Abstract Body: Introduction: The prevalence of hepatic steatosis—a central and early phenotype in multi-system metabolic dysfunction—is increasing in parallel with the obesity pandemic, calling for novel approaches for prevention and treatment.
Hypothesis: We hypothesized that the circulating proteome may reflect cell specific mechanisms of hepatic steatosis.
Methods: Using multi-modality hepatic imaging and broad circulating proteomics in approximately 5,000 individuals across 3 diverse cohorts (CARDIA, Cameron County Hispanic Cohort, UK Biobank), we identified proteins implicated in the progression of hepatic steatosis. We tested for a relationship with these proteomic markers of hepatic steatosis with metabolic-related clinical outcomes in UK Biobank. We translated these findings from the circulating proteome to several tissue-based datasets including bulk RNA sequencing, single-cell RNA sequencing, and spatial transcriptomics. To further prove the hepatocyte origin of prioritized proteins, we used a humanized “liver-on-a-chip” model.
Results: We observed proteins implicated in the progression of hepatic steatosis—such as those related to central carbon and amino acid metabolism, hepatocyte regeneration, inflammation, fibrosis, insulin sensitivity—are largely encoded by genes enriched at the transcriptional level in human liver. Circulating multi-protein signatures of hepatic steatosis were strongly associated with a fatty liver disease phenotype and multi-system metabolic outcomes in >26,000 free-living individuals. Moreover, we observed increased activity of transcripts encoding proteins prioritized in clinical studies spatially in areas of steatosis via spatial transcriptomics in human liver, with several top candidates dynamic during progression of steatosis in human liver. Finally, using a humanized “liver-on-a-chip” model, we induced hepatic steatosis, confirming cell-specific expression of targets implicated across tissue and clinical studies at a transcriptional and proteomic level.
Conclusions: These results underscore the utility of a unified approach that combines human studies, multi-omics, and dynamic tissue-on-a-chip experiments to identify a prognostic, functional, dynamic “liquid biopsy” of human liver, with relevance for clinical biomarker discovery and mechanistic research applications.