Identifying and engineering TCR specificity against solid tumor neoantigens

Abstract Neoantigens (NeoAg) offer attractive therapeutic targets for directing a patient’s immune response to the immunogenic subset of mutations expressed exclusively by their cancer cells. Despite the specificity with which NeoAg enable tumor recognition, the majority of approaches for their identification rely on purely predictive methods such as calculating the ability of mutated peptides to bind to a patient’s set of HLA molecules. These methods have met with limited success in revealing natural targets present on tumor cells. We have developed a novel HLA-agnostic functional approach to NeoAg identification which combines genomic sequencing with bioinformatic analysis to nominate mutations for subsequent functional analysis using patient’s own T cells in an effort to identify natural responses generated under physiologic conditions. Using this, we identified a missense mutation (V205I) in the ribosomal protein RPS2 that is recognized by CD8+ T cells from tumor-infiltrating lymphocytes (TIL) of a metastatic HPV16+ Head and Neck Squamous Cell Carcinoma lesion. We then performed adoptive cellular therapy (ACT) using either unseparated TIL or those enriched for RPS2 V205I-specific CD8+ T cells and found the latter to be superior in controlling outgrowth of tumor of a PDX cell line generated from this lesion in NSG mice. Finally, we used single-cell transcriptomics to isolate the genes encoding the RPS2-specific TCR and show that it recognizes the mutated peptide bound to HLA-B*07:02. These results demonstrate that high-affinity NeoAg-specific T cell responses can be identified in cancer patients, that ACT of these cells can control tumor growth, and that the relevant TCR can be isolated for use in TCR engineering-based immunotherapy.