Genomic Analysis of 71 Cardiovascular Proteins in the Framingham Heart Study Identifies Regulatory Networks for Cardiometabolic Risk

Abstract

Aims: Cardiometabolic risk factors often cluster and their coexistence increases cardiovascular disease (CVD) risk. Unraveling molecular mechanisms for the clustering of risk factors and their links to CVD may highlight promising therapeutic approaches. Methods and Results: We performed a multi-stage integrative genomic study of 71 CVD-related proteins in relation to ten CVD risk factors including lipids, adiposity, dysglycemia and blood pressure traits. Leveraging 16,000 protein expression quantitative trait loci (pQTL) with GWAS signals for CVD risk factors, we identified pQTLs on Chromosome 2 (GCKR region), 9 (ABO region), and 12 (SH2B3 region) that colocalized with at least three risk factors. By integrating pQTLs with tissue specific expression quantitative trait loci (eQTLs), we identified complex three-way regulatory relations of proteins with CVD risk factors via expression of transcripts that were distant (trans from the protein coding gene, including the trans-regulation of APOB (protein) in relation to LDL cholesterol (risk factor) through SORT1 (transcript) expression and ANGPTL3 in relation to triglycerides through PCSK7 expression. Using Mendelian randomization, we identified 13 putatively causal proteins for seven risk factors and found three circulating proteins (ANGPTL13, APOB, and NTproBNP) whose levels were associated with the corresponding risk factor when tested cross-sectionally and longitudinally in ~7000 Framingham Heart Study participants. Conclusions: Results of this integrative approach may provide a broader view of network biology underlying cardiometabolic risk and highlight promising therapeutic targets for prevention and treatment of CVD.