Abstract
Whether or not dietary fructose exacerbate hepatic lipogenesis and non-alcoholic fatty liver disease (NAFLD) pathogenesis remains an unresolved question. Furthermore, few studies have utilised systems approaches, notably at a genome-scale level, to investigate specific monosaccharides on hepatic steatosis. Therefore, the aim of this work was to develop a predictive, multi-scale computational model of hepatic monosaccharide transport, signalling and metabolism; validated experimentally to an in vitro model of steatosis. Utilising quasi-steady state Petri nets (QSSPN), a dynamic network of hepatic glucose and fructose transport was reconstructed and integrated with a human hepatocyte-specific genome scale metabolic network constrained by in vitro flux data. Additionally, a regulatory, kinetic model of hepatic insulin signalling was reconstructed and successfully integrated in QSSPN. Together with our novel simulation method, dynamic flux variability analysis, simulations predicted minimum and maximum flux rates allowing the calculation of extracellular glucose and fructose, and triacylglycerol production over time, while also satisfying the demands of a ‘healthy hepatocyte’. Immortalised hepatic cells, HepG2, treated with fatty acids were confirmed as an in vitro model of NAFLD after the characterisation of both viability, as measured by LDH and MTT assays, and intracellular lipid by Nile red staining. Differential effects of glucose and fructose on lipid metabolism were also investigated in an in vitro mixed-nutrient model of glucose, fructose and oleate. Intracellular lipid was assessed by Nile red and the gold standard, gas chromatography mass spectrometry. While oleate induced intracellular lipid (P < 0.05), no statistical differences were detected between sugar treatments. Additionally, intracellular lipid by Nile red was measured in HepG2 cells treated with glucose or fructose ± insulin. Notably, while hepatic monosaccharide uptake in vitro was reproduced by our in silico model, differential lipogenic effects of fructose were not found experimentally, nor predicted by our in silico model. In conclusion, in vitro hepatic monosaccharide uptake and intracellular lipid was reproduced in our in silico model with no differential lipogenic effects of fructose detected. This highlights existing concerns of current public health recommendations targeting single nutrients, e.g. fructose, to combat lifestyle-related metabolic diseases such as NAFLD. Future work should aim to predict the outcomes and investigate underlying molecular and regulatory mechanisms of sugar and lipid metabolism in response to mixed meals containing physiological levels of glucose and fructose without confounding excess energy.