Introduction: Coronary artery calcification (CAC) is a measure of subclinical atherosclerosis that strongly predicts clinical coronary artery disease events. Previous GWAS have identified 4 loci associated with CAC. We used the expanded genomic coverage provided by whole genome sequencing (WGS) data generated by the Trans-Omics for Precision Medicine (TOPMed) program to explore genetic associations with CAC.
Methods: The study sample consisted of 22,404 participants (52% women, mean age 58 years) with WGS and CAC data in TOPMed Freeze 6, including 12,185 European, 8,606 African, 1,010 Hispanic, and 603 East Asian ancestry participants from 10 studies. We used GENESIS to perform pooled single variant analyses adjusting for age, sex, study, ancestry group, 11 ancestry informative principal components, and the kinship matrix. We analyzed rank-normalized residuals of log(CAC+1) as a quantitative trait and CAC>100 as a binary trait. Associations with gene expression levels were examined using GTEx V7.
Results: Variants at 6 loci were significantly associated (P<5E-9) with log(CAC+1), including 4 known loci (9p21, PHACTR1, APOE, and APOB), and 2 novel loci (ARSE, P=2.7E-14; and DCAF4L2, P=3.2E-9). Known loci 9p21, APOE, and APOB were also associated with CAC>100, as was novel locus DCAF4L2. The index variant at ARSE was suggestively associated with CAC>100 (P=8.0E-7). rs5982944 in ARSE is common among African ancestry participants, but rare in European ancestry participants. The allele associated with increased CAC is also associated with increased expression of ARSE, which is located on the X chromosome and is involved in bone mineralization. rs13268080 near DCAF4L2 is common across ancestry groups, and the allele associated with increased CAC is also associated with decreased expression of MMP16, which encodes for a protein involved in the matrix remodeling of blood vessels through interactions with fibronectin and collagens.
Conclusions: Using WGS we identified two novel loci for CAC that provide valuable insights into the etiology of CAC. We plan to perform gene-based aggregate analyses in order to increase the power to detect rare variant associations, and to replicate associations at novel loci in independent samples.

Carotid intima-media thickness (CIMT) and presence of carotid plaque, evaluated non-invasively using ultrasound, are markers of atherosclerosis and useful in detecting subclinical cardiovascular disease. The largest published genome-wide association study (GWAS) found 11 CIMT and 5 plaque loci. Utilizing whole genome sequencing (WGS) generated by the National Heart, Lung, and Blood Institute’s Trans-Omics for Precision Medicine (TOPMed) program, we sought to investigate potentially novel genetic associations with CIMT and carotid plaque that may have been missed by GWAS based on genotyping arrays. WGS data was available for 22,443 participants from 9 studies who had CIMT measured and 20,254 participants from 7 studies who had carotid plaque evaluated. Participants were of European, African-American, Hispanic, or East Asian ancestry. CIMT was transformed through inverse rank-based normalization. Linear and logistic mixed models accounting for relatedness, implemented in SAIGE, were used to perform single variant tests for CIMT and plaque, respectively. Analyses were adjusted for age, sex, study, ancestry group, and principal components and limited to variants with a minor allele frequency (MAF) >0.1%. Using a significance threshold of P<5E-8, we identified two loci (APOE and CCBE1) associated with CIMT and one locus (HNF4G) with plaque. While the APOE locus has been previously linked to CIMT, CCBE1 and HNF4G represent potential novel associations. CCBE1 plays a role in extracellular matrix remodeling and is important to the formation of lymphatic vessels. HNF4G is a part of a family of transcription factors known to play important roles in development, homeostasis, and metabolism. The index variant for CCBE1 (rs747874008) was rare in all ancestry groups (MAF=0.12%). In contrast, the index variant for HNF4G (rs1817002) was intergenic and common across all ancestry groups (MAF=26.4%). Population specific analyses suggest associations with both variants are primarily driven by European participants. Through our analysis, we confirmed an association between APOE and CIMT, and identified two potentially novel loci for CIMT and plaque. The CCBE1 variant may have been missed by previous GWAS due to its low MAF. However, the common HNF4G variant was evaluated in prior array-based GWAS and was not associated despite being well imputed. Replication is required in order to confirm or refute the association of these novel loci with carotid atherosclerosis.