Unraveling the Mechanisms of Neoantigen Vaccine-Induced Epitope Spreading

Project Abstract The advent of immunotherapy has reshaped our approach to cancer treatment, with lasting cures achieved in a subset of patients. Tumor neoantigen vaccines have been shown to effectively prime tumor-directed T cell responses and induce epitope spreading (ES), a phenomenon linked to improved clinical outcome. ES occurs when vaccine-induced tumor cell lysis releases additional tumor antigens, which subsequently primes a new wave of tumor-specific T cells. However, technical challenges have restricted the evaluation of ES to only small sets of known target antigens, while it remains unknown how T cells primed through vaccines differ phenotypically from naturally primed (via ES) T cells, and what types of T cell – target cell interactions are favored with respect to their functional avidities. In addition, ES to novel non-canonical antigens, such as non-canonical open reading frames (nuORFs), structural variants (SVs), and transposable elements (TEs) may also contribute to anti-tumor T cell responses when conventional antigens (derived from point mutations and short insertions and deletions) are scarce. In recent years, the Wu and Schumacher groups have developed a set of immunogenomics-based tools, which provide new approaches to address these unresolved questions. Specifically, we have developed neoantigen prediction pipelines facilitating the identification of antigens from whole-exome sequencing data and mass spectrometry. Further, we have refined our antigen discovery abilities, by focusing on the detection of potential antigens derived from non-canonical antigens. Finally, we have conceived HANSolo and PAIR-Scan, novel library-based tools capable of generating functional, paired TCR- antigen information at scale. With these capabilities, I propose to thus systematically dissect neoantigen vaccine- induced ES in two distinct settings: In Specific Aim 1, I will systematically evaluate ES in melanoma, a tumor type characterized by a high tumor mutational burden (TMB). I hypothesize that vaccine-induced ES generates T cells of memory-like phenotype characterized by a defined spectrum of functional avidities towards their respective target antigens, resulting in enhanced tumor control. In Specific Aim 2, I will evaluate the immunogenicity of non-canonical antigens in the setting of high-grade serous ovarian cancer, a tumor with low TMB, yet rich in nuORFs, SVs, and TE. I will furthermore assess how potential ES to these non-canonical antigens may contribute to disease control. By relating features underlying ES to clinical biomarkers, I hope to determine how this phenomenon is related to positive clinical outcomes during cancer immunotherapy.