1 The cytotoxicity of metabolites generated from phenytoin, sorbinil and mianserin by human and mouse liver microsomes was assessed by co-incubation with human mononuclear leucocytes as target cells. Cytotoxicity was determined by trypan blue dye exclusion. 2 Phenytoin and sorbinil were metabolised by NADPH-dependent murine microsomal enzymes to cytotoxic metabolites. Cytotoxicity produced by both drugs was significantly enhanced by the epoxide hydrolase inhibitor trichloropropane oxide (TCPO). No significant cytotoxicity was observed in the presence of human liver microsomes. 3 Mianserin was metabolised by both human and mouse liver microsomes to a cytotoxin. Cytotoxicity was greater in the presence of human liver microsomes (13.7 ± 2.2%; mean + s.d. for four livers, compared with 6.0 ± 2.4%, mean ± s.d., n = 4, with mouse liver microsomes), and was unaffected by pretreatment with TCPO. 4 Stable metabolites were quantified by radiometric high performance liquid chromatography. Phenytoin and sorbinil were metabolised to 5-(p-hydroxyphenyl)-5-phenylhydantoin (0.3-0.5% of incubated radioactivity) and 2-hydroxysorbinil (0.4-2.7% of incubated radioactivity), respectively, by both human and mouse liver microsomes. 5 Mianserin was metabolised to 8-hydroxymianserin and desmethylmianserin by both human and mouse liver microsomes. Desmethylmianserin was the major product in incubations with human liver microsomes (32.3 ± 12%, mean ± s.d. for four livers), whereas 8-hydroxymianserin was the predominant metabolite generated by mouse liver microsomes (25.9 ± 1.5%, mean ± s.d., n = 4). 6 Generation of electrophilic metabolites was assessed by determination of the amount of radiolabelled material which became irreversibly bound to protein. Only mouse liver microsomes activated phenytoin to a chemically reactive metabolite, whereas both mouse and human liver microsomes generated reactive metabolites from sorbinil and mianserin. 7 These studies show that drug cytotoxicity can be mediated by low concentrations (circa F.M) of metabolites generated by NADPH-dependent hepatic microsomal enzymes; however demonstration of cytotoxicity in vitro has not been established as a means of predicting in vivo toxicity.

No abstract

The formation of cytotoxic metabolites from the anticonvulsants phenytoin and carbamazepine was investigated in vitro using a hepatic microsomal enzyme system and human mononuclear leucocytes as target cells. Both drugs were metabolised to cytotoxic products. In order to assess the structural requirements for this bioactivation, a series of structurally related compounds was investigated. It was found that molecules which contain either an amide function or an aryl ring may undergo activation in vitro, but only the metabolism‐dependent toxicity of the latter is potentiated by pre‐treatment of the target cells with an epoxide hydrolase inhibitor. Taken collectively, these data are consistent with the concept that reactive epoxide metabolites of both phenytoin and carbamazepine may produce toxicity in individuals with an inherited deficiency in epoxide hydrolase.

1 A two compartment system, comprising two adjacent teflon chambers separated by a semi-permeable membrane, has been devised with which to investigate the generation of drug metabolites that are toxic to human cells in vitro. 2 Compartment A contained a drug-metabolising system (human liver microsomes ± NADPH) and compartment B contained target cells (human mononuclear leucocytes). The semi-permeable membrane retained protein (m.w. > 10,000) but allowed equilibration (within 1 h) of drug and drug metabolites, during which time cells remained viable.3 Incubation of dapsone (100 ,UM) with human microsomal protein (2 mg mI-') and NADPH (1 mM) in compartment A caused cell death (8.7 ± 1.8%) in compartment B, which was reduced significantly (P < 0.05) by the addition of glutathione (500 ,.M). Dapsone in the absence of NADPH was not cytotoxic. 4 Chemical analysis showed the presence of dapsone hydroxylamine as the only stable metabolite in both compartment A (5.2 ± 0.4% incubated drug) and compartment B (3.5 ± 0.5%).5 Irreversible binding of dapsone to cells was significantly (P < 0.05) reduced by omission of NADPH (85 ± 13 pmol/106 cells) or addition of glutathione (103 ± 9) compared with control values (153 ± 51).

1. The metabolism of the enantiomers of mianserin to stable, chemically reactive and cytotoxic metabolites by human liver microsomes has been investigated in vitro. 2. Both enantiomers were metabolised to three major oxidation products: 8‐hydroxymianserin, desmethylmianserin and mianserin 2‐oxide. Hydroxylation occurred more readily with the S‐ enantiomer, whereas desmethylmianserin was always the major metabolite of the R‐enantiomer. 3. The generation of chemically reactive metabolites exhibited a marginal degree of stereoselectivity, as assessed by irreversible binding of drug to microsomal protein (S greater than or equal to R; P less than or equal to 0.05). 4. The formation of metabolites which were cytotoxic towards human mononuclear leucocytes was greater (P less than or equal to 0.001] for R(‐)‐ mianserin than for S(+)‐mianserin and showed a significant correlation with N‐demethylation (r = 0.84, P less than or equal to 0.01).