tife Sciences, Vol. 60, No. 22, pp. 19531960,1597 Cop&M 0 1597 E!ls&er Science Inc. Printed in the USA. All rights resewed 0024-3205197 s17.M) t .a0 ELSEVIER
`
`PI1 SOO24-3205(97)00160-4
`
`INVOLVEMENT OF CYTOCHROME P450 3A4 IN N-DEALKYLATION OF BUPRENORPHINE IN HUMAN LIVER MICROSOMES Christelle Iribame, Daniel Picart, Yvonne Drkano, Jean-Pierre Bail* and Fraqois Berthou’ Laboratoires de Biochimie-Nutrition - Equipe d’ Accueil948 - Faculte de Medecine BP. 8 15 - 29285 BREST-Cedex - France * Service de chirurgie digestive - CHU La Cavale Blanche - BREST - France.
`
`(Received in final form February u), 1997)
`
`(F =
`
`Buprenorphine is a long acting analgesic of the opiate family. Recently, it has been proposed for the opioid dependency treatment at a large scale. The drug is extensively metabolized by the hepatic cytochrome P450 in man, yielding a N- dealkylated metabolite, norbuprenorphine. The specific forms of P450 involved in this oxidative N-demethylation were examined in a panel of 18 human liver microsomal preparations previously characterized with respect to their P450 contents. Buprenorphine was N-dealkylated with an apparent Km of 89 & 45 pM (n = 3). The metabolic rates were 3.46 f. 0.43 nmol/(min x mg of protein). This metabolic pathway was strongly correlated with 6 catalytic activities specific to P450 3A4 and with the immunodetectable P450 3A content of liver microsomal samples
`0.87). Buprenorphine metabolism was 62-71% inhibited by three mechanism-based inhibitors (TAO, erythralosamine, gestodene), by nifedipine as competitive inhibitor (Ki = 129 l&l) and by ketoconazole 0.6 @I (25% residual activity), all these inhibitors specific to P450 3A. Among 10 heterologously expressed P45Os tested, only P450 3A4 was able to dealkylate buprenorphine with a turnover number of 9.6 min -‘. Morever, this catalytic activity was inhibited up to 80% (vs control) by anti-rat P450 3A antibody. Taken together, all these data demonstrate that P450 3A4 is the major enzyme involved in hepatic buprenorphine N-dealkylation. Key
`Buprenorphine (BPN, N-cyclopropylmethyl-7a-[ l-(5) hydroxy- 1,2,2-trimethyl- propyl]-6,14-endo-ethano-6,7,8,14-tetrahydronororipavine; Temgesi? or Subutex@ specialities) is a derivative of the opiate thebaine. It is a long acting analgesic with both narcotic agonist and antagonist actions. Buprenorphine (Temgesi?) is often prescribed for the treatment of chronic post-operative pain and for terminal cancer patients (1). Recently, buprenorphine has been proposed for the management of opioid addicts (2-4). Prescribing and delivery of buprenorphine as Subutex@ speciality can be made at the request of addicts by all the medicine doctors in France (5). Indeed, of greatest importance is the very positive impact of the effective BPN maintenance program on preventing the spread of HIV or hepatitis C viruses. Efficiency of treatment of these
`
`Wonic
`
`buprenorphine dealkylation, P450 3A4, human liver microsomes
`
`’ Corresponding
`
`Page 1
`
`RB Ex. 2019
`BDSI v. RB PHARMACEUTICALS LTD
`IPR2014-00325
`
`author: Dr. F. Bertbou; Fax (33) 98 01 66 03; E-mail: Francois.BerthouBuniv-brest.fr
`

`
`1954
`
`Bupreaorphine De&$&n by P4SO 3A4 Vol. 60, No. 22, 1997 patients requires a knowledge of the drug metabolic pathways and, in particular, the nature of the enzyme(s) involved in its biotransformation. BPN is metabolized by N-dealkylation of its cyclopropyl methyl group (Figure 1) into norbuprenorphine (NBPN) and subsequent conjugation with glucuronic acid of both BPN and NBPN (6). Buprenorphine is so extensively metabolized by the hepatic first-pass that less than 20% of BPN administered per OS remains unchanged into blood circulation. Thus, to avoid this effect of first-pass metabolism by the liver, BPN is sublingually administered in treatment of heroin addicts (7). Despite of its widespread use as a substitutive treatment for opiate abuse, however, there ate no direct studies on the nature of enzymes involved in the major metabolic pathway of BPN. Preliminary experiments suggest that cytochrome P450 enzymes are involved in BPN metabolism. The aim of this in vitro study was to identify the hepatic P450 enzyme(s) involved in BPN dealkylation. Numerous approaches were used: 1) correlation between activities of isoform-selective substrates with use of a bank of 18 human liver samples, 2) use of isoform selective inhibitory probes, 3) determination of drug metabolism by P450 heterologously produced, 4) immuno-inhibition by specific inhibitory antibodies. It is now established that such a combined approach should provide a higher degree of certainty for the identification of P450 isoforms responsible for the metabolism of buprenorphine. That should allow to predict drug-drug interactions (8). I III1 d- c (CJ$ j3 L OH O-C%
`
`BUPRENORPHINE
`
`FIG. 1 Structural formula of buprenorphine. The arrow indicates the site of enzymatic attack by P450 leading to a N-dealkylated metabolite, named norbuprenorphine Material and methods Buprenorphine HCl and its dealkylated metabolite, norbuprenorphine, were gifts from Schering-Plough (Herouville-St-Clair, France). NADPH, troleandomycin (TAO), 7,8 - benzoflavone (a-naphthoflavone), ketoconazole and quinidine were provided by Sigma (St Louis,
`
`Page 2
`
`

`
`(lo),
`N-demethylation
`1 A,
`
`toremifene
`(12)
`
`Vol. 60, No. 22,1997 Buprenorphirre De&yl&o by P450 3A4 1955 MO, USA). Sulfaphenazole was from Ultrafine Chemicals (Manchester, U.K.). Gestodene was given by Schering SA (Lys-lez-Lannoy, France) and e@hdOSamiIIe Was gifi ficxn Dr. *laforge (URA-CNRS 400, paris). Solvents were of the highest quality available from Merck (Darmstadt, Germany). B * s ‘cro om ‘vities Human liver samples were obtained as previously described (9). The samples termed ‘( BrX >> were hepatic fragments obtained from patients with SeCOIKbIy tumors While SqleS termed << FHX bb were obtained from multiorgan transplant donors within 30 mm of clinical death after traffic accidents. The specific forms of P450 have been previously characterized acmrciing to methods reported elsewhere (9-12). These reactions included tamoxifene N- demethylation (9~ 6h-hydroxylation of testosterone and 2-hydroxylation of estmdiol
`2D6,2El and 2C9 and 2C19 have been previously characterized for the bank of human liver microsomes (12) Det ’ ‘n of buvreno&ine metabolisq ermmatw The final standard 1 mL incubation mixture contained 1 mg of microsomal proteins, 100 mM potassium phosphate buffer (pH 7.4) and 0.2 mM buprenorphine HCI. After 5 mm of preincubation at 37”C, the reaction was started by addition of 1 mM NADPH. After 15 mm shaking at 37°C the reaction was stopped by addition of 1 mL sodium carbonate 0.5 M pH 8.7. Extraction was performed by 2 x 5 mL diethyl ether. The organic layer was evaporated at 40°C under nitrogen stream. The dried residue was dissolved in 1 mL of HPLC mobile phase. Control incubations were run as described above except that microsomal proteins or NADPH was omitted. For the determination of kinetic parameters, BPN was added to the reaction mixture in the range 10-1000 PM. Overall biotransformation was calculed by calibrating with known amounts of NBPN added to the reaction mixture and extracted as described above. Linearity of NBPN formation was checked by varying protein concentration (0.5-3 mg) and time (lo-30 min) with BPN concentration of 0.2 mM. The HPLC analysis was performed on a Nucleosil C18, 5 ).rm particle diameter, 250 x 4.6 mm (Interchim, France). The mobile phase consisted of acetonitrile/water (30/70, v/v) containing 0.5% (v/v) triethylamine, pH adjusted to 3 with orthophosphoric acid, with a flow rate of 1 mumin. The analytes were detected at 210 nm at a sensitivity of 0.01 absorbance. Peaks were quantified by PC integration Pack Software (Kontron, France). Peaks were identified by their retention times and W absorption spectra in comparison with standard compounds. Morever, BPN and NBPN were detected by fluorescence detection (excitation 210 mm, emission 346 nm; Flo 2000 TSP, Les Ulis, France). For the BPN metabolism inhibition study, two types of inhibitions were used: (i) competitive inhibitors such as nifedipine, (ii) mechanism-based inhibitors such as gestodene, TAO and erythralosamine. The selective mechanism-based inhibitors were used in the range of concentrations determining their u windows of selectivity >> (8, 14), namely 0.25 mM TAO, 0.1
`
`mM
`
`gestodene. As the latter inhibitions require NADPH dependent complexation for inactivation (15), TAO, gestodene and 0.1 mM erythralosamine were preincubated in the presence of 1 mM NADPH and 1 mg microsomal protein at 37°C for 20 mm in a volume of 0.2 mL. The incubation medium was then diluted 5 fold in 0.1 mM potassium phosphate buffer containing 30 uM BPN and 1 mM NADPH. After 15 min incubation, BPN and its metabolites were extracted and analysed by HPLC as described above. Inhibitors sparingly soluble in water were added to the reaction medium by means of organic solvents as vehicle, either dimethylsulfoxide for quinidine and erythralosamine, acetone for nifedipine, methanol for a-naphthoflavone, TAO and gestodene. Control reactions were performed by adding to the reaction medium the same amount of solvent that was needed for the addition of inhibitors. The volume of solvent never exceeded 0.2% (V/V). Sulfaphenazole was added in a phosphate buffer/ KOH 60 mM mixture (50/50, V/V).
`
`Page 3
`
`N-demethylation (1 l), erythromycin N-demethylation, methadone
`and mfedipine oxidation (13). Other monoxygenase activities specific to P450
`

`
`Buprenorphiee Dealkylation by P450 3A4 Vol. 60, No. 22,
`
`1997
`
`Immunoblot analvsis
`
`(20 pg)
`were separated by electrophoresis and transferred to nitrocellulose sheets (Hybond-C, Amersham, UK) according to usual procedures (16, 17). A human monoclonal anti-P450-NIF (18) antibody was used as previously described
`(9,
`10). This anti-human P450 3A4 did not distinguish between the three members of the 3A family. Thus, the blots quantified by means of image processing scan analysis (Bioprofil, Wilber-Lourmat, France) represented the amount of total P450 3A in liver microsomes.
`
`Immune-inhibition of BPN metabolism
`
`Microsomal preparation from Br022 liver containing 70 pmol of total P450 in 0.1 mM potassium phosphate buffer (pH 7.4) in the absence of BPN and NADPH was incubated at room temperature for 20 min, in the presence of 50 pL of rabbit serum polyclonal antibody raised against rat P450 3A2 (Gentest, Wobum, MA, USA) or non-immune rabbit serum. Reaction was started by 0.07 mM BPN and 1 mM NADPH additions and processed as described above. This antibody did not cross react with other P450 families other than 3A family, but cross-reacted with all the P450 3A enzymes in all mammalian species according to the manufacturer.
`
`Heterolopouslv enDressed P4.W Droteins.
`
`Cell microsomes containing pure human P45Os were purchased from Gentest. Human B-lymphoblastoid cell lines, transfected with human P450 lA1, 1A2, 2A6, 2B6, 2C8, 2C9, 2Cl9, 2D6, 2El cDNA expressed high specific enzymatic activities. Three microsomal preparations of human P450 3A4 heterologously expressed were used, each characterized by its testosterone 6B-hydroxylation activity expressed as turnover number (TN, rnin~‘). Microsomes containing P450 3A4 and P450 reductase with a TN of 30 mini, overexpressed P450 3A4 and P450 reductase with a TN of 6.4 min.’ and overexpressed P450 3A4, P450 reductase and cytochrome b5 with a TN of 173 min“ were purchased from Gentest.
`
`&tistical analvsis
`
`r,
`
`r,
`p
`0.05.
`were considered to be statistically significant when
`Km
`
`were calculated using an ANOVA table by the least squares regression analysis from the raw data. As a quite normal Gaussian distribution in the population was observed (skewness = 0.45, n = 18) correlation coefficients were calculated by including all the raw data. Correlation coefficients,
`and Vm were calculated by means of a non linear regression analysis based upon Wilkinson method analysis (19) by using the Enzpack 3 Software (Biosoft, Cambridge, UK). This method uses a non graphical method for estimating kinetic parameters. Provisional estimates of Km and Vm were calculated by using a weighted tit of data to a straight line plot. These provisional estimates were then adjusted by means of a non linear regression analysis to fit the best hyperbola to data.
`
`Results
`
`Metabolism of bwrenorohiw
`
`phase to 3 HPLC conditions were optimized, in particular, by adjusting the pH of mobile and adding triethylamine to mobile phase. One main metabolite was identified when buprenorphine was incubated with human liver microsomes supplemented with NADPH (Figure 2. Its identity, N-dealkylated buprenorphine, was based upon its retention time, UV and fluorescence spectra identical to pure compound. Preliminary experiments suggested the involvement of P450 in this dealkylation reaction. This evidence included localization of catalytic
`
`Page 4
`
`Protein samples
`Correlation coefficients,
`was less than
`

`
`Vol. 60, No. 22,1!397 Buprenorphine Deakylation by P450 3A4 1957 activitv in the microsomal subcellular fraction and its absolute dependence on the presence of NAD$H (Fizure 2) ((A) 7 I 1
`
`0
`
`0
`
`2omin FIG. 2 Chromatographic metabolic profiles of 0.2 mM buprenorphine (BPN) incubated with microsomal preparation of BrO56 liver sample, (A) in presence, (B) in absence of 1 mM NADPH, (C) incubation of 0.2 mM buprenorphine with a microsomal preparation containing heterologously overexpressed P450 3A4, P450 reductase and cytochrome b5. NBPN : norbuprenorphine; Xl, X2 : unknown metabolites. Kinetic mmmeters Figure 3 shows the double reciprocal plot of norbuprenorphine formation from BPN by microsomal preparation from the human liver BrO56. Km was determined to he 85 + 16 pM with a Vm 3.14 f 0.33 nmol/(min x mg of protein). The use of a concentration as low as 10 @I did not allow to detect biphasic steady state kinetics. Km was determined on three samples (FH3, BrO55 and BrO56) as 89 + 45 uM (mean It SD). In order to establish the metabolic rates of BPN-dealkylation in the panel of 18 liver microsomes the drug was therefore incubated at the saturating concentration of 0.2 mM so that the intra-individual rate differences reflected the level of P450 enzyme(s) involved.
`
`Page 5
`
`2omin
`

`
`1958
`
`Vol.
`
`60, No. 22, 1997 4- 2 Vm = 3.14 nmol/min/mg 8 _ Km=85uM T 2 - 3- 2- l- y = 0.3 18 + 27.07 x 0, -0.04 0.04 I- 0.08 0.12 11 [BPN] uM FIG. 3 Double reciprocal plot of velocity (nmol/(min x mg of protein) against buprenorphine concentration (pM_‘) based on liver microsomes from Br056 sample. Km and Vm were calculated by using a weighted tit of data to a straight line plot, by means of a non linear regression analysis based upon Wilkinson method analysis. co elation between butwenolphine metabolism and d$7krent monooxvgenase activities in human liveTmicrosome8 If two reactions are catalyzed by the same enzyme, then the rates should be correlated to each other when compared in a wide panel of microsomal preparations containing varying levels of the enzyme. The Ndealkylation of BPN presented a 5-fold inter-individual variation ranging from 0.93 to 4.28 nmo&nin x mg of protein) with a mean of 2.30 f 1.18 mnoY(min x mg of protein). The Ndealkylation of BPN correlated significantly with 6 catalytic activities known to be mediated by the P450 3A4 enzyme, namely nifedipine oxidation (r = 0.83), testosterone 68 hydroxylation (r = 0.70), erythromycin N-demethylation (r = 0.84), methadone N-demethylation (r = 0.69), tamoxifene and totemifene N-demethylation (r = 0.70 and r = 0.92, respectively). Conversely, no significant correlations were found for substrate markers of P450 1 A2 (phenacetin 0-deethylation r = 0.33; methoxyresomfin Odemethylation r = 0.37), P450 2El (chlorzoxazone 6-hydroxylation r = 0.06, nitmsodimethylamine demethylation, r = 0.16), P450 2C9 (S- mephenytoin 4-hydroxylation, r = 0.35 or tolbutamide hydroxylation r = 0.52), P450 2A6 (cournarin 7-hydroxylation r = 0.10) and P450 2D6 (dextromethorphan Odemethylation, r = 0.02). The N-dealkylation of BPN correlated significantly with the amount of total P450 3A immunodetected by a monoclonal anti-human P450 3A4 antibody (r = 0.87, p<O.OOl) by means of Western blot technique. The regression line failed to intersect the y axis at zero. This lack of zero interception may be explained in part by the assay that is insensitive to detect traces (Figure 4).
`
`Page 6
`
`Buprenorphine &alky&m by P450 3A4
`

`
`Vol. 60, No. 22,1997 Buprenorphine Deakylaticm by P450 3A4 1959
`
`0
`
`1000 2600 3000 4000 BPN deakylation (pmol/minlmg) FIG. 4 Correlation between N-dealkylation of buprenorphine (pmo&n.in x mg of protein)) and hepatic content of immunodetectable P450 3A family (relative arbitrary units determined by image processing quantification). Correlation coefficient, r, was calculated by the least squares regression method. . . .
`
`nhdxtron of BPN dealkvlation bv different substrates
`
`To access further that the main metabolic pathway of BPN is catalyzed by the P450 3A enzymes, as suggested by the correlation studies, the inhibitory effect of different substrates or inactivators of P450 was tested on liver microsomal samples. w reports the percentage of BPN dealkylation inhibition by three mechanism-based inhibitors of P450 3A4, namely TAO, gestodene and erythralosamine. Inhibition was between 62 and 7 1% with respect to controls. oestodene Erythral TAO Inhibitors FIG. 5 Percent residual activity (vs controls) of buprenorphine N-dealkylation in three microsomal human liver samples (FIT1 m, PH3 (cid:144)d and BrO43 B) by three mechanism-based inhibitors of P450 3A4 (gestodene, erythralosamine, termed erythral, and TAO). Inhibitors were pre-incubated at different concentrations (TAO at 0.25 mM, gestodene and erythralosamine at O.lmM) in presence of NADPH. Numbers above graphs represent mean f SD of the three microsomal sample used.
`
`Page 7
`
`

`
`Dealkylation by P450 3A4
`
`Vol. 60, No. 22, 1997
`
`Ketoconazole, a potent and specific inhibitor of P450 3A4, inhibited buprenorphine N- dealkylation up to about 75% (vs control) when used up to 40 fold its Ki, i.e. 0.6 PM (Table I). TABLE I Effect of ketoconazole on buprenorphine dealkylation. Results are expressed as residual activities (mean + SD % vs control). Human liver microsomal fractions (PH3, BrO55 and Br056 samples) were incubated with buprenorphine 7OjtM, 1 mM NADPH and increasing concentrations of ketoconazole. Ketoconazole (ClM) Residual buprenorphine dealkylation (% vs control) 0.2 0.4 0.6 1.5 63.9 f 11.8 41.5 + 12.7 27.6 f 9.9 24 + 3 Nifedipine, a specific probe of P450 3A4, was shown to be a competitive inhibitor of BPN dealkylation with an apparent Ki of 129 uM (&rue 6). Chemical inhibition of P450 1A (a- naphthoflavone), P450 2C (sulfaphenazole) and P450 2D6 (quinidine) and P450 2El (DMSO 7OmM) did not significantly inhibit BPN metabolism (data not shown).
`
`260
`[I]
`
`.
`
`’ 300
`
`nifedipine m PIG. 6 Competitive inhibition of buprenorphine N-dealkylation by nifedipine. Liver microsomal preparation from BrO56 sample was assayed for buprenorphine (0 35@Q, (0 70 l&i) and (m 140 ltM) jmmuno-inhibition qf BPN deallqvlation In order to confirm that P450 3A enzymes are involved in BPN dealkylation, immuno-inhibition studies with anti-rat P450 3A2 antibody were carried out. This study showed that inhibition was up to 78% vs control activity (-7). This antibody was able to inhibit nifedipine oxidation, specific to P450 3A4, up to 84% under the same conditions as for BPN.
`
`Page 8
`
`Buprenorphiae
`

`
`1997
`
`Bupreaorphiae Dea&yhhon by P450 3A4
`
`1961
`
`This anti-rat P450 3A2 was shown to cross-react with all P450 3A isoforms but not with other P450 families. Non-immune Immune FIG. 7 Immuno-inhibition of buprenorphine N-dealkylation (a) and nifedipine oxidation (Cl) by rabbit antibody raised against rat P450 3A2. Microsomes from Br022 liver sample were preincubated with 50 pL of rabbit nonimmune serum or anti-P450 3A2. Results are for individual data. J3 I Q
`MicrosomaJ preparations of human cells genetically engineered for stable expression of 10 human P45Os were used to test their capacity to catalyze BPN dealkylation. P450 3A4 was the only isoform able to dealkylate BPN (&re 10 Three preparations of P450 3A4 were tried (J&j&&L Presence of NADPH P450 reductase (OR) and cytochrome b5 (b5), both involved in the electrons transfer to P450 in microsomal preparations, dramaticahy increased the velocity of BPN dealkylation. By taking into account the mean content of immunodetectable P450 3A4K45 isoforms in human liver (20, 21), i.e. about 300 pmol/(min x mg of protein), BPN dealkylation can be assumed to be about 2886 pmoV(min x mg of protein) in human liver microsomes. TABLE II Buprenorphine dealkylation activity by human heterologously expressed P45Os. 0.5 mg of protein were incubated for 30 minutes at 37°C with 0.2 mM BPN and 1 mM NADPH. Cell expressing Total P450 = BPN dealkylation b Turnover number” BPN dealkylation d 3A4 + OR 444.4 1264 2.84 852 3A4+OR+b5 256.4 2470 9.63 2886 ’ pmol/(min x mg of proteins). D Activity expressed as pmoY(min x mg of protein). ’ Turnover number expressed as pmoY(min x pmol of P450) or mm-‘. d hepatic activity calculated from the turnover number x 300 pmol/ mg of protein, mean content of immunodetectable P450 3A4/3A5 isoforms in human liver microsomes, expressed as pmoY(min x mg of protein).
`
`‘n
`
`sed P
`
`roteins
`
`Page 9
`
`Vol. 60, No. 22,
`r .
`

`
`Buprenorphiae Dealirylation by P450 3A4 Vol. 60, No. 22, 1997 Discussion Buprenorphine is a semi-synthetic molecule derived from thebaine, an opioid alkaloid. Since the discovery of its analgesic properties, BPN was shown to have pharmacological properties very different from other morphinic molecules. Thus, BPN is a mixed agonist antagonist molecule that has tendered it an effective treatment for narcotic addiction (2). Although it was an (< orphan drug >p, it constitutes a basic molecular for substitute treatment for opiate abuses (4). Facing to public health problems due to the deleterious effects of heroin dependence associated with transmission of HIV or hepatitis C viruses, some governmental organizations approved it as maintenance drug for opioid-dependent patients. Buprenorphine’s long duration of action may result more from its slow dissociation from opioid receptors than from its plasma half- life. However, buprenorphine metabolism may regulate its pharmacological effects. Despite of its widespread use as a substitutive treatment for opioid abuse, there are not direct studies on the nature of enzymes involved in BPN metabolism. BPN has previously been shown to be extensively metabolized in the liver by N- dealkylation and glucuro-conjugation of drug itself and its N-dealkylated metabolite (6). No free parent drug was detected in urine. The present study demonstrates that BPN is efficiently N- dealkylated by microsomal preparations from human liver with metabolic rates about 4 fold those measured for methadone (12). Only norbuprenorphine was detected as metabolite of buprenorphine when incubated with liver microsomal preparations supplemented with NADPH. The compound, resulting from a molecular rearrangement involving loss of methanol followed by ring formation between the side chain and methoxy group (22), was never detected in the analytical conditions reported in this study. BPN exhibited monophasic steady state kinetics with a Km of 89 @I in man. As the buprenorphine plasma level was about 10 nM during chronic treatment (23), the involved enzyme catalyses reaction at non saturating substrate concentration. The present work demonstrates that human liver P450 3A4 has a major role in the N-dealkylation of BPN. Such a conclusion is based on four kinds of results: correlation studies, P450 3A4-selective chemical inhibitions, immuno-inhibition by antibody raised against P450 3A2 and metabolism by heterologously expressed P45Os. Indeed the N-dealkylation of buprenorphine, which varied 5-fold between individuals, was shown to be significantly correlated with six P450 3A specific monooxygenase activities in a large panel of 18 human livers. As the BPN metabolism was measured at a saturating substrate concentration of 0.2 mM, the differences in reaction rates reflect the level of the involved P450 isoform. The use of various P450 3A selective chemical inhibition similarly points out the involvement of P450 3A4 in BPN metabolism. Gestodene (15), erythralosamine (24) and TAO (25), well-known to be mechanism-based inhibitors of P450 3A4, inhibited the N-dealkylation of BPN by up to 70 %. TAO has been considered as selective inhibitor as polyclonal antibodies directed against P450 3A enzyme (25). Furthermore, nifedipine, a prototype substrate of P450 3A4 (13) competitively inhibited BPN deall@ation with an apparent Ki of 129 PM. Conversely, chemical inhibition specific to P450 1A (a-naphthoflavone), 2C (sulfaphenazole), 2D6 (quinidine), 2El (dimethylsulfoxide) did not significantly inhibit the BPN metabolism. The extent to which a biotransformation mediated by a human liver microsomal preparation is reduced by an isoform-selective inhibitory probe indicates the extent to which that isoform is involved in the total metabolic activity. Ketoconazole was demonstrated to be a potent and specific inhibitor of P450 3A4 when used up to 40 fold its Ki (26). A very potent concentration-dependent inhibition of buprenorphine dealkylation by ketoconazole was observed. This result suggest that at least 75% of buprenorphine dealkylation is supported by P450 3A4 in human liver fractions. Finally, the use of 10 heterologously expressed P450 enzymes confiied the involvement of P450 3A4 in BPN dealkylation. On the basis of the relative level of P450 isoforms
`
`Page 10
`
`

`
`Vol.
`
`Buprenorphk Dakyhth
`
`by P450 3A4
`
`1963
`
`in human liver microsomes, it can be estimated that P450 3A4, which contributes 30-50% of the total P450 pool in liver (27), is the major enzyme involved in BPN dealkylation. Such an assumption, however, is limited by some caveats. Fist, estimates of P450 content in human liver are based on immunodetectable proteins and not on active P450 enzyme content (20,21). Second, it must be borne in mind that rough estimates of BPN dealkylation activity calculated, i.e. 9.8 min. ’ x 300 pmole P450 3A/mg of protein = 2886 pmoW(min x mg of protein), represent an average. However, as this calculated value is in full agreement with the experimental value, namely 3446 f 430 pmol/(min x mg of protein), it can be asserted that P450 3A4 is the only P450 isoform involved in BPN dealkylation, at least at concentrations of 50-100 w. Furthermore, the turnover number of BPN dealkylation dramatically increased when incubation media contained NADPH P450 rcductase and cytochrome b5. Such a result confii the involvement of these two components of the P450-dependent drug metabolism system at the step of electron transfer to P450 3A4. The drugs which have been characterized as being substrates or inducers of human P450 3A4 include erythromycin, nifedipine, testosterone, cortisol, diltiazem, cyclosporine, diazepam, propafenone, midazolam, imipramine, rifampicin, tamoxifene and dexamethasone (for a complete list, see ref. 28). It can therefore be speculated that any of these drugs might lead to interaction with BPN when given in association. However, it should be kept in mind that the extent of such interactions is likely to be modulated by various factors such as dose, biodisponibility and relative Km or Ki of both drugs. Thus, it has to be emphasized that BPN therapy may be complicated when given in association with drugs known to be inducers or inhibitors of P450 3A enzymes. Intrinsic clearance was calculated as the Vm/Km ratio (29). Vm was scaled for the amount of total enzyme present in the whole liver, assuming that a human liver contains about 50 g of microsomal proteins (29). Such scaling factors have been shown to be applicable for a wide range of drugs (30). Thus, intrinsic clearance was estimated to be around 1,600 mumin. From pharmacodynamic data of buprenorphine (31, 32; t,,Z = 2 hr,
`200 L) the intrinsic clearance was calculated as 1,200 mL/min. Thus, the in vitro determination of pharmacokinetic parameters was correlated with the in vivo data reported in this study. Such high clearance of buprenorphine (>l,OOO m.Umin) approaches hepatic blood flow and agrees with the substantial first-pass effect of liver and its low oral biodisponibility. The first-pass effect by the liver is so substantial that BPN must be administered by a sublingual route to avoid its hepatic metabolism. In conclusion, the knowledge of nature of P450 involved in bupmnorphine dealkylation should contribute to improve treatment for opioid dependence. Indeed, buprenorphine holds great promise for such a treatment with acceptable safety and efficacy. Acknowledgements This work was financially supported by the Ministi?re de 1’Enseignement Sup&ieur et de la Recherche (Equipe d’accueil948) and by Programme Hospitalier de Recherche Cliniclue (CHU- Brest). The authors wish to thank Dr. M. Delaforge (URA CNRS 400, Universite Rend Descartes, Paris), Schering SA (Lys-Lez-I-annoy, France) and Schering-Plough (Herouville- Saint-Clair, France) for their gifts of erythralosamine, gestodene and BPN and NBPN, respectively. References 1. R.C. HEEL, R.N. BRGGDEN, T.M. SPEIGHT and G.S. AVERY. Drugs 17 81-110 (1979). 2. N.K. MELLO. and J.H. MENDELSON. Science 2122 657-659 (1980).
`
`Vd =
`
`Page 11
`
`60, No. 22 l!W7
`

`
`1964 Buprenorpbhe DealkyWon by P4SO 3A4 Vol. 60, No. 22, 1997 3. 4. Z: 7. :: 10. 11. 12. 13. 14. 15. 16. 17. 18. ::: 21. 22. ;:: 25. W.K. BICKEL, M.L. STITZER, G.E. BIGELOW, I.A. LIEBSON, D.R. JASINSKI and R.E. JOHNSON. Clin. Pharmacol.Ther. & 72-78 (1988). A. COWAN and J.W. LEWIS. Buprenorphine combatting drug abuse with a unique opioid. Wiley-Liss, New York (1995). A. HUGUET-LEVET. Ann. Pharm. 5a. 124-130 (1995). E.J. CONE, C.W. GORODETZKY, D. YOUSEFNEJAD, W.F. BUCHWALD and R.E. JOHNSON. Drug Metab. Dispos. 12.577-581 (1984). D.S. WEINBERG, C.E. INTURRISI, B. REIDENBERG, D.E. MOULIN, T.J. NIP, S . WALLENSTEIN, R.W. HOUDE and K.M. FOLEY. Clin. Pharmacol. Ther. 44 335-342 (1988). A.D. RODRIGUES. B&hem. Pharmacol. 12 2147-2156 (1994). F. JACOLOT, I. SIMON, Y. DREANO, P.H. BEAUNE, C. RICHE and F. BERTHOU. B&hem. Pharmacol. 41 1911-1919 (1991). V. KERLAN, Y. DREANO, J.P. BERCOVICI, P.H. BEAUNE, H.H. FLOCH and F. BERTHOU. Biochem. Pharmacol. 4 1745-1756 (1992). F. BERTHOU, Y. DREANO, C. BELLOC, L. KANGAS, J.C. GAUTIER and P.H. BEAUNE. B&hem. Pharmacol. a 1883-1895 (1994). C. IRIBARNE, F. BERTHOU, S. BAIRD, Y. DREANO, D. PICART, J.P. BAIL, P.H. BEAUNE and J.F. MENEZ. Chem. Res. Toxicol. 9 365-373 (1996). F.P. GUENGUERICH, M.V. MARTIN, P.H. BEAUNE, P. KREMERS, T. WOLFF and D.J. WAXMAN. J. Biol. Chem. a505 l-5060 (1986). D.J. NEWTON, R.W. WANG and A.H.Y. LU. Drug Metab. Dispos. 22 154-158 (1995). F.P. GUENGUERICH. Chem. Res. Toxicol. 3 363-371 (1990). W.K. LAEMMLI. Nature 227 680-685 (1970). H. TOWBIN, T. STAEHELIN and J. GORDON. Proc. Natl. Acad. Sci. USA 26 4350- 4354 (1970). P. BEAUNE, P. KREMERS, F. LETAWE-GOUJON and J. GIELEN. B&hem. Pharmacol. &j 3547-3552 (1985). G.N. WILKINSON. Biochem. J. 8a 324-332 (1961). F.P. GUENGUERICH and C.G. TURVY. J. Pharmacol. Exp. Ther. 256 1189-1194 (1991). T. SHIMADA, H. YAMAZAKI, M. INORA, Y. INUI and F.P. GUENGUERICH. J. Pharmacol. Exp. Ther. 270 414-423 (1994). E.J. CONE, C.W. GORODETZKY, W.D. DARWIN and W.F. BUCHWALD.
`
`Pharmac. Sci. 22 243-246 (1984). SUBUTEX - monographie. (SCHERING-PLOUGH, FRANCE). E. SARTORI, M. DELAFORGE, D. MANSUY and P. BEAUNE, Biochem. Biophys. Res. Comm. 128 1431-1439 (1985). M.J. NAMKUNG, H.L. YANG, J.E. HULLA and M.R. JUCHAU. Mol. Pharmacol. 34 628-637 (1988). 26. M. BOURRIE. V. MEUNIER. Y. BERGER and G. FABRE. J. Pharmacol. Exn. Ther. m 321-332 (1996). I 27. V. CARRIERE, F. BERTHOU, S. BAIRD, C. BELLOC, P. BEAUNE and I. DEWAZIERS. Pharmacogenetics, 6 203-211 (1996). 28. A.P. LI, D.L. KAMINSKI and A. RASMUSSEN. Toxicology 104 l-8 (1995). 29. K.A. HAYES, B. BRENNAN, R. CHENERY and J.B. HOUSTON. Drug Metab. Discos. 22 349-353 (1995). 30. J.B. HOUSTON. Biochem. Pharmacol. &Z 1469-1479 (1994). 3 1. R.E. BULLINGHAM, H.J. MC QUAY and R.A. MOORE. Clin. Pharmacokinet. & 332- 343 (1983). 32. H.J. MC QUAY, R.A. MOORE and R.E.S. BULLINGHAM. Adv. Pain Res. Ther. 8 271- 283 (1986).
`
`J .
`
`Page 12