Computational and experimental methods for scalable identification of oncogenic non-coding regions

Project Abstract In this project, we will identify context-specific enhancers, whose activity is essential for either the growth of primary tumors or for evolution of treatment-resistance in metastatic tumors by modulation of target gene expression. We will identify open chromatin sites in metastatic treatment-resistant breast tumors using whole-genome sequencing of cell-free DNA. Cell-free (cfDNA) enzymatic in plasma predominantly originates from nucleosome-protected parts of DNA after processing. We will develop innovative computational methods that will analyze the nucleosomal signal obtained from whole-genome sequencing of cfDNA from plasma of metastatic cancer patients. We will develop scalable methods to create catalogues of `oncogenic' enhancers that are essential for tumor growth, which may or may not be due to DNA sequence variants at these regions. We will decipher the impact of subtype specific enhancer perturbation on tumor initiation (using CRIPSRa) and continued growth (using CRISPRi) in primary breast cancers. We will test the impact of repressing ~15,000 accessible sites in breast cancer on tumor growth in pooled CRISPR screens in multiple cell lines for each major breast cancer subtype. Importantly, we will also study the impact of activation of the same enhancers on tumor initiation in normal (and malignant) breast cell lines, and we will identify `tumor-suppressive' enhancers that when silenced drive tumor growth. We will also identify oncogenic mutations by integration of 4,427 breast cancer whole-genomes and CRISPR base editing screens. We will focus on specific mutations at cis-regulatory elements, CREs (enhancers, promoters and untranslated regions). For the top 5000 putative noncoding drivers, we will first insert precise mutations via base editing and then, for the top 500 with the most dramatic impact on growth, we will use single-cell sequencing coupled with base editing to decipher their impact on gene expression in cis and in trans. We will also examine the interaction between the top noncoding drivers and aromatase inhibitors, which is a therapy prescribed in many breast cancers.