Integrated Approach to the Assessment of Acetaminophen Hepatoxicity: Reduction of “Omics” and New Technology to Practice

The Safety Sciences group at Pfizer has developed an integrated strategy for the analysis of toxicological data to reduce early compound attrition. This strategy helps to identify screening approaches and biomarkers for assessing the potential toxicity of candidate drugs by determining if they are structurally or mechanistically similar to known hepatotoxicants. These approaches are highly dependent on the emerging knowledge of genes, proteins, and molecular events in the mediation of adverse effects. Understanding the biology of adverse events is facilitated by a plethora of new technologies and bioinformatics tools available to the safety scientist. Presented here is a case scenario in which acetaminophen (APAP)-induced hepatotoxicity was studied using predictive in silico, in vitro, and in vivo techniques, including structure-activity relationships, biochemical evaluations, quantitative immunohistochemistry, gene expression profiling, and proteomics. Using in silico databases to search chemical specific structure-activity relationships, it was predicted that the hydroxyaniline moiety of APAP has the ability to form a quinone imine, which in turn can react with hepatic proteins to potentially cause toxicity. In primary cultures of rat hepatocytes and intact rats APAP caused depletion of glycogen prior to membrane leakage/necrosis. Induction of CYP3A increased APAP toxicity in rat but not human hepatocytes. Treatment of human hepatocytes with the combination of APAP and barbiturates resulted in toxicity associated with inhibition of APAP conjugation, indicating drug-drug interaction due to compromised APAP glucuronidation. This response was potentiated in cells prepared from individuals deficient in UDPGT activity. In intact rats, the necrosis predicted in silico and in vitro and was confirmed via histopathology and enzyme chemistry. Insight into the mechanism of APAP-mediated hepatotoxicity was achieved by showing increased expression of Heat Shock Protein 70 (HSP-70) using laser scanning cytometry, which suggested abnormal protein folding. It was also shown using enzyme chemistry (glutathione reductase, glucose 6-phosphate dehydrogenase, GSH) that APAP caused depletion of the glutathione system and secondary production of oxidative species. In addition, using gene expression-based statistical models, the livers from rats exposed to acetaminophen were predicted to be responding to a compound that causes necrosis, and more specifically had gene expression profiles similar to carbon tetrachloride and acetaminophen. Potentially predictive serum biomarkers were developed using Proteomics (SELDI TOF-MS) conducted on serum of animals experienced different severity of centrilobular necrosis as determined by histopathology. Controls were separated in different cluster from treated animals and some increases in proteins were directly proportional to the degree of damage detected via histopathology. In conclusion, an integrated approach can be used to identify potential toxic liability of compounds using multiple technologies to cross validate the endpoints results in identification of most efficient screening tools, drug-drug interaction and potential biomarker that contribute to toxicological decision-making during drug developmental process.