Abstract 2706: Effects of germline and somatic mutations on protein expression in tumor and adjacent normal tissues in breast, ovarian, and colorectal tumors

Abstract

Abstract The National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium (CPTAC) provides proteomic data that accurately quantify protein expression in tumors. This work improves upon previous estimates of expression, i.e., RNA-seq and or microarray, because it directly measures protein quantity using mass spectrometry. These data are critical for establishing more accurate estimates of a mutation's impact on tumor cell microenvironment and molecular processes. Here we address the association of both germline and somatic mutations with protein expression in three cancer types: breast cancer (BRCA), ovarian cancer (OV) and colorectal cancer (CO). Additionally, we analyze both primary tumors and adjacent normal pairs that provide insight into tumor etiology. CPTAC2 prospective proteome and phosphoproteome established a dataset of ~450 samples. For analysis we require samples to have reliable protein and phosphoprotein measures. This filtering strategy resulted in a dataset with measurements for 8,000-12,000 proteins and over 30,000 phosphosites for these three cancer types. We first performed analysis on tumors with likely predisposition germline mutations. We defined germline-predisposed samples as tumor samples with germline mutations in one or more of the following DNA repair genes: BRAC1/2, MSH2/6, PMS2. When analyzing samples with and without germline predisposition, an altered protein expression profile was found, albeit less extreme. We observed overexpressed genes in predisposed samples that include C9orf16, PRDX5, SERPINB8, and CMPK1, and found genes to have lower expression that include TULP1, MAEL, KMT2B, and HIST1H1D. Additionally, we performed differential gene expression analysis using samples with adjacent normal tissue biopsies. We examined both the proteome and phosphoproteome levels, comparing tumor vs. adjacent normal samples, with and without germline predisposition mutations. Preliminary analysis for BRCA between tumor and normal samples showed altered protein expression profile in tumors, with about 60% of the genes showing higher or lower protein expression in the tumor. This pattern is recapitulated when restricting our analysis to known cancer driver genes. Genes found to be overexpressed in tumor samples include GNB1, SERPINA1, CDKN2C, IGF1, ERG, AZGP1, and H3F3A, while genes found to have lower expression include PRKDC, DDX5, NUP93, CTCF, ATRX, MYD88, SMARCA4, KDM6A, SF3B1, and MED12. Proteomic/phosphoproteome data deliver reliable results on both cis and transmutational effects on protein expression at both the germline and somatic level. Furthermore, these data provide a glimpse into the tissue microenvironment of adjacent normal tissue and indicate biologic stresses of germline mutations on tissues. Citation Format: Matthew H. Bailey, Daniel C. Zhou, Yige Wu, Matthew A. Wyczalkowski, Liang-Bo Wang, Fernanda Martins Rodrigues, Gordon Mills, Samuel Payne, David Fenyo, Li Ding. Effects of germline and somatic mutations on protein expression in tumor and adjacent normal tissues in breast, ovarian, and colorectal tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2706.