Genome-wide analysis of hepatic gene expression in patients with non-alcoholic fatty liver disease and in healthy donors in relation to hepatic fatty acid composition and other nutritional factors

In nonalcoholic fatty liver disease (NAFLD), hepatic gene expression and fatty acid (FA) composition have been reported independently but a comprehensive gene expression profiling in relation to FA composition is lacking. The aim was to assess this relationship. In a cross-sectional study, hepatic gene expression (Illumina Microarray) was first compared among 20 patients with simple steatosis (SS), 19 with nonalcoholic steatohepatitis (NASH), and 24 healthy controls (HC). FA composition in hepatic total lipids was compared between SS and NASH, and associations between gene expression and FA were examined. Gene expression differed mainly between HC and patients (SS and NASH), including genes related to unsaturated FA metabolism. Twenty-two genes were differentially expressed between NASH and SS; most of them correlated with disease severity and related more to cancer progression than to lipid metabolism. Biologically long-chain polyunsaturated FA (PUFA) (eicosapentaenoic acid + docosahexaenoic acid, arachidonic acid) in hepatic total lipids were lower in NASH than in SS. This may be related to overexpression of FADS1, FADS2, and PNPLA3. The degree and direction of correlations between PUFA and gene expression were different among SS and NASH which may suggest that low PUFA content in NASH modulates gene expression in a different way compared with SS or, alternatively, gene expression influences PUFA content differently depending on disease severity (SS versus NASH). Conclusion: Well-defined subjects with either healthy liver, SS or NASH showed distinct hepatic gene expression profiles including genes involved in unsaturated FA metabolism. In NASH, hepatic PUFA were lower and associations with gene expression were different than in SS. Cross-sectional study, including patients with non-alcoholic fatty liver disease (25 SS, 23 NASH) and 24 healthy living liver donors as HC. Nine samples (5 SS, 4 NASH) were excluded during quality control. The final analysis inclulded 63 participants (20 SS, 19 NASH, 24 HC).