Abstract Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive cancer with poor survival rates. Towards understanding the underlying molecular alterations that drive the PDAC oncogenesis, we conducted the first comprehensive proteogenomic analysis of 76 pancreatic tumors and 44 adjacent normal tissues (NATs). Proteomics, whole-genome sequencing, whole-exome sequencing, methylation, and RNA-seq were performed on these tissues to facilitate an integrated proteogenomic analysis and determine the impact of genomic alterations on protein expression. Genomic analysis showed that 89% of tumor samples had a detectable KRAS hotspot mutation, close to the mutation rate reported by TCGA. TP53 (56%), SMAD4 (18%), and CDKN2A (17%) rates were lower, but also close to those in the TCGA cohort, suggesting adequate tumor purity percentages in most of our samples. Additionally, we applied several deconvolution methods, including ESTIMATE and EDec, to estimate tumor purity and cell type composition using transcriptomics and DNA methylation data. These purity estimates correlated with KRAS variant allele frequency (VAF). RNA-based subtyping identified a balanced collection of subtypes consistent with previous reports in PDAC. However, gene expression and protein abundance were not correlated with the mean Spearman correlation of 0.23. Highly correlated pathways included pancreatic secretion, protein digestion and absorption, and allograft rejection, while pathways with low correlation encompassed oxidative phosphorylation and cytoskeleton organization during mitosis. Interestingly, mutations did not have a strong cis impact on the overall proteome level. However, we observed significant increases in BAX and DNMT1 protein levels associated with TP53 mutations. Tumor-normal analyses revealed tumor-specific gene and protein expression, including EMT, cell adhesion, and keratin genes. Proteomic analysis identified several druggable events in ERBB2, SYK, PAK1, PDPK1, and EPHA2. Using cell surface marker gene expression, we identified and classified our samples into four immune subtypes: NK- excluded, NK+ excluded, immune desert, and inflamed. PLEK, WAS, CD48, ALDOA, and ASPH had highly differential protein levels in inflamed tumors, among others. We further identified that protein levels of INTS14, SULF2, and PTPMT1, among others, were associated with patient survival, demonstrating the prognostic value of proteomics over RNA alone. Additionally, we are characterizing an additional 81 pancreatic tumors and 58 NATs using our deep proteogenomic analysis and post-translational modifications, aiming to verify the preliminary findings and to investigate the roles of cell signaling network and cell surface receptors on PDAC oncogenesis. This integrated proteogenomic characterization of PDAC will be a valuable resource for the community, paving the way for the discovery of novel therapeutic targets. Citation Format: Kyung Cho Cho, Ralph Hruban, Hui Zhang, Chen Huang, Bing Zhang, Daniel Cui Zhou, Li Ding, Emily Boja, Investigators from the Clinical Proteomic Tumor Analysis Consortium. Integrated proteogenomic characterization of pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2019 Sept 6-9; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2019;79(24 Suppl):Abstract nr A60.
Read full abstract