Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Cytochrome P-450 (CYP) monoxygenases are probably the most important enzymes for hepatic drug metabolism, which is crucial for the elimination of many therapeutic drugs. The activity of this group of enzymes or a single CYP can determine a patient's response to drug therapy. Therefore, modulation of the activity of CYPs by a given drug is a critical issue for the assessment of safety and efficacy of a drug. Especially inhibition of CYP can increase systemic exposure, thereby causing severe toxic side effects of the drug or another concomitantly given medication that is metabolized by the respective CYP(s) (Jurima-Romet et al., 1994; Wandel et al., 1998). Due to the recent progress in CYP enzymology and biochemistry, such interactions based on enzyme inhibition can now be straightforwardly investigated using in vitro technologies with microsomes, expressed enzymes, or cell systems (Pichard et al., 1990; Von Moltke et al., 1994). Whereas many reports are available on in vitro inhibition of P-450, the opposite effectactivation of CYPsis much less frequently encountered. Effects of substrate activation, which is defined as enzyme activation by the substrate increasing its own rate of metabolism, as well as activation by one compound affecting the metabolism of another compound were both reported for CYP 3A enzymes (Schwab et al., 1988). -Naphthoflavone is the best known activator of CYP 3A (Buening et al., 1981).

During previous investigations with human liver microsomes on the oxidative metabolism of meloxicam, we observed that quinidine markedly increased the rate of CYP-dependent metabolism. Meloxicam is a new nonsteroidal anti-inflammatory drug used for the treatment of rheumatic disease and acting by selective inhibition of cyclooxygenase-2 (Engelhardt et al., 1995). In vitro and in vivo, it is mainly metabolized to a 5'-hydroxymethyl metabolite (Fig. 1) that is further converted to a 5'-carboxy metabolite (Schmid et al., 1995). The 5'-hydroxylation of meloxicam is predominantly catalyzed by CYP 2C9 and with a minor contribution by CYP 3A4 (Chesne et al., 1998). As a consequence, biphasic enzyme kinetics were observed in experiments with human liver microsomes (concentration of meloxicam was 1.25-1000 µM). Average K_m values were 14 µM and 380 µM for CYP 2C9 and CYP 3A4, respectively (Chesne et al., 1998). CYP 3A4 also catalyzes hydroxylation and N-oxidation of quinidine, which is a therapeutically used class 1A antiarrhythmic.

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Chemical structure of meloxicam. The arrow indicates the site of hydroxylation.

The present study was performed to further elucidate the in vitro interaction between meloxicam and quinidine. Enzyme kinetics on the activation of meloxicam hydroxylation were performed, and, vice versa, effects of meloxicam on quinidine metabolism were also investigated. In addition, we investigated the activation of meloxicam metabolism by compounds structurally related to quinidine. Those data are of interest to get insight into the possible mechanism of activation and to assess possible consequences in vivo.

It was the aim of our studies to obtain information on the mechanism of CYP 3A4 activation, which is discussed in several recent publications and might help to better understand some of the special characteristics of CYP 3A4. Another focus of our work was to assess the potential of in vivo drug-drug interactions between concomitantly dosed meloxicam and quinidine.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Chemicals. Meloxicam (4-hydroxy-2-methyl-N-(5-methyl-2-thiazolyl)-2H-1,2-benzothiazine-3-carboxamide-1,1-dioxide) was prepared by the Department of Medicinal Chemistry. [¹⁴C]Meloxicam (radiochemical purity >99%, HPLC) and its 5'-methylhydroxylated metabolite (radiochemical purity >99.5%, HPLC), labeled at the carbonyl carbon, were prepared by the radiochemistry group at Boehringer Ingelheim Pharma KG (Biberach, Germany). NADP, glucose 6-phosphate dehydrogenase, and glucose 6-phosphate were from Boehringer (Mannheim, Germany). Quinidine was purchased from Aldrich (Steinheim, Germany). NADPH was from Sigma Chemie (Deisenhofen, Germany). Hydroxychloroquine was purchased from Acros Organics (Geel, Belgium); sulfaphenazole, oxidized nifedipine, and ketoconazole were purchased from Salford Ultrafine Chemicals (Manchester, UK). Quinidine N-oxide was synthesized by peracid oxidation according to Guengerich et al. (1986b). All other reagents and solvents were products at the highest quality available.

Biological Materials Pooled human liver microsomes of donor 6 and 7 and individual human liver microsomes of donor 2 were supplied by Human Biologics Inc. (Phoenix, AZ). Individual microsomes of donor 55 were purchased from the International Institute for the Advancement of Medicine (Exton, PA). Microsomes containing recombinant human liver CYP isoforms were obtained from Gentest Corp. (Woburn, MA). These included microsomes from a B-lymphoblastoid cell line expressing CYP 3A4 or CYP 2C9 and control microsomes isolated from the same cell line as recombinant CYPs, but without the human liver CYP cDNA insert. Human liver microsomes were characterized in respect to protein concentration and CYP 3A4 and CYP 2C9-selective enzyme activities (testosterone 6-hydroxylation and tolbutamide hydroxylation, respectively) by the supplier.

Meloxicam 5'-Hydroxylation. Microsomal incubations contained meloxicam (1.25-1000 µM), microsomal protein (0.5-1.0 mg protein/ml), magnesium chloride (6 mM), NADPH (1.2 mM), or an NADPH-regenerating system (final concentrations 1.2 mM NADP, 0.7 U glucose 6-phosphate dehydrogenase/ml, 8 mM glucose 6-phosphate), and activator/inhibitor in a total incubation volume of 0.5 ml of 0.1 M Tris buffer, pH 7.4. Substrates/activators/inhibitors were added from freshly prepared stock solutions containing dimethylsulfoxide, the final concentration of which never exceeded 0.5% (v/v) in the incubation mixture. The concentration of organic solvent was kept constant in experiments that were directly compared with each other. After preincubation for 3 min at 37°C, the reaction was initiated by addition of NADPH or regenerating system and allowed to proceed for 30 to 60 min. 5'-Hydroxymeloxicam formation was linear up to an incubation period of 60 min. The reaction was terminated by freezing the incubates in dry ice/methanol or addition of 125 µl of ice-cold acetonitrile and vortex mixing. After centrifugation (5 min at 10,000g) an aliquot of the supernatant was directly injected into the HPLC system.

Testosterone 6-Hydroxylation (CYP 3A). Incubations contained [¹⁴C]testosterone (50 µM, 0.05 µCi), CYP 3A4 (1.0 mg of microsomal protein/ml), magnesium chloride (6 mM), and NADPH (12 mM) in a total volume of 100 µl of 0.1 M HEPES buffer, pH 7.4. The reaction was initiated by addition of NADPH and stopped after 20 min by addition of 50 µl of dimethylsulfoxide-acetone (10:0.2, v/v) at 4°C and vortex mixing. After centrifugation (5 min at 10,000g), 2 µl of the supernatant were directly spotted onto an high performance thin-layer chromatography plate.

Nifedipine Oxidation (CYP 3A). Nifedipine oxidation was performed as described by Guengerich et al. (1986a) with minor modifications as follows. Incubations contained nifedipine (200 µM), CYP 3A4 (1.0 mg of microsomal protein/ml), magnesium chloride (5 mM), and an NADPH regenerating system as described above in a total volume of 0.5 ml of 0.1 M potassium phosphate buffer, pH 7.8, and was performed at 37°C for 20 min. The reaction was stopped by addition of 250 µl of acetonitrile at 4°C and vortex mixing. After centrifugation (5 min at 10,000g), the supernatant was directly injected into the HPLC system. All manipulations involving nifedipine solutions were performed under dim light.

Quinidine 3-Hydroxylation. Incubations contained quinidine (0.5-160 µM), added from a stock solution in 2.5% acetic acid (v/v), CYP 3A4 (1.0 mg of microsomal protein/ml), magnesium chloride (5 mM), and NADPH (1 mM) in 0.1 M Tris buffer, pH 7.4, in a total volume of 0.5 ml. The reaction was initiated by addition of NADPH and stopped after 60 min at 37°C by the addition of 250 µl of acetonitrile and vortex mixing. After centrifugation (5 min at 10,000g) the supernatant was directly injected into the HPLC system. 3-Hydroxyquinidine formation was linear up to an incubation period of 60 min.

Chromatographic Conditions. Metabolites of meloxicam, nifedipine, and quinidine were analyzed using HPLC equipment (Hewlett Packard, Waldbronn, Germany) with precolumn enrichment (Schmid and Roth, 1987) on reversed phase columns, Bondesil C18, 40 µm, 17 mm × 4.6 mm i.d. (ICT Handels GmbH, Germany). For meloxicam, 3 min of enrichment with 1% aqueous ammonium formate solution (w/v) was followed by separation on Hypersil ODS (Shandon, Astmoor Runcorn, UK), 5 µm, 125 × 4.6 mm i.d. slurry packed and protected by a 17-mm guard column of the same material with a combination of step and linear gradient of 1% aqueous ammonium formate (w/v)-methanol (0-95%, v/v), a flow of 1 ml/min, and UV detection at 363 nm. ¹⁴C off-line measurement of meloxicam and its metabolites was performed by collecting the eluate in fractions (300 µl, corresponding to 18 s) in 24-well micro plates, addition of Microscint 40 (Canberra Packard, Germany) at a ratio of 3:1 (v/v), and followed by liquid scintillation counting (Topcount, Canberra Packard, Germany). The data of the liquid scintillation counting were processed with the CHROMI V1 software (Department of Pharmacokinetics and Drug Metabolism, Boehringer Ingelheim Pharma KG, Germany), and formation of 5'-hydroxymethylmeloxicam was calculated from the ratio of labeled metabolite to the total radioactivity. For nifedipine, enrichment with bidistilled water was followed by separation on Hypersil ODS (as described above) isocratically with methanol-water (55:45, v/v) and UV detection at 254 nm. Under these conditions, oxidized nifedipine and nifedipine had retention times of 8 and 12 min, respectively. Formation of metabolites (5'-hydroxymethylmeloxicam and oxidized nifedipine) was quantified by comparing the peak areas with those of authentic standards. For quinidine, enrichment with bidistilled water was followed by separation on Kromasil 100 C18, 5 µm, 120 × 2 mm i.d. protected by a 10-mm guard column of the same material (Grohm, Herrenberg-Kayh, Germany) with a linear gradient of formic acid (100 mM)-methanol (0-30%, v/v) and a flow of 0.25 ml/min. After addition of methanol-water-phosphoric acid (85%, w/v) (600:200:200, v/v/v) with a flow of 0.1 ml/min, peaks were detected fluorimetrically (excitation, 350 nm; emission, 450 nm) and were eluted after 7.2 min (3-hydroxyquinidine), 8.7 min (quinidine N-oxide), and 11.3 min (quinidine). At the time the experiments were performed, 3-hydroxyquinidine was not available. Therefore, formation of 3-hydroxyquinidine in the presence of meloxicam was calculated as percent of control values without meloxicam under identical incubation conditions. Incubations containing [¹⁴C]testosterone were analyzed by high performance thin-layer chromatography. Two microliters of the supernatant were spotted on high performance thin-layer chromatography plates (silica gel 60 F 254, 10 × 10 cm, Merck, Germany), developed in eluent I (dichlormethane-acetone, 40:10, v/v), dried, and developed in eluent II (chloroform-ethyl acetate-ethanol, 40:10:7, v/v/v). After drying, plates were analyzed by a phosphor imager (BioImaging analyzer Fuji, type BAS 2000, Fuji Photo Film Co., Germany). Formation of 6-hydroxytestosterone was calculated from the ratio of labeled metabolite to the total radioactivity.

Analysis of Data. K_m and V_max values were obtained initially by graphical analysis of Eadie-Hofstee plots. The resulting values were used as first estimates for iterative nonlinear regression analysis using the solver subprogram, which is implemented in Excel 5.0 to calculate values of K_m and V_max according to simple one enzyme-one substrate Michaelis-Menten-type enzyme kinetics:
where V_max is the limiting maximal velocity, K_m is the Michaelis-Menten constant, and S is the concentration of substrate.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Influence of Quinidine on Metabolism of Meloxicam. Investigations on meloxicam hydroxylation by human liver microsomes and expressed microsomal CYP 2C9 and CYP 3A4 in the presence of quinidine revealed that quinidine increased the rate of meloxicam hydroxylation by human liver microsomes to 140 and 280% compared with the control (control activities without quinidine: 12.3 pmol/min/mg protein) (Table 1). Quinidine (10 and 100 µM) had no effect on meloxicam hydroxylation by expressed CYP 2C9 (control activity without quinidine, 2.3 pmol/min/mg protein) but increased the rate of meloxicam hydroxylation by CYP 3A4 in a dose-dependent manner to 160% (10 µM) and 510% (100 µM) of control experiments (control activity without quinidine, 1.85 pmol/min/mg protein). Quinidine had no effect on control microsomes (microsomes isolated from the same cell line as recombinant CYPs but without the human liver cDNA insert). Incubating meloxicam with CYP 3A4 in the presence of quinidine, ketoconazole (5 µM), a potent competitive CYP 3A4 inhibitor (Maurice et al., 1992), completely inhibited quinidine-mediated activation of meloxicam hydroxylation and CYP 3A4-mediated meloxicam hydroxylation. Two different test reactions for CYP 3A activity, viz. testosterone 6-hydroxylation and nifedipine oxidation, showed no or only minute responses upon quinidine activation compared with control incubations (control activities, 390 pmol/min/mg protein and 1200 pmol/min/mg protein for testosterone 6-hydroxylation and nifedipine oxidation, respectively). In contrast, a reduction of substrate oxidation was found for nifedipine oxidation, which is consistent with an inhibition of CYP 3A4-dependent nifedipine oxidation by quinidine as reported earlier (Guengerich et al., 1986b).

Influence of Quinidine-Related Compounds on Metabolism of Meloxicam. The effect of quinoline derivatives that are structurally related to quinidine and -naphthoflavone, a known activator of CYP 3A4, on meloxicam hydroxylation was investigated. During incubation experiments with human liver microsomes, quinidine, quinine, and hydroquinidine (100 µM) increased the rate of meloxicam hydroxylation to 280%, 200%, and 480% of control incubations, respectively (Table 2) (control activity, 12.3 pmol/min/mg protein). Inhibition of the CYP 2C9 component of meloxicam hydroxylation by sulfaphenazole (10 µM), a selective CYP 2C9 inhibitor (Baldwin et al. 1995), resulted in a more pronounced activation (quinine, 610%; quinidine, 940%; and hydroquinidine, 2100%). This effect was due to "unmasking" the CYP 3A4-dependent activation (control value, 4.8 pmol/min/mg microsomal protein). Meloxicam hydroxylation was not activated by -naphthoflavone. Instead, a slight inhibition was observed that was consistent with an unspecific CYP inhibition of -naphthoflavone at concentrations of 10 and 100 µM. For such high concentrations, inhibition of CYP 2C9 and CYP 3A4 was observed (Chang et al., 1994; Newton et al., 1995). The activation of meloxicam hydroxylation by expressed CYP 3A4 was more pronounced than by human liver microsomes. Quinidine, quinine, and hydroquinidine increased the rate of meloxicam hydroxylation to 510%, 260%, and 1200%, respectively (Tables 1 and 3). The control activity without activator was 1.7 pmol/min/mg microsomal protein.

Kinetic Analysis of Meloxicam Hydroxylation by Human CYP 3A4 in the Presence of Quinidine and Hydroquinidine. Meloxicam hydroxylation by expressed CYP 3A4 was investigated in the presence of quinidine and hydroquinidine over a concentration range from 10 to 750 µM meloxicam and 0 to 100 µM quinidine and hydroquinidine. Michaelis-Menten parameters K_m and V_max with and without activator were calculated from V/S plots by nonlinear regression analysis. In the absence of activator, 5'-hydroxymethylmeloxicam formation followed simple one enzyme Michaelis-Menten kinetics and showed no deviation from linearity in the V/(V/S) plot. Upon increasing concentrations of quinidine or hydroquinidine, a decrease in K_m and an increase in V_max was observed. This resulted in a marked increase of the intrinsic clearance that is represented by V_max/K_m. The kinetic parameters are given in Table 4. Subsequently, the total set of experimentally determined data was fitted to the velocity equation for mixed-type nonessential activation as described under analysis of data (Fig. 2). The calculated values of V_max, the maximal velocity without activator (Table 5), were comparable with V_max values obtained by fitting the data to simple one enzyme-one substrate Michaelis-Menten type enzyme kinetics without activator (Table 4). Apparent K_m values for expressed CYP 3A4 (550 µM, Table 4) or the CYP 3A4 component of human liver microsomes (380 µM; Chesne et al., 1998) were also in a similar range as to K_s as calculated for the mixed-type nonessential activation model (Fig. 3), which was 310 µM and 270 µM for activation by quinidine and hydroquinidine, respectively (Table 5).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Effect of quinidine (A, Q) or hydroquinidine (B, HQ) on meloxicam metabolism by human CYP 3A4. Meloxicam (10-750 µM) was incubated in 0.1 M Tris buffer, pH 7.4, at 37°C with human CYP 3A4 (0.5 mg of protein/ml), an NADPH-generating system (consisting of 1.2 mM NADP, 0.7 U of glucose 6-phosphate dehydrogenase/ml, 8 mM glucose 6-phosphate), and in the presence of quinidine (1-100 µM) for 60 min (mean of duplicate or quadruplicate experiments). The complete set of measured values was fitted to the equation for mixed-type nonessential activation. A, -, 100 µM Q; +, 50 µM Q; , 10 µM Q; , 1 µM Q; , 0 µM Q. B, -, 100 µM HQ; +, 10 µM HQ; , 5 µM HQ; , 1 µM HQ; , 0 µM HQ.

View this table: [in this window] [in a new window]   TABLE 5 Kinetic parameters for activation of meloxicam metabolism by human CYP 3A4 by quinidine and hydroquinidine Experimental data obtained in the presence of activator (A) were fitted to the velocity equation for mixed-type nonessential activation. Vmax, maximal velocity of the unactivated reaction; Ks, dissociation constant of the enzyme-substrate complex; Ka, dissociation constant of the enzyme-activator complex; , factor by which Vmax changes when A occupies the enzyme; , factor by which Ks changes when A occupies the enzyme.

View larger version (8K): [in this window] [in a new window]   Fig. 3.   General scheme for nonessential activation. A, activator; E, enzyme; P, product; S, substrate; Ks, dissociation constant of the enzyme-substrate complex; Ka, dissociation constant of the enzyme-activator complex; kp, maximal velocity in the absence of effector; , factor by which Vmax changes when A occupies the enzyme; , factor by which Ks changes when A occupies the enzyme.

Kinetic Analysis of Quinidine Metabolism by Human CYP 3A4 in the Presence of Meloxicam. 3-Hydroxyquinidine formation and formation of quinidine N-oxide by human CYP 3A4 was investigated in the presence of meloxicam over a concentration range from 1 to 160 µM for quinidine and 0 to 750 µM for meloxicam. At the time of the study, no 3-hydroxyquinidine reference compound was available. Therefore, the linearity of the HPLC detector response was assessed using various dilutions of 3-hydroxyquinidine, which was obtained from an incubation experiment. Because linearity of detector response was satisfactory (correlation coefficient, 0.994; six calibration samples each assayed as duplicate, data not shown), the amounts of 3-hydroxyquinidine were expressed in arbitrary units (peak area). Proof of chemical structure of the major in vitro quinidine metabolite as 3-hydroxyquinidine was provided by HPLC-MS/MS measurements. A second metabolite of quinidine was identified as quinidine N-oxide by comparison with authentic standard. K_m of 3-hydroxyquinidine formation obtained with human liver microsomes was 84 µM; K_m of quinidine N-oxide formation was 160 µM. Michaelis-Menten parameters K_m and V_max for quinidine 3-hydroxylation in the presence and without meloxicam were calculated from V/S plots by nonlinear regression analysis. A decrease of K_m and V_max (Table 6) with increasing meloxicam concentrations was found, indicating uncompetitive inhibition. Subsequently, the total set of experimental data was fitted to the velocity equation for uncompetitive inhibition. This resulted in a V_max of 9.3 arbitrary units, a K_s of 84 µM, and a K_i of 410 µM (B = 0.9871). The Cornish-Bowden plot of the experimental data is shown in Fig. 4, confirming an uncompetitive inhibition mechanism. Formation of quinidine N-oxide was also influenced by meloxicam in a similar manner. K_m values decreased from 160 µM without meloxicam to 64 µM in the presence of 750 µM meloxicam (B values in the range of 0.9261-0.9959) without changing V_max/K_m (further data not shown).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Effect of meloxicam on quinidine metabolism by human CYP 3A4. Cornish-Bowden plot, S/V against I at various quinidine concentrations (5-160 µM) are shown. The complete set of measured values was fitted to the equation for uncompetitive inhibition (Ki = 410 µM, Ks = 84 µM). , 5 µM; , 10 µM; , 20 µM; , 40 µM; , 80 µM; , 160 µM.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The predominant P-450 enzyme involved in meloxicam metabolism is CYP 2C9, with a minor contribution of CYP 3A4 (Chesne et al., 1998). There was clear evidence that of the two cytochrome P-450 enzymes involved, CYP 2C9 was not affected by quinidine, and activation occurred exclusively with the CYP 3A4-catalyzed component. Several cases of activation of CYP 3A4-catalyzed reactions have been described in the literature, such as activation involving two different compounds or positive cooperative binding of more than a single substrate molecule (Ueng et al., 1997). The latter effect was not observed with meloxicam; there was no sigmoidicity in the Michaelis-Menten plots of meloxicam 5'-methylhydroxylation.

Different mechanisms of CYP 3A4 activation have been described in the literature. Shou et al. (1994) described simultaneous binding of substrate and activator in the active site. In a recent publication of Korzekwa et al. (1998), important aspects concerning cooperative binding of multiple substrates (effector is also substrate) were reported. The authors deduced that the inability of a substrate to displace the effector from the active site is indicative for simultaneous binding to the P-450 active site. Because of the effects on K_m in case of the meloxicam-quinidine system, the possibility of two different substrate and effector binding sites should be also discussed. ¹H NMR measurement with equimolar concentrations of meloxicam and quinidine in aqueous solution gave no indication of complex formation between the two compounds (data not shown). Huang et al. (1981) described another mechanism of cytochrome activation by enhancing the interaction between cytochrome P-450 and cytochrome P-450 reductase. However, no effect of the activator on K_m was observed in this model. To explain the effects of the meloxicam-quinidine and quinidine-meloxicam interaction, an allosteric mechanism with two distinct binding sites was assumed. The decrease of K_m most likely indicated an increase of the affinity of CYP 3A4 for meloxicam in the presence of quinidine or hydroquinidine. This can be explained by an effector binding site that, upon binding of an effector molecule, alters the character of the substrate binding site as outlined by Ueng et al. (1997). Interestingly, the vacant effector binding site has apparently no affinity for binding of meloxicam because no cooperativity (substrate activation) was observed. In turn, uncompetitive inhibition of quinidine metabolism by meloxicam indicated that meloxicam binds exclusively to the enzyme-substrate complex, not to the free enzyme, yielding an inactive enzyme-substrate-inhibitor complex (Segel, 1975). In the work by Ueng et al. (1997), which investigated the activation of aflatoxin B₁ metabolism by -naphthoflavone, there was a lack of activation or inhibition of -naphthoflavone 5,6-epoxidation by aflatoxin B₁. In contrast, mutual effects were observed for the meloxicam-quinidine system. We believe that our data are not conclusive enough to definitively discriminate between allosteric binding to two distinct binding sites or simultaneous binding to the P-450 active site. Both compounds, meloxicam and quinidine, are markedly different from the CYP 3A4 activator -naphthoflavone in respect to their chemical properties. Quinidine is a fairly strong base, whereas meloxicam is an acid. This might explain why there was no activation of meloxicam hydroxylation by -naphthoflavone and, in turn, quinidine did not activate testosterone 6-hydroxylation or nifedipine oxidation. We could show activation of meloxicam hydroxylation by quinidine, and preliminary results indicated a similar effect on the metabolism of piroxicam (activation by up to 500% in the presence of 100 µM quinidine). However, quinidine activation of CYP 3A4 might not be restricted to metabolism of the oxicam class of nonsteroidal anti-inflammatory drugs, but other CYP 3A substrates could be also affected. In a recent publication on the metabolism of seratrodast (Kumar et al., 1997), a clear increase (150%) of 5'-hydroxylation was observed in the presence of 2 µM quinidine in microsome samples of CYP 3A4 rich liver samples. Fentanyl N-dealkylation by CYP 3A4 was also increased in the presence of quinidine (Feierman and Lasker, 1996).

Our experiments have shown that meloxicam 5'-hydroxylation was activated by quinidine and structurally related compounds. This indicated that the hypothetical effector site accepts a variety of structurally different basic compounds and that there might be the potential for other classes of compounds to act in a similar way. Our results showed that hydroquinidine was the most potent activator of meloxicam metabolism. Stimulation by hydroquinidine could be of practical relevance, because several pharmacopoeias (e.g., British Pharmacopoeia, Pharmacopea Europea, and United States Pharmacopeia) allow a content of up to 15% and 20% hydroquinidine in quinidine preparations. Under the assumption of concentrations in the liver similar to blood plasma concentrations, after therapeutic doses of quinidine, concentrations of up to 18 µM quinidine (calculated as free base) (Jack, 1992) and 3 µM hydroquinidine could be reached. It was therefore interesting to assess the effects of therapeutic quinidine and hydroquinidine concentrations on meloxicam hydroxylation. The relative participation of CYP 3A4 and CYP 2C9 in relation to quinidine was exemplified using a human liver microsome sample with above average CYP 3A4 and CYP 2C9 enzyme activities (Fig. 5, top panel). A switch from CYP 2C9 as the major metabolizing enzyme to CYP 3A4 could be clearly seen (Fig. 5, bottom panel); a minor participation of 15 to 20% of CYP 3A4 at therapeutic meloxicam concentration (1-20 µM) changes to 60 to 70% in the presence of 18 µM quinidine. A clinically relevant effect on meloxicam drug therapy is not expected because this effect would probably cause a slight increase of metabolic clearance and thus lower meloxicam plasma concentration. Coadministration of meloxicam to patients receiving quinidine or other CYP 3A4 substrates should not result in relevant in vivo drug interactions because the K_i (410 µM) and K_m of meloxicam for CYP 3A4 (380 µM) (Chesne et al., 1998) are well in excess of therapeutic meloxicam steady-state concentrations (6.7 µM, after multiple peroral doses of 15 mg) (Tuerck et al., 1997). Quinine has recently regained interest in the therapy of malaria infections. Quinine plasma concentrations of 9 to 22 µM (calculated as free base) (Jack, 1992) are observed, which should result in similar but quantitatively even lower effects as therapeutic quinidine concentrations. The minute amounts of quinine present in various soft drinks (e.g., Indian tonic water, bitter lemon) have no relevance in respect to CYP 3A activation.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Relative participation of CYP 2C9 (solid line) and CYP 3A4 (dashed line) to meloxicam metabolism by human liver microsomes with above average activities of CYP 2C9 and CYP 3A4 (A, without quinidine; B, in the presence of 18 µM quinidine). Basal enzyme activities for CYP 2C9 (tolbutamide hydroxylation) and CYP 3A4 (testosterone 6-hydroxylation) were 0.25 and 10.5 nmol/min/mg protein, respectively. Km and Vmax values were used to calculate the contribution of CYP 2C9 and CYP 3A4 to meloxicam metabolism. Vertical lines are according to minimal and maximal blood plasma concentrations after single and multiple once-daily peroral meloxicam doses adjusted to 15-mg doses.

Taken together, our results give additional information on the unique properties of CYP 3A4 in respect to enzyme-substrate-effector interactions. We believe that an allosteric model is the most adequate to describe our results, even if no deductions on the proximity of the substrate and effector binding sites can be made. In addition, our results show that there is the possibility of substrate-effector interactions in respect to enzyme activation for therapeutically used drugs.