Adoptive T cell therapy is a promising therapeutic approach with the potential to eradicate otherwise incurable cancers. Through T cell receptor (TCR) gene therapy, donor- or patient-derived T cells can be reprogrammed to recognize tumor-associated antigens (Ags); the ability of these engineered T cells to eliminate tumor cells is largely dictated by the affinity of the transferred TCR. As thymic selection eliminates T cells with high avidity for self, most tumor-reactive T cells in the native repertoire have limited antitumor efficacy. While TCRs can be mutated to enhance affinity, such TCRs carry the clinically-validated risk of off-target or on-target/off-tissue toxicity. Therefore, one of the key challenges of T cell-based immunotherapies is to identify TCRs with sufficient affinity to mediate effective anti-tumor activity without compromising safety.

Based on the premise that a wide variability in thymus-vetted, tumor/self antigen-specific T cell avidity exists among individuals, we established a high-throughput method to simultaneously affinity rank and quickly retrieve the highest affinity TCRs from any number of pooled matched donors. We hypothesized high-affinity clonotypes would be preferentially enriched by binding to limiting concentrations of antigen-specific peptide/major histocompatibility complex (pMHC) multimers, such that TCR clonotype enrichment under increasingly stringent conditions would identify the highest affinity TCRs from within heterogenous T cell populations.

Using this approach, we isolated TCRs specific for HLA A2-restricted epitopes of over-expressed self Ags (Wilms’ Tumor Antigen 1 [WT1], Mesothelin, Melanoma Antigen Recognized by T cells 1 [MART1]), cancer testis (CT) Ags (Melanoma-associated antigen 1 [MAGE A1]), as well as for viruses, including HIV and the Merkel cell polyoma virus. High-throughput TCRα/β chain pairing was obtained by PairedSeq (Adaptive Biotechnologies) or by single-cell sequencing (10X Genomics).

Our results show that 1) clonotype Enrichment Ratios positively correlated with TCR affinity, thus validating our method; 2) the highest affinity T cell clones targeting self-Ags (but not viral Ags) were present at very low frequencies in native TCR repertoires, suggesting impaired antigen-driven expansion in vitro and precluding detection by conventional cloning methods; 3) many of the highest-affinity TCRs targeting viral, CT or self Ags bound pMHC complexes independent of CD8 co-receptor, confirming that tumor/self Ag-specific T cells are not entirely deleted by negative selection in healthy donors; and 4) select class I TCRs thus identified activated CD4+ (class II HLA-restricted) T cells to produce cytokines and lyse cells expressing cognate antigen.

This method is poised to rapidly (~6 weeks) identify for clinical translation rare, native, high-affinity tumor-reactive TCRs with the potential to engage CD4+ T cells and enhance anti-tumor CD8+ T cell activity, but that have a low likelihood of off-target toxicity.

Citation Format: Thomas M. Schmitt, Megan McAfee, Harlan Robins, Lara Kropp, Philip D. Greenberg, Aude G. Chapuis. High-throughput method identifies rare, high-affinity, thymus-vetted T cell receptors (TCRs) for clinical translation [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2017 Oct 1-4; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2018;6(9 Suppl):Abstract nr A51.