One of the critical steps to developing curative and tumor-specific immunotherapy is the identification and selection of antigens with tumor-restricted expression in order to avoid undesirable immune responses against normal tissues. Proteins with tumor-specific mutations have the potential to be the most restricted of all tumor antigens. However, the comprehensive identification of such 'neoantigens' has only become feasible recently using next-generation massively parallel sequencing technologies, providing unprecedented genetic information about cancer cells. The current project seeks to apply tumor sequence information to study this under-evaluated class of tumor-specific antigens. We hypothesize that immune responses against chronic lymphocytic leukemia (CLL) - observed in patients treated with hematopeotic stem cell transplant (HSCT), donor lymphocyte infusion (DLI) and leukemia whole cell vaccination - are often targeted against neoantigens generated from somatic tumor mutations. We have recently completed whole genome or exome sequencing of paired leukemia and normal tissue from 7 individuals with CLL, and have discovered many missense mutations, splice-site changes and gene fusions per individual. The current proposal focuses on defining the immunologic consequence of these genetic changes. We propose to systematically identify mutated peptides that are expressed, processed and presented by MHC molecules on CLL tumor cells (Aims 1-2). We will then test whether CD8+ T cells targeting these mutated peptides show selective lysis of CLL cells ex vivo and expand in number in HSCT/DLI-treated or vaccinated patients concurrent with the anti-leukemia response (Aim 3). The identification of tumor-specific neoantigens and T cell responses against them would help explain the success of HSCT/DLI in CLL and provide a motivation for the development of individualized, highly-specific and safer tumor vaccines.