Editors' ChoiceLiver disease

Richer data with personalized GEMs

See allHide authors and affiliations

Science Translational Medicine  15 Mar 2017:
Vol. 9, Issue 381, eaam9865
DOI: 10.1126/scitranslmed.aam9865


Personalized genome-scale metabolic models augment clinical hepatic steatosis data and motivate treatment based on glutathione repletion.

Genome-scale metabolic models (GEMs), consisting of mathematically defined collections of thousands of biochemical reactions, have been developed with increasing detail and efficiency for over a decade. GEMs allow metabolic networks to be computationally simulated based on a limited number of constraining measurements, which then enables unmeasured metabolic pathway activities to be inferred. Although the vast majority of research using such complex models has been applied to controlled in vitro experiments, GEMs have recently shown promise for interpreting clinical data and applying personalized medicine to metabolic disease.

In a recent study, Mardinoglu and colleagues use a GEM constructed specifically for hepatocytes to uncover mechanisms of nonalcoholic fatty liver disease and its characteristic feature, hepatic steatosis (HS), a common condition defined by accumulation of fat in the liver. There are few pharmacological treatment options for HS and the mechanisms controlling progression to more severe diseases such as nonalcoholic steatohepatitis and hepatocellular carcinoma need further clarification. Impressively, the authors computed personalized GEMs for 73 individual patients with varying degrees of HS.

To accomplish this, typical clinical parameters such as body mass index were combined with kinetic studies of very low density lipoprotein production in the liver, using administered stable isotopes and serial blood draws. Computed GEMs for each patient were then correlated with their HS severity, revealing the disease to be particularly associated with an unbalanced redox state and increased demand for glutathione (GSH) and NAD+. Serum metabolomics provided further evidence for GSH deficiency: the GSH precursor glycine and its associated metabolites, including readily interconverted serine, were inversely correlated with disease severity.

These results ultimately motivated the test of whether serine supplementation could improve disease progression. Experiments in mice showed efficacy, and a small clinical trial with six patients also had a positive outcome. In subjects with high HS, daily ~20 g oral l-serine supplementation significantly reduced liver fat percentage from 27 to 20% in 2 weeks, which is an encouraging motivation for larger trials. Overall, Mardinoglu et al. show how personalized GEMs can provide a rich mechanistic perspective that guides feasible therapeutic strategies.

Highlighted Article

Stay Connected to Science Translational Medicine

Navigate This Article