PHARMACOGENETICS and GENOMICS Effect of the Gene Dosage of CYP2C19 on Diazepam Metabolism in Chinese Subjects

______www.paper.edu.cn

PHARMACOGENETICS AND GENOMICS Effect of the gene dosage of CYP2C19 on diazepam metabolism in Chinese subjects

Objective: To determine whether the gene dosage of CYP2C19 affects the metabolism of diazepam and desmethyldiazepam in healthy Chinese subjects. Subjects and methods: Eighteen unrelated adult men were recruited for the study from a total of 101 healthy Chinese volunteers who had been screened for CYP2C19 phenotype and genotype. All subjects received a single oral dose (5 mg) of diazepam, and the pharmacokinetics of diazepam and desmethyldiazepam were compared in six m1 homozygotes (m1/m1), six m1 heterozygotes (wt/m1), and six wild-type homozygotes (wt/wt). Results: The plasma elimination half-life values of diazepam (84.0 ± 13.7 hours) and desmethyldiazepam (176.0 ± 28.9 hours) in subjects of m1/m1 were significantly longer than those (62.9 ± 9.8 hours for diazepam; 132.1 ± 24.9 hours for desmethyldiazepam; both P < .01) in subjects of wt/m1 or those (20.0 ± 10.8 hours for diazepam; 99.2 ± 21.7 hours for desmethyldiazepam; both P < .01) in subjects of wt/wt. A significant difference in the corresponding half-life values existed between the wt/m1 and wt/wt subjects (P < .01). As expected, the slowest mean clearance of diazepam was observed in the m1/m1 subjects (2.8 ± 0.9 mL/min) and the fastest in the wt/wt subjects (19.5 ± 9.8 mL/min), with the wt/m1 heterozygotes having an intermediate value (7.2 ± 2.6 mL/min). Conclusion: The presence of a single-nucleotide polymorphism (G681A) of the CYP2C19 gene cosegregates with the impaired metabolism of diazepam and desmethyldiazepam among Chinese subjects in a gene-dosage effect manner. (Clin Pharmacol Ther 1999;66:642-6.)

Xu-Ping Qin, MD,a Hong-Guang Xie, MD, PhD,b Wei Wang, MS, Nan He, MS, Song-Lin Huang, BS, Zhen-Hua Xu, MD, PhD,c Dong-Sheng Ou-Yang, MS, Yong-Jin Wang, MD,a and Hong-Hao Zhou, MD Changsha, Hunan, People’s Republic of China From the Pharmacogenetics Research Institute, Basic and Clinical A well-known genetic polymorphism that affects drug Pharmacology Institute, Hunan Medical University. metabolism is characterized by the impaired 4′-hydrox- Supported by grant F39330230 from the Natural Science Foundation ylation of S-mephenytoin.1,2 S-Mephenytoin 4′-hydroxy- of China and grants 92-568 and 99-697 from the China Medical Board of New York. lase has recently been identified as cytochrome P450 Received for publication May 24, 1999; accepted Sept 24, 1999. (CYP) 2C19.3,4 Individuals can be classified as the phe- Reprint requests: Hong-Hao Zhou, MD, Pharmacogenetics Research notype of either extensive or poor metabolizers on the Institute, Basic and Clinical Pharmacology Institute, Hunan Med- basis of their levels of the CYP2C19 activity. A marked ical University, Changsha, Hunan 410078, China. interracial difference in the frequency of the CYP2C19 aCurrent address: Department of Experimental Pharmacology, Changzhi Medical College, Changzhi, Shanxi 046000, China. poor metabolizer phenotype existed between the Asian bCurrent address: Division of Clinical Pharmacology, Departments (13% to 23%) and white (1% to 6%) populations.1,2,5-9 of Medicine and Pharmacology, Vanderbilt University School of Accordingly, such a genetic polymorphism may be of Medicine, Nashville, TN 37232-6602. more clinical concern in the former than in the latter, c Current address: Department of Pharmacology, Mayo Clinic, because pronounced genetically determined differences Rochester, MN 55905. Copyright © 1999 by Mosby, Inc. were observed in the metabolism and clinical conse- 0009-9236/99/$8.00 + 0 13/1/103379 quences of the CYP2C19-catalyzed substrates, such as

642 中国科技论文在线______www.paper.edu.cn CLINICAL PHARMACOLOGY & THERAPEUTICS VOLUME 66, NUMBER 6 Qin et al 643

diazepam,10-14 omeprazole,15,16 certain barbiturates,17,18 and chloroguanide (INN, proguanil).19,20 However, there was an issue surrounding the reason diazepam N-demethylation and polymorphic S-mephenytoin 4′- hydroxylation are associated with each other in white and Korean subjects but not in Chinese subjects. Bertilsson and Kalow14 hypothesized that this may be attributable to the higher proportion of CYP2C19 allelic heterozygotes in Chinese extensive metabolizers than in white extensive metabolizers. As expected, it is true.21 Because the Chinese and Korean subjects are Orientals with a similar ethic origin residing in the same geographic area, a further expanded question is why the findings in the native Chinese11 were inconsistent with those in Koreans.12 Therefore an attempt should be made to answer such a question by reassessing the elimination of diazepam in healthy Chinese subjects and its relationship to the genetically polymorphic oxidation of S-mephenytoin. In our preliminary data,13 we found

1 that the plasma clearance and elimination half-life (t ⁄2)

1 of diazepam and the t ⁄2 of its N-demethylated metabolite desmethyldiazepam (nordiazepam) were significantly different between extensive and poor metabolizers of S-mephenytoin oxidation, in good agreement with that in Koreans12 and white subjects.10 Furthermore, we needed to know whether or not the gene dosage of the CYP2C19 gene affects the metabolism and disposition of diazepam and desmethyldiazepam in healthy Chi- nese. This study was designed to address that topic.

METHODS Fig 1. Plasma concentrations of diazepam (top panel) and desmethyldiazepam (bottom panel) after a single oral 5 mg Subjects. Eighteen unrelated adult men were recruited dose of diazepam in subjects of three different genotypes for the study from a total of 101 healthy Chinese vol- (open triangles, homozygous wt; open squares, heterozygous unteers who had been screened for CYP2C19 phenotype wt/m1; open circles, homozygous m1). The error bar indi- and genotype.9 Six individuals were homozygous for cates SD. the wild-type (wt/wt), six were heterozygous for allele m1 (wt/m1), and six were homozygous for allele m1 (m1/m1). All participants were screened by use of a after administration. Plasma was obtained and kept careful medical history, physical examination, and rou- frozen at Ð20¡C until analyzed. tine laboratory tests before entry into this study. Writ- Plasma diazepam and desmethyldiazepam assay. ten informed consent was obtained from each subject Plasma concentrations of diazepam and desmethyl- according to the study protocol, which was approved by diazepam were determined by gas chromatographic the Ethical Committee of Hunan Medical University. No method developed in our laboratory. We added 0.5 mL medications, including alcohol, were permitted for at sodium hydroxide (0.5N, pH 13) and 100 µL least 2 weeks before and during the study. A single oral clomipramine (internal standard) to 1 mL of the plasma dose of 5 mg diazepam (HoffmannÐLa Roche, Basel, sample. The plasma sample was diluted with 5 mL of Switzerland) with 100 mL water was given to the sub- absolute ether as eluate, mixed and shaken for 5 minutes, jects in the morning after overnight fasting, and 10 mL and then followed by centrifugation (2000g × 10). The venous blood samples were collected into heparinized transferred ether phase was evaporated at 45¡C and the tubes from a forearm vein of each subject immediately residue dissolved in 20 µL methanol (HPLC reagent), of before dosing (as a blank or drug-free sample) and at 1, which 3 µL was assayed by gas chromatography. Refer- 2, 4, 8, 12, and 24 hours and then 2, 3, 6, and 12 days ence diazepam and desmethyldiazepam were purchased 中国科技论文在线______www.paper.edu.cn CLINICAL PHARMACOLOGY & THERAPEUTICS 644 Qin et al DECEMBER 1999

Table I. Demographic data and pharmacokinetic parameters of oral diazepam in healthy Chinese subjects P Values wt/wt (A) wt/m1 (B) m1/m1 (C) A versus B A versus C B versus C Age (y) 28 ± 9 21 ± 2 24 ± 4 NS NS NS Weight (kg) 59 ± 3 58 ± 7 58 ± 5 NS NS NS Diazepam AUC (µg á h/L) 5.3 ± 2.7 13.1 ± 5.2 32.4 ± 10.1 <.01 <.01 <.01 CL (mL/min) 19.5 ± 9.8 7.2 ± 2.6 2.8 ± 0.9 <.01 <.01 <.05

t1 (h) 20.0 ± 10.8 62.9 ± 9.8 84.0 ± 13.7 <.01 <.01 <.01 ⁄2 Desmethyldiazepam AUC (µg á h/L) 9.5 ± 1.9 13.4 ± 2.3 22.5 ± 9.7 <.01 <.01 <.05

t1 (h) 99.2 ± 21.7 132.1 ± 24.9 176.0 ± 28.9 <.05 <.01 <.01 ⁄2

Data are expressed as mean values ± SD; n = 6 in each group. NS, Not statistically significant (P > .05).

AUC, Area under the concentrationÐtime curve; CL, clearance; t1 , half-life. ⁄2 from Sigma Chemical Company (St Louis, Mo). RESULTS Clomipramine was provided by Ciba Geigy Company Each of the three genotypes had distinct time pro- (Basel, Switzerland). The chromatographic apparatus files of plasma diazepam and desmethyldiazepam (Fig consisted of an HP-17 capillary column and analyzed 1). The demographic characteristics and pharmacoki- with a nitrogen-phosphorus detector. Conditions of the netic parameters of oral diazepam in healthy Han Chi- operation for separation of the diazepam and its metabo- nese subjects of different CYP2C19 genotypes are lite desmethyldiazepam were an oven temperature of shown in Table I. There were intergenotypic differences

1 260¡C, an injector temperature of 279¡C, a detector tem- in the plasma t ⁄2 values of diazepam and its active perature of 275¡C, nitrogen flow rate through the column metabolite desmethyldiazepam between the different of 15 mL/min, hydrogen of 3.0 mL/min, and air flow rates genotypic groups, with the subjects homozygous for

1 of 110 mL/min. The intraday and interday coefficients of the allele m1 (m1/m1) having the longest t ⁄2 and the sub- variation of both diazepam and desmethyldiazepam were jects homozygous for the wild-type allele (wt/wt) hav- less than 7.7% and 9.7% over the concentration range ing the shortest one for both compounds. As expected, from 8 to 1000 ng/mL, respectively. the slowest mean clearance of diazepam existed in the Pharmacokinetic analysis. The diazepam and des- subjects with m1/m1, or vice versa, and the fastest was methyldiazepam elimination rate constants (β) were cal- observed in the wt/wt subjects. As hypothesized, the culated by log-linear regression analysis of the terminal heterozygotes (wt/m1) had the intermediate values of

1 phase of the concentrationÐtime curve. The area under the systemic clearance and elimination t ⁄2. concentrationÐtime curve (AUC) was estimated by the log- trapezoidal rule to the last measured concentration point DISCUSSION

1 and then extrapolated to infinity. The elimination t ⁄2 values The major genetic defect in the poor metabolizers of of both compounds of interest were calculated as follows: S-mephenytoin was identified to be a single nucleotide polymorphism in the coding sequence of exon 5 of t1 = ln(2)/β ⁄2 CYP2C19 (G681→A).22 Such a single nucleotide poly- Assuming complete bioavailability (F = 1), the oral morphism introduces a cryptic splice site in the exon, clearance (CL) of diazepam was estimated based on the alters the reading frame of the messenger ribonucleic equation: acid (mRNA) starting with amino acid 215, and produces a premature stop code, finally resulting in a CL = Dose/AUC nonfunctional protein. In theory, a certain drug-metab- Statistical analysis. Data are expressed as mean olizing enzyme is a gene product of the expression of values ± SD. One-way ANOVA and the Student t test its same or different alleles, thus its activity should be was applied to compare the intergenotypic differences regulated by the gene dose. In fact, the gene dosage of in the demographic variables or pharmacokinetic para- CYP2C19 can affect the enzyme activity of CYP2C19 meters. A P value of <.05 was considered to be statis- in healthy Chinese populations,8,9 Japanese subjects,23 tically significant. black subjects,24 and white subjects16 and can also 中国科技论文在线______www.paper.edu.cn CLINICAL PHARMACOLOGY & THERAPEUTICS VOLUME 66, NUMBER 6 Qin et al 645

affect the metabolism of some CYP2C19 substrates, 2. Wilkinson GR, Guengerich FP, Branch RA. Genetic poly- such as omeprazole,23 chloroguanide,25 and lansopra- morphism of S-mephenytoin hydroxylation. Pharmacol zole.26 This study was in complete agreement with the Ther 1989;43:53-76. above observations and was the first definite evidence 3. Wrighton SA, Stevens JC, Becker GW, VanderBranden that the gene dosage of CYP2C19 can markedly affect M. Isolation and characterization of human liver the metabolism and disposition of diazepam and des- cytochrome P450 2C19: correlation between 2C19 and S-mephenytoin 4′-hydroxylation. Arch Biochem Biophys methyldiazepam in humans. 1993;306:240-5. Diazepam is extensively metabolized in the liver 4. Goldstein JA, Faletto MB, Romkes-Sparks M, Sullivan through P450-catalyzed reactions, that is, ~60% of a T, Kitareewwan S, Raucy JL, et al. Evidence that given dose is N-demethylated to produce a major metabo- CYP2C19 is the major (S)-mephenytoin 4′-hydroxylase lite desmethyldiazepam,10,27 and the remaining is con- in humans. Biochemistry 1994;33:1743-52. verted to temazepam through C3-hydroxylation. In prepa- 5. Alván G, Bechtel P, Iselius L, Gundert-Remy U. Hydrox- rations of human liver microsome28 and complementary ylation polymorphisms of debrisoquine and mephenytoin deoxyribonucleic acidÐexpressed recombinant enzymes in European populations. Eur J Clin Pharmacol 1990;39: (CYP2C P450s),29 CYP2C19 has been identified to be a 533-7. principal contributor to diazepam N-demethylation at low 6. Bertilsson L. Geographic/interracial differences in poly- substrate concentrations. Furthermore, there is a good morphic drug oxidation: current state of knowledge of correlation between the levels of CYP2C19 activity and cytochrome P450 (CYP) 2D6 and 2C19. Clin Pharma- the formation of desmethyldiazepam after administra- cokinet 1995;29:192-209. 7. Xie HG, Xu ZH, Luo X, Huang SL, Zeng FD, Zhou HH. tion of diazepam in white subjects10 and in Korean12 11,13 Genetic polymorphisms of debrisoquine and S-mepheny- subjects. However, conflicting results were toin oxidation metabolism in Chinese populations: a reported on the in vivo diazepam N-demethylation in meta-analysis. Pharmacogenetics 1996;6:235-8. Chinese subjects. The results of our study may partly 8. de Morais SMF, Goldstein JA, Xie HG, Huang SL, Lu explain why diazepam metabolism and polymorphic S- YQ, Xia H, et al. Genetic analysis of the S-mephenytoin mephenytoin hydroxylation are associated with each polymorphism in a Chinese population. Clin Pharmacol other in white and Korean populations but not in Chi- Ther 1995;58:404-11. nese populations as reported by Zhang et al.11 More- 9. Xiao ZS, Goldstein JA, Xie HG, Blaisdell J, Wang W, over, several previous lines of evidence indicated a Jiang CH, et al. Differences in the incidence of the greater interindividual variation in the clearance of CYP2C19 polymorphism affecting the S-mephenytoin diazepam, showing a dependence on age, gender, and phenotype in Chinese Han and Bai populations and iden- liver disease.30 We recently observed that the effect of tification of a new rare CYP2C19 mutant allele. J Phar- gender on S-mephenytoin 4′-hydroxylase activity is macol Exp Ther 1997;281:604-9. 10. Bertilsson L, Henthorn TK, Sanz E, Tybring G, Säwe EJ, dependent on the subject’s genotype of CYP2C19.31 Villén T. Importance of genetic factors in the regulation Interestingly, both male and female subjects were of diazepam metabolism: relationship to S-mephenytoin, 11 included in the study of Zhang et al. When only male but not debrisoquine, hydroxylation phenotype. Clin subjects were included, an agreeable conclusion on the Pharmacol Ther 1989;45:348-55. in vivo diazepam N-methylation was obtained in 11. Zhang Y, Reviriego J, Lou YQ, Sjöqvist F, Bertilsson L. Korean subjects12 and Chinese subjects.13 This may Diazepam metabolism in native Chinese poor and exten- suggest that such an inconsistent conclusion on sive hydroxylators of S-mephenytoin: interethnic differ- diazepam metabolism in Chinese subjects may also be ences in comparison with white subjects. Clin Pharma- attributable to the gender of subjects studied. col Ther 1990;48:496-502. In summary, we documented that the metabolism of 12. Sohn DR, Kusaka M, Ishizaki T, Shin SG, Jang IJ, Shin diazepam and its major metabolite desmethyldiazepam JG, et al. Incidence of S-mephenytoin hydroxylation defi- in Chinese subjects is dependent on both the phenotype ciency in a Korean population and the interphenotypic of S-mephenytoin 4′-hydroxylation and genotype of differences in diazepam pharmacokinetics. Clin Pharma- col Ther 1992;52:160-9. CYP2C19. 13. Wan J, Xia H, He N, Lu YQ, Zhou HH. The elimination of diazepam in Chinese subjects is dependent on the mephenytoin oxidation phenotype. Br J Clin Pharmacol References 1996;42:471-4. 1. Küpfer A, Preisig R. Pharmacogenetics of mephenytoin: 14. Bertilsson L, Kalow W. Why are diazepam metabolism a new drug hydroxylation polymorphism in man. Eur J and polymorphic S-mephenytoin hydroxylation associ- Clin Pharmacol 1984;26:753-9. ated with each other in white and Korean populations but 中国科技论文在线______www.paper.edu.cn CLINICAL PHARMACOLOGY & THERAPEUTICS 646 Qin et al DECEMBER 1999

not in Chinese populations? [letter]. Clin Pharmacol Ther K, et al. Pharmacokinetics of omeprazole (a substrate of 1993;53:608-10. CYP2C19) and comparison with two mutant alleles, 15. Andersson T, Regárdh CG, Dahl-Paustinen ML, Bertilsson CYP2C19m1 in exon 5 and CYP2C19m2 in exon 4, in L. Slow omeprazole metabolizers are also poor S-mepheny- Japanese subjects. Clin Pharmacol Ther 1996;59:647-53. toin hydroxylators. Ther Drug Monit 1990;12:415-6. 24. Edeki TI, Goldstein JA, de Morais SMF, Hajiloo L, Bul- 16. Chang M, Dahl ML, Tybring G, Gotharson E, Bertilsson ter M, Chapdelaine P, et al. Genetic polymorphism of S- L. Use of omeprazole as a probe drug for CYP2C19 phe- mephenytoin 4′-hydroxylation in African-Americans. notype in Swedish Caucasians: comparison with S- Pharmacogenetics 1996;6:357-60. mephenytoin hydroxylation phenotype and CYP2C19 25. Hoskins JM, Shenfield GM, Gross AS. Relationship genotype. Pharmacogenetics 1995;5:358-63. between proguanil metabolic ratio and CYP2C19 geno- 17. Küpfer A, Branch RA. Stereoselective mephobarbital type in a Caucasian population. Br J Clin Pharmacol hydroxylation cosegregates with mephenytoin hydrox- 1998;46:499-504. ylation. Clin Pharmacol Ther 1985;38:414-8. 26. Katsuki H, Nakamura C, Arimori K, Fujiyama S, Nakano 18. Adedoyin A, Prakash C, O’Shea D, Blair IA, Wilkinson M. Genetic polymorphism of CYP2C19 and lansoprazole GR. Stereoselective disposition of hexobarbital and its pharmacokinetics in Japanese subjects. Eur J Clin Phar- metabolites: relationship to the S-mephenytoin polymor- macol 1997;52:391-6. phism in Caucasian and Chinese subjects. Pharmacoge- 27. Jack ML, Colburn WA. Pharmacokinetic model for netics 1994;4:27-38. diazepam and its major metabolite desmethyldiazepam 19. Birkett DJ, Rees D, Andersson T, Gonzalez FJ, Miners following diazepam administration. J Pharm Sci 1983;72: JO, Veronese ME. In vitro proguanil activation to 1318-23. cycloguanil by human liver microsomes is mediated by 28. Yasumori T, Li QH, Yamazoe Y, Ueda M, Tsuzaki T, Kato CYP3A4 isoforms as well as by S-mephenytoin hydrox- R. Lack of low Km diazepam N-demethylase in livers of ylase. Br J Clin Pharmacol 1994;37:413-20. poor metabolizers for S-mephenytoin 4′-hydroxylation. 20. Funck-Brentano C, Becquemont L, Leneveu A, Roux A, Pharmacogenetics 1994;4:323-31. Jaillon P, Beaune P. Inhibition by omeprazole of proguanil 29. Jung F, Richardson TH, Raucy JL, Johnson EF. Diazepam metabolism: mechanism of the interaction in vitro and metabolism by cDNA-expressed hyman 2C P450s: inden- prediction of in vivo results from the in vitro experiments. tification of P4502C18 and P4502C19 as low Km J Pharmacol Exp Ther 1997;280:730-8. diazepam N-demethylase. Drug Metab Dispos 1997;25: 21. Xie HG. Direct evidence for the higher frequency of 133-9. CYP2C19 allelic heterozygotes in Chinese subjects than in 30. Greenblatt DJ, Harmatz JS, Shader RI. Factors influenc- white subjects [letter]. Clin Pharmacol Ther 1997;62:691-2. ing diazepam pharmacokinetics: age, sex and liver dis- 22. de Morais SMF, Wilkinson GR, Blaisdell J, Nakamura K, ease. Int J Clin Pharmacol Biopharm 1978;16:177-9. Meyer UA, Goldstein JA. The major genetic defect 31. Xie HG, Huang SL, Xu ZH, Xiao ZS, He N, Zhou HH. responsible for the polymorphism of S-mephenytoin Evidence for the effect of gender on activity of (S)- metabolism in humans. J Biol Chem 1994;269:15419-22. mephenytoin 4′-hydroxylase (CYP2C19) in a Chinese 23. Ieiri I, Kubota T, Urae A, Kimura M, Wada Y, Mamiya population. Pharmacogenetics 1997;7:115-9