Epigenome-wide methylation study identified two novel CpGs associated with T2DM risk and a network of co-methylated CpGs capable of patient's classifications.

Prevention of Type 2 diabetes mellitus (T2DM) pandemic needs markers that can precisely predict the disease risk in an individual. Alterations in DNA methylations due to exposure towards environmental risk factors are widely sought markers for T2DM risk prediction. To identify such individual DNA methylation signatures and their effect on disease risk, we performed an epigenome-wide association study (EWAS) in 844 Indian individuals of Indo-European origin. We identified and validated methylation alterations at two novel CpG sites in MIR1287 (cg01178710) and EDN2-SCMH1 (cg04673737) genes associated with T2DM risk at the epigenome-wide-significance-level (P < 1.2 × 10-7). Further, we also replicated the association of two known CpG sites in TXNIP, and CPT1A in the Indian population. With 535 EWAS significant CpGs (P < 1.2 × 10-7) identified in the discovery phase samples, we created a co-methylation network using weighted correlation network analysis and identified four modules among the CpGs. We observed that methylation of one of the module associates with T2DM risk factors (e.g. BMI, insulin and C-peptide) and can be used as markers to segregate T2DM patients with good glycemic control (e.g. low HbA1c) and dyslipidemia (low HDL and high TG) from the other patients. Additionally, an intronic SNP (rs6503650) in the JUP gene, a member of the same module, associated with methylation at all the 14 hub CpG sites of that module as methQTL. Our network-assisted EWAS is the first to systematically explore DNA methylation variations conferring risks to T2DM in Indians and use the identified risk CpG sites for patient segregation with different clinical outcomes. These findings can be useful for better stratification of patients to improve the clinical management and treatment effects.

Multifaceted genome-wide study identifies novel regulatory loci in SLC22A11 and ZNF45 for body mass index in Indians.

Obesity, a risk factor for multiple diseases (e.g. diabetes, hypertension, cancers) originates through complex interactions between genes and prevailing environment (food habit and lifestyle) that varies across populations. Indians exhibit a unique obesity phenotype with high abdominal adiposity for a given body weight compared to matched white populations suggesting presence of population-specific genetic and environmental factors influencing obesity. However, Indian population-specific genetic contributors for obesity have not been explored yet. Therefore, to identify potential genetic contributors, we performed a two-staged genome-wide association study (GWAS) for body mass index (BMI), a common measure to evaluate obesity in 5973 Indian adults and the lead findings were further replicated in 1286 Indian adolescents. Our study revealed novel association of variants-rs6913677 in BAI3 gene (p = 1.08 × 10-8) and rs2078267 in SLC22A11 gene (p = 4.62 × 10-8) at GWAS significance, and of rs8100011 in ZNF45 gene (p = 1.04 × 10-7) with near GWAS significance. As genetic loci may dictate the phenotype through modulation of epigenetic processes, we overlapped genetic data of identified signals with their DNA methylation patterns in 236 Indian individuals and performed methylation quantitative trait loci (meth-QTL) analysis. Further, functional roles of discovered variants and underlying genes were speculated using publicly available gene regulatory databases (ENCODE, JASPAR, GeneHancer, GTEx). The identified variants in BAI3 and SLC22A11 genes were found to dictate methylation patterns at unique CpGs harboring critical cis-regulatory elements. Further, BAI3, SLC22A11 and ZNF45 variants were located in repressive chromatin, active enhancer, and active chromatin regions, respectively, in human subcutaneous adipose tissue in ENCODE database. Additionally, these genomic regions represented potential binding sites for key transcription factors implicated in obesity and/or metabolic disorders. Interestingly, GTEx portal identify rs8100011 as a robust cis-expression quantitative trait locus (cis-eQTL) in subcutaneous adipose tissue (p = 1.6 × 10-7), and ZNF45 gene expression in skeletal muscle of Indian subjects showed an inverse correlation with BMI indicating its possible role in obesity. In conclusion, our study discovered 3 novel population-specific functional genetic variants (rs6913677, rs2078267, rs8100011) in 2 novel (SLC22A11 and ZNF45) and 1 earlier reported gene (BAI3) for BMI in Indians. Our study decodes key genomic loci underlying obesity phenotype in Indians that may serve as prospective drug targets in future.

Genome-Wide Association Study of Metabolic Syndrome Reveals Primary Genetic Variants at CETP Locus in Indians.

Indians, a rapidly growing population, constitute vast genetic heterogeneity to that of Western population; however they have become a sedentary population in past decades due to rapid urbanization ensuing in the amplified prevalence of metabolic syndrome (MetS). We performed a genome-wide association study (GWAS) of MetS in 10,093 Indian individuals (6,617 MetS and 3,476 controls) of Indo-European origin, that belong to our previous biorepository of The Indian Diabetes Consortium (INDICO). The study was conducted in two stages-discovery phase (N = 2,158) and replication phase (N = 7,935). We discovered two variants within/near the CETP gene-rs1800775 and rs3816117-associated with MetS at genome-wide significance level during replication phase in Indians. Additional CETP loci rs7205804, rs1532624, rs3764261, rs247617, and rs173539 also cropped up as modest signals in Indians. Haplotype association analysis revealed GCCCAGC as the strongest haplotype within the CETP locus constituting all seven CETP signals. In combined analysis, we perceived a novel and functionally relevant sub-GWAS significant locus-rs16890462 in the vicinity of SFRP1 gene. Overlaying gene regulatory data from ENCODE database revealed that single nucleotide polymorphism (SNP) rs16890462 resides in repressive chromatin in human subcutaneous adipose tissue as characterized by the enrichment of H3K27me3 and CTCF marks (repressive gene marks) and diminished H3K36me3 marks (activation gene marks). The variant displayed active DNA methylation marks in adipose tissue, suggesting its likely regulatory activity. Further, the variant also disrupts a potential binding site of a key transcription factor, NRF2, which is known for involvement in obesity and metabolic syndrome.

Common variants of ARID1A and KAT2B are associated with obesity in Indian adolescents.

Obesity involves alterations in transcriptional programs that can change in response to genetic and environmental signals through chromatin modifications. Since chromatin modifications involve different biochemical, neurological and molecular signaling pathways related to energy homeostasis, we hypothesize that genetic variations in chromatin modifier genes can predispose to obesity. Here, we assessed the associations between 179 variants in 35 chromatin modifier genes and overweight/obesity in 1283 adolescents (830 normal weight and 453 overweight/obese). This was followed up by the replication analysis of associated signals (18 variants in 8 genes) in 2247 adolescents (1709 normal weight and 538 overweight/obese). Our study revealed significant associations of two variants rs6598860 (OR = 1.27, P = 1.58 × 10-4) and rs4589135 (OR = 1.22, P = 3.72 × 10-4) in ARID1A with overweight/obesity. We also identified association of rs3804562 (ß = 0.11, P = 1.35 × 10-4) in KAT2B gene with BMI. In conclusion, our study suggests a potential role of ARID1A and KAT2B genes in the development of obesity in adolescents and provides leads for further investigations.

DNA methylation profiling reveals the presence of population-specific signatures correlating with phenotypic characteristics.

Phenotypic characteristics are known to vary substantially among different ethnicities around the globe. These variations are mediated by number of stochastic events and cannot be attributed to genetic architecture alone. DNA methylation is a well-established mechanism that sculpts our epigenome influencing phenotypic variation including disease manifestation. Since DNA methylation is an important determinant for health issues of a population, it demands a thorough investigation of the natural differences in genome wide DNA methylation patterns across different ethnic groups. This study is based on comparative analyses of methylome from five different ethnicities with major focus on Indian subjects. The current study uses hierarchical clustering approaches, principal component analysis and locus specific differential methylation analysis on Illumina 450K methylation data to compare methylome of different ethnic subjects. Our data indicates that the variations in DNA methylation patterns of Indians are less among themselves compared to other global population. It empirically correlated with dietary, cultural and demographical divergences across different ethnic groups. Our work further suggests that Indians included in this study, despite their genetic similarity with the Caucasian population, are in close proximity with Japanese in terms of their methylation signatures.

Genome wide association study of uric acid in Indian population and interaction of identified variants with Type 2 diabetes.

Abnormal level of Serum Uric Acid (SUA) is an important marker and risk factor for complex diseases including Type 2 Diabetes. Since genetic determinant of uric acid in Indians is totally unexplored, we tried to identify common variants associated with SUA in Indians using Genome Wide Association Study (GWAS). Association of five known variants in SLC2A9 and SLC22A11 genes with SUA level in 4,834 normoglycemics (1,109 in discovery and 3,725 in validation phase) was revealed with different effect size in Indians compared to other major ethnic population of the world. Combined analysis of 1,077 T2DM subjects (772 in discovery and 305 in validation phase) and normoglycemics revealed additional GWAS signal in ABCG2 gene. Differences in effect sizes of ABCG2 and SLC2A9 gene variants were observed between normoglycemics and T2DM patients. We identified two novel variants near long non-coding RNA genes AL356739.1 and AC064865.1 with nearly genome wide significance level. Meta-analysis and in silico replication in 11,745 individuals from AUSTWIN consortium improved association for rs12206002 in AL356739.1 gene to sub-genome wide association level. Our results extends association of SLC2A9, SLC22A11 and ABCG2 genes with SUA level in Indians and enrich the assemblages of evidence for SUA level and T2DM interrelationship.