As retinol was not inducing any of the
As retinol was not inducing any of the isoforms of CYP450 responsible for the bioactivation of paracetamol, we decided to examine retinol's effect on renal and hepatic glutathione to determine whether retinol was potentiating hepatotoxicity through a mechanism of glutathione depletion. Studies, utilising the classical glutathione depletor diethyl maleate and a glutathione synthesis inhibitor buthionine sulfoximine, have demonstrated that under conditions of hepatic glutathione depletion potentiation of paracetamol-induced hepatotoxicity occurs (James et al., 1993). The authors were also able to conclude that phenylpropanolamine potentiated paracetamol-induced hepatotoxicity in mice through a mechanism of moderate (30–50%) glutathione depletion. Similarly, a decrease in renal glutathione content is associated with paracetamol-induced nephrotoxicity (McMurtry et al., 1978, Richie et al., 1992). As retinol potentiated the hepatotoxicity but not the nephrotoxicity of paracetamol, our glutathione investigation would further characterise this organ-specific response. Our results showed that hepatic and renal total reduced glutathione levels were not different from control following retinol treatment. Therefore, retinol is acting through a mechanism independent of both glutathione depletion and CYP450 induction. We can conclude that the mechanism of retinol's potentiation of paracetamol-induced hepatotoxicity does not involve an increased bioactivation of paracetamol through an induction of CYP450 and it also does not involve a decrease in NAPQI conjugation through glutathione depletion. Furthermore, neither of these systems is responsible for the differential response observed between the liver and the kidney. We propose that retinol may be causing potentiation through a mechanism of glucuronide saturation, which would limit the capacity of this major conjugation pathway, and are currently investigating this aspect. However, it is conceivable that retinol treatment impairs glutathione S-transferase activity which, if decreased, could lead to higher levels of NAPQI in the presence of sufficient reduced glutathione. Alternatively, since retinol has been shown to activate Kupffer guanylyl cyclase in rats (Badger et al., 1997, Hoglen et al., 1997), retinol may increase the production of reactive oxygen species in mice which would lead to a subsequent increase in the progression of paracetamol-induced hepatotoxicity. This type of mechanism may also explain the organ-specific response between the liver and the kidney and should also be investigated.
Introduction As our knowledge and understanding of the species differences in the regulation and substrate specificity of the mammalian CYP450 enzymes has broadened significantly, the need for a human-relevant in vitro hepatic model system has become more apparent than ever before. Pharmaceutical scientists have attempted to utilize a number of liver-derived model systems to study drug disposition in vitro, including liver slices, immortalized cell lines, and primary hepatocytes. With the lack of phenotypic gene expression in nearly all immortalized cell lines and the limitations of liver slices, such as short-term viability and diffusional barriers, primary cultures of human hepatocytes have become the ‘gold standard’ for the in vitro testing of drugs. For nearly two decades primary cultures of human hepatocytes have been utilized for toxicological and pharmacological studies (Ferrini et al., 1997, Guillouzo, 1998, Guillouzo et al., 1997, Kern et al., 1997). Since that time an overabundance of literature has emerged that describes the various applications and methods for which they have been utilized, especially for studying drug metabolism and the induction of liver cytochrome P450 enzymes. Within the literature, one can find a number of different in vitro approaches that have been applied successfully for assessing the metabolism and induction potential of drugs (Donato et al., 1995, Maurel, 1996a, Kern et al., 1997, Guillouzo, 1998, Li et al., 1997b, Silva et al., 1998). However, for the novice who is attempting to identify those methods and conditions that are most appropriate for a particular type of study, this may appear initially to be an overwhelming task. Likewise, there are few resources available for obtaining essential information needed to determine which is the best model system and methods for a particular purpose or type of study. This commentary attempts to address some of the more important issues and caveats which must be considered when utilizing primary cultures of human hepatocytes for drug evaluation, especially for long-term studies of gene expression (e.g., CYP450 induction or suppression). The effects of different culture conditions on the restoration and maintenance of normal hepatic structure and function in vitro will be presented, especially as they relate to testing the potential of new drugs to affect liver enzyme expression. The relationship between induction of CYP3A4 in human hepatocytes and the activation of orphan nuclear receptors, such as pregnane X receptor (PXR) and constitutively activated receptor (CAR), by xenobiotics is discussed in light of both species differences in CYP450 regulation and predicting drug interactions in humans. Specific issues and important considerations pertinent to testing NCEs in vitro utilizing primary hepatocyte cultures are discussed. Finally, the development of future guidelines for the standardization of the practices and protocols utilizing this in vitro model system is proposed.