Proc. Natl. Acad. Sci. USA Vol. 93, pp. 2257-2259, March 1996 Review Viruses, cytokines, antigens, and autoimmunity Roberto Gianani and Nora Sarvetnick* Department of Neuropharmacology, The Scripps Research Institute, 10666 North Torrey Pines Road, La Jolla, CA 92037

ABSTRACT To explain the pathogen- Autoantigens in Type I Diabetes of antigens since most self-responsive lym- esis of autoimmunity, we hypothesize that phocytes are eliminated within the thy- following an infection the immune re- Humoral and cellular autoimmunity to mus. Additionally, the organism has the sponse spreads to tissue-specific autoan- islet antigens has been detected both in ability to counterregulate such destructive tigens in genetically predisposed individ- new-onset diabetic patients and in their processes in the periphery through feed- uals eventually determining progression first degree relatives (3-6) and in animal back within intricate cytokine networks. to disease. Molecular mimicry between models of this disease (7, 8). In particular, However, in some small subset of geneti- viral and self antigens could, in some immune responsiveness to each of several cally predisposed individuals these "stop instances, initiate autoimmunity. Local individual islet constituents, such as glu- gap" measures of the body fail, allowing elicitation of inflammatory cytokines fol- tamic acid decarboxylase (GAD), has been the effectors generated to destroy self. lowing infection probably plays a pivotal linked repeatedly to type I diabetes. Two role in determining loss of functional research teams (9, 10) have shown that an Viruses as Triggers of Autoimmunity tolerance to self autoantigens and the immune response to GAD precedes the destructive activation of autoreactive onset of reactivity to other islet molecules, The question of what determines the ini- cells. We also describe the potential role such as insulin and carboxypeptidase, in tial loss of tolerance to a self antigen still of interleukin 10, a powerful B-cell acti- nonobese diabetic (NOD) mice. These remains to be answered. We favor the vator, in increasing the efficiency of authors also found that intrathymic or theory that infection can act as a trigger of epitope recognition, that could well be intravenous injection of GAD into such organ-specific autoimmunity in predis- crucial to the progression toward disease. mice precluded the spreading of reactivity posed individuals. An interesting example to other islet molecules and prevented ofthe association between a viral infection diabetes. Moreover, a T-cell clone specific and autoimmunity is offered by subacute Organ-Specific Autoimmunity for an islet-secreted protein accelerated thyroiditis, a disease whose clinical pre- diabetes in young NOD mice (11), as did sentation in humans strongly suggests a Autoimmune diseases are characterized by insulin-specific T-cell clones (12). Oral viral pathogenesis (17). Although sub- tissue destruction and functional impair- feeding with insulin (13) or immunization acute thyroiditis is not an autoimmune ment caused by autoreactive cells or anti- with insulin (14) along with both isoforms disease, these patients have autoantibod- bodies. Although the target antigens and of GAD (15, 16) also prevented diabetes ies and T-cell reactivity to thyroid autoan- genetic bases of several autoimmune dis- in this mouse strain. All these data indi- tigens (18, 19). These autoimmune phe- eases are now better understood, the initial cate that antigen-specific T-cell reactivity nomena are transient, and most of the events leading to loss of tolerance toward has a fundamental role in the effector patients recover completely. But, a few of self-constituents remain unknown. Typical phase of type I diabetes. these individuals progress from subacute of organ-specific autoimmunity is the selec- However, we believe that antigen- thyroiditis to autoimmune thyroid disease tive targeting of a single organ or individual specific T cells and antibodies represent (20-22). cell type, whereas gross anomalies of the epiphenomenons of a primary destructive The clinical historyofsubacute thyroiditis immune system are usually absent. event (such as a viral infection) that results may represent a general paradigm of organ- in local inflammation of the islet. These specific autoimmunity with autoimmune Type 1 Diabetes: Antigens and antigen-primed effectors are probably re- disease following viral infections in only a Autoimmunology of the Pancreas sponsible for the majority of the islet subset ofpatients. A potential role ofviruses destruction. Conceivably, as an aftermath and other infectious agents as inducers of Type I diabetes is among the most-studied of local cytokine secretion, a number of autoimmunity is also suggested by the asso- of the organ-specific autoimmune dis- self molecules in the pancreas become ciation of chronic infections such as malaria, eases. This disease develops as insulin- antigenic. This hypothesis would explain tuberculosis, syphilis, and schistosomiasis producing beta cells of the pancreatic why such an assortment of molecules as- with autoimmune phenomena (23). How- islets are infiltrated by macrophages and sumes antigenicity in type I diabetes and ever, the loss of tolerance to self antigens in lymphocytes (1, 2). Ultimately, the pro- why it is possible to induce disease with these patients does not usually lead to au- cess completely destroys the beta cells. several T-cell clones, specific for different toimmune disease, suggesting that other fac- But, years before type I diabetes is clini- molecules. The prevention of this disease tors must participate. cally evident, usually as an acute and by both GAD and insulin could well result The main difficulty in assigning a patho- sudden event, autoantibodies to islet an- from a bystander suppression of the im- genic role to any virus in type I diabetes is tigens can be identified in individuals who mune response. If so, the protective effect that during the long prodromal phase pre- subsequently progress to this disease. of GAD or insulin would not necessarily Therefore, a beta cell-specific virus is reflect a primary pathogenic role for the Abbreviations: GAD, glutamic acid decarbox- unlikely to be the direct cause of this islet immune response to these molecules. ylase; NOD, nonobese diabetic; IFN-,y, inter- destruction. Additionally, the fine speci- Of course, epidemiology studies tell us feron y. ficity of beta cell destruction in type I that the vast majority ofvirus infections do *To whom reprint requests should be addressed at: Department of Neuropharmacology, diabetes is so precise that the adjacent not lead to devastating autoimmune dis- CVN-10, The Scripps Research Institute, non-insulin-secreting endocrine cells of eases such as diabetes. The potential for 10666 North Torrey Pines Road, La Jolla, CA the pancreas are spared. autoimmunity is limited to a small subset 92037. 2257 Downloaded by guest on September 27, 2021 2258 Commentary: Gianani and Sarvetnick Proc. Natl. Acad. Sci. USA 93 (1996) ceding overt type I diabetes the infecting antigens of the infected tissue. We have Autoimmunity and Diversity agent would have disappeared. That is, investigated, in particular, the role of cyto- any related viral infection in the pancre- kines secreted within target tissues in deter- The search for target antigens in insulin- atic islets that triggered a long-term, mining the loss of tolerance to self antigens. dependent diabetes mellitus has identified chronic autoimmune response would have The expression of interferon ry (IFN--y) in a large and increasing number of pancre- occurred years earlier. Subsequently, the islet cells of transgenic mice led to destruc- atic islet molecules. The anti-islet re- virus would have been eliminated by an tion of the cells and development of type I sponse that causes diabetes is diverse from anti-viral immune response, making its diabetes (35). Local expression of IFN-,y in the point of T cells as well, with no T-cell direct identification at the clinical onset of the islets abolished tolerance to islet cell receptor predominating the islet reactive diabetes virtually impossible. antigens in a specific manner, since the repertoire. Thus, significant diversifica- However, suggestive evidence for a viral IFN-y transgenic mice rejected histocom- tion occurs and is probably essential for pathogenesis of type I diabetes comes from patible islet grafts but not pituitary trans- enough damage for clinical disease to its development as a sequel of congenital plants. Loss oftolerance to selfantigens can, develop. This notion is supported by the rubella infection (24). Interestingly, several therefore, occur not only as a result of a fact that induction of immune sensitiza- other endocrine diseases (25-27), thought systemic immune defect but also as a result tion by immunization to any of these islet to be autoimmune, can also develop follow- of local cytokine secretion. While the vast antigens individually does not lead to di- ing congenital rubella infection. Autoanti- majority of self reactive, high-affinity T cells abetes. Thus, from the initial "insult" bodies to tissue-specific antigens have been are destroyed in the thymus, some lower- there is a gradual recognition of increas- identified in the sera of patients with rubel- affinity clones escape selection. Indeed, un- ing numbers of islet antigens and recruit- la-associated endocrine disease, suggesting der normal conditions, these autoreactive ment of more islet reactive T-cell speci- their autoimmune pathogenesis (28, 29). cells reside in a quiescent state in the pe- ficities into the pancreas. Evidence for the Several years after intrauterine infec- riphery (36). Possibly, the induction of co- inter- and intramolecular spreading pro- tion, type I diabetes was diagnosed in stimulatory signals by local secretion of cy- cess in insulin-dependent diabetes melli- patients expressing the HLA class II DR3 tokines could activate these otherwise qui- tus has been presented recently (9, 10). or DR4 haplotypes (30); these haplotypes escent T cells. Our theory of this process is The molecular processes governing deter- also confer an increased risk of developing that, following viral infection and local cy- minant spreading in insulin-dependent di- type I diabetes when rubella is not in- tokine secretion, immune reactivity under- abetes mellitus are of significant interest, volved (31). Because congenital rubella goes dysregulation in specific target tissues since they allow a unique axis for inter- infection can directly induce immunologic of genetically predisposed individuals. In vention. B lymphocytes may be critical for abnormalities (32), rubella-associated au- this manner, a relatively common viral in- this diversification process since they have toimmunity may result. However, this fection with tropism for a given tissue could the unique capability to concentrate pro- pathogenic mechanism would not explain produce organ-specific autoimmune disease teins and present nondominant determi- why a dysfunction so induced would cause in some individuals. nants to T lymphocytes. A role for B organ-specific rather than systemic auto- Regional factors within tissues are also lymphocytes in the development of auto- immunity. clearly an important factor. IFN-,y is a immune diabetes in the NOD mouse was An alternative hypothesis is that an Thl-promoting cytokine that activates T demonstrated by experiments where B initial immune response directed to ru- cells but no detectable humoral response lymphocytes were abrogated by anti-IgM bella virus-infected cells could activate to islet antigens (37). However, the ex- administration (41). Factors that regulate reactivity not only to the viral antigens but pression of IFN--y in the motor end plate activation of B lymphocytes and enhance- also to self antigens on the cells. The led to development of antibodies to motor ment of their antigen presenting ability presence of particular HLA class I or class end plates and a disease clinically similar may be very important in the disease II molecules could affect the ability of to human myasthenia gravis (38). There- process (42). Interestingly, the cytokine antigen presenting cells to present self fore, this factor induces a distinct response interleukin 10 is a potent B-cell activator peptides specific for the target tissue. This, pathway in different target tissues. The and enhances major histocompatibility in turn, would determine the efficiency of importance of tissue-specific factors in the complex II expression on B cells. We have an immune response to these autoanti- induction of autoimmunity has been found that this cytokine accelerates dis- gens, a response that presumably could stressed by Barrett and coworkers (39). ease when overexpressed in the pancreatic destroy the target tissue. Compelling data These authors expressed H/K ATPase 13 islets (43) and it has also been found that for this "mimicry" mechanism as a poten- subunit, an autoantigen in thymectomy- it is essential for disease to progress (M. S. tial explanation for multiple sclerosis (33) induced autoimmune gastritis of BALB/c Lee, R. Mueller, L. S. Wicker, L. B. Peter- has recently been presented. T-cell clones mouse. Three days after undergoing son, and N.S., unpublished data). We now to the immunodominant myelin basic pro- thymectomies, the transgenic mice they hypothesize that this molecule is impor- tein were activated by a variety ofviral and used developed autoimmune gastritis and tant in the diversification process that bacterial peptides. Interestingly, align- periinsulitis but not diabetes. The induc- precedes disease. A further understanding ment of the primary amino acid sequences tion of insulitis in this model was antigen of cytokine regulation of diversification, did not indicate the potential for activa- specific and absolutely dependent on the processing enzymes and antigen present- tion in most of the cases. This work ele- T-cell response to the transgene product ing cells involved will be an important area gantly establishes the fact that molecular in the islets. Lack of diabetes in the pres- of upcoming investigation. mimicry may be an underappreciated ence of periinsulitis strongly suggests that mechanism for activation of subsets of T tissue-specific factors determine the This work was supported by a Diabetes In- cells that can mediate autoimmune tissue pathogenicity of a local immune response. terdisciplinary Research Program from the Ju- destruction (34). The inducibility of an autoimmune re- venile Diabetes Foundation International and sponse could be critically affected by a National Institutes of Health HD Grant 29764 Pathogenic Events Leading to Loss of localized feature, making the target cells to N.S. a more Self Tolerance Following Viral visible for the activated immune 1. Gepts, W. (1965) Diabetes 14, 619-633. Infection system. Results from a recent study (40) 2. Foulis, A. K., Liddle, C. N., Farquharson, showed the hyperinducibility of class II M. A., Richmond, J. A. & Weir, R. S. Several mechanisms have been proposed to molecules in thyrocytes obtained from (1986) Diabetologia 29, 267-274. explain how an initial anti-viral or bacterial patients with Graves disease and cultured 3. Baekkeskov, S., Aanstoot, H. J., Christ- response could allow the recognition of self in the presence of IFN-y. gau, S., Reetz, A., Solimena, M., Cascalho, Downloaded by guest on September 27, 2021 Commentary: Gianani and Sarvetnick Proc. Natl. Acad. Sci. USA 93 (1996) 2259 M., Folli, F., Richter-Olesen, H. & De 15. Pleau, J. M., Fernandez-Saravia, F., Es- 31. Thomson, G., Robinson, W. P., Kuhner, Camilli, P. (1990) Nature (London) 347, ling, A., Homo Delarche, F. & Dardenne, M. K., Joe, S., MacDonald, M. J., 151-156. M. (1995) Clin. Immunol. Immunopathol. Gottschall, J. L., Barbosa, J., Rich, S. S., 4. Pietropaolo, M., Castano, L., Babu, S., 76, 90-95. Bertrams, J. & Baur, M. P. (1988) Am. J. Buelow, R., Kuo, Y. L., Martin, S., Mar- 16. Elliot, J. F., Qin, H. Y., Bhatti, S., Smith, Hum. Genet. 43, 799-816. tin, A., Powers, A. C., Prochazka, M., D. K., Singh, R. K., Dillon, T., Lauzon, J. 32. Rabinowe, S. L., George, K. L., Loughlin, Naggert, J., Leiter, E. H. & Eisenbarth, & Singh, B. (1994) Diabetes 43, 1494- R., Soeldner, J. S. & Eisenbarth, G. S. G. S. (1993) J. Clin. Invest. 92, 359-371. 1499. Am. J. 5. Palmer, J. P., Asplin, C. M., Clemons, P., 17. Tomer, Y. & Davies, T. F. (1993) Endocr. (1986) Med. 81, 779-782. Lyen, K., Tatpati, O., Raghu, P. K. & Rev. 14, 107-120. 33. Wucherpfennig, K. W. & Strominger, Paquette, T. L. (1983) Science 222, 1137- 18. Volpe, R., Row, V. V. & Ezrin, C. (1967) J. L. (1995) Cell 80, 695-795. 1139. J. Clin. Endocrinol. Metab. 27, 1275-1284. 34. Oldstone, B. A. (1987) Cell 50, 819-820. 6. Roep, B. O., Kallan, A. A., Hazenbos, 19. Lio, S., Pontecorvi, A., Caruso, M., Mo- 35. Sarvetnick, N., Shizuru, J., Liggitt, D., W. L., Bruining, G. J., Bailyes, E. M., Ar- naco, F. & D'Armiento, M. (1984) Acta Martin, L., McIntyre, B., Gregory, A., den, S. D., Hutton, J. C. & de Vries, R. R. Endocrinol. 106, 67-70. Parslow, T. & Stewart, T. (1995) Nature (1991) Lancet 337, 1439-1441. 20. Wartofsky, L. & Schaaf, M. (1987) Am. J. (London) 346, 844-847. 7. Luhder, F., Woltanski, K. P., Hamann, J., Med. 81, 761-764. 36. Mitchinson, N. A. (1968) Immunology 15, Kloting, I., Ziegler, B. & Ziegler, M. 21. Werner, S. C. (1979) Arch. Intern. Med. 531-547. (1992) Diabetes Res. 20, 97-107. 139, 1313-1315. 37. Lee, M. S., Von Herrath, M., Reiser, H., 8. Gelber, C., Paborski, L., Singer, S., Mc- 22. Weetman, A. P., Smallridge, R. C., Nut- Oldstone, M. & Sarvetnick, N. (1995) J. Ateer, D., Tisch, R., Jolicour, C., Buelow, mann, T. B. & Burman, K. D. (1987) J. Clin. Invest. 95, 486-492. R., McDevitt, H. 0. & Fathman, C. G. Clin. Lab. Immunol. 23, 1-6. (1994) Diabetes 43, 33-39. 23. Abu-Shakra, M. & Shoenfeld, Y. (1991) 38. Gu, D., Wogensen, L., Calcutt, N. A., Xia, 9. Tisch, R., Yang, X. D., Singer, S. M., Immunol. Ser. 55, 285-313. C., Zhu, S., Merlie, J. P., Fox, H. S., Lind- Libleu, R. S., Fugger, L. & McDevitt, 24. Forrest, J. M., Menser, M. A. & Burgess, strom, J., Powell, H. C. & Sarvetnick, N. H. 0. (1993) Nature (London) 366, 72-75. J. A. (1971) Lancet 2, 332-334. (1995) J. Exp. Med. 181, 547-557. 10. Kaufman, D. L., Clare-Salzer, M., Tian, J., 25. Ziring, P. R., Fedun, B. A. & Cooper, 39. Barrett, S. P., van Drief, I. R., Tan, S. S., Forsthuber, T., Ting, G. S., Robinson, P., L. Z. (1975) J. Pediatr. 87, 1002. Alderuccio, F., Toh, B. H. & Gleeson, Atkinson, M. A., Sercarz, E. E., Tobin, 26. AvRuskin, T. W., Brakin, M. & Juan, C. P. A. (1995) Eur. J. Immunol. 25, 2686- A. J. & Lehmann, P. V. (1993) Nature (1982) Pediatrics 69, 495-496. 2694. (London) 366, 69-72. 27. Comas, A. P. (1976) J. Pediatr. 88, 1065- 40. Sospedra, M., Obiols, G., Barbi, L. F., 11. Haskins, K. & McDuffie, M. (1990) Sci- 1066. Tolosa, E., Vargas, F., Roura-Mir, C., ence 249, 1433-1436. 28. Schopfer, K., Matter, L., Flueler, U. & Lucas-Martin, A., Ercilla, G. & Pujol- 12. Daniel, D., Gill, R. G., Schloot, N. & Werder, E. (1982) Lancet 2, 159. Borrell, R. (1995) J. Immunol. 154, 4213- Wegmann, D. (1995) Eur. J. Immunol. 25, 29. Ginsberg-Fellner, F., Witt, M. E., Yagi- 4222. 1056-1062. chashi, S., Dobersen, M. J., Taub, F., Fe- 41. Forsgren, S., Anderson, A., Hillorn, V., 13. Zheng, C. F., Davidson, L., Eisenbarth, dun, B., McEvoy, R. C., Roman, S. H., G. S. & Weiner, H. (1991) Proc. Natl. Davies, T. F., Cooper, L. Z., Rubinstein, Soderstrom, A. & Holmberg, D. (1991) Acad. Sci. USA 88, 10252-10256. P. & Notkins, A. L. (1984) Diabetologia Scand. J. Immunol. 34, 445-451. 14. Muir, A., Peck, A., Clare-Salzler, M., 27, 87-89. 42. Mamula, M. J. & Craft, J. (1994) Curr. Song, Y. H., Cornelius, J., Luchetta, R., 30. Rubinstein, P., Walker, M. E., Fedun, B., Opin. Immunol. 6, 882-886. Krischer, J. & McLaren, N. (1995) J. Clin. Witt, M. E., Cooper, L. Z. & Ginsberg- 43. Wogensen, L., Lee, M. S. & Sarvetnick, N. Invest. 95, 628-634. Fellner, F. (1982) Diabetes 31, 1088-1091. (1994) J. Exp. Med. 179, 1379-1384. Downloaded by guest on September 27, 2021