ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 342, No. 2, June 15, pp. 329–337, 1997Article No. BB970125

Reconstitution of Recombinant Cytochrome P450 2C10(2C9)and Comparison with Cytochrome P450 3A4 and OtherForms: Effects of Cytochrome P450–P450and Cytochrome P450–b5 Interactions1

Hiroshi Yamazaki,* Elizabeth M. J. Gillam,† Mi-Sook Dong,‡,2 William W. Johnson,‡F. Peter Guengerich,‡ and Tsutomu Shimada*,3

*Osaka Prefectural Institute of Public Health, Nakamichi, Higashinari-ku, Osaka 537, Japan; †Department of Physiologyand Pharmacology, University of Queensland, St. Lucia 4072, Australia; and ‡Department of Biochemistry and Center inMolecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232

Received February 4, 1997, and in revised form March 27, 1997

systems. Neither apo- nor holo-b5 increased bufuralol1*-hydroxylation activity by CYP1A1 or 2D6 or theoph-Tolbutamide methyl hydroxylation and S-warfarinylline 8-hydroxylation by CYP1A2. Interestingly, we7-hydroxylation activities were reconstituted in sys-found that testosterone 6b-hydroxylation by CYP3A4tems containing recombinant human cytochrome P450was stimulated by CYP1A2 (and also by a modified(P450 or CYP) 2C10(2C9) and the optimal conditionsform in which the first 36 residues of the native humanfor the systems were compared with those of bufuralolprotein were removed) and CYP1A1 as well as by b5,1*-hydroxylation by CYP1A1, theophylline 8-hydroxyl-and such stimulations were not seen when other P450ation by CYP1A2, bufuralol 1*-hydroxylation byproteins (e.g., CYP2C10, 2D6, or 2E1) were added to theCYP2D6, chlorzoxazone 6-hydroxylation by CYP2E1,reconstituted systems. In contrast, substrate oxida-and testosterone 6b-hydroxylation by CYP3A4. CYP2C10tions by CYP2C10 and CYP2E1 were not stimulated byrequired cytochrome b5 (b5) for optimal rates of tolbu-other P450 proteins. The present results suggest thattamide and S-warfarin oxidations and b5 could be re-there are differences in optimal conditions for recon-placed by apo-b5; apo-b5 and b5 effects on the reconsti-stitution of substrate oxidations by various forms oftuted systems have already been reported in systemshuman P450 enzymes, and in some P450-catalyzed re-containing CYP3A4 for the oxidation of testosteroneactions protein–protein interactions between P450and nifedipine and for the rapid reduction of CYP3A4and b5 and other P450 proteins are very important inby NADPH–P450 reductase (H. Yamazaki et al., 1996,some oxidations catalyzed by CYP2C10, 2E1, and 3A4.J. Biol. Chem. 271, 27438–27444). Stopped-flow studies,q 1997 Academic Presshowever, suggested that apo-b5 as well as b5 did not

Key Words: human P450; b5; CYP2C10; CYP2C9;cause stimulation of the reduction of CYP2C10 byCYP3A4; CYP2E1; CYP2D6; CYP1A2; CYP1A1; reconsti-NADPH–P450 reductase, while the reduction ratestution; xenobiotics; drug oxidation.were dependent on the substrates in reconstituted sys-

tems. Chlorzoxazone 6-hydroxylation by CYP2E1 wasstimulated by b5, but not by apo-b5, in reconstituted

Multiple forms of cytochrome P450 (P450 or CYP)4

1 Supported in part by grants from the Ministry of Education, Sci- are found in liver microsomes and these P450 formsence, and Culture of Japan, the Ministry of Health and Welfare of play important roles in the oxidation of structurallyJapan, the Developmental and Creative Studies from Osaka Prefec-

diverse xenobiotic chemicals such as drugs, toxic chem-tural Government, and by United States Public Health Serviceicals, and carcinogens as well as endobiotic chemicalsGrants CA44353 and ES00267.

2 Current address: Korea Institute of Science and Technology, P.O.Box 131, Cheongryang, Seoul 130–650, South Korea.

4 Abbreviations used: P450 or CYP, cytochrome P450; b5, cyto-3 To whom correspondence and reprint requests should be ad-dressed at Osaka Prefectural Institute of Public Health, 3–69 Na- chrome b5; DLPC, L-a-dilauroyl-sn-glycero-3-phosphocholine; DOPC,

dioleoyl-sn-glycero-3-phosphocholine; PS, L-a-phosphatidyl-L-serine;kamichi 1-chome, Higashinari-ku, Osaka 537, Japan. Fax: (81)6-972-2393. BSA, bovine serum albumin; HSA, human serum albumin.

3290003-9861/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID ABB 0125 / 6b38$$$$61 05-09-97 23:59:53 arca

330 YAMAZAKI ET AL.

including steroids, fatty acids, fat-soluble vitamins, these recombinant human P450 enzymes are also re-ported.and prostaglandins (1). Major P450 enzymes in human

liver microsomes identified to date include CYP1A2,MATERIALS AND METHODS2A6, 2B6, 2C8, 2C9/105 (2–4), 2C19, 2D6, 2E1, 3A4,

and 3A5 (and 3A7 in fetal livers) (5). Large interindi- Chemicals. Theophylline, tolbutamide, chlorzoxazone, testoster-one, and nifedipine were purchased from Sigma Chemical Co. (St.vidual variations exist in the levels of each of theseLouis, MO), and S-warfarin and ({)-bufuralol were from UltraFineP450 enzymes, and these variations are considered oneChemicals Co. (Manchester, UK). Other chemicals used were fromof the major factors contributing to different suscepti-the same sources as described previously or of highest qualities com-

bilities of humans toward actions and toxicities of mercially available (5).drugs, toxic chemicals, and carcinogens (5). Recently, Enzymes and lipids. Recombinant CYP1A1, 1A2, 2C10, 2D6,many investigators have examined P450 proteins puri- 2E1, and 3A4 were purified to gel-electrophoretic homogeneity from

membranes of Escherichia coli into which respective cDNA clonesfied from organisms in which native and modified hu-containing certain N-terminal modifications had been introduced, asman P450 cDNAs have been introduced, in order todescribed elsewhere (13–19). The turnover numbers (nmol productsdetermine the molecular mechanisms for genetic poly- formed/min/nmol P450) for marker substrates of these P450 enzymes

morphisms and induction by chemical agents of indi- in the optimal reconstituted systems were 12 for 7-ethoxyresorufinvidual P450 enzymes in humans (6). O-deethylation by CYP1A1, 1.4 for theophylline 8-hydroxylation by

CYP1A2, 6.3 for tolbutamide methylhydroxylation by CYP2C10, 0.75Conditions for reconstitution of liver microsomalfor bufuralol 1*-hydroxylation by CYP2D6, 9.0 for chlorzoxazone 6-P450 proteins with NADPH–P450 reductase in syn-hydroxylation by CYP2E1, and 14 for testosterone 6b-hydroxylationthetic phospholipids in order to obtain maximal cata- by CYP3A4. Liver microsomal CYP1A2 was purified from liver micro-

lytic activities have been shown to depend upon the somes of rats (20), rabbits (21), or humans (22). Rabbit NADPH–P450 reductase (23) and rabbit b5 and apo-b5 (24) were purified fromP450 proteins employed (7). In some cases, particularlyliver microsomes using the methods described.when CYP3A enzymes are used for reconstitution, sev-

Liver lipids were extracted from human liver microsomes as de-eral factors such as lipid membrane environment andscribed (25).other components including b5, divalent metal ions

Enzyme assays. Standard incubation mixtures (final volume ofsuch as Mg2/, and reduced GSH have been shown to be 0.20 ml) for bufuralol 1*-hydroxylation by CYP1A1 and 2D6 andrequired for maximal catalytic activity (8–10). Other theophylline 8-hydroxylation by CYP1A2 contained P450 (5 pmol),

NADPH–P450 reductase (10 pmol), and cholate (100 nmol) and lipidhuman P450 enzymes which have been shown to re-mixture (4 mg of a mixture of DLPC, DOPC, and PS, 1/1/1, w/w/w)quire b5 and particular lipid environments for optimalin 100 mM potassium phosphate buffer (pH 7.4) (25 mM whenreconstitution include CYP2E1 and possibly CYP2C9/ CYP2D6 was used) containing an NADPH-generating system (0.10

10 and CYP2C19, but the mechanisms underlying mmol of NADP/, 1 mmol of glucose 6-phosphate, and 0.05 unit ofstimulation by b5 have been reported to differ de- glucose-6-phosphate dehydrogenase) and bufuralol (40 nmol) or the-

ophylline (100 nmol). Product formation was determined by HPLCpending on the P450 enzymes examined (11, 12).with a C18 5-mm analytical column (4.6 1 150 mm; Kanto Chemical,In this study, we examined optimal conditions forTokyo).reconstitution of tolbutamide methyl hydroxylation Methyl hydroxylation of tolbutamide and 7-hydroxylation of S-

and S-warfarin 7-hydroxylation activities by recombi- warfarin (both final volume of 0.20 ml) were determined in systemsnant human CYP2C10 and compared them with those containing CYP2C10 (5 pmol), NADPH–P450 reductase (10 pmol),

b5 (10 pmol), sodium cholate (100 nmol), and the lipid mixture (4 mg)of bufuralol 1*-hydroxylation by CYP1A1 and 2D6, the-in 50 mM potassium phosphate buffer (pH 7.4) containing theophylline 8-hydroxylation by CYP1A2, chlorzoxazoneNADPH-generating system and tolbutamide (500 nmol) or S-warfa-6-hydroxylation by CYP2E1, and testosterone 6b-hy- rin (20 nmol). The separation of hydroxylated tolbutamide and warfa-

droxylation and nifedipine oxidation by CYP3A4. Dif- rin metabolites by HPLC was carried out with a C18 5-mm analyticalcolumn.ferences in the effects of buffer concentrations and b5

Incubation mixture (final volume of 0.20 ml) for chlorzoxazone 6-and apo-b5 on the reconstituted catalytic activities byhydroxylation consisted of CYP2E1 (5 pmol), NADPH–P450 reduc-tase (10 pmol), b5 (10 pmol), sodium cholate (100 nmol), chlorzoxa-zone (100 nmol), and the lipid mixture (4 mg) in 100 mM potassiumphosphate buffer (pH 7.4) containing the NADPH-generating system.5 The coding sequence of the genes termed CYP2C9 and CYP2C10Product formation was determined by HPLC with a 4.6 1 150-mmdiffer in only two amino acids, 358 and 417 (2, 3). CYP2C9 has TyrNucleosil octylsilyl (C8) reverse-phase column (Chemco Scientific,and Gly, while CYP2C10 has Cys and Asp, respectively. TheOsaka).CYP2C10 sequence corresponds to the first cDNA clone we isolated

Standard incubation conditions (final volume of 0.25 ml) for testos-from this family (4). We reported that the two cDNAs now termedterone 6b-hydroxylation and nifedipine oxidation consisted ofCYP2C9 (MP-4) and 2C10 (MP-8) differed considerably in their 3 *-CYP3A4 (10 pmol), NADPH–P450 reductase (20 pmol), and b5 (20noncoding sequences, and oligonucleotide probes were used to iden-pmol) in 50 mM potassium Hepes buffer (pH 7.4) containing 0.5 mMtify both groups of sequences in the mRNA of a single liver samplecholate, 20 mg/ml of the lipid mixture, 3.0 mM GSH, 30 mM MgCl2, the(2). It is conceivable that the existence of the two cDNA clones (withinNADPH-generating system, and 100 mM testosterone or nifedipine.an expression library generated from a single individual human) isProduct formations were estimated by HPLC as described previouslyan artifact of the library construction or that the sequences recog-(7, 26).nized by the probes are parts of other genes in the CYP2C subfamily.

Nevertheless, CYP2C9 and CYP2C10 are treated here as the prod- Other assays. P450 and b5 were estimated spectrally by the meth-ods of Omura and Sato (27). Kinetic analysis of CYP2C10 reductionucts of individual genes.

AID ABB 0125 / 6b38$$$$61 05-09-97 23:59:53 arca

331RECONSTITUTION OF HUMAN P450 OXIDATIONS

TABLE I

Tolbutamide Methyl Hydroxylation and S-Warfarin 7-Hydroxylation by CYP2C10 in Reconstituted Systems

System

Protein added Tolbutamide methyl hydroxylation S-Warfarin 7-hydroxylation(10 pmol) Lipid Buffer MgCl2 (nmol/min/nmol P450) (pmol/min/nmol P450)

— Mix Kpi 0 2.6 { 0.2 12 { 1b5 Mix Kpi 0 6.3 { 0.7 55 { 5apo b5 Mix Kpi 0 5.5 { 0.8 50 { 5BSA Mix Kpi 0 16 { 3HSA Mix Kpi 0 18 { 2Cytochrome c Mix Kpi 0 15 { 2Hemoglobin Mix Kpi 0 18 { 1Catalase Mix Kpi 0 17 { 2b5 DLPC Kpi 0 3.4 { 0.6 31 { 3b5 Mix Hepes 0 5.1 { 0.9 42 { 5b5 Mix Kpi / 5.9 { 0.6 38 { 3b5 Mix Hepes / 6.2 { 0.5 53 { 6

Note. Incubation mixture (0.2 ml) contained CYP2C10 (0.025 mM), NADPH–P450 reductase (0.05 mM), b5 or other proteins (0.05 mM),either lipid mixture (20 mg/ml) plus 0.5 mM sodium cholate or DLPC (20 mg/ml), MgCl2 (30 mM), and tolbutamide (2.5 mM) or S-warfarin (0.10mM) in 50 mM potassium phosphate (Kpi) or potassium Hepes (pH 7.4) buffer. Results are presented as means of duplicate determinations {SD (range).

by reconstituted systems containing b5 and NADPH–P450 reductase methyl hydroxylation and S-warfarin 7-hydroxylationwas determined by the methods described previously (7). Protein activities by CYP2C10 were also dependent on the con-concentrations were estimated by the method of Lowry et al. (28). centrations of the buffer up to 50 mM, both activities

decreased when the buffer concentrations were furtherRESULTS increased. In contrast, CYP3A4-dependent testoster-

Reconstitution of tolbutamide methyl hydroxylation one 6b-hydroxylation activity was increased with in-and S-warfarin 7-hydroxylation activities by CYP2C10. creasing buffer concentrations of at least up to 200 mM,Both tolbutamide methyl hydroxylation and S-warfa- though the Hepes buffer was more useful for CYP3A4-rin 7-hydroxylation by CYP2C10 were found to be en- catalytic activities than the potassium phosphatehanced by apo-b5 as well as by native b5 in reconstituted buffer at the optimal CYP3A4-reconstituted conditions,systems (Table I). Other external proteins such as BSA, as described previously (7, 11).HSA, cytochrome c, hemoglobin, and catalase slightly Effects of b5 and apo-b5 on the reconstituted drugincreased the 7-hydroxylation of S-warfarin, but these oxidation activities by CYP1A1, 1A2, 2C10, 2D6, 2E1,effects were much less than those of b5. When the lipid and 3A4 were determined (Fig. 2). Neither b5 nor apo-mixture was replaced by DLPC, both catalytic activi- b5 had any effects on the bufuralol 1*-hydroxylation byties by CYP2C10 were decreased by about 40% (from CYP1A1 or 2D6 or the theophylline 8-hydroxylation bythe maximal activities). Replacement of potassium CYP1A2 in reconstituted systems. As reported pre-phosphate buffer by potassium Hepes buffer also de- viously (24), CYP3A4-dependent testosterone 6b-hy-creased the catalytic activities; addition of MgCl2 did droxylation and nifedipine oxidation activities were en-not cause enhancement of the activities. hanced by b5 or apo-b5. Interestingly, apo-b5 was as

effective as b5 in stimulating tolbutamide methyl hy-Effects of buffer concentration and b5 and apo-b5 ondroxylation and S-warfarin 7-hydroxylation catalyzedthe hydroxylation of prototypic substrates by CYP1A1,by CYP2C10. In contrast, chlorzoxazone 6-hydroxyla-1A2, 2C10, 2D6, 2E1, and 3A4 in reconstituted systems.tion was enhanced only by b5, but not by apo-b5, inBufuralol hydroxylation activities by CYP1A1 and 2D6reconstituted CYP2E1 systems.were decreased with increasing buffer concentrations

in reconstituted systems (Fig. 1). Activities of theophyl- Effects of synthetic phospholipid compositions on tes-tosterone 6b-hydroxylation by CYP3A4. It has pre-line 8-hydroxylation catalyzed by CYP1A2 and chlor-

zoxazone 6-hydroxylation by CYP2E1 were increased viously been shown that the phospholipid compositionaffects the activities of testosterone 6b-hydroxylationwith increasing the concentrations of potassium phos-

phate buffer in reconstituted systems and the maximal and nifedipine oxidations catalyzed by CYP3A4 in re-constituted systems (7, 11, 24, 29). We further exam-activities were obtained when the buffer concentration

wasÇ100 mM. Although reconstitutions of tolbutamide ined the roles of synthetic phospholipids and the lipid

AID ABB 0125 / 6b38$$$$61 05-09-97 23:59:53 arca

332 YAMAZAKI ET AL.

FIG. 1. Effects of buffer concentrations on bufuralol 1*-hydroxylation by CYP1A1 (A), theophylline 8-hydroxylation by CYP1A2 (B),tolbutamide methyl hydroxylation (C) and S-warfarin 7-hydroxylation (D) by CYP2C10, bufuralol 1*-hydroxylation by CYP2D6 (E), chlorzoxa-zone 6-hydroxylation by CYP2E1 (F), and testosterone 6b-hydroxylation by CYP3A4 (G) in reconstituted systems. The experimental condi-tions for each of the drug oxidations by P450 enzymes are described under Materials and Methods. In the case of testosterone 6b-hydroxyla-tion (G), both potassium phosphate (open circle) and potassium Hepes (closed circle) buffers were examined, and in other cases (A to F)only the effects of potassium phosphate buffer were determined. Data represent means of duplicate or triplicate determinations.

FIG. 2. Effects of concentrations of b5 (s) and apo-b5 (l) on bufuralol 1*-hydroxylation by CYP1A1 (A), theophylline 8-hydroxylation byCYP1A2 (B), tolbutamide methyl hydroxylation (C) and S-warfarin 7-hydroxylation (D) by CYP2C10, bufuralol 1*-hydroxylation by CYP2D6(E), chlorzoxazone 6-hydroxylation by CYP2E1 (F), and testosterone 6b-hydroxylation (G) and nifedipine oxidation (H) by CYP3A4 inreconstituted systems. The experimental conditions for each of the drug oxidations by P450 enzymes are described under Materials andMethods. Data represent means of duplicate or triplicate determinations.

AID ABB 0125 / 6b38$$0125 05-09-97 23:59:53 arca

333RECONSTITUTION OF HUMAN P450 OXIDATIONS

TABLE II Effects of addition of purified P450 proteins on testos-terone 6b-hydroxylation by CYP3A4 in reconstitutedEffects of Phospholipid Compositions on Testosterone 6b-

Hydroxylation by CYP3A4 in Reconstituted Systems systems. Effects of recombinant CYP1A1, 1A2, 2C10,2E1, and 2D6 on testosterone 6b-hydroxylation by

Testosterone 6b- CYP3A4 were determined (Table III). Previously weLipid hydroxylation found that although b5 and apo-b5 stimulate CYP3A4-

b5 composition Cholate Mg (nmol/min/nmol P450) catalyzed testosterone 6b-hydroxylation in reconstitu-ted systems, other proteins such as BSA, HSA, cyto-/ DLPC 0 / 1.8 { 1.2chrome c, hemoglobin, and catalase do not cause such/ DLPC / / 3.4 { 1.1

/ DLPC / 0 1.4 { 0.9 effects (24). We also determined whether other human0 DLPC / / 2.4 { 1.1 P450 proteins (e.g., CYP1A1, CYP1A2, CYP2C10,/ DLPC/DOPC / / 7.9 { 1.5 CYP2D6, and CYP2E1) can alter CYP3A4-dependent/ DLPC/PS / / 9.8 { 0.7

testosterone 6b-hydroxylation activities (Table III)./ DOPC/PS / / 8.2 { 0.5None of the enzymes examined here had appreciable/ DLPC/DOPC/PS / / 15.3 { 1.8

/ DLPC/DOPC/PS / 0 1.4 { 0.3 levels of testosterone 6b-hydroxylation activity, except0 DLPC/DOPC/PS / / 5.1 { 1.2 for CYP3A4 (Table III). However, CYP1A2 and 1A1,/ Human lipids / / 14.9 { 1.3 when added to the reconstituted systems containing/ Human lipids / 0 1.8 { 0.3

CYP3A4, caused increases in testosterone 6b-hydroxyl-0 Human lipids / / 4.7 { 0.8ation activities, by about two- to threefold. CYP2C10,

Note. Incubation mixtures (0.25 ml) contained CYP3A4 (10 pmol), 2D6, and 2E1 did not show such stimulatory effects.NADPH–P450 reductase (20 pmol), b5 (20 pmol), lipids (20 mg/ml), Seven CYP1A2 mutants, purified from membranescholate (0.5 mM), MgCl2 (30 mM), and GSH (3.0 mM) in 50 mM potas-sium Hepes buffer (pH 7.4). Results are means of duplicate or tripli-cate determinations { SD.

extracts isolated from human liver microsomes on tes-tosterone 6b-hydroxylation catalyzed by CYP3A4 (Ta-ble II). As reported previously, the lipid mixture con-taining three synthetic phospholipids was required tohave optimal catalytic activities of testosterone 6b-hy-droxylation activities by CYP3A4 and we found thatremoval of each of the DOPC or PS from the completesystems caused marked decreases in catalytic activi-ties. Replacement of the lipid mixture by human lipidsdid not cause decreases in the testosterone 6b-hydrox-ylation activities when both MgCl2 and cholate werepresent.

Kinetic analysis of reduction rates of CYP2C10.CYP2C10 was reduced only very slowly by NADPH–P450 reductase, under the conditions used in the recon- FIG. 3. Reduction of CYP2C10. The general experimental plan is

described elsewhere (17). One syringe of the stopped-flow spectropho-stitution systems, when no substrate was presenttometer contained an anaerobic solution of CYP2C10 (0.35 mM),(Figs. 3A and 3B). The rate was õ10 min01 and wasNADPH–P450 reductase (0.70 mM), the standard phospholipid mix-unaffected by the addition of b5 or apo-b5. When tolbu-ture (20 mg/ml), sodium cholate (0.5 mM), and 50 mM potassium

tamide (2.5 mM) was present, a rapid burst of reduction phosphate buffer (pH 7.4) under CO. The other syringe containedoccurred, with a rate of Ç250 min01, followed by a NADPH (0.5 mM) in 50 mM potassium phosphate buffer (pH 7.4)

under CO. Both syringes also contained a mixture of glucose oxidase,slower second phase (9 min01) (Fig. 3C). The same bi-catalase, and glucose to facilitate anaerobiosis. Reduction was moni-phasic pattern was seen when b5 (Fig. 3D) or apo-b5tored at 450 nm by formation of the Fe2/

rCO complex. (A) Mixingwas added; the two rates of the biexponential reduction of components as indicated above. (B) Mixing as in A, except thatwere 160 and 6 min01. When the substrate S-warfarin the P450 syringe contained 0.35 mM b5. (C) Mixing as in A, exceptwas included, similar rates were observed in the ab- that 5 mM sodium tolbutamide was included in the NADPH syringe

(final concentration 2.5 mM). (D) As in C, except that 0.35 mM b5 wassence or presence of b5 (as with tolbutamide). Thus,included in the CYP2C10 syringe. (In D, 4000 points were collectedthe presence of substrate appears necessary to achieveinstead of 400.) The residual analysis for a first-order plot is shownrapid reduction (of part) of CYP2C10. b5 did not stimu- at the bottom of C and D. In A and B, the reduction rate is õ10

late reduction (in the presence of the substrate tolbuta- min01. In C and D, the fit (solid) is to a curve with rates of 250 and9 min01. In D, the fit is to rates of 160 and 6 min01.mide).

AID ABB 0125 / 6b38$$$$61 05-09-97 23:59:53 arca

334 YAMAZAKI ET AL.

TABLE III an increase in S-warfarin 7-hydroxylation activities byCYP2C10.Effects of P450 Proteins on Testosterone 6b-Hydroxylation

by CYP3A4 in Reconstituted Systems Effects of addition of several proteins on chlorzoxa-zone 6-hydroxylation by CYP2E1 in reconstituted sys-

Testosterone 6b- tems. Several proteins including BSA, HSA, cyto-CYP3A4 hydroxylation chrome c, hemoglobin, and catalase and human P450(pmol) Protein (pmol) (nmol/min/nmol P450)

proteins (CYP1A2, 2C10, 2D6, and 3A4) were added tothe reconstituted systems containing CYP2E1, and the10 — 4.4 { 0.3

10 b5 10 14.6 { 0.8 effects of these proteins on chlorzoxazone 6-hydroxyla-10 Apo b5 10 13.8 { 1.9 tion activities were determined (Table VII). These pro-— CYP1A1 20 õ0.02 teins did not show any effects on the chlorzoxazone 6-10 CYP1A1 20 9.0 { 0.8

hydroxylation activity of CYP2E1, and b5 was the only— CYP1A2 10 õ0.02protein that increased the chlorzoxazone 6-hydroxyla-10 CYP1A2 10 7.3 { 0.4

10 CYP1A2 20 11.5 { 0.9 tion by CYP2E1.— CYP2C10 10 õ0.0210 CYP2C10 10 5.2 { 0.7

DISCUSSION10 CYP2C10 20 5.7 { 0.9— CYP2D6 20 õ0.02 It has been reported that several human liver P45010 CYP2D6 20 4.4 { 0.7 enzymes require b5 for maximal catalytic activities in— CYP2E1 20 õ0.0210 CYP2E1 20 5.7 { 0.8

TABLE IVNote. Incubation mixtures (0.25 ml) contained CYP3A4 (10 pmol),NADPH–P450 reductase (20 pmol), b5 or other P450 enzymes as Effects of CYP1A2 Mutants on Testosteroneindicated, lipid mixture (20 mg/ml), cholate (0.5 mM), MgCl2 (30 mM), 6b-Hydroxylation by Reconstitutedand GSH (3.0 mM) in 50 mM potassium Hepes buffer (pH 7.4). Results Systems Containing CYP3A4are presented as means of duplicate or triplicate determinations{ SD.

Testosterone 6b-CYP3A4 hydroxylation(pmol) Addition (pmol) (nmol/min/nmol P450)

of E. coli into which modified CYP1A2 cDNA sequences10 — 4.0 { 0.5had been introduced (30, 31), were analyzed for their10 b5 10 13.1 { 1.0effects on testosterone 6b-hydroxylation catalyzed by — CYP1A2 20 õ0.02

CYP3A4 (Table IV). Characteristics of these mutated 10 CYP1A2 20 8.3 { 1.310 CYP1A2a 20 2.4 { 0.1forms of CYP1A2 have been reported previously (30,10 CYP1A2b 20 3.2 { 0.631). Among the CYP1A2 mutants examined, CYP1A2/10 CYP1A2c 20 4.6 { 0.5M1D increased testosterone 6b-hydroxylation by10 CYP1A2/M1 20 5.2 { 1.2CYP3A4 by about twofold. Other mutant forms showed — CYP1A2/M1D 20 õ0.02

no effects or slightly enhanced the hydroxylation activi- 10 CYP1A2/M1D 20 8.9 { 0.310 CYP1A2/M2 20 4.0 { 0.3ties (though less so than CYP1A2 and CYP1A2/M1D).10 CYP1A2/M2D 20 4.4 { 0.5The increases in testosterone 6b-hydroxylation activ-

ities by CYP3A4 were also determined when liverNote. Incubation mixtures (0.25 ml) contained CYP3A4 (10 pmol),CYP1A2 purified from human, rat, or rabbit was added NADPH–P450 reductase (20 pmol), lipid mixture (20 mg/ml), cholate

to the incubation mixtures (Table V). (0.5 mM), MgCl2 (30 mM), and GSH (3.0 mM) in 50 mM potassiumHepes buffer (pH 7.4) with or without b5 or CYP1A2 or mutantEffects of addition of purified P450 proteins on tolbu-CYP1A2 (20 pmol). Results are presented as means of duplicate ortamide methyl hydroxylation and S-warfarin 7-hy-triplicate determinations { SD. The nomenclature used for the

droxylation by CYP2C10 in reconstituted systems. CYP1A2 proteins is as follows. CYP1A2 is the recombinant humanCYP1A2, 2E1, and 3A4 were added to the reconstituted protein expressed in E. coli (plasmid 1024, with residues 3–13 of the

native protein removed and other changes at codons 18–20) (14).systems containing CYP2C10 and the effects of theseRandom mutants CYP1A2a, 1A2b, and 1A2c involve changes in co-external P450 proteins on tolbutamide methyl hydrox-dons 3–5 of plasmid 1024 (31). CYP1A2/M1 is a modified version ofylation and S-warfarin 7-hydroxylation activities were CYP1A2 (plasmid 1024) containing a thrombin-sensitive site; this

determined (Table VI). None of these three human has been designated 1A2//12 (30) and mutant 1 (31). In CYP1A2/P450 enzymes increased tolbutamide methyl hydroxyl- M1D the N-terminal peptide has been removed by treatment with

thrombin (and repurification); this was designated 1A2//12-31 pre-ation by CYP2C10 in reconstituted systems. Althoughviously (30). CYP1A2/M2 is thrombin-sensitive site-containing mu-CYP1A2 had considerable S-warfarin 7-hydroxylationtant 2 (31) and D indicates cleavage by thrombin. The sequences ofactivity, the activities were not enhanced synergisti- all mutants were confirmed by Edman degradation (31). For further

cally when CYP2C10 and CYP1A2 were present in re- information on the retention of the N-formyl Met group in the variousrecombinant CYP1A2 proteins, see Ref. 31.constituted systems. CYP2E1 and 3A4 did not cause

AID ABB 0125 / 6b38$$$$61 05-09-97 23:59:53 arca

335RECONSTITUTION OF HUMAN P450 OXIDATIONS

TABLE V

Effects of CYP1A2 on CYP3A4-Dependent Testosterone 6b-Hydroxylation in Reconstituted Systems

CYP3A4 b5 Testosterone 6b-hydroxylation(pmol) (pmol) CYP1A2 (pmol) (nmol/min/nmol P450)

10 — — 4.7 { 0.210 10 — 11.4 { 0.2— — Escherichia coli 1A2 20 õ0.0210 — Escherichia coli 1A2 20 8.1 { 0.910 10 Escherichia coli 1A2 20 14.0 { 1.4— — Human 1A2 20 õ0.0210 — Human 1A2 20 7.5 { 1.110 10 Human 1A2 20 12.6 { 1.8— — Rat 1A2 20 õ0.0210 — Rat 1A2 20 6.5 { 0.710 10 Rat 1A2 20 10.2 { 0.7— — Rabbit 1A2 20 õ0.0210 — Rabbit 1A2 20 8.2 { 0.510 10 Rabbit 1A2 20 16.7 { 1.3

Note. Incubation mixtures (0.25 ml) contained CYP3A4 (10 pmol), NADPH–P450 reductase (20 pmol), b5 (10 pmol) or one of severalCYP1A2 preparations (20 pmol) when indicated, lipid mixture (20 mg/ml), cholate (0.5 mM), MgCl2 (30 mM), and GSH (3.0 mM) in 50 mM

potassium Hepes buffer (pH 7.4). Results are presented as means of duplicate or triplicate determinations { SD. Escherichia coli 1A2indicates human CYP1A2 expressed in E. coli with an N-terminal truncation as described elsewhere, using plasmid 1024 (14). Human 1A2refers to the protein isolated from human liver microsomes.

reconstitution systems, e.g., CYP3A4, CYP3A5, In this study, we found that CYP2C10-dependent tol-butamide methyl hydroxylation and S-warfarin 7-hy-CYP2E1, CYP2C19, and CYP2C9 (11, 12, 32–35).

CYP17 has also been shown to require b5 for maximal droxylation activities were also stimulated by apo-b5 aswell as by b5 in reconstituted systems. The mechanismscatalytic activities (36). However, the mechanisms for

such stimulation of P450-dependent substrate oxida- underlying stimulation of CYP2C10- and CYP3A4-de-pendent substrate oxidations by b5 might be different,tions by b5 have been shown to differ depending upon

the P450 enzymes used (11, 12). For example, apo-b5 because rapid reduction of P450 is stimulated by b5 orapo-b5 in systems containing CYP3A4 but notas well as native b5 has been shown to be effective in

stimulation of testosterone 6b-hydroxylation and nifed- CYP2C10. In addition, differences exist in optimal re-constitution conditions for testosterone 6b-hydroxyla-ipine oxidation by CYP3A4, while such apo-b5-depen-

dent stimulation could not be seen when chlorzoxazone tion by CYP3A4 and tolbutamide methyl hydroxylationand S-warfarin 7-hydroxylation by CYP2C10. For ex-6-hydroxylation activities were determined in reconsti-

tuted systems containing CYP2E1 (11, 12, 24). ample, the former enzyme requires MgCl2 (for the re-

TABLE VI

Effects of b5 and Several P450 Enzymes on Tolbutamide Methyl Hydroxylation and S-Warfarin 7-Hydroxylationby Reconstituted Systems Containing CYP2C10

CYP2C10 Tolbutamide methyl hydroxylation S-Warfarin 7-hydroxylation(pmol) Protein (pmol) (nmol/min/nmol P450) (pmol/min/nmol P450)

5 — 2.7 { 0.3 11 { 25 b5 5 6.3 { 0.6 53 { 45 Apo b5 5 5.9 { 0.7 47 { 45 CYP1A2 5 3.1 { 0.4 13 { 20 CYP1A2 5 õ0.1 4 { 15 CYP2E1 5 2.7 { 0.3 11 { 30 CYP2E1 5 õ0.1 0.4 { 0.15 CYP3A4 5 3.0 { 0.5 12 { 20 CYP3A4 5 õ0.1 0.7 { 0.6

Note. Experimental details are described under Materials and Methods. Values represent means of duplicate determinations{ SD (range).

AID ABB 0125 / 6b38$$$$61 05-09-97 23:59:53 arca

336 YAMAZAKI ET AL.

TABLE VII CYP2E1-dependent substrate oxidations were not af-fected by the addition of other P450 proteins, althoughEffects of b5 and Other Proteins on Chlorzoxazone 6-

Hydroxylation by CYP2E1 in Reconstituted Systems these two P450 enzymes both require b5 for optimalrates. At this point it is not clear whether the same

Chlorzoxazone 6- mechanism is involved in the stimulation of CYP3A4CYP2E1 Proteins hydroxylation activity by b5 and CYP1A2. Although both can stimu-(pmol) added (pmol) (nmol/min/nmol P450) late activity, the effects appeared to be additive in the

study reported in Table V. Also, we have not previously10 — 1.7 { 0.1found CYP1A2 activity to be stimulated by b5 (14). We10 b5 20 9.0 { 1.1

10 BSA 20 2.0 { 0.3 have not examined the effect of CYP3A4 on the cata-10 HSA 20 1.6 { 0.1 lytic activity of CYP1A2. Whether these interactions10 Cytochrome c 20 1.8 { 0.3 result in enhanced catalytic activity in a cellular envi-10 Hemoglobin 20 1.8 { 0.2

ronment is also unknown.10 Catalase 20 1.9 { 0.1Chlorzoxazone 6-hydroxylation catalyzed by a recon-— CYP1A2 20 õ0.05

10 CYP1A2 20 2.0 { 0.1 stituted system containing CYP2E1 was stimulated by— CYP2C10 20 õ0.05 native b5 (but not by apo-b5) in a concentration-depen-10 CYP2C10 20 1.7 { 0.1 dent manner. The increased electron flow from— CYP2D6 20 õ0.05

NADPH–P450 reductase to CYP2E1 via b5 seems to10 CYP2D6 20 1.4 { 0.1be a cause of such stimulation of CYP2E1-dependent— CYP3A4 20 õ0.05

10 CYP3A4 20 1.9 { 0.1 chlorzoxazone 6-hydroxylation activities, and themechanisms of b5 stimulation might be different in the

Note. The standard reaction mixtures (0.20 ml) described under cases of reconstituted systems containing CYP2C10Materials and Methods were used for the assay of chlorzoxazone 6-and CYP3A4 (11, 12). In this connection, it is interest-hydroxylation, except that several proteins were included as indi-ing to note the results of Patten et al. (33) who showedcated. Data are presented as means of duplicate and triplicate deter-

minations { SD. that native b5, but not apo-b5, is required for the N-demethylation and denitrosation of N-nitrosodimethy-lamine by rat CYP2E1 in reconstituted monooxygenasesystems. N-Nitrosodimethylamine catalytic activitiesconstitution of testosterone 6b-hydroxylation andwere restored when hemin was added with apo-b5 tonifedipine oxidation activities) while the latter doesthe reconstituted systems (33).not. In addition, 50 mM Hepes buffer was required for

Among recombinant human P450 enzymes exam-optimal reconstitution of CYP3A4-dependent reac-ined, CYP1A1, 1A2, and 2D6 did not show any re-tions, while a 50 mM potassium phosphate buffer wasquirements of b5 and apo-b5 for maximal catalyticbetter for the reconstitution of CYP2C10-dependent ac-activities. These phenomena have already been re-tivities. However, it remains unclear whether or notported previously in reconstituted systems con-there are different mechanisms in the stimulation oftaining human, rat, and rabbit liver P450 enzymessubstrate oxidations by b5 in CYP3A4- and CYP2C10-(14, 16, 20, 22, 38–40).reconstituted systems.

Ionic strength has previously been shown to affectCYP1A2 and CYP1A1, as well as b5, were found incatalytic activities in P450-catalyzed drug oxidation re-this study to cause an increase in CYP3A4-dependentactions, probably due to the increased electron flowtestosterone 6b-hydroxylation activities. Other P450from NADPH–P450 reductase to P450 (41, 42). How-enzymes examined here did not have such enhancingever, potassium phosphate buffer decreased bufuraloleffects on testosterone 6b-hydroxylation by CYP3A4.1*-hydroxylation catalyzed by CYP1A1 and CYP2D6In addition, we also found that when recombinantand tolbutamide methyl hydroxylation and S-warfarinCYP1A2 was replaced by human, rat, and rabbit7-hydroxylation catalyzed by CYP2C10 in reconstitu-CYP1A2 or recombinant CYP1A2/M1D mutant, theted systems (Fig. 1). Optimal concentrations of potas-testosterone 6b-hydroxylation activity of CYP3A4 wassium phosphate were found to be 25 mM for CYP2D6-increased. Although the detailed mechanisms are notcatalyzed reactions and 50 mM for CYP2C10-dependentknown, this is the first report to show stimulation ofactivities. In contrast, CYP3A4 testosterone 6b-hy-P450 activities by other P450 proteins in reconstituteddroxylation activity was increased by raising the buffersystems (37).6 In contrast to CYP3A4, CYP2C10- andconcentration, to at least 200 mM (which is not a physi-ological concentration in the cell), although the potas-

6 Cawley et al. (37) reported that rabbit CYP1A2 yielded an in- sium Hepes buffer was more effective than the phos-crease in the benzphetamine N-demethylation activity of (rabbit) phate buffer.CYP2B4. However, the stimulation (23%) was seen only at a single

It has been reported that the phospholipid environ-CYP1A2 concentration (Fig. 1 of that report) and was not observedfor another reaction, pentoxyresorufin O-dealkylation. ment is very important to reconstitute the substrate

AID ABB 0125 / 6b38$$$$61 05-09-97 23:59:53 arca

337RECONSTITUTION OF HUMAN P450 OXIDATIONS

17. Gillam, E. M. J., Guo, Z., and Guengerich, F. P. (1994) Arch. Bio-oxidations by purified rat and human CYP3A enzymeschem. Biophys. 312, 59–66.(7, 11, 24, 25, 29, 43). Recently, we found that the 7-

18. Gillam, E., Baba, T., Kim, B. R., Ohmori, S., and Guengerich,ethoxycoumarin O-deethylation activity of human re-F. P. (1993) Arch. Biochem. Biophys. 305, 123–131.

combinant CYP2E1 was increased when DLPC was re-19. Guengerich, F. P., Gillam, E. M. J., and Shimada, T. (1996) Crit.

placed by a phospholipid mixture in reconstituted sys- Rev. Toxicol. 26, 551–583.tems (12). The present results further showed that this 20. Kuwahara, S., Harada, N., Yoshioka, H., Miyata, T., and Omura,phospholipid mixture was also useful for the reconsti- T. (1984) J. Biochem. (Tokyo) 95, 703–714.tution of tolbutamide methyl hydroxylation and S-war- 21. Shimada, T., Imai, Y., and Sato, R. (1981) Chem. Biol. Interact.

38, 29–44.farin 7-hydroxylation by CYP2C10, and that the lipid22. Distlerath, L. M., Reilly, P., Martin, M. V., Davis, G. G., Wilkin-extracts from human liver microsomes are useful for

son, G. R., and Guengerich, F. P. (1985) J. Biol. Chem. 260,the reconstitution of drug oxidations by P450 enzymes.9057–9067.These results suggest that these substrate oxidations

23. Guengerich, F. P., Wang, P., and Mason, P. S. (1981) Biochemis-of P450 enzymes that are stimulated by b5 appear to try 20, 2379–2385.require some specific lipid environment to optimize pro- 24. Yamazaki, H., Johnson, W. W., Ueng, Y.-F., Shimada, T., andtein–protein interactions. Guengerich, F. P. (1996) J. Biol. Chem. 271, 27438–27444.

25. Eberhart, D. C., and Parkinson, A. (1991) Arch. Biochem. Bio-phys. 291, 231–240.REFERENCES

26. Yamazaki, H., Mimura, M., Sugahara, C., and Shimada, T.1. Guengerich, F. P., and Shimada, T. (1991) Chem. Res. Toxicol. (1994) Biochem. Pharmacol. 48, 1524–1527.

4, 391–407. 27. Omura, T., and Sato, R. (1964) J. Biol. Chem. 239, 2370–2378.2. Ged, C., Umbenhauer, D. R., Bellew, T. M., Bork, R. W., Srivas- 28. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

tava, P. K., Shinriki, N., Lloyd, R. S., and Guengerich, F. P. (1951) J. Biol. Chem. 193, 265–275.(1988) Biochemistry 27, 6929–6940. 29. Imaoka, S., Imai, Y., Shimada, T., and Funae, Y. (1992) Biochem-

3. Srivastava, P. K., Yun, C. H., Beaune, P. H., Ged, C., and Guen- istry 31, 6063–6069.gerich, F. P. (1991) Mol. Pharmacol. 40, 69–79. 30. Dong, M.-S., Yamazaki, H., Guo, Z., and Guengerich, F. P. (1996)

Arch. Biochem. Biophys. 327, 11–19.4. Umbenhauer, D. R., Martin, M. V., Lloyd, R. S., and Guengerich,F. P. (1987) Biochemistry 26, 1094–1099. 31. Dong, M.-S., Bell, L. C., Guo, Z., Phillips, D. R., Blair, I. A., and

Guengerich, F. P. (1996) Biochemistry 35, 10031–10040.5. Shimada, T., Yamazaki, H., Mimura, M., Inui, Y., and Guenger-ich, F. P. (1994) J. Pharmacol. Exp. Ther. 270, 414–423. 32. Shimada, T., Misono, K. S., and Guengerich, F. P. (1986) J. Biol.

Chem. 261, 909–921.6. Gonzalez, F. J. and Korzekwa, K. R. (1995) Annu. Rev. Pharma-33. Patten, C. J., Ning, S. M., Lu, A. Y. H., and Yang, C. S. (1986)col. Toxicol. 35, 369–390.

Arch. Biochem. Biophys. 251, 629–638.7. Yamazaki, H., Ueng, Y.-F., Shimada, T., and Guengerich, F. P.34. Lu, A. Y. H. and West, S. B. (1980) Pharmacol. Rev. 31, 277–(1995) Biochemistry 34, 8380–8389.

295.8. Brian, W. R., Sari, M. A., Iwasaki, M., Shimada, T., Kaminsky,35. Richardson, T. H., Jung, F., Griffin, K. J., Wester, M., Raucy,L. S., and Guengerich, F. P. (1990) Biochemistry 29, 11280–

J. L., Kemper, B., Bornheim, L. M., Hassett, C., Omiecinski,11292.C. J., and Johnson, E. F. (1995) Arch. Biochem. Biophys. 323,9. Shet, M. S., Fisher, C. W., Holmans, P. L., and Estabrook, R. W. 87–96.(1993) Proc. Natl. Acad. Sci. USA 90, 11748–11752.

36. Katagiri, M., Kagawa, N., and Waterman, M. R. (1995) Arch.10. Shet, M. S., Faulkner, K. M., Holmans, P. L., Fisher, C. W., and Biochem. Biophys. 317, 343–347.

Estabrook, R. W. (1995) Arch. Biochem. Biophys. 318, 314–321.37. Cawley, G. F., Batie, C. J., and Backes, W. L. (1995) Biochemis-

11. Yamazaki, H., Nakano, M., Imai, Y., Ueng, Y.-F., Guengerich, try 34, 1244–1247.F. P., and Shimada, T. (1996) Arch. Biochem. Biophys. 325, 174– 38. Ryan, D. E., and Levin, W. (1990) Pharmacol. Ther. 45, 153–182. 239.

12. Yamazaki, H., Nakano, M., Gillam, E. M. J., Bell, L. C., Guenger- 39. Buters, J. T. M., Shou, M., Hardwick, J. P., Korzekwa, K. R., andich, F. P., and Shimada, T. (1996) Biochem. Pharmacol. 52, 301– Gonzalez, F. J. (1995) Drug Metab. Dispos. 23, 696–701.309.

40. Miners, J. O., Wing, L. M. H., and Birkett, D. J. (1985) Aust.13. Guo, Z., Gillam, E., Ohmori, S., Tukey, R. H., and Guengerich, N. Z. J. Med. 15, 348–349.

F. P. (1994) Arch. Biochem. Biophys. 312, 436–446. 41. Schenkman, J. B., Voznesensky, A. I., and Jansson, I. (1994)14. Sandhu, P., Guo, Z., Baba, T., Martin, M. V., Tukey, R. H., and Arch. Biochem. Biophys. 314, 234–241.

Guengerich, F. P. (1994) Arch. Biochem. Biophys. 309, 168–177. 42. Voznesensky, A. I., and Schenkman, J. B. (1994) J. Biol. Chem.15. Sandhu, P., Baba, T., and Guengerich, F. P. (1993) Arch. Bio- 269, 15724–15731.

chem. Biophys. 306, 443–450. 43. Ingelman-Sundberg, M., Hagbjork, A.-L., Ueng, Y.-F., Yamazaki,H., and Guengerich, F. P. (1996) Biochem. Biophys. Res. Com-16. Gillam, E. M. J., Guo, Z., Martin, M. V., Jenkins, C. M., and

Guengerich, F. P. (1995) Arch. Biochem. Biophys. 319, 540–550. mun. 221, 318–322.

AID ABB 0125 / 6b38$$$$61 05-09-97 23:59:53 arca