Download pdf - Expression, Purification, and Characterization of a Catalytically Active Human Cytochrome P450 1A2:Rat NADPH-Cytochrome P450 Reductase Fusion Protein

PROTEIN EXPRESSION AND PURIFICATION 9, 346–354 (1997)ARTICLE NO. PT970721

Expression, Purification, and Characterization of aCatalytically Active Human Cytochrome P450 1A2:RatNADPH-Cytochrome P450 Reductase Fusion Protein

Asit Parikh and F. Peter Guengerich1

Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine,Nashville, Tennessee 37232

Received October 1, 1996, and in revised form December 20, 1996

pharamaceuticals, carcinogens, pesticides) and endoge-An enzymatically active human cytochrome P450 nous compounds (e.g., steroids, eicosanoids) undergo

(P450) 1A2:rat NADPH-P450 reductase fusion protein Phase I (oxidative) hepatic metabolism. Composed ofwas purified and partially characterized following heter- several hundred identified members, this enzyme su-ologous expression in Escherichia coli. A cDNA was engi- perfamily has attracted considerable attention for itsneered to include the coding sequence for human P450 role in essential biochemical pathways (1–4), normal1A2 at its 5* end (up to but not including the stop codon) and aberrant drug metabolism (5,6), and carcinogene-fused in-frame to the coding sequence for a truncated sis (7–9).(soluble) rat NADPH-P450 reductase at its 3* end via an Detailed structure–function analysis of the cyto-oligonucleotide sequence encoding the hydrophilic di- chromes P450, while of considerable interest to re-peptide Ser–Thr. This fusion plasmid was expressed in searchers, has been impeded by the difficulty of purify-E. coli and the recombinant protein was purified from ing these enzymes from tissue in quantities sufficientthe detergent-solubilized membrane fraction via sequen- for study. While a number of P450 enzymes have beential DEAE, ADP–agarose, and hydroxylapatite chromato- isolated from mammalian microsomes, purificationsgraphies. The purified protein has the spectral charac-

have been complicated by limited yield and, at times,teristics of human P450 1A2 and cytochrome c reductionlow activity (10). Today, the field has progressed to aactivity comparable to rabbit NADPH-P450 reductase.stage where heterologous expression systems are re-The fusion protein catalyzed 7-ethoxyresorufin O-deethy-quired to answer a number of pertinent biochemicallation and phenacetin O-deethylation to appreciable lev-questions.els in the presence of NADPH and phospholipid. While

Yeast expression systems have been explored andthese activities were comparable to those of other suchhave certain advantages including the capacity for ex-P450:NADPH-P450 reductase fusion proteins, they werepression of recombinant mutants and the presence anlower than those of the system reconstituted from its

individual hemoprotein and flavoprotein components. endogenous NADPH-P450 reductase which functionsNevertheless, the production of a functional, catalyti- as an electron donor for heterologously expressed P450cally self-sufficient monooxygenase in E. coli enhances (11). Notable disadvantages, however, include lowthe prospect of using bacterial systems for production yields of recombinant protein, slow growth of organ-and characterization of human P450 drug metabolites as isms, and moderate cost.well as for biodegradation of chemicals in the environ- Although not totally devoid of drawbacks, bacterialment. q 1997 Academic Press expression systems have been employed successfully

for the expression of a number of microsomal P450 en-zymes. In many instances, they have provided an idealThe cytochromes P450 (P450)2 constitute the major

pathway by which a great number of xenobiotics (e.g.,

line; DTT, DL-dithiothreitol; FMN, flavin mononucleotide; HPLC,1 To whom correspondence and reprint requests should be ad-dressed. high-performance liquid chromatography; IPTG, isopropyl b-D-thio-

galactopyranoside; P450, cytochrome P450; PMSF, phenylmethylsul-2 Abbreviations used: aNF, a-napthoflavone (7,8-benzoflavone);BCA, bicinchoninic acid; BSA, bovine serum albumin; d-ALA, d- fonyl fluoride; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide

gel electrophoresis; TLC, thin-layer chromatography.aminolevulinic acid; DLPC, L-a-dilauroyl-sn-glycero-3-phosphocho-

346 1046-5928/97 $25.00Copyright q 1997 by Academic Press

All rights of reproduction in any form reserved.

AID PEP 0721 / 6q0f$$$121 03-11-97 20:12:06 pepa AP: PEP

P450 1A2:NADPH-P450 REDUCTASE FUSION PROTEIN 347

combination of low cost, facility of handling, and high [ring-3H]Phenacetin was a gift from Dr. F. F. Kadlu-bar (National Center for Toxicological Research, Jef-enzyme yields (12). Recreating a functional microbial

system capable of catalyzing P450-mediated reactions ferson, AR).presents the added possibility of employing genetic se- Construction of a human P450 1A2:rat NADPH-P450lection (12). Important considerations include modi- reductase fusion plasmid The expression vector pCWfying the N-terminus of the protein for optimal expres- (Ori/) was a gift from Professor F. W. Dahlquist (Uni-sion and providing a source of electrons for the heterol- versity of Oregon, Eugene). The P450 1A2 cDNA, a giftogously expressed enzyme. This last issue can be from Professor R. H. Tukey (University of California,addressed in (at least) three ways: (i) reliance on bacte- San Diego), was modified for expression in E. coli asrial flavoproteins, (ii) coexpression of mammalian described in an earlier publication (33).NADPH-P450 reductase, and (iii) expression of fusion The fusion plasmid was generated by first digestingproteins composed of a P450 domain and a truncated the modified P450 1A2 expression plasmid with SalI(soluble) reductase domain. The first approach has and HindIII to remove the last 85 bp of the codingproven limited (13,14) and the second has, to date, been region (Fig. 1). This region of the clone was replacedcomplicated by the inherent difficulty of constructing/ with six overlapping synthetic oligonucleotide linkers,maintaining two-plasmid or bicistronic systems. The which restored the 3 * end of the coding region andfinal approach has been employed successfully by oth- obliterated the internal SalI site and stop codon. Thisers and ourselves for the expression of P450 and cloning step also created a new SalI site, which encodesNADPH-P450 reductase as part of a single polypeptide the Ser–Thr dipeptide linking the P450 and reductasechain (15–18). domains, and a HindIII restriction site following the

The fusion proteins P450 102 (BM-3, Bacillus mega- last amino acid of P450 1A2. This new P450 1A2 con-terium) (19) and the mammalian P450-like nitric oxide struct, designated 1A2 SalI0, was digested with NdeIsynthase enzymes (20,21), which occur naturally, con- and SalI to release the entire coding region. This frag-tain both heme and flavin domains. Electron transfer ment was ligated in place of P450 3A4 into a humanseems to function effectively in these systems. Ohkawa P450 3A4:rat NADPH-P450 reductase plasmid gener-and associates were the first to express an artificial ated previously in our laboratory using the same re-fusion protein, patterned after the naturally occuring striction sites (18). The nucleotide sequence of the re-precedents, of rat P450 1A1 and yeast NADPH-P450 gion spanning the oligonucleotide linkers up to andreductase (22). They also expressed the bovine steroido- including the junction of the P450 and reductase do-genic P450s 17A and 21A in the same manner (23,24). mains was confirmed by dideoxy sequencing.The heterologous expression of a number of fusions of

Expression of fusion plasmid in E. coli. A singlehuman P450s with the rat NADPH-P450 reductase hasampicillin-resistant colony of E. coli transformed withbeen described by Estabrook and co-workers (15–17)the fusion plasmid expression construct describedas well as by our own laboratory (18).above was grown overnight to saturation at 377C inP450 1A2 has long been of interest to researchers forLuria–Bertani (LB) medium containing 100 mg ampi-the diversity of reactions it catalyzes. In addition tocillin ml01. A 10-ml aliquot was used to inoculate eachits well-documented role in caffeine and theophyllineliter of Terrific Broth (TB) containing 0.20% bactopep-metabolism (25,26), it activates a number of arylam-tone (w/v). TB was supplemented with 100 mg ampicil-ines and heterocyclic amines (25), and is inducible bylin ml01, 1.0 mM thiamine, trace elements (34), 0.15 mMcompounds found in cigarette smoke (27), charbroiledFeCl3, 0.5 mM d-ALA, and 0.10 mg riboflavin ml01 (35).meat, and cruciferous vegetables (28). A primarily he-Cultures were grown at 287C with shaking at 125 rpmpatic enzyme, human P450 1A2 appears to be subjectfor 48 h in an Innova shaking incubator (New Bruns-to genetic polymorphism (26). Its relevance to cancerwick, NJ). After 4 and 24 h, protein expression wasrisk has been the subject of much study (29).induced with 1.0 mM IPTG. Membranes were preparedOur understanding of this important P450 enzymefrom bacteria as previously described (33,34,36). Yieldhas been furthered by experiments involving its heter-at each stage was quantitated by Fe2/ vs Fe2/–CO dif-ologous expression in mammalian cells (30,31) and inference spectra (37).Escherichia coli (32,33). To obtain a catalytically self-

Purification of human P450 1A2:rat NADPH-P450sufficient P450 1A2-like protein, this enzyme was cho-reductase fusion protein. Membranes were diluted tosen as a target for expression as part of a fusion proteina protein concentration of approximately 3 mg ml01 inwith rat NADPH-P450 reductase.10 mM potassium phosphate buffer (pH 7.5) containing20% glycerol (v/v), 0.5 mM EDTA, 0.10 mM DTT, 0.5

MATERIALS AND METHODS mM PMSF, 1.0% Emulgen 911 (w/v), 0.5% sodium cho-late (w/v), and 30 mM aNF. The solubilized membranesChemicals. 7-Ethoxyresorufin (ethylresorufin) was

purchased from Sigma Chemical Co. (St. Louis, MO). were stirred for 1 h and centrifuged at 105g for 45 min.


PARIKH AND GUENGERICH348

FIG. 1. Construction of the pCW*/human P450 1A2:rat NADPH-P450 reductase fusion plasmid. Overlapping oligonucleotides encodingthe 3 * end of P450 1A2 (hatched lines) were designed to eliminate the internal SalI site by changing the last nucleotide of the recognitionsequence. The new SalI site encodes the Ser–Thr dipeptide, which bridges the P450 domain and the truncated (soluble) reductase domainin the final construct.

The clarified supernatant was applied to a 2.5 1 9-cm mM PMSF. The column was washed three times with50 ml of equilibration buffer containing 25, 300, andDE-52 column (Whatman, Fairfield, NJ) equilibrated

with 10 mM potassium phosphate buffer (pH 7.5) con- 25 mM potassium phosphate, respectively. The proteinwas then eluted with 500 ml of equilibration buffertaining 20% glycerol (v/v), 0.5 mM EDTA, 0.10 mM DTT,

1.0 mM FMN, 0.20% Emulgen 911 (w/v), 0.5% sodium using a linear gradient of 2 *-AMP (0–10 mM). Homoge-nous fractions, as judged by SDS–PAGE, were appliedcholate (w/v), and 30 mM aNF. The fusion protein bound

to the resin and was eluted with a 1-liter linear gradi- to a 1.0 1 1.0-cm Bio-Gel HTP hydroxylapatite column(Bio-Rad, Hercules, CA). Following extensive washingent of NaCl (0–0.5 M) in the same buffer. Fractions in

which fusion protein was detectable by SDS–PAGE to remove aNF and detergent, the protein was elutedwith 500 mM potassium phosphate buffer (pH 7.4) con-were pooled and applied to a 11 7-cm 2 *,5*-ADP–agar-

ose affinity column (Sigma) equilibrated with 25 mM taining 20% glycerol (v/v), 0.20 mM EDTA, and 1.0 mM

DTT, concentrated by ultrafiltration using an Amiconpotassium phosphate buffer (pH 7.7) containing 20%glycerol (v/v), 0.10 mM EDTA, 0.10 mM DTT, 1.0 mM Centriprep 30 device (Amicon, Beverly, MA), and

stored in small aliquots at 0207C.FMN, 0.5% sodium cholate (w/v), 15 mM aNF, and 0.5



Purification of other proteins. Recombinant human Optimal expression temperature (287C) and time (48h) yielded approximately 35 nmol of spectrally detect-P450 1A2 was purified from E. coli as described pre-

viously (33). NADPH-P450 reductase was purified from able P450 1A2 (liter culture)01 in bacterial whole cellliver microsomes of phenobarbital-treated rabbits (38,39). lysate. This yield is in sharp contrast to the 250 nmol

(liter culture)01 we observe upon expression of P450Other procedures and assays. Oligonucleotides1A2 at 307C for 24 h (33). Even slight elevations of theused for plasmid construction were synthesized in thegrowth temperature and shaking speed (e.g., 307C andVanderbilt facility using a PerSeptive Biosystems 8909200 rpm, respectively) reduced expression of the fusionDNA synthesizer (Framingham, MA) and partially pu-protein to spectroscopically undectectable levels.rified using ChromaSpin30 columns (Clontech, Palo

Additives for the culture medium included the hemeAlto, CA). Deoxynucleotide sequencing was performedusing the method of Sanger et al. (40). Proteins were and flavin prosthetic group precursors d-ALA and ribo-separated by electrophoresis on 7.5% SDS–polyacryl- flavin. Prior experiments have demonstrated that in-amide gels (w/v) according to the method of Laemmli cluding d-ALA in the culture medium boosts expression(41) as modified elsewhere (39). Proteins were visual- of P450s to varying extents. For example, expressionized by staining with ammonical silver (42). levels of P450s 1A1, 2C9, 2D6, and 2E1 are enhanced

P450 spectra were recorded with an Aminco DW2/ several fold by d-ALA, whereas those of P450s 1A2 andOLIS instrument (On-Line Instrument Systems, Bo- 3A4 remain relatively constant (12,47). We have notgart, GA) at ambient temperature. Fusion protein was investigated the effect of varying riboflavin concentra-quantitated by measuring P450 using the spectral tion on fusion protein expression but elected to includemethod of Omura and Sato (37). Wavelength maxima it at low concentations as described for the expressionwere determined using the peak finder or Aminco/OLIS of rat NADPH-P450 reductase (35,48).second-derivative software. NADPH-cytochrome c re- Purification of human P450 1A2:rat NADPH-P450duction was determined as described (38). reductase from E. coli membranes. The purificationEthoxyresorufin O-deethylation activity was deter- procedure utilized was based on that published formined independently by fluorescence and spectrophoto- NADPH-P450 reductase (38), as modified by others andmetric methods as described elsewhere (43,44). Phen- ourselves for fusion proteins (18,49). Recovery for thisacetin O-deethylation was determined by radio-TLC as protein was as described in Table 1. Individual purifi-described previously (45,46). cation fractions, separated by 7.5% SDS–PAGE, areProtein concentrations were estimated using the bi-

shown in Fig. 2.cinchoninic acid (BCA) method according to supplier’sMembranes were first solubilized in buffer con-recommendations (Pierce Chemical Co., Rockford, IL)

taining Emulgen and the P450 1A2 inhibitor aNF. Hu-using BSA as a standard. The protein concentration ofman P450 1A2 is very unstable in the presence of deter-the final preparation was determined by quantitativegent required for its solubilization, as evidenced by aamino acid analysis in the Vanderbilt facility, usingcomplete loss of P450 spectral properties minutes afterHPLC of phenythiohydantoins prepared after hydroly-addition of detergent (33). aNF has been shown to sta-sis of the protein in 6 N HCl for 24 h. The calculatedbilize the protein throughout the purification until theconcentration was not corrected for tryptophan, whichfinal chromatographic step, at which point it can becomposes 2.5 mol% of the protein.washed out along with detergent. For this reason, aNF

RESULTS AND DISCUSSION was used for purification of the fusion protein as well.Solubilized membranes were sequentially subjectedConstruction of a human P450 1A2:rat NADPH-P450

to DEAE anion exchange (in the presence of Emulgen),reductase expression plasmid and expression in E. coli.2 *,5*-ADP-agarose affinity (in the presence of sodiumElimination of the internal SalI site in our P450 1A2cholate), and hydroxylapatite chromatographies. ThecDNA facilitated subcloning of this fragment into aADP–agarose step purified the fusion protein to essen-(truncated) reductase expression vector. The only re-tial homogeneity (Fig. 2) by exploiting the affinity ofmaining SalI site in the fusion plasmid partially en-the reductase domain of the fusion protein for pyridinecodes the Ser–Thr dipetide linker, which bridges thenucleotides resembling the cofactor, NADP/. A bacte-P450 and reductase domains of the fusion protein. Thisrial protein with an Mr of 60–70 kDa and appreciablestrategy is based on described precedents which havecytochrome c reduction activity also bound to the ADP–been employed successfully for the expression of a num-agarose column and initially complicated both the puri-ber of human P450:rat NADPH-P450 reductase fusionfication and the precise activity measurements. Thisproteins in E. coli (15–18). To date, attempts to expresscontaminant, however, could easily be separated byfusion plasmids using the human NADPH-P450 reduc-eluting with a slow gradient of 2 *-AMP.tase have proven unsuccessful. Construction of the fu-

Considerable losses of purified protein were incurredsion plasmid was carried out as described under Mate-rials and Methods and is depicted in Fig. 1. during hydroxylapatite chromatography. While we have



TABLE 1

Purification of Human P450 1A2:rat NADPH-P450 Reductase Fusion Protein from E. coli Membranes

Protein P450 P450 specific content Yield Cytochrome c reductionPurification step (mg) (nmol) (nmol/mg protein) (%) (nmol/min/mg protein)

Solubilized membranes 3870 225 0.058 100 480DE52 342 98 0.29 44 850ADP–agarose 164 38 0.23 17 120Hydroxylapatite 0.78a 5.2 6.7 2.3 27,000

a Value calculated from quantitative amino acid analysis carried out as described under Materials and Methods.

not investigated the necessity of this last step in any binant human P450 1A2 (33). The absolute spectra ofthe Fe3/ and Fe2/ forms were consistent with a proteingreat detail, results using purified protein with cholate

added at concentrations representative of those required containing both heme and flavin prosthetic groups (Fig.4B). Furthermore, the content of P450 heme (per milli-to elute the fusion protein from an ADP–agarose column

seem to indicate that this step can be eliminated, thereby gram protein) indicated that little apoprotein was pres-ent (Table 1). These spectra indicate the presence ofincreasing yield. Preincubation of the purified protein

with concentrations of cholate ranging from 0.1 to 0.3% mainly high-spin iron in the preparation, as noted bydiagnostic absorption bands at 389 and 641 nm andactually stimulated 7-ethoxyresorufin O-deethylation ac-

tivity by Ç50% (results not shown). In lieu of past find- coalesced a,b peaks. Second-derivative spectroscopy ofthe Soret band of the Fe3/ form (52) confirmed thisings that low concentrations of cholate enhance many

P450-mediated reactions (50,51), these results are not result, revealing a strong negative peak at 386 nm (Fig.4C). This band is not due only to flavin, because a com-completely surprising. It might, therefore, be advisable

to remove aNF by washing while the protein is bound to plementary band was not seen at 450 nm. As is thecase for P450 1A2, there is evidence for some low-spinthe ADP–agarose column and elute in as low a concentra-

tion of cholate as possible. iron, as noted by the negative peak at 412 nm, but itrepresents a small portion of the preparation (õ10%).Following the precedent of recombinant P450 1A2, the

potassium phosphate concentration of the final deter- Extinction coefficients for wavelength maxima of thepurified protein were calculated and are presented ingent-free preparation was kept relatively high (ú100

mM) to prevent precipitation (33). The purified protein Table 2. These values are based upon De450–490 Å 91mM01 cm01 for the Fe2/–CO vs Fe2/ difference spec-comigrated with other P450:NADPH-P450 reductase pro-

teins (18) produced in this laboratory (Fig. 3). trum (37,53).Catalytic activities of purified human P450 1A2:ratSpectral properties. The purified protein had a typi-

cal Fe2/–CO vs Fe2/ difference spectrum, with little NADPH-P450 reductase. The purified protein cata-lyzed 7-ethoxyresorufin O-deethylation and phenacetinevidence for cytochrome P420 (Fig. 4A). The absorption

maximum was at 446 nm, identical with that of recom- O-deethylation activities (Table 3). Unlike similarlyconstructed human P450 3A4:rat NADPH-P450 reduc-tase fusion proteins (17,18), enzymatic activity could

FIG. 3. Purified recombinant human P450:rat NADPH-P450 re-FIG. 2. Purification of recombinant human P450 1A2:rat NADPH-P450 reductase from E. coli membranes. Lane 1, Mr markers (Gibco ductase fusion proteins. Lane 1, Mr markers (Gibco); lane 2, P450

3A4:NADPH-P450 reductase fusion protein (18); lane 3, P450BRL, Grand Island, NY); lane 2, whole cell lysate; lane 3, 104g su-pernant; lane 4, membranes; lane 5, 105g supernatant; lane 6, DE- 1A2:NADPH-P450 reductase fusion protein; lane 4, P450

1A1:NADPH-P450 reductase fusion protein (18); lane 5, Mr markers.52 fraction; lane 7, ADP–agarose fraction; lane 8, Mr markers.



FIG. 4. Absorbance spectra of purified human P450 1A2:rat NADPH-P450 reductase fusion protein. Spectra were recorded with 0.93 mM

fusion protein in 500 mM potassium phosphate buffer (pH 7.4) containing 20% glycerol (v/v), 1.0 mM EDTA, and 0.20 mM DTT. (A) Fe2/–CO vs Fe2/ difference spectrum. (B) Absolute spectra: Fe3/, Fe2/, and Fe2/–CO forms, as indicated (reduction with excess Na2S2O4). (C)Second-derivative spectrum of Fe3/ form.

not be stimulated by the addition of equimolar amounts to 2-fold higher than we obtained here (33), we haveof purified rabbit NADPH-P450 reductase, implying in- consistently obtained the presented values with thistramolecular electron transfer. Nonetheless, the mea- enzyme preparation (18). The corresponding values forsured activities are significantly less than those of the the fusion protein range from 40 to 60% of those of thesystem reconstituted from its hemoprotein and flavo- reconstituted system, suggesting that the structuralprotein components. While we previously reported O- constraints imposed by the linkage between the do-deethylation activities for the reconstituted system (ra- mains may be less than optimal, at least as comparedtio NADPH-P450 reductase:P450 Å 1:1) that were 1.5- to the naturally occuring fusion proteins, P450 102

(BM-3, Bacillus megaterium), and the mammalianP450-like nitric oxide synthase enzymes. When ex-pressed as a percentage of activity of the reconstitutedTABLE 2system, however, our values are comparable to thoseSpectral Characteristics of Purified Human P450 1A2:Ratreported for other such human P450:rat NADPH-P450NADPH-P450 Reductase Fusion Proteina

reductase fusion proteins, in the absence of addedNADPH-P450 reductase (17,18).Wavelength

maximum e Interestingly, we detected a fairly marked effect ofIron valence, ligand (nm)b (mM01 cm01) buffer choice on catalytic activity of both recombinant

P450 1A2 and the fusion protein (results not shown).Fe3/ 389 180Replacing the potassium phosphate assay buffer (100641 8.9

Fe2/ 414 120 mM potassium phosphate, pH 7.4) with a Tris buffer548 23 (25 mM Tris-HCl, 100 mM KCl) reduced catalytic activi-

Fe2/rCO 446 140 ties by Ç25%. Addition of 10 mM MgCl2 to the Tris551 22

buffer inhibited both systems by another Ç25%. ThisFe2/rCO vs Fe2/ (difference) 446 91c

last observation is in contrast to the stimulatory effecta All spectra were recorded in 0.50 M potassium phosphate buffer of MgCl2 on other P450-mediated reactions, e.g., P450

(pH 7.4) containing 1.0 mM EDTA and 20% glycerol (v/v). 3A4 (13).b Determined using the peak finder program of the Aminco DW2/The NADPH-P450 reductase domain of the fusionOLIS system.

c Assumed in making calculations (37,53). protein was catalytically active, as judged by the capac-



TABLE 3 supplied with bacterial flavoproteins as electron donors(13). Furthermore, this study presents continued evi-7-Ethoxyresorufin O-Deethylation and Phenacetin

O-Deethylation Activities of Purified Human P450 dence that mammalian detoxication enzyme systems1A2:Rat NADPH-P450 Reductase Fusion Protein can be engineered for bacterial expression into single

catalytically self-sufficient polypeptides. Such proteinsCatalytic activity (nmol product provide potential means for determining metabolite

formed/min/nmol P450) profiles of new pharmaceuticals and for carrying outbiodegradation of chemicals in the environment. TheySystem 7-Ethoxyresorufin Phenacetinmay also be useful for mutagenesis studies of struc-

Reconstituted system ture–function relationships.completea 0.42 0.26

Fusion proteinCompletea 0.24 0.094 ACKNOWLEDGMENTSPlus NADPH-P450 reductase 0.24 ND

The authors thank Dr. Y.-J. Chun for the pCW*3A4rOR construct,Plus P450 1A2 0.22 NDDr. M-S. Dong for helpful discussions, Dr. T. Kuwabara for criticalreading of the manuscript, and Dr. O. Boutaud for technical assis-a 7-Ethoxyresorufin (5 mM) or [ring-3H]phenacetin (20 mM) was in-tance with figures. This work was supported in part by United Statescubated with either equimolar amounts of purified recombinant hu-Public Health Service Grants R35 CA 44353 and P30 ES 00267man P450 1A2 and rabbit NADPH-P450 reductase (25 or 100 nM of(F.P.G.) and Medical Scientist Training Program Grant GM 07347each enzyme, respectively) or human P450 1A2:rat NADPH-P450(A.P.).reductase fusion protein (25 or 100 nM, respectively). The incubation

mixtures also included 30 mM DLPC, 100 mM potassium phosphatebuffer (pH 7.4), and an NADPH-generating system (39). The capacity

REFERENCESof free NADPH-P450 reductase or P450 1A2 (25 nM incubation) tostimulate 7-ethoxyresorufin O-deethylation activity of the fusion pro-

1. Keeney, D. S., and Waterman, M. R. (1993) Regulation of steroidtein was determined where indicated. All values represent means ofhydroxylase gene expression: Importance to physiology and dis-duplicate experiments. ND, not determined.ease. Pharmacol. Ther. 58, 301–317.

2. Higashi, Y., Hiromasa, T., Tanae, A., Miki, T., Nakura, J., Kondo,T., Ohura, T., Ogawa, E., Nakayama, K., and Fujii-Kuriyama,ity of this protein to reduce cytochrome c. As shown inY. (1991) Effects of individual mutations in the P-450 (C21) pseu-

Table 1, the purified protein displayed a rate of 27,000 dogene on the P-450 (C21) activity and their distribution in thenmol cytochrome c reduced (min)01 (mg fusion pro- patient genomes of congenital steroid 21-hydroxylase deficiency.

J. Biochem. 109, 638–644.tein)01. This value is based upon measurement of pro-3. Miller, W. L., and Morel, Y. (1989) The molecular genetics of 21-tein concentration by quantitative amino acid analysis

hydroxylase deficiency. Annu. Rev. Genet. 23, 371–393.rather than the BCA method. Past and current results4. White, P. C., Dupont, J., New, M. I., Leiberman, E., Hochberg,(18) have demonstrated that the latter method often

Z., and Rosler, A. (1991) A mutation in CYP11Ba (Arg 448—grossly overestimates protein concentration of a puri-His) associated with steroid 11b-hydroxylase deficiency in Jewsfied preparation. of Moroccan origin. J. Clin. Invest. 87, 1664–1667.

As mentioned previously, a contaminating protein 5. Smith, R. L., Idle, J. R., Mahgoub, A. A., Sloan, T. P., and Lan-bound to ADP–agarose and possessed enzymatic activ- caster, R. (1978) Genetically determined defects of oxidation atity toward cytochrome c. The observed decrease in cyto- carbon centres of drugs. Lancet i, 943–944.chrome c reduction activity following DE52 chromatog- 6. Woosley, R. L., Chen, Y., Freiman, J. P., and Gillis, R. A. (1993)

Mechanism of the cardiotoxic actions of terfenadine. J. Am. Med.raphy probably is explained best by purification of theAssoc. 269, 1532–1536.fusion protein away from other protein(s) present in

7. Guengerich, F. P. (1988) Roles of cytochrome P-450 enzymes insolubilized membranes which are capable of catalyzingchemical carcinogenesis and cancer chemotherapy. Cancer Res.this activity.48, 2946–2954.

Conclusions. We report here the successful con- 8. Guengerich, F. P. (1990) Characterization of roles of human cyto-struction and heterologous expression of a human P450 chrome P-450 enzymes in carcinogen metabolism. Asia Pacific

J. Pharmacol. 5, 327–345.1A2:rat NADPH-P450 reductase fusion plasmid, aswell as the purification and spectral and catalytic char- 9. Guengerich, F. P. (1992) Metabolic activation of carcinogens.

Pharmacol. Ther. 54, 17–61.acterization of the fusion protein it encodes. Expression10. Shimada, T., Yun, C-H., Yamazaki, H., Gautier, J-C., Beaune,levels of this protein were fairly low and its measured

P. H., and Guengerich, F. P. (1992) Characterization of humanactivities in vitro were less than those of the combina-lung microsomal cytochrome P-450 1A1 and its role in the oxida-tion of recombinant human P450 1A2 and rabbit tion of chemical carcinogens. Mol. Pharmacol. 41, 856–864.

NADPH-P450 reductase, emphasizing the need for in-11. Sakaki, T., Oeda, K., Yabusaki, Y., and Ohkawa, H. (1986) Mono-

vestigating new and better methods for heterologous oxygenase activity of Saccharomyces cerevisiae cells transformedexpression of detoxication enzymes. It should be noted, with expression plasmids carrying rat cytochrome P-450MC

cDNA. J. Biochem. 99, 741–749.however, that catalytic activities of this protein sig-nificantly exceed those of recombinant human P450s 12. Guengerich, F. P., Gillam, E. M. J., and Shimada, T. (1996) New



applications of bacterial systems to problems in toxicology. Crit. 26. Butler, M. A., Lang, N. P., Young, J. F., Caporaso, N. E., Vineis,P., Hayes, R. B., Teitel, C. H., Massengill, J. P., Lawsen, M. F.,Rev. Toxicol. 26, 551–583.and Kadlubar, F. F. (1992) Determination of CYP1A2 and NAT213. Yamazaki, H., Ueng, Y-F., Shimada, T., and Guengerich, F. P.phenotypes in human populations by analysis of caffeine urinary(1995) Roles of divalent metal ions in oxidations catalyzed bymetabolites. Pharmacogenetics 2, 116–127.recombinant cytochrome P450 3A4 and replacement of iron-sul-

fur proteins and oxygen surrogates. Biochemistry 34, 8380– 27. Sesardic, D., Boobis, A. R., Edwards, R. J., and Davies, D. S.(1988) A form of cytochrome P450 in man, orthologous to form8389.d in the rat, catalyses the O-deethylation of phenacetin and is14. Dong, M-S., Yamazaki, H., Guo, Z., and Guengerich, F. P. (1996)inducible by cigarette smoking. Br. J. Clin. Pharmacol. 26, 363–Recombinant human cytochrome P450 1A2 and an N-terminal372.truncated form: Construction, purification, aggregation proper-

ties, and interactions with flavodoxin, ferredoxin, and NADPH- 28. Conney, A. H., Buening, M. K., Pantuck, E. J., et al. (1980) Regu-lation of human drug metabolism by dietary factors In ‘‘Environ-cytochrome P450 reductase. Arch. Biochem. Biophys. 327, 11–

19. mental Chemicals, Enzyme Function and Human Disease,’’ CibaFoundation Symposium 76, pp. 147–167, Excerpta Medica, Am-15. Shet, M. S., Fisher, C. W., Arlotto, M. P., Shackleton, C. H. L.,sterdam.Holmans, P. L., Martin-Wixtrom, C. A., Saeki, Y., and Esta-

brook, R. W. (1994) Purification and enzymatic properties of a 29. Kadlubar, F. F., Butler, M. A., Kaderlik, K. R., Chou, H. C., andLang, N. P. (1992) Polymorphisms for aromatic amine metabo-recombinant fusion protein expressed in Escherichia coli con-

taining the domains of bovine P450 17A and rat NADPH-P450 lism in humans: Relevance for human carcinogenesis. Environ.Health Perspect. 98, 69–74.reductase. Arch. Biochem. Biophys. 311, 402–417.

16. Fisher, C. W., Shet, M. S., Holmans, P. L., and Estabrook, R. W. 30. Crespi, C. L., Steimel, D. T., Aoyama, T., Gelboin, H. V., andGonzalez, F. J. (1990) Stable expression of human cytochromeExpression of fusions between cytochrome P450 and cytochrome

P450 reductase in E. coli in ‘‘8th International Conference on P450IA2 cDNA in a human lymphoblastoid cell line: Role of theenzyme in the metabolic activation of aflatoxin B1. Mol. Carcino-Cytochrome P450: Biochemistry, Biophysics, and Molecular Biol-

ogy,’’ Lisbon, Portugal, 24–28 October, 1993, p. 81. gen. 3, 5–8.

31. Glatt, H. R., Pauly, K., Wolfel, C., et al. (1993) Stable expression17. Shet, M. S., Fisher, C. W., Holmans, P. L., and Estabrook, R. W.(1993) Human cytochrome P450 3A4: Enzymatic properties of of heterologous cytochromes P450 in V79 cells: Mutagenicity

studies with polycyclic aromatic hydrocarbons in ‘‘Polycyclic Aro-a purified recombinant fusion protein containing NADPH-P450reductase. Proc. Natl. Acad. Sci. USA 90, 11748–11752. matic Compounds: Synthesis, Properties, Analytical Measure-

ments, Occurrence and Biological Effects: Proceedings of the18. Chun, Y-J., Shimada, T., and Guengerich, F. P. (1996) Construc-Thirteenth International Symposium on Polynuclear Aromatiction of a human cytochrome P450 1A1:rat NADPH-P450 reduc-Hydrocarbons,’’ October 1–4, 1991, Bordeaux, France (Gar-tase fusion protein cDNA, expression in Escherichia coli, purifi-rigues, P., and Lamotte, M., Eds.), pp. 1167–1174, Gordon andcation, and catalytic properties of the enzyme in bacterial cellsBreach, New York.and after purification. Arch. Biochem. Biophys. 330, 48–58.

32. Fisher, C. W., Caudle, D. L., Martin-Wixtrom, C., Quattrochi,19. Ruettinger, R. T., Wen, L. P., and Fulco, A. J. (1989) Coding nu-L. C., Tukey, R. H., Waterman, M. R., and Estabrook, R. W.cleotide, 5* regulatory, and deduced amino acid sequences of(1992) High-level expression of functional cytochrome P450 1A2P-450BM-3, a single peptide cytochrome P-450:NADPH-P-450in Escherichia coli. FASEB J. 6, 759–764.reductase from Bacillus megaterium. J. Biol. Chem. 264, 10987–

10995. 33. Sandhu, P., Guo, Z., Baba, T., Martin, M. V., Tukey, R. H., andGuengerich, F. P. (1994) Expression of modified human cyto-20. White, K. A., and Marletta, M. A. (1992) Nitric oxide synthasechrome P450 1A2 in Escherichia coli: Stabilization, purification,is a cytochrome P-450 type hemoprotein. Biochemistry 31, 6627–spectral characterization, and catalytic activities of the enzyme.6631.Arch. Biochem. Biophys. 309, 168–177.

21. McMillan, K., Bredt, D. S., Hirsch, D. J., Snyder, S. H., Clark,34. Sandhu, P., Baba, T., and Guengerich, F. P. (1993) ExpressionJ. E., and Masters, B. S. S. (1992) Cloned, expressed rat cerebel-

of modified cytochrome P450 2C10 (2C9) in Escherichia coli, pu-lar nitric oxide synthase contains stoichiometric amounts ofrification, and reconstitution of catalytic activity. Arch. Biochem.heme, which binds carbon monoxide. Proc. Natl. Acad. Sci. USABiophys. 306, 443–450.89, 11141–11145.

35. Shen, A. L., Porter, T. D., Wilson, T. E., and Kasper, C. B. (1989)22. Murakami, H., Yabusaki, Y., Sakaki, T., Shibata, M., and Oh-Structural analysis of the FMN binding domain of NADPH-cyto-kawa, H. (1987) A genetically engineered P450 monooxygenase:chrome P-450 oxidoreductase by site-directed mutagenesis. J.Construction of the functional fused enzyme between rat cyto-Biol. Chem. 264, 7584–7589.chrome P450c and NADPH-cytochrome P450 reductase. DNA 6,

189–197. 36. Gillam, E. M. J., Baba, T., Kim, B-R., Ohmori, S., and Guenger-ich, F. P. (1993) Expression of modified human cytochrome P45023. Shibata, M., Sakaki, T., Yabusaki, Y., Murakami, H., and Oh-3A4 in Escherichia coli and purification and reconstitution ofkawa, H. (1990) Genetically engineered P450 monooxygenases:the enzyme. Arch. Biochem. Biophys. 305, 123–131.Construction of bovine P450c17/yeast reductase fused enzymes.

DNA Cell Biol. 9, 27–36. 37. Omura, T., and Sato, R. (1964) The carbon monoxide-bindingpigment of liver microsomes. I. Evidence for its hemoprotein24. Sakaki, T., Shibata, M., Yabusaki, Y., Murakami, H., and Oh-nature. J. Biol. Chem. 239, 2370–2378.kawa, H. (1990) Expression of bovine cytochrome P450c21 and

its fused enzymes with yeast NADPH-cytochrome P450 reduc- 38. Yasukochi, Y., and Masters, B. S. S. (1976) Some properties of adetergent-solubilized NADPH-cytochrome c (cytochrome P-450)tase in Saccharomyces cerevisiae. DNA Cell Biol. 9, 603–614.reductase purified by biospecific affinity chromatography. J.25. Butler, M. A., Iwasaki, M., Guengerich, F. P., and Kadlubar,Biol. Chem. 251, 5337–5344.F. F. (1989) Human cytochrome P-450PA (P-450IA2), the phenac-

etin O-deethylase, is primarily responsible for the hepatic 3- 39. Guengerich, F. P. (1994) Analysis and characterization of en-zymes in ‘‘Principles and Methods of Toxicology’’ (Hayes, A. W.,demethylation of caffeine and N-oxidation of carcinogenic ary-

lamines. Proc. Natl. Acad. Sci. USA 86, 7696–7700. Ed.), pp. 1259–1313, Raven Press, New York.



40. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) DNA sequenc- two prototypes for genetic polymorphism in oxidative drug me-tabolism. J. Biol. Chem. 260, 9057–9067.ing with chain-terminatiing inhibitors. Proc. Natl. Acad. Sci.

USA 74, 5463–6467. 47. Gillam, E. M. J., Guo, Z., Martin, M. V., Jenkins, C. M., andGuengerich, F. P. (1995) Expression of cytochrome P450 2D6 in41. Laemmli, U. K. (1970) Cleavage of struatural proteins duringEscherichia coli, purification, and spectral and catalytic charac-the assembly of the head of bacteriophage T4. Nature 227, 680–terization. Arch. Biochem. Biophys. 319, 540–550.685.

48. Porter, T. D., Wilson, T. E., and Kasper, C. B. (1987) Expression42. Wray, W., Boulikas, T., Wray, V. P., and Hancock, R. (1981)of a functional 78,000 dalton mammalian flavoprotein, NADPH-Silver staining of proteins in polyacrylamide gels. Anal. Bio-cytochrome P-450 oxidoreductase, in Escherichia coli. Arch. Bio-chem. 118, 197–203.chem. Biophys. 254, 353–367.43. Burke, M. D., and Mayer, R. T. (1983) Differential effects of phe-

49. Fisher, C. W., Shet, M. S., Martin-Wixtrom, C., and Estabrook,nobarbitone and 3-methylcholanthrene induction on the hepaticR. W. (1992) High-level expression in Escherichia coli of enzy-microsomal metabolism and cytochrome P-450-binding of phe-matically active fusion proteins containing the domains of cyto-noxazone and a homologous series of its n-alkylethers (alkoxyre-chromes P450 and NADPH-P450 reductase flavoprotein. Proc.sorufins). Chem.-Biol. Interact. 45, 243–258.Natl. Acad. Sci. USA 89, 10817–10821.44. Burke, M. D., and Mayer, R. T. (1975) Inherent specificities of

50. Imaoka, S., Imai, Y., Shimada, T., and Funae, Y. (1992) Role ofpurified cytochromes P-450 and P-448 toward biphenyl hydroxyl-phospholipids in reconstituted cytochrome P450 3A forms andation and ethoxyresorufin deethylation. Drug Metab. Dispos. 3,mechanism of their activation of catalytic activity. Biochemistry245–253.31, 6063–6069.45. Larrey, D., Distlerath, L. M., Dannan, G. A., Wilkinson, G. R.,

51. Kaminsky, L. S., Dunbar, D., Guengerich, F. P., and Lee, J. J.and Guengerich, F. P. (1984) Purification and characterization(1987) Detergents as probes of reconstituted rat liver cytochromeof the rat liver microsomal cytochrome P-450 involved in the 4-P-450 function. Biochemistry 26, 1276–1283.hydroxylation of debrisoquine, a prototype for genetic variation

in oxidative drug metabolism. Biochemistry 23, 2787–2795. 52. O’Haver, T. C., and Green, G. L. (1976) Numerical error analysisof derivative spectrometry for the quantitative analysis of mix-46. Distlerath, L. M., Reilly, P. E. B., Martin, M. V., Davis, G. G.,tures. Anal. Chem. 48, 312–318.Wilkinson, G. R., and Guengerich, F. P. (1985) Purification and

characterization of the human liver cytochromes P-450 involved 53. Omura, T., and Sato, R. (1967) Isolation of cytochromes P-450and P-420. Methods Enzymol. 10, 556–561.in debrisoquine 4-hydroxylation and phenacetin O-deethylation,
