Hepatic steatosis is a multifactorial condition related to obesity that may contribute to the development of virus-associated and non-alcoholic fatty liver disease (NAFLD) in humans, in which patients usually start to show symptoms when the disease is in its advanced stages. Although the gut microbiome has recently evolved as a new and important player in the pathophysiology of liver diseases, the mechanisms involved have not yet been fully elucidated.

A new study, led by Dr. Marc-Emmanuel Dumas from the Department of Surgery and Cancer at Imperial College London (United Kingdom), has found that the gut microbiome along with the host transcriptome and metabolome interact to drive hepatic steatosis progression in non-diabetic obese women.

The researchers studied two independent cohorts of Italian (n = 61) and Spanish (n = 44) non-diabetic morbidly obese women recruited to the FLORINASH study, who were negative for viral hepatitis. Clinical variables, fecal metagenomes, plasma and urine metabolomes, and liver transcriptomes (called molecular phenomics) were studied through an integrative multi-omics approach. Clinical variables identified as significant confounders and thus included in statistical analyses include age, cohort and body mass index.

Analyses of fecal metagenomics and molecular phenomics revealed a relationship between the gut microbiome and host gene expression and metabolic pathways and the degree of hepatic steatosis. Metagenomic signatures of hepatic steatosis included a negative association between gene count (a measure of microbial gene richness) and liver steatosis progression and several markers of liver function (including g-glutamyltransferase, alanine aminotransferase and C-reactive protein). Regarding the role of gut microbiota on these correlations, the phyla Proteobacteria, Actinobacteria and Verrucomicrobia were associated with liver steatosis and other related clinical parameters. Besides this, positive associations were found between hepatic steatosis and microbial carbohydrate, lipid and amino acid metabolism. Of particular interest was the positive correlation between lipopolysaccharide (LPS) and peptidoglycan biosynthesis—notably from Proteobacteria—with liver steatosis, which had been previously reported in rodents. On the whole, these findings suggest that microbial composition and metabolism contribute to liver function in morbidly obese women.

The gut microbiome effect on the hepatic steatosis phenome was also reported by an association between liver steatosis-associated urine and plasma aromatic (AAAs, tryptophan, tyrosine and phenylalanine) and branched-chain (BCAAs, valine, leucine and isoleucine) amino acids and low microbial gene richness. While the increase in the level of BCAA had been previously confirmed in the context of obesity and insulin resistance, non-steatotic patients exhibited a high microbial gene richness that was significantly correlated with several gut-derived microbial metabolites.

Based on these findings, the transfer of fecal material from patients with liver steatosis grade 3 to mice led to increased hepatic lipid accumulation and metabolic phenotypes, including increased hepatic triglycerides, circulating BCAAs and trimethylamine N-oxide (TMAO). These findings suggest a causal role of the human fecal microbiota in the triggering of hepatic steatosis, which may contribute to hepatic comorbidities.

Phenylacetic acid (PAA)—a microbial product of aromatic amino acid metabolism— showed the strongest positive association with steatosis and low microbial gene richness and the researchers sought to investigate its potential for affecting the hepatic steatosis phenome. Fecal microbiota transplants and chronic treatment with PAA on both primary cultures of human hepatocytes and in mice led to increased lipid accumulation, the expression of genes involved in steatosis—such as lipoprotein lipase and fatty acid synthase genes—and increased BCAA utilization.

Finally, the study quantified the extent to which the crosstalk between gut microbiome, clinical phenotypes, liver transcriptome and urine and plasma metabolomes was robust. A strong correlation was found between metagenomic and phenomic data, with urinary metabolome and clinical parameters showing the weaker association. According to the researchers, a predictive model using metagenomic, transcriptomic and metabolomic data might provide a more robust signature for predicting hepatic steatosis.

In conclusion, this study provides an in-depth picture of the contribution of the interactions between gut metagenome and host molecular phenome data as a means of identifying subgroups of patients with different degrees of hepatic steatosis. The results show the existence of a metabolic phenotype associated with hepatic steatosis and low microbial gene richness, with elevated BCAA, AAA and microbial metabolite levels linked to an altered liver function. According to what the researchers explained in a press release, these findings open up a potential way of using metabolomics to detect liver steatosis before any significant liver damage takes place, whereas current blood tests and ultrasound scans only detect this condition once significant liver damage has occurred.




Hoyles L, Fernández-Real JM, Federici M, et al. Molecular phenomics and metagenomics of hepatic steatosis in non-diabetic obese women. Nat Med. 2018; doi: 10.1038/s41591-018-0061-3.