Discovering the spectrum of genetic architecture influencing coronary artery calcification

Coronary artery calcification (CAC) is directly correlated to the amount of atherosclerotic plaque with atherosclerosis being a major cause of coronary heart disease. The development of CAC is influenced by both genetic and non-genetic risk factors. This work tested for the genetic variants influencing the extent of CAC via three approaches including the interaction with an environmental variable, smoking, as well as testing for common and rare variants. Three sources of genetic data were analyzed in European American participants from the Genetic Epidemiology Network of Arteriopathy (GENOA). The three aims of the study were 1) identify gene by smoking (GxS) interactions associated with the extent of CAC in sibships (N=915) in Genome-Wide Association studies in GENOA and attempt replication in the Framingham Heart Study (N=1025), 2) determine the association of CAC with the oxidative stress pathway by testing common and rare variants in 40 candidate genes and 5 gene families obtained by targeted sequencing (N=270), and 3) identify novel and known associations with CAC in the extreme tails of the distribution (N=228) in common and rare variants of the exome chip array. Analysis of the GxS GWAS utilized a generalized estimating equations (GEE) model to account for the correlation within sibships with interaction tested via cigarette smoking status stratification followed by a t test of the differences between strata. Replication was attempted by identifying gene concordance between suggested genome-wide significant discovery SNPs of GENOA and followed up within a +/- 250 kb window in the Framingham sample, then corrected for multiple testing by the number of linkage disequilibrium blocks within a region. To test associations with genes in the oxidative stress pathway, a suite of rare variant tests were conducted to determine associations of particular gene families as well as if several rare variant tests consistently provided similar results in candidate genes. Finally, participants selected from the extreme tails of the CAC distribution were analyzed with SCORE-Seq for common and rare variants available on the exome chip. Results from GxS GWAS interaction replicated genes included TBC1D4 (p=6.9x10-5) and ADAMTS9 (p=7.1x10 -5). SNPs/genes associated with oxidative stress were a common variant in FBG (p=6.43x10-4) and gene-based tests of rare variants were consistent for NOS1AP (Madsen-Browning Weighted Sum Test p=9.50x10-5) of the nitric oxide synthase family as well as the peroxidase genes (Madsen-Browning Weighted Sum Test p=1.02x10-4). The exome chip results yielded a common SNP in PNPLA2 (p=5.06x10-6) and gene-based tests of rare variants yielded ATP6AP1 (EREC test p=1.53x10 -3). Genetic differences in smokers and nonsmokers may help target associations with the extent of CAC unique to a subgroup. In addition, analyses of rare variants may provide novel insights into disease mechanism. Genes involved in inflammatory pathways were associated with CAC and as smoking induces inflammation, it follows that CAC development has inflammatory mediators. To establish causation, future functional studies would be necessary. Genes of the NF-κB axis influencing the OPG/RANKL signaling pathway of bone regulation are promising drug targets for future treatment to reduce CAC, especially in targeted groups with increased risk of cardiovascular disease, type 2 diabetes, rheumatoid arthritis, and chronic kidney disease