Abstract P099: A Methylome- and Transcriptome-Wide Study of Dietary Fructose Intake in Humans
Background: Animal model studies reported robust associations between fructose intake, a known determinant of metabolic health across species, and DNA methylation/ gene expression patterns. To date, these associations have not been comprehensively investigated in humans.
Methods: Using DNA methylation data on ~470,000 cytosine-phosphate-guanine (CpG) sites in the Genetics of Lipid Lowering Drugs and Diet Network (GOLDN, n=991) quantified in CD4+ T-cells, we have interrogated the cross-sectional relationships between fructose intake (ascertained using a validated food frequency questionnaire) and epigenome-wide methylation. We fit linear mixed models adjusted for age, sex, study site, technical covariates (fixed effects) and family relatedness (random effect). We have supplemented the methylation analyses with a transcriptome-wide scan on 98 unrelated GOLDN participants with available RNASeq data in whole blood. For the RNASeq data analysis, we used negative binomial models of read counts as a function of fructose intake, adjusted for age, sex, study site, technical covariates (including differential white blood cell counts).
Results and Conclusions: The top loci differentially methylated with respect to fructose intake mapped to SNED1 (regression coefficient +/- SE= 0.002 +/- 0.0005, P=5x10-6) and FARSB (regression coefficient +/- SE= 0.002 +/- 0.0004, P=5x10-6) on chromosome 2. SNED1 is an insulin responsive transcription factor that has previously been shown to be downregulated in the setting of metabolic syndrome. FARSB encodes an aminoacyl-tRNA synethetase implicated in cell growth. Additionally, two genes reached transcriptome-wide significance in the RNASeq analysis: MTATP8P1 (P=1x10-7) and PDXDC1 (P=6x10-7). Sequence variation in PDXDC1 (pyridoxal-dependent decarboxylase isoform) has previously been linked to lipid metabolism, while MTATP8P1 is a pseudogene for a mitochondrially encoded ATP synthase. Upon future replication and functional validation, our preliminary findings offer promising insights into the link between fructose intake and metabolic health.
Author Disclosures: S. Aslibekyan: None. A.N. Do: None. M.R. Irvin: None. B.A. Hidalgo: None. D. Zhi: None. H.K. Tiwari: None. D.M. Absher: None. J.M. Ordovas: None. D.K. Arnett: None.
- © 2017 by American Heart Association, Inc.