Mathematical modelling reveals compound-specific stress pathway activity

Abstract

Drug-induced liver injury (DILI) is a major problem for the drug development industry. It has been suggested that activation of stress pathways within cells is an important indicator of DILI. In this project, we aimed to develop a mathematical model of invoked stress responses by three compounds with high DILI liability: nitrofurantoin, diclofenac and ketoconazole. To this end, we used imaging data from HepG2 cells and cell-associated compound and intracellular glutathione measurements. We initially developed a model for the integrated and oxidative stress responses following nitrofurantoin exposure. Subsequently, we expanded this to simulate responses to diclofenac and ketoconazole. To apply the model to these compounds multiple parameters required recalibration, yet the structure of the model was unchanged. Our analysis shows that the magnitude of interactions between transcription factors and downstream targets can differ even when the activated pathways are the same.