Development of CRISPR-Based Screening Methods to Identify Cis-Regulatory Elements that Control Complex Cellular Phenotypes

Abstract

Genome-wide association studies have identified thousands of DNA variants associated with specific phenotypes, yet it remains unknown which variants are causal. Additionally, more than 90 percent of common variants occur in noncoding genomic regions, further complicating efforts to predict their function. Large consortia efforts have leveraged functional genomics assays to characterize the genome-wide and epigenome-wide features of noncoding regions and common candidate cis-regulatory elements. However, these predictions are largely based on DNA sequence or correlation of epigenome marks alone, and do not provide information for which genes and broader regulatory networks are controlled by these candidate cis-regulatory elements. Advancements in CRISPR/Cas9-based genome and epigenome editing tools and multiplexed screening assays have enabled systematic perturbation of candidate cis-regulatory elements. We first sought to establish principles for performing noncoding CRISPR screens. We next applied these principles to investigate a subclass of cis-regulatory elements that respond to mechanical stimuli. Finally, we combined noncoding CRISPR screening approaches with single cell transcriptome profiling to clarify the regulatory landscape in diverse cell types in the Major Histocompatibility (MHC) Locus, one of the most complex regions of the human genome. Through these efforts, we establish guidelines for noncoding CRISPR screen design, execution, and analysis, identify mechanosensitive cis-regulatory elements and their role in complex cellular processes, and reveal cell-type specific and shared regulatory mechanisms governing gene expression in the MHC locus. Collectively, these studies provide experimental and computational frameworks for future investigation of cis-regulatory element function and will enable further dissection of variant-gene-phenotype relationships.