Does CYP2E1 play a major role in the aggravation of isoniazid toxicity by rifampicin in human hepatocytes?

Abstract

Isoniazid and rifampicin are first-line anti-tubercular drugs. In a recent paper, Shen et al. provided interesting findings that rifampicin exacerbated isoniazid toxicity in human hepatocytes but not in rat hepatocytes. The main conclusion was that the difference in cytochrome P450 2E1 (CYP2E1) induction by rifampicin between rat and human hepatocytes accounted for the difference in exacerbation of isoniazid hepatotoxicity by rifampicin. 4-Nitrophenol hydroxylase (4-NP) activity was the only probe of CYP2E1 activity used in the paper. The authors presented data showing that rifampicin enhanced 4-NP activity and CYP2E1 mRNA expression in human hepatocytes, but not in rat hepatocytes. However, CYP3A also makes a significant contribution to 4-NP activity in humans and rats, which has been confirmed by both CYP3A-specific inducer and inhibitors. Rifampicin is a strong inducer of human CYP3A; thus, the increase in 4-NP activity in human hepatocytes could be due to the induction of CYP3A. Rifampicin did not increase 4-NP activities in rat hepatocytes, which could reflect a lack of the induction of rat CYP3A by rifampicin. Additionally, more experiments are needed to support the conclusion that rifampicin increased CYP2E1 mRNA expression in human hepatocytes because of the small sample size and the limitations of semi-quantitative RT-PCR. The study by Shen et al. suggests that another drug-metabolizing enzyme rather than CYP2E1 could be involved in the aggravation of isoniazid toxicity by rifampicin in human hepatocytes.

The 2008 paper by Shen et al. is available from http://www3.interscience.wiley.com/cgi-bin/fulltext/121667602/PDFSTART

Isoniazid and rifampicin are the first-line drugs for the treatment of tuberculosis. As isoniazid and rifampicin are frequently given together to patients, hepatotoxicity is the major concern during the treatment with these drugs. Cytochrome P450s (CYPs) are responsible for the biotransformation of xenobiotics and endogenous compounds. The regulation of CYPs is the most common mechanism that can lead to drug–drug interactions (Kalgutkar et al., 2007). Cytochrome P450 2E1 (CYP2E1) produces free radicals independent of a ligand, which can cause cell damage from lipid peroxidation and DNA strand breaks (Caro and Cederbaum, 2004). Previous studies suggest that hepatic CYP2E1 plays an essential role in isoniazid-induced hepatotoxicity through generation of free radicals (Huang et al., 2003; Yue et al., 2004; Shen et al., 2006).

In a recent paper, Shen et al. (2008) provided interesting findings that rifampicin exacerbated isoniazid toxicity in human hepatocytes but not in rat hepatocytes. The main conclusion was that a difference in CYP2E1 induction by rifampicin between rat and human hepatocytes accounted for the difference in exacerbation of isoniazid hepatotoxicity by rifampicin. This was derived from the observation that rifampicin enhanced 4-nitrophenol hydroxylase (4-NP) activity and CYP2E1 mRNA expression in human hepatocytes, but not in rat hepatocytes. 4-NP was the only probe of CYP2E1 activity used in the paper; however, the specificity of 4-NP is questionable. Earlier work has revealed that another CYP, CYP3A, also makes a significant contribution to 4-NP activity in humans and rats (Zerilli et al., 1997; Zerilli et al., 1998; DiPetrillo et al., 2002; Kobayashi et al., 2002). The role of CYP3A in 4-NP activity was confirmed by CYP3A-specific inducers and inhibitors in intact rats and rat hepatocytes. Dexamethasone is a potent inducer of CYP3A, but not CYP2E1, which can significantly increase hepatic CYP3A protein and CYP3A1 mRNA in rats. The activity of 4-NP was markedly increased in dexamethasone-treated rats, which can be inhibited up to 50% by the specific CYP3A inhibitors, ketoconazole and troleandomycin. Rifampicin is a strong inducer of human CYP3A, which was mentioned in the discussion by Shen et al. (2008). Thus, the induction of CYP3A by rifampicin could be responsible for the increase in 4-NP activity in human hepatocytes. Previous studies have reported that there are species differences in the effects of rifampicin on hepatic CYP3A. In humans and dogs, rifampicin is a potent inducer of CYP3A, but not in rat and mouse, whereas dexamethasone can strongly induce CYP3A in both human and rat (Graham et al., 2002; Martignoni et al., 2004; Vignati et al., 2004; Nishimura et al., 2007). Rifampicin did not increase 4-NP activities in rat hepatocytes, which could be due to the lack of induction of rat CYP3A by rifampicin. This further supports the hypothesis that the induction of human CYP3A by rifampicin would contribute to the increase in 4-NP activity in human hepatocytes.

Another finding by Shen et al. (2008) was that rifampicin increased CYP2E1 mRNA expression in human hepatocytes. In the paper, the hepatocytes for mRNA expression experiment were derived from one donor. Semi-quantitative conventional RT-PCR instead of real-time RT-PCR was used for the measurement, and the times for the repeating experiments have not been mentioned. There are some inconsistent reports that the levels of CYP2E1 mRNA and protein were not increased by rifampicin in human hepatocytes (Mattes and Li, 1997; Rae et al., 2001; Raucy et al., 2004). Because of the small sample size and the limitations of the technical approach, we consider that no firm conclusion can be drawn from this finding.

Does CYP2E1 play a major role in the aggravation of isoniazid toxicity by rifampicin in human hepatocytes? CYP2E1 shows a strong conservation among species (Martignoni et al., 2006). Rat is thought to be a good experimental model of CYP2E1-mediated metabolism, despite some discrepancies (rat CYP2E1 shares 80% identity to human CYP2E1) (Lieber, 1999; Morel et al., 1999; Zuber et al., 2002; Martignoni et al., 2006). The extrapolation for CYP2E1 between species appears to be fairly good. In contrast, the rat is not a good model of CYP3A4-dependent metabolism (Zuber et al., 2002; Martignoni et al., 2006). Shen et al. (2008) found that rifampicin did not induce 4-NP activities in rat hepatocytes and that there was no exacerbation of isoniazid toxicity by the addition of rifampicin. This is consistent with the in vivo results that rifampicin co-administration failed to aggravate isoniazid hepatotoxicity after 21 day treatment by the attenuation of isoniazid-induced CYP2E1 activities (Yue et al., 2004). The study by Shen et al. (2008) suggests that another drug-metabolizing enzyme rather than CYP2E1 could be involved in the aggravation of isoniazid toxicity by rifampicin in human hepatocytes.

In summary, although the finding for the differential effect of rifampicin on isoniazid toxicity in human and rat hepatocytes is certainly welcome, the experiments do not support a clear conclusion that the discrepancy results from a simple difference in CYP2E1 regulation by rifampicin. This paper highlights the difficulties inherent in the choice of both the enzyme probes for catalytic activity studies and the most relevant animal species to humans. Particular care is needed when comparing data on drug metabolism between species.