Liver-On-Chip as an Early Prediction-Selection Tool for the Toxicological Evaluation of Candidate Pharmaceutical Substances

Organ-on-chip assemblies are next-generation, microfluidic cell culture systems. They induce the cells to an overall closer to tissue/organ level behavior and also provide an improved monitoring capacity of key cell functions. Given the failure of animal models to accurately predict pharmacological end-points relevant for humans and the immense resource spending required for new drug development, the presented liver-on-chip aspires to provide an optimized alternative for toxicological candidate-drug assessment. The platform allows real-time analysis of minute shifts in metabolic activity. Mitochondrial respiration is monitored using two-frequency phase modulation of organoid-embedded phosphorescent microprobes, while the Central Carbon Metabolism is monitored using enzyme-based, off-chip, electrochemical measurements of glucose, lactate, glutamine and glutamate. We quantify the dynamics of cellular adaptation (ATP redistribution/metabolite production), upon exposure to several FDA-approved and candidate drugs, unveiling new insights unavailable to previous methods. We revealed that valproate, a common epilepsy treatment, increases lipid production, long before observable damage and troglitazone, an anti-diabetic drug, causes a metabolic shift from oxidative phosphorylation to anaerobic glycolysis. Thapsigargin, a candidate oncologic treatment, induces a similar glycolytic shift even on sub-threshold exposures, which further shifts to glutaminolysis, an indication of mitochondrial dysfunction. Acetaminophen, a common analgesic and antipyretic drug, induces a two-step toxicity, immediate and prolonged, indicating 2 separate mechanisms. Furthermore, the platform uniquely allows us to assess oxygen consumption change and its Time to Onset (TtO) with respect to drug administration, revealing a pattern of TtO for every different drug concentration, as an early indicator of cellular stress response. Overall, we present a robust, metabolic profile-based early predictor of drug safety, unveiling a brand new aspect to pharmacodynamics and instrumental in early discrimination between promising and unsuitable candidates.