Human DNA methylation trajectories of the postprandial metabolic response to food

Background and objectives: Diet and postprandial markers of metabolic flexibility are major determinants of cardiometabolic health. The postprandial response is highly individualised, and molecular mechanisms underlying this variability are not well characterised. DNA methylation (DNAm) is a key regulator of gene function with potential to modulate the personalised metabolic responses after food intake. Within the JPI-DIMENSION consortium, we characterised human DNAm changes over the postprandial response, and explore their impacts on genes linked to metabolism and disease. Methods: We profiled whole blood DNAm using the Infinium MethylationEPIC Kit in 360 samples from 120 individuals in the PREDICT1 study of postprandial metabolic responses in healthy individuals. Samples were profiled at baseline and after food intake, comparing changes in DNAm from fasting state (t0) to peak glycemia (t30m), and to peak lipemia (t4h). Changes specific to peak lipemia were also explored in the baseline postprandial phase of the CORDIOPREV study, a randomized clinical trial involving patients with coronary disease. We profiled DNAm in 300 samples from 150 CORDIOPREV individuals at fasting and 4h. Changes in DNAm trajectory postprandially were tested in both samples at >756,000 CpG-sites genome-wide, using linear mixed-effects models adjusted for sex, age, BMI, smoking, blood cell composition, technical covariates, and family structure. Results are presented after multiple testing correction (FDR=10%). Results: Overall, 19 differentially methylated positions (DMPs) showed postprandial trajectory changes in PREDICT1. The peak signal was in the CAMTA1ge ne, with decreased methylation at 30m after test meal (cg23021268; b=0.311±0.050, p=2.40E-08). PREDICT1 and CORVIOPREV meta-analysis identified 125 DMPs across 94 genes that changed postprandially at 4h, including in the cholesterol efflux gene ABCG1(cg27518648: b=0.078±0.016, p=1.11E-06). Pathway analysis identified enrichment in the PKA pathway, as well as in pathways of cardiac and metabolic signalling. Ongoing analysis include sex-specific components and transcriptomic profiling of 50 individuals of the PREDICT1 sample which link specific expression changes to DNAm trajectories. Conclusions: In summary, we observe significant changes in DNAm levels at target genomic regions postprandially. Signals include both changes at peak glycemia and peak lipemia, and target metabolically relevant genes, with insights towards mechani