Transcriptomic Profiling Across the Nonalcoholic Fatty Liver Disease Spectrum Reveals Gene Signatures for Steatohepatitis and Fibrosis

The mechanisms that drive nonalcoholic fatty liver disease (NAFLD) remain incompletely understood. This large multicenter study characterized the transcriptional changes that occur in liver tissue across the NAFLD spectrum as disease progresses to cirrhosis to identify potential circulating markers. We performed high-throughput RNA sequencing on a discovery cohort comprising histologically characterized NAFLD samples from 206 patients. Unsupervised clustering stratified NAFLD on the basis of disease activity and fibrosis stage with differences in age, aspartate aminotransferase (AST), type 2 diabetes mellitus, and carriage of , a genetic variant associated with NAFLD. Relative to early disease, we consistently identified 25 differentially expressed genes as fibrosing steatohepatitis progressed through stages F2 to F4. This 25-gene signature was independently validated by logistic modeling in a separate replication cohort ( = 175), and an integrative analysis with publicly available single-cell RNA sequencing data elucidated the likely relative contribution of specific intrahepatic cell populations. Translating these findings to the protein level, SomaScan analysis in more than 300 NAFLD serum samples confirmed that circulating concentrations of proteins AKR1B10 and GDF15 were strongly associated with disease activity and fibrosis stage. Supporting the biological plausibility of these data, in vitro functional studies determined that endoplasmic reticulum stress up-regulated expression of , , and , whereas GDF15 supplementation tempered the inflammatory response in macrophages upon lipid loading and lipopolysaccharide stimulation. This study provides insights into the pathophysiology of progressive fibrosing steatohepatitis, and proof of principle that transcriptomic changes represent potentially tractable and clinically relevant markers of disease progression.