PROJECT SUMMARY / ABSTRACT Epithelial ovarian cancer (OC) is the deadliest gynecological cancer in the US. It consists of several histotypes, each biologically distinct with different clinical challenges. Clear cell OC (CCOC) represents a much understudied histotype marked by low response rates to standard chemotherapies and a lack of effective therapeutic options. Endometrioid OC (ENOC) is a closely related histotype. CCOC and ENOC share a common tissue of origin (endometriosis) and key genetic mutations but differ markedly in their clinical behavior. Our recently published study showed that most of these mutations were detectable in untransformed endometriosis, suggesting further distinct mechanisms of malignant transformation. We seek to delineate the contributions of altered epigenetics and transcriptional control in this process. The most prevalent histotype, high grade serous OC (HGSOC) harbors epigenetic inactivation of key tumor suppressor genes BRCA1 and RAD51C in the homologous DNA repair (HR) pathway, a pathway highlighted for its therapeutic relevance. These tumors often recur within five years despite initial good response to platinum therapy. Past epigenetic studies of primary OC have usually been conducted without stratification of histotypes, or heavily biased towards the most common histotype HGSOC. Even for HGSOC, epigenetic profiling has been performed with only limited genome coverage, focused primarily on gene promoters. Enhancers have incurred much interest in recent literature as the most dynamically used compartment of the genome, but enhancer studies of primary human OC samples, especially those of distinct histotypes, are generally lacking. This is attributable in part to limitations of existing technology. To address this knowledge gap, we propose to implement an innovative cost-effective tool, compatible with primary human samples, to profile enhancers and other regulatory elements using a targeted technology that jointly profiles DNA methylation and nucleosome occupancy (Target-NOMe Seq). We will also develop the associated bioinformatic pipeline needed to apply this technology (Aim 1). We will use Target-NOMe Seq and transcriptome profiling to analyze 300 bulk OC tumor samples, as well as microdissected tumor and supportive stromal compartments on a subset of samples (Aim 2). With this rich dataset we hope to address the research and clinical questions described above (Aim 3). We will use enhancer and promoter epigenetic states, and a new category of non-coding RNA - enhancer RNA (eRNA), as well as the expression levels of transcription factors and candidate target genes to define transcriptional regulatory networks. By analyzing which networks are altered in the different histotypes, we will gain a better understanding of the distinct molecular makeup of each of these histotypes. The relatively high sequencing depth of our focused Target-NOMe Seq technology will also allow us to assess intertumor and subclonal hete...