- Email: [email protected]
T cells causing immunological disease: immunopathology or autoimmunity?
42nd. FORUM
310
IN Ih4MUNOLOG
Y
hemopoietic cells in the thymus. J. exp. Med., 172, 1305. Speiser, D.E., Lees, R.K., Hengartner, H., Zinkernagel, R.M. &MacDonald, H.R. (1989), Positive and negative selection of T cell receptor VP domains controlled by distinct cell populations in the thymus. J. exp. Med., 170, 2165. Webb, S., Morris, C. & Sprent, J. (1990), Extrathymic tolerance of mature T cells: clonal elimination as a consequence of immunity. Cell, 63, 1249. Wyllie, A.H., Kerr, J.R.F. & Currie, A.R. (1980), Cell death: the significance of apoptosis. Znt. Rev. Cytol., 68, 251. Wyllie, A.H., Morris, R.G., Smith, A.L. & Dunlop, D. (1984), Chromatin cleavage in apoptosis : association with condensed chromatin morphology and dependence on macromolecular synthesis. J. Path., 142,67.
& Matis, L.A. (1990). In vivo induction of anergy in peripheral V88+ T cells by staphylococcal enterotoxin B. J. exp. Med., 172, 1091. Roberts, J.L., Sharrow, SO. & Singer, A. (1990), Clonal deletion and clonal anergy in the thymus induced by cellular elements with different radiation sensitivities. J. exp. Med., 171, 935. Shortman, K., Vremec, D. & Egerton, M. (1991), The kinetics of T cell antigen receptor expression by subsets of CD4+CD8+ thymocytes: delineation of CD4+8+3++ thymocytes as post-selection intermediates leading to mature T cells. 1. exp. Med., 173,323. Speiser, D.E., Chvatchko, Y., Zinkernagel, R.M. & MacDonald, H.R. (1990), Distinct fates of self specific T cells developing in irradiation bone marrow chimeras: clonal deletion, clonal anergy or in vifro responsiveness to self MIS-la controlled by
T cells causing immunological disease : immunopathology or autoimmunity ? R.M. Departement
Zinkernagel
of Pathology,
University
Evidence is summarized that genetically encoded self-peptides may not be considered immunologically as self when expressed solely extrathymically on non-lymphohaemopoietic cells ; nevertheless, they are antigenic and are recognized by induced effector T cells. An immune response is readily induced against such non-immunological self (as against foreign) by an appropriate presentation of these nonimmunological self peptides on proper antigenpresenting cells. If substantial, such an immune response causes a disease resembling an autoimmune disease which pathogenetically may, however, be called more appropriately an immunopathological T-cell mediated disease. Definition self
of immunological
vs. non-immunological
All antigenic epitopes or peptides coded for by the host genome represent the “genetic self”. Those
(*) For reprint requests.
(*) and H. Hengartner Hospital,
8091 Ziirich
(Switzerland)
genetic self peptides that induce clonal abortion of maturing T cells in the thymus constitute the “immunological self”. Genetic self peptides that are not expressed in the thymus by thymic epithelial cells or do not reach the thymus in soluble form or on migrating lymphohaemopoietic cells (including antigenpresenting cells) do not induce tolerance in the thymus ; similarly, if genetic self peptides are not expressed on lymphohaemopoietic cells including antigen-presenting cells in the periphery, they cannot induce an immune responses (as cannot foreign antigens introduced into the host under this rule). Therefore, these peptides do not exist from the immune system’s point of view and they may be defined as “non-immunological self”. In conclusion, genetic self = immunological self + non-immunological self (Zinkernagel et al., 1991). Self-peptides presented in the thymus render these epitopes tolerogenic in the thymus. T cells specific for such self antigens get clonally eliminated in the thymus and therefore are not present in the periph-
CLONAL
DELETION
AND ANERGY:
ery and cannot be induced. T cells specific for antigens/peptides that are genetically encoded by the host, but that are immunologically not seen do not get triggered ; similarly, T cells specific for foreign peptides cannot be activated by this foreign peptide unless it is presented by antigen-presenting cells in association with class I or II. Thus there is no need for T cells to be unresponsive (neither by suppression nor anergy) to non-immunological self peptides in the periphery because these antigens are only presented in a form that is neither inducing unresponsiveness nor an immune response. Normally, if nonimmunological self antigens/peptides get forced into/onto professional antigen-presenting cells above a certain threshold, a T-cell response may be readily induced. If substantial enough, such a T-cell response against non-immunological self may cause pathology. Whether such T-cell reactivities should better be called immunopathological T-cell responses rather than “autoimmune” T-cell reactions may be regarded as a question of semantics. However, thinking pathogenesis, T-cell autoimmunity might better be reserved for T-cell immune reactivity to immunological self (i.e. antigens/peptides that are immunogenically and tolerogenically processed and presented). Therefore, pathological or disease-causing T-cell reactivity to non-immunological self and to foreign antigens might better be called immunopathological T-cell responses. We do not know of a convincing example of human or animal disease where true T-cell autoimmunity as defined here is of pathogenetic relevance. In contrast, there are many diseases that are caused by T-cell-mediated immunopathologies (Zinkernagel et al., 1991). Experimental
evidence
To study cytotoxic T-cell reactivity and tolerance, we have used transgenic mice expressing a T-cell receptor specific for a viral peptide (amino acids 32-42 of LCMV-GP (lymphocytic choriomeningitis virus/glycoprotein) presented by the class I transplantation antigen Db and other transgenic mice expressing the viral glycoprotein (LCMV-GP including the target peptide of the transgenic T-cell receptor) only in pancreatic islet cells to study cytotoxic T-cell reactivity and tolerance in single and double transgenic mice (Ohashi et al., 1991). The GP transgenic mice do not delete the specific thymocytes and T cells in the thymus, and do not develop diabetes up to the age of one year. However, infection with LCMV induces an antivirally protective cytotoxic T-cell response and the GP mice develop diabetes exactly in parallel with the induced LCMV-specific CTL response. Our results therefore document an example of absence of specific peripheral tolerance to a class-I-MHC-presented extrathymic genetic self antigen. This model of inducing a T-cell-mediated
FROM
MODELS
TO REALITY
311
immunopathology against (genetic but) nonimmunological self therefore raises the general question of whether pathogenic T-cell responses against truely immunological self exist at all. As widely documented, clonal abortion in the thymus is expected to render this possibility highly unlikely, particularly since tolerance is more sensitive than antigen stimulation (Pircher et al., 1991). The examples presented here and possibly other examples of T-cell epitope mimicry (where potentially trivial infectious agents express epitopes that are similar to self (Fujinami et al., 1983; Oldstone et al., 1986) operationally/mechanistically cannot be distinguished from T-cell-mediated immunopathology against foreign antigenic epitopes (Bianchi, 1981; Doherty and Zinkernagel, 1974; Hotchin, 1962; Mims, 1982; Mondelli and Eddleston, 1984; Stott and Taylor, 1985). Tolerance and responsiveness of B cells to a biologically relevant antigen, the neutralizing antibody response to vesicular stomatitis virus (VSV), was studied in mice expressing VSV-Indiana glycoprotein as a cell-internally-synthesized protein (Zinkernagel et al., 1990, 1991). These mice immunized with purified VSV-Indiana glycoprotein mixed with a strong adjuvant failed to produce IgG antibodies with VSVIND neutralizing activity. They also failed to respond to infection with a vaccinia recombinant virus encoding the VSV-Indiana glycoprotein, because the VSV glycoprotein is not accessible for antibodies on the infecting recombinant vaccinia virus (Zinkernagel et al., 1990). But after infection with VSV-Indiana wildtype virus, VSV transgenic mice promptly mounted an IgG autoantibody response. The response to the membrane-associated viral antigen demonstrates that reactive B cells have not been eliminated in these GP transgenic mice. They can be induced to produce not only IgM, but also IgG if direct, contact-mediated T help to a linked new helper determinant by other VSV proteins is provided. Thus, unresponsiveness of B cells is maintained by absence of T help, a concept that has been previously documented as “split tolerance” (reviewed in Kake and Mitchison, 1977 ; Weigle, 1973). Obviously, these experiments and their interpretations are not all readily compatible with those using allo-H-2 as test antigens (Goodnow et al., 1990; Moller, 1991; Nemazee and Burki, 1989; Zinkernagel et al., 1991). The high density of H-2 may be an experimental situation not usually applying to nonMHC antigens. Both sets of experiments may serve to support the notion that T cells are clonally eliminated in the thymus by exposure to antigen presented locally or by immigrating antigen-transporting and -presenting cells (these antigens define immunological self). Peripheral T cells bearing a TCR receptor reactive to non-immunological self become effector cells if
42nd FORUM
312
properly induced by professional antigen presenting cells in secondary lymphoid organs (LCMV-reactive CTL in transgenic LCMV-GP expressing pancreatic islet cells). Alternatively, T helper cells against foreign viral antigens may induce true IgG autoantibody (i.e. against VSV transgenic glycoprotein) if the two antigens are properly linked (Lake and Mitchison, 1977), as is the case in the wild type virus. Three illustrations
of the concept
Three examples may illustrate the concept. Several viruses avoid inducing an immune response early after infection by replicating exclusively in epithelial or mesenchymal cells. Because viral antigens are not presented properly to immune cells under these conditions, no T-cell induction occurs. Only after viral antigens are released by cells as a consequence of cytopathogenicity of the viral infection, viral antigens will be taken up by antigen-presenting cells which are capable of efficiently inducing T cells. Rabies virus staying initially in neurons or papilloma viruses keeping to keratinocytes of an advanced differentiation stage and therefore staying away from Langerhans cells illustrate these tactics well (Field, 1985). Similarly, carcinomas or sarcomas may well express tumour-associated antigens but cannot induce a protective immune response, because antigen presentation is not on professional antigen-presenting cells (Prehn, 1990; Zinkernagel, 1982). Once tumour cells decay because of insufficient vascularization, antigen may be processed and presented immunogenitally to induce an immune response, at a time point when the tumour cell burden may already be too high. Third, studies using purely epithelial tissue and organ grafts have revealed that such grafts can be transplanted and are accepted across major transplantation antigen barriers. If the host is subsequently exposed to lymphohaemopoietic cells of the same donor, rejection is promptly induced (reviewed in Lafferty et al., 1983) proving that, for transplantation antigens as well, the same rules apply. Induction
IN IMMUNOLOGY
forced otherwise into appropriate antigen-presenting cells, such non-immunological self antigens or “cross-reactive” or “self-mimicking” foreign antigenie fragments may induce an immune response. This may happen experimentally by mixing such nonimmunological self antigen-fragments with complete Freund’s adjuvant (for example in allergic encephalomyeolitis (EAE) (Schluesener and Wekerle, 1985 ; Wekerle et al., 1989) or by chronic inflammatory processes involving destruction of tissues expressing this non-immunological self antigen, leading to repeated chronic immunogenic presentation by professional antigen-presenting cells in these granulomatous lesions. The quantitative and kinetic requirements are not clear as yet but it is reasonable to assume that they follow the same general rules as T-cell activation and reactivity to foreign antigens ; unfortunately, even these rules are only poorly understood. An additional possibility may be that intracellular parasites including viruses may alter normal processing of immunological self on antigenpresenting cells in the periphery by producing new determinants. Such a possibility has been documented in vitro (Gammon and Sercarz, 1990) but may well be possible in vivo. An implication of the proposal may be that T-celldependent immunopathologies to non-immunological self (or foreign) are amenable to preventive measures similar to those against foreign infectious agents or antigen. Accordingly, vaccination with low quantities of the antigen involved may be capable of favorably shifting the overall balance between immunoprotection and immunopathology; this possibility has been documented in experimental allergic encephalomyelitis (EAE) (reviewed in Weigle, 1973 ; Zamvil and Steinman, 1990) and is now being tested in our models. Acknowledgements Supported by Swiss National Sciencefoundation Grants and the Kanton Ziirich.
of immunopathology
Induction of T cells reactive against genetic self that normally is immunologically not considered as self because it is not presented appropriately to immune cells may cause T-cell-mediated immunopathology. How could such T cells be induced? One may extend the old concept of immunologically privileged sites to non-immunological self antigens, or fragments thereof. In contrast to older versions, we are not referring to anatomic barriers, but rather to the absence of immunologically inducible antigen presentation, i.e. non-immunological self peptides normally cannot induce a T-cell response. If introduced either as part of an infectious agent or when
References Bianchi, L. (1981), The immunopathology of acute type B hepatitis. Springer Semin. Immunopath., 3, 421438.
Doherty, P.C. & Zinkernagel. R.M. (1974), T cell-mediated immunopathology in viral infections. Transplant. Rev., 19, 89-120. Fields, B.N. (1985), Virology. Raven Press, New York. Fujinami, R.S., Oldstone, M.B.A., Wroblewska, Z., Frankel, M.E. & Koprowski, H. (1983), Molecular mimicry in virus infection : cross-reaction of measles phosphoprotein or of herpes simplex virus protein with human intermediate filaments. Proc. nat. Acad. Sci. (Wash.), 80, 2346-2350.
CLONAL
DELETION
AND ANERGY:
Gammon, G. & Sercarz, E. (1990), Does the presence of self-reactive T cells indicate the breakdown of tolerance? Clin. Immunol. Immunopafh., 56, 287-291. Goodnow, CC., Adelstein, S. & Basten,A. (1990), The needfor centraland peripheraltolerancein the B cell repertoire. Science, 248, 1373-1379. Hotchin, J. (1962), The biology of lymphocytic choriomeningitisinfection : virus inducedimmunedisease. Cold Spr. Harb.
Symp. quant. Biol.,
21, 419-499.
Lafferty, K.J., Prowse, S.J. & Simeonovic,C.J. (1983). lmmunobiologyof tissuetransplantation: a return to the passengerleukocyte concept. Ann. Rev. Immunol., 1, 143-173. Lake, P. & Mitchison, N.A. (1977), Regulatory mechanismsin the immuneresponseto cell-surfaceantigens. Cold Spr. Harb.
Symp. quant. Biol., 41, X39-595.
Mims, C.A. (1982),Pathogenesis of infectiousdisease. Ed. 2 (pp. 10-160).Academic Press,London, New York. Mondelli, M. & Eddleston,A.L.W.F. (1984),Mechanisms of liver cell injury in acute and chronic hepatitis B. Semin. Liver Dis., 4, 47-58.
Moller, G. (1991), Transgenicmice and immunological tolerance. Immunol. Rev., 122, 133-171. Nemazee,D.A. & Biirki, K. (1989),Clonal deletion of B lymphocytesin a transgenicmousebearinganti-MHC classI antibody genes.Nature (Land.), 337,562-566. Ohashi,P.S., Oehen,S., Biirki, K., Pircher, H.P., Ohashi, C.T., Odermatt, B., Malissen,B., Zinkernagel, R. & Hengartner, H. (1991),Ablation of “tolerance” and induction of diabetesby virus infection in viral antigen transgenicmice. Cell, 65, 305-317. Oldstone, M.B.A., Schwimmbeck, P., Dyrberg, T. & Fujinami, R. (1986), Mimicry by virus of host molecules: implications for autoimmune disease. Progr.
Immunol.,
6, 787-195.
Pircher, H.P., Hoffmann Rohrer, U., Moskophidis, D.,
Post-thymic
FROM
MODELS
TO REALITY
Zinkernagel, R.M. & Hengartner, H. (1991),Lower receptor avidity required for thymic clonal deletion than for effector T cell function. Nature (Lond.), 351, 482-485. Prehn, R.T. (1990),Immunologicaltoleranceto many self epitopesmay be unnecessary.Stand. J. Immunol., 32, 293-296. Schluesener,H.J. & Wekerle, H. (1985), Autoaggressive T lymphocyte linesrecognizingthe encephalitogenic regionof myelin basicprotein : in vitro selectionfrom unprimed rat T lymphocyte populations. J. Immunol., 135, 3128. Stott, E.J. & Taylor, G. (1985), Respiratory syncytial virus. Arch. Virol., 84, l-8. Weigle, W.O. (1973), Immunological unresponsiveness. Advanc. Immunol., 16, 61-122. Wekerle, H., Pette, M., Fujita, K., Nomura, K. & Meyermann, R. (1989),Autoimmunity in the nervoussystem: functional properties of an encephalitogenic protein. Progr. Immunol., 7, 813-820. Zamvil, S.S. & Steinman,L. (1990),The T lymphocyte in experimentalallergic encephalomyelitis.Ann. Rev. Immunol., 8, 579-621. Zinkemagel,R.M. (1982),How to escapeimmunesurveillance? Springer Semin. Immunopath., 5, 107-I12. Zinkernagel, R.M., Cooper, S., Chambers,J., Lazzarini, R.A., Hengartner, H. 8cArnheiter, H. (1990),Virus induced autoantibody responseto a transgenicviral antigen. Nature (Lond.), 344, 68-71. Zinkernagel, R.M., Pircher, H.P., Ohashi, P.S., Oehen, S., Odermatt, B., Mak, T.W., Arnheiter, H., Burki, K. & Hengartner, H. (1991),T and B cell tolerance andresponses to viral antigensin transgenicmice: implications for the pathogenesisof autoimmuneversusimmunopathological disease. Immunol. Rev., 122, 133-171.
T-cell tolerance or indifference that is the question J.F.A.P.
The Walter and Eliza Hall Institute
Miller
and W.R.
:
Heath
of Medical Research, Post Office Royal Melbourne Victoria 3050 (Australia)
It is likely that several mechanisms operate to neu-
313
Hospital,
tralize autoreactive cells in a system as vitally impor-
mechanisms should operate to hold in check any selfreactive T lymphocytes that may sneak through
tant as that governing self-tolerance. The most efficient way to ensure this is the deletion of clones of self-reactive cells. In the case of T cells, it has only recently been experimentally demonstrated to occur intra-thymically. It seems logical that other fail-safe
extra-thymic autoantigens. The entire T-cell repertoire must be generated intra-thymically since there appears to be no postrearrangement diversification of the TCR genes (Fink
thymic censorship or that may be reactive to unique
310
IN Ih4MUNOLOG
Y
hemopoietic cells in the thymus. J. exp. Med., 172, 1305. Speiser, D.E., Lees, R.K., Hengartner, H., Zinkernagel, R.M. &MacDonald, H.R. (1989), Positive and negative selection of T cell receptor VP domains controlled by distinct cell populations in the thymus. J. exp. Med., 170, 2165. Webb, S., Morris, C. & Sprent, J. (1990), Extrathymic tolerance of mature T cells: clonal elimination as a consequence of immunity. Cell, 63, 1249. Wyllie, A.H., Kerr, J.R.F. & Currie, A.R. (1980), Cell death: the significance of apoptosis. Znt. Rev. Cytol., 68, 251. Wyllie, A.H., Morris, R.G., Smith, A.L. & Dunlop, D. (1984), Chromatin cleavage in apoptosis : association with condensed chromatin morphology and dependence on macromolecular synthesis. J. Path., 142,67.
& Matis, L.A. (1990). In vivo induction of anergy in peripheral V88+ T cells by staphylococcal enterotoxin B. J. exp. Med., 172, 1091. Roberts, J.L., Sharrow, SO. & Singer, A. (1990), Clonal deletion and clonal anergy in the thymus induced by cellular elements with different radiation sensitivities. J. exp. Med., 171, 935. Shortman, K., Vremec, D. & Egerton, M. (1991), The kinetics of T cell antigen receptor expression by subsets of CD4+CD8+ thymocytes: delineation of CD4+8+3++ thymocytes as post-selection intermediates leading to mature T cells. 1. exp. Med., 173,323. Speiser, D.E., Chvatchko, Y., Zinkernagel, R.M. & MacDonald, H.R. (1990), Distinct fates of self specific T cells developing in irradiation bone marrow chimeras: clonal deletion, clonal anergy or in vifro responsiveness to self MIS-la controlled by
T cells causing immunological disease : immunopathology or autoimmunity ? R.M. Departement
Zinkernagel
of Pathology,
University
Evidence is summarized that genetically encoded self-peptides may not be considered immunologically as self when expressed solely extrathymically on non-lymphohaemopoietic cells ; nevertheless, they are antigenic and are recognized by induced effector T cells. An immune response is readily induced against such non-immunological self (as against foreign) by an appropriate presentation of these nonimmunological self peptides on proper antigenpresenting cells. If substantial, such an immune response causes a disease resembling an autoimmune disease which pathogenetically may, however, be called more appropriately an immunopathological T-cell mediated disease. Definition self
of immunological
vs. non-immunological
All antigenic epitopes or peptides coded for by the host genome represent the “genetic self”. Those
(*) For reprint requests.
(*) and H. Hengartner Hospital,
8091 Ziirich
(Switzerland)
genetic self peptides that induce clonal abortion of maturing T cells in the thymus constitute the “immunological self”. Genetic self peptides that are not expressed in the thymus by thymic epithelial cells or do not reach the thymus in soluble form or on migrating lymphohaemopoietic cells (including antigenpresenting cells) do not induce tolerance in the thymus ; similarly, if genetic self peptides are not expressed on lymphohaemopoietic cells including antigen-presenting cells in the periphery, they cannot induce an immune responses (as cannot foreign antigens introduced into the host under this rule). Therefore, these peptides do not exist from the immune system’s point of view and they may be defined as “non-immunological self”. In conclusion, genetic self = immunological self + non-immunological self (Zinkernagel et al., 1991). Self-peptides presented in the thymus render these epitopes tolerogenic in the thymus. T cells specific for such self antigens get clonally eliminated in the thymus and therefore are not present in the periph-
CLONAL
DELETION
AND ANERGY:
ery and cannot be induced. T cells specific for antigens/peptides that are genetically encoded by the host, but that are immunologically not seen do not get triggered ; similarly, T cells specific for foreign peptides cannot be activated by this foreign peptide unless it is presented by antigen-presenting cells in association with class I or II. Thus there is no need for T cells to be unresponsive (neither by suppression nor anergy) to non-immunological self peptides in the periphery because these antigens are only presented in a form that is neither inducing unresponsiveness nor an immune response. Normally, if nonimmunological self antigens/peptides get forced into/onto professional antigen-presenting cells above a certain threshold, a T-cell response may be readily induced. If substantial enough, such a T-cell response against non-immunological self may cause pathology. Whether such T-cell reactivities should better be called immunopathological T-cell responses rather than “autoimmune” T-cell reactions may be regarded as a question of semantics. However, thinking pathogenesis, T-cell autoimmunity might better be reserved for T-cell immune reactivity to immunological self (i.e. antigens/peptides that are immunogenically and tolerogenically processed and presented). Therefore, pathological or disease-causing T-cell reactivity to non-immunological self and to foreign antigens might better be called immunopathological T-cell responses. We do not know of a convincing example of human or animal disease where true T-cell autoimmunity as defined here is of pathogenetic relevance. In contrast, there are many diseases that are caused by T-cell-mediated immunopathologies (Zinkernagel et al., 1991). Experimental
evidence
To study cytotoxic T-cell reactivity and tolerance, we have used transgenic mice expressing a T-cell receptor specific for a viral peptide (amino acids 32-42 of LCMV-GP (lymphocytic choriomeningitis virus/glycoprotein) presented by the class I transplantation antigen Db and other transgenic mice expressing the viral glycoprotein (LCMV-GP including the target peptide of the transgenic T-cell receptor) only in pancreatic islet cells to study cytotoxic T-cell reactivity and tolerance in single and double transgenic mice (Ohashi et al., 1991). The GP transgenic mice do not delete the specific thymocytes and T cells in the thymus, and do not develop diabetes up to the age of one year. However, infection with LCMV induces an antivirally protective cytotoxic T-cell response and the GP mice develop diabetes exactly in parallel with the induced LCMV-specific CTL response. Our results therefore document an example of absence of specific peripheral tolerance to a class-I-MHC-presented extrathymic genetic self antigen. This model of inducing a T-cell-mediated
FROM
MODELS
TO REALITY
311
immunopathology against (genetic but) nonimmunological self therefore raises the general question of whether pathogenic T-cell responses against truely immunological self exist at all. As widely documented, clonal abortion in the thymus is expected to render this possibility highly unlikely, particularly since tolerance is more sensitive than antigen stimulation (Pircher et al., 1991). The examples presented here and possibly other examples of T-cell epitope mimicry (where potentially trivial infectious agents express epitopes that are similar to self (Fujinami et al., 1983; Oldstone et al., 1986) operationally/mechanistically cannot be distinguished from T-cell-mediated immunopathology against foreign antigenic epitopes (Bianchi, 1981; Doherty and Zinkernagel, 1974; Hotchin, 1962; Mims, 1982; Mondelli and Eddleston, 1984; Stott and Taylor, 1985). Tolerance and responsiveness of B cells to a biologically relevant antigen, the neutralizing antibody response to vesicular stomatitis virus (VSV), was studied in mice expressing VSV-Indiana glycoprotein as a cell-internally-synthesized protein (Zinkernagel et al., 1990, 1991). These mice immunized with purified VSV-Indiana glycoprotein mixed with a strong adjuvant failed to produce IgG antibodies with VSVIND neutralizing activity. They also failed to respond to infection with a vaccinia recombinant virus encoding the VSV-Indiana glycoprotein, because the VSV glycoprotein is not accessible for antibodies on the infecting recombinant vaccinia virus (Zinkernagel et al., 1990). But after infection with VSV-Indiana wildtype virus, VSV transgenic mice promptly mounted an IgG autoantibody response. The response to the membrane-associated viral antigen demonstrates that reactive B cells have not been eliminated in these GP transgenic mice. They can be induced to produce not only IgM, but also IgG if direct, contact-mediated T help to a linked new helper determinant by other VSV proteins is provided. Thus, unresponsiveness of B cells is maintained by absence of T help, a concept that has been previously documented as “split tolerance” (reviewed in Kake and Mitchison, 1977 ; Weigle, 1973). Obviously, these experiments and their interpretations are not all readily compatible with those using allo-H-2 as test antigens (Goodnow et al., 1990; Moller, 1991; Nemazee and Burki, 1989; Zinkernagel et al., 1991). The high density of H-2 may be an experimental situation not usually applying to nonMHC antigens. Both sets of experiments may serve to support the notion that T cells are clonally eliminated in the thymus by exposure to antigen presented locally or by immigrating antigen-transporting and -presenting cells (these antigens define immunological self). Peripheral T cells bearing a TCR receptor reactive to non-immunological self become effector cells if
42nd FORUM
312
properly induced by professional antigen presenting cells in secondary lymphoid organs (LCMV-reactive CTL in transgenic LCMV-GP expressing pancreatic islet cells). Alternatively, T helper cells against foreign viral antigens may induce true IgG autoantibody (i.e. against VSV transgenic glycoprotein) if the two antigens are properly linked (Lake and Mitchison, 1977), as is the case in the wild type virus. Three illustrations
of the concept
Three examples may illustrate the concept. Several viruses avoid inducing an immune response early after infection by replicating exclusively in epithelial or mesenchymal cells. Because viral antigens are not presented properly to immune cells under these conditions, no T-cell induction occurs. Only after viral antigens are released by cells as a consequence of cytopathogenicity of the viral infection, viral antigens will be taken up by antigen-presenting cells which are capable of efficiently inducing T cells. Rabies virus staying initially in neurons or papilloma viruses keeping to keratinocytes of an advanced differentiation stage and therefore staying away from Langerhans cells illustrate these tactics well (Field, 1985). Similarly, carcinomas or sarcomas may well express tumour-associated antigens but cannot induce a protective immune response, because antigen presentation is not on professional antigen-presenting cells (Prehn, 1990; Zinkernagel, 1982). Once tumour cells decay because of insufficient vascularization, antigen may be processed and presented immunogenitally to induce an immune response, at a time point when the tumour cell burden may already be too high. Third, studies using purely epithelial tissue and organ grafts have revealed that such grafts can be transplanted and are accepted across major transplantation antigen barriers. If the host is subsequently exposed to lymphohaemopoietic cells of the same donor, rejection is promptly induced (reviewed in Lafferty et al., 1983) proving that, for transplantation antigens as well, the same rules apply. Induction
IN IMMUNOLOGY
forced otherwise into appropriate antigen-presenting cells, such non-immunological self antigens or “cross-reactive” or “self-mimicking” foreign antigenie fragments may induce an immune response. This may happen experimentally by mixing such nonimmunological self antigen-fragments with complete Freund’s adjuvant (for example in allergic encephalomyeolitis (EAE) (Schluesener and Wekerle, 1985 ; Wekerle et al., 1989) or by chronic inflammatory processes involving destruction of tissues expressing this non-immunological self antigen, leading to repeated chronic immunogenic presentation by professional antigen-presenting cells in these granulomatous lesions. The quantitative and kinetic requirements are not clear as yet but it is reasonable to assume that they follow the same general rules as T-cell activation and reactivity to foreign antigens ; unfortunately, even these rules are only poorly understood. An additional possibility may be that intracellular parasites including viruses may alter normal processing of immunological self on antigenpresenting cells in the periphery by producing new determinants. Such a possibility has been documented in vitro (Gammon and Sercarz, 1990) but may well be possible in vivo. An implication of the proposal may be that T-celldependent immunopathologies to non-immunological self (or foreign) are amenable to preventive measures similar to those against foreign infectious agents or antigen. Accordingly, vaccination with low quantities of the antigen involved may be capable of favorably shifting the overall balance between immunoprotection and immunopathology; this possibility has been documented in experimental allergic encephalomyelitis (EAE) (reviewed in Weigle, 1973 ; Zamvil and Steinman, 1990) and is now being tested in our models. Acknowledgements Supported by Swiss National Sciencefoundation Grants and the Kanton Ziirich.
of immunopathology
Induction of T cells reactive against genetic self that normally is immunologically not considered as self because it is not presented appropriately to immune cells may cause T-cell-mediated immunopathology. How could such T cells be induced? One may extend the old concept of immunologically privileged sites to non-immunological self antigens, or fragments thereof. In contrast to older versions, we are not referring to anatomic barriers, but rather to the absence of immunologically inducible antigen presentation, i.e. non-immunological self peptides normally cannot induce a T-cell response. If introduced either as part of an infectious agent or when
References Bianchi, L. (1981), The immunopathology of acute type B hepatitis. Springer Semin. Immunopath., 3, 421438.
Doherty, P.C. & Zinkernagel. R.M. (1974), T cell-mediated immunopathology in viral infections. Transplant. Rev., 19, 89-120. Fields, B.N. (1985), Virology. Raven Press, New York. Fujinami, R.S., Oldstone, M.B.A., Wroblewska, Z., Frankel, M.E. & Koprowski, H. (1983), Molecular mimicry in virus infection : cross-reaction of measles phosphoprotein or of herpes simplex virus protein with human intermediate filaments. Proc. nat. Acad. Sci. (Wash.), 80, 2346-2350.
CLONAL
DELETION
AND ANERGY:
Gammon, G. & Sercarz, E. (1990), Does the presence of self-reactive T cells indicate the breakdown of tolerance? Clin. Immunol. Immunopafh., 56, 287-291. Goodnow, CC., Adelstein, S. & Basten,A. (1990), The needfor centraland peripheraltolerancein the B cell repertoire. Science, 248, 1373-1379. Hotchin, J. (1962), The biology of lymphocytic choriomeningitisinfection : virus inducedimmunedisease. Cold Spr. Harb.
Symp. quant. Biol.,
21, 419-499.
Lafferty, K.J., Prowse, S.J. & Simeonovic,C.J. (1983). lmmunobiologyof tissuetransplantation: a return to the passengerleukocyte concept. Ann. Rev. Immunol., 1, 143-173. Lake, P. & Mitchison, N.A. (1977), Regulatory mechanismsin the immuneresponseto cell-surfaceantigens. Cold Spr. Harb.
Symp. quant. Biol., 41, X39-595.
Mims, C.A. (1982),Pathogenesis of infectiousdisease. Ed. 2 (pp. 10-160).Academic Press,London, New York. Mondelli, M. & Eddleston,A.L.W.F. (1984),Mechanisms of liver cell injury in acute and chronic hepatitis B. Semin. Liver Dis., 4, 47-58.
Moller, G. (1991), Transgenicmice and immunological tolerance. Immunol. Rev., 122, 133-171. Nemazee,D.A. & Biirki, K. (1989),Clonal deletion of B lymphocytesin a transgenicmousebearinganti-MHC classI antibody genes.Nature (Land.), 337,562-566. Ohashi,P.S., Oehen,S., Biirki, K., Pircher, H.P., Ohashi, C.T., Odermatt, B., Malissen,B., Zinkernagel, R. & Hengartner, H. (1991),Ablation of “tolerance” and induction of diabetesby virus infection in viral antigen transgenicmice. Cell, 65, 305-317. Oldstone, M.B.A., Schwimmbeck, P., Dyrberg, T. & Fujinami, R. (1986), Mimicry by virus of host molecules: implications for autoimmune disease. Progr.
Immunol.,
6, 787-195.
Pircher, H.P., Hoffmann Rohrer, U., Moskophidis, D.,
Post-thymic
FROM
MODELS
TO REALITY
Zinkernagel, R.M. & Hengartner, H. (1991),Lower receptor avidity required for thymic clonal deletion than for effector T cell function. Nature (Lond.), 351, 482-485. Prehn, R.T. (1990),Immunologicaltoleranceto many self epitopesmay be unnecessary.Stand. J. Immunol., 32, 293-296. Schluesener,H.J. & Wekerle, H. (1985), Autoaggressive T lymphocyte linesrecognizingthe encephalitogenic regionof myelin basicprotein : in vitro selectionfrom unprimed rat T lymphocyte populations. J. Immunol., 135, 3128. Stott, E.J. & Taylor, G. (1985), Respiratory syncytial virus. Arch. Virol., 84, l-8. Weigle, W.O. (1973), Immunological unresponsiveness. Advanc. Immunol., 16, 61-122. Wekerle, H., Pette, M., Fujita, K., Nomura, K. & Meyermann, R. (1989),Autoimmunity in the nervoussystem: functional properties of an encephalitogenic protein. Progr. Immunol., 7, 813-820. Zamvil, S.S. & Steinman,L. (1990),The T lymphocyte in experimentalallergic encephalomyelitis.Ann. Rev. Immunol., 8, 579-621. Zinkemagel,R.M. (1982),How to escapeimmunesurveillance? Springer Semin. Immunopath., 5, 107-I12. Zinkernagel, R.M., Cooper, S., Chambers,J., Lazzarini, R.A., Hengartner, H. 8cArnheiter, H. (1990),Virus induced autoantibody responseto a transgenicviral antigen. Nature (Lond.), 344, 68-71. Zinkernagel, R.M., Pircher, H.P., Ohashi, P.S., Oehen, S., Odermatt, B., Mak, T.W., Arnheiter, H., Burki, K. & Hengartner, H. (1991),T and B cell tolerance andresponses to viral antigensin transgenicmice: implications for the pathogenesisof autoimmuneversusimmunopathological disease. Immunol. Rev., 122, 133-171.
T-cell tolerance or indifference that is the question J.F.A.P.
The Walter and Eliza Hall Institute
Miller
and W.R.
:
Heath
of Medical Research, Post Office Royal Melbourne Victoria 3050 (Australia)
It is likely that several mechanisms operate to neu-
313
Hospital,
tralize autoreactive cells in a system as vitally impor-
mechanisms should operate to hold in check any selfreactive T lymphocytes that may sneak through
tant as that governing self-tolerance. The most efficient way to ensure this is the deletion of clones of self-reactive cells. In the case of T cells, it has only recently been experimentally demonstrated to occur intra-thymically. It seems logical that other fail-safe
extra-thymic autoantigens. The entire T-cell repertoire must be generated intra-thymically since there appears to be no postrearrangement diversification of the TCR genes (Fink
thymic censorship or that may be reactive to unique