ABSTRACT The HIV/AIDS epidemic continues to have enormous medical, societal and economic implications worldwide. While combination anti-retroviral therapy (cART) has greatly reduced the global burden of HIV, the ability of the virus to establish a persistent reservoir within the body requires that HIV-infected individuals remain on lifelong treatment. As a result, new modalities that can suppress the viral reservoir and thereby limit the requirement of HIV treatment are greatly needed. Recent efforts have been focused on the induction of cytotoxic T cells (CTLs) by therapeutic vaccines. However, the accumulation of CTL escape mutations in chronically infected, cART-suppressed patients has greatly limited the ability of CTLs to successfully prevent viral rebound following cART cessation. Thus, in order to counteract this viral escape, this DP2 proposal will focus on the study of CTL responses to a new set of targets, known as `highly networked' epitopes, to determine whether they can form the basis of a novel therapeutic CTL-based vaccine for HIV. These highly networked epitopes were identified using an innovative approach known as structure-based network analysis. By applying network theory principles to HIV protein structure data, the approach was able to identify a set of epitopes that are intolerant to mutation, but which are also presented by a broad array of HLA alleles. Moreover, the targeting of highly networked epitopes by functional CTL responses was shown to strongly distinguish individuals who naturally control HIV from those with progressive disease. Thus, the goal now is to determine whether CTLs directed against highly networked epitopes can also suppress viral outgrowth following cART cessation in the remaining ~99% of chronically-infected, cART-treated individuals. This will be accomplished by: (i) deep sequencing highly networked epitopes in proviral DNA derived from peripheral blood and gastrointestinal tissue and (ii) determining whether CTLs targeting highly networked epitopes can suppress latent virus outgrowth both ex vivo and in a humanized mouse model. Demonstrating the effectiveness of CTL-mediated responses to highly networked epitopes will confirm the value of the structure-based network analysis approach to guide the rational design of a effective, therapeutic CTL-based vaccine for HIV.