Habitual Aerobic Exercise Preserves Selective “Modules” of DNA Methylation Loci During Aging: Implications for Maintaining Vascular Endothelial Function

• Published in: The FASEB journal 2016-04, Vol.30 (S1)
• Main Authors: Martens, Christopher R, Lubieniecki, Kara, McNamara, Molly, Bohr, Adam D, McQueen, Matthew B, Seals, Douglas R
• Format: Article
• Language: English
• ISSN: 0892-6638; 1530-6860
• DOI: 10.1096/fasebj.30.1_supplement.lb680

Abstract only Habitual aerobic exercise (AEx) training reduces cardiovascular disease risk during aging, in part, by preventing age‐related impairments in vascular endothelial function; however, the underlying molecular mechanisms are not completely understood. DNA methylation is altered with aging and is modifiable by AEx training; therefore, we sought to determine if habitual AEx prevents age‐related changes in DNA methylation, and whether these changes are associated with preserved vascular endothelial function as assessed by brachial artery flow‐mediated dilation (FMD). 12 young sedentary (Y‐SED; 22 ± 1 yrs; FMD = 8.3 ± 0.7%; VO 2 max = 44.6 ± 2.6 ml/kg/min), 15 older sedentary (O‐SED; 60 ± 1 yrs; FMD = 5.0 ± 0.7%; VO 2 max = 29.0 ± 1.2 ml/kg/min) and 11 older AEx trained (O‐AEx; 58 ± 2 yrs; FMD = 7.7 ± 0.7%; VO 2 max = 44.5 ± 2.2 ml/kg/min) individuals were included in the analysis. DNA was isolated from peripheral blood mononuclear cells (PBMCs) and used to determine genome‐wide methylation of 485,512 individual methylation loci using the Illumina™ Human 450K Methylation Array. After normalization and pre‐processing of the raw methylation data, 450,855 autosomal CpG methylation loci were categorized into distinct co‐expression “modules” using Weighted Gene Correlation Network Analysis (WGCNA). Linear regression methods were used to identify modules that were differentially methylated by aging and/or AEx training, with the Y‐SED group set as a reference control. Each module was assigned an arbitrary color for naming purposes. Of the 54 modules identified, significant aging‐related differences were observed between O‐SED vs. Y‐SED groups in three modules (modules: brown [n=65,338 loci], pink [n=6,831 loci], grey60 [n=680 loci]; p < 0.0001). Two of these “aging modules” (brown, grey60) were not different between O‐AEx vs. Y‐SED suggesting that AEx preserved the methylation status of loci within these modules. Significant differences were observed between O‐AEx vs. Y‐SED in two of the 54 modules (modules: green [n = 29,976 loci], salmon [n = 1,898 loci]; p < 0.001) in the absence of any differences between O‐SED vs. Y‐SED groups, suggesting that AEx also exerts separate changes in DNA methylation that are independent of aging. In a secondary regression analysis, six of the 54 modules were associated with FMD, independent of group assignment. Three of these “vascular modules” were also the same “aging modules” described in the primary aging analysis (brow