Suppressor T cells: some answers but more
questions Martin E. Dorf, Vijay K. Kuchroo and Mary Collins
The concept that suppressor T (T s) cells play an important role in regulating immune responses has been challenged over the past several years. However, it is clear that, although clonal deletion plays a major role in the development of tolerance to self antigens, mechanisms of peripheral tolerance exist. Both suppression and anergy have been postulated to fulfill this role. Although anergic T cells can be identified in the periphery, the anergy model does not account for the findings that adoptive transfer of T cells from mice tolerant to a given antigen reduces the immune response to that same antigen in syngeneic recipients. The best explanation for this result is thatT cells exist that can suppress immune responses. It has been difficult to characterize T s cells and the factors made by T s hybridomas that substitute for T s cells. However, progress has been achieved on several points, including the definition of (1) the requirements for induction of T s cell activity, (2) the relationship between T s cell receptors and the antigen-binding factors released by T s hybridomas (TsFs), (3)the specificity o f T s cell responses, and (4) effector mechanisms. Induction of T s cells Requirements for the induction of T s cells are distinct from those for the induction of helper (TH) cells. Tenfold fewer antigen-presenting cells (APCs) are required for the induction of T s than T H cells. The most efficient APC in inducing T s cells is the macrophage; such macrophages are operationally distinguished from those which induce T H cells by the criteria of ultraviolet (UV) and cyclophosphamide sensitivity I. The mechanisms whereby these treat-
Many properties of suppressor T cells and the antigen-bindingfactors derived from them have evaded molecular genetic definition. Here, Martin Dorf and colleagues discuss recent data in the context of a growing awareness of the and 1-E. A recombinational hotspot was located within the E13 gene but molecules and principles involved. ments affect APC function are unclear but may involve interference with costimulatory molecules or cytokine production 2. Major histocompatibility complex (MHC) class 11 molecule expression on APCs is necessary for T~ cell induction, and gamma-interferon (IFN-y), which induces expression of MHC class II, enhances the capacity of macrophages to generate T s cells. Induction of T H and T s cell responses can be mediated by the same MHC class-II-bearing macrophage cloned cell lines <4. Accumulating evidence suggests that APC-T s cell interactions are primarily restricted by antigen in association with either I-E (in mice) or DQ (in humans) ',6. Macrophages or related cells residing in privileged sites, such as the brain and the eye, selectively induce routine T s cell responses, thereby making these tissues more resistant to immune attack. APCs from these sites are capable of inducing tolerance, which is mediated by T s cell activity i,7. Transforming growth factor 13 (TGF-13) is another important microenvironmental cytokine that confers on APCs the capacity to induce T s cells (Ref. 8). TGF-13 is a normal constituent of several privileged sites, inchlding aqueous humor and cerebrospinal and amniotic fluids. The I-J question The problem that accompanies the study of T-cell-mediated suppression in the murine system is l-J. I-J is defined as the genetic difference between the two inbred strains, B10.A(3R) and B10.A(SR); it was initially mapped to H-2, between I-A
the sequence around this hotspot in B 10.A(3R) and B10.A(SR) indicated that identical recombination events occurred in both strains. Thus, the sequence in this region does not account for differences in 1-J phenotypC. The possibility that additional crossovers, mutations or gene conversion in 3R mice account for I-.1 must be considered. Numerous reports demonstrate that antibodies raised in the 3R-SR combination bind TsFs, react with T s cells and interfere with T s cell bioactivity. Tada and associates > characterized surface molecules with anti-I-J antibodies but the relationship of these receptor-like molecules expressed by most T cells to the I-J determinants involved in suppression is unknown. In addition, the l-J phenotype of the APC controis the genetic restrictions of T s cells and TsFs I. TsFs produced by effector T s cell hybridomas are typically restricted by an 1-.1 allele and do not function in both 3R and 5R mice. On the basis of Ts cell genetic restriction data, it seems likely that therc are two distinct molecules in I-J disparate strains: first a receptorlike molecule expressed by effector T s cells and TsFs, and second, a restricting element expressed on APCs I. It is unclear whether I-J on macrophages is derived from known MHC genes or is encoded bv an unknown genc. The former hypothesis is supported by the observation that I-J polymorphism is influenced by Ec~ alleles in transgenic mice II. The discovery of proteosomc and transporter genes within the MHC illustrates how additional MHC genes serve critical functions in antigen processing and presentation,
© 1992, ElsevierScience Publishers Ltd, UK.
TRENDS and in control of immune responsiveness. Some of these genes are upregulated following IFN-y treatment, as would be predicted of T scell-inducing molecules expressed by APCs. T s cell biology It is important to note that delayed-type hypersensitivity (DTH) and plaque-forming cell (PFC) assays are primarily employed to demonstrate TsF activity. Even in these systems suppression is generally quantitative, resulting in about 50% suppression, although there is greater suppression of high-affinity PFC responses. Attempts to observe TsF-mediated suppression of proliferative responses in vitro have generally failed. The phrase 'immune deviation' better describes the phenomenon, since only selected parameters of the immune response are suppressed 7. One interpretation of these findings is that a limited subset of T-cell functions are inhibited by T s cell activity, perhaps because TH1 cells are more susceptible to suppression and/or anergy. This is consistent with the isolation of T s cell clones from patients with lepromatous leprosy, a disease characterized by poor cellular immune responses and intact humoral immunity u,13. The class of T cells that mediate DTH responses and are suppressed by Ts cells are likely to be important in autoimmunity and cancer. Autoimmune diseases such as experimental autoimmune encephalomyelitis (EAE) or uveitis and selected tumors, especially those induced by agents such as UV light that alter APC function, can be readily modulated by T s cells 14-16. Phenotypic analyses of T s cell populations indicate that both CD4 + and CD8 + subsets are required for suppression. CD4 +, MHC class-II-restricted inducer T s cells are required early in an immune response but usually do not directly mediate suppression. CD8 + cells serve as the effector T s cell population and regulate ongoing immune responses. In several systems an additional, intermediate, T s cell population is observed. These generalities apply to murine Ts cells induced by administration of antigen-modified syngeneic cells and antigen inoculated into privileged sites, such as the
anterior chamber of the eye. Until recently there were no phenotypic markers to distinguish T s from other T-cell subsets; however, candidate monoclonal antibodies for both human and mouse T s cells have recently been described 12,17,J~. Networks of interacting T cells that regulate mouse immune responses have been defined for several model antigens. Expression of complementary anti-idiotypic receptors on some T s cells led to the notion that T s cells are part of a self-regulating idiotypic network. Additional evidence in support of this concept has been obtained from human T s cell systems 19. Limited diversity among T-cell receptor (TCR) 0~ chains, reported in one T s cell system, should facilitate network communications 2°, but additional studies are required to establish whether Ts cells in general use a restricted repertoire. The activity of T s cells is claimed to be antigen specific as responses to irrelevant antigens are not suppressed. However, bystander suppression occurs after antigen-dependent T s cell activation, indicating that the final step in the suppression mechanism is nonspecific 1. Some cloned MHC-restricted T cells also suppress immune responses by antigen-dependent non-specific mechanisms mediated by cytokines such as interleukin 4 (IL-4), TGF-IB and IL-10 (Refs 13, 14, 21, 22). Cytokine-mediated suppression represents only one potential mechanism for regulation of immune responses. Another potential mechanism involves antigen-binding effector TsF molecules that are targeted to APCs and that could interfere with APC function 1. The effector mechanisms of cytokines such as IL-10 also involve APC intermediates 23, suggesting that distinct suppressive mediators may function through common pathways. Several groups have prepared hybridomas representing each of the critical cell components in T s cell responses. T s cell hybridomas are defined by their ability to release antigen-specific TsFs. Conditioned media from T s cell hybridomas substitutes for T s cells in suppression assays. TsFs released by these T s cell hybridomas specifically bind to antigen columns. In addition, those T s
cell hybridomas that express TCR complexes often bind antigens in the absence of MHC molecules, although activation of the cells remains MHC dependent. Direct antigen binding by the TCR is not unique to T s cells: similar observations have been noted with T H cells and cytotoxic T lymphocytes (CTL). This property of MHCindependent antigen binding is peculiar to antigens that express high densities of identical epitopes, such as haptens. T suppressor factors Contrary to initial speculation, numerous studies have demonstrated that T s cells use a[3 TCRs to recognize antigen s,6,22,24-28. Reexamination of the original T s cell hybridomas made with the BW5147 fusion partner indicated that they expressed unique TCR o~ chain rearrangements and expressed the BW5147-derived 13 chain. Complementation by the BW5147 TCR 13 chain resulted in TCR-mediated, non-MHC-restricted antigen recognition 26. Although exceptions have been reported 25, in most systems expression of TCRs on the cell surface is not required for TsF release. Hybridomas that lack TCR expression because they have lost their TCR 13 chain can retain TsF bioactivity26, 27. Accumulating evidence indicates that TCR oLchain is a critical component of TsF. TsFs derived from hybridomas and T cell clones are bound by H28.710 (Refs 22, 2428), an antibody against the TCR oL chain. Variants of T s cell hybridomas that fail to express the TCR (x chain do not make TsF, and transfection of the TCR ~ chain cDNA into these cells restores TsF bioactivity 27,28. In contrast, variants of T s hybridomas that fail to express TCR ~3 chain mRNA produce normal levels of TsFs. The absence of surface CD3-TCR on these variants indicates that Ts F are not a shed form of cell surface TCR. Although the precise molecular definition of TsF is unknown, there is ample evidence that TCR ~xchain is a component of both inducer and effector TsFs. However, inducer and effector TsFs are not functionally equivalent, as they do not substitute for each other in assays. This suggests that at
Vol. 13 No. 7 1992
TRENDS least one of the factors is likely to be composed of something more than the TCR ~ chain, a possibility supported by the fact that only the effector molecules are MHC restricted. In addition, the factors can be partially purified by passage over specific antigen or antibody columns, suggesting that they are not composed of TCR o~chain and a second molecule that is released independently into the media. Monomeric, homodimeric and heterodimeric models of TsF have been proposed. It is possible that a second polypeptide is associated with TCR c~chain; the role of this second component may be either to facilitate release of TCR (x chain from the cell and/or to confer suppressive bioactivity22. Alternatively, it is possible that TsF activity is solely derived from the TCR c~ chain and that differences in processing of TCR c~ chain by the inducer and effector hybridomas result in their distinct phenotypes. Complete biochemical characterization of TsFs is necessary to distinguish among these possible models. The notion that TCR (x chains are critical for TsF activity has implications for network theories of immunoregulation. The TCR chain may serve as a stimulus for anti-idiotypic CTL or T s cells. Indeed, it has been shown that T s cell interactions are dependent on network recognition events which, presumably, are established outside the thymic microenvironment. The implication of these observations is that vaccination with TCR peptides may induce regulatory T s cells. In the few systems where this has been attempted there has been regulation of DTH-like responses that mediate autoimmune disease>. TCR o~chain release may represent the physiologi-
cal mechanism responsible for maintaining homeostasis of the immune system. Disturbing this balance could result in autoimmunity or susceptibility to the development of tumors. The desire to understand T s cells and TsFs is accompanied by the hope of manipulating the immune system, especially in situations of clinical utility. Martin E. Dorf and Vijay K. Kuchroo are at the Department of Pathology, Harvard Medical School, 200 Longwood Avenue, Boston, MA 02115, USA; and Mary Collins is at Genetics Institute, 87 Cambridge Park Drive, Cambridge, MA 02140, USA.
References 1 Ishikura, H., Kuchroo, V., Abromson-Leeman, S. and Dorf, M.E. (1988) lmmunol. Rev. 106, 93-114 2 Simon, J.C., Tigelaar, R.E., Bergstresser, P.R., Edelbaum, D. and Cruz, P.D.J. (1991)J. lmmunol. 146, 485-491 3 Kuchroo, V.K., Minami, M., Diamond, B. and Doff, M.E. (1988) J. lmmunol. 141, 10-16 4 Jayaraman, S., Mensi, N., Webb, D.R. and Doff, M.E. (1991) J. Immunol. 146, 4085-4091 5 Mitchison, N.A. (1989) Prog. Immunol. 7, 845-852 6 Salgame, P., Convit, J. and Bloom, B.R. (1991) Proc. Natl Acad. Sci. USA 88, 2598-2602 7 Wilbanks, G.A., Mammolenti, M. and Streilein, J.W. (1992) Eur. J. Immunol. 22, 165-173 8 Wilbanks, G.A. and Streilein, J.W. Eur. J. Immunol. (in press) 9 Kobori, J.A., Strauss, A.E., Minard, K. and Hood, L. (1986) Science 234, 173-179 10 Nakayama, T., Kubo, R.T., Kishimoto, H., Asano, Y. and Tada, T. (1989) Int. Immunol. 1, 50-58 11 Flood, P.M., Benoist, C., Mathis, D. and Murphy, D.B. (1986) Proc. Natl Acad. Sci. USA 83, 8308-8312
12 Li, S.G., Ottenhoff, T.H.M., Van den Elsen, P. et al. (1990) Eur. I. Immunol. 20, 1281-1288 13 Salgame, P., Abrams, J.S., Clayberger, C. et al. ( 199 l) Science 253,279-282 14 Miller, A., Lider, O., Roberts, A.B., Sporn, M.B. and Weiner, t1.L. (1992) Proc. Natl Acad. Sci. USA 89, 421-425 15 Hara, Y., Caspi, R.R., Wiggert, B. et al. (1992) J. lmmunol. 148, 1685-1692 16 Alcalay, J. and Kripke, M.L. (1991) J. Immunol. 146, 1717-1721 17 Torimoto, Y., Rothstein, D.M., Dang, N.M., Schlossman, S.F. and Morimoto, C. (1992)1. lmmunol. 148, 388-396 18 Devens, B.H., Koontz, A.W., Kapp, J.A., Pierce, C.W. and Webb, D.R. (1991) J. lmmunol. 146, 1394-1401 19 Mohagheghpour, N., Damle, N.K., Takada, S. and Engleman, E.G. (1986) J. Exp. Med. 164, 950-955 20 Koseki, H., Imai, K., lchikawa, T., Hayata, I. and Taniguchi, M. (1989) Int. Immunol. 1,557-564 21 Moore, K.W., Vieira, P., Fiorentino, D.F. et al. (1990) Science 248, 1230-1234 22 lwata, M., Katamura, K., Mori, A. et al. (1990) 1. lmmunol. 145, 3578-3588 23 Malefyt, R.W., Abrams, J., Bennett, B., Figdor, C.G. and de Vries, J.E. (1991)J. Exp. Med. 174, 1209-1220 24 Takata, M., Maiti, ILK.,Kubo, R.T. et al. (1990).I. lmmunol. 145, 2846-2853 25 Fairchild, R.L., Kubo, R.T. and Moorhead, J.W. (1990) J. Immunol. 145, 2001-2009 26 Collins, M., Kuchroo, V.K., Whitters, M.J. et al. (1990) J. lmmunol. 145, 2809-2819 27 Kuchroo, V.K., Byrne, M.C., Atsumi, Y. et al. (1991) Proc. Natl Acad. Sci. USA 88, 8700-8704 28 Green, D.R., Bissonnette, R., Zheng, H. et al. (1991) Proc. Natl Acad. Sci. USA 88, 8475-8479 29 Offner, H., Vainiene, M., Gold, D.P. etal. (1992)]. lmmunol. 148, 1706-1711
The Budapest Bulletin: Congress News Number 6 Workshop on M o l e c u l a r a n d G e n e t i c I m m u n o p r o b e s for Biotechnology Call for papers Participants of the 8th International Congress of Immunology who are interested in this Workshop should contact Professor E.C. de Macario, Wadsworth Center for Laboratories and Research, New York State Department of Health, Empire State Plaza, Albany, NY 12201-0509, USA. Phone and fax (518) 474 1213.
N o 7 x99e