Merkel cell polyomavirus HLA class I epitopes for generating therapeutic T cell-based cancer immunotherapy

Summary Merkel cell carcinoma (MCC) is a highly aggressive skin cancer with an incidence of 3,000 USA cases/year. MCC carries a 2-year mortality of 40% with clear unmet medical need. Majority (65%) of MCC are virus-positive. The Merkel cell polyomavirus (MCPyV) is the main driver of virus-positive MCC tumors. Viral T antigens are

highly immunogenic because they are foreign to the immune system. Viral positive MCC can thus be an ideal setting to develop targeted immunotherapies including cancer vaccines. Several challenges confront our ability to develop a T cell-based MCC vaccine; i) the lack of knowledge of conserved immunogenic MCPyV-

derived T cell epitopes that bind to diverse human HLA alleles to provide immunity; ii) the lack of a vaccine platform for generating persistent memory T cells against cancer cells. To overcome these barriers, I have led efforts for immunopeptidome analysis, leading to the generation of monoallelic class I expressing cell lines

across 95 different HLA alleles. Using this resource, we identified >185,000 HLA class I-binding peptides, from which we trained highly accurate epitope prediction algorithms (HLAthena) 6,7. We developed robust protocols to establish patient-derived MCC lines; application of HLAthena to these lines has yielded new MCPyV epitopes,

which we confirmed by LC-MS/MS analysis (Lee JCI, in press). Finally, in collaboration with Dr. Moon we developed lipid nanodisc immunization to target professional antigen-presenting cells (APCs) that are capable of raising mucosal memory CD8 T cells to reject tumors.8 We hypothesize that cancer-driving MCPyV viral

antigens can yield highly immunogenic epitopes. We expect that such knowledge can fuel the development of an effective off-the-shelf cancer vaccine against MCC. Our aims are thus to:Aim 1. Systematically determine class I presented MCPyV epitopes across human HLA alleles that could provide global population

coverage for a T cell-based cancer vaccine. We have generated 95 single HLA allele-expressing B721.221 cell lines that provide global HLA coverage 7. We will transfect MCPyV LT and ST and seek to identify class I HLA-presented MCPyV epitopes by LC-MS/MS analysis. We will also predict epitopes using HLAthena. Epitopes

identified from the MS analysis and predictions will be further analyzed to identify conserved promiscuous epitopes which could be candidates for a universal off-the-shelf T cell-based MCPyV vaccine.Aim 2. Confirm the immunogenicity of the candidate MCPyV epitopes. Immunogenic validation and characterization of

MCPyV epitopes will enable the discovery of candidate protective MCC-reactive T cell epitopes. We will utilize already banked HLA typed naïve and MCC patient human PBMCs in these studies to determine T cell immune memory, de-novo T cell responses, TCR avidity, and cytotoxic activity against MCC tumor cell lines. Aim 3. Test

the impact of T cell-based therapies directed against MCPyV in in vivo models. We will devise multi-epitope vaccines using the nanodisc platform. These candidate MCPyV-targeting vaccines will be tested in MCC GEM model for tumor control. We will further characterize MCC GEM model tumor microenvironment +/- vaccine.