Autoimmune diabetes is caused by selective loss of insulin-producing pancreatic ␤-cells. The main factors directly implicated in ␤-cell death are autoreactive, cytotoxic (islet-antigen specific) T-lymphocytes (CTL), and inflammatory cytokines. In this study, we have used an antigen-specific model of virally induced autoimmune diabetes to demonstrate that even high numbers of autoreactive CTL are unable to lyse ␤-cells by perforin unless major histocompatibility complex class I is upregulated on islets. This requires the presence of inflammatory cytokines induced by viral infection of the exocrine pancreas but not of the ␤-cells. Unexpectedly, we found that the resulting perforin-mediated killing of ␤-cells by autoreactive CTL is not sufficient to lead to clinically overt diabetes in vivo, and it is not an absolute prerequisite for the development of insulitis, as shown by studies in perforin-deficient transgenic mice. In turn, destruction of ␤-cells also requires a direct effect of ␥-interferon (IFN-␥), which is likely to be in synergy with other cytokines, as shown in double transgenic mice that express a mutated IFN-␥ receptor on their ␤-cells in addition to the viral (target) antigen and do not develop diabetes. Thus, destruction of most ␤-cells occurs as cytokine-mediated death and requires IFN-␥ in addition to perforin. Understanding these kinetics could be of high conceptual importance for the design of suitable interventions in prediabetic individuals at risk to develop type 1 diabetes. Diabetes

To asses the requirement of interleukin (IL)‐10 for peripheral CD4 T cell tolerance, the IL‐10 knockout (KO) was introduced into a T cell receptor‐transgenic mouse model (TCR1) specific for SV40 T antigen (Tag). IL‐10‐deficient TCR1‐transgenic mice failed to establish antigen‐specific T cell tolerance following sequential injections with Tag peptide. Nevertheless, IL‐10 was not required for the establishment of CD4 T cell tolerance in double transgenic RT2/TCR1 mice in which Tag is expressed endogenously under control of the insulin promoter. However, in contrast to stable anergy in wild‐type RT2/TCR1 mice, tolerant T cells in RT2/TCR1/Il‐10KO mice could be driven into vigorous proliferation by exogenous antigenic stimulation in vivo. The observed reactivation of anergic T cell populations in IL‐10‐deficient mice was only seen after in vivo but not in vitro peptide priming, reflecting an important regulatory function of IL‐10 in the context of the living organism. Taken together, these results demonstrate that IL‐10 is required to maintain T cell tolerance following exposure to enhanced antigenic stimuli but is not essential for the induction of self‐tolerance.