Predicting Changes in Renal Metabolism after Compound Exposure with a Genome-Scale Metabolic Model

The kidneys are metabolically active organs that are important for several physiological tasks such as the secretion of soluble wastes into urine. Other functions of the kidneys include synthesizing glucose and oxidizing fatty acids for energy in fasting (non-fed) conditions. Once damaged, the metabolic capability of the kidneys becomes altered. Here, we define metabolic tasks in a computational modeling framework to capture kidney function in an update to the iRno network reconstruction of rat metabolism using literature-based evidence. To demonstrate the utility of iRno for predicting kidney function, we exposed primary rat renal proximal tubule epithelial cells to five compounds with varying levels of nephrotoxicity (acetaminophen, carbon tetrachloride, gentamicin, 2,3,7,8-tetrachlorodibenzodioxin, and trichloroethylene for six and twenty-four hours), and collected transcriptomics and metabolomics data to measure the metabolic effects of compound exposure. We observed changes in fatty acid metabolism and amino acid metabolism, consistent with changes in existing markers of kidney function such as Lcn2 (lipocalin), Clu (clusterin), and the carnitine transporter Octn2 (solute carrier family 22, member 5). The iRno metabolic network reconstruction was able to predict alterations in these same pathways after integrating transcriptomics data, and was able to distinguish between select compound-specific effects on the proximal tubule epithelial cells. Genome-scale metabolic network reconstructions with coupled omics data can be used to predict changes in metabolism to provide a step towards identifying novel biomarkers of kidney function and dysfunction. Examination of RNA from compound-treated primary rat renal proximal tubule epithelial cells for 6 conditions (including the control) with 3 replicates for each condition.