The T cell response to the 65-kD mycobacterial heat-shock protein (Bhsp65) has been implicated in the pathogenesis of autoimmune arthritis. Adjuvant arthritis (AA) induced in the Lewis rat (RT-1(l)) by injection of Mycobacterium tuberculosis serves as an experimental model for human rheumatoid arthritis (RA). However, the immunological basis of regulation of acute AA, or of susceptibility/resistance to AA is not known. We have defined the specificity of the proliferative T cell responses to Bhsp65 during the course of AA in the Lewis rat. During the early phase of the disease (6-9 d after onset of AA), Lewis rats raised T cell responses to many determinants within Bhsp65, spread throughout the molecule. Importantly, in the late phase of the disease (8-10 wk after onset of AA), there was evidence for diversification of the T cell responses toward Bhsp65 carboxy-terminal determinants (BCTD) (namely, 417-431, 441-455, 465-479, 513-527, and 521-535). Moreover, arthritic rats in the late phase of AA also raised vigorous T cell responses to those carboxy-terminal determinants within self(rat) hsp65 (Rhsp65) that correspond in position to the above BCTD. These results suggest that the observed diversification is possibly triggered in vivo by induction of self(Rhsp65)-reactive T cells. Interestingly, another strain of rat, the Wistar Kyoto (WKY/NHsd) rat (RT-1(l)), with the same major histocompatibility complex class II molecules as the Lewis rat, was found to be resistant to AA. In WKY rats, vigorous responses to the BCTD, to which the Lewis rat responded only in the late phase of AA, were observed very early, 10 d after injection of M. tuberculosis, Strikingly, pretreatment with the peptides comprising the set of BCTD, but not its amino-terminal determinants, provided significant protection to naive Lewis rats from subsequent induction of AA. Thus, T cell responses to the BCTD are involved in regulating inflammatory arthritis in the Lewis rat and in conferring resistance to AA in the WKY rat. These results have important implications in understanding the pathogenesis of RA and in devising new immunotherapeutic strategies for this disease.