Strategy for Hepatotoxicity Prediction Induced by Drug Reactive Metabolites Using Human Liver Microsome and Online 2D-Nano-LC-MS Analysis.

Hepatotoxicity is a leading cause of drug withdrawal from the market; thus, the assessment of potential drug induced liver injury (DILI) in preclinical trials is necessary. More and more research has shown that the covalent modification of drug reactive metabolites (RMs) for cellular proteins is a possible reason for DILI. Unfortunately, so far no appropriate method can be employed to evaluate this kind of DILI due to the low abundance of RM-protein adducts in complex biological samples. In this study, we proposed a mechanism-based strategy to solve this problem using human liver microsomes (HLMs) and online 2D nano-LC-MS analysis. First, RM modification patterns and potential modified AA residues are determined using HLM and model amino acids (AAs) by UHPLC-Q-TOF-MS. Then, a new online 2D-nano-LC-Q-TOF-MS method is established and applied to separate the digested modified microsomal peptides from high abundance peptides followed by identification of RM-modified proteins using Mascot, in which RM modification patterns on specific AA residues are added. Finally, the functions and relationship with hepatotoxicity of the RM-modified proteins are investigated using ingenuity pathway analysis (IPA) to predict the possible DILI. Using this strategy, 21 proteins were found to be modified by RMs of toosendanin, a hepatotoxic drug with complex structure, and some of them have been reported to be associated with hepatotoxicity. This strategy emphasizes the identification of drug RM-modified proteins in complex biological samples, and no pretreatment is required for the drugs. Consequently, it may serve as a valuable method to predict potential DILI, especially for complex compounds.