There are a number of mechanisms which cooperate to produce and maintain T-cell tolerance. First, and perhaps most important, is the clonal deletion in the thymus of T cells with high affinity for self antigens. However, to ensure that a wide repertoire of T cells is available in the periphery to combat foreign antigens, the threshold of clonal deletion may be set low enough so that T cells whose TCR's have sub-threshold affinity for self antigens mature and migrate to the periphery. T cells which recognize self antigen-derived peptides not expressed or presented in the thymus will also fail to be deleted. For those self-reactive T cells which are not deleted in the thymus, other mechanisms may produce tolerance, including an undefined alteration of signalling pathways which produces clonal anergy, and lowering the avidity of the TCR for its ligand by downregulating coreceptor and accessory molecules. Active suppression of T-cell responses in another well-described phenomenon whose mechanism is undefined. From our observations with the model systems discussed here, we have observed three distinct mechanisms by which T-cell tolerance can be circumvented, allowing autoimmune phenomena to occur. These mechanisms may have relevance for different types of autoimmune diseases seen in humans. In gld mice, the autoimmune disease seems to be related to a global defect in T-cell differentiation and function, which allows for the expansion of autoimmune B cells. While we showed that clonal deletion of V beta-bearing T cells is appropriate in certain cases, aberrant lymphokine secretion by the abnormal T cells or disruption of immune system regulation are most probably responsible for allowing autoantibody production. While human lupus erythematosis shares much of the pathology of lpr and gld mice, there is no expansion of T cells with a similar phenotype in human lupus. There are environmental factors which must play a role in the development of human lupus, since the incidence of the disease does not follow an absolute genetic pattern. The escape from clonal deletion and subsequent reactivation of autoimmune T cells which we observed in V beta 8.1 TCR-transgenic mice can be a model for human autoimmune diseases such as multiple sclerosis and type I diabetes, in which T cells are directed against a specific autoantigen. According to this model, susceptibility loci for autoimmune disease such as the MHC would function by producing different repertoires of T cells which in some cases could gain autoreactivity following activation.(ABSTRACT TRUNCATED AT 400 WORDS)