney et a.!., 1987). Glycoprotein C-1 reacts with a monoclonal antibody raised to CRl (Kubota et a l . , 1987), contains short stretches of substantial amino acid sequence homology to RCA proteins (V. M. Holers, unpublished data), although no typical SCR structure, and harbors decay-accelerating activity for the alternative pathway convertase (Fries et al., 1986). Herpes simplex virus type 2 encodes a protein homologous to HSV-1 glycoprotein C-1 (Zazulak et al., 1984), termed HSV-2 glycoprotein C-2, that also binds C3b (McNearney et al., 1987). The presence of either the HSV-1 glycoprotein C-l gene or the HSV-2 glycoprotein C-2 gene in HSV recombinants affords protection against in vitro complement-mediated neutralization (McNearney et al., 1987) of the virion. In a similar manner one would predict that glycoprotein C-1 or glycoprotein C-2 on cell surfaces could provide protection for the infected cell. In addition to encoding their own complement-regulating proteins, pathogens may also manipulate the regulation of the host RCA proteins. In the case of SV40, transformation of fetal lung fibroblast lines has resulted in a five-fold increase in MCP when assayed by either the appearance of surface-labeled protein or by the occurrence of cofactor activity (T. A. McNearney and J. P. Atkinson, unpublished observations). In addition, the ratio of the two MCP polymorphic forms changes markedly. IX. Conclusion
Complement activation is modulated by a family of at least six regulatory genes. The regulators of complement activation proteins encoded by these genes exert their influence mainly through interactions with C3b and C4b, these proteins being essential for marking target tissues and immune complexes for disposal and destruction. Three major interactions of the RCA proteins have been described: decay acceleration hastens the dissociation of the C3 convertases that produce the C3b molecules; cofactor activity irreversibly inhibits C3b and C4b effects through I-mediated proteolysis of target-bound and fluid-phase C3b and C4b; and C3b and C4b receptors are involved in the binding, transport, and clearance of foreign material in the spleen and liver, the mediation of endocytosis, and in signal transduction. Through these biochemical interactions the RCA proteins play major roles in the protection of cells from complement (and, thus, in
REGULATORS OF COMPLEMENT ACTIVATION
411
distinguishing self from nonself), in regulation of complement activation, in processing of immune complexes, microbes, and other foreign materials, and in the activation of immune cells. Recently, molecular analysis has established several remarkable features of the RCA genes: All six genes are clustered at one chromosomal location and all protein products are composed primarily of a single repeated 60-amino acid domain. Interactions with C3b and C4b and their derivatives must occur in these repeats. Most repeated domains are each encoded by a single exon and evolution of the RCA genes must have involved repeated gene duplication. Current research is underway to understand the structure and evolution of the RCA active sites, and to learn how from a common structural framework there evolved a complex family of genes, each with different capacities for biochemical and cellular activity. ACKNOWLEDGMENTS We thank our past and present collaborators for their numerous critical contributions to our own efforts in this field, R. W. Strunk, T. A. McNearney, and N. Bora for allowing us to refer to their unpublished observations, Ted Post for valuable discussions, and Pat Parvin and Lorraine Whiteley for excellent secretarial assistance.
REFERENCES Abrahamson, D. R., and Fearon, D. T.(1983). Lab. Invest. 48, 162. Alsenz, J., Schulz, T. E, Lambris, J. D., Sim, R. B., and Dierich, M. .’F (1985). Bi0chem.J. 232. 841. Asch, A . , Kinoshita, T., and Nussenzweig, V. (1985).J. Erp. Med. 162, 221. Atkinson, J. P., and Eisen, H. (1989). In “Microbiology”(H. Eisen, ed.). Lippincott, Philadelphia, Pennsylvania (in press). Atkinson, J. P., and Farries, T. (1987). Immunol. Today 8, 212. Ballard, L., Seya, T., Teckman, J. Lublin, D. M., and Atkinson, J. l? (1987).J.Zmmunol.
138, 3850.
Barnum, S., Keeney, J., Kristensen, T., Noack, D., Selfin, M., D’Eustachio, P., Chaplin, D., and Tack, B. F. (1987). Complement 4, 131. Becherer, J. D., and Lambris, J. D. (1987). Complement 4, 132. Bentley, D. R. (1986). Bi0chem.J. 239, 339. Berger, M., and Cross, A. S. (1984). Immunology 51, 431. Bohnsack, J. F., Takahashi, T., and Brown, E. J. (1986).J. Immunol. 136, 3793. Bora, N. S., Lublin, D. M., Kumar, B. V.,Hockett, R. D., Holers, V. M., and Atkinson, J. P. (1988).J. Exp. Med. (in press). Campbell, R. D., Bentley, D. R., and Morley, B. J. (1984).Philos. 7hzns. R . SOC. London, Sez B. 306, 367. Campbell, R. D., Law, S. K. A., Reid, K. B. M., and Sim, R. B. (1988). Annu. Rev. Immunol. 6, 161. Caras, I. W., Davitz, M. A,, Rhe, L., Weddell, G . , Martin, D. W., Jr., and Nussenzweig, V. (1987). Nature (London) 325, 545. Carroll, M. C., Alicot, E. M., Katzman, l? J., Klickstein, L. B., Smith, J. A., andFearon, D. T.(1988).J. Exp. Med. 167, 1271.
412
DENNIS HOURCADE ET A L .
Carter, R. H., Spycher, M. O., Ng, Y. C., Hoffman, R., and Fearon, D. T. (1988). J. Immunol. 141, 457. Changelian, P. S . , and Fearon, D. T. (1986).J. Exp. Med. 163, 101. Cheung, N. V., Walter, E. I., Smith-Mensah, W. H., Ratnoff, W. D., Tykocinski, M. L., and Medof, M. E. (1987).J. Clin. Invest. 81, 1122. Chung, L. P., and Reid, K. B. M. (1985). Biosci. Rep. 5, 855. Chung, L. P., Gagnon, J.. and Reid, K. B. M. (1985a). Mol. Immunol. 22, 427. Chung, L. P., Bentley, D. R., and Reid, K. B. M. (1985b). Biochem. J. 230, 133. Cines, D. B., Lyss, A. P., Bina, M., Corkey, R., Kefalides, N. A., andFriedman, H. M. (1982).J. Clin, Invest. 69, 123. Cole, J., Houseley, G . A., Jr., Dykman, T. R., MacDermott, R. P., and Atkinson, J. P. (1985). Proc. Natl. Acad. Sci. U.S.A. 83, 859. Cooper, N. R., Moore, M. D., andNemerow, G . R. (1988). Annu. Rev. Immunol. 6,85. Dahlback, B. (1983). Biochem. J. 209, 847. Dahlback, B., and Muller-Eberhard, H. J. (1984).J. Biol. Chem. 259, 11631. Dahlback, B.. and Stenflo, J. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 2512. Dahlback, B., Smith, C. A., and Muller-Eberhard, H. J. (1983).Proc. Natl. Acad. Sci. U.S.A. 80, 3461. Davis, C. G., Elhammer, A., Russell, D. W., Schneider, W. J., Kornfeld, S., Brown, M. S., and Goldstein, J. L. (1986).J. Biol. Chem. 261, 2828. Davis, L. S.,Patel, S.S., Atkinson, J. P., and Lipsky, P. E. (1988).J. Immunol. 141,2246. Davitz, M., Low, M., and Nussenzweig, V. (1986).J. Exp. Med. 163, 1150. de Cordoba, S. R., and Rubinstein, P. (1986).J. Ex$. Med. 164, 1274. de Cordoba, S . R., Lublin, D. M., Rubinstein, P., and Atkinson, J. P. (1985).J. Exp. Med. 161, 1189. Dykman, T. R., Cole, J. L., Iida, K., and Atkinson, J. P. (1983a).J.Ex#. Med. 157,2160. Dykman, T. R.,Cole, J. L., Iida, K., and Atkinson, J. P. (1983b).Proc. Natl. Acad. Sci. U.S.A. 80, 1698. Dykman, T. R., Hatch, J. A., and Atkinson, J. P. (1984).J. Ex#. Med. 159, 691. Dykman, T. R., Hatch, J. A,, Aqua, M. S., and Atkinson, J. P. (1985).J. Immunol. 134, 1787. Early, P., Rogers, J., Davis, M., Calame, K., Bond, M.,Wall, R, and Hood, L. (1980). Cell 20, 313. Eden, A,, Miller, G. W., and Nussenzweig, V. (1973).J. Clin. Invest. 52, 3239. Fearon, D. T. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 5867. Fearon, D. T. (1980).J. Exp. Med. 152, 20. Fearon, D. T. (1985).J. Invest. Dermatol. 85, 53s. Fearon, D. T., and Collins, L. A. (1983).J. Immunol. 130, 370. Fearon, D. T., Kaneko, I., and Thomson, G . G. (1981).J. Exp. Med. 153, 1615. Ferreira, A., Takahashi, M., and Nussenzweig, V. (1977).J. Exp. Med. 146, 1001. Ferreira, A., Weisz-Carrington, P., Gigli, I., and Nussenzweig, V. (1978).J. Immunol. 120, 1772. Fingeroth, J. D., Weis, J. J., Tedder, T. F., Strominger, J. L., Biro, P. A., and Fearon, D. T. (1984). Proc. Natl. Acad. Sci. U.S.A. 81, 4510. Fischer, E.,'Appay, M. D., Cook, J., and Kazatchkine, M. D. (1986).J. Immunol. 136, 1373. Frade, R., Crevon, M. C., Barel, M., Vazquez, A., Krikorian, L., Charriaut, C., and Galanaud, l? (1985). Eur. J. Immunol. 15, 73. Frank, M. M., Joiner, K. A., and Brown, E. J. (1984). h o g . Immunol. 5, 381. Friedman, H. M.,Cohen, G. H., Eisenberg, R. J., Seidel, C. A., andCines, D. B. (1984). Nature (London) 309, 633.
REGULATORS OF COMPLEMENT ACTIVATION
413
Fries, L. F., Friedman, H. M., Cohen, G. H., Eisenberg, R. J., Hammer, C. H., and Frank, M. M. (1986).J. Immunol. 137, 1636. Fukisaku, A., Harley, J. B., Frank, M. B., Gruner, B. A., Frazier, B., and Holers, V. M. (1989).J. Biol. Chem. (in press). Gilbert, W. (1978). Nature (London) 271, 501. Gilbert, W. (1979). ICN-UCLA Symp. Mol. Cell. Biol. 14, 1-12. Goldberger, G . , Bruns. G. A. P., Rits, M . , Edge, M. D., andKwiatkowski, D. J. (1987). J. Biol. Chem. 262, 10065. Greene, W. C., and Leonard, W. J. (1986). Annu. Rev. Immunol. 4, 69. Griffin, F. M.,Jr., and Mullinax, P.J. (1981).J. Ex$. Med. 154, 291. Hebert, L. A.,and Cosio, F. G. (1987). Kidney Int. 31, 877.,. Hoffman, E. M. (1969a). Zmmunochemisty 6, 391. Hoffman, E. M. (1969b). Immunochemisty 6, 405. Holers, V. M., Cole, J. L., Lubli, D. M., Seya, T., and Atkinson, J. P.(1985).Immunol. Today 6, 188. Holers, V. M., Seya, T., Brown, E., O’Shea, J. J., and Atkinson, J. P. (1986). Complement 5, 63. Holers, V. M., Chaplin, D. D., Leykam, J. F., Gruner, B. A., Kumar, V., and Atkinson, J. P. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 2459. Hourcade, D.,Miesner, D. R., Atkinson, J. P., and Holers, V. M. (1988).J. Exp. Med. 168, 1255. Ichinose, A., McMullen, B. A., Fujikawa, K., and Davie, E. W. (1986). Biochemistry 25, 4633. lida, K., and Nussenzweig, V. (1981).J. fip. Med. 153, 1138. Iida, K., Mornaghi, R., and Nussenzweig, V. (1982).J. Exp. Med. 155, 1427. Iida, K.,Nadler, L., and Nussenzweig, V. (1983).J. Exp. Med. 158, 1021. Ishihara, A., Hou, Y., and Jacobson, K. (1987). Proc. Natl. Acad. Sci. U.S.A. 84, 1290. Janatova, J., and Reid, K. B. M. (1987). Complement 4, 173. Joiner, K. A., Brown. E. J., and Frank, M. M. (1984). Annu. Rev. Immunol. 2, 461. Joumet, A., and Tosi, M. (1986). Biochem. J. 240, 783. Kai, S., Fujita, T., Gigli, I., and Nussenzweig, V. (1980).J. Immunol. 125, 2409. Kaidoh, T., Natsuume-Sakai, and Takahashi, M. (1981).J. Immunol. 126, 463. Katz, Y., and Strunk, R. C. (1988).J. Immunol. 141, 559. Kingsley, D. M., Kozarsky, K. E,Hobbie, L., and Krieger, M. (1986). Cell (Cambridge, Mass.) 44, 749. Kinoshita, T., Rosenfeld, S. I., and Nussenzweig, V. (1987).J. Immunol. 138, 2994. Klickstein, L. B., Rabson, L. D., Wong, W. W., Smith, J. A., and Fearon, D. T. (1987a). Complement 4, 180. Klickstein, L. B., Wong, W. W., Smith, J. A., Weis, J. H., Wilson, J. G, and Fearon, D. T. (1987b).J. Exp. Med. 165, 1095. Kornfeld, R., and Kornfeld, S. (1980). In “The Biochemistry of Glycoproteins and Proteoglycans” (W.Lennarz, ed.). p. 85. Plenum, New York. Kotwal, G. J., and Moss, B. (1988). Nature (London) 335, 176. Kristensen, T., and Tack, B. E (1986). Proc. Natl. Acad. Scz. U.S.A. 83, 3963. Kristensen, T., Wetsel, R. A., and Tack, B. F. (1986).J. Immunol. 136, 3407. Kristensen, T., Ogata, R. T., Chung, L. P., Reid, K. B. M., and Tack, B. F. (1987). Biochemistry 26, 4668. Kubota, Y., Gaither, T. A., Cason, J., O’Shea,J. J., and Lawley, T. J. (1987).J Immunol. 138, 1137. Kurovsky, A., Barnett, D. R., Lee, T.H., Touchstone, B., Hay, R. E., Arnott, M. S., Bowman, B. H., and Fitch, W. M. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 3388.
414
DENNIS HOURCADE ET A L
Lachmann, P. J.. and Walport, M. J. (1986). In “Immunobiology of the Complement System” (G. D. Ross, ed.), p. 237. Academic Press, Orlando, Florida. Low, M. G., and Saltiel, A. R. (1988). Science 239, 268. Lozier, J., Takahashi, N., and Putnam, F. W. (1984). Proc. Natl. Acad. Sci. U.S.A.
81,3640.
Lublin, D. M., and Atkinson, J. P. (1989). Annu. Rev. Immunol. 7 (in press). Lublin, D. M., and Atkinson, J. P. (1989). In “Biochemistry and Molecular Biology of Complement” (R. B. Sim, ed.) (in press). Lublin, D. M., Krsek-Staples, J., Pangburn, M. K., and Atkinson, J. P. (1986). J. Immunol. 137, 1629. Lublin, D.M., Lemons, R. S., LeBeau, M. M., Holers, V. M., Tykocinski, M. L.. Medof, M. E., and Atkinson, J. P. (1987).J. Ex#. Med. 165, 1731. Lublin, D. M., Liszewski, M. K., Post, T. W., Arce, M. A., LeBeau, M. M., Rebentisch, M. B., Lemons, R. S., Seya, T., and Atkinson, J. P. (1988).J.Exfi. Med. 168, 181. McNearney, T. A., Odell, C., Holers, V. M., Spear, P. G., and Atkinson, J. P. (1987). J. Exp. Med. 166, 1525. McNearney, T. A., Ballard, L., Seya, T., and Atkinson, J. P. (1988). Submitted for publication. Maeda, N. (1985).J Biol. Chem. 260, 6698. Malhotra, V., and Sim, R. B. (1985). E u z J . Immunol. 15, 935. Mather, E. L., Nelson, K. J., Haimovich, J., and Perry, R. P. (1984). Cell (Cambridge, Mass,) 36, 329. Mayer, M. M., Michales, D. W., Ramm, L. E., Whitlow, M. B., Willoughby, J. B., and Shin, M. K. (1981). CRC C d . Rev. Immunol. 2, 133. Medof, M. E. (1988). In “Cytolpic Lymphocytes and Complement Effectors of the Immune System” (E. R. Podack, ed.), p. 57. CRC Press, Boca Raton, Florida. Medof, M. E., and Nussenzweig, V. (1984).J. Exp. Med. 159, 1669. Medof, M. E., Iida, K., Mold, C., and Nussenzweig, V. (1982).J. Ex$. Med. 156, 1739. Medof, M. E., Walter, E. I., Rutgers, J. L., Knowles, D. M., and Nussenzweig, V. (1987a). J. Exp. Med. 165, 848. Medof, M. E., Lublin, D. M., Holers, V. M., Ayers, D. J., Getty, R. R., Leykam, J. F., and Atkinson, J. P. (1987b). Proc. Natl. Acad. Sci. U.S.A. 84, 2007. Melamed, J., Arnaout, M. A., and Colten, H. R. (1982).J Immunol. 128, 2313. Moldenhauer, F., David, J., Fielder, A. H. L., Lachmann, P.J., and Walport, M. J. (1987). Arthritis Rheum. 30, 961. Moore, J. C., Frank, M. M., and Miiller-Eberhard, H . J. (1985).J. Ex$. Med. 162. 1182. Moore, M. D.,Cooper, N. R., Tack, B. F., and Nemerow, G. R. (1987). Proc. Natl. Acad. Sci. LJ.S.A. 84, 9194. Morley, B. J., and Campbell, R. D. (1984). EMBO J. 3, 153. Miiller-Eberhard, H. J. (1984). Springer Semin. Immunopathol 7, 93. Nelson, R. A. (1953). Science 118, 733. Nemerow, G.R., McNaughton, M. E., and Cooper, N. R. (1985a).J.Immunol. 135,
3068.
Nemerow, G. R., Wolfert, R., McNaughton, M. E., and Cooper, N. R. (1985b).J Virol.
55, 347.
Nicholson-Weller, A,, Burge, J., Fearon, D. T., Weller, P. F., and Austin, K. F. (1982). J. Immunol. 129, 184. Nicholson-Weller, A., Russian, D. A., and Austen, K. F. (1986).J. Immunol. 137, 1275. Nickells, M. W., Seya, T., Holers, V. M., and Atkinson, J. P. (1986). Mol. Immunol.
23, 661.
Noda, M., Yoon, K.,Rodan, G. A , , and Koppel, D. E. (1987).J. Cell Biol. 105, 1671.
REGULAMRS OF COMPLEMENT ACTIVATION
415
Nunez, D., Charriaut-Marlangue, C., Barel, M., Benveniste, J., and Frade, R. (1987). Euz J. Immunol. 17, 515. Oglesby, T. J., Accavitti, M. A. and Volanakis, J. E. (1988).J. Immunol. 141, 926. O’Shea, J. J., Brown, E. J., Seligmann, B. E., Metcalf, J. A., Frank, M. M., and Gallin, J. I. (1985a).J. Immunol. 134, 2580. O’Shea, J. J., Brown, E. J., Gaither, T. A., Takahashi, T.. and Frank, M. M. (1985b). I. Immunol. 135. 1325. Pangburn, M . K. ,and.Miiller-Eberhard, H. J. (1984). Springer Semin. Immunopathol.
7, 164.
Perkins, S. J., Chung, L. P., and Reid, K. B. M. (1986). Bi0chem.J. 233, 799. Perkins, S.J., Haris, P.I., Sim, R. B., and Chapman, D. (1988).Biochemistry 27,4004. Pommier, C. G., O’Shea, J., Chused, T., Yancey, K., Frank, M. M., Takahashi, T., and Brown, E. J. (1984).J. Ex$. Med. 159, 137. Porteu, F., Mir, A., and Halbwachs-Mecarelli, L. (1987). Euz J. Immunol. 17, 629. Reid, K. B. M. (1986). Essays Biochem. 22, 69. Reid, K. B. M., and Gagnon, J. (1982). FEBS Lett. 137, 75. Reid, K. B. M., Bentley, D. R., Campbell, R. D., Chung, L. P., Sim, R. B., Kristensen, T., and Tack, B. F. (1986). Immunol. Toduy 7, 230. Rey-Campos, J., Rubinstein, P., and de Cordoba, S. R. (1988).J. Ex$. Med. 167,664. Reynes, M.,Aubert, J. P., Cohen, J. H. M., Audouin, J., Tricottet, V., Diebold, J., and Kazatchkine, M. D. (1985).J. Immunol. 135, 2687. Ripoche, J., Day, A. J., Harris, T. J. R., and Sim, R. B. (1988). Bi0chem.J. 249, 593. Ross, G.D. (ed.) (1986). “Immunobiology of the Complement System: An Introduction for Research and Clinical Medicine.” Academic Press, Orlando, Florida. Ross, G. D., and Medof, M. E. (1985). Adv. Immunol. 37, 217. Ross, G. D., Polley, M. J., Rabellino, E. M., and Grey, H. M. (1973).J. Exp. Med. 138,
798.
Scharfstein, J., Ferreira, A., Gigli, I., and Nussenzweig, V. (1978).J.f i p . Med. 148,207. Schulz, T. F., Schwable, W., Stanley, K. K., Weiss, E., and Dierich, M. P. (1986). Eur J. Immunol. 16, 1351. Schulz, T. E,Pernegger, G., Petzer, A., Eigentler, A., Myones, B. L., andDierich, M. P. (1987). Complement: Int. Complement Workshop, 12th p. 223. Schwaeble, W., Schulz, T. E , Zwirner, J., Dierich, M. P., and Weiss, E. H. (1987). Complement 4, 224. Seya, T., Turner, J., and Atkinson, J. P. (1986). J. Exp. Med. 163, 837. Seya, T., Farries, T., Nickells, M., and Atkinson, J. l? (1987).J. Immunol. 139, 1260. Seya, T., Ballard, L. L., Bora, N. S., Kumar, V., Cui, W., and Atkinson, J. P. (1988). Eur J. Immunol. 18, 1289. Shimuzu, A., Kondo, S., Takeda, S., Yodoi, J., Ishida, N., Sabe, H., Osawa, H., Diamanstein, T., Nikaido, T., and Honjo, T. (1985). Nucleic Acids Res. 13, 1505. Sim, R. B., and DiScipio, R. G. (1982). Biochem. J. 205, 285. Smiley, M. L.,and Friedman, H. M. (1985).J. Virol. 55, 857. Smiley, M. L.,Hoxie, J. A., and Friedman, H. M. (1985).J. Immunol. 134, 2673. Stafford, H. A., Tykocinski, M. L., Lublin, D. M., Holers, V. M., Rosse, W. F., Atkinson, J. F’., and Medof, M. E. (1988). R o c . Natl. Acad. Sci. U.S.A. 85, 880. Tack, B. F. (1985).In “Complement” (H. J. Muller-Eberhard and P. A. Meischer, eds.), p. 49. Springer-Verlag, New York. Tanner, J., Weis, J., Fearon, D., Whang, Y., and Kieff, E. (1987). Cell (Cambridge, Mass.) 50, 203. Tedder, T. F., Clement, L. T., and Cooper, M. D. (1984). J. Immunol. 133, 678. Thomas, J., Webb, W., Davitz, M . , and Nussenzweig, V. (1987). Bi0phys.J. 51,522 (abstr).
416
DENNIS HOURCADE ET A L
Tosi, M., Duponche, C., Meo, T., and Julier, C. (1987). Biochemuty 26, 8516. Ueda, A., Keamey, J. E,Roux, K. H., and Volanakis,J. E. (1987).J. Immunol. 138, 1143. Van Dyne, S., Holers, V. M., Lublin, D. M., and Atkinson. J. P. (1987). Clin. Exp. Immunol. 68, 570. Vik, D. P., Keeney, J. B., Bronson, S., Westlung, B., Kristensen, T., Chaplin, D. D., and Tack, B. F. (1987). Complement 4, 235. Waldmann, T. A. (1986). Science 232, 727. Weis, J. H., Morton, C. C., Bruns, G . A. P., Weis, J. J., Klickstein, L. B., Wong, W. W., and Fearon, D. T. (1987).J. Immunol. 138, 312. Weis, J. J., and Fearon, D. T. (1985). J. Biol. Chem. 260, 1384. Weis, J. J., Toothaker, L. E., Smith, J. A., Weis, J. H., and Fearon, D. T. (1988).J. Ex$. Med. 167, 1047. Weitkamp, L. R., and Lamm, L. U. (1982). Cytogenet. Cell Genet. 32, 130. Whaley, K., and Ruddy, S. (1976). Science 193, 1011. Wilson, B. S., Platt, J. L., and Kay, N. E. (1985). Blood 66, 824. Wilson, J. G . , Wong, W. W., Schur, P. H., and Fearon, D. T. (1982). N. Eng1.J. Med. 307, 981. Wilson, J. G . , Murphy, E. E., Wong, W. W., Klickstein, L. B., Weis, J. H., and Fearon, D. T. (1986a). J. Ex$. Med. 164, 50. Wilson, J. G.,Ratnoff, W. D., Schur, l? H., and Fearon, D. T. (1986b). Arthritis Rheum. 29, 739. Wong, W. W., Wilson, J. C., and Fearon, D. T. (1983). J. Clin. Invest. 72, 685. Wong, W. W., Jack, R. M., Smith, J. A., Kennedy, C. A., and Fearon, D. T. (1985). J. Immunol. Methods 82, 303. Wong, W. W., Kennedy, C. A., Bonaccio, E. T., Wilson, J., Klickstein, L., Weis, J. H., and Fearon, D. T. (1986). J. Exp. Med. 164, 1531. Wong, W. W., Cahill, J., Kennedy, C. A., Bonaccio, E., Wilson, J. G., Klickstein, L. B., Rabson, K., and Fearon, D. T. (1987). Complement 4, 240. Wright, S. D., and Meyer, B. C. (1985a). J. Immunol. 136, 1759. Wright, S. D., and Meyer, B. C. (1985b).J. Ex$. Med. 162, 762. Wright, S. D., Craigmyle, L. S., and Silventein, S. C. (1983).J. Exfi. Med. 158, 1338. Wright, S. D., Detmers, P. A., Jong, M. T. C., and Meyer, B. C. (1986).J. Exp. Med. 163, 1245. Yoon, S. H., and Fearon, D. T. (1985).J. Immunol. 138, 312. Yoshida, K., Yukiyama, Y., and Miyamoto, T. (1987). J. Rheumatol. 14, 490. Yu, G.. Holers, V. M., Seya, T., Ballard, L., and Atkinson, J. P. (1986).J. Clin. Invest. 78, 494. Zezulak, K. M.,and Spear, P. G . (1984). J. Vzrol. 49, 741.
ADVANCES IN IMMUNOLDCY,VOL. 45
Origin and Significance of Autoreactive T Cells MAURICE ZAUDERER Cancer Center. Unlwmlty of Rochester Medical Center. Rochester. New York 14642
I. Introduction
The mature peripheral T cell population is restricted to recognize foreign antigens in association with only those MHC-encoded molecules expressed in the thymic differentiation environment. Since foreign antigens are not normally present in that environment, this clearly means that at an early stage of differentiation T cell precursors are self-MHC reactive. It is presumed that some process for negative selection prevents seeding any very high-affinity self-reactiveclones in the mature population. Antigen-specific, MHC-restricted peripheral T cells must, nevertheless, derive from thymic precursors that expressed a functional affinity for self-MHC, perhaps in association with other self-molecules.The basis for this transition from an autoreactive thymic precursor to an antigendependent, MHC-restricted peripheral T cell is not understood. Present evidence suggests that this transition cannot be explained by somatic mutation of rearranged T cell receptor genes (Chien et al., 1984; Barth et al., 1985; Arden et al., 1985; Becker et al., 1985). It would seem, therefore, that T cell maturation is accompanied by physiological changes that modulate T cell activation signals or activation thresholds such that self-MHC or self peptide-MHC recognition is not, in general, sufficient to induce clonal expansion of resting peripheral T cells. Such changes could reflect properties of mature T cells relative to thymic precursors as well as properties of the stimulatory environment in secondary lymphoid tissues relative to the thymus. The nature of these physiological changes is not known and their stability cannot be predicted. A considerable body of evidence suggests, however, that some mature peripheral T cells either retain or reacquire the ability to be stimulated by selfMHC in the absence of foreign antigens. Autoreactive T cells and T cell lines derived from normal individuals have been independently described in many laboratories. Such T cells are activated in the absence of any identifiable foreign antigen by class I1 MHC syngeneic stimulators, but not MHC allogeneic stimulators. In particular, autoreactive responses have been demonstrated in the absence of xenogeneic serum components (Yamashita and Shevach, 1980; 417 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
418
MAURICE ZAUDERER
Glimcher et al., 1981; Glimcher and Shevach, 1982; Zauderer et a l . , 1984; Finnegan et al., 1984; Coeshott et al., 1986). Several independent lines of evidence developed in our own and other laboratories suggest that autoreactive T cells derive from antigen-stimulated precursors. We favor, therefore, the interpretation that “normal” autoreactivity represents a particular physiological state that recapitulates the conditions of thymic selection and is induced in many antigen-specific, MHC-restricted T cells as a result of normal antigen-dependent activation. As will be discussed below, this interpretation is further supported by evidence that in the presence of selected stimulators, autoreactive T cells have unique activation requirements distinct from the activation requirements of antigenspecific, MHC-restricted T cells. Even if autoreactivity is a transient physiological state in peripheral T cells, it could make a substantial contribution to clonal expansion and amplification of immune functions. Since we must, in any case, start with the assumption that a mechanism is already in place in uivo for driving T cells out of a developmental self-MHC-reactivestate and into a state of mature antigen-dependent activation, it is possible that these same mechanisms regulate normal autoreactive physiology in the periphery. In the absence of normal feedback controls, T cells driven to a self-MHC-reactivestate could give rise to a variety of chronic systemic and organ-specificautoimmune responses. This and other evidence concerning the origin, physiological significance, and in vivo regulation of self-Ia-specific autoreactive T cells will be further elaborated below. II. Origin
Within the framework suggested above, autoreactive T cells derive from mature antigen-dependent precursors that have undergone a physiological transition which restores their ability to respond to self. This could be a normal and reversible transition that frequently accompanies specific T cell activation. Evidence for T cell autoreactivity derives from autologous or syngeneic mixed lymphocyte responses (AMLRs and SMLRs, respectively) in complex populations and more recently from the isolation and characterization of cloned autoreactive T cell lines maintained continuously in Vitro. The autologous mixed lymphocyte reaction has been previously reviewed in this series (Weksler et al., 1981). This evidence is consistent with the interpretation of autoreactivity elaborated here, since activated T cells in unfractionated populations may give rise to proliferative responses in AMLRs. Among cloned T cell lines, those that manifest autoreactive specificity may have stabilized the physiological transition postulated here during
AUTOREACTIVE T CELLS
419
in vitro adaptation to a continuous line. Although analysis of the properties of these T cell lines has been very informative, it cannot be assumed that they are typical of normal resting T cells. This exception may be especially important in considerations of their activation requirements and regulation.
T CELLSDERIVE A. SOMEAUTOREACTIVE FROM ANTIGEN-STIMULATED PRECURSORS Dos Reis and Shevach (1981)reported that negative selection through in vitro exposure to bromouridine deoxyribose (BUdR) and light during a syngeneic mixed lymphocyte reaction greatly reduces the secondary proliferative response to specific antigen in subsequent cultures of insulin and purified protein derivative (PPD)-primed guinea pig lymph node T cells. Conversely, positive selection for autoreactive T cells in SMLR cultures of antigen-primed lymph node T cells gives rise to an enhanced secondary proliferative response upon restimulation with specific antigen. These workers concluded that the autoreactive T cells responding in SMLRs include diverse antigen-specific T cell clones induced to proliferate in the absence of nominal antigen. One possible objection to this conclusion is that autoreactive T cells may participate in amplifying the proliferative response of independent antigen-specificclones. This could account for the observation that primed (A X B) F1 T cells selected in a SMLR culture with parent A stimulators give rise to a more vigorous antigen-specific proliferative response in subsequent culture with parent A antigen-presenting cells (APCs) than with parent B APCs. A different approach to analyzing the relationship between autoreactive and antigen-specificT cell clones has been described by Faherty et al. (1985),who show that, under well-definedconditions, the specificity of autoreactive T cell clones selected from antigen-primed populations reflects the predominant MHC restriction specificity of T cells specific for the immunogen. Thus, I-E subregion-specific autoreactive T cells are detected at a much higher frequency in populations primed with the I-E-restricted antigen G L 8 than in populations primed with the predominantly I-A-restrictedantigen keyhole limpet hemocyanin (KLH) (Sproviero et al., 1981). A striking confirmation and extension of these observations has been noted in the analysis of T cell receptor gene expression in autoreactive T cell clones selected from pigeon cytochrome c-stimulated populations. The diversity of germ line T cell receptor variable region genes expressed in the pigeon cytochrome c-specific T cell repertoire is relatively limited (Fink et al., 1986; Sorger et al., 1987). The Vp16 T cell receptor gene predominates in pigeon cytochrome c-specific, I-Ek-restricted T cells that have also been selected for
420
MAURICE ZAUDERER
1-Ef:Eb alloreactivity. It was observed that autoreactive T cells also arose in these T cell lines, and that the Vp16 T cell receptor gene also predominates in this autoreactive T cell subset (Louis Matis, unpublished observations). These results strongly support the conclusion that autoreactive T cells can derive from antigen-stimulated precursors.
B. SOMEANTIGEN-SPECIFIC T CELLSALSO MANIFEST AUTOREACTIVE SPECIFICITY Further evidence that autoreactive specificity is a normal property of mature peripheral T cells derives from the frequent observation that T cell clones and hybridomas with well-defined antigen specificity also manifest a significant autoreactive response to MHC -syngeneicstimulators in the absence of the nominal antigen (Glimcher and Shevach, 1982; Rock and Benacerraf, 1983; Dekruyff et al., 1985; Dos Reis and Shevach, 1985). Activation of many other cloned T cell lines is, of course, strictly antigen dependent. As suggested above, however, it is possible that these clones have a lower intrinsic affinity for self-Ia and cannot be activated in antigen-independent fashion in vitro. Alternatively, a physiological state characteristic of self-reactive T cells may be stabilized in some but not all in vitro-adapted T cell lines. Dos Reis and Shevach (1985) have characterized the Ia fine specificity of two pork insulin-specific, Ia-restricted T cell clones that also give rise to significant autoreactive proliferation in response to syngeneic stimulators in the absence of antigen. Blocking studies employing a panel of Ia-specific monoclonal antibodies demonstrated that although different Ia epitopes were required for activation of the two clones, for each clone the same Ia epitope was involved in both the autoreactive and pork insulin-specific response. Additional experiments suggest that autoreactive T cells are subject to the same MHC-directed thymic selection as mature antigen-specific T cells. The specificity of the syngeneic mixed lymphocyte response in semiallogeneic or fully allogeneic radiation-induced bone marrow chimeras is determined by the differentiation environment (Glimcher et al., 1981). The MHC fine specificity of cloned autoreactive T cell lines from antigen-stimulated populations was examined by Faherty et al. (1985). They report that autoreactive clones can be identified with specificity for unique F1 hybrid determinants of (BALB.K X BALB.B) F1, and for mutant I-Ab determinants of the B.6C-H-2bm12(bm12) strain. This indicates a similar MHC fine specificity for autoreactive and antigen-specific, MHC-restricted T cells.
AUTOREACTIVE T CELLS
421
C. SOMEAUTOREACTIVE T CELLSMAYDERIVE FROM IMMATURE PRECURSORS Evidence cited above suggests that autoreactive T cells can derive from antigen-stimulatedprecursors. It has also been suggested that autoreactive T cells may represent an immature subset of precursors, perhaps even of prethymic origin. De Talance et al. (1986) and Morisset et al. (1988) have described a population of Lyt-e-, L3T4- peripheral T cells which are unresponsive to either syngeneic or allogeneic stimulators. These double-negativeprecursors can be induced to proliferate by concanavalin A (Con A) and give rise to bright Thy-1 + , Lyt-2-, L3T4+ T lymphocytes which give a vigorous class 11-restrictedmixed lymphocyte response in secondary cultures with either syngeneic or allogeneic stimulators. It is not known, however, whether these Con A-reactive double-negative precursors of autoreactive and alloreactive L3T4 T cells represent an immature population or derive from mature T cells which have reversibly down regulated L3T4 expression. +
111. Activation Requirements and Specificity
A. THYMIC STROMAL CELLSARE EFFICIENT ANTIGENPRESENTING CELLSBUT Do NOT STIMULATE AUTOREACTIVE T CELLS in Vitro In an attempt to identify physiological properties that might distinguish autoreactive T cells from antigen-specific, MHC-restricted T cells, we have compared the ability of different types of Ia-positive stimulators/APCs to induce proliferation of cloned autoreactive and antigen-specificT cell lines. Because the thymus is the major site of both positive and negative MHC-directed selection, thymic stromal cells were thought to be a particularly interesting population of stimulators. Thymic stroma was prepared by dissociation of thymus with collagenase and treatment with anti-Thy-1,anti-Lyt-2, anti-LST4, and complement (C’). The results in Table I show that thymic stromal cells are more efficient than similarly treated spleen cells for antigen presentation to a keyhole limpet hemocyanin-specific, I-Ad-restricted cloned T cell line. This is representative of results obtained for several different cloned T cell lines with specificity for different nominal antigens and different MHC restriction. In sharp contrast, spleen cells but not thymic stroma efficiently stimulate proliferation of I-Ad-specific (CLJl) and I-Ed-specific (FGB1) autoreactive T cell lines.
422
MAURICE ZAUDERER
TABLE I THYMIC STROMAL CELLSDo NOT EFFICIENTLY STIMULATE AUTOREAC~IVE T CELLCLONESO
Stimulators*
+
Anti-Thy-1 C’-treated spleen 105 5 x 104 2.5 x 104 104 Thymic stroma 105 5 x 104 2.5 x 104 104 None
CKMl KLH
None
33,187 33,954 31,020 21,623
74,709 68,847 38,788 39,503
802 628 419 238
9176 3201 879 667
2080 1528 844 715
152,692 137,593 62,715 29,421
499 893 733 486
2 84
460
479
FGBl
CLJl
58,437 56,927 39,827 17,993
+
n 2 X 104 T cells of cloned autoreactive T cell line FGBl (I-Ed specific) or CLJl (I-Ad specific) or antigen-specific CKMl (KLH specific, I-Ad restricted) were cultured for 48 h with the indicated mitomycin C-treated stimulators. Thymic stroma was prepared by digestion with collagenase and treatment with anti-Thy-1, anti-Lyt-2, and complement. KLH (40 pg/ml) was added to CKMl cultures only. [SHIThymidine incorporation was determined during the final 16 h of culture. * Cells/culture. CCounts/minute/culture; KLH, keyhole limpet hemocyanin.
Several factors could account for the failure of thymic stroma to stimulate autoreactive T cells. 1. Thymic stroma may induce tolerance in Ia-specific T cells. 2. Autoreactive T cells may not be specific for native Ia, but may be specific for Ia in association with other self-molecules that are not expressed on thymic epithelial cells. 3. Activation requirements for autoreactive T cells may be different than for antigen-specific T cells.
It does not appear that tolerance induction can account for the failure of thymic stroma to stimulate autoreactive T cells, since stimulation of these in vitro-adapted autoreactive T cell lines with thymic stroma does not suppress the response to syngeneic spleen cells (Table 11). It may indeed be the case that autoreactive T cells are specific for Ia in association with other self-molecules.These selfsame combinations
423
AUTOREACTIVE T CELLS
TABLE I1 THYMIC STROMAL CELLSDo NOT SUPPRESS STIMULATION OF AUTOREACTIVE T CELLSn
T cellsb Thymic stroma (cells/culture) 105
5 x 104 2.5 x 104 None
FGBl
CKMl KLH
+
FGBl
174,982 153,961 51,586 824
1812 628 410 334
+ 105 splenic stimulators
None
46,811 49,008 49,112 50,821
235 295 395
-
"Autoreactive FGBl T cells were cocultured with thymic stroma in the presence or absence of 1 X l o 5 splenic stimulators. Antigenspecific CKMl was cultured with thymic stroma and KLH alone. Proliferation assays were carried out as described in the footnote to Table I. Similar results were obtained when addition of splenic stimulators to cultures of FGBl and thymic stroma was delayed 24 h. bCounts/minute/culture.
of self-molecules might, in fact, also serve as the elements that select the MHC-restricted T cell repertoire during thymic differentiation in the absence of foreign antigens. This would then still be consistent with the suggestion above that autoreactivity in normal T cells reflects the same specificity that is the basis for positive selection in the thymus. However, since the thymic epithelium plays an important role in selection of the MHC-restricted repertoire, these elements must be expressed in the thymic stroma. We undertook, therefore, to determine whether conditions could be defined in which the thymic stroma would efficiently stimulate autoreactive T cell clones. As shown in Table 111, this proved to be the case when cultures were supplemented with recombinant interleukin 1 (IL-1). We conclude, therefore, that the determinants recognized by autoreactive T cells can be expressed in the thymic stroma and that failure of stimulation is due to a requirement for additional activation signals. The difference in the signals required for activation of autoreactive and antigen-specific T cells could be either quantitative or qualitative. This difference might again reflect different physiological states in the two T cell populations. Alternatively, it may be that the consequences of interactions between T cells and stimulators are different when they are mediated by high-affinity recognition of a low-density of processed
424
MAURICE ZAUDERER
TABLE 111 IL-1 ENHANCES STIMULATION OF A U T O R E A ~ ITVCELL E CLONE FGBl BY THYMIC STROMAL CELLS' Stimulatorsb Thymic stroma 5 x 104 2.5 x 104 104
Splenic stimulators 5 x 104 2.5 x 104
104
No addition IL-1 (10 U/ml) (cpm/culture) 5703 1558 467
48,656 22,419 6617
33,067 27,894 9486
N.D.c N.D. N.D.
'Autoreactive FGBl T cells were cultured 48 h with indicated stimulators in the presence or absence of IL-1 (10 U/ml). b Cellslculture. cN.D., Not determined.
antigen on the surface of APCs or lower affinity recognition of highdensity Ia on the surface of autoreactive stimulators. Independent evidence for different sequelae to Ia recognition during the activation of autoreactive and antigen-specific T cells has been recently described by Griffith et al. (1988). These workers report that a mutant Ia-positive stimulator with a truncated cytoplasmic domain in the I-A, chain functions efficiently in antigen presentation to specific T cell hybridomas but is unable to stimulate a n autoreactive T cell hybridoma. An even more striking differential loss of the ability to stimulate a larger panel of autoreactive T cell hybrids was observed in further experiments employing mutant stimulators truncated in cytoplasmic domains of both I-A, and I-A0 (L.H.Glimcher, personal communication). These results suggest that activation of autoreactive T cells requires that a signal be transduced through Ia recognition on the stimulator, which is not required for activation of antigen-specific T cells. It is a seeming paradox that a stimulator population that is thought to mediate positive selection of MHC-specific thymic precursors would not stimulate cloned autoreactive T cell lines. It is important to note, however, that a signal equivalent or identical to IL-1 may not be limiting in the thymus in situ. It has been reported that murine thymocytes induce IL-1 secretion when cultured with irradiated Ia-positive accessory cells
AUTOREACTIVE T CELLS
425
in mlro (Rock and Benacerraf, 1984). Activation of autoreactive precursors may, therefore, be a normal event in that environment. If however the equivalent signal is limiting in peripheral lymphoid tissues, then this could contribute to regulation of autoreactive T cell expansion in uiuo. This might also explain the markedly increased frequency with which autoreactive precursors are activated during any ongoing antigen-specific immune response (Glimcher and Shevach, 1982;Imperiale et a l . , 1982; Zauderer et al., 1984; Saito and Rajewsky, 1985). B. AUTOREACTIVE SPECIFICITY 1. Recognition of l a Alone or la-Self Complexes
Autoreactive T cell clones are, in general, efficiently activated by MHC syngeneic stimulators with diverse background genes. This suggests that polymorphic background genes do not contribute significantly to autoreactive determinants. It does not, however, exclude specificity for nonpolymorphic gene products in association with self-Ia. Experiments of Coeshott et al. (1986)have demonstrated stimulation of IL-2 secretion by both autoreactive and alloreactive T cell hybridomas in response to glass beads coated with liposomes incorporating affinity-purified Ia. This suggests that these T cells are specific for Ia alone rather than for Ia in association with other self-antigens. However, in view of the recent observation that processed peptides may remain associated even with crystalline MHC class I molecules (Bjorkman et al., 1987), it cannot be ruled out that similar peptides are associated with affinity-purifiedclass I1 molecules and contribute to autoreactive determinants. On the other hand, in a novel set of experiments, Parham et al. (1987)have demonstrated that small class I peptides can inhibit recognition of target cells by class I-specific alloreactive cytolytic T lymphocytes (CTLs). This has suggested that alloreactive T cells may be specific for processed MHC antigens in association with native MHC. Class 11-specificautoreactive T cells might, by analogy, be specific for processed self-Ia. 2. Recognition of Cryptic Antigens A perennial challenge to the identification of autoreactive T cells has been the suggestion that they might be specific for xenogeneic serum components. Huber et al. (1982)and Kagan and Choi (1983)reported that efficient induction of the human autologous mixed lymphocyte response (AMLR) was dependent on exposure to either sheep red blood cells (SRBCs) or xenogeneic serum during fractionation procedures. A curious observation in these experiments was that T cells primed in AMLRs under any of these conditions could be restimulated with
426
MAURICE ZAUDERER
non-T cells exposed to xenogeneic serum from any of several other species (Huber et al., 1982). It is possible that this T cell response is not serum specific, but that, under the conditions of these experiments, exposure to mitogenic factors was required to activate competent stimulators of autoreactive precursors. In contrast to the above results for human AMLRs, Yamashita and Shevach (1980) reported that a primary guinea pig SMLR could be induced between responder T cells and peritoneal stimulators never exposed to xenogeneic serum components. More definitive characterization of autoreactive T cells has followed from the isolation of autoreactive T cell clones and hybridomas. Maintenance of these continuous lines in vitro is, like maintenance of virtually all other mammalian cell lines, fetal calf serum (FCS) dependent. It has, nevertheless, been possible in many instances to rule out immunological specificity for FCS components by carrying out shortterm assays in serum-free medium (Glimcher and Shevach, 1982; Zauderer et al., 1984; Finnegan et al., 1984; Clayberger et al., 1984; Coeshott et al., 1986). Two recent reports (Julius and Heusser, 1986; Pullen and Munro, 1988) describe a limited number of T cell lines that could not be restimulated in the absence of FCS. The conclusions that might follow from this observation are essentially trivial, since there is no reason to exclude the possibility that some T cells are specific for xenogeneic serum components. It should be noted, however, that the putative serum component for which these T cell lines are specific has not been identified. These T cell lines may, in fact, also be self-Ia specific but have an increased serum dependence - perhaps to maintain T cells in a physiologically responsive state or because they require activated stimulators (see below). It has been suggested that autoreactive T cells might be specific for other cryptic in vitro antigens. This is difficult to exclude entirely, but would appear unlikely in view of the apparently high frequency and ease of isolation of autoreactive precursors (Glimcher and Shevach, 1982; Imperiale et al., 1982; Zauderer et al., 1984; Saito and Rajewsky, 1985) as well as the demonstration that autoreactive T cells function in adoptive transfer in vivo (Saito and Rajewsky, 1985; Shiohara et al., 1987b,c; Saito et al., 1986). It may be possible in the near future to definitively characterize the fine specificity of at least some putative autoreactive clones through the use of small MHC peptides as described by Parham et al. (1987). Experience suggests, however, that those of our colleagues who cannot accommodate this phenomenon within their idiotypic gestalt will rise to the occasion with novel objections. It seems to me, therefore, most useful to emphasize that there are important questions to address on both sides of this issue. If autoreactive T cells in normal individuals
AUTOREACTIVE T CELLS
427
are viewed as a necessary consequence of MHC-directed positive selection, which may also be useful in the amplification of immune responses, then the mechanisms that limit the expansion of such T cells in Vivo must be clearly identified. If the presence of class I1 MHC-specific autoreactive T cells in normal individuals is viewed as problematic, then it must be explained how, in the absence of somatic mutation, thymic precursors selected for self-reactivity give rise to unreactive progeny.
IV. Physiological Significance
A. NONSPECIFIC HELPER FUNCTION The functional capabilities of autoreactive T cells have been investigated in a number of laboratories. In general, they have been found to provide the same physiological signals as other CD4-positiveT cells. Both lymphotoxin and interferon-y-producing autoreactive Thl (Clayberger et al., 1984; Shiohara et a l . , 1987a) and IL-4-secreting autoreactive Th2 (D. Burstyn and M. Zauderer, unpublished observation) have been isolated. Examination of the helper functions of both murine and human autoreactive T cell clones has clearly demonstrated that these T cells can provide helper signals that serve to enhance B cell proliferation and maturation to Ig secretion (Zauderer et al., 1984; Bensussan et al., 1984; Clayberger et al., 1984; Finnegan et al., 1984; Saito and Rajewsky, 1985; Quintans et al., 1986; Kotani et al., 1986). In all instances T cell activation was found to be MHC specific. The T cell clones characterized differed, however, in their ability to provide nonspecific bystander help to MHC-allogeneic B cells. Zauderer et al. (1984) and Clayberger et al. (1984) found that following T cell activation, enhancement of the antibody response to TI-2 antigens was not MHC restricted. Finnegan et al. (1984) reported that autoreactive T-B interaction was not MHC restricted for a polyclonal IgM response but was MHC restricted for an antigen-dependent specific IgG response. Saito and Rajewsky (1985)found that autoreactive T cell help for both polyclonal and antigen-dependent specific antibody responses was MHC restricted. In the case of human autoreactive T cell clones, Bensussan et al. (1984) and Kotani et al. (1986) reported that induction of polyclonal antibody responses was not MHC restricted. There may be important qualitative and quantitative differences in helper functions of different autoreactive T cell clones. Some of these
428
MAURICE ZAUDERER
differences may arise during the process of in vilro adaptation. Others, however, may be characteristic of different in w'uo populations. It needs to be determined, for example, whether some of the autoreactive T cell clones described above are representative of Thl and others of Th2. A second important variable in these experiments is the state of activation of the indicator B cell populations. Saito and Rajewsky (1985) determined by limiting dilution analysis that autoreactive T cells directly stimulate only 1 in 100 splenic B cells in vitro. The number of antigenspecific B precursors in this responsive population was significantly enhanced by stimulation with antigen in mlro and was absolutely antigen dependent in an adoptive transfer with cloned autoreactive T cells in vivo. This autoreactive T cell clone was evidently able to target antigen-activated B cells even in the presence of much larger numbers of resting Ia-positive precursors. In the absence of a mechanism for carrier-focused interaction with hapten-specific B cells, this requires that T cells discriminate target Ia antigens on activated and resting B cells or secrete readily diffusible MHC-specific factors. There is currently no evidence for the latter. In support of the former, Clayberger et al. (1984) and Quintans et al. (1986) have each characterized an autoreactive T cell clone that can be induced to proliferate by activated but not by resting B cells. The question of which Ia-positive cells are competent to interact with autoreactive T precursors is also relevant to a consideration of mechanisms that may regulate in vivo expansion of autoreactive T cells. We have, therefore, carried out a more extensive analysis of the ability of small resting B cells fractionated by centrifugal elutriation to stimulate a proliferative response in a panel of seven different cloned autoreactive T cell lines 0. Moynihan and M. Zauderer, unpublished observations). Of these seven lines, six were efficiently stimulated by resting B cells. Figure 1 shows the proliferative response of one such clone, KNA2, to small (Fxn 1) and large (Fxn 3) stimulator fractions. Also shown is the proliferative response of the one I-Ed-specific Th2-type T cell clone, FGB1, that failed to be stimulated by resting B cells. FGBl did, however, respond to this same small stimulator cell fraction following lipopolysaccharide (LPS) activation. An independent I-E-specific alloreactive Th2 clone was, in contrast to FGB1, efficiently stimulated by resting B cells (data not shown). This suggests that a requirement for activated stimulators is not a general property of the Th2 T cell subset, but is an intrinsic property of certain T cell clones perhaps related to their MHC avidity. B. PARTICIPATION IN TUMOR RESISTANCE AND INFLAMMATORY RESPONSE The apparently high frequency with which autoreactive precursors are isolated from antigen-primed populations (Glimcher and Shevach,
429
AUTOREACTIVE T CELLS
lo5
KNA2
* -t-
Fxn 1 Fxn3
Stimulator cellslwell
\
FGBI .1.3
-3 5 I L E,
* Fxn 1 + Fxn3
lo3
102
!
I
105
I
33x10 4
I
1
104
Stimulator cells/well FIG. 1. Small resting B cells fail to stimulate some cloned lines of autoreactive T cells; 3 X 104 T cells of the indicated cloned lines, KNA2 (I-Ak specific) and FGB1.1.3 (I-Ed specific), were cultured for 48 h with varied numbers of small (Fxn 1) and large (Fxn 2) anti-Thy-1 and complement and mitomycin C-treated syngeneic stimulator cells fractionated by centrifugal elutriation. Staining with anti-Ia, antimouse Ig, and mithramycin confirmed that Fxn 1 is constituted almost entirely of small resting B cells, while Fxn 3 includes activated B cells and other large non-B cells. [SHIThymidine incorporation was determined during the final 16 h of culture.
430
MAURICE ZAUDERER
1982; Imperiale et al., 1982; Zauderer et al., 1984; Saito and Rajewsky, 1985) and their well-documented nonspecific helper activity has suggested that they may serve an important amplifier function in specific humoral responses. Autoreactive T cells have, in addition, been shown to secrete a number of other lymphokines, including lymphotoxin and interferon-?, which would allow them to play an important role in inflammatory responses. The antitumor activity of an autoreactive T cell clone in vitro and in vivo has been characterized by Shiohara et al. (1987a,b). It was found that injection of these autoreactive T cells at the site of syngeneic tumor inoculation inhibited tumor growth. Autoreactive T cells but not unrelated antigen-specific T cells also markedly reduced pulmonary metastases when injected intravenously immediately after injection of melanoma cells. Saito et al. (1986) and Shiohara et al. (1987a) have also reported that in vivo injection of autoreactive T cell clones in footpads of syngeneic mice induced delayed-type hypersensitivity reactions in the dermis. In addition, autoreactive T cell clones with cytotoxic activity (lymphotoxin secretion) induced skin lesions, with infiltration of lymphocytes in the epidermis and epidermal cell damage. Some autoreactive T cell clones induced Ia expression on epidermal keratinocytes and a morphologic change in Langerhans cells without skin lesions. It is possible that a related cycle of events is involved in abnormal keratinocyte proliferation in psoriasis (Valdimarsson et al., 1986).
C. RELEVANCE TO AUTOIMMUNE DISEASE Class I1 MHC molecules have been shown to be aberrantly expressed on target tissues in a number of organ-specific autoimmune diseases. Bottazzo et al. (1983) demonstrated aberrant Ia expression on thyroid epithelium in Graves’ disease and discussed the possibility that this might be important in presentation of self-antigens that would not otherwise be immunogenic to T cells. Aberrant Ia expression has since also been found on /3 cells of the pancreas in type I diabetes (Bottazzo et al., 1985; Foulis and Farquharson, 1986) and on bile duct cells in primary biliary cirrhosis (Ballardini et al., 1984). Induction of abnormal Ia expression by IFN-y in uitro has been demonstrated for thyroid epithelium (Todd et al., 1984) and also for glial cells that may be important in autoimmune reactions to brain (Hirsch et al., 1983; Wong et al., 1984). These in vitroinduced Ia-positive cells have been shown to be competent to present antigen to specific T cells (Londei et al., 1984, 1985) and to stimulate mixed lymphocyte responses (Wong et al., 1985; Takiguchi et al., 1985). This has suggested a cycle of events that could give rise to chronic inflammatory reactions typical of autoimmune diseases. A specific precipitating
AUTOREAClXVE T CELLS
431
event, perhaps viral infection, leads to tissue infiltration by specific T cells. Lymphokine secretion by these activated T cells induces aberrant Ia expression. This in turn leads to further stimulation of autoreactive T cells and continued induction of Ia expression. T cells recruited into this response may include autoreactive clones specific for tissue antigens that would not normally be presented in association with Ia as well as self-Ia-specific autoreactive T cells. Both thyroid-specific T cell clones and T cell clones that proliferate in response to autologous Ia-positive peripheral blood mononuclear cells have been isolated from thyroid glands of two patients with Graves’ disease (Londei et al. 1985). Among other autoimmune diseases, it has been reported that self-Iaspecific autoreactive T cells are enriched in cerebral spinal fluid of patients with multiple sclerosis (Birnbaum et al., 1984) and in synovial fluid of patients with rheumatoid arthritis (Schlesier et al., 1984). Since autoreactive T cells are often recruited in ongoing specific immune responses, it has been difficult to establish the role of these T cells in the autoimmune disease. As noted above, however, Saito et al. (1986) and Shiohara et al. (1987~)were able to induce autoimmune skin lesions by injection of self-Ia-specific autoreactive T cell clones. In general, normal self-Ia-specificautoreactive T cells might escape regulatory feedback because of chronic stimulation in an environment in which Ia is aberrantly expressed or because of a concomitant defect in suppressor T cells. They might then contribute to either systemic or to organ-specific autoimmune disease. V. Regulation
In spite of the weight of evidence supporting the existence of autoreactive T cells and the important role they may play in the amplification of specific immune responses, their contribution to normal and abnormal responses cannot be evaluated until the mechanisms that must limit their expansion in vivo have been identified. A. SUPPRESSOR T CELLS Several early studies reported induction of suppressor activity in autologous mixed lymphocyte responses (Smith and Knowlton, 1979; Innes et al., 1979). More recently, autoreactive T cell clones have been described which either help or suppress in witro antibody responses, depending on the conditions of activation and T cell dose (Clayberger et al., 1984; Kotani et al., 1986; Quintans et al., 1986). These suppressor activities could play a role in limiting the amplification of immune responses mediated by autoreactive T cells. It is not clear, however,
432
MAURICE ZAUDERER
whether any of these suppressor mechanisms could also directly limit expansion of autoreactive T cell precursors. Sano et al. (1987) have described induction of L3T4+ T suppressor cells in mvo following intraparenteral injection of a cloned autoreactive T cell line. These suppressor T cells appear to act directly on helper T cells with the same Ia restriction specificity as the inducing autoreactive T cell clone. They could, therefore, serve to limit expansion of both antigen-stimulated and self-Ia-reactive T cells in mbo. A role for classical Lyt-2+ T suppressor cells in regulation of autoreactive T cells is suggested by experiments of Nagarkatti et al. (1988),who found that SMLRs were enhanced by depletion of Lyt-2+ cells from spleen. The syngeneic graft-versus-host disease (GVHD) is a particularly intriguing experimental system in which the role of regulatory cells in limiting autoreactive responses has been demonstrated. Glazier et al. (1983)reported that graft-versus-host disease develops in syngeneic or autologous bone marrow-reconstituted, irradiated, cyclosporin A-treated rats following withdrawal of cyclosporin. Syngeneic GVHD could be adoptively transferred by spleen cells from affected animals to syngeneic irradiated recipients but not to unirradiated recipients. This implicates a normal radiation-sensitive mechanism in the regulation of this autoreactive response.
B. STIMULATORS It has been suggested (Glimcher et al., 1981;Clayberger et al., 1984) that only a subset of activated Ia-positivecells are competent stimulators of autoreactive T precursors. Expansion of autoreactive T cells in mvo might then be regulated, at least in part, by the availability of activated stimulators. Our results described above indicate, however, that activated stimulators are required to induce a proliferative response for only some cloned autoreactive T cell lines. The majority of T cell lines we characterized were efficiently stimulated to proliferate by small resting B cells. This has a clear parallel in the results of an earlier study by Ashwell et al. (1984),who found that purified resting B cells efficiently presented antigen to 10 of 14 antigen-specific MHC-restricted cloned T cell lines. These results contrast, however, with results of Krieger et al. (1986),who reported that resting B cells are unable to function efficiently as stimulators of a primary mixed leukocyte reaction, as well as with results of Frohman and Cowing (1985),who found that resting cells were relatively inefficient antigen-presenting cells for a secondary in mlro proliferative response; and with results of Nussenzweig and Steinman (1980), who found that low-density cells are more than 100 times as efficient
AUTOREACI’IVE T CELLS
433
as Ig-positive cells for induction of primary SMLRs. All these studies could be reconciled if the activation requirements of in vitro-adapted T cell lines, such as used in our own experiments and those of Ashwell et al. (1984) and Gosselin et al. (1988), are, in general, less stringent than requirements for activation of T cells that are not tissue culture adapted. These continuously growing in vitro lines may rarely revert to fully resting T cells, or they may have been inadvertently selected for ease of stimulation by unfractionated spleen cells. It remains possible, therefore, that some regulation of autoreactive responses in vivo is at the level of activated stimulators.
VI. Conclusion
AUTOREACTIVE T CELLSRECAPITULATE CONDITIONS OF THYMIC SELECTION It is suggested here that “normal” autoreactivity is an intrinsic property of T cells originally selected in the thymus and recalled or revealed following antigen-dependent activation in the periphery. This induced autoreactive state might be transient or relatively stable. If transient, then the problem of regulation of the autoreactive response is identical to regulation of normal antigen-specific immune responses. Any mechanism that returns T cells to a resting state would eliminate autoreactive expansion until another cycle of stimulation is induced by a specific foreign antigen. This possibility is not inconsistent with available evidence. The best characterized autoreactive responses are those of in vitro-adapted continuous T cell lines, which may never revert to a fully resting state. The autoreactive responses manifested by some antigen-specificT cell lines might reflect the proportion of that population which is in an activated state. In contrast, those antigen-specific, MHC-restricted T cell lines that consistently fail to manifest any autoreactive component may have a relatively low affinity for self, sufficient for positive selection in the thymus but not for in vitro stimulation. Even if T cell autoreactivity is normally a transient physiological state in the periphery, it could still contribute significantly to clonal expansion and lymphokine secretion when antigen is limiting. Under conditions of chronic stimulation, when high levels of Ia are continuously expressed and/or immune regulation is abnormal, autoreactive T cells could give rise to severe and persistent inflammation.
434
MAURICE ZAUDERER
REFERENCES Arden, B., Klotz, J., Siu, G., and Hood, L. (1985). Diversity and structure of genes of the a family of mouse T-cell antigen receptor. Nature (London) 316, 783. Ashwell, J. D., DeFranco, A. L., Paul, W. E., and Schwartz, R. H. (1984). Antigen presentation by resting B cells. J. Ex#. Med. 159, 881. Ballardini, G., Mirakian, R.,Biachi, F. B., Pisi, E., Doniach, D., and Bottazzo, G. F. (1984). Aberrant expresion of HLA-DR antigens on bile duct epithelium in primary biliary cirrhosis: Relevance to pathogenesis. Lancet 2, 1009. Barth, R.,Kim, B.,Lan, N., Hunkapiller, T., Sobieck, N., Winoto, A., Gershenfeld, H., Okada, C., Hamburg, D., Weissman, I., and Hood, L. (1985). The murine T-cell receptor employs a limited repertoire of expressed Vg gene segments. Nature (London) 316, 517. Becker, D., Patten, P., Chien, Y. H., Yokota, T., Eshhar, Z., Gieldlin, M., Gascoigne, N. R.J., Goodenow, C., Wolf, R.,Arai, K. I., and Davis, M. M. (1985). Variability and repertoire size in T-cell receptor V, gene segments. Nature (London) 317, 430. Bensussan, A., Meuer, S. C., Schlossman, S. F.,and Reinherz, E. L. (1984). Delineation of an immunoregulatory amplifier population recognizing autologous Ia molecules: Analysis with human T cell clones. J. Ex#. Med. 159, 559. Bimbaum, G., Kotilinek, L., Schwartz, M., and Stemad, M. (1984). Spinal fluid lymphocytesresponsive to autologous and allogeneic cells in multiple sclerosis and control individuals. J. Clin. Invest. 74, 1307. Bjorkman, P.J.. Saper, M. A., Samraoui, B., Bennett, W. S.,Strominger, J. L., and Wiley, D. C. (1987). The foreign antigen binding site and T cell recognition regions of class 1 histocompatibility antigens. Nature (London) 329, 512. Bottazzo, G. E, Pujol-Borrell, R.,Hanafusa, T., and Feldmann, M. (1983). Role of aberrant HLA-DR expression and antigen presentation in induction of endocrine autoimmunity. Lancet 2, 1115. Bottazzo, G. F., Dean, B. M., McNally, J. M., MacKay, E. H., Swift, P. G. E. and Gamble, D. R. (1985). In situ characterization of autoimmune phenomena and expression of HLA molecules in the pancreas in diabetic insulitis. N. Eng1.J.Med. 313, 353. Chien, Y. H., Gascoigne, N. R. J., Kavaler, J., Lee, N. E., and Davis, M. M. (1984). Somatic recombination in a murine T cell receptor gene. Nature (London) 309, 322. Clayberger, C., DeKruyff, R. H., and Cantor, H. (1984). Immunoregulatory activities of autoreactive T-cells: An la-specific T cell clone mediates both help and suppression of antibody responses. J. Immunol. 132, 2237. Coeshott, C. M., Chesnut, R. W., Kubo, R. T., Grammer, S. F., Jenis, D. M., and Grey, H. M. (1986). la-specific mixed leukocyte reactive T cell hybridomas: Analysis of their specificity by using purified class I1 MHC molecules in a synthetic membrane system. J. Immunol. 136, 2832. DeKruyff, R. H., Clayberger, C., and Cantor, H. (1985). Monoclonal helper T cells induce B cell responses to T-independent antigens: Antigen-specific T cells are directly stimulated by activated B cells in the absence of antigen. J. Immunol. 134, 86. De Talance, A., Regnier, D., Spinella, S., Morisset, J., and Seman, J. (1986). Origin of autoreactive T helper cells. I. Characterization of Thy-1 , I+-, LST4- precursors in the spleen of normal mice. J. Immunol. 137, 1101. +
AUTOREACTIVE T CELLS
435
Dos Rejs, G. A., and Shevach, E. M. (1981). The syngeneic mixed leukocyte reaction represents polyclonal activation of antigen-specific T lymphocytes with receptors for self Ia antigens. J. Immunol. 127, 2456. Dos Reis, G. A., and Shevach, E. M. (1985). Analysis of autoreactive 1 region-restricted T cell colonies isolated from the guinea pig syngeneic mixed leukocyte reaction and from immune responses to conventional foreign antigens. Eur J. Immunol. 15, 466.
Faherty, D. A., Johnson, D. R., and Zauderer, M. (1985). Origin and specificity of autoreactive T cells in antigen-induced populations. J. Ex$. Med. 161, 1293. Fink, P.,Matis, L., McElligott, D..Bookman, M., and Hedrick, S. (1986). Correlations between T-cell specificity and the structure of the antigen receptor. Nature (London) 321, 219. Finnegan, A., and Hodes, R. J. (1986). Antigen-induced T suppressor cells regulate the autoreactive T helper-B cell interaction. J. Immunol. 136, 793. Finnegan, A., Needleman, B., and Hodes, R. J. (1984). Activation of B cells by autoreactive T cells: Cloned autoreactive T cells activate B cells by two distinct pathways. J. Immunol. 133, 78. Foulis, A. K., and Farquharson, M. A. (1986). Aberrant expression of HLA-DR antigens by insulin containing beta cells in recent onset type I (insulin dependent) diabetes mellitus. Diabetes 35, 1215. Frohman, M., and Cowing, C. (1985). Presentation of antigen by B cells: Functional dependence on radiation dose, interleukins, cellular activation, and differential glycosylation.J. Immunol. 134, 2269. Glazier, A,, Tutschka, P. J., Farmer, E. R.,and Santos, G. W. (1983). Graft-versus-host disease in cyclosporin A-treated rats after syngeneic and autologous bone marrow reconstituti0n.J. Exp. Med. 158, 1. Glimcher, L. H., and Shevach, E.M. (1982). Production of autoreactive I region restricted T cell hybrid0mas.J. Ex$. Med. 156, 640. Glimcher, L. H., liongo, D.L.. Green, I., and Schwartz, R. H. (1981). Murine syngeneic mixed lymphocyte response. I. Target antigens are self Ia molecu1es.J. Ex$.Med. 154, 1652.
Gosselin, E.J., Tony, H. -P., and Parker, D.C. (1988). Characterization of antigen processing and presentation by resting B lymphocytes. J. Immunol. 140, 1408. Griffith, I. J., Ghogawala, Z.., Nabavi, N., Golan, D. E., Myer, A., McKean, D.J , , and Glimcher, L. H. (1988). Cytoplasmic domain affects membrane expression and function of an l a molecule. Proc. Natl. Acad. Sci. U.S.A. 85, 4847. Hinch, M.-R., Weitzerbin, J., Pierres, M., and Garichs, C. (1983). Expression of Ia antigens by cultured astrocytes treated with gamma-interferon. Neurosci. Lett. 41, 199.
Huber, C., Merkenschlager, M., Gattringer, C., Royston, I., Fink, U.,and Braunsteiner, H. (1982). Human autologous mixed lymphocyte reactivity is primarily specific for xenoprotein determinants adsorbed to antigen-presenting cells during rosette formation with sheep erythrocytes. J. Ex$. Med. 155, 1222. Imperiale, M. J., Faherty, D.A., Sproviero,J. E and Zauderer, M. (1982). Functionally distinct helper T cells enriched under different culture conditions cooperate with different B ce1ls.J. Immunol. 129, 1843. Innes, J. B., Kuntz, M. M., Kim, Y. T., and Weksler, M. E. (1979). Induction of suppressor activity in the autologous mixed lymphocyte reaction and in cultures with concanavalin A.J. Clin. Invest. 64, 1608.
436
MAURICE ZAUDERER
Julius, M. H., and Heusser, C. H. (1986). Autoreactiveor self restricted?EuzJ. Immunol. 16, 867. Kagan, J., and Choi, Y. S. (1985). Failure of the human autologous mixed lymphocyte reaction in the absence of foreign antigens. Eur. J. Immunol. 13, 1031. Kotani, H., Mitsuya, H., Jarret, R. E , Yenokida, G. G., James, S. P., and Strober, W. (1986). An autoreactive T cell clone which can be activated to provide both helper and suppressor function. J. Immunol. 136, 1951. Kotani, H., Mitsuya, H., Benson, E.,James, S. P., andStrober, W. (1988). Activation and function of an autoreactive T cell clone with dual immunoregulatory activity. J. Immunol. 140, 4167. Krieger, J. I., Chesnut, R. W., and Grey, H. M. (1986). Capacity of B cells to function as stimulators of a primary mixed leukocyte reaction. J. Immunol. 137, 5117. Londei, M., Lamb, J. R., Bottazzo, G. E, and Feldmann, M. (1984). Epithelial cells expressing aberrant MHC class I1 determinants can present antigen to cloned human T cells. Nature (London) 312, 639. Londei, M., Bottazzo, G. F., and Feldmann, M. (1985). Human T cell clones from autoimmune thyroid glands: Specific recognition of autologous thyroid cells. Science 228, 85. Morisset, J., Trannoy, E., De Talance, A., Spinella, A., Debre, P., Godet, P., and Seman, M . (1988). Genetics and strain distribution of concanavalin A-reactive Ly-2-, LST4-, peripheral precursors of autoreactive T cells. Eur. J. Immunol. 18, 587. Nagarkatti, P. S., Nagarkatti, M., Mann, L. W., Jones, L. A.,and Kaplan, A. M. (1988). Characterization of an endogenous Lyt-2 T-suppressor-cellpopulation regulating autoreactive T cells in Mtro and in tnim. Cell. Immunol. 112, 64. Nussenzweig, M. C., and Steinman, R. M. (1980). Contribution of dendritic cells to stimulation of the murine syngeneic mixed leukocyte reaction. J. Ex#. Med. 151, 1196. Parham, P., Clayberger, C., Zorn, S. L., Ludwig, D. S., and Krensky, A. M. (1987). Inhibition of alloreactive cytotoxic T lymphocytes by peptides from the apdomain of HLA-A2. Nature (London) 325, 625. Pullen, A. M., and Munro, A. J. (1988). Fetal calf serum stimulates autoreactive T-cell hybridomas. Immunology 63, 255. Quintans, J., Suzuki, H., Sosman, J. A., and Shah, P. D. (1986). Immunoregulation by T cells. I. Characterization of the IEk-specific Lbd self-reactive T cell clone that helps, suppresses and contrasuppresses B cell responses.J. Immunol. 136, 1974. Rock, K. L., and Benacerraf, B. (1983). The role of Ia molecules in the activation of T lymphocytes. 111. Antigen-specific, Ia-restricted, interleukin 2-producing T cell hybridomas with detectable affinity for the restricting I-A molecu1e.J. Exp. Med. 157, 559. Rock, K. L., and Benacerraf, B. (1984). Thymic T cells are driven to expand upon interaction with self-class I1 major histocompatibility complex gene products on accessory cells. Proc. Natl. Acad. Sci. U.S.A. 81, 1221. Saito, K., Tamura, A., Narimatsu, H., Tadakuma, T., and Nagashima, M. (1986). Cloned auto-la-reactiveT cells elicit lichen planus-likelesion in the skin of syngeneic mice. J. Immunol. 137, 2485. Saito, T., and Rajewsky, K. (1985). A self-la reactive T cell clone directly stimulates every hundredth B cell and helps antigen-specificB cell responses. Euz J. Immunol. 15, 927. Sano, K., Fujisawa, I., Abe, R., Asano, Y.,and Tada, T. (1987). MHC-restrictedminimal +
AUTOREACTIVE T CELLS
437
regulatory circuit initiated by a class 11-autoreactive T cell clone. J. Exp. Med. 165, 1284. Schlesier, M., Ramb-Lindhauer, C., Gartner, M., and Peter, H. H. (1984). Analysis of T-cell cultures and clones from a patient with classic rheumatoid arthritis-Evidence for the existence of autoreactiveT cell clones in blood and synovial fluid. Rheumatol. Int. 4, Suppl., 1. Shiohara, T., Ruddle, N. H., Horowitz, M., Moellmann, G. E., and Lerner, A. B. (1987a). Anti-tumor activity of class I1 MHC antigen-restricted cloned autoreactive T cells. I. Destruction of B16 melanoma cells mediated by bystander cytolysis in vitro. J. Immunol. 138, 1971. Shiohara, T., Moellmann, G . , Jacobson, K., Kuklinska , E., Ruddle, N. H., and Lemer, A. B. (1987b). Anti-tumor activity of class I1 MHC antigen-restricted cloned autoreactive T cells. 11. Novel immunotherapy of B16 melanomas by local and systemic adoptive transfer. J. Immunol. 138, 1979. Shiohara, T., Moriya, N., Mochizuki, T., and Nagashima, M. (1987~).Lichenoid tissue reaction (LTR) induced by local transfer of Ia-reactive T-cell clones. 11. LTR by epidermal invasion of cytotoxic lymphokine-producingautoreactive T cells.J. Invest. Dennatol. 89, 8 . Smith, J. B., and Knowlton, R. P. (1979). Activation of suppressor cells in human autologous mixed lymphocyte culture. J. Immunol. 123, 419. Sorger, S. B., Hedrick, S. M., Fink, P. J., Bookman, M. A., and Matis, L. A. (1987). Generation of diversity in the T cell receptor repertoire specific for pigeon cytochrome c. J. Ex$. Med. 165, 279. Sproviero, J. F., Imperiale, M. J., and Zauderer, M.(1981). Clonal analysis of F1 hybrid helper T cells. I-A subregion-encoded hybrid determinants restrict the activity of keyhole limpet hernocyanin-specific helper T cells. J. Ex$. Med. 154, 1255. Takiguchi, M., Ting, J. P.-Y., Buessow, S. C., Boyer, C., Gillespie, Y.,and Frelinger, J. A. (1985). Response of glioma cells to interferon-gamma: Increase in class I1 RNA, protein and mixed lymphocyte reaction-stimulatingability. EucJ. Immunol. 15, 809. Todd, I., Pujol-Borell, R., Hammond, L. J., Bottazzo, G. E and Feldmann, M. (1984). Interferon-y induces HLA-DR expression by thyroid epithelium. Clin. Ex$. Immunol. 61, 265. Valdimarsson, H., Baker, B. S.,Jonsdottir, I., and Fry, L. (1986). Psoriasis: A disease of abnormal keratinocyte proliferation induced by T lymphocytes. Immunol. T o h y 7 , 256. Weksler, M. E., Moody, C. E., Jr., and Kozak, R. W. (1981). The autologous mixedlymphocyte reaction. Adv. Immunol. 31, 271. Wong, G. H. W., Bartlett, P.F., Clark-Lewis, I., Battye, E, andschrader, J. W. (1984). Inducible expression of H-2 and Ia antigens on brain cells. Nature (London) 310, 688. Wong, G. H. W., Bartlett, P. F., Clark-Lewis, I., McKimm-Breschkin, J. L., and Schrader, J. W. (1985). Interferon-y induces the expression of H-2 and Ia antigens on brain cells. J. Neuroimmunol. 7 , 255. Yamashita, U.,and Shevach, E. M. (1980). The syngeneic mixed leukocyte reaction: The genetic requirements for the recognition of self resemble the requirements for the recognition of antigen in association with self. J. Immunol. 124, 1773. Johnson, D. R., and Seman, M. (1984). Helper functions Zauderer, M., Campbell, H., of antigen-induced specific and autoreactive T-cell co1onies.J. Mol. Cell. Immunol. 1, 65.
Index
A
Accessory molecules, T cell receptor and, 108-1 10 Agretopes synthetic T and B cell sites and, 201, 202,210 T cell receptor and antigen processing, 122, 123 structure-function relationships, 135, 139 T cell repertoire, 166 AIDS humoral immune response and, 78 synthetic T and B cell sites and, 219, 264 virus-induced immunosuppression and, 335, 336, 351-357, 360, 363, 365 Alloreactivity autoreactive T cells and, 420, 421, 425 humoral immune response and, 40 synthetic T and B cell sites and, 234 T cell receptor and, 148-157 homogeneity, 128, 130 MHC molecules, 116 peptides, 138 polymorphic residues, 158, 159, 162 structure-function relationships, 136, 140 T cell repertoire, 165 Amino acids humoral immune response and, 18, 20 regulators of complement activation and, 381, 411 genes, 405, 406 protein expression, 399, 400, 402 RCA-like protein utilization, 409,410 short consensus repeat, 391, 392, 394 sporozoite malaria vaccine and antibodies, 300 CS proteins, 292, 294-300 CS-specific T cells, 311-313 human trials, 315 Plasmodium vivax, 317 synthetic T and B cell sites and, 196, 255, 256, 258, 262
antigens, 232, 233 bacterial antigens, 230, 231 candidate synthetic peptide vaccines, 253, 255, 256, 258 globular protein antigens, 210, 216 immunological considerations, 200 prediction, 251 viral antigens, 219-221, 227, 228 T cell receptor and alloreactivity, 151-153 antigen processing, 120-123 epitopes, 143, 145, 148 peptides, 138 polymorphic residues, 159 T cell repertoire, 165 virus-induced immunosuppression and, 350, 362, 364 Angiotensin 11. synthetic T and B cell sites and, 232, 233 Anopheles, sporozoite malaria vaccine and, 283, 284 Antibodies autoreactive T cells and, 427 humoral immune response and, 1-6 antigens, 10, 12 cellular interactions in vivo, 78, 81-83, 85 helper T cells, 24-30, 32 interleukins, 62, 67, 72, 74 physical interaction, 38, 40-54 regulators of complement activation and complement pathways, 381 protein expression, 396, 397 protein roles, 389, 390 sporozoite malaria vaccine and, 320, 321 CS proteins, 292, 299, 300 CS-specific T cells, 306-313 endemic areas, 315-317 human trials, 313-315 immunity, 286-292 interferon-y, 301, 304, 306 Plasmodium vivax, 317, 318 role, 300, 301
439
440
lNDEX
synthetic T and B cell sites and, 196, 260-264 antigens, 233, 236 bacterial antigens, 231 candidate synthetic peptide vaccines, 253, 255-259 globular protein antigens, 215, 216 immunological considerations, 197 parasitic antigens, 228, 229 peptides, 237-239 prediction, 251 viral antigens, 220-223,225, 227 T cell receptor and accessory molecules, 108, 109 alloreactivity, 152, 154 homogeneity, 128, 130 peptides, 138 polymorphic residues, 158, 160 structure-function relationships, 136,140 T cell repertoire, 164 virus-induced immunosuppression and human immunodeficiency virus, 351-353, 356, 357, 363, 365, 368 measles, 340, 341, 343, 345-347 Antigen-presenting cells (APCs) autoreactive T cells and activation, 421-427 origin, 419 regulation, 432 humoral immune response and, 5-8 B cells, 10-12 cellular interactions in vivo, 79-81, 84 class I1 molecules, 18-20 helper T cells, 24, 30, 31 interleukins, 55, 62-64, 71, 73 physical interaction, 35,36,41,42,45,53 processing, 8, 9, 12, 13 types, 9, 10 in vitm, 21 in vivo, 21-24 sporozoite malaria vaccine and, 287, 305, 307, 316 synthetic T and B cell sites and, 197, 201,202 antigens, 232, 237 globular protein antigens, 215 prediction, 252 viral antigens, 220, 225 T cell receptor and, 107 accessory molecules, 108, 109 alloreactivity, 153
antigen processing, 118, 120, 123, 124,127 epitopes, 143, 144 homogeneity, 129 MHC molecules, 116 peptides, 132, 138 T cell repertoire, 165 virus-induced immunosuppression and, 366 Antigens autoreactive T cells and, 417, 418, 433 activation, 421-426 origin, 418-421 physiology, 427, 428,430, 431 regulation, 431, 432 humoral immune response and, 2, 5-7 cell surface molecules, 41-46 cellular interactions in vivo, 79, 80, 82-85 helper T cells, 24-33, 35 lymphokines, 46-54 physical interaction, 35-41 presentation, 9-12, 15-24 processing, 8, 9, 12-15, 17, 18 regulators of complement activation and, 381, 389 sporozoite malaria vaccine and, 283, 319-322 CS proteins, 299 CS-specific T cells, 307-311 endemic areas, 317 human trials, 314 immunity, 287, 291, 292 interferon-y, 304-3015 Plasmodium vivax, 318 synthetic T and B cell sites and, 195, 197, 260-264 bacterial antigens, 230-232 candidate synthetic peptide vaccines, 253-259 globular protein antigens, 203-216 immunological considerations, 197-202 parasitic antigens, 230 peptides, 203, 232-250 prediction, 251 viral antigens, 216-228 T cell receptor and, see T cell receptor virus-induced immunosuppression and human immunodeficiency virus, 351-353, 355-358, 360-362, 365-367 measles, 339, 344, 346, 347
441
INDEX
Antipain, T cell receptor and, 124 Apamin, synthetic T and B cell sites and, 232 Autoantibodies, synthetic T and B cell sites and, 247, 248 Autologous mixed lymphocyte responses, autoreactive T cells and, 418, 425, 426
Autoreactive T cells, 417, 418, 433 activation specificity, 425-427 thymic stromal cells, 421-425 origin, 418, 419 antigen-specific T cells, 420 antigen-stimulated precursors, 419, 420 immature precursors, 421 physiology autoimmune response, 430, 431 nonspecific helper function, 427-429 tumor resistance, 428, 430 regulation, 431 stimulators, 432, 433 suppressor T cells, 431, 432 B
B cell activation factor I, humoral immune response and, 67 B cell growth factor I, humoral immune response and, 65, 66 Bacteria regulators of complement activation and, 388, 409 T cell receptor and, 125 Bacterial antigens humoral immune response and, 82, 83 synthetic T and B cell sites and, 230-232, 245
Brain autoreactive T cells and, 430 virus-induced immunosuppression and human immunodeficiency virus, 349, 351, 360, 365
measles, 346
C Caprine arthritis encephalitis virus (CAEV), virus-induced immunosuppression and. 349, 351, 354, 365
CD3 cells, T cell receptor and, 109, 110 CD4 cells humoral immune response and, 20, 44, 45 sporozoite malaria vaccine and, 303, 304, 319
T cell receptor and, 109, 110, 127 virus-induced immunosuppression and, 349-358, 360-363, 366
CD8 cells humoral immune response and, 20 sporozoite malaria vaccine and, 319 CS-specific T cells, 312 interferon-?, 303-306 T cell receptor and, 109, 110 virus-induced immunosuppression and human immunodeficiency virus, 352, 353, 356-358, 366
measles, 339 cDNA humoral immune response and, 63, 66, 72, 74, 75
regulators of complement activation and genes, 403, 405 protein expression, 396, 398, 399 short consensus repeat, 391 virus-induced immunosuppression and, 349- 35 1
Chloroquine humoral immune response and antigens, 13-15, 17, 20 physical interaction, 38, 51 synthetic T and B cell sites and, 223 Chromosomes humoral immune response and, 32 regulators of complement activation and, 381, 403, 404, 411
T cell receptor and, 110, 115 Circumsporozoite (CS) proteins sporozoite malaria vaccine and, 292-294, 319-322
antibodies, 300 cloning, 294-296 endemic areas, 315-317 human trials, 315 immunity, 291, 292 interferon-?, 306 Plasmodium vivax, 317, 3 18 repeats, 2%-300 synthetic T and B cell sites and, 228, 258 Circumsporozoite reaction, sporozoite malaria vaccine and. 288, 290
INDEX
Clones autoreactive T cells and, 418, 433 activation, 421-426 origin, 418-420 physiology, 427, 428, 430, 431 regulation, 431, 432 humoral immune response and antigens, 16. 17, 24 cellular interactions in vivo, 80, 83 helper T cells, 25-28, 30-32 interleukins, 62, 63, 66, 69, 74, 75 physical interaction, 35-37 regulators of complement activation and, 391, 396, 398, 399
sporozoite malaria vaccine and antibodies, 300 CS proteins, 292, 294-296, 299 CS-specific T cells, 310-312 synthetic T and B cell sites and, 261, 262 bacterial antigens, 230-232 globular protein antigens, 211, 214 parasitic antigens, 229 viral antigens, 218-220, 222-224 T cell receptor and accessory molecules, 111 alloreactivity, 149, 151-156 antigen processing, 119, 120, 125, 126 epitopes, 143 experimental systems, 131 polymorphic residues, 159-162 structure-function relationships, 135, 140, 141
virus-induced immunosuppression and human immunodeficiency virus, 349-351, 366 measles, 346
Conalbumin. humoral immune response and, 37 Concanavalin A autoreactive T cells and, 421 humoral immune response and, 45 virus-induced immunosuppression and, 357, 362
Cyclosporin A, humoral immune response and, 47 Cytochrome c autoreactive T cells and, 419 synthetic T and B cell sites and bacterial antigens, 240 globular protein antigens, 203, 204, 210, 211, 214
immunological considerations, 201 T cell receptor and antigen processing, 121, 127 epitopes, 143 experimental systems, 131, 132 structure-function relationships, 135, 136 Cytokines, humoral immune response and, 1, 55-61, 73
Cytotoxic T cells (CTLs), synthetic T and B cell sites and, 262, 264 antigens, 234, 235 immunological considerations, 199 viral antigens, 218-220, 226, 221 Cytotoxic T lymphocytes (CTLs) autoreactive T cells and, 425 human immunodeficiency virus and, 356, 358, 365, 366
measles and, 341, 345, 346 sporozoite malaria vaccine and, 312 T cell receptor and accessory molecules, 110 alloreactivity, 151-157 antigen processing, 118-120, 125-127 experimental systems, 137 MHC molecules, 116 peptides. 137, 138 polymorphic residues, 159, 162 structure- function relationships, 139-141 T cell repertoire, 166
D Decay-acceleration factor (DAF), regulators of complement activation and genes, 403, 408 protein expression, 395-402 protein interactions, 383 protein roles, 385-388, 390, 391 Delayed-type hypersensitivity (DTH), humoral immune response and, 24, 27, 79, 85 Dinitrophenyl (DNP), synthetic T and B cell sites and, 199 DNA humoral immune response and, 6, 32, 35 sporozoite malaria vaccine and antibodies, 301 CS proteins, 297 immunity, 291 interferon-y, 302
443
INDEX
Plasmodium vivax, 317
synthetic T and B cell sites and, 231 T cell receptor and, 109, 165 virus-induced immunosuppression and human immunodeficiencyvirus, 354,355 measles, 344, 348 E
Electron microscopy humoral immune response and, 37 regulators of complement activation and, 392, 393
Endoplasmic reticulum regulators of complement activation and, 400 T cell receptor and, 123, 125 Epitopes autoreactive T cells and. 420 humoral immune response and antigens, 8, 19 cellular interactions in vivo, 80 helper T cells, 28 physical interaction, 44 sporozoite malaria vaccine and, 283, 319-322
CS proteins, 292, 299 CS-specific T cells, 307, 308, 310-313 endemic areas, 315 human trials, 313 interferon-?, 306
Plasmodium vivax, 311, 318
synthetic T and B cell sites and, 195-197, 260-264
antigens, 236 bacterial antigens, 230-232 globular protein antigens, 210-214 immunological considerations, 201, 202 parasitic antigens, 228-230 peptides, 250 vaccines, 252, 255-258 viral antigens, 217-220, 222, 223 T cell receptor and, 142 amphipathic character, 144, 145 antigen processing, 121, 122 conformation, 145, 146 experimental systems, 132 helical conformation, 142-144 location, 146-148 peptides, 137
structure-function relationships, 135
T cell repertoire, 164-166
virus-induced immunosuppression and, 363-365. 367
Epstein-Barr virus (EBV) humoral immune response and, 16, 17 regulators of complement activation and, 388, 409
synthetic T and B cell sites and, 226, 240, 241 T cell receptor and, 138 virus-induced immunosuppression and, 351, 352, 356, 358
Erythrocytes humoral immune response and, 17, 63, 65 regulators of complement activation and protein expression, 395, 396, 399, 400,402
protein roles, 388, 389 Escherichia coli sporozoite malaria vaccine and, 315 synthetic T and B cell sites and, 257 T cell receptor and, 126, 165 Exoerythrocytic forms (EEFs), sporozoite malaria vaccine and, 284, 319 antibodies, 301 CS-specific T cells, 307, 312 human trials, 314 immunity, 285, 286, 289-291 interferon-y, 301-303, 305, 306 Exons humoral immune response and, 32 regulators of complement activation and, 404-406,411 T cell receptor and, 136 Experimental allergic autoimmune encephalomyelitis, synthetic T and B cell sites and, 235, 236
F
Fetal calf serum, autoreactive T cells and, 426 Fibroblasts regulators of complement activation and protein expression, 395, 397, 399 protein roles, 387 RCA-like protein utilization. 410 synthetic T and B cell sites and, 217 T cell receptor and, 121
444
lNDEX
Follicular dendritic cell, humoral immune response and, 9,24, 79, 80 Foot and mouth disease virus, synthetic T and B cell sites and, 227, 253, 254
G Globular protein antigens, synthetic T and B cell sites and, 203-216, 240 Glycoprotein humoral immune response and, 42 regulators of complement activation and, 399, 400, 410 synthetic T and B cell sites and, 225, 226,255 virus-induced immunosuppression and, 350 human immunodeficiency virus, 361, 364-366 measles. 339, 346 Glycosylation humoral immune response and, 42 regulators of complement activation and, 400-403 Graft-versus-host disease, autoreactive T cells and, 432
H Hapten-carrier effect, humoral immune response and, 3, 4, 6,7 Hemagglutinin sporozoite malaria vaccine and. 307 synthetic T and B cell sites and peptides, 241, 242 vaccine, 253-255 viral antigens, 216, 219 T cell receptor and, 122, 125, 133, 135 virus-induced immunosuppression and, 346 Hen egg lysozyme, synthetic T and B cell sites and globular protein antigens, 204, 211-214 peptides, 240 Hepatitis B virus (HBV), synthetic T and B cell sites and, 196, 261, 263 bacterial antigens, 230 globular protein antigens, 206, 212, 214 immunological considerations, 201, 202 peptides, 202, 239, 242, 243
vaccine, 254-256 viral antigens, 220, 221, 224 Herpes simplex virus regulators of complement activation and, 409,410 synthetic T and B cell sites and, 225, 243,256 T cell receptor and, 132 virus-induced immunosuppression and, 367 HLA-A2 synthetic T and B cell sites and, 219, 223, 234, 235 T cell receptor and alloreactivity, 150, 151, 153, 155, 156 epitopes, 142, 144 homogeneity, 129 MHC molecules, 111, 112, 115, 116 peptides, 137, 138 polymorphic residues, 158, 160, 161 structure-function relationships, 139-141 Homogeneity, T cell receptor and, 128-131 Hormones sporozoite malaria vaccine and, 305, 312 synthetic T and B cell sites and, 232, 246, 247, 259 Human chorionic gonadotropin, synthetic T and B cell sites and, 259 Human fibrinopeptide B, synthetic T and B cell sites and, 233 Human immunodeficiency virus humoral immune response and, 78 sporozoite malaria vaccine and, 308-310 synthetic T and B cell sites and, 196, 263 globular protein antigens, 207 peptides, 243 viral antigens, 219, 220 virus-induced immunosuppression and, 335, 336, 355-363 history of infection, 352-355 immune response, 363-368 virology, 347-352 Humoral immune response, 1 antibody response, 4 antigen-bridging model, 7, 8 antigen presentation, 9-12,22 B cells, 15-17 class I1 molecules, 18-21 in virtu, 21 in vivo, 21, 23, 24 antigen processing, 12, 22 B cells, 17, 18
445
INDEX
intracellular events, 12-14 macrophages, 14, 15 requirements, 8, 9 antigen-specific receptors, 6, 7 cellular interactions in vivo, 78, 79, 85 bacterial antigens, 82, 83 lymphoid organ, 79-82 parasites, 84, 85 viral antigens, 83, 84 hapten-carrier effect, 3, 4 helper T cells, 24, 25 activation, 29-31 clones, 29 differences, 27. 28 heterogeneity, 25-27 isotype switching in B cells, 32-35 subtype surface markers, 28, 29 in uivo, 31, 32 interleukins, 54-62, 75-77 ILl, 72, 73 IL2, 67-70 IL3,73, 74 IL4, 62-65 IL5, 65-67 IL-6, 74, 75 interferon-r, 70, 71 models, 1-3 physical interaction, 35, 36 biochemical events, 38, 39 cell surface Ig, 40, 41 cell surface molecules, 41-46 functional outcome, 38 intracellular events, 39 lymphokines, 46-54 model systems, 36-38 monogamous nature, 39, 40 suppression, 5 Hybridization autoreactive T cells and, 420, 424 regulators of complement activation and. 405 T cell receptor and, 135. 139, 140 virus-inducedimmunosuppressionand, 354 Hybridomas autoreactive T cells and, 420, 424-426 humoral immune response and antigens, 15 interleukins, 63, 65, 74, 75 physical interaction, 36-39, 44,50 synthetic T and B cell sites and antigens, 232, 234, 237
globular protein antigens, 213, 215, 216 immunological considerations, 200 viral antigens, 217, 225, 228 T cell receptor and, 151
I Icoccomes, humoral immune response and, 24 Immune complexes, regulators of complement activation and, 388, 389, 410, 411 Immune response (Ir) genes humoral immune response and, 5 sporozoite malaria vaccine and, 308-310 synthetic T and B cell sites and antigens, 232 globular protein antigens, 210, 211. 214 immunological considerations, 198-201 viral antigens, 221 Immunoglobulins autoreactive T cells and, 427 human immunodeficiency virus and, 351, 353, 356, 362, 365 humoral immune response and, 4, 6, 8 antigens, 10, 12, 13, 16-18, 21 cellular interactions in uiuo, 80, 82-84 helper T cells, 26, 27, 29, 32-35 interleukins, 55, 63-72, 76 physical interaction, 40,43, 44, 49, 50, 52 measles and, 340, 341 regulators of complement activation and, 382, 390, 398 sporozoite malaria vaccine and, 286, 307 T cell receptor and accessory molecules, 109 antigen processing, 118 experimental systems, 131 homogeneity, 129 MHC molecules, 111 Immunosuppression, virus-induced, see Virus-induced immunosuppression Inflammatory response, autoreactive T cells and, 428-430, 433 Influenza sporozoite malaria vaccine and, 307 synthetic T and B cell sites and antigens, 234 vaccine, 253-255 viral antigens, 216-219
446
INDEX
T cell receptor and antigen processing, 119, 122, 125, 126 epitopes, 143 homogeneity, 128 peptides, 133, 137 structure-function relationships, 135, 141 virus-induced immunosuppression and, 356 Insulin autoreactive T cells and, 420 synthetic T and B cell sites and, 202, 209, 214, 237 T cell receptor and, 117, 124, 131 Intercellular adhesion molecule-I (ICAM-I), humoral immune response and, 43 Interferon T cell receptor and, 117 virus-induced immunosuppression and, 343, 361 Interferon-a regulators of complement activation and, 397 sporozoite malaria vaccine and, 302 virus-induced immunosuppression and, 339, 344 Interferon-@ regulators of complement activation and, 397 sporozoite malaria vaccine and, 302 Interferon-y autoreactive T cells and, 427, 430 humoral immune response and, 61, 62, 68, 70-72, 76 cellular interactions in vivo, 82, 84, 85 helper T cells, 26-30, 33 regulators of complement activation and, 397 sporozoite malaria vaccine and, 319 CS-specific T cells, 307, 312 liver stages, 301-306 T cell receptor and, 117, 119, 129 virus-induced immunosuppression and, 356, 357, 360, 361, 363, 366 Interleukin, humoral immune response and, 54-62, 75-77 Interleukin-I autoreactive T cells and, 423, 424 humoral immune response and, 56, 68, 70-75, 80 regulators of complement activation and. 397 virus-induced immunosuppression and, 357
Interleu kin-2 autoreactive T cells and, 425 humoral immune response and, 57, 62, 66-71, 73, 76 cellular interactions in vivo, 84 helper T cells, 26, 28-33 physical interaction, 44 regulators of complement activation and, 394, 402, 404 virus-induced immunosuppression and human immunodeficiency virus, 356, 358, 360, 362. 363, 366, 367 measles, 342, 344 Interleukin-3, humoral immune response and. 26, 57, 65, 73-75 Interleukind autoreactive T cells and. 427 humoral immune response and, 58, 62-67, 69-71, 75, 76 cellular interactions in vivo, 82-84 helper T cells, 26-34, 42, 44 Interleukin-5, humoral immune response and cellular interactions in vivo, 82-84 helper T cells, 26, 28, 32, 33 physical interaction, 48 Interleukin-6, humoral immune response and, 60,74, 75, 82
K Keyhole limpet hemocyanin autoreactive T cells and, 419, 421 humoral immune response and, 10, 11, IS, 33 sporozoite malaria vaccine and, 301 synthetic T and B cell sites and, 227, 255, 256 virus-induced immunosuppression and, 361, 362 Kidney regulators of complement activation and, 389 virus-induced immunosuppression and, 337
L
Latent membrane protein, synthetic T and B cell sites and, 226
447
INDEX
LentiviruseS,virus-induced immunosuppression and, 347-349, 351,355
Leupeptin synthetic T and B cell sites and, 223 T cell receptor and, 124 LFA-1,humoral immune response and, 39, 41, 43, 45, 51, 52, 54 Lipids, T cell receptor and, 120, 124 Lipopolysaccharide autoreactive T cells and, 428 humoral immune response and helper T cells, 27, 33, 34 interleukins, 63, 64, 66, 68, 70 physical interaction, 44, 46, 53 regulators of complement activation and, 397 virus-induced immunosuppressionand, 357 Liver regulators of complement activation and, 410 protein expression, 397, 399 protein interactions, 384 protein roles, 388, 389 sporozoite malaria vaccine and, 319 antibodies, 301 CS proteins, 300 CS-specific T cells, 312 immunity, 287, 289, 290 interferon-7, 301-306 Plasmodium v i v a , 317 synthetic T and B cell sites and, 228 T cell receptor and, 129 Low-density lipoprotein, regulators of complement activation and, 402 L3T4, humoral immune response and, 25,
26, 44,45
Luteinizing hormone, synthetic T and B cell sites and, 259 Lymph nodes autoreactive T cells and, 419 humoral immune response and antigens, 23, 24 cellular interactions in vivo, 79, 81 physical interaction, 54 sporozoite malaria vaccine and, 305 T cell receptor and, 134 virus-induced immunosuppression and, 350, 353, 354, 360 Lymphocytic choriomeningitis virus (LCMV),synthetic T and B cell sites and. 226
Lymphoid organ, humoral immune response and, 79-82 Lymphokine-activated killer cells, humoral immune response and, 68, 70 Lymphokines autoreactive T cells and, 430, 431, 433 humoral immune response and, 5 cellular interactions in vivo, 85 helper T cells, 24-35 interleukins, 55, 62, 65, 67, 69, 71-76 physical interaction, 36, 39, 41, 46-54 regulators of complement activation and, 389 sporozoite malaria vaccine and, 303,304,
306. 312
synthetic T and B cell sites and, 219, 248 Lysosomes humoral immune response and antigens, 12-14, 20, 21 cellular interactions in vivo, 79 synthetic T and B cell sites and, 215 Lysozyme synthetic T and B cell sites and, 204, 211,
213, 214
T cell receptor and
antigen processing, 122, 123 experimental systems, 131 homogeneity, 129 peptides, 133, 134 structure-function relationships, 135, 136 M
Major histocompatibility complex (MHC) autoreactive T cells and, 417, 418,433 activation, 421, 423, 427 origin, 418-420 physiology, 428, 430 regulation, 432 human immunodeficiency virus and, 349,
361, 365-367
humoral immune response and, 5-8, 80 antigens. 8, 10, 19, 20 helper T cells, 25, 26 interleukins, 62, 66,67 physical interaction, 37, 41. 43, 46-48,
52-54
measles and, 345, 346 regulators of complement activation and. 404
448
lNDEX
sporozoite malaria vaccine and CS-specific T cells, 307, 310 immunity, 287 interferon-? 304, 305 synthetic T and B cell sites and, 197, 263 antigens, 209, 212-214, 234-236 immunology, 198-200. 202 prediction, 252 viral antigens, 217, 218, 221, 225, 226 T cell receptor and, see T cell receptor Malaria regulators of complement activation and, 389 synthetic T and B cell sites and globular protein antigens, 207 parasitic antigens, 228, 229 vaccine, 254 viral antigens, 222 T cell receptor and, 132, 138 vaccine, see Sporozoite malaria vaccine Measles T cell receptor and, 124, 127 virus-induced immunosuppression and, 335-347, 358 Member cofactor protein, regulators of complement activation and genes, 403, 404,408 protein expression, 395-398,400-402 protein interactions, 383 protein roles, 386, 387, 390 RCA-like protein utilization, 410 short consensus repeat, 391 PMicroglobulin T cell receptor and alloreactivity, 155, 156 MHC molecules, 111, 115 structure-function relationships, 140 virus-induced immunosuppression and, 353 Microtubule organizing centers, humoral immune response and, 37. 38, 42 Mitochondria, T cell receptor and, 123 Mitogen autoreactive T cells and, 426 humoral immune response and, 67, 70, 72 antigens, 23 helper T cells, 26, 29, 33-35 interleukins, 64,73, 75 physical interaction, 43 virus-induced immunosuppression and
human immunodeficiency virus, 355, 356, 358, 359,361, 362 measles, 338, 339. 342, 344 Mixed lymphocyte reaction (MLR), virusinduced immunosuppression and, 341 Monoclonal antibodies autoreactive T cells and. 420 regulators of complement activation and, 390, 396, 397, 410 sporozoite malaria vaccine and, 292, 293, 306, 312 synthetic T and B cell sites and, 238,239 T cell receptor and, 130, 141, 153, 156 virus-induced immunosuppression and, 346. 357, 362 Mononucleosis, virus-induced immunosuppression and, 352 mRNA humoral immune response and, 40, 62 regulators of complement activation and, 397-400 sporozoite malaria vaccine and, 294 T cell receptor and. 165 virus-induced immunosuppression and, 344, 361 Multiple-antigen peptide (MAP), sporozoite malaria vaccine and. 321 Mutation autoreactive T cells and, 417,420,424, 427 sporozoite malaria vaccine and, 298, 311. 312 synthetic T and B cell sites and, 217, 228, 229, 261 T cell receptor and alloreactivity, 150-152, 154, 156, 157 homogeneity, 128, 129 structure-function relationships, 135, 139-142 T cell repertoire, 164, 165 virus-induced immunosuppression and, 349-351 Myelin basic protein, synthetic T and B cell sites and, 209, 235-237 Myoglobin humoral immune response and, 19 synthetic T and B cell sites and globular protein antigens, 204, 211, 212, 215 peptides, 203, 240, 250 T cell receptor and, 123, 131, 134
INDEX
449
N Natural killer cells humoral immune response and, 68, 70, 83, 84
regulators of complement activation and, 395, 3%
virus-induced immunosuppression and, 341, 357, 358, 360, 363
Nephritis regulators of complement activation and, 389 virus-inducedimmunosuppressionand, 338 Neuraminidase, humoral immune response and, 42, 44 Neutralization, virus-induced immunosuppression and, 365 Nucleoprotein synthetic T and B cell sites and, 261, 263 antigens, 234 bacterial antigens, 230 viral antigens, 217, 218 T cell receptor and antigen processing, 119, 125 homogeneity, 128 peptides, 137 structure-function relationships, 139 Nucleotides, T cell receptor and, 109
0 Oncogenes regulators of complement activation and, 395 synthetic T and B cell sites and, 248, 249 virus-induced immunosuppressionand, 348 Ovalbumin humoral immune response and, 11 synthetic T and B cell sites and globular protein antigens, 204, 210, 214-216
immunological considerations, 200 viral antigens, 228 T cell receptor and alloreactivity. 153 antigen processing, 123, 125 experimental systems. 131 peptides, 133
P Parasites, humoral immune response and, 84. 85
Parasitic antigens, synthetic T and B cell sites and, 228-230, 245 Paroxysmal nocturnal hemoglobinuria (PNH), regulators of complement activation and, 387, 400 Peptides autoreactive T cells and, 425, 426 humoral immune response and, 7, 18-21 regulators of complement activation and, 389, 394, 399, 400 sporozoite malaria vaccine and, 320-322 antibodies, 304, 305 CS proteins, 296 CS-specific T cells, 307, 308, 310, 311 human trials, 314, 315 immunity, 287 synthetic T and B cell sites and, see Synthetic T and B cell sites T cell receptor and, 107, 108 alloreactivity, 148-157 antigen processing, 118-123, 125, 127, 128 binding, 132-134 epitopes, 142-148 experimental systems, 131, 132, 137 homogeneity, 129-131 MHC molecules, 112, 115, 116 polymorphic residues, 158-161 selection, 137-139 structure-function relationships, 135, 139-142
T cell repertoire, 162, 163, 165-167 virus-induced immunosuppression and human immunodeficiency virus, 350, 351, 362, 363, 366-368
measles, 337 Peripheral blood mononuclear cells (PBMCs) human immunodeficiency virus and, 350, 352. 354-362, 367
measles and, 338, 340, 342, 346 Phosphatidyl inositol-specificphospholipase C, regulators of complement activation and, 399 Phytohemagglutinin, virus-induced immunosuppression and human immunodeficiency virus, 356-358, 361-363
INDEX
measles, 338 Plasma, regulators of complement activation and, 387, 388, 397 Plasma cells, humoral immune response and, 81-83 Plasmacytomas, humoral immune response and, 74, 75 Plasmodium, sporozoite malaria vaccine and, 296, 297, 299 Plasmodium betghei, sporozoite malaria vaccine and CS-specific T cells, 308, 309 immunity, 285-290 interferon-y, 302, 306 Plasmodiumfdcipatum sporozoite malaria vaccine and, 283, 317, 320-322
CS-specific T cells, 308-311 endemic areas, 315, 316 human trials, 313 immunity, 285, 286, 288, 290, 292 interferon-y, 306 synthetic T and B cell sites and, 229, 258 T cell receptor and, 143 PIasmodium malariae, sporozoite malaria vaccine and, 283, 293, 294, 297 Plasmodium vivax, sporozoite malaria vaccine and, 283 CS proteins, 297-299, 317, 318 CS-specific T cells, 308-310, 313 immunity, 286, 288, 292 interferon-y, 302 Pokeweed mitogen, virus-induced immunosuppression and human immunodeficiency virus, 356, 362 measles. 340-342, 346 Polyadenylation, regulators of complement activation and, 407, 408 Polyclonal B cell activation (PBA), virusinduced immunosuppression and, 356, 357, 362
Poly(L-lysine) (PLL), synthetic T and B cell sites and, 198, 199 Polymorphism regulators of complement activation and, 388, 396, 407
sporozoite malaria vaccine and, 311-313 synthetic T and B cell sites and, 229 T cell receptor and, 108 alloreactivity, 149. 154, 156 homogeneity, 129
MHC molecules, 112, 115-117 peptides, 132 residues, 157-162 structure-function relationships, 140 T cell repertoire, 162, 167 Polymorphonuclear cells, regulators of complement activation and, 389, 390, 403, 410
Polypeptides regulators of complement activation and, 392, 405
sporozoite malaria vaccine and CS proteins, 294, 295, 300 human trials, 315 Plasmodium vivax, 318 synthetic T and B cell sites and bacterial antigens, 238 prediction, 251 vaccine, 253, 255 viral antigens, 221, 224 T cell receptor and accessory molecules, 109 antigen processing, 120 homogeneity. 130 MHC molecules, 112 T cell repertoire, 165 Properdin, sporozoite malaria vaccine and, 296 Protein humoral immune response and antigens, 10, 14, 15, 17, 18 helper T cells, 35 interleukins. 72, 74, 75 regulators of complement activation and, 381-383, 410, 411
expression, 394-403 genes, 403, 405, 408 interactions, 383, 384 roles, 384-390 short consensus repeat, 391-394 utilization, 409, 410 sporozoite malaria vaccine and, 319-322 antibodies, 300 CS proteins, 292-300 CS-specific T cells, 306-313 immunity, 291, 292 interferon-y, 306 Plasmodium vivax, 317, 318 synthetic T and B cell sites and, 195, 197, 259-264
antigens, 234
45 1
INDEX
globular protein antigens, 203-216 immunological considerations, 201, 202 parasitic antigens, 228-230 peptides, 203, 230-232, 237-239 prediction, 250-252 vaccine, 255-258 viral antigens, 218,219,221,223,227,228 T cell receptor and, 107 accessory molecules, 110 alloreactivity, 149 antigen processing, 118, 119, 124. 125, 127
epitopes, 143-148 experimental systems, 131, 132, 137 homogeneity, 128, 129 MHC molecules, 117 peptides, 138, 139 polymorphic residues, 160 structure-function relationships, 139 T cell repertoire, 165, 166 virus-induced immunosuppression and human immunodeficiency virus, 348, 349, 361-363, 365-367
measles, 337-339, 346 Protein kinase C, humoral immune response and, 43 Purified protein derivative (PPD) autoreactive T cells and, 419 humoral immune response and, 16 virus-induced immunosuppression and,
tissue-specific, 394-398 protein interactions, 383, 384 protein roles autologous tissue, 386, 387 history, 384, 385 immune complexes, 388, 389 plasma, 387, 388 transmembrane functions, 389, 390 RCA-like protein utilization, 409, 410 short consensus repeat common structural motif, 391, 392 localization, 392-394 Retrovirus, virus-induced immunosuppression and human immunodeficiency virus, 348, 349, 361, 365
measles, 344 Reverse transcriptase, virus-induced immunosuppression and, 358, 360, 361, 366
RNA regulators of complement activation and, 396, 398. 399, 407
sporozoite malaria vaccine and, 302 T cell receptor and, 116 virus-induced immunosuppression and human immunodeficiency virus, 348, 354, 355, 366
measles, 336, 344, 346
355, 356
S R
Rabies, synthetic T and B cell sites and, 226, 244
Regulators of complement activation, 381, 410, 411
complement pathways, 381-383 genes divergent function, 408, 409 exon structure, 404, 405 gene duplication, 405 intragenic duplication, 406,407 organization, 403, 404 polyadenylation, 407, 408 protein expression alternate RNA processing, 398, 399 anchoring systems, 399, 400 differential glycosylation, 400-403
Sclerosing panencephalitis (SSPE), virusinduced immunosuppression and, 346 Sheep red blood cells (SRBCs), humoral immune response and, 1, 2, 68, 70, 72 Short consensus repeat, regulators of complement activation and common structural motif, 391, 392 genes, 404-408 localization, 392-394 protein expression, 399 protein utilization, 409,410 Signal transduction regulators of complement activation and, 384, 410
T cell receptor and, 110 Spleen autoreactive T cells and activation, 421, 422
452
INDEX
physiology, 428 regulation, 432, 433 humoral immune response and, 2 antigens, 11, 16, 17 helper T cells, 27, 34 physical interaction, 40,53 regulators of complement activation and, 384, 388, 389, 410 sporozoite malaria vaccine and, 287, 305,306 synthetic T and B cell sites and, 199 T cell receptor and, 134 virus-induced immunosuppression and, 354 Sporozoite malaria vaccine, 283, 284 antibodies, 300, 301 CS proteins, '92-294 cloning, 294-296 evolution of repeats, 2%-299 function of repeats, 299, 300 CS-specific T cells, 306-308 epitopes, 310- 313 Ir gene control, 308-310 endemic areas, 315-317 human trials, 313-315 immunity, 285, 286 specificity, 287-292 viability, 286, 287 interferon-y, 301-306 perspectives, 319-322 Plasmodium v i v a , 317, 318 Steroids, virus-induced immunosuppression and, 338 Stromal cells, thymic, autoreactive T cells and, 421-425 Supernatants, humoral immune response and, 5 antigens, 15 interleukins, 62, 65, 67, 69, 70 physical interaction, 54 Suppressor T cells, autoreactive T cells and, 431, 432 Synctium formation, virus-induced immunosuppression and, 349, 350, 358, 359 Syngeneic mixed lymphocyte responses, autoreactive T cells and activation, 426 origin, 418-420 regulation, 432, 433
Synthetic T and B cell sites, 159-164, 195-197 B cell sites, synthetic peptides and, 237-250 candidate synthetic peptide vaccines, 252-259 immunological considerations, 197-202 prediction B cell, 250, 251 T cell, 251, 252 T cell sites, synthetic peptides and, 202, 203 antigens, 232-237 bacterial antigens, 230-232 globular protein antigens, 203-216 parasitic antigens, 228-230 viral antigens, 216-228
T
T cell receptor, 107, 108, 131 accessory molecules, 108, 109 CD3 complex, 109, 110 alloreactivity, 148-157 antigen processing class I molecules, 118, 119 class I1 molecules, 117, 118 exceptions, 119, 120 factors, 122-124 molecular mechanisms, 127, 128 pathways, 124-127 peptides, 120, 121 recycling, 121, 122 autoreactive T cells and, 417, 419 epitopes, 142 conformation, 145, 146 helical conformation, 142-144 location, 146-148 experimental systems, 131, 132, 137 homogeneity peptidic self model, 128, 129 self-component, 128 self-peptides, 129-13 1 humoral immune response and, 5, 6 antigens, 18, 19 physical interaction, 36, 38, 41, 44-46, 52-54 MHC molecules, 110, 111 antigen presentation, 115-117 class I1 molecules, 112-115
453
INDEX
HLA-A2, 111 models, 111, 112 peptides binding, 132-134 selection, 137-139 polymorphic residues, 157-160 modeling, 160, 161 recognition, 161, 162 structure-function relationships, 135, 136, 139-142
synthetic T and B cell sites and antigens, 233, 234, 236 globular protein antigens, 210, 211, 213, 216
immunological considerations, 200, 201 prediction, 252 viral antigens, 219 T cell repertoire, 162 positive selection in thymus, 163-167 recognition, 167, 168 T cell replacement factors, humoral immune response and, 65, 66, 68 Tetanus toxoid humoral immune response and, 13 sporozoite malaria vaccine and antibodies, 300, 301 human trials, 313-315 synthetic T and B cell sites and, 229, 232, 255, 258
virus-induced immunosuppression and, 355, 357, 361, 362
Thy-1, humoral immune response and, 46,74 Thymic stromal cells, autoreactive T cells and, 421-425 Thymus autoreactive T cells and, 417, 421, 423,433
humoral immune response and, 2, 3 cellular interactions in vivo, 83 helper T cells, 32 interleukins, 64 synthetic T and B cell sites and, 197 T cell receptor and, 115, 162-168 Tissue specificity, regulators of complement activation and, 394-398 Tobacco mosaic virus protein synthetic T and B cell sites and. 227, 244 T cell receptor and, 132 Transferrin T cell receptor and, 124
virus-induced immunosuppression and, 344, 356
Trinitrophenyl-antigen-bindingcells (TNP-ABCs) helper T cells, 27, 33 interleukins, 62-64, 66 physical interaction, 37-40, 42, 44, 47, 50, 51, 53, 54
Trinitrophenyl-memory-antigen-bindingcells (TNP-MABCs), 50, 51 Tubemlosis, virus-induced immunosuppression and, 335, 336, 338, 339 Tumor autoreactive T cells and, 428-430 humoral immune response and, 7, 24 regulators of complement activation and, 395, 397, 402
T cell receptor and, 117, 118 'hmor necrosis factor, regulators of complement activation and, 397
v Vaccine sporozoite malaria, see Sporozoite malaria vaccine synthetic T and B cell sites and, 195, 197, 260-264
bacterial antigens, 231, 232 candidate synthetic peptide, 252-259 globular protein antigens, 213 parasitic antigens, 228, 230 peptides, 239 prediction, 251, 252 viral antigens, 217-219,222, 223,225,226 virus-induced immunosuppression and, 336, 339, 340
Vaccinia virus, regulators of complement activation and, 409 Viral antigens, synthetic T and B cell sites and, 216-228, 240, 241 Virus autoreactive T cells and, 431 humoral immune response and, 24, 78, 82, 83
regulators of complement activation and, 395, 409
synthetic T and B cell sites and, 195. 219, 260, 262
454 T cell receptor and alloreactivity, 156 antigen processing, 119, 125, 127 experimental systems, 131, 137 structure-function relationships, 141 Virus-induced immunosuppression, 335, 336, 368, 369
INDEX
human immunodeficiency virus, 355-363 history of infection, 352-355 immune response, 363-368 virology, 347-352 measles, 338-347 virology, 336-338
Erratum The third paragraph on page 286 slioirld rvatl: Clydeandother investigatorsvaccinatedhumans with sporozoites( 16-2 I). The method utilized was unique (22). that is, the volunteers were subjected to the bite of hundreds of infected mosquitoes which had been y-irradiated. (Sporozoites obtained by dissection of salivary glands of mosquitoes are heavily contaminated with mosquito tissue and cannot be injected into humans. When the radiation-attenuated sporozoites are delivered by mosquito bite, the amount of mosquito-contaminating material is kept to a minimum.) A total of 11 volunteers were vaccinated against P falciparimi. Only after multiple exposures to the infected and irradiatedmosquitoes,over periods of time ranging from 3 to 10 months, were five volunteers protected against challenge with the same or different strains of sporozoites of P. fakiparum. Two out of five volunteers were protected against P. viiwr infection when immunized by the bite off. i3iiVa.r-infected mosquitoes (23).
