High-throughput Screening and CRISPR-Cas9 Modeling of Causal Lipid-associated Expression Quantitative Trait Locus Variants

Avanthi Raghavan,Wang Xf,Peter Rogov,Wang L,Xiaolan Zhang,Tarjei S. Mikkelsen,Kiran Musunuru

DOI: https://doi.org/10.1101/056820

2016-01-01

bioRxiv

Abstract:Genome-wide association studies have identified a number of novel genetic loci linked to serum cholesterol and triglyceride levels. The causal DNA variants at these loci and the mechanisms by which they influence phenotype and disease risk remain largely unexplored. Expression quantitative trait locus analyses of patient liver and fat biopsies indicate that many lipid-associated variants influence gene expression in a cis-regulatory manner. However, linkage disequilibrium among neighboring SNPs at a genome-wide association study-implicated locus makes it challenging to pinpoint the actual variant underlying an association signal. We used a methodological framework for causal variant discovery that involves high-throughput identification of putative disease-causal loci through a functional reporter-based screen, the massively parallel reporter assay, followed by validation of prioritized variants in genome-edited human pluripotent stem cell models generated with CRISPR-Cas9. We complemented the stem cell models with CRISPR interference experiments in vitro and in knock-in mice in vivo. We provide validation for two high-priority SNPs, rs2277862 and rs10889356, being causal for lipid-associated expression quantitative trait loci. We also highlight the challenges inherent in modeling common genetic variation with these experimental approaches.