We have obtained evidence that in humans the profile of antigen-specific immune responses associated with active disease differs from that observed with latent infection. HLA-A2-restricted, Th1-type cytotoxic T lymphocyte responses against mycobacterial alanine dehydrogenase and glutamine synthetase can be detected only in non-vaccinated, tuberculin skin test-positive asymptomatic individuals but not in active tuberculosis patients. In the mouse lung, Th1 immunity is the principal cause for changes in the Mycobacterium tuberculosis growth from exponential to non-replicating persistence during infection. The shift in bacterial growth in response to Th1 immunity is accompanied by changes in the transcription pattern of immunodominant bacterial proteins. These data imply that some similarities may exist between antigenic profiles expressed by M. tuberculosis in human and mouse infections. We propose that T cell responses elicited against alanine dehydrogenase and glutamine synthetase are protective. Additionally, vaccination against these bacterial proteins expressed during chronic infection in mouse lung may be more effective than vaccination against other targets expressed predominantly at initial phase of infection. This hypothesis will be tested in a low-dose aerosol infection model employing mice transgenic for human HLA-A2 allele. The timing of expression of these bacterial proteins relative to the onset of Th1 immunity, and the kinetics of expression of antigen-specific T cell responses will be characterized during infection with M. tuberculosis. The protective potential of T cell-mediated immune responses against these proteins will be determined in DNA vaccinated and challenged mice. Finally, the molecular targets of polyclonal cytotoxic T lymphocyte responses elicited by DNA vaccines against the whole protein will be identified in challenged mice. This knowledge is expected to open a new approach in designs of vaccines against M. tuberculosis. ? ?