AMPK subunits harbor largely non-overlapping genetic determinants for body fat mass, glucose- and cholesterol metabolism

Abstract Context AMP-activated protein kinase (AMPK) is a heterotrimeric enzyme and central regulator of cellular energy metabolism. The impact of single nucleotide polymorphisms (SNPs) in all seven subunit genes on adiposity, glucose- and lipid metabolism has not been systematically studied yet. Objective To analyze the associations of common SNPs in all AMPK genes, and of different scores thereof, with adiposity, insulin sensitivity, insulin secretion, blood glucose, total-, LDL- and HDL-cholesterol and triglycerides. Study Design and Methods A cohort of 2789 non-diabetic subjects from the Tübingen Family study of type-2 diabetes, metabolically characterized by oral glucose tolerance test and genotyped by genome-wide SNP array was analyzed. Results We identified largely non-overlapping SNP sets across four AMPK genes (PRKAA1, PRKAA2, PRKAG2, PRKAG3) associated with adiposity, insulin sensitivity, insulin secretion, blood glucose, total-/LDL-cholesterol or HDL-cholesterol, respectively. A genetic score of body-fat-increasing alleles revealed per-allele effect sizes on BMI of +0.22 kg/m² (p=2.3·10-7), insulin sensitivity of -0.12·1019 L²/mol² (p=9.9·10-6) and 2-h blood glucose of +0.02 mmol/L (p=0.0048). Similar effects on blood glucose were observed with scores of insulin-sensitivity-reducing, insulin-secretion-reducing and glucose-raising alleles, respectively. A genetic cholesterol score increased total- and LDL-cholesterol by 1.17 mg/dL per allele (p=0.0002 and p=3.2·10-5, respectively), and a genetic HDL score decreased HDL-cholesterol by 0.32 mg/dL per allele (p=9.1·10-6). Conclusions We describe largely non-overlapping genetic determinants in AMPK genes for diabetes-/atherosclerosis-related traits which reflect the metabolic pathways controlled by the enzyme. Formation of trait-specific genetic scores revealed additivity of allele effects, with body-fat-raising alleles reaching a marked effect size.