THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 264, No. 13, Issue of May 5, pp. 7584-7589,1989 Printed in U.S.A.

Structural Analysis of the FMN Binding Domain of NADPH- Cytochrome P-450 Oxidoreductase by Site-directed Mutagenesis*

(Received for publication, November 28, 1988)

Anna L. Shen, Todd D. Porter$, Thomas E. Wilson$, and Charles B. Kasperll From the McArdle Laboratory for Cancer Research, University of Wisconsin, Madison, Wisconsin 53706

Comparison of the amino acid sequence of rat liver NADPH-cytochrome P-450 oxidoreductase with that of flavoproteins of known three-dimensional structure suggested that residues Tyr-140 and Tyr-178 are in- volved in binding of FMN to the protein. To test this hypothesis, NADPH-cytochrome P-450 oxidoreduc- tase was expressed in Escherichia coli using the expression-secretion vector PIN-111-ompA3, and site- directed mutagenesis was employed to selectively alter these residues and demonstrate that they are major determinants of the FMN-binding site.

Bacterial expression produced a membrane-bound 80-kDa protein containing 1 mol each of FMN and FAD per mol of enzyme, which reduced cytochrome c at a rate of 51.5 Mmol/min/mg of protein and had ab- sorption spectra and kinetic properties very similar to those of the rat liver enzyme. Replacement of Tyr-178 with aspartate abolished FMN binding and cytochrome c reductase activity. Incubation with FMN increased catalytic activity to a maximum of 8.6 pmol/min/mg of protein. Replacement of Tyr-140 with aspartate did not eliminate FMN binding, but reduced cytochrome c reductase activity about 5-fold, suggesting that FMN may be bound in a conformation which does not permit efficient electron transfer. Substitution of phenylala- nine at either position 140 or 178 had no effect on FMN content or catalytic activity. The FAD level in the Asp-178 mutant was also decreased, suggesting that FAD binding is dependent upon FMN; FAD incor- poration may occur co-translationally and require prior formation of an intact FMN domain.

Microsomal NADPH-cytochrome P-450 oxidoreductase is a 78,225-dalton flavoprotein which catalyzes electron transfer from NADPH to cytochrome P-450, as well as to other micro- somal enzymes and various artificial electron acceptors (1-3). The protein contains 1 mol each of FMN and FAD (4), and electron transfer proceeds from NADPH to FAD to FMN to the heme of cytochrome P-450 (5-7). Five functional domains of NADPH-cytochrome P-450 oxidoreductase have been iden- tified. These include an amino-terminal membrane-binding domain which is also required for reduction of cytochrome P-

*This research was supported by Grant CA22484 from the Na- tional Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address: Dept. of Biological Chemistry, University of Michigan Medical School, Ann Arbor, MI 48109-0010.

Present address: Washington University School of Medicine, St. Louis, MO 63110.

ll To whom correspondence and reprint requests should be ad- dressed.

450 (8) and regions involved in binding of FMN, FAD, NADPH, and cytochrome P-450 (9,lO).

Comparison of the amino acid sequence of rat liver reductase’ deduced from the cDNA sequence (11) with that of flavoproteins of known three-dimensional structure reveals regions of sequence homology corresponding to residues in- volved in prosthetic group binding (9). The amino-terminal portion of the protein (residues 90 to 210) shows regions of homology with the FMN-containing bacterial flavodoxins, while the carboxyl-terminal half of the molecule has regions of homology with ferredoxin-NADP+ reductase and glutathi- one reductase, which were proposed to be the FAD- and NADPH-binding domains. Identification of these regions of sequence homology led to the proposal that NADPH-cyto- chrome P-450 oxidoreductase arose through a fusion of the ancestral flavodoxin and ferredoxin-related genes, with sub- sequent insertion of a region between the FMN- and FAD- binding domains to facilitate electron transfer between FMN and FAD or to cytochrome P-450 (9).

X-ray crystallographic analysis of Desulfouibrw vulgaris flavodoxin reveals that FMN is bound between Trp-60 and Tyr-98 in the D. vulgaris protein (12). The FMN isoalloxazine ring is co-planar with Tyr-98, permitting stacking interactions between these two aromatic groups, while the Trp-60 side chain is tilted approximately 45” from the plane of the isoal- loxazine ring. Sequence comparisons indicate that Tyr-98 corresponds to Tyr-178 of the rat liver reductase, while Trp- 60 can be aligned with Tyr-140 (9). Conservation of Tyr-178 and maintenance of the 38-amino acid spacing between the putative FMN-binding residues in the D. vulgaris protein and NADPH-cytochrome P-450 oxidoreductase suggests that these 2 tyrosyl residues are essential in the binding of FMN to the reductase. Chemical modification studies have also implicated tyrosyl residues in the binding of FMN to rabbit liver reductase (13).

To test the hypothesis that Tyr-178 and Tyr-140 are in- volved in FMN binding, site-directed mutagenesis was em- ployed to selectively alter these residues, and the mutant cDNAs for rat liver reductase were expressed in E. coli (14). Substitution with phenylalanine at one or both of these po- sitions preserves aromaticity and would be expected to have minimal effects on FMN binding, while introduction of an alkyl side chain bearing a negative charge (e.g. aspartate) should greatly alter the character of the FMN-binding do- main. Accordingly, the following single and double mutants were constructed Phe-140, Phe-178, Asp-140, Asp-178, Phe- 140/Phe-178, and Asp-140/Asp-178. The effects of these sub- stitutions on FMN binding and catalytic activity are pre- sented.

The abbreviations used are: reductase, NADPH-cytochrome P- 450 oxidoreductase; SDS, sodium dodecyl sulfate.

7584

F M N Binding Domain of NADPH-Cytochrome P-450 Oxidoreductase 7585

MATERIALS AND METHODS

Construction of NADPH-Cytochrome P-450 Oxidoreductase Expression-Secretion Plasmid-The secretion-expression vector PIN- 111-ompA3 (15) was used for expression of rat liver reductase in Escherichia coli. Fig. 1A shows the steps in construction of the reductase expression plasmid. The 2.4-kilobase pair BamHI-Hind111 fragment of pORFL (14), which contains the complete reductase coding sequence plus 398 base pairs of pBR322 sequence, was cloned into pUC8 (16) to generate pOR261. pOR262 was constructed by cloning the 2.4-kilobase pair EcoRI-Hind111 fragment of pOR261 into PIN-111-ompA3. To remove the 3”BamHI site, the HindIII-Sal1 fragment of pOR262 was replaced with a HindIII-XhoI fragment from pACYC177 (17), producing pOR263. pOR263 in E. coli C-1A was used for all expression and mutagenesis studies, although pOR262 produces an identical protein.

Mutagenesis-The template for mutagenesis was prepared by clon- ing the 748-base pair BamHI-Sac1 fragment from pORFL into M13mp19. Mutagenesis was carried out by the two-primer method of Zoller and Smith (18) using the (-20) M13 sequencing primer and the following oligonucleotides synthesized at the University of Wis- consin Biotechnology Center: 5’-TCTCCGTCTGTGGCC-3’(Asp-

GTGGC-3’ (Phe-140), and 5’-GTGCTCAAAGGTCTT-3’ (Phe-178). Double mutants were constructed by mutagenizing the Asp-178 or Phe-178 mutant phages with the appropriate oligonucleotides.

Mutant phage were selected by plaque hybridization using the synthetic oligonucleotides as probes and confirmed by sequencing. The pOR263 BamHI-Sac1 fragment was replaced with the mutant BamHI-Sac1 fragment. Mutant plasmids were mapped and sequenced across the mutation and cloning sites to verify that the mutation was present and that no other rearrangements had occurred. All sequenc- ing was carried out by the method of Sanger (19).

Purification of NADPH-Cytochrome P-450 Oxidoreductase-E. coli C-1A carrying pOR263 or one of the mutant plasmids was grown at 37 “C in LB broth containing 1 pg/ml riboflavin and 50 pg/ml ampicillin to an Asso of approximately 0.8. Isopropyl-1-thio-0-D- galactopyranoside was added to a final concentration of 0.5 mM, and the cells were grown for an additional 16 h. Cells were harvested by centrifugation at 3,000 X g for 15 min at 4 “C and resuspended in 75 mM Tris, pH 8.0, 0.25 M sucrose, 0.25 mM EDTA, and 0.02 mg/ml lysozyme. After incubation for 20 min at 4 “C, spheroplasts were pelleted by centrifugation at 3,000 X g for 30 min, resuspended in 50 mM Tris, pH 8.0, 0.5 mM EDTA, 0.01 mg/ml aprotinin, and lysed by sonication. The membrane fraction was isolated by centrifugation at 73,000 X g for 45 min, resuspended in affinity buffer (50 mM Tris, pH 7.7, 0.1 mM EDTA, 0.05 mM dithiothreitol, 10% glycerol), and solubilized with Triton X-100 (final concentration 0.1%). Unsolubil- ized material was removed by centrifugation at 73,000 X g for 45 min, and the supernatant was applied to a 2’,5’-ADP Sepharose (20) column equilibrated with affinity buffer containing 0.1% Triton X- 100. The column was washed with 5 mM adenosine in affinity buffer containing 0.1% Triton X-100 followed by elution of the reductase with 2 mM 2’-AMP in affinity buffer containing 0.1% Triton X-100. SDS-polyacrylamide gel electrophoresis and Western blotting were performed as previously described (14). Protein was measured by the method of Lowry et al. (21). NADPH-cytochrome P-450 reductase was prepared from the livers of male Sprague-Dawley rats by the method of Yasukochi and Masters (20) as described by Vermilion and Coon (22).

Enzymatic reduction of cytochrome c was carried out at 28 ‘C using a Beckman DU-50 spectrophotometer. Reactions contained 0.3 M potassium phosphate, pH 7.7, 50 p~ NADPH, and 50 p~ cytochrome c. After a 2-min preincubation at 28 “C, reactions were initiated with either NADPH (liver, pOR263, and Phe-178 and Asp-178 mutants) or cytochrome c (Phe-140, Asp-140, Phe-140/Phe-l78, and Asp-140/ Asp-178 mutants). Assays containing added FMN were preincubated for 4 min. Ferricyanide reactions were carried out in a similar manner in 0.3 M potassium phosphate, pH 7.7, 100 p~ NADPH, and 500 p~ potassium ferricyanide (6). FMN and FAD were measured as de- scribed previously (14). FMN-depleted enzyme was prepared by the method of Vermilion and Coon (22), except that salt exchange was carried out in a Centriprep-30 unit (Amicon), and fluorescence titra- tion was carried out on the FMN-depleted enzyme at 20 “C using an Aminco-Bowman spectrofluorometer with an excitation wavelength of 450 nm and an emission wavelength of 535 nm (22). Amino- terminal sequencing of the purified proteins was done on an Applied

140), 5”TGCTCATCGGTCTTG-3’ (Asp-178), 5”CTCTCCGAAT

Biosystems 477A protein sequencer at the Protein and Nucleic Acid Shared Facility, Medical College of Wisconsin, Milwaukee, WI.

RESULTS AND DISCUSSION

Expression of NADPH-Cytochrome P-450 Oxidoreductase in E. coli.-The cDNA for rat liver reductase was cloned in the expression-secretion vector PIN-111-ompA3 (15, 23) and the protein expressed in E. coli C-1A. The expression plasmid pOR263 consists of the bacterial lpp promoter, the lac pro- moter-operator sequence, and the ribosome binding site and coding sequence for the ompA signal peptide, followed by 24 bases added during the cloning steps and the reductase cDNA sequence (Fig. 1B). These features were confirmed by DNA sequencing. Translation initiated at the ompA ribosome bind- ing site should produce a precursor protein which, when transported to the periplasmic space, will be processed to yield a mature reductase protein with 8 extra amino acids before the amino-terminal methionine of the rat liver enzyme.

Analysis of total cellular proteins by Western blotting using an antibody directed against the rat liver enzyme showed a polypeptide of approximately 80,000 daltons in isopropyl-l- thio-/3-D-galactopyranoside-induced cells transformed by

A

5000 pORFL 2 000 puce

soc I

2000

1000

3000 “ D O 1

7000 pOR 263

1000

r / 6000

B G C T A C C G T A G C G C A G G C C G G A A T T C C C G G G G A T C C G A C C A C AlaThrValAlaGlnAl 1yIleProGlyAspProThrAsnMetGlyAsp T

FIG. 1. Construction of the NADPH-cytochrome P-450 ox- idoreductase expression-secretion plasmid, pOR263. A, plas- mid construction. The cDNA coding for the reductase is indicated by the shaded region and the vector ampicillin resistance (Amp) and lac1 genes by arrows. Coordinates are given in base pairs, and the size of the plasmid in base pairs is shown in parentheses. B, sequence of the amino-terminal region of the bacterially expressed reductase. The sequence begins with the last 6 amino acids of the ompA signal peptide. f indicates the signal peptide cleavage site and is followed by 8 amino acids added during cloning. The amino-terminal sequence of the rat liver reductase is shown in italics.

7586 FMN Binding Domain of NADPH-Cytochrome P-450 Oxidoreductase

FIG. 2. SDS-polyacrylamide gel of purified wild-type and mutant reductase proteins. The gel is stained with Coomassie Brilliant Blue. Lunes 1 and 10 contain purified rat liver reductase (78

pOR263 (80 kDa); lune 3, Asp-178 mutant (81 kDa); lune 4, Phe-178 kDa); lunes 2 and 9, bacterially expressed reductase from cells carrying

mutant (80 kDa); lune 5, Asp-140 mutant (84 kDa); lune 6, Phe-140 mutant (79 kDa); lune 7, Asp-140/Asp-178 mutant (85 kDa); and lune 8, Phe-140/Phe-178 mutant (79 kDa). Lane 11 contains molecular weight standards: phosphorylase B, 92,500; bovine serum albumin, 66,200; and ovalbumin, 45,000.

pOR263 or any of the mutant plasmids. This band was en- riched in the membranes prepared from spheroplasts and could be solubilized with Triton X-100, suggesting that the protein was associated with the inner membrane. Hence, although the ompA signal peptide directs transport of the newly synthesized oxidoreductase protein out of the cyto- plasm, the hydrophobic amino-terminal sequence of the re- ductase, which anchors the rat liver protein to the nuclear envelope and endoplasmic reticulum, prevents its release from the membrane after translocation to the periplasmic side.

Chromatography of Triton X-100-solubilized membranes on 2',5'-ADP-Sepharose yielded a protein which on SDS- polyacrylamide gels migrated as a single band of approxi- mately 80,000 daltons (Fig. 2). The molecular weight calcu- lated from the sequence, including the 8 additional amino acids at the amino terminus is 79,100 daltons. The amino- terminal sequence of the purified protein was found to be: Gly-Ile-Pro-X-Asp-Pro-Thr-Asn-Met-Gly-Asp-Ser-His-Glu- Asp-Thr-Ser-Ala-Thr-Met-Pro-Glu-Val-Ala,2 indicating that the signal peptide had been properly cleaved by the E. coli signal peptidase.

Porter et ul. (14) have previously expressed NADPH-cyto- chrome P-450 oxidoreductase in E. coli using the expression vector pCQV2, in which the reductase cDNA was fused to the cro ribosome binding site and translation initiation site and expressed under the control of the X rightward promoter. The protein expressed in this system was identical with the liver enzyme in flavin content, spectral characteristics, and biolog- ical activity, but was present at a low level because of problems with degradation. To avoid this problem, the expression- secretion vectors PIN-111-ompA were selected. These vectors utilize the ompA signal peptide to direct transport of the expressed protein out of the cytoplasm into the periplasmic space, where proteolytic activity is reduced (23). Although the reductase protein expressed by pOR263 remains associated with the membrane, protection from proteolysis does result.

Mutant proteins were expressed and purified in the same way as the wild-type (pOR263) protein. Fig. 2 shows a Coo- massie Brilliant Blue-stained SDS-polyacrylamide gel of the

*Glycine at position 4 was not established unambiguously by protein sequencing and is designated as X.

purified wild-type and mutant proteins with the apparent molecular weights indicated in the legend. Substitutions at positions 140 and 178 resulted in detectable alterations in the electrophoretic mobilities. Most notable is the substitution of aspartic acid at position 140, in which case the mutant pro- teins migrated with an apparent molecular weight of 84,000 to 85,000 (Fig. 2, lunes 5 and 7). In contrast, replacement of tyrosine-178 with phenylalanine did not produce anomalous migration (Fig. 2, lune 4 ) , and the Phe-140 mutants had only slightly altered electrophoretic mobilities (Fig. 2, lanes 6 and 8). Restriction mapping and sequencing of the Asp-140 mu- tant expression plasmid failed to reveal changes that could account for such a shift in molecular weight. The amino- terminal sequence of the Asp-140 mutant was found to be identical with the wild-type sequence, indicating that the protein had been correctly processed; furthermore, the change in apparent molecular weight was also seen in the trypsinized form of the protein, indicating that alterations in the first 64 amino acids were not responsible. Hence, a most likely expla- nation is that a single amino acid substitution has dramati- cally altered the electrophoretic behavior of the protein in SDS-polyacrylamide gels. Single amino acid substitutions have been reported to alter the apparent molecular weights of subtilisin (24) and the a-crystallin A chain (25) and changes in the apparent molecular weight of cytochrome P-450 have been reported with minor changes in composition (26).

Catalytic Activities-Table I shows the specific activities of the wild-type and mutant enzymes with cytochrome c as the electron acceptor. The bacterially expressed reductase protein catalyzed cytochrome c reduction at a rate of 51.5 pmol/min/ mg of protein, which is similar to that obtained for the rat liver enzyme (53.3 pmol/min/mg of protein) and in the range of values reported previously for rat liver reductase (5, 27). Substitution of phenylalanine at either the 140 or 178 position did not significantly affect the rate of cytochrome c reduction; however, substitution of aspartic acid at either position greatly reduced enzyme activity. The Asp-140 mutant pos- sessed about 20% of wild-type activity, while the activities of the Asp-178 and Asp-140/Asp-178 mutants were less than 1% of wild-type.

In order to further evaluate the transfer of electrons from NADPH to FAD, ferricyanide was used as an acceptor, since ferricyanide reduction involves direct transfer of electrons from FAD and does not require FMN (6, 28). Unexpectedly, the wild-type enzyme as well as the Asp-140 and the three phenylalanine mutants reduced ferricyanide at approximately twice the rate of the liver enzyme (Table I). As a control, an affinity-purified fraction from cells transformed with the PIN- 111-ompA3 vector (no cDNA insert) had <1 pmol/min/mg of ferricyanide reductase activity. The basis for this enhanced

TABLE I Activities of NADPH-cytochrome P-450 oxidoreductase mutants

Enzyme Cytochrome c Ferricyanide

Liver pOR263 Phe-178 Phe-140 ASP-178 ASP-140 Phe-140/Phe-178 ASP-140fAsp-178

pmolfminfmg protein" 53.3 f 2.7 (3)b 52.9 f 2.51 (3) 51.5 f 9.0 (13)b 102 f 7.9 (4) 47.4 f 10.5 (5)b 113 f 22.1 (3) 55.2 f 9.2 (4)' 94.8 f 18.5 (3) 0.44 f 0.36 (6)b 48.1 f 9.5 (4) 11.0 f 3.3 (11)' 102 f 2.5 (3) 46.0 f 3.2 (4)' 94.2 f 9.6 (3) 0.28 f 0.22 (3)' 51.0 f 11.6 (3)

'Mean f S.D. ( N ) . Reactions were initiated with NADPH after preincubation with

cytochrome c.

with NADPH. Reactions were initiated with cytochrome c after preincubation

FMN Binding Domain of NADPH-Cytochrome P-450 Oxidoreductase 7587

activity is unknown, but may reflect a conformational differ- ence that facilitates electron transfer through the FAD site. Both the Asp-178 and Asp-140/Asp-178 mutants exhibited an approximately 50% decrease in ferricyanide reductase ac- tivity which correlates roughly with their reduced FAD levels (see below).

Flavin Content-Table I1 presents the flavin contents of the wild-type and mutant reductase proteins. Like the rat liver reductase, the wild-type enzyme contains 1 mol each of FMN and FAD per mol of protein. Substitution of phenylal- anine for tyrosine at either position 140 or 178 had no effect on either FMN or FAD content. Substitution of aspartate at position 140 had no effect on FMN or FAD levels; however, Asp-178 mutants had (0.04 mol of FMN and approximately 0.5 mol of FAD per mol of protein.

If Tyr-178 of the reductase is in fact a homologue of Tyr- 98 of D. vulgaris flavodoxin as previously suggested (9), Tyr- 178 would be expected to play a major role in the binding of FMN. This expectation is based on the high resolution crys- tallographic analysis of the flavodoxin protein (12) showing Tyr-98 to be co-planar to and in close association with the isoalloxazine ring of FMN. Substitution of a nonaromatic side chain at position 178 would be expected to markedly reduce FMN binding and catalytic activity, but substitution with phenylalanine would preserve the stacking interactions be- tween the protein and the isoalloxazine ring and would have minimal effects on FMN binding and enzymic activity. As expected, the Phe-178 mutant protein was found to contain the normal complement of FMN (Table 11), while the low levels of FMN contained in the Asp-178 and Asp-l40/Asp- 178 mutants indicate that an aromatic residue at position 178 is required for binding of FMN.

It is interesting to note that, in mutants exhibiting loss of FMN binding, a reduction in FAD content is also detected. This observation suggests that proper incorporation of FAD into the protein to form a catalytically active center depends upon the existence of an intact FMN domain. Although little is known about the mechanisms of flavin incorporation, flavin insertion may occur co-translationally, with binding of FMN required for correct FAD insertion. Alternatively, lack of FMN may decrease the affinity of the protein for FAD. This is consistent with the observation that FAD is more easily removed from the rat liver reductase in the absence of FMN (6).

FMN Binding-The ability of FMN to bind to the mutant reductases was determined indirectly by measuring cyto- chrome c reductase activity after preincubation of the FMN- depleted enzyme with varying concentrations of FMN. The rate of cytochrome c reduction was taken as a measure of the amount of reconstituted enzyme. Since the level of FMN in the Asp-178 and Asp-140/178 mutants was below 0.04 mol/ mol of protein, no attempt was made to remove residual FMN prior to reconstitution.

TABLE I1 Flavin contents of NADPH-cytochrome P-450

oxidoreductase mutants Enzyme FMN FAD

pOR263 Phe-178 Phe-140

pmol/pmol protein" 1.1 f 0.2 (10) 0.9 f 0.1 (10) 1.1 f 0.2 (3) 1.0 f 0.1 (3) 1.1 f 0.3 (3) 1.0 f. 0.2 (3)

Asp-178 <0.04 (5) 0.5 f 0.2 (5) Asp-140 Phe-140/Phe-178 0.9 k 0.1 (3)

1.2 f 0.2 (8) 1.2 f 0.2 (8) 1.0 f 0.1 (3) 0.4 f 0.2 (5) Asp-140/Asp-178 C0.04 (5)

Mean f S.D. ( N ) .

Table 111 shows the dependence of cytochrome c reduction on FMN. The concentration of FMN required for half-maxi- mal activity was slightly higher for the FMN-depleted Asp- 140 mutant compared to the wild-type enzyme (0.07 uersus 0.01). Addition of FMN to the Asp-178 and Asp-140/Asp-178 mutants significantly increased the cytochrome c reductase activity, with 340- and 26,500-fold higher FMN concentra- tions required, respectively, for half-maximal activity. For each of these aspartyl mutants, however, maximal activities were no more than 19-22% of the wild-type activity, although these increases did represent a 20- to 26-fold elevation over

TABLE 111 FMN dependence of cytochrome c reduction

The results are an average of two experiments.

Enzyme half-maximal activity reductase activity FMN required for Maximum cytochrome c

P M pmol/min/mg Liver 0.05 47.0 pOR263 0.01 38.5 Asp-178 3.40 8.6 Asp-140 0.07 10.6 As~-140/As~-178 265 1.2

.3 ,

-"I I

W A V E L E N G T H lnd

FIG. 3. Visible absorption spectra of wild-type and mutant reductases. The oxidized (-) and reduced (---) spectra are shown. Protein concentrations were: wild-type, 5.4 pM; liver, 12.6 pM; Phe-178,1.2 PM; Phe-140/Phe-178,6.2 p ~ ; Asp-l78,7.5 PM; and Asp- 140,2.0 p ~ . The semiquinone forms were prepared by adding NADPH to a final concentration of 50 mM and letting the sample equilibrate for 10 min at room temperature. Samples were scanned on a Beckman DU-50 spectrophotometer at room temperature.

TABLE IV K, values for NADPH and cytochrome c

K , Enzyme

NADPH Cytochrome c

P@

Liver 6.6 (2) pOR263 6.4 f 1.0 (3) 21.1 f 2.5 (3) Phe-178 7.1 f 0.3 (3) 14.0 f 4.9 (5) Phe-140 7.8 f 1.2 (4) 18.8 f 2.7 (5) Asp-178' 9.5 f 1.5 (3) 18.5 f 3.8 (4)

Phe-140/Phe-178 11.4 f 1.4 (4) 13.1 f 4.2 (5)

17.8 (2)

ASP-140 6.6 f 0.9 (3) 19.1 f 3.7 (4)

Mean f S.D. ( N ) . Assays were carried out in the presence of 16.5 PM FMN.

7588 FMN Binding Domain of NADPH-Cytochrome P-450 Oxidoreductase

TABLE V Effect of NADPH preincubation o n rate of cytochrome c reduction

Specific activity Enzyme

Ratio NADPH Cytochrome c NADPH/

preincubation" preincubation* cytochrome

prnollrninlrng protein pOR263 Phe-178

53.0 53.1 56.0

1.0 58.4

Phe-140 52.8 34.6 1.0 1.5

<0.7 10.5 2.9

Phe-l40/Phe-178 3.6

46.5 29.9 1.6

ASP-178 C0.7 ASP-140

Asp-140/Asp-178 <0.6 <0.6 "Reactions were preincubated for 2 min at 28 "C with 50 p~

Reactions were preincubated for 2 min at 28 "C with 50 p~ NADPH and initiated by addition of 50 p~ cytochrome c.

cytochrome c and initiated by addition of 50 p~ NADPH.

the activities of the non-FMN-supplemented mutant proteins (Table I). A Kd value for FMN was found by fluorescence titration to be 0.04 WM for the wild-type bacterially expressed enzyme, which compares with a value of 0.01 WM previously reported for the rat liver enzyme (7, 28). The Kd for the Asp- 140 mutant was 0.18 WM.

In contrast to the position 178 mutants, substitution of aspartate at position 140 did not affect FMN content of the enzyme (Table 11) and the affinity of the protein for FMN was only slightly reduced. The residue in D. vulgaris flavo- doxin corresponding to Tyr-140 in the oxidoreductase se- quence, Trp-60, is tilted 45" from the plane of the flavin ring (12). The nearly 100-fold higher FMN concentration required for half-maximal cytochrome c reductase activity of the Asp- 140/Asp-178 mutant relative to the Asp-178 mutant suggests that Tyr-140 also has an important role in FMN binding. However, the major effect of removing the aromatic side chain at position 140 is to decrease the maximal activity of the enzyme, even in the presence of saturating levels of FMN, suggesting that FMN may be bound in a conformation which cannot participate effectively in electron transfer.

Absorption Spectra-The visible absorption spectra of the bacterially expressed and rat liver reductases, as well as the Phe-178, Asp-178, Asp-140, and Phe-140/Phe-178 mutants, are shown in Fig. 3. The spectra of the oxidized liver and bacterial proteins are very similar, with peaks at 383 and 454 nm. Addition of NADPH to either protein followed by air reoxidation resulted in a decrease in the absorption at 454 nm and the appearance of a broad absorption band at 585 nm characteristic of the air-stable semiquinone form. The oxi- dized and reduced spectra of the phenylalanine mutants and the Asp-140 mutant were indistinguishable from that of the wild-type enzyme. The spectrum of the oxidized Asp-178 mutant had a higher baseline and a decreased absorbance at 454 nm compared to wild-type, which is consistent with the decreased FMN and FAD content of the protein. Addition of NADPH produced a transient semiquinone absorption band, but no semiquinone absorption was seen after air reoxidation. This is similar to the spectra obtained for the FMN-depleted rat liver enzyme (7, 28).

Kinetic Properties of the Mutant Reductases-The K, val- ues for NADPH and cytochrome c were determined for the wild-type enzyme and five of the mutant enzymes (Table IV). The wild-type enzyme had K, values of 6.4 ~ L M for NADPH and 21.1 KM for cytochrome c which are similar to those found for the liver enzyme and in the range of values published previously for the liver enzyme (22, 27, 29).

The K, values of the Phe-140, Asp-140, and Asp-178 mu- tants for cytochrome c were similar to wild-type, while the

Phe-178 and Phe-140/Phe-178 mutants had a slightly de- creased K, for cytochrome c. Although the Asp-178 mutant was unable to bind FMN, the structural requirements for cytochrome c binding have been preserved. Position 178 is adjacent to a cluster of negatively charged amino acids (Asp- 207, Asp-208, Asp-209, Glu-213, Glu-214, Asp-215) which may be involved in cytochrome c binding. Nisimoto (10) has pro- posed charge pairing between these residues and lysyl residues of cytochrome c with the FMN isoalloxazine ring and tyro- sine-178 being in close proximity to the heme group for efficient electron transfer. Substitution at position 178 could alter the K, for cytochrome c by altering cytochrome c binding or the rate of electron transfer from flavin to heme.

NADPH K, values for the Asp-140, Phe-140, and Phe-178 mutants were similar to those of wild-type, while the Phe- 140/Phe-178 and Asp-178 mutants had slightly higher NADPH K, values. Considering that the NADPH binding site is located in the carboxyl-terminal region of the protein, it is interesting that mutations within the amino-terminal segment of the enzyme (Phe-140/Phe-178 and Asp-178) re- sulted in perturbation of the K, for NADPH.

Table V shows the effect of preincubation with either co- factor or substrate on the rate of cytochrome c reduction. Preincubation of the wild-type enzyme or the Asp-178 or Phe- 178 mutants with either cytochrome c or NADPH had no effect on the rate of cyrochrome c reduction. However, prein- cubation of the position 140 mutants with NADPH produced a 1.5- to 3.6-fold increase in the rate of cytochrome c reduction over that obtained when the reaction was initiated by addition of NADPH. The proposed reaction cycle for reduction of cytochrome c by reductase includes rate-limiting reduction of FAD by NADPH, followed by rapid intramolecular 1-electron transfers between FAD and FMN, and transfer of electrons from FMNHz to cytochrome c (30, 31). The l-electron-re- duced form of the enzyme, the air-stable semiquinone (FAD/ FMNH.), is unable to reduce cytochrome c in the absence of NADPH. Stopped-flow studies of cytochrome c reduction by Masters et al. (32) demonstrated a lag of 20 to 30 ms after addition of NADPH which was interpreted as the time re- quired for formation of the FMNHz form of the enzyme. The slightly increased K, for NADPH seen in the Phe-l4O/Phe- 178 mutant and the requirement for preincubation with NADPH seen in both the Asp-140 and Phe-140 mutants may be due to either decreased affinity of the enzyme for NADPH and/or a decrease in the rate of FAD reduction by NADPH, or a shift in the rate-limiting step from reduction of FAD by NADPH to the transfer of electrons from FAD to FMN. Since the position 140 mutants reduce ferricyanide at the same rate as the wild-type enzyme, it is likely that the rate of reduction of FAD by NADPH is unchanged. A decrease in the rate of formation of the electron-donating species (FMNH2) in the mutants, as a result of a change in the redox potential of the bound FMN or a requirement for a tyrosyl residue for electron transfer, could produce the changes observed here.

In summary, we have altered, by site-directed mutagenesis, two putative FMN-binding residues in the enzyme NADPH- cytochrome P-450 oxidoreductase and have shown that Tyr- 178 is required for binding of FMN to the enzyme. Replace- ment of Tyr-178 by aspartate virtually abolishes FMN binding and cytochrome c reductase activity. Replacement of another putative flavin-binding residue, Tyr-140, by aspartate does not significantly affect the amount of FMN bound, but cata- lytic activity is reduced about &fold. Although FMN does bind to the Asp-140 mutant enzyme, it may be bound in a conformation which does not permit efficient electron trans- fer from either FAD to FMN or from FMN to substrate.

FMN Binding Domain of NADPH.

Substitution of phenylalanine for tyrosine at either position 140 or 178 does not affect FMN binding or catalytic activity. The effects of these substitutions on FMN binding and cata- lytic activity are in accord with those predicted on the basis of amino acid sequence comparisons of the reductase with FMN-containing proteins of known structure (9). Although the most dramatic changes in these mutants occurred in the FMN-binding properties, other alterations in the properties of the reductase were noted. Substitution of phenylalanine at position 178 slightly decreased the K,,, for cytochrome c, while the Asp-178 and Phe-140/Phe-178 mutants had a slightly increased K,,, for NADPH. Position 140 substitutions either altered the interaction with NADPH or the kinetics of the reaction cycle, as evidenced by an enhanced enzymic activity upon preincubation with NADPH. Furthermore, loss of FMN binding (Asp-178 and Asp-140/Asp-178 mutants) surprisingly resulted in reduced incorporation of FAD into the newly synthesized mutant protein. The basis for these interesting changes will be the subject of future investigations.

Acknowledgments-We are grateful to Dr. Masayori Inouye for the PIN-111-ompA3 vector, to Lorelei Manney and Susan Zjaba for tech- nical assistance, and to Kristen Adler for preparation of this manu- script.

REFERENCES 1. Masters, B. S. S. (1980) in Enzymatic Basis of Detoxification

2. Lu, A., and Coon, M. (1968) J. Biol. Chem. 243,1331-1332 3. Ernster, L., Lind, C., Nordenbrand, K., Thor, H., and Orrenius,

S. (1982) in Oxygenases and Oxygen Metabolism (Nozaki, M., Yamamoto, S., Ishimura, Y., Coon, M., Ernster, L., and Esta- brook, R., eds) pp. 357-369, Academic Press, Orlando, FL

4. Iyanagi, T., and Mason, H. S. (1973) Biochemistry 12,2297-2308 5. Vermilion, J., Ballou, D., Massey, V., and Coon, M. (1981) J.

6. Kurzban, G., and Strobel, H. (1986) J. Biol. Chem. 261 , 7825-

(Jakoby, W., ed) pp. 183-200, Academic Press, Orlando, FL

Biol. Chem. 266,266-277

7830

-Cytochrome P-450 Oxidoreductase 7589

7. Iyanagi, T., Makino, R., and Anan, F. (1981) Biochemistry 2 0 ,

8. Black, S., French, J., Williams, C., and Coon, M. (1979) Biochem.

9. Porter, T., and Kasper, C. (1986) Biochemistry 26, 1682-1687 10. Nisimoto, Y. (1986) J. Biol. Chem. 2 6 1 , 14232-14239 11. Porter, T., and Kasper, C. (1985) Proc. Natl. Acad. Sci. U. S. A.

1722-1730

Biophys. Res. Commun. 91, 1528-1535

82,973-977

Acad. Sci. U. S. A. 7 0 , 3857-3860

Y. (1984) J . Biol. Chem. 259, 2480-2483

12. Watenpaugh, K., Sieker, L., and Jensen, L. (1973) Proc. Natl.

13. Nisimoto, Y., Hayashi, F., Akutsu, H., Kyogoku, Y., and Shibata,

14. Porter, T., Wilson, T., and Kasper, C. (1987) Arch. Biochem.

15. Ghrayeb, J., Kimura, H., Takahara, M., Hsiung, H., Masui, Y.,

16. Vieira, J., and Messing, J. (1982) Gene (Amst.) 19 , 771-780 17. Chang, A., and Cohen, S. (1978) J. Bacteriol. 134,1144-1150 18. Zoller, M., and Smith, M. (1984) DNA ( N Y ) 3,479-488 19. Saneer. F.. Nicklen. S.. and Coulson. A. (1977) Proc. Natl. Acad.

Biophys. 264,353-367

and Inouye, M. (1984) EMBO J. 3,2437-2442

. . Ss . U. S. A. 7 4 , 5463-5467

20. Yasukochi, Y., and Masters, B. (1976) J. Biol. Chem. 261,5337- 5344

(1951) J. Biol. Chem. 193, 265-275

2704

21. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

22. Vermilion, J., and Coon, M. (1978) J. Biol. Chem. 253 , 2694-

23. Duffaud. G.. March. P.. and Inouve. M. (1987) Methods Enzvmol. ~~ , I " , . .

153,4921507 24. Carter. P.. and Wells. J. (1987) Science 237.394-399 25. de Jong, W., Zweers, A., and Cohen, L. (1978) Biochem. Biophys.

26. Levin, W., Shively, J., Yuan, P., and Ryan, D. (1984) Biochem.

27. Dignam, J., and Strobel, H. (1977) Biochemistry 16 , 1116-1123 28. Vermilion, J., and Coon, M. (1978) J. Biol. Chem. 2 5 3 , 8812-

29. Nadler, S., and Strobel, H. (1988) Arch. Biochem. Biophys. 261,

30. Oprian, D., and Coon, M. (1982) J. Biol. Chem. 257,8935-8944 31. Yasukochi, Y., Peterson, J., and Masters, B. (1979) J. Biol. Chem.

32. Masters, B., Kamin, H., Gibson, Q., and Williams, C. (1965) J.

Res. Commun. 82,532-539

SOC. Trans. 12,62-68

8819

418-429

254,7097-7104

Biol. Chem. 240,921-931