Modeling the impact of mutations in ubiquitin ligase genes on transcriptional programs in endometrial cancer Abstract: Endometrial cancer is the most common gynecological cancer in the United States and has been the subject of two TCGA studies, a recently published characterization of uterine corpus endometrial carcinomas (ECs) and an ongoing study of uterine carcinosarcomas (UCs), a rare and aggressive subtype. A pan-cancer analysis reveals a strikingly high frequency of mutations of specific ubiquitin pathway genes in uterine cancers: FBXW7, SPOP, and RNF43, all encoding substrate recognition proteins in distinct E3 ubiquitin ligase complexes, and altered by somatic mutations in ~44% of UCs and ~25% of ECs based on TCGA analysis. While these genes have known or proposed tumor suppressor functions in other cancers, they are little studied in endometrial cancer. We have recently developed a novel computational approach for exploiting parallel phosphoproteomics (reverse-phase protein array) and mRNA expression (microarray or RNA-seq) data available for large tumor sets through TCGA to link dysregulation of upstream signaling pathways with altered transcriptional response through the transcriptional circuitry. Importantly, our approach provides a statistically principled framework for interpreting the impact of mutations and copy number events in terms of altered transcription factor and signaling activity. We propose to develop and apply our integrative modeling approach to interpret the role of frequently mutated ubiquitin ligase genes in endometrial cancer. We will pursue a major methodological advance in the computational model by incorporating detailed epigenomic information from appropriate cell line models in order to represent transcription factor binding signals in enhancers. We will also use comparative modeling of two genomically similar tumor types, serous ovarian and serous endometrial carcinomas, to potentially link the higher frequency of FBXW7 mutations in endometrial tumors to chemotherapeutic resistance. More broadly, this work will provide general methodological tools to enable the study of other classes of somatically altered genes across tumor types.
Large-scale cancer genomics studies have catalogued the somatic alterations across tumors in numerous cancers, together with parallel data measuring mRNA expression and now phosphoprotein levels. Here we present a new computational approach to leverage these multiple sources of data in order to interpret the impact of mutations on gene expression programs in tumors. We develop and apply this approach in endometrial cancer to study the role of a number of frequently altered genes that have not been well studied in this malignancy.
| Osmanbeyoglu, Hatice U; Toska, Eneda; Chan, Carmen et al. (2017) Pancancer modelling predicts the context-specific impact of somatic mutations on transcriptional programs. Nat Commun 8:14249 |
