Abstract P258: Alterations in DNA Methylation of Circadian-related Genes are Associated with Metabolic Traits: A Gene-set Analysis in Monozygotic Twins

Abstract

Background: Circadian disruption adversely affects the metabolic system and increases the risk of metabolic disorders. Epigenetic factors, especially DNA methylation, are essential for the regulation of both circadian clock and metabolic function, but whether and how altered DNA methylation of circadian-related genes affects metabolic traits is poorly understood. Moreover, previous DNA methylation analysis largely focused on individual CpG sites in single genes, but the joint contribution of methylation variation at multiple CpG sites in multiple genes may explain a larger proportion of disease variability compared to single-CpG single-gene analysis. Objective: To assess the joint contribution of 84 CpG sites in six circadian-related genes to interindividual variability in metabolic traits in a well-matched monozygotic (MZ) twin sample. Methods: We analyzed 69 apparently healthy, male-male MZ twin pairs recruited by the Emory Twins Heart Study. All twins were middle-aged (mean age 55) veterans from the Vietnam Era Twin Registry, and were free of self-reported histories of CVD and diabetes at the time of enrollment. Peripheral blood DNA methylation at 84 CpG sites in the promoter regions of six circadian-related genes, including ARNTL, CLOCK, MC4R, PER1, PER2, and PER3, were quantified by bisulfite pyrosequencing. We first examined the association of DNA methylation at individual CpG site in each candidate gene with metabolic phenotypes (waist circumference, systolic blood pressure, HbA1c, HOMA-IR, HDL, LDL, and hsCRP) by regressing the intra-pair difference in each trait on the intra-pair difference in DNA methylation at each CpG site, adjusting for twin age and intra-pair differences in smoking (pack-year), alcohol consumption (g/day), level of physical activity and Beck depression scale. We then conducted promoter-based association analysis by combining P-values of all CpG sites within a gene based on single CpG analysis using a weighted truncated product method (TPM) with regression coefficient as weight. Gene-set analysis was performed by combining P-values of all six genes using TPM. Multiple testing was controlled by false discovery rate. Results: Individual CpG site analysis showed that, although methylation variation at multiple CpG sites in multiple genes was marginally associated with one or more metabolic traits, only a few sites survived multiple testing. Gene-set analysis, however, revealed that methylation variation in all six circadian-related genes was significantly associated with all studied metabolic traits (all P’s≤0.0009). Conclusion: DNA methylation variation in circadian-related genes jointly contributes to metabolic function, independent of genes and early family environment. These findings may shed light on the role of DNA methylation in the circadian control of metabolic functions.