We studied the mechanisms for T cell recognition of antigens in association with major histocompatibility complex (MHC)-encoded molecules, and applications to the design of synthetic vaccines for AIDS and cancer. We developed synthetic vaccines for HIV using broadly reactive HIV helper, cytotoxic T lymphocyte (CTL) and neutralizing antibody epitopes, and showed the importance of an intrinsic or covalently linked helper epitope for induction of CTL. We developed ways to steer responses with cytokines. Using an adjuvant that we found allowed induction of CTL, TH1 help, and neutralizing antibodies, we commenced a human phase I immunotherapy trial with these vaccine constructs. Initial results show induction of a new HIV peptide-specific CTL response in one patient not present prior to vaccination, the first induction of CTL with a peptide vaccine in humans. We found in murine studies that peptide vaccines for HIV can be made more potent or more broadly effective by selective introduction of mutations that improve binding to MHC or T-cell receptors ("epitope enhancement") and demonstrated proof of principle with second generation synthetic vaccine constructs in mice. To apply these approaches to human vaccines we studied the binding of HIV envelope and reverse transcriptase peptides to human HLA molecules, raised HIV peptide-specific human CTL, and mapped the residues involved in binding to HLA-A2 and to the T-cell receptor (TCR). We showed that avidity of CTL can be selected and plays a profound role in efficacy for adoptive immunotherapy of viral infection. However, high concentrations of antigen can induce apoptosis of high avidity CTL by a TNF-dependent, fas-independent mechanism. We showed that free peptide inhibits CTL, by a self-veto-like mechanism involving dual engagement of MHC and TCR on the TCR. We showed the profound effect of cytokine imbalance on CTL activity and identified cells that suppress the CTL response. We identified several CTL epitopes in proteins of the hepatitis C virus that causes liver cancer using a novel approach, and developed model vaccine constructs. We showed an inverse correlation between IL-2 response to human papillomavirus peptides and stage of cervical neoplastic disease, the first specific T-cell response correlated with progression of human neoplasia. We developed peptide cancer vaccines inducing CTL immunity to mutant p53 expressed in cancer cells. We induced murine CTL against fusion proteins from chromosomal translocations in pediatric tumors, that protect mice. 23 patients have been entered in a phase I/II clinical trial of the mutant p53/ras peptide vaccine.