Project 1: Functional characterization of candidate cancer genes identified by genomic analysis of serous endometrial tumors During the previous reporting period, we published one of the first whole exome sequencing studies of serous endometrial tumors (Le Gallo et al., Nature Genetics 2012;44:1310-5). The major finding of that study was our discovery that in serous endometrial tumors the FBXW7, CHD4, and SPOP genes are mutated at statistically significantly higher rates than the background mutation rate. This observation strongly suggests that somatic mutations in FBXW7, CHD4, and SPOP are pathogenic driver mutations that confer a selective advantage in serous endometrial tumorigenesis. However, proof of this hypothesis requires functional studies of the mutant proteins. Towards the end of the last reporting period, we initiated functional studies to determine how the mutations in candidate driver genes affect the encoded proteins. In ongoing and planned studies that will extend into the next reporting period, we continue to characterize the function of the significantly mutated genes using biochemical and cell-based approaches. We will focus our efforts on recurrent hotspots mutations that are present in tumors of multiple unrelated patients, and are therefore likely to be pathogenic driver mutations that contribute to endometrial tumorigenesis. Project 2. Identification of somatic mutations that underlie clear cell endometrial cancer During the previous reporting period, we subjected DNAs from clear cell endometrial tumors and from the matched normal tissues to whole exome sequencing at the NIH Intramural Sequencing Center. During the current reporting period, we analyzed the exome sequencing data by: 1. Filtering germline variants (present in both tumor and normal DNAs) from somatic variants (present exclusively in the tumor DNA). 2. Subjecting the somatic variants to orthogonal validation using Sanger sequencing to distinguish true somatic mutations from false-positive calls. 3. Prioiritizing a subset of nonsynonymously, somatically mutated genes, which we have deemed candidate cancer genes, for further analysis in a mutation prevalence screen. The mutation prevalence screen was initiated within the current reporting period and will extend into the next reporting period. In this screen we are sequencing the coding exons of the candidate cancer genes from a larger set of primary clear cell endometrial tumors. Based on the number of mutations identified in the discovery and prevalence screens, will then calculate the mutation rate of the individual genes in clear cell endometrial cancer. If a gene is mutated at a significantly higher rate than the background mutation rate (which we will empirically determine) this provides strong genetic evidence that the mutated gene is likely to be a pathogenic driver gene that confers a selective advantage in clear cell endometrial tumorigenesis. Should these studies reveal a novel, significantly mutated gene, we plan to initiate a limited number of biochemical and cell-based studies to determine the functional consequences of the observed mutations. Project 3. Interrogation of somatic copy number alterations in serous and clear cell endometrial cancer In parallel to the mutational analyses described in Project 1 and 2, we are searching for somatic copy number alterations in serous and clear cell endometrial tumor genomes. In past reporting periods, we used high-density SNP arrays to catalogue somatic genomic copy number alterations (gains and losses) that occur in serous and clear cell endometrial tumors. Within the current reporting period we have used an independent method to search for copy number alterations in these tumors. Specifically, in collaboration with the Mullikin lab (NHGRI), we have used the ExomeCNV statistical methodology to identify copy number alterations within the short sequence reads we generated for serous and clear cell endometrial tumor exomes (Project 1, and Project 2 respectively). In the incoming reporting period we will compare the SNP array and ExomeCNV data to identity concordant regions of copy number gain or loss identified by these two independent methods.
Showing the most recent 10 out of 21 publications
