ABSTRACT
We investigated five methylation markers recently linked to body mass index, for their role in the neuropathology of obesity. In neuroimaging experiments, our analysis involving 23 participants showed that methylation levels for the cg07814318 site, which lies within the KLF13 gene, correlated with brain activity in the claustrum, putamen, cingulate gyrus and frontal gyri, some of which have been previously associated to food signaling, obesity or reward. Methylation levels at cg07814318 also positively correlated with ghrelin levels. Moreover, expression of KLF13 was augmented in the brains of obese and starved mice. Our results suggest the cg07814318 site could be involved in orexigenic processes, and also implicate KLF13 in obesity. Our findings are the first to associate methylation levels in blood with brain activity in obesity-related regions, and further support previous findings between ghrelin, brain activity and genetic differences.
Subject(s)
Cell Cycle Proteins/genetics , DNA Methylation , Ghrelin/metabolism , Kruppel-Like Transcription Factors/genetics , Neurons/metabolism , Obesity/genetics , Obesity/metabolism , Orexins/metabolism , Repressor Proteins/genetics , Animals , Appetite Regulation , Brain/cytology , Brain/diagnostic imaging , Brain/metabolism , Cell Cycle Proteins/metabolism , Feeding Behavior/physiology , Functional Neuroimaging , Gene Expression Regulation , Humans , Kruppel-Like Transcription Factors/metabolism , Male , Mice , Obesity/physiopathology , Receptors, Ghrelin/metabolism , Repressor Proteins/metabolism , RewardABSTRACT
Genome-wide association studies have identified a number of single-nucleotide polymorphisms (SNPs) that are associated with psychiatric diseases. Increasing body of evidence suggests a complex connection of SNPs and the transcriptional and epigenetic regulation of gene expression, which is poorly understood. In the current study, we investigated the interplay between genetic risk variants, shifts in methylation and mRNA levels in whole blood from 223 adolescents distinguished by a risk for developing psychiatric disorders. We analyzed 37 SNPs previously associated with psychiatric diseases in relation to genome-wide DNA methylation levels using linear models, with Bonferroni correction and adjusting for cell-type composition. Associations between DNA methylation, mRNA levels and psychiatric disease risk evaluated by the Development and Well-Being Assessment (DAWBA) score were identified by robust linear models, Pearson's correlations and binary regression models. We detected five SNPs (in HCRTR1, GAD1, HADC3 and FKBP5) that were associated with eight CpG sites, validating five of these SNP-CpG pairs. Three of these CpG sites, that is, cg01089319 (GAD1), cg01089249 (GAD1) and cg24137543 (DIAPH1), manifest in significant gene expression changes and overlap with active regulatory regions in chromatin states of brain tissues. Importantly, methylation levels at cg01089319 were associated with the DAWBA score in the discovery group. These results show how distinct SNPs linked with psychiatric diseases are associated with epigenetic shifts with relevance for gene expression. Our findings give a novel insight on how genetic variants may modulate risks for the development of psychiatric diseases.
Subject(s)
DNA Methylation , Glutamate Decarboxylase/genetics , Histone Deacetylases/genetics , Mental Disorders/genetics , RNA, Messenger/metabolism , Adaptor Proteins, Signal Transducing/genetics , Adolescent , Adult , Female , Formins , Gene Expression , Genetic Predisposition to Disease , Humans , Linear Models , Male , Orexin Receptors/genetics , Polymorphism, Single Nucleotide , Quantitative Trait Loci , Risk , Tacrolimus Binding Proteins/genetics , Young AdultABSTRACT
The ability of statins to strongly reduce low-density lipoprotein cholesterol (LDL-C) varies interindividually and is partially influenced by genetic variants. Based on a comprehensive analysis of 23 single nucleotide polymorphisms (SNPs) known to be associated with pharmacokinetics and dynamics of statins, we developed a genetic risk score to study its impact on the therapy outcome in elderly individuals under at least 5 years statin therapy. The study was performed in a population-based cohort of 1016 elderly individuals, which comprised 168 statin users investigated at age 75 and 80. Using random forest models, the major variants influencing LDL-C levels were summarized in a weighted GRS (wGRS). The wGRS was tested with lipid and glucose outcomes and validated in an independent population-based cohort including 221 statin users. Four SNPs within the APOE cluster (rs7412, rs4420638), ABCC2 (rs2002042) and CELSR/SORT1/PSRC1 (rs646776), displayed a major impact on statin efficacy. The wGRS was significantly associated with lower LDL-C at age 75 and 80. This association was replicated displaying similar results. GRS analysis is a powerful tool to evaluate the additive effects of genetic variants on statin response and to estimate the magnitude of LDL-C reduction to a considerable extent in the older population.