HIV-specific cytotoxic T lymphocytes (CTL) constitute a critical component of the protective immune response early in infection by eliminating virus infected cells. CTL recognize a limited number of peptides in both structural and regulatory proteins presented by class I molecules. Induction of CTL is dependent on both the primary peptide sequence and cellular cofactors that regulate antigen processing and presentation. Equally important are extracellular factors responsible for CTL expansion. The goals of this proposal are to determine the relative contribution of properties intrinsic to an epitope itself and its molecular context for class I presentation, CTL induction and memory. Four fundamental questions are addressed: First, do peptides with high class I affinity alter immunodominance? CTL responses to three gp160-derived epitopes with differential class I affinity restricted to the same class I molecule will be monitored in the presence and absence of competing peptides. Second, is processing of epitopes derived from secretory proteins masked by glycosylation? Mutations abrogating glycosylation will be introduced into a secreted form of the gp160 V3 loop and the variants compared for CTL recognition and induction. Third, do flanking residues exert a global, rather than epitope-specific, effect on class I presentation. A global relevance of inhibiting flanking residues will be analyzed by placing in vivo challenge model will confirm a critical role of epitope context in vaccine design. The fourth question addresses the survival of epitope-specific CTLp following primary stimulation and the effects of restimulation on the existing CTL memory pool. The basic experimental approach uses minigenes as tools for genetically linking well characterized viral CTL epitopes. The multi-valent peptides will be expressed using both recombinant vaccinia and Sindbis viruses. The kinetics of presentation from chimeric antigens will be monitored by HIV-specific CTL recognition; the efficacy of CTL induction will be determined by analysis of CTL precursor frequencies. This strategy allows the differentiation of effects attributed to peptide/class I affinity and molecular context-dependent processing. A precise analysis of CTL responses in a controlled model system will reveal protein-specific parameters contributing to immunodominance. Furthermore, these data will contribute to the design of multiple T cell epitope vaccines for the induction of heterologous virus-specific CD8+ CTL.
