906 news & views NATURE MEDICINE | VOL 24 | JULY 2018 | 899–907 | www.nature.com/naturemedicine (ref. 9 ), and genes encoding spliceosome components 10 , as well as recent preclinical advances in restoring TET2 function 11,12 . The ultimate aim of such integrated models would be to test interventions that may mitigate the adverse consequences of CHIP. These new reports provide another useful step toward such models and the rational design of interventional studies. ❐ Rob S. Sellar 1,2,3 , Siddhartha Jaiswal 4 and Benjamin L. Ebert 1,2,5 * 1 Brigham and Women’s Hospital, Boston, MA, USA. 2 Broad Institute of MIT and Harvard, Cambridge, MA, USA. 3 UCL Cancer Institute, University College London, London, UK. 4 Department of Pathology, Stanford University, Stanford, CA, USA. 5 Dana-Farber Cancer Institute, Boston, MA, USA. *e-mail: bebert@partners.org Published online: 9 July 2018 https://doi.org/10.1038/s41591-018-0114-7 References 1. Young, A. L., Challen, G. A., Birmann, B. M. & Druley, T. E. Nat. Commun. 7, 12484 (2016). 2. Steensma, D. P. et al. Blood 126, 9–16 (2015). 3. Genovese, G. et al. N. Engl. J. Med. 371, 2477–2487 (2014). 4. Jaiswal, S. et al. N. Engl. J. Med. 371, 2488–2498 (2014). 5. Jaiswal, S. et al. N. Engl. J. Med. 377, 111–121 (2017). 6. Abelson, S. et al. Nature http://doi.org/10.1038/s41586-018-0317- 6 (2018). 7. Desai, P. et al. Nat. Med. http://doi.org/10.1038/s41591-018- 0081-z (2018). 8. Papaemmanuil, E. et al. N. Engl. J. Med. 374, 2209–2221 (2016). 9. Amatangelo, M. D. et al. Blood 130, 732–741 (2017). 10. Lee, S. C. & Abdel-Wahab, O. Nat. Med. 22, 976–986 (2016). 11. Agathocleous, M. et al. Nature 549, 476–481 (2017). 12. Cimmino, L. et al. Cell 170, 1079–1095 (2017). Competing interests The authors declare no competing interests. MICROBIOME Microbiome metabolomics reveals new drivers of human liver steatosis An integrative multiomics approach in nondiabetic obese women identifies phenylacetate as a microbial metabolite contributing to the accumulation of lipids in the liver and hence to nonalcoholic steatohepatitis. Nathalie M. Delzenne and Laure B. Bindels U p to a quarter of patients with nonalcoholic fatty liver disease (NAFLD) develop a progressive inflammatory liver disease termed nonalcoholic steatohepatitis (NASH) that may progress toward cirrhosis and hepatocellular carcinoma. However, the biological events driving the progression of NAFLD are not clearly elucidated. The gut microbiota—that is, the microorganisms living in the intestine—is able to modulate host metabolism and immunity, mainly through the release of metabolites and bioactive components 1 , and clearly contributes to the metabolic products in blood plasma—the plasma metabolome. Although tremendous efforts have recently been dedicated to the identification of bacterial by-products 2 , the mechanism behind the beneficial or detrimental effects of key microbial metabolites on host health in the field of NAFLD remains to be studied 3 . In this issue, Hoyles et al. 4 explored the plasma and urinary metabolome of nondiabetic obese women and integrated these data with fecal metagenomics and hepatic transcriptome data to unravel the molecular pathways linking the gut microbiota to hepatic steatosis. The authors focused on a cohort of 56 morbidly obese, weight-stable, nondiabetic women recruited in Italy and Spain who did not receive any hypoglycemic drug treatment (among numerous exclusion criteria). Across the cohort, the women presented variable levels of steatosis. This homogenous cohort allowed the authors to identify confounding factors to the analyses that were then taken into account and to focus on potential key explanatory factors for variability in the progression to steatosis. They found that the degree of steatosis and also markers of liver dysfunction were negatively associated with microbial gene richness and were positively associated with levels of Proteobacteria, Actinobacteria and Verrucomicrobia at the phyla level in the gut. With regards to metabolite association with steatosis, the authors’ characterization of microbial functions revealed, among other findings, a positive association of hepatic steatosis with gut microbial amino acid metabolism. Accordingly, the authors’ fecal metabolome analysis uncovered a link between elevated branched chain and aromatic amino acid level and steatosis, as well as an increase in phenylacetate that strongly correlated with steatosis. Phenylacetate is the main phenolic compound found in fecal samples and can be derived from plant secondary compounds as well as from microbial fermentation of aromatic amino acids (mainly phenylalanine) 5 . Interestingly, Bacteroides spp. is likely the major contributor to microbial phenylacetate production in humans 5 . In the mammalian host, phenylacetate production from phenylalanine through hepatic transamination is normally low (except in the case of phenylketonuria), so it can be assumed that the level seen by these authors is from microbial metabolism (Fig. 1). Then, the authors transferred the fecal material from obese women to antibiotic- pretreated mice and observed an increased accumulation of lipids in the livers of mice transplanted with the gut microbiota from the patients with grade 3 NASH as compared to mice transplanted with the gut microbiota from those with grade 0 disease. This clarified the causal role played by the gut microbiota of patients with steatosis in hepatic lipid accumulation. In a second set of experiments, the authors fed mice a diet including phenylacetate, and this lead to increased hepatic lipid content. The authors further corroborated the causal effect of phenylacetate on steatosis in in vitro experiments in primary human hepatocytes. By doing so, the authors established that production of phenylacetate is a key component in the mechanisms through which the gut microbiota contributes to hepatic steatosis. However, as pointed out by the authors, phenylacetate administration did not fully recapitulate the full steatosis