Abstract B036: Exploration of the human leukocyte antigen immunopeptidome in pancreatic adenocarcinoma

Abstract

Abstract Pancreatic adenocarcinoma (PDAC) is a lethal disease with an 11% five-year survival rate and frequent distant metastasis at diagnosis. The immune escape mechanisms driving this unchecked proliferation in PDAC remain unclear. The HLA system plays a crucial role in immune-mediated recognition of neoplasms. HLA class I presents endogenously processed peptides to circulating CD8+ T-cells, while HLA class II presents exogenous peptides to CD4+ T-cells, with the collective sum of these peptides known as the immunopeptidome. Most immunopeptidome studies rely on peptides derived solely from cell lines, with few peptides directly reported from patients. Characterization of the immunopeptidome in cancer patients is critical for understanding peptide presentation and neoantigen recognition in PDAC. We hypothesize that isolating HLA class I and II peptides directly from PDAC donors will identify novel neoantigen-derived peptides presented to the immune system with high fidelity for specific HLA alleles. Whole Exome Sequencing (WES) was performed on nine PDAC donors from the UNMC Rapid Autopsy Program. Somatic mutations were identified using best practices published by the BROAD Institute. HLA alleles were identified with the HLA-HD algorithm. Large-scale immunopurification of HLA class I complexes and associated peptides was performed on PDAC liver metastases. Data-Dependent-Acquisition MS/MS with PEAKS analysis identified individual peptide sequences. Similar strategies are being employed for HLA class II peptide isolation from matched donor spleens. To date 39,639 HLA class I peptides and 2,995 HLA class II peptides have been identified from nine donors. Of HLA class I isolated peptides, 9,987 unique sequences have been traced and mapped to the human reference genome. Of the remaining 29,652 unmapped peptides, we have manual confirmation of the presence of neoantigen-derived peptides within that pool, with a pipeline for large scale alignment of these peptides to the annotated mutated genome pending. The gold standard HLA-peptide predictive binding algorithm, netMHC, confirms that 24.2% of our isolated HLA class I peptides are predicted binders for their patient-specific HLA alleles. A notable finding is that 9% of identified HLA class II peptide sequences contained sequences identical to those isolated during independent HLA class I isolations from the same donor. These data establish the high-fidelity of our pipeline for identification of peptides presented by the HLA system directly from PDAC donors. The quantity of HLA-associated peptides discovered sets this study on track to surpass all previously reported patient-derived immunopeptidome studies. The observed conservation of HLA class I and II sequences from independent isolations within the same donor implicates a novel mechanism by which specific combinations of HLA class I and II alleles shape the immunopeptidome. Further work identifying neoantigens in our unmatched isolated HLA class I and II peptides will be continued to uncover potential novel targets for future immunotherapies. Citation Format: William R. Miklavcic, Paul Grandgenett, Michael A. Hollingsworth. Exploration of the human leukocyte antigen immunopeptidome in pancreatic adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr B036.