Epigenetic Regulation of Gene Expression in NAFLD

Abstract

BACKGROUND: Epigenetic modifications play an important role in the regulation of gene expression through response to environmental and developmental cues. Gene-environment interactions may impact which patients with nonalcoholic fatty liver disease (NAFLD) develop hepatic injury and progressive fibrosis (advanced NAFLD) rather than simple steatosis (mild NAFLD). The aim of this study was to identify sites of epigenetic regulation that affect liver gene expression patterns in NAFLD. METHODS: RNA and DNA were extracted from liver biopsies of 56 patients with NAFLD (33 mild NAFLD, F0-1; 23 advanced NAFLD, F3-4). Gene expression data were generated using Affymetrix human genome U133 Plus 2.0 arrays and Illumina Infinium HumanMethylation450 BeadChips were used to profile DNA methylation of >485,000 CpG dinucleotides (CpGs) from the same biopsy specimens. Methylation intensity was assessed via log2 ratio of methylated to unmethylated bead intensities. Differentially methylated CpGs were identified using an ANOVA model, adjusting for age and gender. After correcting for multiple comparisons, a q-value of<0.05 was considered significant. Differentially methylated CpGs inversely related with gene expression (r<-0.3; p-value<0.02) were determined. Generalized topological overlapping measure (GTOM) analysis was used to identify groups of genes showing coordinate epigenetic regulation, referred to as modules. RESULTS: There were 1143 genes with significant differential expression and 76,833 CpGs with differential methylation in mild vs. advanced NAFLD. Among these, 4,849 CpGs were negatively correlated with gene expression indicating a potential regulatory relationship between methylation status and transcription. GTOM analysis grouped the genes into 10 modules relating those with similar biologic function. Relative to mild NAFLD, 705 genes (22.2%) were hypermethylated and down-regulated (hyperM/down) in advanced NAFLD while 885 (27.9%) were hypomethylated and up-regulated (hypoM/up). Ingenuity Pathway Analysis revealed modules with hyperM/down genes to be involved in PXR/RXR activation (CYP1A2, CYP2C19, NR1I2, RXRA ) and metabolic processes (ADH6, DGAT2, MAT1A, OAT, PPP1R3B) while modules with hypoM/up genes are involved in cell signaling, matrix remodeling (CD44,COL1A1, ITGB2) and T cell immune responses (IL12R, NFKB STAT1). Further analysis revealed social alcohol, coffee and statin consumption was associated with methylation patterns in advanced NAFLD similar to those of non-consumers with mild NAFLD. CONCLUSIONS: We report the first genome-wide methylation study in patients with NAFLD. Integrated analysis of gene expression and methylation data obtained using high-throughout technologies is valuable for identification of relevant molecular events involved in advanced NAFLD and may reveal potential diagnostic and therapeutic biomarkers.