Understanding the full spectrum of epigenetic vulnerability in cancer through the delineation of DNA methylation function in gene 3' end

摘要

Abstract DNA methylation abnormalities occur in all genomic contexts throughout the cancer genome, and studies of such aberrant epigenetic patterns have led to seminal discoveries regarding tumor suppressor gene silencing by promoter hypermethylation. While promoter hypermethylation causes transcriptional silencing, the functions of non-promoter DNA methylation are poorly defined. Such lack of knowledge severely limits our ability to contextualize the effects of abnormal DNA methylation on cancer biology and to realize the full potential of epigenetic-based cancer therapy. In this proposal, we propose to investigate the impact of non-promoter DNA methylation on the transcriptome and will focus on studying the functions of DNA methylation near gene 3' ends. Using a pair of isogenic cancer cells (HCT116 and DKO cells) that differ specifically in their ability to maintain DNA methylation, we discovered a robust association between gene 3' end differential DNA methylation and alternative cleavage and polyadenylation (APA) events. Briefly, pre-mRNAs undergo cleavage and polyadenylation as part of normal mRNA 3' end formation, and alternative sites of cleavage and polyadenylation can be utilized to produce transcripts with varying regulatory sequences in the 3' untranslated regions (3' UTRs) or protein isoforms via APA within coding sequences. Previous studies have demonstrated that cancer cells can hijack the APA pathway to skew expression of short 3' UTRs in oncogenes to evade negative regulation, highlighting APA as an important process involved in cancer initiation and progression. By leveraging the Cancer Genome Atlas (TCGA) data, we could also observe the correlation between gene 3' DNA methylation and APA at select loci in cancer patient samples. We hypothesize that differential DNA methylation in gene 3' ends can result in cancer-promoting expression patterns through regulation of APA. In Aim 1, we will define the mechanism of DNA methylation-regulated APA using a combination of computational, biochemical, molecular biology, and genomic approaches. In Aim 2, we will validate our model and in vitro data by mapping polyadenylation site usage in additional cancer cell lines and testing for disease-relevant APA events across different cancer types using publically available RNA-seq and DNA methylation data from the International Cancer Genome Consortium (ICGC). The results of our study will improve overall functional understanding of non-promoter DNA methylation, provide a novel mechanism for APA regulation, and ultimately accelerate discovery of novel targets for cancer management. We also envision that our findings here can have broad impact on our knowledge of how epigenetic regulation shapes the transcriptome in cancer as well as in other human healthy and pathological conditions.