Genomic Analyses of Endometrial Cancer

National Human Genome Research Institute

Investigators

Linked publications, trials & patents

Abstract

Summary Uterine corpus cancer is the 6th leading cause of cancer death amongst American women. The American Cancer Society predicts that in 2021 there will be 66,570 newly diagnosed cases of uterine corpus cancer and 12,940 deaths from this disease in the United States. The vast majority (98%) of these cancers are endometrial carcinomas (ECs), which arise from the inner epithelial lining (the endometrium) of the uterus. Although the prognosis for many women with newly diagnosed EC is good, certain disease subgroups including high-grade endometrioid ECs, late-stage endometrioid ECs, and aggressive histological subtypes such as serous EC and uterine carcinosarcoma have generally unfavorable outcomes (recently reviewed in: Urick & Bell, Nature Reviews Cancer, 2019; and Bell & Ellenson, Annual Review of Pathology: Mechanisms of Disease, 2019). During the last decade, this laboratory has successfully used genomic tools to identify somatic mutations in clinically aggressive forms of EC, to home in on potential pathogenic driver mutations that contribute to tumor development. In the current reporting period, we prepared a manuscript reporting our identification of KLF3 and PAX6 as statistically significantly mutated genes in late-stage but not early-stage endometrioid EC (medRxiv preprint: https://medrxiv.org/cgi/content/short/2021.04.26.21256125v1). These findings are hypothesis-generating, raising the possibility that KLF3 and PAX6 mutations may contribute to endometrial tumor progression. The focus of the laboratory has now shifted to determine: 1) how mutations in candidate driver genes impact the function of the encoded protein, and 2) whether mutant proteins can be leveraged as therapeutic targets in EC cells. This report describes ongoing projects in the Bell laboratory, which specifically aim to determine the functional consequences and potential druggability of somatic mutations in the FBXW7, ALK2, and ALK5 genes and related pathways in EC. Determining the functional consequences and potential druggability of FBXW7 mutations in EC: The protein encoded by FBXW7 is a tumor suppressor and a critical component of a ubiquitin ligase complex that degrades numerous substrate proteins involved in tumorigenesis. We previously discovered frequent and recurrent somatic mutations of FBXW7 in ECs (Nature Genetics 2012). Moreover, we found that FBXW7 mutations are particularly frequent in serous EC. We thus hypothesized that recurrent FBXW7 mutations disrupt the proper function of the encoded protein in serous EC. In 2018, we reported novel functional effects of recurrent FBXW7 mutations in serous EC (Urick and Bell, Molecular Carcinogenesis, 2018). Specifically, we showed that recurrent mutations in FBXW7 cause increased levels of phosphorylated Cyclin E1, SRC-3, c-MYC, Rictor, GSK3, P70S6, and AKT proteins in serous EC cell lines. By performing drug sensitivity assays in a serous EC cell-line that was CRISPR/Cas9-edited to knock in a series of FBXW7 mutations, we found that FBXW7-mutant serous EC cells are more sensitive to two drugs, SI-2 (a SRC3 inhibitor-2) and dinaciclib (a cyclin dependent kinase CDK inhibitor), than parental unedited cells (Urick and Bell, Molecular Carcinogenesis, 2018). Collectively these studies shed novel insights into the functional consequences of FBXW7 mutations in serous EC. Many proteins that are regulated by FBXW7 are transcription factors that in turn regulate the expression of other genes. We thus hypothesize that FBXW7 mutations impact the levels of a multitude of proteins within the cell. Therefore, we sought to determine the repertoire of protein alterations that result from mutation of FBXW7 in EC cells. In the previous reporting period, we systematically determined the global proteomic and phosphoproteomic changes in serous EC cells CRISPR/Cas9 edited to knock in FBXW7 mutations, as compared to the parental FBXW7-nonmutated cells. This approach yielded new insights into proteomic changes resulting from FBXW7 mutations in serous EC cells including increased levels of the PADI2 (peptidyl arginine deiminase 2) protein, which is potentially druggable (Urick and Bell, Cancer Medicine, 2020). We also demonstrated an association between FBXW7 mutation and increased PADI2 levels in endometrioid EC cells, which raises the possibility that FBXW7 mutations may have similar effects in divergent histotypes of EC. In the current reporting period, we extended our proteomic analyses to HEC-50B, a high-grade endometrioid EC cell line and three derivative FBXW7 mutation knock-in lines (R465C, R479Q, R505C) generated by CRISPR/Cas9 editing. We identified 278 proteins and 1304 phosphoproteins with significantly (P<0.05) different levels in at least one FBXW7 mutation knock-in cell line compared to parental cells (Urick et al, Cancer 2021, 127(16):2905-2915. doi: 10.1002/cncr.33567). L1CAM (L1 cell adhesion molecule) and TGM2 (transglutaminase 2) levels were significantly altered in all FBXW7-mutant HEC-50B cell lines. Immunoblotting showed increased levels of L1CAM and TGM2 in FBXW7-mutated versus FBXW7-wildtype HEC50B cells. This association held true in grade 2 endometrioid EC cell lines whereas in serous EC cell lines FBXW7 mutations were, for the most part, associated with decreased L1CAM and TGM2 levels. Our findings delineated proteomic changes resulting from FBXW7 mutations in a high-grade endometrioid EC cell line and provided the first direct evidence that FBXW7 mutation affects levels of L1CAM, a known clinically relevant biomarker in EC, and of TGM2, a druggable protein. Determining the functional consequences and potential druggability of mutations in ALK2 and ALK5 in EC: ALK2 and ALK5 are two members of the ALK serine-threonine kinase family. Both ALK2 and ALK5 are critical regulators of transcription via both SMAD-dependent and SMAD-independent pathways. ALK2 mutations occur in 5% of non-ultramutated endometrial tumors and are preferentially located in the TGFbeta/GS and kinase domains of the protein. The ALK2-R206H somatic hotspot mutation is recurrent in EC and is identical to a germline pathogenic variant that causes Fibrodysplasia Ossificans Progressiva (FOP) due to a gain-of function effect. Moreover, several non-recurrent ALK2 mutations in EC are identical to germline pathogenic variants that cause variant-FOP. Based on these collective observations we hypothesize that somatic ALK2 mutations in EC are gain-of-function mutants that disrupt SMAD-dependent and/or SMAD-independent signaling and are potentially druggable. Functional studies designed to test these hypotheses are ongoing. ALK5 mutations occur in 3% of human ECs, but their functional effects are currently unknown. Based on in silico predictions and structural modeling results, we hypothesize that a subset of ALK5 mutations found in human endometrial tumors perturb signal transduction via the TGFbeta-ALK5-SMAD2/3 pathway. To test this hypothesis, we are assessing the biochemical and cellular consequences of ALK5 mutants in EC. If mutations prove to be activating, we will conduct follow-up studies to determine whether such mutants confer increased sensitivity to small molecule inhibitors that target ALK5.