Abstract
Cytochrome P450 2D6 (CYP2D6) is an enzyme of potential importance for the metabolism of drugs used clinically, and it exhibits genetic polymorphism with interindividual differences in metabolic activity. To date, 21 CYP2D6 allelic variants have been identified in the Japanese population. The aim of this study was to investigate the functional characterization of CYP2D6 variants identified in Japanese subjects. Wild-type CYP2D6 and its variants, namely, CYP2D6.2, CYP2D6.10, CYP2D6.14A, CYP2D6.14B, CYP2D6.18, CYP2D6.27, CYP2D6.36, CYP2D6.39, CYP2D6.47, CYP2D6.48, CYP2D6.49, CYP2D6.50, CYP2D6.51, CYP2D6.53, CYP2D6.54, CYP2D6.55, and CYP2D6.57 were transiently expressed in COS-7 cells, and enzymatic activities of the CYP2D6 variant proteins were characterized using bufuralol and dextromethorphan. Functional characterization of 17 CYP2D6 variants revealed an absence of enzyme activity in four (CYP2D6.14A, CYP2D6.36, CYP2D6.47, and CYP2D6.57), low activity in eight (CYP2D6.10, CYP2D6.14B, CYP2D6.18, CYP2D6.49, CYP2D6.50, CYP2D6.51, CYP2D6.54, and CYP2D6.55), and high activity in one (CYP2D6.53) compared with the wild type. Analysis of CYP2D6 variant proteins can be useful for predicting CYP2D6 phenotypes and could be applied to personalized drug therapy.
Cytochrome P450 2D6 (CYP2D6) is an important drug-metabolizing enzyme involved in the metabolism of many therapeutic agents, including antidepressants, β-adrenergic antagonists, antiarrhythmics, and opioids (Zanger et al., 2004). The genetic polymorphisms of CYP2D6 are defined in 67 allelic variants, many of which are associated with increased, decreased, or abolished enzyme functions (Human CYP Allele Nomenclature Web site at http://www.cypalleles.ki.se/cyp2d6.htm). These polymorphisms result in differences of up to 30 to 40-fold in substrate drug clearance, leading to drug concentrations of the therapeutic range in treated patients. Consequently, such differences in CYP2D6 activity would lead not only to severe adverse effects in clinical therapy (Kirchheiner et al., 2004) but also to nonresponse to medications, e.g., no observable analgesic effects of prodrugs such as codeine in poor metabolizers (PMs) (Gasche et al., 2004).
CYP2D6 activity widely differs among ultrarapid metabolizer, extensive metabolizer, intermediate metabolizer (IM), and PM phenotypes. The PM phenotype is due to two nonfunctional (null) alleles, whereas the extensive metabolizer phenotype is due to one or two alleles with normal function, such as CYP2D6*1 and CYP2D6*2. An IM phenotype is usually observed in populations harboring a combination of one CYP2D6 null allele and another allele with impaired expression and/or function, such as CYP2D6*10 (Zanger et al., 2004). The frequency of PMs is 5 to 10% in the Caucasian population and less than 1% in the Asian population (Sohn et al., 1991; Dahl et al., 1992; Tateishi et al., 1999). Among the variant alleles reported thus far, CYP2D6*3, CYP2D6*4, CYP2D6*5, and CYP2D6*6 account for approximately 97% of the PM alleles in the Caucasian population (Sachse et al., 1997). On the other hand, studies conducted by us and other groups have identified 21 variant alleles of the CYP2D6 gene in the Japanese population (Fig. 1) (Yokoi et al., 1996; Chida et al., 1999; Nishida et al., 2000; Yamazaki et al., 2003; Soyama et al., 2004, 2006; Ebisawa et al., 2005). A number of these variant alleles have also been described in other Asian populations (Table 1). Among them, CYP2D6*4, CYP2D6*5, CYP2D6*14A, CYP2D6*18, CYP2D6*21, CYP2D6*36, and CYP2D6*44 are associated with the PM phenotype (Kagimoto et al., 1990; Gaedigk et al., 1991, 2006; Yokoi et al., 1996; Chida et al., 1999; Wang et al., 1999; Yamazaki et al., 2003). The catalytic properties of CYP2D6*10 and CYP2D6*2 have been studied extensively in vivo and in vitro using recombinant expression systems and reaction phenotyping. The frequency of CYP2D6*10 is relatively high in Asians, and individuals with an IM phenotype and with CYP2D6*10/*10 or CYP2D6*10/*null genotypes, respectively, exhibit lower catalytic activities toward typical CYP2D6 substrates such as dextromethorphan (Tateishi et al., 1999), metoprolol (Huang et al., 1999), and nortriptyline (Yue et al., 1998) than wild-type homozygous individuals. Subjects harboring CYP2D6*2/*2 exhibit metabolic activity similar to that of wild-type individuals for various substrates (Johansson et al., 1993; Dahl et al., 1995). Functional alterations conveyed by sequence variations in CYP2D6*14B, CYP2D6*27, CYP2D6*39, CYP2D6*47, CYP2D6*48, CYP2D6*49, CYP2D6*50, CYP2D6*51, CYP2D6*53, CYP2D6*54, CYP2D6*55, and CYP2D6*57, which have been identified in the Japanese population, remain unknown.
The purpose of this study was to investigate the functional characterization of the CYP2D6 variants identified in Japanese subjects. We generated 17 expression constructs (CYP2D6*2, CYP2D6*10, CYP2D6*14A, CYP2D6*14B, CYP2D6*18, CYP2D6*27, CYP2D6*36, CYP2D6*39, CYP2D6*47, CYP2D6*48, CYP2D6*49, CYP2D6*50, CYP2D6*51, CYP2D6*53, CYP2D6*54, CYP2D6*55, and CYP2D6*57), which were transfected into COS-7 cells. The enzymatic properties of CYP2D6 variant proteins were characterized using the specific substrates bufuralol and dextromethorphan.
Materials and Methods
Construction of Expression Plasmids.CYP2D6 cDNA fragments were amplified from the human liver cDNA library (TaKaRa; Takara, Otsu, Japan) using polymerase chain reaction with the forward primer 5′-CACCATGGGGCTAGAAGCAC-3′ and the reverse primer 5′-CTAGCGGGGCACAGCACAAAG-3′ in the GeneAmp High Fidelity polymerase chain reaction system (Applied Biosystems, Foster City, CA). The amplified fragments were subcloned into the pENTR/D-TOPO vector (Invitrogen, Carlsbad, CA). The underlined sequence was introduced for directional TOPO cloning. Plasmids carrying the CYP2D6*2 cDNA were used as the template to generate 17 CYP2D6 constructs (CYP2D6*1, CYP2D6*10, CYP2D6*14A, CYP2D6*14B, CYP2D6*18, CYP2D6*27, CYP2D6*36, CYP2D6*39, CYP2D6*47, CYP2D6*48, CYP2D6*49, CYP2D6*50, CYP2D6*51, CYP2D6*53, CYP2D6*54, CYP2D6*55, and CYP2D6*57) with a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) according to the manufacturer's instructions. All constructs were sequenced to confirm successful mutagenesis. The wild-type and variant CYP2D6 cDNAs were subsequently subcloned into the pcDNA-DEST40 mammalian expression vector (Invitrogen).
Expression of Variant CYP2D6 Proteins in COS-7 Cells. COS-7 cells were cultured in Dulbecco's modified Eagle's medium (Invitrogen) containing 10% fetal bovine serum under 5% CO2 at 37°C. The cells were plated at 1.8 × 106 cells in 100-mm dishes 24 h before transfection. Subsequently, Opti-MEM medium (Invitrogen) was added to the culture medium, and the cells were transfected with 5 μg of CYP2D6 plasmids by using TransFectin lipid reagent (Bio-Rad, Hercules, CA), according to the manufacturer's instructions. Cells were incubated for 24 h at 37°C and then scraped off the dish after being washed once with cold phosphate-buffered saline. Cells were then resuspended in a homogenization buffer containing 10 mM Tris-HCl (pH 7.4), 0.25 M sucrose, and 1 mM EDTA. Microsomal fractions were prepared by differential centrifugation at 9000g for 20 min followed by centrifugation of the resultant supernatant at 105,000g for 60 min. The microsomal pellet was resuspended in 10 mM Tris-HCl (pH 7.4), 10% glycerol, and 1 mM EDTA and stored at -80°C. The protein concentration was measured using a BCA protein assay kit (Pierce Chemical, Rockford, IL).
Determination of Protein Expression Levels by Immunoblotting. Immunoblotting was performed according to standard procedures with 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and 5 μgofmicrosomal protein per lane. Recombinant CYP2D6 Baculosome reagent (Invitrogen) was coanalyzed on each gel as the standard (range, 0.125–0.5 pmol) and for the quantification of the CYP2D6 apoprotein content. CYP2D6 protein was detected using rabbit anti-human CYP2D6 antibody (Daiichi Pure Chemicals, Tokyo, Japan) and horseradish peroxidase-conjugated goat anti-rabbit IgG (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Immunoblots were developed using SuperSignal West Pico chemiluminescence substrate (Pierce Chemical). Chemiluminescence was quantified using the Lumino Imaging Analyzer FAS-1000 (Toyobo Engineering, Osaka, Japan) and the Gel-Pro Analyzer (Media Cybernetics, Inc., Silver Spring, MD).
Enzymatic Properties of Wild-Type and CYP2D6 Variants.Bufuralol 1′-hydroxylation. The activity of bufuralol 1′-hydroxylation was measured by the method reported by Marcucci et al. (2002) with minor modifications. The incubation mixture contained 1 to 640 μM bufuralol (Sigma-Aldrich, St. Louis, MO), microsomal fraction (25 μg) obtained from COS-7 cells, 1.3 mM NADP, 3.3 mM glucose 6-phosphate, 0.4 U/ml glucose-6-phosphate dehydrogenase, 3.3 mM MgCl2, and 100 mM potassium phosphate buffer (pH 7.4) in a final volume of 75 μl. The reactions were carried out at 37°C for 15 min and terminated by the addition of 75 μl of methanol. The samples were then centrifuged at 10,000g for 3 min to obtain a protein pellet before high-performance liquid chromatography (HPLC) analysis. The HPLC system consisted of a Waters 2695 Separations Module, Waters 2475 multi λ fluorescence detector (Waters, Milford, MA), and a Novapak Phenyl column (3.9 mm × 15 cm; particle size, 4 μm; temperature, 40°C; Waters). The mobile phase was composed of 82% buffer (20 mM potassium phosphate and 20 mM hexane sulfonic acid, pH 4.0) and 18% acetonitrile for the first 12 min, which was subsequently replaced by 60% buffer and 40% acetonitrile from 12 to 17 min, and the mobile phase was finally restored to 82% buffer and 18% acetonitrile from 17 to 27 min. The flow rate was set at 1.0 ml/min with fluorescence detection at excitation and emission wavelengths of 252 and 302 nm, respectively. Under these conditions, the retention times of 1′-hydroxybufuralol and bufuralol were 9 and 17 min, respectively.
Dextromethorphan O-demethylation. The O-demethylation activity of dextromethorphan was measured by the method reported by Marcucci et al. (2002) with minor modifications. The reaction mixture containing 20 μM dextromethorphan (Wako Pure Chemicals, Osaka, Japan) was incubated for 10 min at 37°C. The reactions were terminated by adding 75 μl of methanol containing 1 nM levallorphan tartrate (Sigma-Aldrich), which was used as the internal standard. Other conditions for incubation and deproteinization were identical to those used in the assay of bufuralol 1′-hydroxylation activity. Dextrorphan was detected using the same HPLC system used for the determination of 1′-hydroxybufuralol. The mobile phase consisted of 75% buffer (20 mM potassium phosphate and 20 mM hexane sulfonic acid, pH 4.0) and 25% acetonitrile at a flow rate of 1.2 ml/min with fluorescence detection at excitation and emission wavelengths of 280 and 310 nm, respectively. Under these conditions, the retention times of dextrorphan and dextromethorphan were 4 and 16 min, respectively.
Data Analysis. Kinetic parameters such as the Km and Vmax for bufuralol 1′-hydroxylation were determined using Eadie-Hofstee plots. The intrinsic clearance (CLint) values were determined by the ratio Vmax/Km. All values were expressed as the mean ± S.D. of three independent transfection experiments. Each assay was performed in triplicate. Statistical analyses were performed using the Student's t test, and a value of P < 0.05 was considered statistically significant.
Results
Expression of Wild-Type and Variant CYP2D6 in COS-7 Cells. The CYP2D6 variants were expressed in COS-7 cells, and the protein levels were measured by immunoblot analysis (Fig. 2). All CYP2D6 proteins expressed were recognized with an antibody against CYP2D6. Of the 17 variant proteins, the following 13 had expression levels reduced to 33 to 81% of that of CYP2D6.1: CYP2D6.10, CYP2D6.14A, CYP2D6.27, CYP2D6.36, CYP2D6.39, CYP2D6.47, CYP2D6.49, CYP2D6.50, CYP2D6.51, CYP2D6.53, CYP2D6.54, CYP2D6.55, and CYP2D6.57. The expression level of CYP2D6.1 was 65.7 ± 18.7 pmol/mg microsomal protein.
Enzymatic Properties of Wild-Type and Variant CYP2D6. Bufuralol 1′-hydroxylation activity at substrate concentrations of 80 μM and dextromethorphan O-demethylation activity at substrate concentrations of 20 μM were determined on the basis of microsomal CYP2D6 of the wild-type protein and the 17 variant CYP2D6 proteins (Fig. 3). The bufuralol 1′-hydroxylation by CYP2D6.10, CYP2D6.18, CYP2D6.36, CYP2D6.47, CYP2D6.49, CYP2D6.50, CYP2D6.51, CYP2D6.54, CYP2D6.55, and CYP2D6.57 was reduced to 2 to 36% of that of CYP2D6.1. The dextromethorphan O-demethylation activity of CYP2D6.10, CYP2D6.18, CYP2D6.49, CYP2D6.50, CYP2D6.54, and CYP2D6.55 was reduced to 7 to 36% of that of CYP2D6.1. CYP2D6.14A had no detectable enzyme activity toward bufuralol and dextromethorphan.
Figure 4 shows the Michaelis-Menten curves and kinetic parameters of bufuralol 1′-hydroxylation by the wild-type and the 17 variants. The Vmax and CLint (Vmax/Km) values for bufuralol 1′-hydroxylation were normalized to those of the CYP2D6 protein levels. CYP2D6.1 had an apparent Km value of 8.4 μM (the average value from four experiments). CYP2D6.2, CYP2D6.10, CYP2D6.14B, CYP2D6.18, and CYP2D6.55 had significantly higher apparent Km values that varied from 11.4 to 72.8 μM. In contrast, CYP2D6.49, CYP2D6.50, and CYP2D6.53 had apparent Km values lower than that of CYP2D6.1, ranging from 2.5 to 3.5 μM. The Vmax values of CYP2D6.10, CYP2D6.49, and CYP2D6.54 were 20, 17, and 6% of that of CYP2D6.1, respectively. The CLint values of CYP2D6.10, CYP2D6.14B, CYP2D6.18, CYP2D6.49, CYP2D6.54, and CYP2D6.55 were lower than that of CYP2D6.1 (7–35% of that of CYP2D6.1). On the other hand, CYP2D6.53 had an apparent CLint value 4 times higher than that of CYP2D6.1. CYP2D6.36, CYP2D6.47, CYP2D6.51, and CYP2D6.57 showed extremely low activity, and the kinetic parameters could not be determined. CYP2D6.14A activity was not detected at the substrate concentrations used.
Discussion
CYP2D6 is a clinically important enzyme that metabolizes numerous therapeutic drugs (Zanger et al., 2004). The present study provides comprehensive data regarding functional alterations of CYP2D6 variant proteins in the Japanese population. To assess the effects of 17 CYP2D6 variant alleles on the protein expression levels and enzymatic activity, the wild-type and 17 variant CYP2D6 proteins were transiently expressed in COS-7 cells.
Among the 17 variant CYP2D6 alleles, CYP2D6*14A, CYP2D6*18, and CYP2D6*36 are associated with PM phenotype. There was a subject carrying CYP2D6*5/*14A, which was shown to be the PM phenotype (Wang et al., 1999). Shiraishi et al. (2001) have reported that a combination of G169R and P34S decreased bufuralol 1′-hydroxylation and dextromethorphan O-demethylation activities. In this study, CYP2D6.14A harboring the P34S, G169R, R296C, and S486T substitutions showed no detectable activity. On the other hand, the activity of CYP2D6.14B harboring G169R, R296C, and S486T substitutions was higher than that of CYP2D6.14A. These results suggest that the relatively high activity of CYP2D6.14B may be attributed mainly to the absence of the P34S substitution. Yokoi et al. (1996) have reported the kinetic parameters for CYP2D6.18-mediated bufuralol 1′-hydroxylation activity. The Km of CYP2D6.18 showed a 236-fold increase from the wild-type Km (990 μM versus 4.2 μM), a finding that is in agreement with the results we obtained in the COS-7 expression system. The COS-7 cells that expressed CYP2D6.36 exhibited decreased protein levels and bufuralol 1′-hydroxylation activities compared with those that expressed the wild type; these results were consistent with those of a previous report (Johansson et al., 1994). The CYP2D6*36 alleles have been described to be in a tandem arrangement with CYP2D6*10 as CYP2D6*36–*10 alleles. In recent studies, the duplicate-type *36×2 in the Japanese and the single-type *36 in African Americans and Asians were found in subjects displaying the PM phenotype (Chida et al., 2002; Gaedigk et al., 2006; Soyama et al., 2006). Because CYP2D6.36 shows extremely low activities toward bufuralol and dextromethorphan, the combined activities of CYP2D6.36 and CYP2D6.10 could be comparable with the activity of CYP2D6.10 alone.
Of the 12 functionally unknown variant proteins (Fig. 1), CYP2D6.47, CYP2D6.51, and CYP2D6.57 exhibited a drastic decrease in enzymatic activity, retaining <5% of the wild-type activity, and these variants may cause as the PM phenotype. CYP2D6.47 has R25W, P34S, and S486T substitutions. Arg25 is located within the transmembrane domain and is considered to act as a halt transfer signal, which prevents the translocation of the protein into the lumen of the endoplasmic reticulum (Sakaguchi et al., 1987). CYP2D6.51 has R296C, E334A, and S486T substitutions. Based on the crystal structure of human CYP2D6 (Rowland et al., 2006), Glu334 is located on the J helix and conserved within P450s, further suggesting that this amino acid is critical for maintaining P450 function. The enzymatic activity of CYP2D6.57 (P34S, R62W, and seven amino acid changes in exon 9) was approximately 4% of that of the wild-type, and this decreased activity level was similar to that of CYP2D6.36. Hence, loss of CYP2D6.57 activity is probably attributable to the presence of the P34S substitution and the amino acids introduced by the exon 9 conversion. CYP2D6*57 alleles have also been described to exist in a tandem arrangement with CYP2D6*10 as CYP2D6*57–*10. The combined activities of CYP2D6.57 and CYP2D6.10 could be similar to the CYP2D6.10 activity and to the sum of the CYP2D6.36 and CYP2D6.10 activities. CYP2D6*47, CYP2D6*51, and CYP2D6*57 have recently been detected using single nucleotide polymorphism analysis, but in vivo data have not yet been obtained.
CYP2D6.53 exhibited a 73% decrease in Km values, resulting in a 4-fold increase in its CLint value compared with that of CYP2D6.1. CYP2D6.53 harbored F120I and A122S substitutions located in substrate recognition site 1. Flanagan et al. (2004) recently suggested that the Phe120 residue within the B–C loop of P450s would play an important role in substrate binding and orientation. Furthermore, Keizers et al. (2004) demonstrated that the mutation of Phe120 to alanine results in a 50% decrease in the Km value compared with the Km value of CYP2D6.1. Likewise, our results revealed that CYP2D6.53 activity exhibited a lower Km value than CYP2D6.1, suggesting that the increased affinity of CYP2D6.53 might be attributed to a natural variant of F120I. Because an increase in the CLint of CYP2D6.53 is considered to be the main alteration affecting substrate binding, individuals with CYP2D6*53 might exhibit the UM phenotype.
CYP2D6.14B, CYP2D6.49, CYP2D6.50, CYP2D6.54, and CYP2D6.55 exhibited intermediate activity similar to that of CYP2D6.10 and could be associated with the IM phenotype. Although CYP2D6.27 (E410K), CYP2D6.39 (S486T), and CYP2D6.48 (A90V) exhibited slightly higher CLint values for bufuralol, the differences were not significant. E410K, S486T, and A90V are located between the K′ and K″ helices, B helix, and β-sheet 3, respectively (Rowland et al., 2006), and these amino acids are not conserved among human P450s; therefore, they may not be critical with regard to CYP2D6 activity.
In Asians, the most abundant variant allele is CYP2D6*10. Several studies have reported that CYP2D6.10 exhibited decreased protein levels and an increased Km value with respect to bufuralol (Fukuda et al., 2000; Hanioka et al., 2006; Shen et al., 2007). CYP2D6.10 harbored the P34S and S486T substitutions; P34S is part of a prolinerich region that is highly conserved among microsomal P450s and may function as a hinge between the hydrophobic membrane anchor and the heme-binding portion of the enzyme (Yamazaki et al., 1993). It has been shown that subjects homozygous for CYP2D6*10/*10 require lower drug dosages of metoprolol and nortriptyline than do CYP2D6*1/*1 subjects to achieve the same therapeutic effect. Thus, the former may be at greater risk for developing dose-dependent adverse effects (Yue et al., 1998; Huang et al., 1999). An important example of cancer therapy is the adjuvant treatment of breast cancer with tamoxifen. Several recent retrospective and prospective studies have demonstrated the significant impact of the CYP2D6 genotype on the plasma concentrations of the active tamoxifen metabolite endoxifen (Goetz et al., 2007). Patients with CYP2D6*10/*10 genotypes have extremely low levels of the endoxifen and thus benefit to a lesser extent from the treatment (Lim et al., 2007).
To characterize and assess the functional effect of variant alleles on CYP2D6 activity, several studies have been performed using various heterologous expression systems, including bacteria, insect cells, and mammalian cells. By using these systems, CYP2D6.10 have been shown to exhibit a decreased catalytic efficiency toward bufuralol than that exhibited by CYP2D6.1, but the kinetic parameters (Km, Vmax, and CLint) reported by the different authors were highly variable (Fukuda et al., 2000; Hanioka et al., 2006; Shen et al., 2007). This discrepancy among the laboratories may be partly due to the use of different expression systems. Therefore, a comparison of the kinetic parameters with previous reports would be difficult. In this study, the catalytic properties of a large number of CYP2D6 allelic variants that were expressed under the same conditions were comprehensively compared with the properties of the wild type. CYP2D6 proteins expressed in the COS-7 cells would undergo both post-translational modification and protein degradation: the mammalian expression system would be an appropriate system for the functional characterization of the P450 enzymes.
In this study, the rates of enzymatic activity alteration between bufuralol and dextromethorphan were similar (Fig. 3). However, several researchers (Bogni et al., 2005; Cai et al., 2006; Shen et al., 2007) have reported that some allelic variants (e.g., CYP2D6*17) are associated with substrate-dependent decreases in catalytic properties. In addition, some CYP2D6 variant proteins may decrease heme incorporation into the CYP2D6 apoprotein, but this possibility could not be assessed in the present study. The levels of CYP2D6 protein expressed in COS-7 cells were too low to be determined by the difference in the carbon monoxide-reduced spectra. Furthermore, in vivo and in vitro studies using other substrates as well as the measurement of the holoprotein level are required to confirm these results.
Most of the 17 variants that were expressed in the COS-7 cells showed a significant decrease in catalytic activity. In routine genotyping, it is important to determine all of these alleles, even if the allele frequencies are low. A CYP2D6 genotyping assay such as the DNA microarray can easily test hundreds of single nucleotide polymorphisms in a single run. The U.S. Food and Drug Administration recently approved the first pharmacogenetic test (the AmpliChip CYP450) that uses a DNA microarray (Roche Molecular Diagnostics, Alameda, CA) to determine the genotypes of CYP2D6 and CYP2C19.
In conclusion, we expressed 17 CYP2D6 variants that have previously been found in the Japanese population. Of these, 13 variants functionally affected CYP2D6 activity in vitro. These data will help to understand the genotype-phenotype relationships of CYP2D6 and provide a foundation for future clinical studies regarding individual variations in drug efficacy and toxicity in Asians.
Footnotes
-
This work was supported by a grant-in-aid from the Japan Research Foundation for Clinical Pharmacology and KAKENHI (20590154).
-
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
-
doi:10.1124/dmd.108.023242.
-
ABBREVIATIONS: PM, poor metabolizer; IM, intermediate metabolizer; HPLC, high-performance liquid chromatography; P450, cytochrome P450.
- Received July 3, 2008.
- Accepted September 8, 2008.
- The American Society for Pharmacology and Experimental Therapeutics