Association between the subcellular localization of host proteins and gut microbiome and metabolome in metabolic dysfunction-associated steatotic liver disease: a pilot study.

BACKGROUND: Metabolic dysfunction-associated steatotic liver disease (MASLD) is estimated to affect 38% of the global population, with limited options for treatment. It could progress to metabolic-associated steatohepatitis (MASH), fibrosis, and hepatocellular carcinoma. Agonists for farnesoid X receptor (FXR), peroxisome proliferation-associated receptors (PPARs), and sirtuin1 (SIRT1) are currently investigated for MASLD treatment. The subcellular localization of those proteins was shown to affect their function and could possibly be affected by different metabolites. Moreover, while those protein targets were found to be affected by the gut microbiome in mice, they have not yet been investigated in humans. Existing evidence independently links the gut microbiome to MASLD onset and demonstrates that host proteins are impacted by the microbiome. Therefore, we aimed at using integrative multi-omics analysis to investigate the interrelationship between the gut microbiome, fecal and serum metabolomes, and those selected protein targets in a cohort of patients with MASLD to identify potential markers differentiating MASLD and MASH.

METHODS: Serum and stool samples were collected from patients with MASLD and healthy controls, while formalin-fixed paraffin-embedded (FFPE) liver biopsies and clinical laboratory tests were obtained from patients only. Expression of the protein targets was analyzed by immunohistochemistry (IHC). Microbiome and metabolome analyses were performed, followed by bioinformatics, correlation, and multivariate and integrated multi-omics analyses.

RESULTS: SIRT1 and FXR subcellular localizations were correlated with multiple bacteria and metabolites, respectively. Three genera (Rothia, Haemophilus, and Acetatifactor) correlated with NAFLD activity score (NAS), and a signature of 20 bacterial genera, 10 fecal and 30 serum metabolites, and 3 host proteins differentiated between MASLD and MASH. Moreover, in silico analysis suggested myristic, lauric, octanoic, and nonanoic acids to putatively affect peroxisome proliferator-activated receptor alpha (PPARA) and FXR, and Coprobacter as an important contributor in our multi-omics model.

CONCLUSION: Our data suggest bacteria and metabolites which potentially affect the subcellular localization, and hence activity, of anti-lipogenic proteins in MASLD patients. We also propose novel discriminatory markers between MASLD and MASH. Our findings form the groundwork for future mechanistic studies of both host and microbial factors possibly contributing to the multifaceted disease outcome and offer potential diagnostic markers.