Massively parallel profiling of human regulatory variants in endothelial cells

Genome-wide association studies (GWAS-s) have linked thousands of genetic variants with complex diseases and traits. However, most genetic variants identified reside in non-coding regions of the genome making it hard to predict and understand the molecular mechanisms underlying causal alleles. We have previously performed molecular Quantitative Trait Locus (molQTL) analysis for transcription factor binding, chromatin accessibility, and H3K27 acetylation across primary Human Aortic Endothelial Cell (HAEC) samples. Here we expand this analysis to study more than 30,000 genetic variants using the massively parallel reporter assay (MPRA) STARR-Seq in immortalized HAEC cells (teloHAEC). We demonstrate that more than 5,000 variants exhibit differential expression between alleles and identify ETS and AP-1 motifs as the strongest sequence attributes—and cell type-specific chromatin accessibility as the strongest epigenetic attribute—associated with allele-specific regulatory activity. Using interleukin 1-beta (IL1b) as a modulator of cell state and environment, we observe robust evidence for context-specific SNP effects, thereby underscoring the prevalence of GxE effects on noncoding variant function. We integrate results from molQTL mapping and STARR-seq with eQTL data from HAECs and GTEx tissues to fine-map functional non-coding SNPs at GWAS loci for vascular diseases. Overall design: STARR-Seq massively parallel reporter assay (MPRA) to study SNP effects on regulatory region activity in human endothelial cells