BACKGROUND Renal and hepatic toxicity of the fluorinated ether volatile anesthetics is caused by biotransformation to toxic metabolites. Metabolism also contributes significantly to the elimination pharmacokinetics of some volatile agents. Although innumerable studies have explored anesthetic metabolism in animals, there is little information on human volatile anesthetic metabolism with respect to comparative rates or the identity of the enzymes responsible for defluorination. The first purpose of this investigation was to compare the metabolism of the fluorinated ether anesthetics by human liver microsomes. The second purpose was to test the hypothesis that cytochrome P450 2E1 is the specific P450 isoform responsible for volatile anesthetic defluorination in humans. METHODS Microsomes were prepared from human livers. Anesthetic metabolism in microsomal incubations was measured by fluoride production. The strategy for evaluating the role of P450 2E1 in anesthetic defluorination involved three approaches: for a series of 12 human livers, correlation of microsomal defluorination rate with microsomal P450 2E1 content (measured by Western blot analysis), correlation of defluorination rate with microsomal P450 2E1 catalytic activity using marker substrates (para-nitrophenol hydroxylation and chlorzoxazone 6-hydroxylation), and chemical inhibition by P450 isoform-selective inhibitors. RESULTS The rank order of anesthetic metabolism, assessed by fluoride production at saturating substrate concentrations, was methoxyflurane > sevoflurane > enflurane > isoflurane > desflurane > 0. There was a significant linear correlation of sevoflurane and methoxyflurane defluorination with antigenic P450 2E1 content (r = 0.98 and r = 0.72, respectively), but not with either P450 1A2 or P450 3A3/4. Comparison of anesthetic defluorination with either para-nitrophenol or chlorzoxazone hydroxylation showed a significant correlation for sevoflurane (r = 0.93, r = 0.95) and methoxyflurane (r = 0.78, r = 0.66). Sevoflurane defluorination was also highly correlated with that of enflurane (r = 0.93), which is known to be metabolized by human P450 2E1. Diethyldithiocarbamate, a selective inhibitor of P450 2E1, produced a concentration-dependent inhibition of sevoflurane, methoxyflurane, and isoflurane defluorination. No other isoform-selective inhibitor diminished the defluorination of sevoflurane, whereas methoxyflurane defluorination was inhibited by the selective P450 inhibitors furafylline (P450 1A2), sulfaphenazole (P450 2C9/10), and quinidine (P450 2D6) but to a much lesser extent than by diethyldithiocarbamate. CONCLUSIONS These results demonstrate that cytochrome P450 2E1 is the principal, if not sole human liver microsomal enzyme catalyzing the defluorination of sevoflurane. P450 2E1 is the principal, but not exclusive enzyme responsible for the metabolism of methoxyflurane, which also appears to be catalyzed by P450s 1A2, 2C9/10, and 2D6. The data also suggest that P450 2E1 is responsible for a significant fraction of isoflurane metabolism. Identification of P450 2E1 as the major anesthetic metabolizing enzyme in humans provides a mechanistic understanding of clinical fluorinated ether anesthetic metabolism and toxicity.