The generation and selection of the T cell repertoire: insights from studies of the molecular basis of T cell recognition.

The molecular basis of T cell corecognition of antigen and products of the major histocompatibility complex (MHC) was examined using the technique of DNA-mediated gene transfer. Introduction of the genes encoding the alpha and beta chains of the clonotypic molecule (Ti) of the pigeon cytochrome c-specific, I-Ek-restricted murine T cell hybridoma 2B4 into the human leukemic T cell Jurkat or its T3- variants led to the surface expression of various dimeric combinations of human and mouse Ti chains. All alpha beta dimers were associated in a 1:1 ratio with the chains of the T3 complex and all Ti-T3 complexes could transmit effective transmembrane signals for IL2 gene activation following binding and cross-linking by anti-T3 or anti-clonotypic antibodies. However, only the reconstituted mouse 2B4 alpha beta dimer mediated functional responses to antigen-presenting cells bearing the appropriate combination of antigen and Ia molecules. The fine specificity of the antigen-MHC molecule responses precisely matched that of the 2B4 T cell gene donor, establishing that the alpha beta clonotypic molecule is both necessary and sufficient to fully define the dual specificity of a T cell. This result, together with similar data from other laboratories, provides direct evidence in favor of the "single receptor" model of T cell recognition. Analysis of past experiments looking for independent recognition of antigen vs. MHC molecules led us to conclude that limitations on either formation of mixed receptor dimers or of appropriate antigen peptide-Ia molecule pairs may have limited the utility of these studies in providing an answer to this question. Therefore, additional gene transfer experiments were carried out to investigate the role of the two chains of the alpha beta heterodimer in controlling antigen vs. MHC molecule specificity. Our results showed that under conditions in which it was assured that antigen-Ia complexes formed, one could produce a receptor molecule composed of the alpha chain of one T cell hybridoma (2H10) and the beta chain of a second hybridoma (2B4) with a mixed antigen and MHC molecule specificity that could be predicted based on the specificities of the parent T cells. We discuss at length the implication of this result for the issue of single- vs. dual-site recognition models of T cell receptor function. In addition to these results looking directly at specificity of receptor molecules, we have also analyzed expression of various combinations of Ti chains. For alpha and beta, clear asymmetries in the efficiency of surface receptor expression were observed for certain pairs of chains.(ABSTRACT TRUNCATED AT 400 WORDS)