Genome editing of peptide loading reveals risks of immunotherapeutic targeting of TAP1 independent peptides

Abstract

Current immunotherapeutic strategies for cancer aim to target HLA class I (HLA-I) presented neo-epitopes. A subgroup of neo-epitopes has been proposed to be presented only in the absence of the TAP transporter. We here identified a novel TAP-independent antigen that was surprisingly effectively presented by TAP-sufficient cells. We engineered a library of HLA-I pathway CRISPR/Cas9-mediated knockout cells on a single genetic background allowing for an optimal comparison of individual component contributions to antigen presentation. Sequence-verified knockouts for HLA-I (-A, -B and -C), B2M, Calnexin, α-Glucosidase II, Calreticulin, ERAP1, TAP1, Tapasin and ERp57 were created. The HLA-I surface expression of these cells depended on both the targeted gene and the detected HLA-I allele, as expected. To investigate the presentation of single antigens, we targeted the knockout library with a panel of peptide-specific T cell clones. Surprisingly, T cells specific for a peptide derived from the endogenously expressed SSR1 protein recognized TAP1-deficient cells at a similar efficiency as wild type cells. These T cells were derived from a patient during graft-versus-host-disease, suggesting that they contributed to the destruction of healthy TAP-sufficient cells. Further experiments demonstrated the functional presentation of the SSR1 antigen by various healthy cells, underscoring that TAP-independent antigens can contribute to unwanted T cell reactivity. In conclusion, we generated and validated a novel cellular toolset that allows comparative research in the HLA-I antigen processing and presentation pathway. Using this toolset we found a TAP-independent peptide that is functionally presented on healthy cells. Our data underscore potential pitfalls of targeting TAP-independent antigens as immunotherapeutic approach