Beta-cell reactive T-cell clones from type I diabetes patients are not beta cell specific and recognize multiple...

Type I diabetes is the result of an autoimmune destruction of pancreatic beta cells. T cells appear to play a key role in this process. Thus far little information is available on the beta cell antigen or antigens recognized by auto-reactive T cells. Previously, we identified a 38 kD T cell antigen that appears to be localized in the membrane of insulin secretory granules and that is recognized by T cells from newly diagnosed type I diabetes patients. Other groups have reported T cell reactivity against glutamic acid decarboxylase (GAD). To obtain an indication of whether or not beta-cell reactive T cells from type I diabetes patients recognize a limited number of beta-cell antigens, we cloned T-cell lines reactive with rat insulinoma (RIN) membranes from two patients and analysed their antigen specificity. We also studied the antigen specificity of one RIN membrane reactive T-cell clone (1C5), previously isolated from a third patient. From the first patient two identical RIN membrane reactive T-cell clones (7A13) were isolated. The second patient yielded two identical (23A19) RIN membrane reactive T-cell clones, and one that was different (234A33). All clones were CD4+ and saw antigen in the context of different HLA class II alleles. The reactivity of the clones was, however, not restricted to beta cells: all clones showed cross-reactivity with one or more rat tissues, with some preference for those of neuroendocrine origin, but the cross-reactivity patterns were all different. All four clones recognized different fractions electro-eluted from RIN membranes: 29-36 kDa (7A13), 120-170 (23A19), 29-41 (23A33) and 56-72 kDa (1C5). The 23A33 clone reacted with the same 38 kDa fraction electro-eluted from insulinoma membranes as a beta-cell reactive clone (1C6) published previously, but none of the other known beta cell antigen preparations tested were recognized by the T-cell clones. Finally, the subcellular localization of the antigens recognized showed at least two different patterns. These data indicate that beta-cell reactive T cells from the peripheral blood of type I diabetes patients are not necessarily beta-cell specific and may be heterogeneous in regard to their antigen specificity and HLA class II restriction.