Single Cell Dissection of Epigenetic and Tumor Ecosystem Dynamics During Pancreatic Cancer Progression

点击翻译按钮获取中文摘要

Project Summary Previous studies of Pancreatic Ductal Adenocarcinoma (PDAC) have failed to inform treatment strategies providing sizeable improvements in patient outcomes, and as such this disease is predicted to be the second leading cause of cancer –related death by the year 2020. A major reason for this shortcoming stems from the extensive cellular heterogeneity that arises during PDAC tumorigenesis, which is largely ignored by bulk genomic studies. I hypothesize that a single cell dissection of the epithelial compartment and tumor ecosystem along disease progression will reveal novel druggable targets and clarify the specific cellular drivers of the disease. I further hypothesize that integration of transcriptional and epigenetic data within computational frameworks will improve the prospects of target detection, as initial evidence suggests a strong impact of epigenetic dysregulation on tumorigenesis. To this end, the proposed project leverages single cell transcriptomic (scRNA-seq) and bulk chromatin accessibility measurements (ATAC-seq) collected in collaboration with the Scott Lowe lab from genetically engineered mice modeling PDAC progression from the moment of initiation through metastasis. For my doctoral research, I propose to develop and apply novel computational methodology to integrate scRNA-seq and ATAC-seq to infer cell type –specific regulatory programs in subpopulations of PDAC, such that we may identify dysregulated mechanisms comparing to normal pancreas epithelium (Aim 1). I then propose to model dynamics of phenotypic shifts over the time course of PDAC progression with a novel method to orient “velocity” of cellular states, again drawing information from both epigenetic and transcriptomic data (Aim 2). This latter aim will allow identification of potential stem cell populations and the phenotypes they give rise to, thus providing a basis for targeting populations driving recurrence or resistance to treatment. In summary, our proposed approach to studying regulation in cancer will provide predictions which are unobtainable with existing methods based on either bulk data or single cell data alone, and which are well-poised to uncover cancer regulators in particular phenotypic niches.