Snyder, R., Burgess, J., Deese-Spruill, J., Myers, C., Wu, S., Fennell, T., & Sumner, S. (2008). Metabolomics: Urinary markers of drug-induced liver injury with correlation with lobe variations in response. Toxicologist, 1917.
This study was conducted to evaluate the use of metabolomics for identifying biomarkers of drug-induced liver injury and associate these markers with specific biochemical pathways to give insight into the mechanism of liver injury. Rats were dosed with vehicle, no effect levels, or effect levels of clofibrate, valproic acid, isoniazid, phenytoin, or acetaminophen: the latter four drugs being the focus of the NIDDK clinical trials for drug-induced liver injury (DILI) and Idiosyncratic Liver Injury (ILIAD). Urine was collected for 24 h (on dry ice using all glass metabolism
chambers) from animals after 1 day of dosing or for up to 14 days of dosing. Liver was processed for examination of pathology for the right, left, and median lobes. The remaining liver was extracted for metabolomic analysis. Serum enzyme levels were also measured. Metabolomics of urine collected from rats exposed to vehicle, no effect doses, or effect doses exhibited different patterns for endogenous metabolites following 1 day or multiple days of dosing. Sub-sets of metabolites could be associated with specific pathology, while serum enzyme levels often failed to correlate with pathology outcomes. Metabolomic profiles of liver extracts were different between vehicle and dose groups, and profiles could be used to differentiate liver lobes. The relevance of metabolites that are significant to the separation of groups based on dose or pathology are being determined through pathway mapping. The application of metabolomics to develop urine based markers that can be used to screen, as an early marker, for drug-induced liver injury and linking markers that are reflective of specific mechanisms of liver injury to biochemical pathways will help gain a mechanistic understanding of the on-set or recovery from liver-injury.
This study is funded by NIGMS grant 5R21GM075903.