Annals of the New York Academy of Sciences 958:3-6 (2002)
ROLF M. ZINKERNAGEL
Institute for Experimental Immunology, University Hospital, CH-8091 Zurich, Switzerland
This review summarizes findings that indicate that immune reactions of T and B cells depend on thresholds of the binding avidity of the receptor, on antigen amounts, on the time period during which antigen is available in secondary lymphatic tissues, as well as on the detection methods used. Usually immunologically ignored, strictly extralymphatic host cells or their antigens may be released by infection or immunopathology to reach lymphatic organs and induce autoimmune diseases. If we know the infection, we call the disease immunopathologically mediated; if we do not recognize or know it, we call the disease autoimmune.
The essential role of the immune system is to protect hosts against acutely pathogenic infections that interfere with reproduction. Specific immune protection is based on important natural or innate resistance mechanisms (e.g., interferon) without which host survival is impossible. Since immunity engages inflammation and causes host cell destruction, immune protection must be balanced coevolutionarily with potentially lethal damage by immune responses—that is, immunopathology, particularly against poorly or noncytopathic infections. Also, the immune system should control infections but should not react against the host's own antigens. Whether the immune system reacts or does not react depends on basic parameters of the system including relative precursor frequencies of T and B cells and the sensitivity of methods used to detect reactivity.
Rules of T Cell Responses
T cells do not react against self-antigens that are present at some level in blood and lymphoid organs including thymus, spleen, lymph nodes, and bone marrow because these T cells have been eliminated by so-called negative selection or deletion.1 Similarly, noncytopathic infections that are transmitted from mother to offspring during pregnancy or at birth delete T cells, as do overwhelming noncytopathic infections or lymphatic tumours in adult immunocompetent hosts.2-4
At the other extreme, those antigens that do not reach organized lymphatic tissue, independent of whether they are self-antigens or foreign, fail to induce an immune response.5,6 Self-antigens exclusively expressed in islet cells or infectious antigens expressed exclusively extralymphatically such as papilloma virus infections of maturing keratinocytes in the skin are immunologically ignored. However, potentially, they can induce immune responses if these antigens reach secondary lymphoid organs. This may result in autoimmune disease or in rejection of infected peripheral cells, respectively.
An efficient immune response is induced if antigen reaches organized secondary lymphoid tissues in a transient fashion for at least 3 to 5 days in sufficient amounts1,6 (see Table 1).
TABLE 1. Immune response against foreign antigens and peripheral self-antigens1,4
Ignored Peripheral Self-Antigens As Targets Of Autoimmune Responses
Cytotoxic T cell reactivities and tolerance were studied in transgenic mice expressing the glycoprotein LCMV GP under the rat insular promoter (RIP) or an artificial self-antigen only in ß islet cells of the pancreas (see Table 2).5 Such RIP GP transgenic mice did not delete antigen-specific T cells and did not develop diabetes up to the age of one year. However, infection with LCMV induced a protective CTL response causing diabetes in RIP GP mice.7 Interestingly, if the same mice were infected with a vaccinia recombinant virus expressing the LCMV GP, the RIP-GP mice were primed for a secondary in vitro CTL response but did not develop diabetes. A semi-quantitative comparison of the CTL responses after LCMV infection versus vaccinia GP infections revealed that the response was about 100 to 1000 times greater on days 7 to 15 in LCMV-infected compared to vaccinia-GP-infected mice. This model situation illustrates that T cell responses against strictly peripheral self-antigens can be induced if sufficient levels of antigen presented on class I MHC antigens reaches draining lymph nodes or spleen. This evidence indicates that quantitative and time thresholds are key to avoiding autoimmunity and immunopathology against self-antigens. These high thresholds preventing autoimmunity must also be viewed within the context of immune surveillance against persisting infections that, despite immunity against infections, are not eliminated completely.4,8,9 They set the stage for maintaining T cell memory, and this fact explains why immune surveillance against peripheral solid tumors usually fails.10
TABLE 2. Induction of diabetes in REP-GPLCMV-transgenic mice5,7
Role of Organized Lymphoid Tissue in Peripheral Target Organs of Autoimmunity
It is an interesting characteristic of many autoimmune diseases including Hashimoto's thyroiditis, Sjögren's disease, and chronic rhumatoid arthritis and of diabetes that lymphoid follicle-like structures eventually form in the target organs of autoimmunity. RIP-GP transgenic mice do not develop diabetes spontaneously, but with an LCMV infection they do. Surprisingly, these mice also become diabetic when repetitively immunized with dendritic cells expressing the LCMV glycoprotein peptide.11 Interestingly, this form of diabetes developed more slowly within 3 to 4 weeks and was accompanied by neoformation of lymphoid structures in and near the islets. This illustrates that neoformation of lymphatic tissues in a target organ may help to maintain a peripheral immune response locally because antigen within the target organ is not going to be eliminated unless the target cells and the target organ are destroyed, as exemplified in juvenile autoimmune diabetes. Thus, physiologically foreign antigen usually has to reach the draining lymph node or spleen in a dose- and time-dependent fashion. The response stops only once the antigen has been eliminated or controlled. In autoimmune disease lymphoid tissue is exported into the peripheral, so-far-ignored antigen source and there maintains a chronic immunopathological autoimmune response.4,11
Analysis of antiviral immunity provides good insights into the biological parameters of immunological tolerance and specificity. The balance between the protective and immunopathological aspects of immunity are vastly on the beneficial side for the host. The costs of some ill-balanced situations leading to immunopathology and autoimmunity are probably the costs of a coevolutionary balance between infectious agents and host immunity.
Of course, if we know the infectious agent and the immune response directed against peripheral infections such as HBV or lepra bacilli, in liver or Schwann's sheaths, respectively, we call the unfavorable balance between host immune response and infection an immunopathology. If we do not know the infectious agent, then we tend to call the ensuing immunopathology an autoimmune disease.12 While this may be an overstatement, it predicts that perhaps many so-called autoimmune diseases will eventually turn out to represent immunopathologies induced, at least initially, by yet-unknown or -unrecognized external or internal (including endogenous retroviral) infections.
© 2002 New York Academy of Sciences