Characterization of ritonavir-mediated inactivation of cytochrome P450 3A4

Ritonavir is a human immunodeficiency virus (HIV) protease inhibitor and an inhibitor of cytochrome P450 3A4, the major human hepatic drug-metabolizing enzyme. Given the potent inhibition of CYP3A4 by ritonavir, subtherapeutic doses of ritonavir are used to increase plasma concentrations of other HIV drugs oxidized by CYP3A4, thereby extending their clinical efficacy. However, the mechanism of inhibition of CYP3A4 by ritonavir remains unclear. To date, data suggests multiple types of inhibition by ritonavir, including mechanism-based inactivation by metabolic-intermediate complex formation, competitive inhibition, irreversible type II coordination to the heme iron, and more recently heme destruction. The results presented here demonstrate that inhibition of CYP3A4 by ritonavir occurs by CYP3A4-mediated activation and subsequent formation of a covalent bond to the apoprotein. Incubations of [(3)H]ritonavir with reconstituted CYP3A4 and human liver microsomes resulted in a covalent binding stoichiometry equal to 0.93 ± 0.04 moles of ritonavir bound per mole of inactivated CYP3A4. The metabolism of [(3)H]ritonavir by CYP3A4 leads to the formation of a covalent adduct specifically to CYP3A4, confirmed by radiometric liquid chromatography-trace and whole-protein mass spectrometry. Tryptic digestion of the CYP3A4-[(3)H]ritonavir incubations exhibited an adducted peptide (255-RM K: ESRLEDTQKHR-268) associated with a radiochromatic peak and a mass consistent with ritonavir plus 16 Da, in agreement with the whole-protein mass spectrometry. Additionally, nucleophilic trapping agents and scavengers of free oxygen species did not prevent inactivation of CYP3A4 by ritonavir. In conclusion, ritonavir exhibited potent time-dependent inactivation of CYP3A, with the mechanism of inactivation occurring though a covalent bond to Lys257 of the CYP3A4 apoprotein.

Fluoxetine- and norfluoxetine-mediated complex drug-drug interactions: in vitro to in vivo correlation of effects on CYP2D6, CYP2C19, and CYP3A4

Fluoxetine and its circulating metabolite norfluoxetine comprise a complex multiple-inhibitor system that causes reversible or time-dependent inhibition of the cytochrome P450 (CYP) family members CYP2D6, CYP3A4, and CYP2C19 in vitro. Although significant inhibition of all three enzymes in vivo was predicted, the areas under the concentration-time curve (AUCs) for midazolam and lovastatin were unaffected by 2-week dosing of fluoxetine, whereas the AUCs of dextromethorphan and omeprazole were increased by 27- and 7.1-fold, respectively. This observed discrepancy between in vitro risk assessment and in vivo drug-drug interaction (DDI) profile was rationalized by time-varying dynamic pharmacokinetic models that incorporated circulating concentrations of fluoxetine and norfluoxetine enantiomers, mutual inhibitor-inhibitor interactions, and CYP3A4 induction. The dynamic models predicted all DDIs with less than twofold error. This study demonstrates that complex DDIs that involve multiple mechanisms, pathways, and inhibitors with their metabolites can be predicted and rationalized via characterization of all the inhibitory species in vitro.

Characterization of inhibition kinetics of (S)-warfarin hydroxylation by noscapine: implications in warfarin therapy

Noscapine is an antitussive and potential anticancer drug. Clinically significant interactions between warfarin and noscapine have been previously reported. In this study, to provide a basis for warfarin dosage adjustment, the inhibition kinetics of noscapine against warfarin metabolism was characterized. Our enzyme kinetics data obtained from human liver microsomes and recombinant CYP2C9 proteins indicate that noscapine is a competitive inhibitor of the (S)-warfarin 7-hydroxylation reaction by CYP2C9. Interestingly, noscapine also inhibited (S)-warfarin metabolism in a NADPH- and time-dependent manner, and removal of unbound noscapine and its metabolites by ultrafiltration did not reverse inhibition of (S)-warfarin metabolism by noscapine, suggesting mechanism-based inhibition of CYP2C9 by noscapine. Spectral scanning of the reaction between CYP2C9 and noscapine revealed the formation of an absorption spectrum at 458 nm, indicating the formation of a metabolite-intermediate complex. Surprisingly, noscapine is a 2- to 3-fold more efficient inactivator of CYP2C9.2 and CYP2C9.3 variants than it is of the wild type, by unknown mechanisms. Based on the inhibitory kinetic data, (S)-warfarin exposure is predicted to increase up to 7-fold (depending on CYP2C9 genotypes) upon noscapine coadministration, mainly due to mechanism-based inactivation of CYP2C9 by noscapine. Together, these results indicate that mechanism-based inhibition of CYP2C9 by noscapine may dramatically alter pharmacokinetics of warfarin and provide a basis for warfarin dosage adjustment when noscapine is coadministered.

Stereoselective inhibition of CYP2C19 and CYP3A4 by fluoxetine and its metabolite: implications for risk assessment of multiple time-dependent inhibitor systems

Recent guidance on drug-drug interaction (DDI) testing recommends evaluation of circulating metabolites. However, there is little consensus on how to quantitatively predict and/or assess the risk of in vivo DDIs by multiple time-dependent inhibitors (TDIs) including metabolites from in vitro data. Fluoxetine was chosen as the model drug to evaluate the role of TDI metabolites in DDI prediction because it is a TDI of both CYP3A4 and CYP2C19 with a circulating N-dealkylated inhibitory metabolite, norfluoxetine. In pooled human liver microsomes, both enantiomers of fluoxetine and norfluoxetine were TDIs of CYP2C19, (S)-norfluoxetine was the most potent inhibitor with time-dependent inhibition affinity constant (KI) of 7 μM, and apparent maximum time-dependent inhibition rate (k(inact,app)) of 0.059 min(-1). Only (S)-fluoxetine and (R)-norfluoxetine were TDIs of CYP3A4, with (R)-norfluoxetine being the most potent (K(I) = 8 μM, and k(inact,app) = 0.011 min(-1)). Based on in-vitro-to-in-vivo predictions, (S)-norfluoxetine plays the most important role in in vivo CYP2C19 DDIs, whereas (R)-norfluoxetine is most important in CYP3A4 DDIs. Comparison of two multiple TDI prediction models demonstrated significant differences between them in in-vitro-to-in-vitro predictions but not in in-vitro-to-in-vivo predictions. Inclusion of all four inhibitors predicted an in vivo decrease in CYP2C19 (95%) and CYP3A4 (60-62%) activity. The results of this study suggest that adequate worst-case risk assessment for in vivo DDIs by multiple TDI systems can be achieved by incorporating time-dependent inhibition by both parent and metabolite via simple addition of the in vivo time-dependent inhibition rate/cytochrome P450 degradation rate constant (λ/k(deg)) values, but quantitative DDI predictions will require a more thorough understanding of TDI mechanisms.

Risk assessment of mechanism-based inactivation in drug-drug interactions

Drug-drug interactions (DDIs) that occur via mechanism-based inactivation of cytochrome P450 are of serious concern. Although several predictive models have been published, early risk assessment of MBIs is still challenging. For reversible inhibitors, the DDI risk categorization using [I]/K(i) ([I], the inhibitor concentration; K(i), the inhibition constant) is widely used in drug discovery and development. Although a simple and reliable methodology such as [I]/K(i) categorization for reversible inhibitors would be useful for mechanism-based inhibitors (MBIs), comprehensive analysis of an analogous measure reflecting in vitro potency for inactivation has not been reported. The aim of this study was to evaluate whether the term λ/k(deg) (λ, first-order inactivation rate at a given MBI concentration; k(deg), enzyme degradation rate constant) would be useful in the prediction of the in vivo DDI risk of MBIs. Twenty-one MBIs with both in vivo area under the curve (AUC) change of marker substrates and in vitro inactivation parameters were identified in the literature and analyzed. The results of this analysis show that in vivo DDIs with >2-fold change of object drug AUC can be identified with the cutoff value of λ/k(deg) = 1, where unbound steady-state C(max) is used for inhibitor concentration. However, the use of total C(max) led to great overprediction of DDI risk. The risk assessment using λ/k(deg) coupled with unbound C(max) can be useful for the DDI risk evaluation of MBIs in drug discovery and development.

Stereospecific metabolism of itraconazole by CYP3A4: dioxolane ring scission of azole antifungals

Itraconazole (ITZ) is a mixture of four cis-stereoisomers that inhibit CYP3A4 potently and coordinate CYP3A4 heme via the triazole nitrogen. However, (2R,4S,2'R)-ITZ and (2R,4S,2'S)-ITZ also undergo stereoselective sequential metabolism by CYP3A4 at a site distant from the triazole ring to 3'-OH-ITZ, keto-ITZ, and N-desalkyl-ITZ. This stereoselective metabolism demonstrates specific interactions of ITZ within the CYP3A4 active site. To further investigate this process, the binding and metabolism of the four trans-ITZ stereoisomers by CYP3A4 were characterized. All four trans-ITZ stereoisomers were tight binding inhibitors of CYP3A4-mediated midazolam hydroxylation (IC(50) 16-26 nM), and each gave a type II spectrum upon binding to CYP3A4. However, instead of formation of 3'-OH-ITZ, they were oxidized at the dioxolane ring, leading to ring scission and formation of two new metabolites of ITZ. These two metabolites were also formed from the four cis-ITZ stereoisomers, although not as efficiently. The catalytic rates of dioxolane ring scission were similar to the dissociation rates of ITZ stereoisomers from CYP3A4, suggesting that the heme iron is reduced while the triazole moiety coordinates to it and no dissociation of ITZ is necessary before catalysis. The triazole containing metabolite [1-(2,4-dichlorophenyl)-2-(1H-1,2,4-triazol-1-yl)ethanone] also inhibited CYP3A4 (IC(50) >15 μM) and showed type II binding with CYP3A4. The dioxolane ring scission appears to be clinically relevant because this metabolite was detected in urine samples from subjects that had been administered the mixture of cis-ITZ isomers. These data suggest that the dioxolane ring scission is a metabolic pathway for drugs that contain this moiety.

Allosteric activation of cytochrome P450 3A4 by α-naphthoflavone: branch point regulation revealed by isotope dilution analysis

Cytochrome P450 3A4 (CYP3A4) is the dominant xenobiotic metabolizing CYP. Despite great interest in CYP enzymology, two in vitro aspects of CYP3A4 catalysis are still not well understood, namely, sequential metabolism and allosteric activation. We have therefore investigated such a system in which both phenomena are present. Here we report that the sequential metabolism of Nile Red (NR) is accelerated by the heterotropic allosteric effector α-naphthoflavone (ANF). ANF increases the rates of formation for NR metabolites M1 and M2 and also perturbs the metabolite ratio in favor of M2. Thus, ANF has as an allosteric effect on a kinetic branch point. Co-incubating deuterium-labeled NR and unlabeled M1, we show that ANF increases k(cat)/k(off) ~1.8-fold in favor of the k(cat) of M2 production. Steady-state metabolic experiments are analyzed using a kinetic model in which the enzyme and substrates are not in rapid equilibrium, and this distinction allows for the estimation of rates of catalysis for the formation of both the primary (M1) and secondary (M2) products, as well as the partitioning of enzyme between these states. These results are compared with those of earlier spectroscopic investigations of NR and ANF cooperativity, and a mechanism of ANF heteroactivation is presented that involves effects on substrate off rate and coupling efficiency.

Evaluation of 6β-hydroxycortisol, 6β-hydroxycortisone, and a combination of the two as endogenous probes for inhibition of CYP3A4 in vivo

An endogenous probe for CYP3A activity would be useful for early identification of in vivo cytochrome P450 (CYP) 3A4 inhibitors. The aim of this study was to determine whether formation clearance (CL(f)) of the sum of 6β-hydroxycortisol and 6β-hydroxycortisone is a useful probe of CYP3A4 inhibition in vivo. In human liver microsomes (HLMs), the formation of 6β-hydroxycortisol and 6β-hydroxycortisone was catalyzed by CYP3A4, and itraconazole inhibited these reactions with half maximal inhibitory concentration (IC(50))(,u) values of 3.1 nmol/l and 3.4 nmol/l, respectively. The in vivo IC(50,u) value of itraconazole for the combined CL(f) of 6β-hydroxycortisone and 6β-hydroxycortisol was 1.6 nmol/l. The greater inhibitory potency in vivo is probably due to circulating inhibitory itraconazole metabolites. The maximum in vivo inhibition was 59%, suggesting that f(m,CYP3A4) for cortisol and cortisone 6β-hydroxylation is ~60%. Given the significant decrease in CL(f) of 6β-hydroxycortisone and 6β-hydroxycortisol after 200-mg and 400-mg single doses of itraconazole, this endogenous probe can be used to detect moderate and potent CYP3A4 inhibition in vivo.

Accurate prediction of dose-dependent CYP3A4 inhibition by itraconazole and its metabolites from in vitro inhibition data

Inhibitory drug metabolites may contribute to drug-drug interactions (DDIs). The aim of this study was to determine the importance of inhibitory metabolites of itraconazole (ITZ) in in vivo cytochrome P450 (CYP) 3A4 inhibition. The pharmacokinetics of ITZ and midazolam (MDZ) were determined in six healthy volunteers in four sessions after administration of MDZ with and without oral ITZ. After doses of 50, 200, and 400 mg of ITZ, the clearance of orally administered MDZ decreased by 27, 74, and 83%, respectively. The in vivo half maximal inhibitory concentration (IC(50)) for ITZ ranged from 5 to 132 nmol/l in the six subjects. The metabolites of ITZ were estimated to account for ~50% of the total CYP3A4 inhibition, with the relative contribution increasing with time after ITZ dosing. Of the total of 18 interactions observed, 15 (84%) could be predicted within a twofold error margin, with improved accuracy observed when ITZ metabolites were included in the predictions. This study shows that the metabolites of ITZ contribute to CYP3A4 inhibition and need to be accounted for in quantitative rationalization of ITZ-mediated DDIs.

Sequential metabolism of secondary alkyl amines to metabolic-intermediate complexes: opposing roles for the secondary hydroxylamine and primary amine metabolites of desipramine, (s)-fluoxetine, and N-desmethyldiltiazem

Three secondary amines desipramine (DES), (S)-fluoxetine [(S)-FLX], and N-desmethyldiltiazem (MA) undergo N-hydroxylation to the corresponding secondary hydroxylamines [N-hydroxydesipramine, (S)-N-hydroxyfluoxetine, and N-hydroxy-N-desmethyldiltiazem] by cytochromes P450 2C11, 2C19, and 3A4, respectively. The expected primary amine products, N-desmethyldesipramine, (S)-norfluoxetine, and N,N-didesmethyldiltiazem, are also observed. The formation of metabolic-intermediate (MI) complexes from these substrates and metabolites was examined. In each example, the initial rates of MI complex accumulation followed the order secondary hydroxylamine > secondary amine >> primary amine, suggesting that the primary amine metabolites do not contribute to formation of MI complexes from these secondary amines. Furthermore, the primary amine metabolites, which accumulate in incubations of the secondary amines, inhibit MI complex formation. Mass balance studies provided estimates of the product ratios of N-dealkylation to N-hydroxylation. The ratios were 2.9 (DES-CYP2C11), 3.6 [(S)-FLX-CYP2C19], and 0.8 (MA-CYP3A4), indicating that secondary hydroxylamines are significant metabolites of the P450-mediated metabolism of secondary alkyl amines. Parallel studies with N-methyl-d(3)-desipramine and CYP2C11 demonstrated significant isotopically sensitive switching from N-demethylation to N-hydroxylation. These findings demonstrate that the major pathway to MI complex formation from these secondary amines arises from N-hydroxylation rather than N-dealkylation and that the primary amines are significant competitive inhibitors of MI complex formation.

Mechanism of formation of the ester linkage between heme and Glu310 of CYP4B1: 18O protein labeling studies

Cytochrome P450s in the CYP4 family covalently bind their heme prosthetic group to a conserved acidic I-helix residue via an autocatalytic oxidation. This study was designed to evaluate the source of oxygen atoms in the covalent ester link in CYP4B1 enzymes labeled with [18O]glutamate and [18O]aspartate. The fate of the heavy isotope was then traced into wild-type CYP4B1 or the E310D mutant-derived 5-hydroxyhemes. Glutamate-containing tryptic peptides of wild-type CYP4B1 were found labeled to a level of 11-13% 18O. Base hydrolysis of labeled protein released 5-hydroxyheme which contained 12.8 +/- 1.9% 18O. Aspartate-containing peptides of the E310D mutant were labeled with 6.0-6.5% 18O, but as expected, no label was transmitted to recovered 5-hydroxyheme. These data demonstrate that the oxygen atom in 5-hydroxyheme derived from wild-type CYP4B1 originates in Glu310. Stoichiometric incorporation of the heavy isotope from the wild-type enzyme supports a perferryl-initiated carbocation mechanism for covalent heme formation in CYP4B1.

Contribution of itraconazole metabolites to inhibition of CYP3A4 in vivo

Itraconazole (ITZ) is metabolized in vitro to three inhibitory metabolites: hydroxy-itraconazole (OH-ITZ), keto-itraconazole (keto-ITZ), and N-desalkyl-itraconazole (ND-ITZ). The goal of this study was to determine the contribution of these metabolites to drug-drug interactions caused by ITZ. Six healthy volunteers received 100 mg ITZ orally for 7 days, and pharmacokinetic analysis was conducted at days 1 and 7 of the study. The extent of CYP3A4 inhibition by ITZ and its metabolites was predicted using this data. ITZ, OH-ITZ, keto-ITZ, and ND-ITZ were detected in plasma samples of all volunteers. A 3.9-fold decrease in the hepatic intrinsic clearance of a CYP3A4 substrate was predicted using the average unbound steady-state concentrations (C(ss,ave,u)) and liver microsomal inhibition constants for ITZ, OH-ITZ, keto-ITZ, and ND-ITZ. Accounting for circulating metabolites of ITZ significantly improved the in vitro to in vivo extrapolation of CYP3A4 inhibition compared to a consideration of ITZ exposure alone.

Sequential metabolism is responsible for diltiazem-induced time-dependent loss of CYP3A

Kinetic parameters (k(inact) and K(I)) obtained in microsomes are often used to predict time-dependent inactivation. We previously reported that microsomal inactivation kinetic parameters of diltiazem underpredicted CYP3A inactivation in hepatocytes. In this study, we evaluated the contributions of inactivation and reversible inhibition of CYP3A by diltiazem and its N-desmethyl (MA) and N,N-didesmethyl (MD) metabolites. In human liver microsomes, MA was a more potent time-dependent inactivator of CYP3A than its parent drug, with apparent k(inact) approximately 4-fold higher than that of diltiazem at a microsomal protein concentration of 0.2 mg/ml. MD did not inactivate CYP3A. Inactivation of CYP3A by diltiazem was dependent on microsomal protein concentration (25, 36, and 41% decrease in CYP3A activity at 0.2, 0.4, and 0.8 mg/ml microsomal protein, respectively, incubated with 10 microM diltiazem over 20 min), whereas inactivation by MA did not seem to be protein concentration-dependent. MA and MD were reversible inhibitors of CYP3A with competitive Ki values of 2.7 and 0.2 microM, respectively. In cryopreserved hepatocytes incubated with diltiazem, time-dependent loss of CYP3A was accompanied by increased formation of MA and MD, with the MA level similar to its K(I) at higher diltiazem concentrations. In addition, the metabolites appeared to be accumulated inside the cells. In summary, time-dependent CYP3A inactivation by MA seems to be the major contributor responsible for the loss of CYP3A in human liver microsomes and human hepatocytes incubated with diltiazem. These findings suggest that prediction of CYP3A loss based solely on microsomal inactivation parameters of parent drug may be inadequate.

Surface plasmon resonance analysis of antifungal azoles binding to CYP3A4 with kinetic resolution of multiple binding orientations

The heme-containing cytochrome P450s (CYPs) are a major enzymatic determinant of drug clearance and drug-drug interactions. The CYP3A4 isoform is inhibited by antifungal imidazoles or triazoles, which form low-spin heme iron complexes via formation of a nitrogen-ferric iron coordinate bond. However, CYP3A4 also slowly oxidizes the antifungal itraconazole (ITZ) at a site that is approximately 25 A from the triazole nitrogens, suggesting that large antifungal azoles can adopt multiple orientations within the CYP3A4 active site. Here, we report a surface plasmon resonance (SPR) analysis with kinetic resolution of two binding modes of ITZ, and the related drug ketoconazole (KTZ). SPR reveals a very slow off-rate for one binding orientation. Multiphasic binding kinetics are observed, and one of the two binding components resolved by curve fitting exhibits "equilibrium overshoot". Preloading of CYP3A4 with the heme ligand imidazole abolishes this component of the antifungal azole binding trajectories, and it eliminates the conspicuously slow off-rate. The fractional populations of CYP3A4 complexes corresponding to different drug orientations can be manipulated by altering the duration of the pulse of drug exposure. UV-vis difference absorbance titrations yield low-spin spectra and K(D) values that are consistent with the high-affinity complex resolved by SPR. These results demonstrate that ITZ and KTZ bind in multiple orientations, including a catalytically productive mode and a slowly dissociating inhibitory mode. Most importantly, they provide the first example of a SPR-based method for the kinetic characterization of binding of a drug to any human CYP, including mechanistic insight not available from other methods.

Sites of covalent attachment of CYP4 enzymes to heme: evidence for microheterogeneity of P450 heme orientation

Typical cytochrome P450s secure the heme prosthetic group with a cysteine thiolate ligand bound to the iron, electrostatic interactions with the heme propionate carboxylates, and hydrophobic interactions with the heme periphery. In addition to these interactions, CYP4B1 covalently binds heme through a monoester link furnished, in part, by a conserved I-helix acid, Glu310. Chromatography, mass spectrometry, and NMR have now been utilized to identify the site of attachment on the heme. Native CYP4B1 covalently binds heme solely at the C-5 methyl position. Unexpectedly, recombinant CYP4B1 from insect cells and Escherichia coli also bound their heme covalently at the C-8 methyl position. Structural heterogeneity may be common among recombinant CYP4 proteins because CYP4A3 exhibited this duality. Attempts to evaluate functional heterogeneity were complicated by the complexity of the system. The phenomenon of covalent heme binding to P450 provides a novel method for assessing microheterogeneity in heme orientation and raises questions about the fidelity of heme incorporation in recombinant systems.

STEREOCHEMICAL ASPECTS OF ITRACONAZOLE METABOLISM IN VITRO AND IN VIVO

Itraconazole (ITZ) has three chiral centers and is administered clinically as a mixture of four stereoisomers. This study evaluated stereoselectivity in ITZ metabolism. In vitro experiments were carried out using heterologously expressed CYP3A4. Only (2R,4S,2'R)-ITZ and (2R,4S,2'S)-ITZ were metabolized by CYP3A4 to hydroxy-ITZ, keto-ITZ, and N-desalkyl-ITZ. When (2S,4R,2'R)-ITZ or (2S,4R,2'S)-ITZ was incubated with CYP3A4, neither metabolites nor substrate depletion were detected. Despite these differences in metabolism, all four ITZ stereoisomers induced a type II binding spectrum with CYP3A4, characteristic of coordination of the triazole nitrogen to the heme iron (K(s) 2.2-10.6 nM). All four stereoisomers of ITZ inhibited the CYP3A4-catalyzed hydroxylation of midazolam with high affinity (IC(50) 3.7-14.8 nM). Stereochemical aspects of ITZ pharmacokinetics were evaluated in six healthy volunteers after single and multiple oral doses. In vivo, after a single dose, ITZ disposition was stereoselective, with a 3-fold difference in C(max) and a 9-fold difference in C(min) between the (2R,4S)-ITZ and the (2S,4R)-ITZ pairs of diastereomers, with the latter reaching higher concentrations. Secondary and tertiary ITZ metabolites (keto-ITZ and N-desalkyl-ITZ) detected in plasma were of the (2R,4S) stereochemistry. After multiple doses of ITZ, the difference in C(max) and C(min) decreased to 1.5- and 3.8-fold, respectively. The initial difference between the stereoisomeric pairs was most likely due to stereoselective metabolism by CYP3A4, including stereoselective first-pass metabolism as well as stereoselective elimination. However, stereoselective elimination was diminished after multiple dosing, presumably as a result of CYP3A4 autoinhibition. In conclusion, the metabolism of ITZ is highly stereoselective in vitro and in vivo.

Cooperative Binding of Midazolam with Testosterone and α-Naphthoflavone within the CYP3A4 Active Site: A NMR T1 Paramagnetic Relaxation Study

Recent studies have indicated that CYP3A4 exhibits non-Michaelis-Menten kinetics for numerous substrates. Both homo- and heterotropic activation have been reported, and kinetic models have suggested multiple substrates within the active site. We provide some of the first physicochemical data supporting the hypothesis of allosteric substrate binding within the CYP3A4 active site. Midazolam (MDZ) is metabolized by CYP3A4 to two hydroxylated metabolites, 1'- and 4-hydroxymidazolam. Incubations using purified CYP3A4 and MDZ showed that both alpha-naphthoflavone (alpha-NF) and testosterone affect the ratio of formation rates of 1'- and 4-hydroxymidazolam. Similar to previous reports, alpha-NF was found to promote formation of 1'-hydroxymidazolam, while testosterone stimulated formation of 4-hydroxymidazolam. NMR was used to measure the closest approach of individual MDZ protons to the paramagnetic heme iron of CYP3A4 using paramagnetic T(1) relaxation measurements. Solutions of 0.2 microM CYP3A4 with 500 microM MDZ resulted in calculated distances between 7.4 and 8.3 A for all monitored MDZ protons. The distances were statistically equivalent for all protons except C3-H and were consistent with the rotation within the active site or sliding parallel to the heme plane. When 50 microM alpha-NF was added, proton-heme iron distances ranged from 7.3 to 10.0 A. Consistent with kinetics of activation, the 1' position was situated closest to the heme, while the fluorophenyl 5-H proton was the furthest. Proton-heme iron distances for MDZ with CYP3A4 and 50 microM testosterone ranged from 7.7 to 9.0 A, with the flourophenyl 5-H proton furthest from the heme iron and the C4-H closest to the heme, also consistent with kinetic observations. When titrated with CYP3A4 in the presence of MDZ, testosterone and alpha-NF resonances themselves exhibited significant broadening and enhanced relaxation rates, indicating that these effector molecules were also bound within the CYP3A4 active site near the paramagnetic heme iron. These results suggest that the effector exerts its cooperative effects on MDZ metabolism through simultaneous binding of MDZ and effector near the CYP3A4 heme.

Oxidation of Caffeine by CYP1A2: Isotope Effects and Metabolic Switching

Caffeine (1,3,7-trimethylxanthine) has previously been shown to undergo metabolic switching in vivo when the N-1 or the N-7 methyl groups were trideuteromethylated [Horning et al. (1976) Proceedings of the Second International Conference on Stable Isotopes, pp 41-54]. We have examined the effect of replacing the N-3 methyl group with a trideuteromethyl group. The corresponding isotope effects can then be used to distinguish the kinetic mechanism by which four primary metabolites can be formed from one substrate by one cytochrome P450 (P450). We have synthesized 3-CD3-caffeine and 3-CD3-7-CD3-caffeine as well as trideuteromethylated analogs of each of the in vitro metabolites formed by cytochrome P4501A2. The observed competitive isotope effects for the metabolites, which do not result from deuterium abstraction (theobromine, theophylline), demonstrate that the nondissociative mechanism applies to caffeine metabolism by cytochrome P4501A2. Thus, there must be equilibration of the kinetically distinguishable activated P450-substrate complexes at rates competitive with hydrogen abstraction. The true isotope effects for the N-3 demethylation of caffeine were derived from the ratios of the amount of paraxanthine relative to the amount of theobromine or theophylline. The resultant ratios indicate that these isotope effects are essentially intrinsic. Observation of the isotope effects on N-3 demethylation was facilitated by branching to the minor in vitro metabolites as well as water formation. Product release is not rate-limiting for this system.

EVALUATION OF TIME-DEPENDENT INACTIVATION OF CYP3A IN CRYOPRESERVED HUMAN HEPATOCYTES

Irreversible CYP3A inhibition by drugs constitutes one of the major causes of inhibition-based drug interactions. We evaluated time-dependent inactivation of CYP3A in cryopreserved human hepatocytes for six structurally diverse compounds known to exhibit this property. Inactivation kinetic parameters were also determined using human liver microsomes. Except for diclofenac, which did not cause CYP3A inactivation either in microsomes or in hepatocytes at concentrations up to 100 microM, time-dependent inactivation was observed in hepatocytes for amprenavir, diltiazem, erythromycin, raloxifene, and troleandomycin. The observed inactivation potency in hepatocytes (observed IC50) was compared with the potency predicted using microsomal parameters (predicted IC50). Despite satisfactory prediction for troleandomycin (1.35 and 2.14 microM for the predicted and observed IC50, respectively), over-prediction of inactivation was observed for raloxifene, amprenavir, and erythromycin (observed IC50 values 6.2-, 55-, and 7.8-fold higher, respectively, than the predicted IC50). By contrast, the observed IC50 for diltiazem in hepatocytes was approximately 4-fold lower than the IC50 predicted from microsomal data (under-prediction). After correcting for factors including nonspecific binding and inactivator consumption, prediction was significantly improved for raloxifene (the observed IC50 then became 2-fold higher than the predicted IC50) and for amprenavir to a lesser extent. A specific P-glycoprotein inhibitor, 4-(6,7-dimethoxy-3,4-dihydroisoquinolin-2(1H)-yl)-N-[2-(3.4-dimethoxyphenyl)ethyl]-6,7-dimethoxyquinazolin-2-amine (CP-100356), modulated the observed CYP3A inactivation potency by erythromycin and troleandomycin. In summary, these studies reveal three important factors that must be considered when microsomal inactivation parameters are used to predict inhibition-based drug interactions in intact cell systems.

ROLE OF ITRACONAZOLE METABOLITES IN CYP3A4 INHIBITION

Itraconazole (ITZ) is a potent inhibitor of CYP3A in vivo. However, unbound plasma concentrations of ITZ are much lower than its reported in vitro Ki, and no clinically significant interactions would be expected based on a reversible mechanism of inhibition. The purpose of this study was to evaluate the reasons for the in vitro-in vivo discrepancy. The metabolism of ITZ by CYP3A4 was studied. Three metabolites were detected: hydroxy-itraconazole (OH-ITZ), a known in vivo metabolite of ITZ, and two new metabolites: keto-itraconazole (keto-ITZ) and N-desalkyl-itraconazole (ND-ITZ). OHITZ and keto-ITZ were also substrates of CYP3A4. Using a substrate depletion kinetic approach for parameter determination, ITZ exhibited an unbound K(m) of 3.9 nM and an intrinsic clearance (CLint) of 69.3 ml.min(-1).nmol CYP3A4(-1). The respective unbound Km values for OH-ITZ and keto-ITZ were 27 nM and 1.4 nM and the CLint values were 19.8 and 62.5 ml.min(-1).nmol CYP3A4(-1). Inhibition of CYP3A4 by ITZ, OH-ITZ, keto-ITZ, and ND-ITZ was evaluated using hydroxylation of midazolam as a probe reaction. Both ITZ and OH-ITZ were competitive inhibitors of CYP3A4, with unbound Ki (1.3 nM for ITZ and 14.4 nM for OH-ITZ) close to their respective Km. ITZ, OH-ITZ, keto-ITZ and ND-ITZ exhibited unbound IC50 values of 6.1 nM, 4.6 nM, 7.0 nM, and 0.4 nM, respectively, when coincubated with human liver microsomes and midazolam (substrate concentration < Km). These findings demonstrate that ITZ metabolites are as potent as or more potent CYP3A4 inhibitors than ITZ itself, and thus may contribute to the inhibition of CYP3A4 observed in vivo after ITZ dosing.

Kiiv, an in vivo parameter for predicting the magnitude of a drug interaction arising from competitive enzyme inhibition

The goal of the study was to test the assumption that a competitive inhibition constant determined in vivo, Kiiv, like its corresponding in vitro counterpart, Ki, is independent of inhibitor concentration. Inhibition of the CYP2C9-dependent formation of (S)-7-hydroxy-warfarin from (S)-warfarin was measured in seven healthy subjects at three different doses of fluconazole. Prothrombin time measurements showed increasing anticoagulant activity with increasing fluconazole dose. The pharmacokinetic parameters calculated from the (S)- and (R)-warfarin plasma levels were consistent with previous studies. Fluconazole reduced the clearance of (S)-warfarin to a greater extent than that of (R)-warfarin. The decrease in clearance of both warfarin enantiomers was fluconazole dose-dependent. The formation of (S)-7-hydroxy-warfarin was inhibited by 31, 55, and 77% at the 100, 200, and 300 mg daily doses of fluconazole, respectively. Kiiv, values calculated from these data based on plasma fluconazole levels at each dose and data from earlier work at 400-mg daily doses of fluconazole were 30.7 +/- 23.7, 19.6 +/- 3.8, 17.9 +/- 7.5, and 19.8 +/- 3.5 microM, respectively. These results confirm the hypothesis that Kiiv is independent of inhibitor concentration.

Comparison of in vitro and in vivo inhibition potencies of fluvoxamine toward CYP2C19

A previous study suggested that fluvoxamine inhibition potency toward CYP1A2 is 10 times greater in vivo than in vitro. The present study was designed to determine whether the same gap exists for CYP2C19, another isozyme inhibited by fluvoxamine. In vitro studies examined the effect of nonspecific binding on the determination of inhibition constant (K(i)) values of fluvoxamine toward CYP2C19 in human liver microsomes and in a cDNA-expressed microsomal (Supersomes) system using (S)-mephenytoin as a CYP2C19 probe. K(i) values based on total added fluvoxamine concentration (K(i,total)) and unbound fluvoxamine concentration (K(i,ub)) were calculated, and interindividual variability in K(i) values was examined in six nonfatty livers. K(i,total) values varied with microsomal protein concentration, whereas the corresponding K(i,ub) values were within a narrow range (70-80 nM). In vivo inhibition constants (K(i)iv) were obtained from a study of the disposition of a single oral dose (100 mg) of the CYP2C19 probe (S)-mephenytoin in 12 healthy volunteers receiving fluvoxamine at 0, 37.5, 62.6, and 87.5 mg/day to steady state. In this population, the ratio of (S)-4-hydroxy-mephenytoin formation clearances (uninhibited/inhibited) was positively correlated with fluvoxamine average steady-state concentration with an intercept of 0.85 (r(2) = 0.88, p < 0.001). The mean (+/-S.D.) values of K(i)iv based on total and unbound plasma concentrations were 13.5 +/- 5.6 and 1.9 +/- 1.1 nM, respectively. Comparison of in vitro and in vivo K(i) values, based on unbound fluvoxamine concentrations, suggests that fluvoxamine inhibition potency is roughly 40 times greater in vivo than in vitro.

Covalent heme binding to CYP4B1 via Glu310 and a carbocation porphyrin intermediate

Recently we found that CYP4B1, and several other members of the CYP4 family of enzymes, are covalently linked to their prosthetic heme group through an ester linkage. In the current study, we mutated a conserved CYP4 I-helix residue, E310 in rabbit CYP4B1, to glycine, alanine, and aspartate to examine the effect of these mutations on the extent of covalent heme binding and catalysis. All mutants expressed well in insect cells and were isolated as a mixture of monomeric and dimeric forms as determined by LC/ESI-MS of the intact proteins. Rates of metabolism decreased in the order E310 > A310 >> G310 > D310, with the A310 and G310 mutants exhibiting alterations in regioselectivity for omega-1 and omega-2 hydroxylation of lauric acid, respectively. In marked contrast to the wild-type E310 enzyme, the G310, A310, and D310 mutants did not bind heme covalently. Uniquely, the acid-dissociable heme obtained from the D310 mutant contained an additional 16 amu relative to heme and exhibited the same chromatographic behavior as the monohydroxyheme species released upon base treatment of the covalently linked wild-type enzyme. Expression studies with H(2)(18)O demonstrated incorporation of the heavy isotope from the media into the monohydroxyheme isolated from the D310 mutant at a molar ratio of approximately 0.8:1. These data show (i) that E310 serves as the site of covalent attachment of heme to the protein backbone of rabbit CYP4B1; (ii) this I-helix glutamate residue influences substrate orientation in the active site of CYP4B1; and (iii) the mechanism of covalent heme attachment most likely involves a carbocation species located on the porphyrin.

Topiramate and phenytoin pharmacokinetics during repetitive monotherapy and combination therapy to epileptic patients

PURPOSE

To evaluate the potential pharmacokinetic interactions between topiramate (TPM) and phenytoin (PHT) in patients with epilepsy by studying their pharmacokinetics (PK) after monotherapy and concomitant TPM/PHT treatment.

METHODS

Twelve patients with epilepsy stabilized on PHT monotherapy were enrolled in this study, with 10 and seven patients completing the phases with 400 and 800 mg TPM daily doses, respectively. TPM was added at escalating doses, and after stabilization at the highest tolerated TPM dose, PHT doses were tapered. Serial blood and urine samples were collected for PK analysis during the monotherapy phase or the lowest PHT dose after taper and the concomitant TPM/PHT phase. Potential metabolic interaction between PHT and TPM also was studied in vitro in human liver microsomal preparations.

RESULTS

In nine of the 12 patients, PHT plasma concentrations remained stable, with a mean (+/-SD) area under the curve (AUC) ratio (combination therapy/monotherapy) of 1.13 +/- 0.17 (range, 0.89-1.23). Three patients had AUC ratios of 1.25, 1.39, and 1.55, respectively, and with the addition of TPM (800, 400, and 400 mg daily, respectively), their peak PHT plasma concentrations increased from 15 to 21 mg/L, 28 to 36 mg/L, and 27 to 41 mg/L, respectively. Human liver microsomal studies with S-mephenytoin showed that TPM partially inhibited CYP2C19 at very high concentrations of 300 microM (11% inhibition) and 900 microM (29% inhibition). Such high plasma concentrations would correspond to doses in humans that are 5 to 15 times higher than the recommended dose (200-400 mg). TPM clearance was approximately twofold higher during concomitant TPM/PHT therapy

CONCLUSIONS

This study provides evidence that the addition of TPM to PHT generally does not cause clinically significant PK interaction. PHT induces the metabolism of TPM, causing increased TPM clearance, which may require TPM dose adjustments when PHT therapy is added or is discontinued. TPM may affect PHT concentrations in a few patients because of inhibition by TPM of the CYP2C19-mediated minor metabolic pathway of PHT.

Fluvoxamine-theophylline interaction: gap between in vitro and in vivo inhibition constants toward cytochrome P4501A2

OBJECTIVE

Several reports indicate that fluvoxamine decreases the clearance of cytochrome P4501A2 (CYP1A2) substrates. This study compared in vitro and in vivo inhibition potencies of fluvoxamine toward CYP1A2 with an approach based on inhibition constants (K(i)) determined in vitro and in vivo.

METHODS

In vitro inhibition constant values were determined with human liver microsomes and complementary deoxyribonucleic acid-expressed CYP1A2 (supersomes). Fluvoxamine in vivo inhibition constants (K(i)iv) for CYP1A2 were obtained from an investigation of single-dose theophylline (250 mg) disposition in 9 healthy volunteers receiving steady-state (9 days) fluvoxamine at 3 doses (0, 25, or 75 mg/d) in a randomized crossover design.

RESULTS

In vitro K(i) values based on total inhibitor concentrations were 177 +/- 56 nmol/L, 121 +/- 21 nmol/L, and 52 +/- 13 nmol/L in human liver microsomes with 1 mg/ml protein and 0.5 mg/ml protein and in supersomes with 0.3 mg/ml protein, respectively. The corresponding in vitro K(i) values based on unbound fluvoxamine concentrations were 35 nmol/L, 36 nmol/L, and 36 nmol/L. The ratio of 1-methyluric acid formation clearances (control/inhibited) in 8 subjects was positively correlated with fluvoxamine concentration (r (2) = 0.87; P <.001) with an intercept near 1. Mean values for K(i)iv based on total and unbound plasma concentrations at steady state were 25.3 nmol/L (range, 14-39 nmol/L) and 3.6 nmol/L (range, 2.4-5.9 nmol/L), respectively.

CONCLUSION

Comparison of in vitro and in vivo K(i) values based on unbound fluvoxamine concentrations suggests that fluvoxamine inhibition potency is approximately 10 times greater in vivo than in vitro.

Covalent linkage of prosthetic heme to CYP4 family P450 enzymes

An extensive body of research on the structural properties of cytochrome P450 enzymes has established that these proteins possess a b-type heme prosthetic group which is noncovalently bound at the active site. Coordinate, electrostatic, and hydrogen bond interactions between the protein backbone and heme functional groups are readily overcome upon mild acid treatment of the enzyme, which releases free heme from the protein. In the present study, we have used a combination of HPLC, LC/ESI-MS, and SDS-PAGE techniques to demonstrate that members of the mammalian CYP4B, CYP4F, and CYP4A subfamilies bind their heme in an unusually tight manner. HPLC chromatography of CYP4B1 on a POROS R2 column under mild acidic conditions caused dissociation of less than one-third of the heme from the protein. Moreover, heme was not substantially removed from CYP4B1 under electrospray or electrophoresis conditions that readily release the prosthetic group from other non-CYP4 P450 isoforms. This was evidenced by an intact protein mass value of 59,217 +/- 3 amu for CYP4B1 (i.e., apoprotein plus heme) and extensive staining of this approximately 60 kDa protein with tetramethylbenzidine/H(2)O(2) following SDS-PAGE. In addition, treatment of CYP4B1, CYP4F3, and CYP4A5/7 with strong base generated a new, chromatographically distinct, polar heme species with a mass of 632.3 amu rather than 616.2 amu. This mass shift is indicative of the incorporation of an oxygen atom into the heme nucleus and is consistent with the presence of a novel covalent ester linkage between the protein backbone of the CYP4 family of mammalian P450s and their heme catalytic center.

Persistent inhibition of CYP3A4 by ketoconazole in modified Caco-2 cells

PURPOSE

The intestinal metabolism of some CYP3A substrates can be altered profoundly by co-administration of the potent inhibitor, ketoconazole. The present research was conducted to test the hypothesis that, unlike the inhibition kinetics observed with isolated microsomes, inhibition of CYP3A4 by ketoconazole in an intestinal cell monolayer is time-dependent and slowly reversible.

METHODS

Confluent, 1alpha,25-dihydroxy Vitamin D3-treated Caco-2 cells were exposed to 1 microM ketoconazole for two hours (Phase I) and then washed three times with culture medium containing no inhibitor. This was followed by a second incubation period (Phase II) that varied in the composition of the apical and basolateral culture medium: Condition 1. apical/basolateral differentiation medium (DM); Condition 2, apical/ basolateral DM + basolateral 2g/dL Human Serum Albumin (HSA); Condition 3, apical/basolateral DM + apical/basolateral 2 g/dL HSA. After various lengths of time for the second phase (0 to 4 hours), both apical and basolateral medium were exchanged with fresh DM. Midazolam (6 microM) was included in the apical medium for determination of CYP3A4 activity (Phase III).

RESULTS

Two-way ANOVA of the data revealed persistent inhibition of CYP3A4 under Conditions 1 and 2 (p < 0.001). In contrast, cells treated under Condition 3 exhibited rapid reversal of CYP3A4 inhibition. The level of CYP3A4 activity observed was inversely correlated with the amount of ketoconazole remaining in the cell monolayer at the end of Phase II.

CONCLUSIONS

These studies provide mechanistic evidence that ketoconazole can be sequestered into the intestinal mucosa after oral administration, producing a persistent inhibition of first-pass CYP3A4 activity.

Effect of inhibitor depletion on inhibitory potency: tight binding inhibition of CYP3A by clotrimazole

The purpose of this work was to evaluate the effect of mutual unbound inhibitor and unbound enzyme depletion on the potency of three antifungal cytochrome P-450 (CYP)3A inhibitors with over 1000-fold range in enzyme affinity. Incubations were performed with human liver microsomal protein concentrations that varied from 25 to 1000 microg/ml. The effect of each inhibitor was evaluated using midazolam as a CYP3A probe. Clotrimazole was found to be a tight binding inhibitor of CYP3A with a Ki of 250 pM. Analysis of percent inhibition data by stepwise linear regression for the matrix of inhibitor and enzyme concentrations used showed that protein concentrations predicted the percent inhibition by clotrimazole (r2 = 0.60, p <.001). When clotrimazole concentrations were added to the model, the r2 improved to 0.81, p =.003. Clotrimazole concentrations alone were not a significant predictor of percent inhibition (r2 = 0. 21, p =.08). For ketoconazole, protein concentrations provided a weak prediction of the percent inhibition (r2 = 0.39, p =.006). Conversely, ketoconazole concentrations alone were a good predictor of percent inhibition (r2 = 0.55, p <.001). In contrast to results with clotrimazole and ketoconazole, percent inhibition by fluconazole was not dependent on protein concentrations (r2 = 0.06, p =.39). We conclude that microsomal inhibitory potency can be affected by incubation conditions that deplete the unbound concentration of inhibitor available to the enzyme. This may introduce serious error into a quantitative prediction of an in vivo drug-drug interaction based on an in vitro derived Ki value.

Midazolam metabolism by modified Caco-2 monolayers: effects of extracellular protein binding

It has been suggested that the binding of a drug to plasma proteins will influence the intestinal extraction efficiency when drug is delivered to the mucosal epithelium via either the gut lumen or vasculature. We evaluated this hypothesis using cytochrome P-450 (CYP)3A4-expressing Caco-2 monolayers as a model for the intestinal epithelial barrier and midazolam as a CYP3A-specific enzyme probe. The rate of 1'-hydroxylation was measured following apical or basolateral midazolam administration to monolayers incubated in the presence or absence of 4 g/dl of human serum albumin (HSA) in the basolateral compartment medium. The midazolam-free fraction in culture medium containing HSA was 3.3%. Inclusion of HSA in the basolateral medium decreased peak intracellular midazolam accumulation after an apical midazolam dose (3 microM) by 35% and reduced the 1'-hydroxymidazolam formation rate by approximately 20%. Because of the accelerated diffusion of midazolam through the cell monolayer and into the basolateral compartment, there was a 61% reduction in the first-pass metabolic extraction ratio: 13.3 +/- 0. 12% for control versus 5.2 +/- 1% with HSA. Compared with control, addition of HSA resulted in a 91% decrease in the peak intracellular midazolam level and a 86% decrease in the rate of 1'-hydroxylation after the administration of midazolam into basolateral medium. These findings suggest that, in vivo, binding of a drug to plasma proteins will impact both first-pass and systemic intestinal midazolam extraction efficiency. Furthermore, the effect will be more pronounced for a drug that is delivered to mucosal enterocytes by way of arterial blood, compared with oral drug delivery.

First-pass midazolam metabolism catalyzed by 1alpha,25-dihydroxy vitamin D3-modified Caco-2 cell monolayers

Cytochrome P-450 (CYP) 3A4 accounts for approximately 50% of all P-450s found in the small intestine (Paine et al., 1997) and contributes to the extensive and variable first-pass extraction of drugs such as cyclosporine and saquinavir. We recently demonstrated that CYP3A4 expression in a differentiated Caco-2 subclone is increased when cell monolayers are treated with 1alpha,25-dihydroxy-vitamin-D3 (Schmiedlin-Ren et al., 1997). This improved metabolic capacity permits the in vitro modeling of first-pass intestinal metabolic kinetics. Midazolam (MDZ) 1'-hydroxylation was used as a specific probe for CYP3A-mediated metabolism in modified Caco-2 monolayers. Caco-2 cells were grown to confluence on laminin-coated culture inserts, and then for two additional weeks in the presence of 1alpha,25-dihydroxy vitamin-D3. Cell monolayers were subsequently exposed to MDZ for varying lengths of time and concentrations. The amount of MDZ in the monolayer increased rapidly after apical drug administration, reaching a pseudo steady state within 6 min. The cellular uptake rate was considerably slower after a basolateral dose. By either route of administration, the rate of 1'-hydroxymidazolam formation was stable and linear for 2 h. Under basolateral sink conditions and low apical MDZ dosing concentration (1-8 microM), the first-pass extraction ratio was found to be approximately 15%. Higher dosing concentrations led to saturation of the hydroxylation reaction and reduction in the extraction ratio. The modified Caco-2 cell monolayer is an excellent model for studying drug absorption and first-pass intestinal metabolic kinetic processes. In this system, the selective CYP3A probe MDZ was rapidly absorbed, yet extensively metabolized, as is observed in vivo.

Inhibition of cytochrome P-450 3A (CYP3A) in human intestinal and liver microsomes: comparison of Ki values and impact of CYP3A5 expression

The purpose of this study was to compare the kinetics of intestinal and hepatic cytochrome P-450 3A (CYP3A) inhibition by using microsomal midazolam 1'-hydroxylation as a marker of enzyme activity. The effect of two antifungal agents commonly implicated in CYP3A drug-drug interactions was examined. Inhibition type and affinities were determined for human liver and intestinal microsomes screened for the presence or absence of CYP3A4 and CYP3A5, as well as for cDNA-expressed CYP3A4 and CYP3A5 microsomes. Ketoconazole and fluconazole were found to be noncompetitive inhibitors of both enzymes. Ketoconazole exhibited a Ki for cDNA-expressed CYP3A4 of 26. 7 +/- 1.71 nM, whereas the Ki for cDNA expressed CYP3A5 was 109 +/- 19.7 nM. Corresponding Ki values for fluconazole were 9.21 +/- 0.51 microM and 84.6 +/- 12.9 microM. For liver and intestinal microsomes that contained only CYP3A4, the average ketoconazole Ki was found to be 14.9 +/- 6.7 nM and 17.0 +/- 7.9 nM, respectively, whereas fluconazole yielded mean respective Ki values of 10.7 +/- 4.2 microM and 10.4 +/- 2.9 microM. Liver and intestinal microsomes that contained an equal or greater amount of CYP3A5, in addition to CYP3A4, were less susceptible to inhibition by both ketoconazole and fluconazole. These findings suggest that there can be significant differences in the affinity of these two enzymes for inhibitors. This may further broaden interindividual variability with respect to the magnitude of in vivo drug-drug interactions. We also conclude that there is no significant difference in inhibition type and affinity of ketoconazole and fluconazole for hepatic versus intestinal CYP3A4.

Mechanism-based inactivation of human cytochrome P450 1A2 by furafylline: detection of a 1:1 adduct to protein and evidence for the formation of a novel imidazomethide intermediate

The rapid loss of human CYP1A2 (cytochrome P450 1A2) activity caused by the 8-methylxanthine furafylline is investigated with the aim of determining whether a stable covalent adduct of the xanthine to the enzyme could be identified. Metabolic studies employing expressed CYP1A2 with radiolabeled furafylline and a close analogue, cyclohexylline, where the furan ring is replaced with cyclohexane, indicate that these xanthines are bound in a 1:1 ratio to CYP1A2 protein. This result, combined with earlier kinetic studies, verifies that these compounds are mechanism-based inhibitors of the enzyme. The 8'-methyl carbinols are the only metabolites formed by CYP1A2, and substantial (70-80%) incorporation of oxygen from the medium into the carbinols is observed. Carbinol formation is further characterized by high intramolecular isotope effects (kH/kD > 9) and low intermolecular isotope effects (DV/K < 2). Overall partition ratios are low (5.0 and 7.6, respectively), confirming our previous conclusion that furafylline is an efficient inactivator. By contrast, the N7-methyl-8-methylxanthines are good substrates for CYP1A2 but are not themselves inactivating agents. In addition to other metabolic products, the 8'-methyl carbinols of these N7-methyl-8-methylxanthines are formed in substantial amounts with equally high intramolecular isotope effects; however, the carbinol oxygen is derived exclusively from molecular oxygen. We conclude that oxidation of the 8-methyl group of furafylline and cyclohexylline, but not their N7-methyl analogues, by CYP1A2 promotes a major fraction of the inactivating xanthines to a two electron oxidized intermediate which either terminates enzyme activity by reaction with an active site amino acid or is decomposed by reaction with the medium to give carbinol.

Subnanomolar quantification of caffeine's in vitro metabolites by stable isotope dilution gas chromatography-mass spectrometry

A method for the quantification of subnanomolar levels of in vitro metabolites of caffeine by an isotope dilution gas chromatographic-mass spectrometric (GC-MS) assay has been developed and applied. Trideuteromethylated analogs of each primary metabolite were synthesized and added after incubations of caffeine with human liver microsomes high in cytochrome P4501A2. HPLC separation of the metabolites prior to GC-MS quantification allowed the isolation of theobromine and paraxanthine which coeluted by GC and enabled quantification over a larger dynamic range. Quantitative analysis was performed on the n-propylated derivatives by selected-ion monitoring of either the M+. ions for the dimethylxanthines or [M-C3H6]+. ions for 1,3,7-trimethyluric acid. For the least abundant metabolite (1,3,7-trimethyluric acid), the detection level on column was 200 pg. Replicate analyses exhibited intra- and inter-day variability of 4.2 and 7.9%, respectively. This assay has been successfully used in the quantification of caffeine's primary metabolites in more than 180 incubations, at varying substrate concentrations and with multiple enzyme sources.

Characterization of interintestinal and intraintestinal variations in human CYP3A-dependent metabolism

Cytochrome P450 3A (CYP3A) metabolizes a diverse array of clinically important drugs. For some of these (e.g., cyclosporine, verapamil, midazolam), CYP3A in the intestinal mucosa contributes to their extensive and variable first-pass extraction. To further characterize this phenomenon, we measured CYP3A content and catalytic activity toward the probe substrate midazolam in mucosa isolated from duodenal, jejunal and ileal sections of 20 human donor intestines. For comparison, the same measurements were performed for 20 human donor livers, eight of which were obtained from the same donors as eight of the intestines. Excellent correlations existed between homogenate and microsomal CYP3A content for the three intestinal regions. Median microsomal CYP3A content was greatest in the duodenum and lowest in the ileum (31 vs. 17 pmol/mg of protein). With respect to midazolam 1'-hydroxylation kinetics, the median Km for each intestinal region was similar to the median hepatic Km, approximately 4 microM. In contrast, the median Vmax decreased from liver to duodenum to jejunum to ileum (850 vs. 644 vs. 426 vs. 68 pmol/min/mg). Intrinsic clearance (Vmax/Km) followed a similar trend for the intestinal regions; median duodenal intrinsic clearance was comparable to hepatic intrinsic clearance (157 and 200 microl/min/mg, respectively). Vmax correlated with CYP3A content for all tissues except the ileum. Duodenal and jejunal Vmax and CYP3A content varied by >30-fold among donors. Microsomes prepared from every other 1-foot section of six intestines were also analyzed for CYP3A as well as for two coenzymes. In general, CYP3A activity, CYP3A content and CYP reductase activity rose slightly from duodenum to middle jejunum and then declined to distal jejunum and ileum. Cytochrome b5 content and cytochrome b5 reductase activity varied little throughout the intestinal tract. Regional intrinsic midazolam 1'-hydroxylation clearance was greatest for the jejunum, followed by the duodenum and ileum (144, 50 and 19 ml/min, respectively). Collectively, these results demonstrate that the upper small intestine serves as the major site for intestinal CYP3A-mediated first-pass metabolism and provides a rationale for interindividual differences in oral bioavailability for some CYP3A substrates.

Influence of stiripentol on cytochrome P450-mediated metabolic pathways in humans: in vitro and in vivo comparison and calculation of in vivo inhibition constants

OBJECTIVE

The spectrum of cytochrome P450 inhibition of stiripentol, a new anticonvulsant, was characterized in vitro and in vivo.

METHODS

Stiripentol was incubated in vitro with (R)-warfarin, coumarin, (S)-warfarin, (S)-mephenytoin, bufuralol, p-nitrophenol, and carbamazepine as probes for CYPs 1A2, 2A6, 2C9, 2C19, 2D6, 2E1, and 3A4, respectively. Caffeine demethylation and the 6 beta-hydroxycortisol/cortisol ratio were monitored in vivo before and after 14 days of treatment with stiripentol as measures of CYP1A2 and CYP3A4 activity, and dextromethorphan O- and N-demethylation were used to measure CYP2D6 and CYP3A4 activity, respectively. In vivo inhibition constants for CYP3A4 were calculated with use of data that previously documented the interaction between stripentol and carbamazepine.

RESULTS

In vitro, stiripentol inhibited CYPs 1A2, 2C9, 2C19, 2D6, and 3A4, with inhibition constant values at or slightly higher than therapeutic (total) concentrations of stiripentol, but it did not inhibit CYPs 2A6 and 2E1 even at tenfold therapeutic concentrations. In vivo inhibition of caffeine demethylation and dextromethorphan N-demethylation were consistent with inhibition of CYP1A2 and CYP3A4, respectively. The 6 beta-hydroxycortisol/cortisol ratio did not provide a reliable index of CYP3A4 inhibition. Inhibition of CYP2D6-mediated O-demethylation was not observed in vivo. With use of carbamazepine, in vivo inhibition constants for CYP3A4 ranged between 12 and 35 mumol/L, whereas the corresponding in vitro value was 80 mumol/L.

CONCLUSIONS

Stiripentol appears to inhibit several CYP450 enzymes in vitro and in vivo. In vivo inhibition constants show that stiripentol inhibition of CYP3A4 is linearly related to plasma concentration in patients with epilepsy.

The role of cytochrome P450 3A4 in alfentanil clearance. Implications for interindividual variability in disposition and perioperative drug interactions

BACKGROUND

There is considerable unexplained variability in alfentanil pharmacokinetics, particularly systemic clearance. Alfentanil is extensively metabolized in vivo, and thus systemic clearance depends on hepatic biotransformation. Cytochrome P450 3A4 was previously shown to be the predominant P450 isoform responsible for human liver microsomal alfentanil metabolism in vitro. This investigation tested the hypothesis that P450 3A4 is responsible for human alfentanil metabolism and clearance in vivo.

METHODS

Nine healthy male volunteers who provided institutionally approved written informed consent were studied in a three-way randomized crossover design. Each subject received alfentanil (20 micrograms/kg given intravenously) 30 min after midazolam (1 mg injected intravenously) on three occasions: control; high P450 3A4 activity (rifampin induction); and low P450 3A4 activity (selective inhibition by troleandomycin). Midazolam is a validated selective in vivo probe for P450 3A4 activity. Venous blood was sampled for 24 h and plasma concentrations of midazolam and alfentanil and their primary metabolites 1'-hydroxymidazolam and noralfentanil were measured by gas chromatography-mass spectrometry. Pharmacokinetic parameters were determined by two-stage analysis using both noncompartmental and three-compartment models.

RESULTS

Plasma alfentanil concentration-time profiles depended significantly on P450 3A4 activity. Alfentanil noncompartmental clearance was 5.3 +/- 2.3, 14.6 +/- 3.8, and 1.1 +/- 0.5 ml.kg-1.min-1, and elimination half-life was 58 +/- 13, 35 +/- 7, and 630 +/- 374 min, respectively, in participants with normal (controls), high (rifampin), and low (troleandomycin) P450 3A4 activity (means +/- SD; P < 0.05 compared with controls). Multicompartmental modeling suggested a time-dependent inhibition-resynthesis model for troleandomycin effects on P450 3A4 activity, characterized as k10(t) = k10[1-phi e-alpha(t-tzero)], where k10(t) is the apparent time-dependent rate constant, k10 is the uninhibited rate constant, phi is the fraction of P450 3A4 inhibited, and alpha is the apparent P450 3A4 reactivation rate. Alfentanil clearance was calculated as V1 k10 for controls and men receiving rifampin, and as V1.average k10(t) for men receiving troleandomycin. This clearance was 4.9 +/- 2.1, 13.2 +/- 3.6, and 1.5 +/- 0.8 ml.kg-1.min-1, respectively, in controls and in men receiving rifampin or troleandomycin. There was a significant correlation (r = 0.97, P < 0.001) between alfentanil systemic clearance and P450 3A4 activity.

CONCLUSIONS

Modulation of P450 3A4 activity by rifampin and troleandomycin significantly altered alfentanil clearance and disposition. These results strongly suggest that P450 3A4 is the major isoform of P450 responsible for clinical alfentanil metabolism and clearance. This observation, combined with the known population variability in P450 3A4 activity, provides a mechanistic explanation for the interindividual variability in alfentanil disposition. Furthermore, known susceptibility of human P450 3A4 activity to induction and inhibition provides a conceptual framework for understanding and predicting clinical alfentanil drug interactions. Finally, human liver microsomal alfentanil metabolism in vitro is confirmed as an excellent model for human alfentanil metabolism in vivo.

The kinetics of aflatoxin B1 oxidation by human cDNA-expressed and human liver microsomal cytochromes P450 1A2 and 3A4

The combined presence of CYP1A2 and 3A4, both of which oxidize aflatoxin B1 (AFB1) to the reactive aflatoxin B1-8,9-epoxide (AFBO) and to hydroxylated inactivation products aflatoxin M1 (AFM1) and aflatoxin Q1 (AFQ1), substantially complicates the kinetic analysis of AFB1 oxidation in human liver microsomes. In the present study, we examine the reaction kinetics of AFB1 oxidation in human liver microsomes (HLMs, N = 3) and in human CYP3A4 and CYP1A2 cDNA-expressed lymphoblastoid microsomes for the purpose of identifying the CYP isoform(s) responsible for AFB1 oxidation at low substrate concentrations approaching those potentially encountered in the diet. AFBO formation by cDNA-expressed human CYP1A2 followed Michaelis-Menten kinetics (Km = 41 microM, Vmax = 2.63 nmol/min/nmol P450). Furthermore, the portion of AFBO formed in HLMs which was eliminated by furafylline, a specific mechanism-based inhibitor of CYP1A2, also followed Michaelis-Menten kinetics (Km = 32-47 microM, Vmax = 0.36-0.69 nmol/min/nmol P450). The formation of AFBO (activation product) and AFQ1 (detoxification product) in cDNA-expressed human CYP3A4 microsomes was sigmoidal and consistent with the kinetics of substrate activation. Accordingly, application of a sigmoid Vmax model equivalent to the Hill equation produced excellent fits to the cDNA-expressed CYP3A4 data and also to the data from HLMs pretreated with furafylline to remove CYP1A2. The Hill model predicted that two substrate binding sites are involved in CYP3A4-mediated AFB1 catalysis and that the average affinity of AFB1 for the two sites was 140-180 microM. Vmax values for AFQ1 formation were 10-fold greater than those for AFBO, and total substrate turnover to both was 67 nmol/min/nmol CYP3A4. Using the derived kinetic parameters for CYP1A2 and 3A4 to model the in vitro rates of AFB activation at low substrate concentrations, it was predicted that CYP1A2 contributes to over 95% of AFB activation in human liver microsomes at 0.1 microM AFB. The important role of CYP1A2 in the in vitro activation of AFB at low substrate concentrations was supported by DNA binding studies. AFB1-DNA binding in control HLMs (reflecting the contribution of CYP1A2 and CYP3A4) and furafylline-pretreated microsomes (reflecting the contribution of CYP3A4 only) catalyzed the binding of 1.71 and 0.085 pmol equivalents of AFB1 to DNA, respectively, indicating that CYP1A2 was responsible for 95% of AFB1-DNA adduct formation at 0.133 microM AFB. These results demonstrate that CYP1A2 dominates the activation of AFB in human liver microsomes in vitro at submicromolar concentrations and support the hypothesis that CYP1A2 is the predominant enzyme responsible for AFBO activation in human liver in vivo at the relatively low dietary concentrations encountered in the human diet, even in high AFB exposure regions of the world. However, because the actual concentrations of AFB in liver in vivo following dietary exposures are uncertain, additional studies in exposed human populations are needed. Quantitative data on the relative rates of AFM1 and AFQ1 excretion (potential biomarkers for CYP1A2 and 3A4 activity, respectively) in humans would be useful to validate the actual contributions of these two enzymes to AFB1 oxidation in vivo.

Relation between plasma and saliva concentrations of enoxacin, ciprofloxacin, and theophylline

To assess the reliability of predicting plasma concentrations of enoxacin, ciprofloxacin, and theophylline from drug concentrations in saliva, six healthy volunteers received single oral doses of enoxacin, ciprofloxacin, and theophylline administered in combination on each of four separate study days, with different, doses separated by at least 5 days. Drug concentrations were determined by a newly developed high-performance liquid chromatography (HPLC) assay, which could measure simultaneously all three drugs in plasma or saliva. Saliva data from the postabsorptive phase after drug administration were used to minimize the effects of variation in absorption. There were good correlations between saliva and plasma concentrations of enoxacin, ciprofloxacin, and theophylline (r = 0.91, 0.88, and 0.98, respectively). The mean (+/-SD) saliva-to-plasma (S/P) ratio for theophylline was 0.63 +/- 0.06 with a coefficient of variation (CV) of 7.9 +/- 2.7%. In contrast, the S/P ratios and CV values for enoxacin and ciprofloxacin were 0.72 +/- 0.21 and 28.9 +/- 11.1%, and 0.58 +/- 0.15 and 25.3 +/- 6.7%, respectively. Because of the large inter- and intraindividual variability, saliva concentrations of enoxacin and ciprofloxacin are not reliable for predicting plasma concentrations. However, saliva may be used reliably for predicting plasma concentrations of theophylline.

Binding of the aflatoxin-glutathione conjugate to mouse glutathione S-transferase A3-3 is saturated at only one ligand per dimer

The binding of two different reaction products (p-nitrobenzyl glutathione and the aflatoxin-glutathione conjugate) to mouse glutathione S-transferase A3-3 (mGSTA3-3) has been measured using equilibrium dialysis and a direct fluorescence quenching technique. As expected, p-nitrobenzyl glutathione was found to bind with a stoichiometry of 2.24 +/- 0.17 mol/mol of dimeric enzyme. However, the much larger aflatoxin-glutathione conjugate, 8, 9-dihydro-8-(S-glutathionyl)-9-hydroxyl-aflatoxin B1 (AFB-GSH), was found to bind with a stoichiometry of 1.12 +/- 0.08 mol/mol of dimeric enzyme. p-Nitrobenzyl glutathione bound mGSTA3-3 with a dissociation constant (Kd) of 59 +/- 17 microM while the aflatoxin-glutathione conjugate bound the enzyme with a Kd of 0.86 +/- 0.19 microM. Glutathione competitively inhibited binding of AFB-GSH to mGSTA3-3 with a Ki of 1.5 mM, suggesting that AFB-GSH was binding to the enzyme active site. Although AFB-GSH bound to mGSTA3-3 with a stoichiometry of 1 mol/mol of dimeric enzyme, AFB-GSH completely inhibited activity toward 1-chloro-2, 4-dinitrobenzene, indicating that AFB-GSH binding to one active site alters affinity for 1-chloro-2,4-dinitrobenzene in the active site of the other subunit. To our knowledge, this is the first report of a glutathione S-transferase reaction product which binds to the enzyme with a stoichiometry of 1 mol/mol of dimer.

First-pass metabolism of midazolam by the human intestine

The in vivo intestinal metabolism of the CYP3A probe midazolam to its principal metabolite, 1'-hydroxymidazolam, was investigated during surgery in 10 liver transplant recipients. After removal of the diseased liver, five subjects received 2 mg midazolam intraduodenally, and the other five received 1 mg midazolam intravenously. Simultaneous arterial and hepatic portal venous blood samples were collected during the anhepatic phase; collection of arterial samples continued after reperfusion of the donor liver. Midazolam, 1'-hydroxymidazolam, and 1'-hydroxymidazolam glucuronide were measured in plasma. A mass balance approach that considered the net change in midazolam (intravenously) or midazolam and 1'-hydroxymidazolam (intraduodenally) concentrations across the splanchnic vascular bed during the anhepatic phase was used to quantitate the intestinal extraction of midazolam after each route of administration. For the intraduodenal group, the mean fraction of the absorbed midazolam dose that was metabolized on transit through the intestinal mucosa was 0.43 +/- 0.18. For the intravenous group, the mean fraction of midazolam extracted from arterial blood and metabolized during each passage through the splanchnic vascular bed was 0.08 +/- 0.11. Although there was significant intersubject variability, the mean intravenous and intraduodenal extraction fractions were statistically different (p = 0.009). Collectively, these results show that the small intestine contributes significantly to the first-pass oxidative metabolism of midazolam catalyzed by mucosal CYP3A4 and suggest that significant first-pass metabolism may be a general phenomenon for all high-turnover CYP3A4 substrates.

Formation of (R)-8-hydroxywarfarin in human liver microsomes. A new metabolic marker for the (S)-mephenytoin hydroxylase, P4502C19

Kinetic studies demonstrate that two forms of human liver cytochrome P450 are responsible for the formation of (R)-8-hydroxywarfarin: a low-affinity enzyme (KM approximately 1.5 mM), previously identified as P4501A2; and a high-affinity enzyme (KM = 330 microM), now identified as P4502C19 on the basis of the following evidence. In crossover inhibition studies with P4501A2-depleted human liver microsomes between (R)-warfarin and (S)-mephenytoin, reciprocal competitive inhibition was observed. Apparent KM values for (S)-mephenytoin-4'-hydroxylation (52-67 microM) were similar to the determined Ki values (58-62 microM) for (S)-mephenytoin inhibition of (R)-8-hydroxywarfarin formation. Similarly, the apparent KM for (R)-warfarin 8-hydroxylation in furafylline-pretreated microsomes (KM = 289-395 microM) was comparable with the Ki values (280-360 microM) for (R)-warfarin inhibition of (S)-4'-hydroxymephenytoin formation. Inhibition studies with tranylcypromine, a known inhibitor of (S)-mephenytoin hydroxylase activity, and either substrate in three different microsomal preparations yielded nearly identical inhibitory constants: Ki = 8.7 +/- 1.6 microM for inhibition of (S)-4'-hydroxymephenytoin formation and 8.8 +/- 2.5 microM for inhibition of (R)-8-hydroxywarfarin formation. In addition, fluconazole, a potent inhibitor of (R)-warfarin 8-hydroxylation, Ki = 2 microM, was found to inhibit (S)-mephenytoin hydroxylation with an identical Ki (2 microM). Finally, a strong correlation between (S)-mephenytoin 4-hydroxylation and (R)-warfarin 8-hydroxylation activities in furafylline-pretreated microsomes was demonstrated in 14 human liver microsomal preparations (r2 = 0.97).

Oral first-pass elimination of midazolam involves both gastrointestinal and hepatic CYP3A-mediated metabolism

OBJECTIVE

To determine in humans the relative roles of intestinal and hepatic metabolism in the oral first-pass elimination of a CYP3A substrate using midazolam as a model compound.

METHODS

Midazolam was administered intravenously (1 mg) or orally (2 mg) to 20 healthy young subjects (10 men and 10 women) in a random fashion, and the disposition of the drug and its 1'-hydroxy metabolite were determined. In separate in vitro studies, the CYP3A-mediated formation of 1'-hydroxymidazolam by human hepatic and intestinal microsomes was investigated.

RESULTS

No gender-related differences were noted in either the systemic (370 +/- 114 ml/min [mean +/- SD]) or oral (1413 +/- 807 ml/min) clearance values of midazolam. Despite complete oral absorption, measured oral bioavailability was on average about 50% less than that predicted on the assumption that only the liver contributed to first-pass metabolism. Pharmacokinetic estimation of the intestinal component indicated an extraction ratio (0.43 +/- 0.24) that was similar to that of the liver (0.44 +/- 0.14). 1'-Hydroxymidazolam was extensively but variably formed in vitro by both hepatic and intestinal microsomes and, although the intrinsic clearance (V(max)/Km) was higher in the liver preparations (540 +/- 747 versus 135 +/- 92 microliters/min/mg protein), this difference was not statistically significant.

CONCLUSIONS

These results show that the small intestine can be a major site for presystemic, CYP3A-mediated metabolism after oral administration. Moreover, it appears that this represents a true first-pass effect. In addition, intestinal and hepatic metabolism may be important factors in interindividual variability in disposition after oral administration of midazolam and similar CYP3A substrates. Finally, intestinal localization of CYP3A may be significant in metabolism-based drug-drug interactions.

Warfarin-fluconazole. I. Inhibition of the human cytochrome P450-dependent metabolism of warfarin by fluconazole: in vitro studies

The antifungal agent fluconazole was found to be a potent inhibitor of cytochrome P450 (P450) 2C9 (Ki = 7-8 microM), the principal enzyme responsible for the clearance (85%) of the more potent anticoagulant (S)-warfarin to the inactive (S)-7- and (S)-6-hydroxywarfarin metabolites in vivo. Fluconazole was also found to be a potent inhibitor of the P4503A4-catalyzed formation of (R)-10-hydroxywarfarin (Ki = 15-18 microM) as well as the low KM P450 enzymes responsible for the formation of (R)-6-, (R)-7-, and (R)-8-hydroxywarfarin (Ki = 2-6 microM). By contrast, experiments with the P4501A2 inhibitor furafylline and cDNA-expressed P4501A2 indicate that fluconazole is a weak inhibitor of this enzyme (Ki > 800 microM), as measured by the inability of fluconazole to significantly suppress the P4501A2-dependent 6-hydroxylation of (R)-warfarin. The prediction generated from these studies, that fluconazole is a potent in vivo inhibitor of warfarin metabolism, , is tested in complementary studies reported in the accompanying article, "Warfarin-Fluconazole II".