DNA hypermethylation of promoter CpG islands is known to be associated with transcriptional silencing in endometrial cancer. We recently identified a unique set of CpG island loci that are hypermethylated in non- recurrent tumors but coordinately hypomethylated (or less methylated) in primary tumors that subsequently recurred. The loci can be highly susceptible to de novo DNA methylation (i.e., a default state) partly attributed to the high expression of DNA methyltransferase 1 (DNMT1) and Polycomb Repressor Complex 2 (PRC2). While these loci, which are implicated in epithelial-mesenchymal transition (EMT), are not usually expressed in normal cells, their chromatin may remain in a bivalent state similar to those observed in stem and progenitor cells. Network-based analysis linking these loci to epidermal growth factor receptor (EGFR) signaling suggests that a low DNA methylation signature and a permissive chromatin state can permit transcriptional activation by this pathway. This was confirmed by our functional studies showing that EGF induced the expression of candidate genes via the epithelial adhesion marker (EpCAM) intracellular domain (EpICD). We hypothesize that the binding of EpICD transcriptional complex at target promoters limits the access of DNMT1/PRC2 to these loci and recruits histone methyltransferases to modify the chromatin into an active state in recurrent tumors. This epigenetic overwriting of the default state is necessary for EMT-mediated progression of endometrial cancer cells.
In Aim 1, we will validate the methylation status of these candidate loci in an independent cohort of endometrial cancer. DNA hypomethylation and over-expression of these loci can be used as biomarkers for predicting risk of endometrial cancer recurrence in patients.
In Aim 2, we will assess whether EGF-induced EMT leads to identification of EpICD target genes delineating novel candidate hypomethylators. Associated mesenchymal characteristics of EGF-induced cells will be examined by molecular means and cellular nanomechanical features indicative of increased invasiveness.
In Aim 3, we will determine whether the EpICD transcription complex is a key factor of this epigenetic reprogramming. EpICD binding to target promoters is expected to diminish the binding of DNMT1/PRC2 and alter combinatorial patterns of histone marks specifying an active chromatin state for EMT-mediate transcription.
Approximately 20% of endometrial cancer patients have identifiable metastases or occult disease that will ultimately recur, and there is an unmet need to identify molecular markers that predict risk for recurrence in these patients. In a preliminary study, we identified a subset of CpG island loci, the hypomethylation of which can be used to stratify recurrent and non-recurrent groups in a patient cohort. The proposed study will validate these potential epigenetic markers, termed hypomethylators, and investigate their regulatory networks associated with epidermal growth factor receptor (EGFR) and epithelial adhesion marker (EpCAM) during tumor progression.
|Hsiao, Tzu-Hung; Chiu, Yu-Chiao; Hsu, Pei-Yin et al. (2016) Differential network analysis reveals the genome-wide landscape of estrogen receptor modulation in hormonal cancers. Sci Rep 6:23035|
|Hsu, Ya-Ting; Osmulski, Pawel; Wang, Yao et al. (2016) EpCAM-Regulated Transcription Exerts Influences on Nanomechanical Properties of Endometrial Cancer Cells That Promote Epithelial-to-Mesenchymal Transition. Cancer Res 76:6171-6182|
|Mao, Zijing; Ma, Chifeng; Huang, Tim H-M et al. (2014) BIMMER: a novel algorithm for detecting differential DNA methylation regions from MBDCap-seq data. BMC Bioinformatics 15 Suppl 12:S6|