Rare variants in a PCSK9 distal-enhancer as drivers of LDL-cholesterol levels

BACKGROUND. Discoveries of genetic causes of dyslipidemia have significantly advanced our understanding of lipoprotein metabolism and its implications in cardiometabolic diseases. However, a significant number of patients with dyslipidemia still have unidentified genetic causes. Leveraging insights from genome-wide association studies, where most common associated variants are located in non-coding regions, we propose that rare non-coding variants located in genomic regulatory regions may be responsible for dyslipidemia. METHODS. Using whole-genome sequencing data from the UK Biobank (n > 200,000), we performed region-based rare-variant association analyses, focusing on liver specific genomic regulatory regions (assessed by transposase-accessible chromatin assay coupled with sequencing (ATAC-seq)), to assess whether the cumulative burden of rare genetic variants (allele frequency < 0.01) in these genomic regulatory regions was associated with LDL-cholesterol levels. To validate our findings, we used an advanced cellular model of human induced pluripotent stem cells (hiPSCs) stably expressing a CRISPR-interference system. This system enables, when combined with guide-RNAs, to target regions of interest and silence regulated genes. RESULTS. We found that the cumulative burden of rare genetic variants in 18 among 53,039 ATAC-seq peaks in human liver, is significantly associated with LDL-cholesterol. We focus our proof-of-concept study on a peak located ~17 kb upstream PCSK9, strongly associated with LDL-cholesterol (p=5.85e-52), apolipoprotein-B (p=3.82e-49) and PCSK9 plasma levels (p=1.15e-67). We showed that this regulatory region is also active in hiPSCs and hiPSC-derived liver organoids. When transfected with guide-RNAs targeting this region, we showed that CRISPRi-hiPSCs present a hypermethylation of the targeted region which in terms leads to a substantial decrease in PCSK9 gene expression (-48%), intracellular protein (-65%) and secretion (-36%) (p<0.001). CONCLUSIONS. This study identifies a non-coding genetic region upstream PCSK9 where rare variants lead to decreased PCSK9 expression and lowered LDL-cholesterol. Such findings advocate for the genetic screening of non-coding regulatory regions in patients with unidentified genetic causes of extreme dyslipidemia and open novel perspectives for genetic regulatory therapeutic intervention.