Exploring metabolic modulations using genome-scale network modelling and omics data in the context of toxicological studies: application for deciphering metabolic shifts occurring during the differentiation of the human hepatic cell line HepaRG

Nathalie Poupin (INRA)
Thursday, February 8, 2018 - 10:30
Room Aurigny
Talk abstract: 

Many man-made chemicals present as contaminants in food and/or water are strongly suspected to induce adverse metabolic effects in Human. Liver is the key organ for xenobiotics biotransformation, and the use of metabolically competent cell lines is essential to explore the mechanisms underlying the metabolic effects of these substances. The hepatic cell line HepaRG, which is increasingly used in toxicity studies, has the particularity to differentiate from progenitor to mature hepatocyte-like cells. We combined multi-omics data and in silico methods in order to better characterize the metabolic capacities of this cell line and to explore the metabolic shifts occurring during this differentiation process. We integrated transcriptomic and metabolomic data in the context of the global human genome-scale metabolic network Recon2, which gathers the metabolic reactions the organism can perform and their associated genes, to compute a relevant sub-network, more specifically representing the functional hepatic metabolic network of HepaRG cells at each developmental stages: day 3 (progenitors) and day 30 (differentiated cells). We used a modified version of the iMAT algorithm developed by Shlomi et al. to identify, based on these data, the sub-networks of reactions specifically active in HepaRG cells at each developmental stage. For each stage, we identified several sub-networks of active reactions, having an equivalent adequacy to experimental data. We applied classification analysis methods to explore intra- and inter-stages variability among these sub-networks. We showed that, for each stage, the heterogeneity between sub-networks was mainly caused by the occurrence of several alternative reactions or the relative low contribution of transcriptomic data in some pathways. To better characterize the systemic metabolic capacities of the cells, we chose, contrary to most approaches, to consider the whole set of similarly adequate sub-networks, since it allows taking into account various metabolic alternatives. Through simulations and pathway enrichment analyses, we predicted that differentiated cells would globally be able to perform a larger number of liver-specific functions (e.g., urea production) and we identified several sets of reactions that were differently active between the two stages. These reactions mostly belong to pathways specific to hepatic activity (e.g., bile acid synthesis) but also to fatty acid synthesis and oxidation pathways. About 50% of the predicted modulated reactions were not evidenced from transcriptomic data and were « newly » inferred by the computational models. Globally, we showed that combining in silico methods with omics data enables to characterize global shifts in the developing hepatic metabolic network.