Oral Presentation 9th GeneMappers Conference 2012

The impact of DNA methylation on variation in triglyceride and cholesterol levels (#1)

Melanie A Carless 1 , Hemant Kulkarni 1 , David L Rainwater 1 , Anthony G Comuzzie 1 , Michael C Mahaney 1 , John Blangero 1
  1. Texas Biomedical Research Institute, San Antonio, Texas, United States

Lipid imbalances are a critical component of obesity and metabolic syndrome. High levels of circulating triglycerides and low-density lipoprotein cholesterol (LDLC), as well as low levels of high-density lipoprotein cholesterol (HDLC) are strong risk factors for cardiovascular disease, the leading cause of death worldwide. Known genetic effects explain only 10-15% of the variance in these highly heritable traits and as such we set out to identify epigenetic factors that might contribute to variance in lipid levels. Using Illumina HumanMethylation450 BeadChips, we generated genome-wide methylation profiles for 1,200 Mexican American individuals from the San Antonio Family Heart Study.  Preliminary analysis of 404 individuals utilized variance-components polygenic models to estimate heritability and sporadic models to investigate association of quantitative DNA methylation levels at 482,419 CpG sites with triglyceride and HDLC concentrations. 5,016 CpG (~1%) sites were heritable, after Bonferroni correction for multiple tests (p<1.04x10-7). We identified several suggestively significant associations between serum triglyceride and HDLC concentrations and DNA methylation. Most notably, STARD13, a gene previously implicated in lipid variation and metabolic syndrome, was associated with both triglyceride (p=6.35x10-6) and HDLC (p=2.32x10-3) concentrations. Similarly, GULP1, a gene implicated in cellular lipid homeostasis, was associated with triglycerides (p=1.56x10-5) and HDLC (p=2.88x10-3). The most significant association seen for HDLC was with methylation within RBPMS (p=6.70x10-7), an RNA binding protein recently associated with heart rate response to exercise training. RBPMS was also associated with triglycerides (p=9.83x10-3). In each case, the direction of effect was opposite for the two traits, as would be expected given their functions. These results suggest that DNA methylation may play an important and overlapping role in lipid variation, possibly contributing to the heritability of lipid traits. We are analyzing additional samples and phenotypes to further investigate the role of DNA methylation in lipid variation, which may have significant implications for cardiovascular disease.