LC-MSMS of the CT26 MHC I immunopeptidome

MHC I-associated peptides (MAPs) presented at the surface of nucleated cells play a central role in CD8 T-cell immunosurveillance. MAPs presented by mature (i.e. MHC IIhi) medullary thymic epithelial cells (mTEChi) are essential to eliminate self-reactive CD8 T cells in a process called central tolerance. On tumor cells, MAPs that do not induce tolerance (i.e. non-tolerogenic MAPs), because absent from mTEChi or any other normal cells, are referred to as tumor-specific antigens (TSAs). Despite their clinical relevance, very few have been identified, even in highly mutated tumor types. Thus, we developed a novel proteogenomic workflow able to characterize the full TSA landscape of any tumor. Briefly, using RNA-seq data, we subtracted the mTEChi from the tumor signal to generate tumor-specific protein databases enriched in non-tolerogenic sequences. Using these databases to analyze the MAP repertoire of two murine cell lines (CT26 and EL4) sequenced by mass spectrometry, we identified a total of 21 TSAs, 90% of which derived from allegedly non-coding regions. Interestingly, our results highlighted that 70% of those TSAs derived from non-mutated yet tumor-restricted sequences, e.g. endogenous retroelements. Focusing on 5 TSAs, we demonstrated that both TSA expression and TSA-specific T-cell frequency in the pre-immune repertoire influenced the overall survival of pre-immunized tumor-bearing mice. In conclusion, this proof-of-concept study demonstrates that non-coding-derived TSAs are frequent and protective in vivo, while they could be shared by several individuals. Altogether, our findings will help expand the repertoire of human TSAs and facilitate their prioritization in the clinic