Substrate selectivity of human cytochrome P450 2C9: importance of residues 476, 365, and 114 in recognition of diclofenac and sulfaphenazole and in mechanism-based inactivation by tienilic acid

of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Abstract

Published:

Views: 37 | Pages: 12

Extension: PDF | Download: 0

Share
Description
Substrate selectivity of human cytochrome P450 2C9: importance of residues 476, 365, and 114 in recognition of diclofenac and sulfaphenazole and in mechanism-based inactivation by tienilic acid
Tags
Transcript
  Substrate selectivity of human cytochrome P450 2C9: importanceof residues 476, 365, and 114 in recognition of diclofenac andsulfaphenazole and in mechanism-based inactivation by tienilic acid Armelle Melet, Nadine Assrir, Pascale Jean, Maria Pilar Lopez-Garcia, 1 Cristina Marques-Soares, Maryse Jaouen, Patrick M. Dansette,Marie-Agn  ees Sari, and Daniel Mansuy * Laboratoire de Chimie et Biochimie Pharmacologiques et Toxicologiques, UMR 8601 CNRS, Universit  ee  Paris V,45 Rue des Saints-P   ee res, 75270 Paris Cedex 06, France Received 29 May 2002, and in revised form 11 September 2002 Abstract A series of six site-directed mutants of CYP 2C9 were constructed with the aim to better define the amino acid residues that play acritical role in substrate selectivity of CYP 2C9, particularly in three distinctive properties of this enzyme: (i) its selective mechanism-based inactivation by tienilic acid (TA), (ii) its high affinity and hydroxylation regioselectivity toward diclofenac, and (iii) its highaffinity for the competitive inhibitor sulfaphenazole (SPA). The S365A mutant exhibited kinetic characteristics for the 5-hydrox-ylation of TA very similar to those of CYP 2C9; however, this mutant did not undergo any detectable mechanism-based inactivationby TA, which indicates that the OH group of Ser 365 could be the nucleophile forming a covalent bond with an electrophilicmetabolite of TA in TA-dependent inactivation of CYP 2C9. The F114I mutant was inactive toward the hydroxylation of dic-lofenac; moreover, detailed analyses of its interaction with a series of SPA derivatives by difference visible spectroscopy showed thatthe high affinity of SPA to CYP 2C9 (  K  s  ¼  0 : 4 l M) was completely lost when the phenyl substituent of Phe 114 was replaced with thealkyl group of Ile (  K  s  ¼  190    20 l M), or when the phenyl substituent of SPA was replaced with a cyclohexyl group(  K  s  ¼  120   30 l M). However, this cyclohexyl derivative of SPA interacted well with the F114I mutant (  K  s  ¼  1 : 6   0 : 5 l M). At theopposite end, the F94L and F110I mutants showed properties very similar to those of CYP 2C9 toward TA and diclofenac. Finally,the F476I mutant exhibited at least three main differences compared to CYP 2C9: (i) big changes in the  k  cat  and  K  m  values for TAand diclofenac hydroxylation, (ii) a 37-fold increase of the  K  i  value found for the inhibition of CYP 2C9 by SPA, and (iii) a greatchange in the regioselectivity of diclofenac hydroxylation, the 5-hydroxylation of this substrate by CYP 2C9 F476I exhibiting a  k  cat of 28 min  1 . These data indicate that Phe 114 plays an important role in recognition of aromatic substrates of CYP 2C9, presumablyvia  P -stacking interactions. They also provide the first experimental evidence showing that Phe 476 plays a crucial role in substraterecognition and hydroxylation by CYP 2C9.   2002 Elsevier Science (USA). All rights reserved. Keywords:  Yeast-expressed CYP 2Cs; Site-directed mutagenesis; Drug metabolism; Hydroxylation; Difference visible spectroscopy; Active-sitetopology;  P  –  P  interactions Cytochromes P450 constitute a superfamily of he-moproteins that play important roles in the oxidativemetabolism of a large variety of xenobiotics and en-dogenous compounds [1]. In order to interpret or topredict various problems that may occur with somedrugs in relation to genetic polymorphism and drug– drug interactions, it is very important to determinewhich human liver cytochrome P450 is mainly involvedin the metabolism of a drug of interest and what are thestructural determinants that are crucial for recognitionof this drug by the protein. Archives of Biochemistry and Biophysics 409 (2003) 80–91www.elsevier.com/locate/yabbi  ABB * Corresponding author. Fax: +33-1-42-86-83-87. E-mail address:  Daniel.Mansuy@biomedicale.univ-paris5.fr (D.Mansuy). 1 Present address: Departemento de Bioquimica y BiologicaMolecular, Universidad de Valencia, Facultad de Farmacia, Burjasot,46100 Valencia, Spain.0003-9861/02/$ - see front matter    2002 Elsevier Science (USA). All rights reserved.PII: S0003-9861(02)00548-9  Cytochromes P450 of the 3A and 2C subfamilies arethe major isoforms in human liver [1]. Of the fourmembers of the human 2C subfamily, CYP 2 2C8, 2C9,2C18, and 2C19, CYP 2C9 is the protein expressed atthe highest level in human liver [2,3]. It is involved in themetabolism of a great number of drugs [4], such asdiclofenac [5] and ( S  )-warfarin [6].Several models describing the structural characteris-tics of the substrates and inhibitors that are required forstrong interaction with CYP 2C9 have been proposed onthe basis of biochemical, UV–visible [7–12], and  1 HNMRresults[10].Allthesemodelsproposedthatatleasttwo major interactions are involved in the binding of substrates to CYP 2C9. Most of the high-affinity sub-strates or inhibitors of CYP 2C9 involve an anionic orpolar site, and the first major interaction with CYP 2C9would imply either an ionic or a hydrogen bond betweenthis site and a CYP 2C9 amino acid residue. The secondmajor interaction would involve an aromatic region of the CYP 2C9 active site establishing a  P -stacking bondwith the aromatic rings that are very often present in theCYP 2C9 substrates and inhibitors. Moreover, several3D models based on homology molecular modeling havebeen proposed for CYP 2C9 [11–18]. However, limiteddata are currently available about the precise nature of the amino acids of the CYP 2C9 active site that aremainlyinvolvedinthespecificrecognitionandbindingof CYP 2C9 substrates and inhibitors.A large number of mutants and chimeras have beenconstructed in order to understand why CYP 2C9 andCYP 2C19 exhibit very different substrate selectivities,despite the fact that only 43 of 490 amino acids differbetween these two closely related hemoproteins [18–21].For instance, CYP 2C19 efficiently catalyzes the 4 0 -hy-droxylation of ( S  )-mephenytoin [22,23] and the 5-hy-droxylation of omeprazole [24,19], while CYP 2C9exhibits very little activity toward either of these sub-strates. Studies of the CYP 2C9 and 2C19 mutants andchimeras showed that residues 99, 220, and 221 as wellas residues 286, 289, 292, and 295 play a key role in thesedifferences in behavior between CYP 2C9 and 2C19 [18– 21]. On the basis of the structure of the closely relatedenzyme CYP 2C5 [25], it was proposed that these resi-dues are unlikely to directly contact the substrate duringcatalysis but are positioned to influence the packing of substrate binding site and likely substrate access chan-nels in the enzyme [18].Returning to CYP 2C9 itself, several site-directedmutants have been constructed to investigate the im-portance of some residues of the 95–114 region, thesubstrate recognition site 1 (SRS1) according to Gotoh[26], in the binding of CYP 2C9 substrates [19,27,28].Mutation of Arg 97 led to a marked increase in the  K  m of diclofenac 4 0 -hydroxylation; however, the mostspectacular result was observed after mutation of Arg108 to Ala, which led to a dramatic decrease in thediclofenac 4 0 -hydroxylation activity [28]. A recent com-bined protein and pharmacophore model for CYP 2C9suggests that Arg 108 is a key residue in binding anionicand polar CYP 2C9 substrates [12], in agreement withmutagenesis results [28]. Arg 108 could thus be theprotein residue responsible for the ionic or hydrogenbond interaction with the anionic or polar site of sub-strates. Mutation of Val 113 and Phe 114 showed thatthese residues should be involved in the hydrophobicbinding of warfarin to CYP 2C9 [27]. Moreover, thecorresponding results are consistent with  P -stacking of Phe 114 with aromatic substrates or inhibitors such asSPA.The objective of the present study was to determinethe residues of the CYP 2C9 active site that play acentral role in three specific characteristics of CYP 2C9:(i) its inactivation by the selective mechanism-based in-hibitor tienilic acid (TA) [29], (ii) its ability to hydrox-ylate diclofenac in a highly regioselective manner, atposition 4 0 [5], and (iii) its particularly high affinity forits selective competitive inhibitor, SPA [30–35]. For thatpurpose, we have constructed some site-directed mu-tants of CYP 2C9 on the basis of previous mechanismstentatively proposed to explain these characteristics of CYP 2C9 and of 3D homology models of CYP 2C9deduced from the X-ray structure of CYP 2C5 [25]. Materials and methods Chemicals All chemicals used were of the highest quality com-mercially available. TA was provided by Anphar– Rolland (Chilly-Mazarin, France), diclofenac and SPAby Sigma (St. Quentin, Fallaviers, France). Physical measurements UV–visible spectra were recorded on a KontronUvikon 860 spectrophotometer equipped with a turbidsample accessory or on an Aminco DW-2 apparatusmodified by Olis, Inc.  1 H NMR spectra were recorded at27  C on a Bruker ARX-250 instrument; chemical shiftsare reported downfield from  ð CH 3 Þ 4 Si and couplingconstants are in Hz. The abbreviations s, d, t, q, m, bs,and dd are used for singlet, doublet, triplet, quadruplet,multiplet, broad singlet, and doublet of doublet, re-spectively. Mass spectra (MS) were performed withchemical ionization (CI) using NH 3  on a Nermag R1010apparatus. 2 Abbreviations used:  CI, chemical ionization; CYP, cytochromeP450; DMF, dimethylformamide; Et 2 O, diethylether; SPA, sulfaphe-nazole; TA, tienilic acid; 3D, three-dimensional; SRS1, substraterecognition site 1. A. Melet et al. / Archives of Biochemistry and Biophysics 409 (2003) 80–91  81  Synthesis of SPA derivatives SPA derivatives  1  and  2  (see formula in Table 4) weresynthesized according to previously described proce-dures [34]. Compound  3  (4-allyl- N  -(2-cyclohexyl-2 H  -pyrazol-3-yl)-benzenesulfonamide) was prepared andcharacterized as follows.2-Cyclohexyl-2 H  -pyrazol-3-ylamine was prepared aspreviously described [36]. 4-Bromobenzenesulfonylchloride (400mg, 1.56mmol) was added to a solution of 2-cyclohexyl-2 H  -pyrazol-3-ylamine (250mg, 1.51mmol)in 2.5ml of anhydrous pyridine. After 10min at roomtemperature, the reaction mixture was heated for 75minat 95  C. The solvent was evaporated and the residuedissolved in 1N NaOH. The aqueous phase was washedwith CH 2 Cl 2 , acidified with 1N HCl, and extracted withCH 2 Cl 2 . The organic phase was dried over MgSO 4  andconcentrated. Three hundred and ten milligrams of 4-bromo- N  -(2-cyclohexyl-2 H  -pyrazol-3-yl)-benzenesulf-onamide was obtained by crystallization and an addi-tional 70mg after purification of the mother liquor bycolumn chromatography (SiO 2 = CH 2 Cl 2  and then +4%Et 2 O) (65% yield); mp 234–235  C;  1 H NMR (CDCl 3 )  d 7.63 (s, 4H), 7.37 (s, 1H), 6.29 (bs, 1H, NH), 5.63 (s,1H), 4.20 (m, 1H), 1.85–1.67 (m, 7H), 1.40–1.18 (m, 3H).Tetrakis(triphenylphosphine)palladium(0) was prepareda day before as previously described [37]. 4-Bromo- N  -(2-cyclohexyl-2 H  -pyrazol-3-yl)-benzenesulfonamide (210mg, 0.55mmol) and the palladium(0) catalyst (36mg)were introduced in a dry schlenck with septum andmagnetic stirring. All the experiments were performedunder argon. Five milliliters of anhydrous and deoxy-genated DMF and 0.19ml of allyltributyltin were in-troduced successively with a syringe. The solution waskept 0.5 h at room temperature, then heated at 110  Cfor 15h. DMF was evaporated. After purification bycolumn chromatography (SiO 2 ; CH 2 Cl 2  þ 5%Et 2 O)140mg of   3  were obtained and recrystallized from Et 2 O;mp 164–165  C;  1 H NMR (CDCl 3 )  d  7.67 (d, 2H, J  ¼ 8.2), 7.35 (d, 1H,  J  ¼ 1.5), 7.31 (d, 2H,  J  ¼ 8.2), 6.19(bs, 1H, NH), 5.91 (m, 1H), 5.66 (d, 1H,  J  ¼ 1.5), 5.10(m, 2H), 4.16 (m, 1H), 3.44 (d, 2H,  J  ¼ 6.6), 1.9–1.66 (m,7H), 1.33–1.16 (m, 3H); MS (CI, NH 3 )  m / z ¼ 346( ½ M  þ  H  þ , 78%); 166 (100%). Biological materials The TG1  Escherichia coli   strain [ D ( lac pro )  supE thi hsdD5 F  0 traD35 proAB LacIq LacZDM15 ] was used forplasmid amplification.  Saccharomyces cerevisiae  strainsW(N), W(R), and W(hR) and the yeast expression vec-tor pYeDP60 were a gift from Pompon (CNRS, Gif-sur-Yvette, France). W303-1B (MAT  a ;  ade 2-1;  his3 -11,-15;  leu2 -3,-112;  ura3 -1;  trp1 -1), designated as W(N),was described previously [38]. Yeast strain W(R) wasconstructed by substitution of the natural promoter of the W303-1B gene encoding cytochrome P450 reductaseby the galactose-inducible  GAL10-CYC1  hybrid pro-moter [39]. Yeast strain W(hR) was obtained from W(R)by replacing the gene of yeast cytochrome P450 reduc-tase with that coding for human cytochrome P450 re-ductase [40]. CYP 2C9 expression vector (2C9V60) wasconstructed by insertion of CYP 2C9 cDNA in thepYeDP60 polylinker as reported previously [41].Growth media for  E. coli   and  S. cerevisiae  were obtainedfrom Difco (COGER, France) or Gibco-BRL (Cergy-Pontoise, France). Oligonucleotides were synthesized byGenome Express (Montreuil, France). Restriction en-donucleases and DNA modification enzymes were of thebest grade available. Plasmid DNA was isolated with theplasmid miniprep kit from Bio-Rad (Marnes la Co-quette, France). NADPH, NADP, glucose 6-phosphate,and glucose-6-phosphate dehydrogenase were purchasedfrom Roche Diagnostics (Meylan, France). Site-directed mutagenesis Site-directed mutagenesis of CYP 2C9 was performedby two successive PCRs followed by homologous re-combination in yeast [42,43]. The primers used for mu-tagenesis are listed in Table 1. Two other primers wereused for PCR, Pro (5 0 -ACGACACATGATCATATGGCATGCATGTGCTCTGTATGTATATAA-3 0 ) andTer (5 0 -TCACACACAGGAAACAGCTATGACCATGATTACGCCAAGCTTTTCGA-3 0 ), annealing witheitherthepromoterregionortheterminatorregionoftheCYP2C9expressionvectorpYeDP60.ThefirstPCRwasachieved with the mutagenic primer and either the pro-moterprimerPro(F94L,F110I,F114I)ortheterminatorprimer Ter (F476I, S365A, S365G). The PCR conditionson a Minicycler (MJ Research) were an initial denatur-ationstepfor2minat94  C,26cyclesof40sdenaturationat92  Cfollowedby2minannealingat50  Cand1.5minextension at 72  C and a final step of 5min at 72  C. TheamplificationproductwasthenusedinasecondPCRwiththe terminator primer Ter (F94L, F110I, F114I) or thepromoter primer Pro (F476I, S365A, S365G) in order toamplify the whole gene. The PCR conditions were 2minpredenaturation at 95  C, followed by 24 cycles of 30s at Table 1Oligonucleotides used in the mutagenesis of CYP 2C9CodonchangePrimer sequenceF94L 5 0 -GCCTCTTCCAGA T AACTCCTCCTCTCCAAG-3 0 F110I 5 0 -GAAAACAATTCCAA T TCCTCTGTTAG-3 0 F114I 5 0 -CTTTCCATTGCTGA T AACAATTCCAAATCC-3 0 F476I 5 0 -AGTTGTCAATGG A TTGCCTCTGTGC-3 0 S365A 5 0 -CTTTCTCCCACCG C CCTGCCCCATGCA-3 0 S365G 5 0 -CTTTCTCCCACCG G CCTGCCCCATGCA-3 0 Changed oligonucleotides are in bold. Codons for the changedamino acids are underlined. The first three primers are reverse oligo-nucleotides.82  A. Melet et al. / Archives of Biochemistry and Biophysics 409 (2003) 80–91  94  C, 3min at 68  C, and 5min at 72  C. The final PCRproduct was cotransformed into yeast strain W(N) withthe double-digested  Bam HI/ Eco RI plasmid expressionvector pYeDP60. The DNA of the gap repair clones wasextractedaccordingtoamethoddescribedpreviously[44]and electroporated into TG1  E. coli  . All the selectedplasmidsweresequencedbyGenomeExpress(Montreuil,France)toconfirmincorporationofthedesiredmutation. Yeast transformation, cell culture, and preparation of the yeast microsomal fraction The expression system used for human liver P450swas based on a yeast strain, W(R), previously described[39], in which yeast cytochrome P450 reductase wasover-expressed. Yeast transformation by the pYeDP60vector containing  CYP   2C9 or mutated cDNAs wasperformed as described previously [39,44]. Yeast cultureand microsome preparation were described elsewhere[45]. Microsomes were homogenized in 50mM Trisbuffer, pH 7.4, containing 1mM EDTA and 20% glyc-erol (v/v), aliquoted, frozen under liquid N 2 , and storedat  ) 80  C until use. Microsomal P450 content was de-termined according to the method of Omura and Sato[46]. The NADPH–cytochrome P450 reductase activitywas measured spectroscopically by the method of Ver-milion and Coon [47], monitoring the reduction of cy-tochrome  c  at 550nm  ð e 550  ¼  21 ; 400M  1 cm  1 Þ . Theprotein content in microsomal suspensions was mea-sured by the Lowry procedure [48] using bovine serumalbumin as the standard. Immunoblotting  Electrophoresis was carried out using a Bio-Rad MiniProtean II device. Proteins were separated on a 10%SDS–PAGE gel [49] and transferred to a nitrocellulosemembrane using a Transblot semidry transfer cell fromBio-Rad. The membrane was blocked and probed aspreviously described [50] with a rabbit polyclonal anti-P450 2C antibody provided by P. Beaune (Inserm U490,Paris). Study of substrate binding to yeast-expressed CYP 2C9by difference visible spectroscopy Difference visible spectra produced by SPA deriva-tives were recorded at room temperature with an Am-inco DW-2 spectrophotometer, modified by Olis, Inc.Yeast microsomes were suspended in a 0.1M phosphatebuffer, pH 7.4, containing 0.1 mM EDTA, to obtain aP450 concentration of 0.1–0 : 15 l M. The solution wasequally divided between two 500 l l quartz cuvettes(1-cm pathlength) and a baseline was recorded. Aliquots(1  –  5 l l) of Me 2 SO solutions containing the studiedcompound were added to the sample cuvette, the samevolume of Me 2 SO being added to the reference cuvette.The difference spectra were recorded between 380 and520nm [51]. Molecular modeling  A 3D model of CYP 2C9 was constructed using aprocedure very similar to that recently reported [16–18]for 3D models of CYP 2C9 based on the structure of CYP 2C5 [25]. These enzymes have a sequence identityof 77% and a similarity of 83%, which makes CYP 2C5 agood template for the modeling of CYP 2C9. The In-sight II software package (Biopolymer, Discover_3)from Molecular Simulations Inc. (San Diego, CA, USA)was used following a previously described protocol [52].The coordinates of CYP 2C9 residues were assigned onthe basis of the corresponding residues of CYP 2C5. Theheme group was transferred from CYP 2C5 and fixed.The other part of the protein was energy minimized,using the module Discover_3 and the esff force field. Thesubstrate was then placed in the distal binding pocket, inan orientation favorable to the experimentally observedproduct. A minimization was finally performed usingthe Discover_3 module. The quality of the model wasassessed using the program Procheck [53], which calcu-lates an overall score of the structure. The authors of this program consider that a protein model is of highquality when the normalized score is lower than 0.5. Thescore found for our CYP 2C9 model was 0.07 (to becompared with the score calculated for the CYP 2C5 X-ray structure: 0.01). Two other parameters calculated bythis program confirmed the quality of the model: (i) thepercentage backbone  /  –  W  angles within the allowedRamachandran region was 73% for the CYP 2C9 model,to be compared with 71.3% for the CYP 2C5 structure,and (ii) the numbers of structural features that differfrom average values in known proteins were only 1 forbond angles, 1 for dihedral angles, and 0 for bond dis-tances in our CYP 2C9 model. Enzyme activity assaysGeneral enzyme assay conditions . Incubations formetabolic activity with yeast microsomes were carriedout at 28  C, using glass tubes in a shaking bath. Theincubation mixtures contained yeast microsomes, thesubstrate, and an NADPH-generating system (1mMNADP þ , 10mM glucose 6-phosphate, and 2 units of glucose-6-phosphate dehydrogenase/ml) diluted in a50mM Tris buffer, pH 7.4, containing 1mM EDTA and8% glycerol (final concentrations). In the particularcases of the S365A and S365G mutants, a 50mM po-tassium phosphate buffer, pH 7.4, containing 0.5mMEDTA and 8% glycerol was used. Activity assays wereroutinely initiated ( t  0  ¼  0min) by addition of theNADPH-generating system to the incubation mixture A. Melet et al. / Archives of Biochemistry and Biophysics 409 (2003) 80–91  83  after 3min of separate preincubation at 28  C for tem-perature equilibration. 5-Hydroxylation of TA . Quantitation of 5-hydroxyTA was based on a spectrophotometric method [54]adapted to yeast microsomes expressing CYP 2C9 [55].Incubations were carried out under the general condi-tions described above, with 0.1–0 : 2 l M P450. At  t  0  andregularly thereafter, aliquots (140 l l) were taken and thereaction was quickly stopped by treatment with 70 l l of a cold CH 3 CN = CH 3 COOH (10/1) mixture.Inhibition studies of CYP 2C9 were performed atvarious concentrations of inhibitors (1  –  100 l M) and fiveconcentrations of TA in the range 5  –  400 l M. The in-hibitor dissolved in Me 2 SO (final concentration 0.5%)and the substrate TA were added simultaneously to theincubation mixture.  K  i  values were derived from analy-ses of Lineweaver–Burk plots corresponding to thevarious enzymatic activities in the presence of increasingconcentrations of inhibitor. Diclofenac hydroxylation assay.  Diclofenac (sodiumsalt) was incubated as described above at 28  C in thepresence of yeast microsomes, containing 0 : 05 l M P450.At  t  0  and regularly thereafter, aliquots (120 l l) were ta-ken and the reaction was quickly stopped by treatmentwith 60 l l of a cold CH 3 CN = CH 3 COOH (10/1) mixturecontaining TA as internal standard (final concentration5 l M). Proteins were precipitated by centrifugation for5min at 10,000rpm and the supernatant was stored at ) 40  C until analysis. Formation of the metabolites wasfollowed by reverse-phase HPLC (Spectra System AS3000 autosampler). Supernatant aliquots were injectedonto an X-Terra MS C18 column (Waters, France)(250   4 : 6mm, 5 l m). The mobile phase (A, 25mMNH 4 HCO 3  buffer, pH 7.8/B, CH 3 CN) was delivered at arate of 1 ml per minute with the following gradient: t   ¼  0 A/B (80/20),  t   ¼  2min A/B (80/20),  t   ¼  18min A/B(72/28),  t   ¼  21min A/B (50/50),  t   ¼  25min A/B (50/50), t   ¼  26min A/B (80/20),  t   ¼  30min A/B (80/20). Moni-toring of the column effluent was performed with ascanning Spectra Focus UV detector between 250 and320nm. In that system, retention times for 4 0 -hydrox-ydiclofenac, 5-hydroxydiclofenac, the internal standardTA and the substrate diclofenac were  R t  ¼ 15, 19, 21,and 24min, respectively. Study of CYP 2C9 inactivation by TA The experimental design to follow enzyme inactiva-tion was based on general procedures previously de-scribed [56,57] that have been applied to demonstrateCYP 2C9 inactivation by TA [29]. Basically, microsomes(equivalent to 0 : 2 l M P450) were preincubated underthe general conditions already outlined with 10 l M TAand an NADPH-generating system. At  t  0  and regularlythereafter (from 0 to 45min), parallel aliquots were re-moved from the incubation medium and were immedi-ately processed to determine TA metabolite formation(140 l l) and residual enzymatic activity (300 l l). Results Choice, construction, and expression of CYP 2C9 mutants Homology models of CYP 2C9 and CYP 2C19 havebeen constructed recently [16–18], on the basis of the X-ray structure of CYP 2C5 [25]. We also constructed sucha tentative 3D model of CYP 2C9 by using almostidentical procedures (see Materials and methods). Themain structural characteristics of the active site of ourmodel were very similar to those of the previously re-ported models [16–18]. An energy minimization wasperformed to place a diclofenac molecule into the activesite of our CYP 2C9 model. This led to a complex inwhich the 4 0 -hydrogen of diclofenac was located ap-proximately 5  AA from the heme iron (Fig. 1), a value inagreement with previous estimates for the iron-hydrox-ylation site distances in CYP 2C9–substrate complexes[10]. Moreover the positioning of diclofenac in closeproximity to the two phenylalanine residues 114 and 476was similar to the positioning of ( S  )-mephenytoin in therecently published model of the CYP 2C19–mepheny-toin complex [18]. Since previous results have suggestedthe great importance of   P  –  P  interactions betweensubstrate aromatic rings and CYP 2C9 aromatic resi-dues in substrate recognition by CYP 2C9 [27,34], Phe114 and 476 appeared as good candidates for such in-teractions (Fig. 1).Mechanism-based inactivation of CYP 2C9 by TA isdirectly related to the covalent binding of an electro-philic metabolite of TA to the protein [29]. The nature of the nucleophilic residue of the protein involved in thiscovalent binding has never been determined. Since otherclosely related members of the CYP 2C subfamily, CYP2C8 and CYP 2C18, hydroxylate TA without any in-activation [58], it was tempting to speculate that thisdifferent behavior is due to the lack of the protein nu-cleophilic residue responsible for covalent binding of CYP 2C9 to TA metabolites in CYP 2C8 and CYP2C18. The only nucleophilic residue that is in goodproximity to the substrate in the model of CYP 2C9(Fig. 1), and that is replaced by nonnucleophilic residuesin CYP 2C8 and CYP 2C18, is Ser 365.For the aforementioned reasons, we decided to con-struct CYP 2C9 mutants at the level of Phe 476, Phe114,and Ser 365. As two other phenylalanines (Phe 94 and110) are present in the same region as Phe 114 (B–Cloop), we also mutated the two former residues to moreclearly establish the key role of Phe 114.The six single-amino-acid mutants F94L, F110I,F114I, F476I, S365A, and S365G were constructedusing a double PCR approach followed by homologous 84  A. Melet et al. / Archives of Biochemistry and Biophysics 409 (2003) 80–91
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x