The shaping of the brain-specific T lymphocyte repertoire in the thymus

We have shown in several distinct experimental systems that the immune system of intact Lewis rats contains T cells which, upon activation, are able to mediate autoimmune brain inflammation. These T cells seem to differentiate within the thymus although the autoantigens are produced (and presumably expressed in a recognizable fashion) within the thymic medulla. Furthermore, an intact fully MHC compatible thymic microenvironment seems to be required for the development of all features of the autoimmune TCR repertoire. Biased utilization of V beta 8.2 gene for the TCR, a hallmark of the Lewis rat T cell response to MBP, is only seen in T cells having matured in thymuses entirely composed of stroma elements of rat origin. It seems that the thymus contains a large spectrum of protein structures, which hitherto had been considered autoantigens specific for "peripheral" tissues, and, most surprisingly, components of the CNS, the classical "sequestered" organ. Deletion of autoreactive T cell clones by many local intrathymic autoantigens is leaky, at best. The reduced expression of CD4 on thymus-derived autoreactive T cells may be construed to reflect abortive efforts of negative selection. Alternatively, however, it may be worthwhile to consider a positive role for intrathymic autoantigens and their complementary T cells clones. It is possible that the requirement of an intact thymus milieu for the typical, V beta 8.2 dominated MBP specific T cell repertoire in the Lewis rat could reflect self peptide presentation by thymus epithelium cells in positive selection stages. In that case, the unusual diversity of thymic autoantigens could indeed have a role in shaping the immune system's TCR diversity, possible in the sense of an "immunological homunculus" as postulated by Cohen (Cohen 1992). Finally, there is a need to explain the mechanisms that in the healthy organism prevent the numerous, potentially autoaggressive T cell clones from attacking the body's own tissues. This is especially important, as T cells reactive against potentially pathogenic autoantigens, e.g. MBP (Ota et al. 1990, Pette et al. 1990b) and acetylcholine receptor (Salvetti et al. 1991, Sommer et al. 1991), are seen at especially high frequency in the human immune repertoire. Clearly, in all experimental paradigms investigated, activation of self-reactive T cells was the critical prerequisite for induction of autoimmune disease. Thus, in principle, prevention of such activation would be one way to maintain self tolerance. The mechanisms that achieve this goal in most individuals remain to be elucidated.