Metabolic detoxification determines species differences in coumarin-induced hepatotoxicity.

Hepatotoxicity of coumarin is attributed to metabolic activation to an epoxide intermediate, coumarin 3,4-epoxide (CE). However, whereas rats are most susceptible to coumarin-induced hepatotoxicity, formation of CE is greatest in mouse liver microsomes, a species showing little evidence of hepatotoxicity. Therefore, the present work was designed to test the hypothesis that detoxification of CE is a major determinant of coumarin hepatotoxicity. CE can either rearrange spontaneously to o-hydroxyphenylacetaldehyde (o-HPA) or be conjugated with gluatathione (GSH). o-HPA is hepatotoxic and is further detoxified by oxidation to o-hydroxyphenylacetic acid (o-HPAA). In vitro experiments were conducted using mouse liver microsomes to generate a constant amount of CE, and cytosols from F344 rats, B6C3F1 mice, and human liver were used to characterize CE detoxification. All metabolites were quantified by HPLC methods with UV detection. In rats and mice, GSH conjugation occurred non-enzymatically and through glutathione-S-transferases (GSTs), and the kinetics of GSH conjugation were similar in rats and mice. In rat liver cytosol, oxidation of o-HPA to o-HPAA was characterized with a high affinity K(m) of approximately 12 microM, and a V(max) of approximately 1.5 nmol/min/mg protein. In contrast, the K(m) and V(max) for o-HPA oxidation in mouse liver cytosol were approximately 1.7 microM and 5 nmol/min/mg protein, respectively, yielding a total intrinsic clearance through oxidation to o-HPAA that was 20 times higher in mouse than in rats. Human cytosols (two separate pools) detoxified CE through o-HPA oxidation with an apparent K(m) of 0.84 microM and a V(max) of 5.7 nmol/min/mg protein, for a net intrinsic clearance that was more than 50 times higher than the rat. All species also reduced o-HPA to o-hydroxyphenylethanol (o-HPE), but this was only a major reaction in rats. In the presence of a metabolic reaction replete with all necessary cofactors, GSH conjugation accounted for nearly half of all CE metabolites in rat and mouse, whereas the GSH conjugate represented only 10% of the metabolites in human cytosol. In mouse, o-HPAA represented the major ring-opened metabolite, accounting for the remaining 50% of metabolites, and in human cytosol, o-HPAA was the major metabolite, representing nearly 90% of all CE metabolites. In contrast, no o-HPAA was detected in rats, whereas o-HPE represented a major metabolite. Collectively, these in vitro data implicate o-HPA detoxification through oxidation to o-HPAA as the major determinant of species differences in coumarin-induced hepatotoxicity.