Abstract
It is estimated that 1 in 3 people in the UK show early warning signs of Non-Alcoholic Fatty Liver Disease (NAFLD), a condition that starts with steatosis, the abnormal accumulation of lipid cells in the liver. Understanding the processes by which this accumulation of lipids occurs could lead to the development of novel drugs to treat steatosis before disease progression can occur. A systems wide computational approach was adopted to understand the interaction between metabolism, cell signalling and gene regulation producing a phenotypic output that resembles steatosis. The in silico model comprises of two different computational formalisms: the human Genome-Scale Metabolic Network (GSMN) Recon2 (analysed using Flux Balance Analysis (FBA)), coupled with the MAP Kinase signalling network; and Nuclear Receptor (NR) gene regulation modelled using Petri Net (PN) theory. These two distinct formalisms are combined using the Quasi-Steady State Petri Net (QSSPN) algorithm. Previously published work uses small-scale static models. This project builds on that research by using quantitative data for the dynamic gene regulatory network, including experimentally determined reaction rates, protein levels and receptor occupancy levels. It also includes the MAP Kinase signalling network created by the conversion of the Reactome dataset, and comprises six MAPK-activated NRs. The results showed that the model can uncover previously unknown metabolic landscapes. It was found that Glucocorticoid Receptor (GR) has an inhibitory effect when coupled with any other NR (excluding the Peroxisome-Proliferator Activated Receptor Alpha (PPARa)), which causes a decrease in the synthesis of triacylglycerol (TAG). This was due to GRs inhibition of IDH1, a gene that upregulates the enzyme isocitrate dehydrogenase. PPARa had a neutralising effect when combined with GR, sustaining a steady state level of TAG. In conclusion, the dynamic mechanistic model developed has the capability to study the metabolic network on a system wide level and identify new biological mechanisms. Supporting data can be found at http://doi.org/10.15126/surreydata.00858237