Mismatch in epitope specificities between IFNγ inflamed and uninflamed conditions leads to escape from T lymphocyte killing in melanoma

Katherine Woods, Ashley J. Knights, Matthew Anaka, Ralf B. Schittenhelm, Anthony W. Purcell, Andreas Behren, Jonathan Cebon

A current focus in cancer treatment is to broaden responses to immunotherapy. One reason these therapies may prove inadequate is that T lymphocytes fail to recognize the tumor due to differences in immunogenic epitopes presented by the cancer cells under inflammatory or non-inflammatory conditions. The antigen processing machinery of the cell, the proteasome, cleaves proteins into peptide epitopes for presentation on MHC complexes. Immunoproteasomes in inflammatory melanomas, and in antigen presenting cells of the immune system, are enzymatically different to standard proteasomes expressed by tumors with no inflammation. This corresponds to alterations in protein cleavage between proteasome subtypes, and a disparate repertoire of MHC-presented epitopes. We assessed steady state and IFNγ-induced immunoproteasome expression in melanoma cells. Using epitope specific T-lymphocyte clones, we studied processing and presentation of three NY-ESO-1 HLA-Cw3 restricted epitopes by melanoma cell lines. Our experimental model allowed comparison of the processing of three distinct epitopes from a single antigen presented on the same HLA complex. We further investigated processing of these epitopes by direct inhibition, or siRNA mediated knockdown, of the immunoproteasome catalytic subunit LMP7. Our data demonstrated a profound difference in the way in which immunogenic T-lymphocyte epitopes are presented by melanoma cells under IFNγ inflammatory versus non-inflammatory conditions. These alterations led to significant changes in the ability of T-lymphocytes to recognize and target melanoma cells. Our results illustrate a little-studied mechanism of immune escape by tumor cells which, with appropriate understanding and treatment, may be reversible. These data have implications for the design of cancer vaccines and adoptive T cell therapies.