Skip to main content

Table 1 Summary of main studies of DNA methylation in obesity, T2D, and RYGB

From: Tissue-specific methylation profile in obese patients with type 2 diabetes before and after Roux-en-Y gastric bypass

Study design Sample Methylation sites and methods Main results
Gene methylation: obesity
 Methylation of TNFα promoter in eutrophic women with decreased central adiposity (n = 40) [25] White blood cells Methylation of 20 CpG regions of TNFα, assessed by Sequenom EpiTyper MassARRAY Women with low central adiposity showed greater methylation in 2 CpG regions, associated with lower BMI and  % of body fat
 Association between methylation profile and BMI in very young children (n = 120) [26] Whole blood Methylation of 10 CpG regions of LEP, assessed by Sequenom EpiTyper MassARRAY An 0.8% decrease in LEP methylation levels was associated with an increase in BMI
 Methylation status of 2 CpG regions of POMC gene in obese (n = 71) and normal-weight (n = 36) adolescents, with a validation cohort in children (n = 154) [27] Whole blood Methylation in 2 CpG regions of POMC gene, assessed by bisulfite sequencing Lean adolescents had lower levels of POMC gene methylation
 Epigenetic modifications in normal-weight (n = 20), overweight (n = 20), and obese women (n = 20) [28] White blood cells Methylation in CpG regions of CLOCK, BMAL1, and PER2 genes, assessed by Sequenom EpiTyper MassARRAY Lean women had lower levels of methylation in CpG regions at BMAL1. Methylations in multiple CpG regions of CLOCK, BMAL1, and PER2 were positively associated with BMI,  % of body fat, and waist circumference (WC)
 Methylation profiles in obesity among monozygotic twins (n = 84) [30] White blood cells Methylation of 20 CpG regions of SLC6A4, assessed by pyrosequencing An increase in SLC6A4 gene methylation was associated with increases in BMI and WC
 Methylation pattern and BMI in elderly men (n = 170) [31] White blood cells Methylation of 8 CpG regions of AR, assessed by Sequenom EpiTyper MassARRAY Methylation of one region of CpG in the AR promoter was positively associated with BMI and  % of body fat
 Methylation profile and BMI in obese adults (n = 34) [32] White blood cells Methylation of HSD2, GR, IGF2, and H19, assessed by pyrosequencing Methylation in several CpG regions in HSD2, GR, and IGF2 was positively associated with WC and BMI
 Methylation profiles of lean (n = 48) and obese (n = 48) adolescents [35] White blood cells Methylation of >450,000 CpG regions, assessed by Illumina Infinium Human Methylation 450 BeadChip Differential methylation was observed in 23,305 CpG regions between lean and obese adolescents
 Methylation profiles of lean (n = 24) and obese (n = 23) female preadolescents [36] Whole blood Methylation of >27,000 CpG regions, assessed by Illumina Infinium Human Methylation 27 BeadChip There was a 15.5% difference in methylation of 20 CpG regions between the lean and obese groups
Genes methylation: T2D
 Methylation profiles in pancreatic islets from T2D (n = 15) and non-T2D (n = 34) donors [43] Pancreatic islets Gene methylation and expression in >450,000 CpG regions, assessed by Illumina Infinium Human Methylation 450 BeadChip and Affymetrix GeneChip Human Gene 1.0 ST array, respectively In T2D islets, 853 genes (i.e., TCF7L2, FTO, and KCNQ1) showed altered methylation profiles, while 102 genes (e.g., CDKN1A, PDE7B, SEPT9, and EXOC3L2) showed altered gene methylation and expression profiles. Methylation of these genes can affect pancreatic cell function
 Methylation profile in T2D (n = 710) and non-T2D patients (n = 459) [45] Whole blood Global methylation, assessed by Affymetrix SNP6 microarray CpG region of FTO gene showed hypomethylation (p < 0.05) in T2D patients compared to control group
 Methylation profile in T2D (n = 15), glucose-intolerant (n = 8), and non-T2D patients (n = 15) [51] Skeletal muscle Methylation and expression of PPARG and PGC1A genes, assessed by bisulfite sequencing and RT-qPCR, respectively PPARG and PGC1A were hypermethylated in T2D patients compared to other groups. Methylation levels were negatively correlated with PGC1A gene expression
 Epigenetic regulation of PPARGC1A in T2D (n = 48) and non-T2D (n = 12) islets [61] Pancreatic islets Methylation and expression of PPARGC1A gene, assessed by sequencing and RT-qPCR, respectively PPARGC1A promoter showed increased methylation (p < 0.05) in T2D patients. PPARGC1A gene expression was reduced 90% (p < 0.005) and correlated with reduced insulin secretion in T2D patient islets
 Methylation profiles in pancreatic islets from T2D (n = 9) and non-T2D (n = 48) donors [62] Pancreatic islets Methylation of 25 CpG regions of INS promoter assessed by Sequenom EpiTyper MassARRAY and pyrosequencing, and expression of INS assessed by RT-qPCR INS promoter methylation was increased in T2D patients and negatively correlated with INS expression (p < 0.05)
 PDX1 methylation profiles in pancreatic islets from T2D (n = 9) and non-T2D (n = 55) donors [63] Pancreatic islets Methylation of 29 CpG regions of PDX1 (15 CpG sites of the human distal PDX1 promoter and the other assay covered 14 CpG sites of the human PDX1 enhancer region) assessed by Sequenom EpiTyper MassARRAY and pyrosequencing, and expression of PDX1 assessed by RT-qPCR In T2D patient islets, 10 CpG regions of PDX1 showed increased methylation compared to non-T2D islets. PDX1 methylation was negatively correlated with gene expression
 MCP1 methylation profiles in T2D (n = 32) and non-T2D patients (n = 15) [64] White blood cells Methylation of MCP1 gene, assessed by PCR MCP1 promoter region was methylated in the control group. MCP1 showed less methylation (p < 0.001) but higher serum levels in T2D patients
Genes methylation: RYGB
 Obese T2D women with weight loss after RYGB (n = 8) compared to non-obese women (n = 9) [29] Skeletal muscle Methylation of PGC1A and PDK4 promoters, assessed by bisulfite sequencing PGC1A and PDK4 promoter methylation levels were modified by obesity and restored to normal levels after weight loss induced by RYGB
 Folate levels and epigenetic alterations in liver from subjects with T2D [46] Liver Methylation and gene expression of >450,000 CpG regions, assessed by Illumina Infinium Human Methylation 450 BeadChip and RT-qPCR, respectively In T2D patients, 251 CpG regions, including regions in GRB10, ABCC3, MOGAT1, and PRDM16, exhibited altered methylation compared to nondiabetic subjects. About 94% of modified CpG regions showed a decrease in hepatic DNA methylation in T2D patients. Hypomethylation was correlated with lower folate levels. T2D patients showed decreased erythrocyte folate levels compared to nondiabetic patients (p < 0.05). Finally, 29 other hepatic genes showed alterations in methylation and gene expression in T2D patients
 DNA methylation and hydroxymethylationin relation to energy-restricted diet (n = 22) or bariatric surgery (n = 14) [48] Whole blood Methylation assessed by PCR Baseline LINE-1 methylation was associated with serum glucose levels. Baseline hydroxymethylation was associated with BMI, WC, total cholesterol, and triglyceride levels. LINE-1 and SERPINE-1 methylation levels did not change after weight loss. IL6 methylation increased after energy restriction and decreased after bariatric surgery. SERPINE-1 methylation was associated with weight loss response
 Promoter methylation after RYGB and VLCD in obese patients (n = 18) [49] Whole blood Methylation of promoter regions of PPARGC1A, PDK4, TFAM, IL1B, IL6, and TNFα genes, assessed by PCR VLCD decreased methylation of PPARGC1A. Methylation levels of PPARGC1A, IL1B, IL6, and TNFα promoters were decreased 2 days after RYGB, but methylation levels of PDK4, IL1B, IL6, and TNFα promoters were increased 12 months after RYGB (p < 0.05)
 Methylation in obese patients (n = 11) 6 months after RYGB compared to health people (n = 16) [52] Whole blood Methylation of >450,000 CpG regions, assessed by Illumina Infinium Human Methylation 450 BeadChip ADK gene methylation was 10% lower 6 months after RYGB. ADK encodes adenosine kinase, a potential regulator of extracellular adenosine and intracellular nucleotide (e.g., adenina) levels. Adenosine can improve insulin secretion and decrease glucose production
 Differential methylation in obesity and T2D genes in siblings born before (n = 531) and after (n = 531) maternal bariatric surgery [53] Whole blood Methylation of >450,000 CpG regions, assessed by Illumina Infinium Human Methylation 450 BeadChip Comparing siblings born before vs. after surgery, 3074 genes were differentially methylated, with an overrepresentation of genes involved in insulin receptor, T2D, and leptin signaling in obesity. HLA-DQA1, HLA-DQB1, and TSPAN18 were the most significantly differentially methylated genes
 DNA methylation analysis in obese patients with NAFLD before (n = 45) and after bariatric surgery (n = 23) [54] Liver Methylation and gene expression of >450.,000 CpG regions, assessed by Illumina Infinium Human Methylation 450 BeadChip and Affymetrix Human Gene 1.1 ST, respectively Gene ontology and transcription factor binding site analyses revealed distinct postsurgery and NAFLD-specific methylation signatures, with >400-fold enrichment of NRF1, HSF1, and ESRRA sites. The findings illustrate treatment-induced epigenetic liver remodeling. PTPRE showed hypermethylation and decreased gene expression after RYGB. This result may represent a key mechanism for reestablishment of hepatic insulin sensitivity during weight loss
 DNA methylation in adipose tissue from obese women (n = 15) before and after weight loss by gastric bypass [55] Subcutaneous and omental adipose tissue Gene methylation and expression assessed for >450,000 CpG regions, by the Illumina Infinium Human Methylation 450 BeadChip and RT-qPCR, respectively Differential methylation was observed in omental and subcutaneous adipose tissue (p < 0.05). A greater proportion of CpG regions were hypermethylated before weight loss. Increased methylation was observed in the 3′ untranslated region and gene bodies relative to promoter regions. Differential methylation was found within genes associated with obesity, epigenetic regulation, and development, such as CETP, FOXP2, HDAC4, DNMT3B, KCNQ1, and HOX clusters. Robust correlations were observed between changes in methylation and clinical traits, including associations between fasting glucose and HDAC4, SLC37A3, and DENND1C in subcutaneous adipose. Genes investigated with differential promoter methylation all showed significantly different levels of mRNA before and after gastric bypass
 Fat cell epigenetic signature in women 2 years after RYGB (n = 16) compared to lean women (n = 14) [56] Subcutaneous adipose tissue Methylation of >450,000 CpG regions and gene expression, assessed by Illumina Infinium Human Methylation 450 BeadChip and Human Gene 1.1 ST, respectively After RYGB, 8504 CpG regions showed methylation in adipose tissue. After RYGB, 3717 genes were overexpressed and associated with cell differentiation pathways. Among the adipogenesis-associated genes, 27% presented altered methylation in patients after RYGB compared to the control group
 Longitudinal genome-wide methylation study in obese patients (n = 11) with hypertension 6 months after RYGB [59] Whole blood Methylation of >450,000 CpG regions, assessed by Illumina Infinium Human Methylation 450 BeadChip There were 24 promoters associated with CpG regions. Data were correlated with systolic blood pressure changes after RYGB. Two CpG loci (cg00875989, cg09134341) were hypomethylated and associated with hypertension