Lipo-proteomic and metabolic fine-mapping of genetic lipid risk loci

Dysregulated blood lipid levels sit at the nexus of cardiometabolic disorders and are major predictors of human cardiovascular health. Using five major lipid traits (HDL-C, LDL-C, non- HDL-C, TC, and TG), a recent genome-wide association study (GWAS) in 1.65 million individuals identified and fine-mapped over 1,000 genetic loci that may be implicated in the etiology of dyslipidemia and related cardiovascular disease. However, a deeper functional understanding of these associations is needed to assess their therapeutic potential as druggable targets. Here we leveraged data from over 98,000 participants of UK Biobank for deep molecular phenotypic refinement and identified 225 lipid risk variants that associated with 168 distinct NMR-derived lipoprotein and metabolic traits, doubling the number of loci that are discoverable when using the five "classical" lipid traits alone. Hypothesis-free testing of >14,000 ratios between metabolite pairs significantly increased statistical power (p-gain) at 72% of the loci, revealing distinct groups of variants with functionally matching NMR-ratios that affect lipoprotein metabolism, transport, and remodeling (LPmtr). We demonstrate how these NMR- trait and -ratio associations can be used in the functional interpretation of the respective lipid risk loci and their evaluation as potential drug targets. Our study reveals a comprehensive picture of the biological roles that the different genetic variants play in LPmtr and supports the emerging view that lipoprotein size and core composition are essential for the understanding, prevention and treatment of lipid-related disorders, beyond the "classical" five major lipid traits currently used in clinical practice.