High-throughput screening for rare antigen-reactive TCRs using natively-paired TCRαβ expression libraries generated from millions diverse primary T cells

Abstract

Abstract A major goal of T cell research is to profile the repertoire of antigen-reactive TCRs targeting specific peptide:MHCs (pMHC). Single cell sorting of pMHC-binding T cells using multimers and ex vivo antigen expansion are the current gold standards for identifying antigen-reactive primary T cells. However, functional validation of the TCRs identified by these approaches requires resource intensive cloning of each individual TCRαβ pair. Additionally, primary T cells, especially tumor infiltrating lymphocytes (TILs), are a limited resource, which restricts the number of antigens that can be screened. To address this, we developed a microfluidic approach to capture and functionally express natively-paired TCRαβ libraries from millions of single T cells. Unlike DNA barcoding approaches that mark single cells by adding a sequence tag, we physically link the TCRα-TCRβ chains to generate sequencing and full-length expression libraries which we introduce into Jurkat cells for functional testing. Using these methods, we captured over 2.9 million TCRαβ clonotypes from six healthy PBMC donors and over 0.5 million from expanded melanoma TIL samples. We applied pMHC binding and cellular activation screens to identify and validate 14 TCRs reactive to common viral and tumor associated antigens with starting library frequencies of <0.0001 – 0.3% and functional avidities ranging from 0.3 – 592nM. We found that including the cellular activation screen reduced our false-positive TCR discovery rate from 65% for pMHC binding alone to 18%. We are further developing these cellular activation enrichment approaches and applying them to broadly profile anti-tumor TCR reactivity from TILs.