Abstract
(3R,4R)-4-Amino-1-((4-((3-methoxyphenyl)amino)pyrrolo[2,1-f][1,2,4]triazin-5-yl)methyl)-3-piperidinol (BMS-690514) is a potent inhibitor of ErbB human epidermal growth factor receptors (HER1, 2, and 4) and vascular endothelial growth factor receptors 1 to 3 that has been under clinical development for solid tumor malignancies. BMS-690514 is primarily cleared by metabolism with the primary metabolic pathways being direct glucuronidation (M6), hydroxylation (M1, M2, and M37), and O-demethylation (M3). In the current investigation, the metabolic drug-drug interaction potential of BMS-690514 was evaluated in a series of in vitro studies. Reaction phenotyping experiments with cDNA-expressed human cytochrome P450 (P450) and UDP-glucuronosyltransferase (UGT) enzymes and human liver microsomes (HLM) in the presence of P450 or UGT inhibitors suggested that CYP3A4, CYP2D6, and CYP2C9 were the major enzymes responsible for the oxidative metabolism of BMS-690514, whereas both UGT2B4 and UGT2B7 were responsible for the formation of M6. BMS-690514 did not cause direct or time-dependent inhibition of P450 enzymes (IC50 values ≥40 μM) in incubations with HLM and probe substrates of CYP1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, or 3A4. The compound also did not substantially induce CYP1A1, CYP1A2, CYP2B6, CYP3A4, or UGT1A1 at concentrations up to 10 μM in cultured human hepatocytes. Considering the submicromolar plasma Cmax concentration at the anticipated clinical dose of 200 mg, BMS-690514 is unlikely to cause clinically relevant drug-drug interactions when coadministered with other medications. In addition, because multiple enzymatic clearance pathways are available for the compound, inhibition of an individual metabolic pathway either via coadministered drugs or gene polymorphisms is not expected to cause pronounced (>2-fold) increases in BMS-690514 exposure.
Introduction
(3R,4R)-4-Amino-1-((4-((3-methoxyphenyl)amino)pyrrolo[2,1-f][1,2,4]triazin-5-yl)methyl)-3-piperidinol) (BMS-690514) is a highly selective and potent inhibitor of multiple human epidermal growth factor receptors (ErbB) and vascular endothelial growth factor receptors (VEGFRs), including EGFR (HER1), HER2, HER4, and VEGFRs 1 to 3. Preclinical and early clinical data have suggested that the combined inhibition of both the VEGFR and ErbB pathways could provide enhanced antitumor efficacy compared with the inhibition of one of these individual pathways (Tabernero, 2007; Tortora et al., 2008). BMS-690514 has been under clinical development as an oral treatment for non–small-cell lung cancer, metastatic breast cancer, and other solid tumor malignancies (Bahleda et al., 2008, 2009).
Previous studies indicated that BMS-690514 underwent extensive metabolism in rats, dogs, and humans (Christopher et al., 2010; Hong et al., 2010; Marathe et al., 2010). In vivo, the primary metabolic pathways included hydroxylation (M1, M2, and M37), O-demethylation (M3), and direct glucuronidation (M6) (Fig. 1) (Christopher et al., 2010; Hong et al., 2010). In humans, the flux through the M3, M6, and combined M2/M37 pathways was estimated to be approximately 20 to 25, 16, and 12 to 16% of the dose, respectively (Fig. 1) (Christopher et al., 2010). The flux through the M1 pathway could not be determined from the human absorption, distribution, metabolism, and excretion (ADME) study because M1 was thermally labile and decomposed into multiple unknown degradation products. It is likely that M1 did not survive in the gastrointestinal tract or was unstable during collection and processing of urine, bile, and fecal samples (Christopher et al., 2010; Hong et al., 2011). However, in vitro studies in human liver microsomes and in vivo human plasma profiles indicated that the flux through M1 could be significant (Christopher et al., 2010; Hong et al., 2011). Thus, the 0.3% of dose detected as M1 in human excreta is likely to be an underestimate of this pathway.
Human liver cytochrome P450 (P450) enzymes are involved in the clearance of a majority of drugs, and alteration of the activity of these enzymes could lead to drug-drug interactions (DDIs). If human in vivo data indicate that P450 enzymes contribute >25% of a drug's clearance, then identifying enzymes responsible for its metabolism becomes important (Guidance for Industry: Drug Interaction Studies—Study Design, Data Analysis and Implications for Dosing and Labeling, 2006; http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM072101.pdf). Likewise, when glucuronidation plays a significant role in the metabolic clearance of the compound, then identification of UDP-glucuronosyltransferase (UGT) enzymes becomes important. The phenotyping of UGT enzymes is only qualitative at this point, due in part to the lack of scaling factors to predict the in vivo DDIs (Bjornsson et al., 2003; Kiang et al., 2005) and a lack of specific probe substrates and inhibitors (Bjornsson et al., 2003).
Preliminary in vitro studies with expressed enzymes had suggested that CYP2D6 and CYP3A4 enzymes were likely to play an important role in the oxidative metabolism of BMS-690514. However, the UGT enzyme(s) responsible for the formation of the glucuronide metabolite (M6) was not identified (Marathe et al., 2010). The objective of the current study was to assess the possible metabolic DDIs of BMS-690514 as both a victim and a perpetrator. To evaluate the victim potential of BMS-690514, reaction phenotyping studies to identify the human enzymes responsible for the formation of primary metabolites (M1, M2/M37, M3, and M6) were conducted. To evaluate the perpetrator potential of BMS-690514, studies were conducted to assess its ability to inhibit the major P450 enzymes in human liver microsomes (HLM) and to induce the expression of P450 enzymes in primary cultures of human hepatocytes.
Materials and Methods
Chemicals.
[14C]BMS-690514 (radiochemical purity 98.96%, specific activity 36.8 μCi/mg) with the 14C uniformly labeled on the carbons of the aromatic ring and [13C6]BMS-690514 (purity 100%) with 13C labeled on the six carbons of the aromatic ring (Supplemental Fig. S1) were synthesized by the Radiochemistry Group at Bristol-Myers Squibb (Princeton, NJ). Unlabeled BMS-690514 and (3S)-(+)-(5-chloro-2-methoxyphenyl)-1,3-dihydro-3-fluoro-6-(trifluoromethyl)-2H-indol-2-one (BMS-204352; MaxiPost) were supplied by the Department of Chemical Development at Bristol-Myers Squibb. Furafylline, tranylcypromine, sulfaphenazole, quinidine, diethyldithiocarbamic acid, orphenadrine, troleandomycin, 1-aminobenzotriazole (ABT), ketoconazole, ticlopidine, 3,4-methylenedioxymethamphetamine, mifepristone, β-NADPH, UDP-glucuronic acid triammonium salt (UDPGA), alamethicin, bilirubin, serotonin, 2,6-diisopropylphenol (propofol), and fluconazole were obtained from Sigma-Aldrich (St. Louis, MO). Hyodeoxycholic acid (HDCA) and hecogenin were obtained from TCI America (Portland, OR). Tienilic acid was purchased from BD Gentest (Woburn, MA).
Pooled HLM (pooled from 20 donors), HLM allelic variant UGT1A1 (*28*28) (from a single donor), and human cDNA-expressed P450 and UGT enzymes (Supersomes) were purchased from BD Biosciences (San Jose, MA).
Incubations with HLM and cDNA-Expressed P450 Enzymes Fortified with NADPH.
[14C]BMS-690514 was incubated at 10 μM with pooled HLM (1 mg/ml) and individually expressed human P450 enzymes (100 pmol/ml each, including CYP1A1, CYP1A2, CYP1B1, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP2E1, CYP3A4, CYP3A5, and CYP4A11) in 100 mM phosphate buffer (pH 7.4), fortified with 1 mM NADPH and 2 mM MgCl2. The total reaction volume was 1 ml. The incubations were conducted at 37°C in a shaking water bath for 30 min. After the incubation period, an equal volume of ice-cold acetonitrile was added to quench the reaction. The samples were vortex-mixed and centrifuged at 13,000 rpm for 10 min. The resulting supernatants were mixed with an equal volume of water, and the samples were injected into the HPLC for metabolite profiling and mass spectrometric analysis.
HLM Incubations in the Presence of P450 Inhibitors.
BMS-690514, at concentrations of 1 and 10 μM, was incubated in triplicate at 37°C for 15 min with HLM (0.25 mg/ml), 1 mM NADPH, 2 mM MgCl2, and chemical inhibitors (various concentrations) in 0.1 M phosphate buffer (pH 7.4), total volume 1 ml. The chemical inhibitors used were furafylline (10 μM) for CYP1A2, tranylcypromine (2 μM) for CYP2A6, ketoconazole (1 μM) and troleandomycin (20 μM) for CYP3A4, orphenadrine (50 μM) for CYP2B6, montelukast (3 μM) for CYP2C8, sulfaphenazole (10 μM) for CYP2C9, benzylnirvanol (1 μM) for CYP2C19, quinidine (1 μM) for CYP2D6, diethyldithiocarbamate (50 μM) for CYP2E1, and ABT (1 mM) for all the P450 enzymes. Among these, time-dependent chemical inhibitors (furafylline, troleandomycin, orphenadrine, diethyldithiocarbamate, and ABT) were preincubated with HLM in the presence of NADPH for 15 min before the addition of the substrates. The stock solutions of these P450 inhibitors were prepared in methanol or DMSO, and the final organic content used in the incubations were <0.2% (v/v). At the end of the incubation period, an equal volume of ice-cold acetonitrile was added to each sample to quench the reaction. Samples were mixed and centrifuged at 13,000 rpm for 10 min. The resulting supernatants were then mixed with an equal volume of 1 μM [13C6]BMS-690514 (IS) prepared in deionized water, and the samples were analyzed by LC-MS/MS for metabolite quantification. HLM control incubations containing only methanol or DMSO solvent (no inhibitors) were also included. The relative percentage metabolite formation was determined by comparing the peak area ratios (metabolite to IS) in test incubations to appropriate solvent control incubations.
Substrate Concentration-Dependent Metabolism of BMS-690514 by P450 Enzymes.
To determine concentration-dependent metabolism of BMS-690514, incubations were conducted in triplicate with pooled HLM (0.25 mg/ml), expressed CYP2D6 (50 pmol/ml), CYP3A4 (50 pmol/ml), or CYP2C9 (50 pmol/ml) at 37°C in 0.1 mM phosphate buffer containing 2 mM MgCl2 and 1 mM NADPH for 7 min. The total reaction volume was 1 ml. At these incubation conditions, M1, M2/M37, and M3 were all formed at linear rates. The BMS-690514 concentrations used were 1, 2, 5, 10, 20, 40, 65, 125, 200, and 600 μM. The sample transferring, dilution, and incubation were done on a Genesis RSP 200 liquid handling workstation (Tecan, Durham, NC) equipped with eight tips and a shaking temperature control system (Yao et al., 2007). After the 7-min incubation, 120 μl of each reaction mixture was transferred to a hydrophilic 96-well filter plate (0.45 μm, polytetrafluoroethylene; Millipore Corporation, Billerica, MA), which was preloaded with 150 μl of ice-cold acetonitrile containing the 13C6-labeled oxidative metabolite internal standards (13C-IS) prepared as described below. The filter plate was then stacked on a 2-ml, 96-well receiving plate (BD Biosciences), which was preloaded with 270 μl of water. The samples were then passed through the filter plate membrane into the receiving plate by centrifugation at 2000g for 5 min. Finally, the receiving plate was vortexed, sealed with a polypropylene film, and the metabolite concentrations in each of the samples were determined by LC-MS/MS. The 14C-labeled calibration standards for M1, M2/M37, and M3 were prepared as described below and analyzed by LC-MS/MS to construct the calibration curves.
The rate of metabolism in incubations with HLM was calculated using the following formula: The rate of metabolism in incubations with expressed P450 enzymes was calculated using the following formula: Km and Vmax values for metabolite formation were determined by fitting the metabolite formation rates versus BMS-690514 concentration to the Michaelis-Menten equation [V = (Vmax · S)/(Km + S)] using a nonlinear regression analysis in Sigma plot (version 10; Systat Software, Inc., San Jose, CA) (Bjornsson et al., 2003).
The reported concentration values of individual P450 enzymes in HLM, 10, 108, and 96 pmol/mg microsomal protein for CYP2D6, CYP3A4, and CYP2C9, respectively (Rodrigues, 1999), were used to convert Vmax of CYP2D6, CYP3A4, and CYP2C9 for the formation of M1, M2/M37, and M3 to the units of picomoles per minute per milligram per protein so that the kinetic results from expressed enzyme studies could be compared directly with the HLM results.
Preparation of 14C-Labeled and 13C-Labeled Oxidative Metabolites As Quantification Standards for Kinetic Studies.
Because synthetic standards of oxidative metabolites (M1, M2/M37, and M3) were not available for use in quantification, the mixture of 14C-labeled oxidative metabolites were generated by incubating [14C]BMS-690514 with P450 enzymes. The concentration of each oxidative metabolite was then determined by LC radioprofiling, using relative peak area and specific activity. To achieve this, [14C]BMS-690514 (at a concentration of 100 μM) was incubated with a mixture of cDNA-expressed P450 enzymes (50 pmol/ml CYP1A1, 100 pmol/ml CYP3A4, 100 pmol/ml CYP2C9, and 10 pmol/ml CYP2D6) with 2 mM MgCl2 and 1 mM NADPH in 100 mM phosphate buffer (pH 7.4) at 37°C for 15 min, total volume 1 ml. After the reaction was quenched with acetonitrile, the mixtures were analyzed by LC radioprofiling, and the concentrations of 14C-labeled M1, M2/M37, and M3 were determined. These amounts corresponded to concentrations of 1.35 μM for 14C-labeled-M1, 0.5 μM for 14C-labeled-M2/M37, and 0.75 μM for 14C-labeled-M3 (Supplemental Fig. S2). The solution of 14C-labeled metabolites was serially diluted and used in the preparation of standard curves for quantification. The concentration ranges for 14C-labeled M1, M2/M37, and M3 standards were 0.023 to 1.35, 0.0021 to 0.5, and 0.0031 to 0.75 μM, respectively.
Oxidative metabolites (M1, M2/M37, and M3), labeled with 13C, were used as internal standards (13C-IS). To generate 13C-IS, [13C6]BMS-690514 (at a concentration of 100 μM) was incubated with a mixture of cDNA-expressed CYP1A1, CYP3A4, CYP2D6, and CYP2C9 at the same conditions as described above for generation of 14C-labeled standards. The resulting solution contained undetermined concentrations of 13C-labeled-M1, 13C-labeled-M2/M37, and 13C-labeled-M3. A constant volume of this mixture was added to each study sample as internal standards to facilitate quantification of each of the metabolites.
Incubations with HLM and Expressed UGT Enzymes Fortified with UDPGA.
[14C]BMS-690514 was incubated with pooled HLM and HLM isolated from genotyped liver tissue carrying the UGT1A1 (*28*28) allelic variant. The incubations contained [14C]BMS-690514 (1 mM) HLM (4 mg/ml) in 100 mM Tris-HCl buffer (pH 7.4) containing 2 mM UDPGA, 10 mM MgCl2, and 50 μg/ml alamethicin in a total volume of 1 ml. On the basis of data from the supplier, the HLM from the allelic variant had approximately 3.3-fold lower estradiol 3-glucuronidation activity than the pooled HLM. [14C]BMS-690514 was also incubated at 1 mM with expressed human UGT enzymes (1.25 mg/ml each, including UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8, UGT1A9, UGT1A10, UGT2B4, UGT2B7, UGT2B15, and UGT2B17) fortified with 2 mM UDPGA, 10 mM MgCl2, and 50 μg/ml alamethicin in 100 mM Tris-HCl buffer (pH 7.4). The incubations were conducted at 37°C in a shaking water bath for 2 h. High levels of substrate, microsomal proteins, or expressed UGT enzymes and long incubation times were used in the incubations because of a very slow rate of formation of M6 in these systems. At the end of the incubation period, an equal volume of ice-cold acetonitrile was added to quench the reaction. The samples were vortex-mixed and centrifuged at 13,000 rpm for 10 min. The resulting supernatants were mixed with an equal volume of deionized water, and the samples were injected into HPLC for metabolite profiling and mass spectral analysis.
Incubations in the Presence of UGT Enzyme Inhibitors.
BMS-690514, at a concentration of 200 μM, was incubated in triplicate with HLM (1 mg/ml), expressed UGT2B4 (0.5 mg/ml), or expressed UGT2B7 (0.5 mg/ml) with 2 mM UDPGA, 10 mM MgCl2, 50 μg/ml alamethicin, and competitive UGT chemical inhibitors in 0.1 M Tris-HCl buffer (pH 7.5), total reaction volume 0.2 ml. The chemical inhibitors of UGT enzymes evaluated in this study were bilirubin (20 μM) for UGT1A1 (Court, 2005), HDCA (600 μM) for UGT2B4/7 (Barre et al., 2007), fluconazole (4 mM) for UGT2B7 (Uchaipichat et al., 2006b), hecogenin (100 μM) for UGT1A4 (Uchaipichat et al., 2006a), propofol (100 μM) for UGT1A9 (Yamanaka et al., 2005), BMS-204352 (100 μM; MaxiPost) for UGT2B7 (Zhang et al., 2004), and serotonin (20 mM) for UGT1A6 (Krishnaswamy et al., 2003). The inhibitors at the selected concentrations used here were shown to have inhibitory activity toward the individual UGTs according to the literature citations listed above. All ingredients except UDPGA were premixed, and the incubation mixture was preincubated at 37°C for 3 min. After the preincubation, UDPGA was added to initiate the reaction, and the reaction was conducted at 37°C for 40 min. The sample transferring, dilution, and incubation were done on a Tecan Genesis RSP 200 liquid handling workstation. After incubation, an equal volume of ice-cold acetonitrile containing 2 μM [13C]BMS-690514 (IS) was added to stop the reaction. The proteins in the sample mixtures were removed by passing the samples through a 96-well filter plate by centrifugation. Then, the filtered sample was injected into LC-MS/MS for analysis. Solvent control incubations without any inhibitors were also included. The relative amount of M6 formed in an incubation was calculated by measuring the peak area ratio of the metabolite to the IS and normalizing the peak area ratios in each incubation to the solvent control. The values were then used to compare the relative amount of metabolite formation between different incubations.
LC-MS/MS Metabolite Quantification and Radiochromatographic Analysis.
An LC-radiochromatographic-mass spectrometry analysis method was used to obtain the metabolite radioprofiles for incubation of [14C]BMS-690514 with HLM and cDNA-expressed P450 enzymes fortified with NADPH as well as with HLM and expressed UGT enzymes fortified with UDPGA. The LC-radiochromatographic-mass spectrometry system consisted of a Surveyor HPLC pump, autosampler, and photodiode array detector and an LCQ Deca-XP Plus ion trap spectrometer (Thermo Fisher Scientific, Waltham, MA). Sample components were separated on a YMC ODS-AQ HPLC column (4.6 × 150 mm, 120 Å; Waters). The mobile phase consisted of two solvents: mobile phase A (10 mM ammonium acetate in water and acetonitrile, 95:5, v/v, pH 5.0) and mobile phase B (100% acetonitrile). The linear gradient program used for HPLC separation was as follows: linear gradient from 0 to 23% B (0–15 min); linear gradient from 23 to 32% B (15–25 min); linear gradient from 32 to 90% B (25–30 min); hold isocratic at 90% B (30–35 min); and re-equilibrate at 0% B for 10 min. The mobile phase flow rate was 1.0 ml/min, and the column eluate was monitored at a wavelength of 307 nm. An accurate postcolumn splitter (LC Packings, San Francisco, CA) was installed after the photodiode array detector, diverting 25% of the eluate to a LCQ Deca XP-Plus ion trap mass spectrometer. The mass spectrometer was equipped with an electrospray ionization source, operated in positive ion mode with a capillary temperature of 350°C and spray voltage of 5 kV. The nitrogen gas flow rate, capillary voltage, and the tube lens voltages were adjusted to give maximum sensitivity for BMS-690514. Product ions were generated via collision-induced dissociation with helium using normalized collision energy of 30% and a precursor ion isolation width of m/z 2.0. The remaining 75% of the eluate was collected in 96-well Wallac ScintiPlates (PerkinElmer Life and Analytical Sciences, Waltham, MA) at a collection rate of 0.13 min/well with a model FC 204 fraction collector (Gilson, Inc., Middleton, WI). After fraction collection, 200 μl of liquid scintillation cocktail (Ecolite; MP Biomedicals, Solon, OH) was added into each well, and the plates were counted for radioactivity (20 min/well) in a Wallac 1450 MicroBeta TriLux Liquid Scintillation and Luminescence Counter (PerkinElmer Life and Analytical Sciences).
An LC-MS/MS quantification method was used to quantify metabolites in samples from kinetic studies and HLM incubations with the presence of P450 inhibitors or UGT inhibitors. The LC-MS/MS system consisted of a Shimadzu LC-10AD high-performance liquid chromatograph system (Shimadzu Scientific Instruments, Columbia, MD), a LEAP Technologies autoinjector (CTC Analytics, Zwingen, Switzerland), and a Sciex Q-Trap 4000 mass spectrometer (Applied Biosystems, Streetsville, ON, Canada). Sample components were separated on a YMC ODS-AQ S-3 HPLC column (2.0 × 50 mm, 120 Å, Waters). The mobile phase consisted of two solvents: mobile phase A (10 mM ammonium acetate, pH 5.0) and mobile phase B (100% acetonitrile). The linear gradient program used for HPLC separation was as follows: hold isocratic at 0% B (0–1 min); linear gradient from 0 to 14% B (1–5 min); linear gradient from 14 to 28% B (5–15 min); linear gradient from 28 to 90% B (15–18 min); and re-equilibrate at 0% B for 10 min. The mobile phase flow rate was 0.3 ml/min. The HPLC eluate was introduced into a Q-Trap 4000 mass spectrometer (Applied Biosystems, Foster City, CA), which was equipped with an electrospray ionization source operating in positive ion mode. Ultra-high-purity nitrogen gas was used as the curtain gas, ion source gas, and collision gas. The capillary temperature was 450°C, the spray voltage was 4.5 kV, and the collision energy was 30 eV. The gas flow, declustering potential voltage and other voltages were adjusted to provide maximum sensitivity for the parent compound.
For semiquantification of oxidative metabolites (M1, M2/M37, and M3) and M6 formed in HLM incubations in the presence of P450 or UGT inhibitors, [13C6]BMS-690514 was added to each sample as internal standard and the following MRM transitions were selected on the basis of the fragmentation of [M + H]+ ions observed in previous MS/MS analyses: 385.2 > 253.3 for M1, 385.2 > 269.3 for M2/M37, 355.2 > 239.3 for M3, 545.2 > 253.3 for M6, and 375.2 > 259.3 for [13C6]BMS-690514. For absolute quantification of M1, M2/M37, and M3 in the kinetic study, 13C-labeled oxidative metabolite mixture prepared as described previously was added to each sample as internal standards. The MRM transitions were 385.2 > 253.3 for M1, 391.2 > 259.3 for [13C6]M1, 385.2 > 269.3 for M2/M37, 391.2 > 275.3 for [13C6]M2/[13C6]M37, 355.2 > 239.3 for M3, and 361.2 > 245.3 for [13C6]M3.
Assessment of Potential of BMS-690514 to Inhibit P450 Enzymes.
The potential for BMS-690514 to inhibit P450 enzymes was assessed according the methods described by Yao et al. (2007). To assess the direct inhibition of BMS-690514 toward P450 enzymes, BMS-690514 (0.0045, 0.018, 0.09, 0.36 1.8, 9, and 45 μM), or prototypical P450 inhibitors (positive controls) were mixed with HLM, probe substrates at concentrations near their respective Km values, and NADPH in 100 mM phosphate buffer (pH 7.4) without preincubation. The total reaction volume was 0.2 ml. Positive control inhibitors included α-naphthoflavone (CYP1A2), tranylcypromine (CYP2A6), orphenadrine (CYP2B8), montelukast (CYP2C8), sulfaphenazole (CYP2C9), (+)-N-3-benzylnirvanol (CYP2C19), quinidine (CYP2D6), and ketoconazole (CYP3A4). The effects of BMS-690514 and positive controls on the rate of phenacetin (45 μM) O-deethylation (CYP1A2), coumarin (0.65 μM) hydroxylation (CYP2A6), bupropion (100 μM) hydroxylation (CYP2B6), dextromethorphan (10 μM) O-demethylation (CYP2D6), paclitaxel (5 μM) hydroxylation (CYP2C8), diclofenac (10 μM) hydroxylation (CYP2C9), S-mephenytoin (55 μM) hydroxylation (CYP2C19), midazolam (5 μM) hydroxylation (CYP3A4), and testosterone (75 μM) hydroxylation (CYP3A4) were evaluated. The LC-MS/MS conditions used for quantification of metabolites of the probe substrates of these P450 enzymes are shown in Supplemental Table S1. When inhibition of P450 enzymes was observed, IC50 values were determined using the transformed Michaelis-Menten equation for competitive inhibition with GraFit (version 5.0; Erithacus Software Limited, London, UK).
To assess the time-dependent inhibition, BMS-690514 (at concentrations of 0, 0.018, 0.09, 0.36, 1.8, 9, and 45 μM) or positive control time-dependent inhibitors were preincubated for 30 min with pooled human liver microsomes in the presence and absence of NADPH. After preincubation, P450-specific probe substrates were added to the incubation mixtures to assess potential time-dependent inhibition. The effects of standard inhibitors and BMS-690514 on the rate of dextromethorphan (5 μM) O-demethylation (CYP2D6), diclofenac (5 μM) hydroxylation (CYP2C9), S-mephenytoin (50 μM) hydroxylation (CYP2C19), midazolam (2 μM) hydroxylation (CYP3A4), and testosterone (50 μM) hydroxylation (CYP3A4) were evaluated. The positive control time-dependent inhibitors included tienilic acid (3 μM) for CYP2C9, ticlopidine (2 μM) for CYP2C19, 3,4-methylenedioxymethamphetamine (10 μM) for CYP2D6, and mifepristone (10 μM) for CYP3A4. When inhibition reached significant levels (>50%), IC50 values were reported. The LC-MS/MS conditions used for quantification of metabolites of the probe substrates of CYP2D6, CYP2C9, CYP2C19, and CYP3A4 are listed in Supplemental Table S2.
Assessment of Potential of BMS-690514 to Induce P450 Enzymes.
The potential of BMS-690514 to induce the expression of mRNA levels and/or P450 enzyme activity of CYP1A1, CYP1A2, CYP2B6, CYP3A4, and UGT1A1 was investigated in primary cultures of freshly isolated human hepatocytes. Human hepatocytes from three tissue donors (lots Hu 680, 681, and 690; CellzDirect, Durham, NC; see Supplemental Table S3 for donor information) were isolated by a collagenase perfusion method as described previously (Wang and LeCluyse, 2003). The hepatocytes from each individual donor were seeded separately on 60-mm Permanox culture dishes at the density of ∼4 × 106 cells/dish in supplemented Dulbecco's modified Eagle's medium. The culture dishes were incubated in a humidified incubator at 37°C with 95%:5% air-CO2. After cell attachment, the Dulbecco's modified Eagle's medium and unattached cells were aspirated, and fresh serum-free supplemented Williams' Eagle's medium were added to the culture dishes. The culture medium was changed on a daily bases thereafter. After an adaptation period of approximately 2 days, the cultured human hepatocytes from three separate human livers were treated once daily for 3 consecutive days with BMS-690514 (0.2, 2, 10, and 30 μM), solvent control (0.1% DMSO) or known prototypical inducers, including 3-methylcholanthrene [(3-MC) 2 μM, positive control inducer for CYP1A2], phenobarbital (1000 μM, positive control inducer for CYP2B6), and rifampicin (10 μM, positive control inducer for CYP3A4). The cell morphology and attachment for each treatment group was examined on a daily basis by phase contrast microscopy, as a preliminary evaluation of cytotoxicity. At 72 h after treatment with the test compounds, cells were harvested. The lactate dehydrogenase (LDH) activity in the culture medium from each treatment group was determined as a more definitive measurement of cytotoxicity. The LDH activity was determined using an LDH assay kit (Promega, Madison, WI), following instructions provided by the manufacturer. Microsomes and cell lysates were prepared from the remaining cells of each culture, on the basis of methods described previously (Wortelboer et al., 1990).
The enzyme activity in microsomal samples was determined by incubating microsomal samples with probe substrates, and the metabolite formation was quantified by LC-MS/MS analysis. The probe substrate concentration and quantity of microsomal protein in each assay were as follows: 100 μM phenacetin and 0.04 mg/ml protein for CYP1A2, 250 μM bupropion and 0.01 mg/ml protein for CYP2B6, and 200 μM testosterone and 0.01 mg/ml protein for CYP3A4. The LC-MS/MS conditions used for quantification of metabolites of the probe substrates of CYP1A2, CYP2B6, and CYP3A4 are listed in Supplemental Table S4. The relative fold induction in enzymatic activity was calculated by comparing the rate of metabolite formation for treatment groups with that of the negative control group (0.1% DMSO).
The total RNA was isolated from cell lysates using an ABI 6100 PrepStation (Applied Biosystems). For reverse transcription, approximately 200 ng of pooled total RNA was converted to cDNA following the manufacturer's procedure for the High-Capacity cDNA Archive Kit (Applied Biosystems). CYP1A1, CYP1A2, CYP2B6, CYP3A4, and UGT1A1 cDNA from human hepatocytes culture was analyzed from each reverse transcription reaction using gene-specific TaqMan primer/probe sets (Applied Biosystems). Amplifications were performed on an ABI 7900 HT Real-Time PCR system in relative quantification mode for 40 amplification cycles using standard conditions for TaqMan-based assays. Threshold cycle (CT) determinations were performed by the ABI 7900HT system software for both P450 and endogenous control genes. Relative fold mRNA content was determined for each treatment group relative to the endogenous control gene expression and 0.1% DMSO vehicle control for each sample.
Results
Incubations of [14C]BMS-690514 with HLM and cDNA-Expressed P450 Enzymes Fortified with NADPH.
The biotransformation of [14C]BMS-690514 (at a concentration of 10 μM) was investigated with human liver microsomes and with a panel of human cDNA-expressed P450 enzymes. In the incubation with HLM (1 mg/ml) at 37°C for 30 min, M1, a hydroxylated metabolite, was the most abundant metabolite, accounting for 17.8% of the sample radioactivity (Fig. 2A). M2, M37, and M3 were minor oxidative metabolites in the HLM incubation. Under the chromatographic conditions used in this study, M2 and M37 (hydroxylated metabolites) coeluted (Hong et al., 2011), and together they accounted for 0.4% of the sample radioactivity. Metabolite M3 (an O-demethylated metabolite) accounted for 1.9% of the radioactivity.
Incubations with human cDNA-expressed P450 enzymes (100 pmol/ml) at 37°C for 30 min indicated that several enzymes were capable of metabolizing BMS-690514. These included CYP1A1, CYP2C9, CYP2C19, CYP2D6, and CYP3A4 (Fig. 2, B–F). CYP2D6 and CYP3A4 metabolized BMS-690514 to form M1, M2/M37, and a dihydroxylated metabolite. The molecular mass of the dihydroxylated metabolite was 32 Da higher than the parent, and it might be a further oxidative product of M1, M2, or M37. CYP2C9, CYP2C19, CYP2D6, and CYP3A4 metabolized BMS-690514 to form metabolite M3. CYP1A1 metabolized BMS-690514 to form M2/M37 and M3. Little or no metabolism was observed for other P450 enzymes, namely CYP1A2, CYP1B1, CYP2A6, CYP2B6, CYP2C8, CYP2E1, and CYP3A5.
HLM Incubations in the Presence of P450 Inhibitors.
The effect of chemical inhibitors of P450 enzymes on the formation of M1, M2/M37, and M3 was evaluated in pooled HLM at two concentrations of BMS-690514 (1 and 10 μM). Because M2 and M37 coeluted under the same chromatographic peak and the two metabolites have identical mass spectrometric fragments, these two metabolites were quantified together. Because no synthetic standards were available for any of these oxidative metabolites, metabolite formation was quantified on the basis of the MRM responses of the metabolites relative to the IS ([13C6]BMS-690514) in each of the incubations. The rate of metabolite formation in the samples was normalized to that of the corresponding HLM solvent control and reported as a percentage metabolite formation relative to the control. The relative percentage metabolite formation in the presence of chemical inhibitors is illustrated in Fig. 3 for the 1 μM substrate concentration and in Supplemental Fig. S3 for the 10 μM substrate concentration.
In the presence of ABT (a general P450 inhibitor), the formation of M1, M2/M37, and M3 at a substrate concentration of 1 μM was completely inhibited. When BMS-690514 (1 μM) was incubated with HLM in the presence of ketoconazole (1 μM, an inhibitor of CYP3A4/3A5), the formation of M1, M2/M37, and M3 was inhibited by 62.3, 51.9, and 19.7%, respectively, whereas in the presence of quinidine (1 μM, an inhibitor of CYP2D6), the formation of M1, M2/M37, and M3 was inhibited by 46.0, 42.2, and 12.0%, respectively. Sulfaphenazole (10 μM, a CYP2C9 inhibitor) inhibited the formation of M3 by 55.3%. At a BMS-690514 concentration of 1 μM, troleandomycin (20 μM), a time-dependent inhibitor of CYP3A4/5 inhibited M1 and M2/M37 formation by 32.5 and 15.9% but did not appear to inhibit M3 formation. At the higher substrate concentration (10 μM), troleandomycin showed greater inhibition of M1 and M2/M37 formation, 58.4 and 41.3%, respectively, and inhibited M3 formation by 10.1% (Supplemental Fig. S3). However, the level of inhibition still did not match the levels observed with ketoconazole. The reason that troleandomycin was less effective than ketoconazole in inhibiting the formation of the metabolites is not known. Perhaps the troleandomycin stock solution used for these studies was not stable over the storage period and had started to degrade. In addition, a large variability in the metabolite formation for the control incubation for troleandomycin (time-dependent DMSO control) (Fig. 3) relative to the corresponding control incubation for ketoconazole (methanol control) (Fig. 3) may affect the accuracy of the calculations for the troleandomycin incubation. There was no appreciable inhibition of the formation of the oxidative metabolites by inhibitors of other P450 enzymes, except for orphenadrine (a CYP2B6 inhibitor). Although orphenadrine (50 μM) inhibited the formation of M1 and M2/M37 in the HLM incubations by 53.4 and 44.2%, respectively, CYP2B6 did not metabolize BMS-690514 in the expressed enzyme incubations, suggesting that CYP2B6 was not involved in the formation of M1, M2, or M37. Similar results were obtained for incubations conducted with 10 μM BMS-690514 (Supplemental Fig. S3).
Substrate Concentration-Dependent Metabolite Formation.
Studies with HLM and chemical inhibitors indicated that CYP2D6, CYP3A4, and CYP2C9 were the major P450 enzymes responsible for the oxidative metabolism of BMS-690514. Therefore, kinetic studies were conducted to determine the Km and Vmax values for M1, M2/M37, and M3 formation in HLM-expressed CYP3A4, CYP2D6, and CYP2C9 incubations. At substrate concentrations of 1, 2, 5, 10, 20, 40, 65, 125, 200, and 600 μM, the concentration-dependent formation of M1, M2/M37, and M3 in HLM CYP3A4, CYP2D6, and CYP2C9 was well fit to the Michaelis-Menten kinetic equation: V = Vmax · S/(Km + S), with R2 > 0.98 (Fig. 4; Supplemental Fig. S4). In HLM (Fig. 4), the Km values for the formation M1, M2/M37, and M3 were 27.2, 82.0, and 125.7 μM, respectively, and the Vmax values were 321.3, 19.9, and 96.7 pmol of M3/mg protein/min. The intrinsic clearance (CLint and Vmax/Km ratios) in HLM suggested that efficiency for metabolite formation followed the rank order M1 > M3 > M2/M37.
In cDNA-expressed P450 enzymes, the catalytic efficiency (Vmax/Km) for the formation of M1 in CYP3A4 incubations was approximately 3-fold greater than for CYP2D6 (Supplemental Fig. S4), whereas the catalytic efficiency (Vmax/Km) for the formation of M2/M37 in CYP2D6 incubations (0.13 μl · min−1 · mg−1) was close to the value for CYP3A4 (0.11 μl · min−1 · mg−1). The Vmax/Km ratio for the formation of M3 in expressed CYP2C9 incubations (0.68 μl · min−1 · mg−1) was similar to the value in HLM (0.77 μl · min−1 · mg−1) and was approximately 3- to 4-fold higher than CYP2D6 (0.16 μl · min−1 · mg−1) or CYP3A4 (0.20 μl · min−1 · mg−1), suggesting that CYP2C9 was the major P450 enzyme involved in the formation M3.
Incubations with HLM and Expressed UGT Enzymes Fortified with UDPGA.
[14C]BMS-690514, at a concentration of 1 mM, was incubated at 37°C for 2 h with a panel of expressed human UGT enzymes (1.25 mg/ml) fortified with UDPGA. BMS-690514 was metabolized to M6 in the incubations with UGT2B4 and UGT2B7 (Table 1). The levels of M6 in the incubation with UGT2B7 were relatively low and could only be detected by LC-MS/MS analysis and not by radioactivity. There was no detectable formation of M6 in incubations with other expressed UGT enzymes, namely UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8, UGT1A9, UGT1A10, UGT2B15, or UGT2B17.
[14C]BMS-690514, at a concentration of 1 mM, was incubated in the presence of UDPGA, with pooled HLM and HLM obtained from an individual donor that expressed the UGT1A1 (*28*28) allelic variant. On the basis of information provided by the manufacturer, the pooled lot of HLM had approximately 3.3-fold higher UGT1A1 activity than the lot of HLM from the individual with a UGT1A1 (*28*28) allelic variant. In both incubations, M6, a glucuronide conjugate of BMS-690514, was the only metabolite formed. Similar levels of metabolite M6, 15.6 and 19.6% of sample radioactivity, respectively, were formed in pooled HLM and HLM from the UGT1A1-deficient donor (Table 1), suggesting that the UGT1A1 enzyme was not involved in the formation of M6.
Incubations in the Presence of UGT Enzyme Inhibitors.
The effect of chemical inhibitors of UGT enzymes on the formation of M6 was evaluated in triplicate in pooled HLM and expressed UGT2B4 and UGT2B7. The relative amount of M6 formed in each incubation was determined by LC-MS/MS analysis. The rate of metabolite formation was normalized to the solvent control and reported as the percentage M6 formation relative to the control sample.
The relative percentage M6 formation in the presence of chemical substrates/inhibitors is illustrated in Fig. 5. In the presence of known UGT2B4 and/or UGT2B7 substrates, HDCA (a UGT2B4 and UGT2B7 substrate) (Barre et al., 2007), fluconazole (a UGT2B7 substrate) (Uchaipichat et al., 2006b), and MaxiPost (a UGT2B7 substrate) (Zhang et al., 2004), the formation of M6 was substantially inhibited (>57%) in incubations with HLM and expressed UGT2B4 and UGT2B7. The inhibitors of other UGT enzymes, namely bilirubin (a UGT1A1 substrate) (Court, 2005), hecogenin (a UGT1A4 substrate) (Uchaipichat et al., 2006a), and serotonin (a UGT1A6 substrate) (Krishnaswamy et al., 2003), did not inhibit the formation of M6 to the same level as observed in incubations with UGT2B4/2B7 inhibitors. Propofol, which is known to be a UGT1A9 substrate but is also reported to be a UGT2B7 inhibitor (Yamanaka et al., 2005), inhibited the formation of M6 in HLM, UGT2B4, and UGT2B7 to an extent similar to that of the UGT2B4/UGT2B7 substrates (>74%). These inhibition results confirmed that UGT2B4 and UGT2B7 were two major UGT enzymes involved in the formation of M6.
Assessment of Potential of BMS-690514 to Inhibit P450 Enzymes.
For all P450 enzymes investigated, the IC50 values for direct inhibition of BMS-690514 (without preincubation, T0) were ≥40 μM (Supplemental Table S5). Because CYP2D6, CYP3A4, and CYP2C9 were the major enzymes involved in the oxidative metabolism of BMS-690514 and CYP2C19 was also capable of metabolizing BMS-690514 to form M3, BMS-690514 was evaluated as a potential time-dependent inhibitor of these enzymes. The experimental results showed that with a 30-min preincubation (T30; Supplemental Table S5), the IC50 values for BMS-690514 were greater than the highest concentration evaluated (>45 μM) for these four P450 enzymes.
Assessment of Potential of BMS-690514 to Induce P450 Enzymes.
The enzyme activities of CYP1A2, CYP2B6, and CYP3A4 and the mRNA levels of CYP1A1, CYP1A2, CYP2B6, CYP3A4, and UGT1A1 were determined in primary cultured human hepatocytes obtained from three individual donors after treatment with BMS-690514 (0.2, 2, 10, and 30 μM) or prototypical inducers, 3-MC (positive control inducer for CYP1A2), phenobarbital (positive control inducer for CYP2B6), and rifampicin (positive control inducer for CYP3A4). At the concentrations tested, BMS-690514 was not cytotoxic, as indicated by the morphological integrity assay and LDH activity assays (data not shown). Treatment with BMS-690514 resulted in dose-dependent increases in CYP1A2 enzyme activity and mRNA levels (Fig. 6). At the two highest concentrations of BMS-690514 (10 and 30 μM), the CYP1A2 enzyme activities were 44.6 and 76.9% of the positive control (3-MC), respectively, reaching significant levels (defined as >40% of the positive control) (Bjornsson et al., 2003; Sinz et al., 2008). The mRNA levels of CYP1A1 also showed dose-dependent increases with BMS-690514 treatment (Fig. 6B). The mRNA content for both CYP2B6 and CYP3A4 generally increased with increasing BMS-690514 concentrations; however, the activity of these two enzymes did not reach significant levels (40% of the positive control) at any of the concentrations tested, and there was no apparent trend for UGT1A1 mRNA levels.
Discussion
The primary metabolic pathways of BMS-690514 in humans included hydroxylation to metabolites M1, M2, and M37, O-demethylation to M3, and glucuronidation to M6 (Fig. 1). In the presence of ABT, a general P450 inhibitor, the formation of M1, M2/M37, and M3 in HLM was completely inhibited, indicating that the formation of these oxidative metabolites was P450-mediated. Initial incubations with cDNA-expressed P450 enzymes and a subsequent study with HLM and selective P450 inhibitors indicated that CYP2D6 and CYP3A4 catalyzed the formation of all oxidative metabolites (M1, M2/M37, and M3), whereas CYP2C9 was also involved in the demethylation of BMS-690514 to form M3. In addition, sulfaphenazole, a CYP2C9 inhibitor, inhibited M3 formation to a higher extent than ketoconazole or quinidine, suggesting that CYP2C9 was the major enzyme responsible for the formation of M3, with CYP3A4 and CYP2D6 playing a minor role. The CLint values (Vmax/Km) for the formation of M3 by CYP3A4, CYP2D6, and CYP2C9 consistently were 0.20, 0.16, and 0.68 μl · min−1 · mg−1, respectively (Supplemental Fig. S4). Although CYP1A1 was capable of metabolizing BMS-690514 to form M2/M37 and M3 in the expressed enzyme incubations, given the low abundance of this enzyme in the human liver microsomes (Rodrigues, 1999), it is not expected to play a significant role in the hepatic metabolism of BMS-690514. Although BMS-690514 was metabolized to form M3 in incubations with cDNA-expressed CYP2C19, its formation in HLM was not inhibited by benzylnirvanol, suggesting that CYP2C19 did not play a significant role in the O-demethylation of BMS-690514. Whereas orphenadrine, a CYP2B6 inhibitor, inhibited the formation of M1 and M2/M37 in HLM, cDNA-expressed CYP2B6 did not metabolize BMS-690514, suggesting that CYP2B6 was not involved in the formation of these metabolites. It has been reported that orphenadrine is not a very specific inhibitor of CYP2B6 and may also inhibit CYP2D6 activity (Sai et al., 2000), which further supports this conclusion.
Compared with P450 enzymes, there are fewer reported clinical DDIs with glucuronidation reactions. This may be due to the nature of the UGT enzymatic reactions (e.g., relative high Km values for many substrates compared with P450 enzymes, availability of multiple UGT enzymes catalyzing the same reaction, and relative high Ki values for inhibitors that are often above the in vivo concentrations) (Kiang et al., 2005; Remmel et al., 2007). Strategies for reaction phenotyping of UGT enzymes are not well established (Hurst et al., 2007). In this investigation, we used a combined approach in which recombinant UGT enzymes and HLM with inhibitors were used to identify the enzymes responsible for the formation of M6. Because of the lack of information about specific inhibitors of UGT enzymes, drugs that have been identified as specific substrates of one or two UGT enzymes were used as competitive inhibitors in the experiments with HLM. The initial screening with expressed UGT enzymes indicated that UGT2B4 and UGT2B7 were capable of metabolizing BMS-690514 to M6. Substrates/inhibitors of UGT2B4 and/or UGT2B7 (hyodeoxycholic acid, fluconazole, and MaxiPost) significantly inhibited the formation of M6 in HLM-expressed UGT2B4 and UGT2B7 incubations. UGT2B4 and UGT2B7 are both expressed in human liver and share a broad range of common substrates (Saeki et al., 2004; Barre et al., 2007). This observation explains the reason that UGT2B7 substrates, namely fluconazole and MaxiPost, not only inhibited the formation of M6 in incubations with expressed UGT2B7 but also inhibited it in incubations with expressed UGT2B4. The Km values for the formation of the M6 in HLM or expressed UGT2B4/7 could not be determined because the projected Km values (>1 mM) were higher than the compound solubility (data not shown). Considering the relatively low Km values for the formation of oxidative metabolites by P450 enzymes in HLM (27–126 μM) (Fig. 4), the potential for drug-drug interactions of BMS-690514 with coadministered UGT2B4/7 inhibitors should be minimal, because multiple P450 enzymes (CYP3A4, CYP2D6, and CYP2C9) might serve as salvage pathways for the metabolic clearance of BMS-690514.
Results from the current experiments with P450 and UGT chemical inhibitors in HLM were used in conjunction with excretion data from the BMS-690514 human ADME study to estimate the flux through each of the enzymatic pathways for BMS-690514 metabolism. On the basis of the observed data in the human ADME study, the flux through the M3, M6, and combined M2 and M37 pathways was estimated to be approximately 20 to 25, 16, and 12 to 16% of the administered dose, respectively (Christopher et al., 2010). Only 0.3% of the dose was recovered as M1 in human excreta (urine, feces, and bile) (Christopher et al., 2010). Because M1 is known to be thermally labile, it may not have survived either transit through the gastrointestinal tract or sample processing. Therefore, the 0.3% is likely to underestimate of the flux through the M1 pathway (Christopher et al., 2010). In contrast to the in vivo data, M1 was the predominant metabolite detected in HLM (Fig. 2A) (Christopher et al., 2010; Hong et al., 2011), and the intrinsic clearance (CLint) data in HLM also suggested that efficiency for metabolite formation followed the order M1 > M3 > M2/M37 (Fig. 4). If all of the unassigned radioactivity in the excreta were assigned to M1, the maximum possible flux through this pathway would be approximately 20%. When these data were integrated with the reaction phenotyping results, the flux through the CYP2C9, CYP2D6, CYP3A4, and UGT2B4/2B7 enzymatic pathways was estimated to be 11 to 17, 14 to 20, 22 to 33, and 16%, respectively. Several factors affected the ability to provide more definitive estimates on the flux through these enzymatic pathways. These included coelution of several metabolites in the radioprofiles from the human ADME study as well as the presence of secondary and tertiary metabolites, which were formed via multiple oxidative reactions (Christopher et al., 2010). In addition, the flux through the UGT2B4/2B7 enzymatic pathways may also be underestimated, because the recovery of M6 in fecal excreta samples from the human ADME study was low as a result of the hydrolysis of M6 back to parent drug by gut microflora (Christopher et al., 2010).
Nevertheless, the data clearly demonstrate that multiple enzymes are involved in the clearance of BMS-690514, with the flux through each enzymatic pathway representing <33% of the dose. Therefore, if BMS-690514 is administered when an individual P450 or UGT enzyme is inhibited or genetically deficient, an appreciable increase in the systematic exposure of BMS-690514 (i.e., <2-fold increase in area under the curve) is not expected (Ogilvie et al., 2008; Nassar et al., 2009). This conclusion was supported by profiling an individual bile sample from a CYP2D6 poor metabolizer, which suggested that when metabolism through the CYP2D6 pathway was reduced, other metabolic pathways were able to participate in BMS-690514 elimination, resulting in similar exposures of the parent compound (Christopher et al., 2010).
To evaluate its potential as a perpetrator of drug-drug interactions, BMS-690514 was evaluated as a direct-acting and time-dependent inhibitor of P450 enzymes in HLMs and as an inducer of P450 enzymes in cultured human hepatocytes. The results from the inhibition studies showed that the IC50 values for both direct inhibition and time-dependent inhibition of the P450 enzymes investigated here were ≥40 μM. Considering a mean steady-state total plasma Cmax of 0.7 μM after a 200-mg daily oral administration of BMS-690514 in patients with cancer (J. S. Zhang, personal communication), the potential for BMS-690514 to inhibit the metabolism of other coadministered P450 substrates is predicted to be extremely low, because the [I]/IC50 ratio is ≤0.02 (Bjornsson et al., 2003).
Although BMS-690514 caused dose-dependent increases in the CYP2B6 and CYP3A4 mRNA content in primary cultures of human hepatocytes, the activity of these two enzymes did not reach levels considered to be significant (40% of the respective positive controls) (Bjornsson et al., 2003; Sinz et al., 2008) at the concentrations evaluated (0.2–30 μM). However, BMS-690514 caused dose-dependent increases in CYP1A2 activity, which reached levels of significance at concentrations of 10 and 30 μM. This increase in enzyme activity was accompanied by a dose-dependent increase in CYP1A2 mRNA content. The mRNA levels of CYP1A1 similarly increased with BMS-690514 dose. Both CYP1A1 and CYP1A2 are regulated by the aryl hydrocarbon receptor and would be expected to be induced concurrently (Zhang et al., 2007). However, because the mean steady-state total plasma Cmax of BMS-690514 is in the submicromolar range, the potential of BMS-690514 to increase clearance of coadministered drugs due to the induction of the evaluated drug-metabolizing enzymes is likely to be minimal.
In summary, BMS-690514 is highly metabolized via multiple enzymatic pathways, which include CYP2C9, CYP2D6, CYP3A4, and UGT2B4/7. Because the clearance through each individual pathway represents <33% of the dose, inhibition of any individual pathway either through coadministration with drugs that are P450 or UGT inhibitors or via enzyme polymorphisms is not expected to cause pronounced (>2-fold) increases in BMS-690514 exposure. In addition, BMS-690514 is unlikely to act as a perpetrator of drug-drug interactions either through the inhibition or induction of drug-metabolizing enzymes. In vivo drug-drug interactions between BMS-690514 and other coadministered medications will be investigated in a clinical DDI study with ketoconazole and in phase II studies.
Authorship Contributions
Participated in research design: Hong, Iyer, Humphreys, and Christopher.
Conducted experiments: Hong, Su, Ma, and Yao.
Performed data analysis: Hong, Iyer, and Christopher.
Wrote or contributed to the writing of the manuscript: Hong, Iyer, Humphreys, and Christopher.
Acknowledgments
We thank Karla Johanning and the staff at CellzDirect for conducting the time-dependent inhibition and induction studies.
Footnotes
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
doi:10.1124/dmd.111.039776.
↵ The online version of this article (available at http://dmd.aspetjournals.org) contains supplemental material.
-
ABBREVIATIONS:
- BMS-690514
- (3R,4R)-4-amino-1-((4-((3-methoxyphenyl)amino)pyrrolo[2,1-f][1,2,4]triazin-5-yl)methyl)-3-piperidinol)
- ErbB
- human epidermal growth factor receptors
- VEGFR
- vascular endothelial growth factor receptor
- ADME
- absorption, distribution, metabolism and excretion
- P450
- cytochrome P450
- DDI
- drug-drug interactions
- UGT
- UDP-glucuronosyltransferase
- HLM
- human liver microsome(s)
- BMS-204352
- (3S)-(+)-(5-chloro-2-methoxyphenyl)-1,3-dihydro-3-fluoro-6-(trifluoromethyl)-2H-indol-2-one
- ABT
- 1-aminobenzotriazole
- UDPGA
- UDP-glucuronic acid
- HDCA
- hyodeoxycholic acid
- HPLC
- high-performance liquid chromatography
- DMSO
- dimethyl sulfoxide
- IS
- internal standard
- LC
- liquid chromatography
- MS/MS
- tandem spectrometry
- MRM
- multiple reaction monitoring
- 3-MC
- 3-methylcholanthrine
- LDH
- lactate dehydrogenase.
- Received March 27, 2011.
- Accepted June 14, 2011.
- Copyright © 2011 by The American Society for Pharmacology and Experimental Therapeutics