Reconstitution of the Fatty Acid Hydroxylation Function of Cytochrome P-450BM-3 Utilizing Its Individual Recombinant Hemo- and Flavoprotein Domains*

Cytochrome P-46oBM.S is a catalytically self-suffi-cient fatty acid w-hydroxylase with two domains. Functional and primary structure analyses of the hemo- and flavoprotein domains of cytochrome P-46oBM.3 and the corresponding microsomal cytochrome P-450 system have shown that these proteins are highly homologous. acid of the fatty acid hydroxylation activity of cytochrome P-460BM.S the recombinantly its hemoprotein counterpart. The rate of fatty acid-dependent oxygen consumption was shown to be linear when increas-ing concentrations of the hemoprotein domain are added to a fixed concentration of the flavoprotein domain and vice versa. The combination of the hemo-and flavoprotein domains in a ratio of 20:l respec-tively, in the reaction mixture, that the mixtures of regioisomers were used instead of the purified products. This method was used to obtain the relative ratio of regioisomers in a mixture, based on the intensity of each molecular ion. By taking the sum of the intensities as 100, the relative percentage of each isomer in the mixture was determined (25).

Cytochrome P-46oBM.S is a catalytically self-sufficient fatty acid w-hydroxylase with two domains. Functional and primary structure analyses of the hemo-and flavoprotein domains of cytochrome P-46oBM.3 and the corresponding microsomal cytochrome P-450 system have shown that these proteins are highly homologous. Prior attempts to reconstitute the fatty acid hydroxylation function of cytochrome P-450B~.s, utilizing the two domains, obtained either by trypsinolysis or by recombinant methods, were unsuccessful. In this paper, we describe the reconstitution of the fatty acid hydroxylation activity of cytochrome P-460BM.S utilizing the recombinantly produced flavoprotein domain (Oster, T., Boddupalli, S . S., and Peterson, J. A. (1991) J. Biol. Chem. 266, 22718-22726) and its hemoprotein counterpart. The rate of fatty acid-dependent oxygen consumption was shown to be linear when increasing concentrations of the hemoprotein domain are added to a fixed concentration of the flavoprotein domain and vice versa. The combination of the hemoand flavoprotein domains in a ratio of 20:l respectively, in the reaction mixture, results in the transfer of 80% of the reducing equivalents from NADPH for the hydroxylation of palmitate at 25 OC. The ratio of the regioisomeric products obtained for lauric, myristic, and palmitic acids was similar to that obtained with the holoenzyme form of cytochrome P-4So~M. 3. The reconstitution of the fatty acid w-hydroxylase activity, using the soluble domains of cytochrome P-460BM-3, without added factors such as lipids, may be useful for structure/function comparisons to their eukaryotic counterparts.
Microsomal cytochromes P-450 and their flavoprotein reductase counterpart have been shown to be involved in the oxidative metabolism of a wide spectrum of substrates including foreign chemicals, endogenous fatty acids, vitamins, steroids, eicosanoids, etc. (1, 2). The cytochrome P-450 gene (CUP)' superfamily consists of an indeterminate number of members; approximately 150 genes have been sequenced to date from organisms ranging from bacteria to humans (3). This superfamily of enzymes can be divided into two classes based on the proximal electron donor for the cytochrome P-450. The first class (Class I) can be arbitrarily described as those enzymes which utilize an iron-sulfur protein as an electron shuttle between an FAD-containing flavoprotein reductase and the cytochrome P-450. The second class (Class 11) of cytochromes P-450 are those which require a unique flavoprotein reductase, NADPH-cytochrome P-450 reductase, containing both FAD and FMN in an equimolar ratio.
Extensive biophysical as well as functional studies have been carried out on the reconstituted camphor monooxygenase system of Pseudomonas putida (4)(5)(6)(7). This Class I cytochrome P-450 system is composed of a hemoprotein (cytochrome P-45OCam), a flavoprotein (putidaredoxin reductase), and an iron-sulfur protein (putidaredoxin). The only high resolution structure of a cytochrome P-450, available to date, is that of cytochrome P-450,- (8). However, the differences between Class I and Class I1 monooxygenase systems indicate that the reconstituted cytochrome P-450,., system may not be the most appropriate model for understanding the enzymatic properties of the microsomal cytochrome P-450-dependent monooxygenase system (9).
The soluble fatty acid w-hydroxylase of Bacillus negaterium, uiz. cytochrome P-450BM.? (Mr 119,OOO), is a catalytically self-sufficient single polypeptide containing heme, FAD, and FMN in an equimolar ratio and requires NADPH as the source of electrons (10)(11)(12)(13)(14). Narhi and Fulco (15) demonstrated that upon limited trypsinolysis, the holoenzyme is cleaved into two portions, one (BMP) (Mr 55,000) retains the heme and binds the substrate and the other (BMR) (Mr 66,000) contains the flavins and carries out an NADPHdependent reduction of cytochrome c. Analysis of the deduced amino acid sequence of cytochrome P-450BM.3 has revealed that this protein contains regions that strongly resemble microsomal o-hydroxylases and NADPH-cytochrome P-450 oxidoreductase (16). Reconstitution of the fatty acid hydroxylation function of cytochrome P-45oBM.3 utilizing either the purified tryptic products (15) or the recombinantly isolated domains (17) has been unsuccessful thus far. The present The nomenclature of cytochromes P-450 is adapted from Nebert et al. (3).
The abbreviations used are: cytochrome P-45bM. 3 paper describes the conditions required to express and purify the hemoprotein domain and the reconstitution of the fatty acid hydroxylation activity using the recombinantly obtained hemo-and flavoprotein domains of cytochrome P -4 5 0~,~.~.
Subcloning of the DNA Region Encoding the Hemoprotein Domain-After isolation by conventional techniques (22), plasmid BM3.zA was digested with BglII and SalI. The resulting 3.2-kilobase pairs fragment, encoding the hemoprotein domain and a portion of the NHz-terminal region of the flavoprotein domain of cytochrome P -4 5 0~~.~, was purified and subcloned into the pIBI24 vector digested with BamHI and SalI as illustrated in Fig. 1. Using a 26-mer primer (5"GGG AAT TCT GAA AGA GGG ATA ACA TG-3'), the 3'-end of which corresponded to the region encoding the NH, terminus of the hemoprotein domain, and another primer matching the region around the PstI site, PCR amplification was performed to synthesize a DNA fragment containing a unique EcoRI site at its 5' end, translation initiation signals, and the unchanged 5"region encoding the hemoprotein domain of cytochrome P -4 5 0~~3 (BMP). The plasmid pP450-1 was digested with EcoRIINarI, and the fragment, corresponding to the vector and the 3' end of the cytochrome P-450 coding region, was purified by agarose gel electrophoresis. The PCRamplified fragment of DNA (153 base pairs), which had been EcoRI/ NarI-digested, was ligated into the purified vector containing the 3' region coding for the hemoprotein domain. This step removed the A second PCR amplification was performed using a primer, corresponding to the region around the PstI site, and a 27-mer primer (5'-GGC TCG AGT TAG CGT ACT TTT TTA GCA-3'), the 3'-end of which was complementary to the region encoding the COOH terminus of the hemoprotein domain. This PCR amplification synthesized a DNA fragment of 851 base pairs containing a stop codon to terminate the translation of the hemoprotein domain and a unique XhoI site. The plasmid, pP450-2, was digested with PstI/XhoI, and the fragment containing the vector and the 5' coding region of cytochrome P- 4 5 0~~. 3 was ligated with the DNA fragment for the 3' end prepared by PCR amplification. Subsequently, clones were selected after small scale DNA purification (23) by restriction analysis. The nucleotide sequence of the DNA fragments obtained by PCR amplification was determined to verify the presence of the desired changes.
The plasmid pP450-3 was used to transform E. coli DH5a competent cells. The cells were grown for 24 h in 2xYT medium containing 50 pg/ml ampicillin at 37 "C which provided the optimal level of expression of BMP.
Lysis of E. coli Cells Containing Expressed BMP-Cells grown to late stationary phase, as described above, were harvested by centrifugation and washed in 20 mM MOPS buffer, pH 7.4, containing 20 mM KC]. The pellet was frozen at -70 "C and then thawed at 4 'C. Freeze-thaw cycles were performed four times, and the cell paste was resuspended as 1 g/4 ml in 20 mM MOPS buffer, pH 7.4, containing 20 mM KCl, 0.1 mM Na,EDTA, and 0.1 mM DTT (buffer A). The cell suspension was treated with lysozyme as previously described (24). After centrifugation of the cell lysate at 40,000 X g for 30 min, the supernatant solution was transferred to a fresh tube. Under these conditions, approximately 75% of BMP in the whole cells was recovered in the supernatant solution. Subsequent purification steps were performed at 0-4 "C, as described in detail under "Results." Oxygen Uptake Determinations-Oxygen consumption was measured with a Gilson 5/6 Oxygraph (Middleton, WI) with an enclosed 2-ml reaction chamber. Stock solutions of the fatty acids (25 mM) were prepared in 50 mM potassium carbonate. The reaction system for performing the reconstitution experiments typically contained 50 mM MOPS buffer (pH 7.4), the desired concentrations of the enzymes (BMP and BMR), and the fatty acid substrate. After a preincubation of reactants for 2 min at room temperature, oxygen consumption was initiated by the addition of NADPH.
Analysis of Fatty Acid Metabolites-The reaction system for monooxygention of fatty acids contained (unless otherwise mentioned) 50 mM MOPS, pH 7.4, 500 p~ fatty acid, BMP, and BMR, and 1 mM NADPH. The substrate was always added first followed by BMP and BMR. After preincubation of the enzymes for about 5 min, the reaction was started by the addition of the reduced pyridine nucleotide. Aliquots of the reaction mixture (0.5 ml) were taken, and the reaction was terminated with 0.1 ml of 1 N HCl. The solutions were extracted twice with 5 ml of ethyl acetate. The combined extracts were washed with 2 ml of water and evaporated with nitrogen gas. The residue was taken up in methanol, and a portion was subjected to reverse phase HPLC on an analytical pBondpak CIS column (30 X 1 cm, Waters Associates). The radioactivity of the eluate was monitored with a radioactive flow detector (Flo-One, Model HP, Radiometrics, Meridian, CT). Metabolites were eluted using a gradient of 50-100% acetonitrile (prepared in 0.1% acetic acid).
Methyl esters of each of the compounds were prepared using an ethereal solution of diazomethane. Diazomethane was prepared using the Mini Diazald apparatus obtained from Aldrich. An aliquot of the metabolites in methanol was treated with a few drops of diazomethane, and after 5 min at room temperature, the sample was evaporated to dryness under a stream of nitrogen and dissolved in a small volume of methanol for mass spectral analysis. Mass spectral analyses of the metabolites were performed by published procedures Gas chromatography followed by chemical ionization was performed for the methyl esters of the metabolites of fatty acids, except that the mixtures of regioisomers were used instead of the purified products. This method was used to obtain the relative ratio of regioisomers in a mixture, based on the intensity of each molecular ion. By taking the sum of the intensities as 100, the relative percentage of each isomer in the mixture was determined (25). (25). coli that were transformed with pP450-3 revealed a protein (M, 55,000) at a much higher level than those of the mocktransformed cells. The protein was specifically detected in immunoblots probed with anti-BMP antiserum and was estimated to correspond to approximately 30% of the soluble proteins in E. coli. This quantitation was in excellent correlation with the content of the cytochrome determined by difference spectrophotometry by measuring the absorbance of the carbon monoxide complex of the ferrous form minus the ferrous form of BMP in cell lysates.

Construction
Purification of BMP-The cell lysates of E. coli that expressed BMP, obtained as described under "Experimental Procedures," were fractionated with ammonium sulfate. The fraction which precipitated between 30 and 60% of saturated ammonium sulfate contained BMP. This precipitate was re-suspended in buffer A and dialyzed for 4 h against a 300-fold excess of buffer A. The dialysate was then chromatographed on a 5 X 25-cm DE52 ion-exchange column, previously equilibrated with buffer A. The resin was washed with 5 bed volumes of buffer A, and the enzyme was eluted with a 50-400 mM linear gradient of KC1 in buffer A. Most of the extraneous proteins were removed by the initial wash, and BMP eluted as a single peak at an ionic strength of approximately 250 mM potassium chloride. The elution of BMP was monitored at 418 nm, which is characteristic of the oxidized, low-spin form of cytochrome P-450. Fractions of 2.5 ml were collected, and those having a specific content of at least 9 nmol of heme per unit of absorbance at 280 nm were pooled and concentrated by ultrafiltration.
The protein was purified to homogeneity by size exclusion chromatography on a 500-ml Ultragel ACA 44 in 20 mM MOPS, 100 mM KC1, 0.1 mM dithiothreitol, and 0.1 mM EDTA. An aliquot corresponding to approximately 1 pg of total protein was taken from each fraction that had a specific cytochrome P-450 content of 15 nmol/unit absorbance a t 280 nm and subjected to SDS-polyacrylamide gel electrophoresis.
Fractions showing a single band after silver staining were pooled and concentrated as described earlier. In a typical purification scheme, approximately 3 pmol of homogeneous BMP (equivalent to 150 mg) starting from 50 g wet weight of E. coli cells was obtained. The NH2-terminal sequence of purified BMP (residues Thr-1 to Pro-8) was identical to that The absorbance spectra of oxidized BMP were found to be similar to those of the native flavocytochrome except for differences which were accountable to the absence of both FAD and FMN that are present in the intact cytochrome P-450~"~. Oxidized BMP had absorbance maxima at 418 nm and well resolved a and @ bands at 570 and 535 nm, respectively. Upon addition of the substrate, potassium palmitate (25 p~) , BMP displayed an absorbance spectrum typical of a mixture of both high and low spin cytochrome P-450 with most of the heme being in the high spin state (26). An absorbance maximum was seen at 396 nm with a broad band appearing from 516 to 526 nm. Addition of sodium dithionite to BMP, in the presence of potassium palmitate, resulted in a shift of the 396-nm absorbance maxima to 409 nm. Bubbling the sodium dithionite-treated enzyme with carbon monoxide resulted in a further shift of the absorbance maximum with peaks at 448 and 548 nm. Hence, the spectral characteristics of the purified, recombinantly produced hemoprotein domain are similar in all respects to those of cytochrome P-450BM.3, except for those attributable to the presence of flavins in the native flavocytochrome.
Substrate Binding Constants-The equilibrium association constants for substrate binding to the enzymes were calculated by analyzing the spectral changes in the Soret region observed during the conversion of the low spin form to the high spin form (6).
A comparison of these constants for various fatty acid substrates is shown in Table I. Of the substrates tested, the association constant for palmitic and pentadecanoic acids was found to be the highest for both the proteins while it was the lowest for lauric acid. The association constants of cytochrome P"i5oBM.3 were found to be uniformly higher than those seen with BMP. The cause for this change remains to be determined.
Reconstitution of the Fatty Acid Hydroxylase Activity of Cytochrome P -4 5 0~~.~-W e had earlier reported that the purified reductase domain of cytochrome P-45oBM.3 had a turnover number of 1600 min-' for NADPH-dependent cytochrome c reduction (27). T o verify whether BMR could carry Of cytochrome P -~~O B M .~ (13).

TABLE I
Comparison of the fatty acid-dependent equilibrium association constants of cytochrome P -4 5 0~~. 3 and BMP To obtain comparable results, the relative concentrations of cytochrome P -4 5 0~~. 3 and BMP in the reaction system were maintained constant (10-12 PM heme). Titrations were performed at 25 "C. Enzymes were diluted in 3 ml of 50 mM MOPS (pH 7.4) buffer to obtain the desired concentration, and the spectra were recorded after adding aliquots of fatty acids. Stock solutions of the fatty acids were prepared in 50 mM potassium carbonate. The equilibrium constants were calculated using a method similar to that described by Peterson (6). For each experiment, the reciprocal of the change in absorbance, for each addition of fatty acid, was plotted uersw the reciprocal of the fatty acid concentration. The intercept on the y axis was taken as the inverse of the maximum absorbance change for each experiment. This value was used to calculate the molar absorbance change for the conversion of substrate-free to substrate-bound enzyme. Using this molar absorbance and the total concentration of enzyme and fatty acid, the concentration of E, ES, and S could be calculated. The out effective electron transfer to its natural redox partner (BMP), oxygen consumption in the presence of an excess of BMP, reduced pyridine nucleotide, and palmitic acid was measured. The reaction was found to be linear with the concentration of BMP added in the reconstituted system. These experiments were repeated with various concentrations of BMR (0.25-5 PM) and BMP (2.5-30 PM) and the rate of oxygen consumption measured. Fig. 2 is a plot of the rate of oxygen consumption versus the product of the concentrations of BMR and BMP.
Similar reconstitution experiments with the two domains were performed to monitor the oxidation of NADPH spectrophotometrically at 340 nm. A plot of the initial rate of oxidation of the reduced pyridine nucleotide was linear with the concentration of BMP in the reaction system. Also, the initial rate of NADPH oxidation was similar to the rate of oxygen consumption seen with equivalent concentrations of BMR and BMP. Similar reconstitution experiments, performed with the purified tryptic product of BMR lacking the amino-terminal 122 residues (27), elicited neither uptake of oxygen nor oxidation of NADPH, confirming that the FMNbinding domain of BMR is essential for successful electron transfer to the cytochrome P-450 domain.
The rate of substrate-dependent oxygen uptake catalyzed by cytochrome P-45oBM.3 is dependent upon the chain length of the fatty acid metabolized (13,24). The rate of oxygen uptake in the reconstitution system was found to be maximal for palmitic and pentadecanoic acids and minimal for lauric acid and were directly dependent upon the concentration of BMP in the reaction system. With a ratio of BMR to BMP of 1:7.5, the turnover number for palmitic acid-dependent oxygen uptake was 180 mol of oxygen/min/mol of BMR. However, the rate of the reaction increased linearly with the concentration of BMP present in the reaction mixture. In view of this, a constant turnover number for the substratedependent oxygen uptake per mol of BMR could not be obtained.
Metabolism of Palmitic Acid-Analysis of the amount of oxygen and NADPH consumed and product formed during the oxidation of palmitic acid to monohydroxy palmitates by the reconstituted system showed that the ratio of the concentration of hydroxylated products formed to that of oxygen consumed was less than 1 ( Table 11). The extent of leakage of electrons was found to be independent of the concentration of BMP in the reconstituted system (Table 11). Other experiments, not shown here, have indicated that the excess oxygen consumed, which is not incorporated into the hydroxylated fatty acid forms hydrogen peroxide. Whether this hydrogen peroxide is formed via the direct two-electron reduction of molecular oxygen or by the intermediate formation of superoxide anion has not been determined. The extent of coupling between the oxygen consumed and substrate oxidized varied with the ratio of BMP to BMR, but under optimal conditions the percentage incorporation into hydroxylated product was observed to be as large as 80% (Table 11). In contrast, the ratio of oxygen consumed to hydroxylated product formed during the metabolism of palmitic acid (500 FM) by cytochrome P-45oBM.3 was always 1 with no detectable hydrogen peroxide formed during the reaction (24).
During the metabolic studies performed with cytochrome P-45OBM-3, we determined the relative abundance of the isomers of the monohydroxylated fatty acids by subjecting the isomeric mixture of the methyl esters to GC followed by chemical ionization mass spectrometry (25). In the reconsti-

I1
Stoichiometry between the oxygen consumed and product formed during palmitate hydroxylation by the reconstituted enzyme system in 50 mM MOPS, pH 7.4, buffer at 25 "C. After preincubating BMP The reaction was carried out in an oxygen electrode vessel (2 ml) (at the concentrations indicated) with [l-'4C]potassium palmitate (600-900 cpm/nmol) for 1 min, BMR was added to the reaction mixture. After a minute, the reaction was initiated with NADPH (1 mM final concentration), and, after 5 min, the amount of oxygen consumed per ml of reaction mixture was recorded. The reaction was stopped, after 5 min, by the addition of 1 ml of 1 N HCl, and the reactants were extracted and subjected to reverse phase HPLC as described under "Experimental Procedures." The ratio is the amount of monohydroxypalmitic acid produced to the amount of oxygen consumed during the reaction. tuted system, the fatty acid was hydroxylated at 0-1, w-2, and w-3 positions in a ratio similar to that observed during oxidation by the holoenzyme (25). During the course of the reaction the ratio of the metabolites remained unaltered significantly. In addition, the ratio of the monohydroxypalmitates did not change as a function of the concentration of the domains and was comparable with that observed with cytochrome P-450BM.3. The preferred carbon for hydroxylation, like the holoenzyme (25), was dependent on the chain length of the fatty acid substrate. While lauric and myristic acids were hydroxylated at the w-1 position, palmitic acid was hydroxylated predominantly on the w-2-methylene group.

DISCUSSION
The objective of these studies was to demonstrate the functionality of the recombinant domains of cytochrome P-450BM.3 by reconstituting the monooxygenation activity seen with the holoenzyme. The intrinsic membrane binding properties of the eukaryotic enzymes, their tendencies to form oligomeric forms, and the requirement for the presence of exogenous lipid in the reaction system have hindered detailed studies on the structure and function of the individual enzymes and mechanism of electron transfer between the microsomal cytochromes P-450 and their reductase counterpart. While cytochrome P-450,,, has proven to be an extremely useful model for the Class I cytochromes P-450, it is not clear yet whether extrapolations from cytochrome P-450,., to microsomal cytochromes P-450 are equally valid in all instances.
Initially studies on cytochrome P -4 5 0~~.~ were important because this enzyme represented a different type of soluble cytochrome P-450, which had been isolated from a Bacillus rather than a catabolic Pseudomonad (28)(29)(30). With the recognition that cytochrome P-45oBM.3 was a catalytically selfsufficient enzyme, which was structurally and functionally analogous to the eukaryotic microsomal cytochromes P-450 and the NADPH-cytochrome P-450 oxidoreductase (16,17,25,27), avenues for utilizing cytochrome P-45oBM-3 as a soluble model system for closely examining the structure/function relationship of the Class I1 type of cytochrome P-450 systems were opened.
Early studies of cytochrome P -4 5 0~~. 3 established that this enzyme would hydroxylate fatty acids and that this reaction had one of the highest turnover numbers of any cytochrome P-450 (14, 24). Narhi and Fulco (13) demonstrated that the holoenzyme was sensitive to trypsin cleavage and that two proteins were formed following trypsinolysis. Although the heme-binding domain had spectrophotometric properties which were similar to those of the holoenzyme, it could not be recombined with the reductase domain (purified from tryptic digests) to reconstitute fatty acid monooxygenase activity. More recently, Li et al. (17) reported the expression of both the flavoprotein and hemoprotein domains of cytochrome P-450BM-3, and while they also exhibit activities which are associated with the individual domains, they were not successful in reconstituting the monooxygenase activity of We have previously reported the heterologous expression of the reductase domain of cytochrome P-45oBM.3 at a level of over 50% of E. coli soluble proteins and characterized this recombinant flavoenzyme (27). Consistent with the similarities in the primary structures (16), the properties of the flavoprotein domain were remarkably comparable with those of liver microsomal cytochrome P-450 reductase. Also, in agreement with the hypothesis that the NH2-terminal region of the flavoprotein participates in the binding of FMN (31), removal of the NHz-terminal 120 residues by limited tryp-cytochrome P -~~O B M .~. sinolysis of the reductase domain abolished binding of the mononucleotide to the protein as well as activities associated with it (27). Therefore, we felt that the use of trypsin to prepare the individual domains of cytochrome P-45oBM.3 may lead to undesirable side effects (as described above) and would hinder the progress of our studies on the mechanism of protein/protein interaction between the two domains.
In our current studies, we have constructed a plasmid which contains the coding sequence for the hemoprotein domain of cytochrome P-450BM.3 and have used this plasmid to transform E. coli strain DH5a. Like cytochrome P"i50BM-3 and the recombinant reductase domain, this protein is expressed at a high level without the addition of either inducers like isopropylthio-P-D-galactose or precursors for the biosynthesis of the cofactors to the cells, and the recombinant enzyme can be readily purified to homogeneity in large amounts. The spectral properties of the purified hemoprotein domain are similar to those of the holoenzyme with the exception of contributions to the spectra attributable to the presence of both FAD and FMN in the holoenzyme. The equilibrium association constants for fatty acid binding to either cytochrome P-450eM.3 or the hemoprotein domain were changed only slightly (Table   Although we and others were unable to obtain reconstitution of the activity of cytochrome P-450BM.3 with the individual domains which had been cleaved by trypsin, our preparations of the recombinant forms of the individual domains are fully active in fatty acid monooxygenation. A plot of the rate of oxygen consumption versus the product of the concentrations of the hemoprotein and flavoprotein domains was linear (Fig. 2). We have interpreted this result to indicate that in the reconstituted enzyme system, the rate-limiting step is the productive collision of the individual domains to form an enzymatically active heterodimer. Since the reaction was performed in the presence of potassium palmitate, the cytochrome P-450 domain is initially in the high spin substratebound form. The solution also contains NADPH so the flavoprotein domain is reduced at least to the three-electronreduced form. Thus, when these domains collide in the proper orientation for the transfer of the first electron required for the monooxygenation reaction, the remainder of the reaction cycle of cytochrome P-450, that is oxygen binding, transfer of the second electron and product formation, will occur prior to separation of these reaction partners.
In summary, we have described in this communication the construction of a plasmid encoding the hemoprotein domain of cytochrome P -~~O B M .~ and the expression of this domain in E. coli. The purification of this domain in large quantity has enabled us to study the reconstitution of these domains in the monooxygenation of fatty acids. The properties of the domain in binding fatty acid substrates and the products formed by the reaction are similar to those of the holoenzyme. Thus, the availability of this enzyme system and its domains in highly pure form, which is catalytically active in the oxidation of fatty acids, will provide us with an additional tool for comparison with eukaryotic microsomal cytochromes P-450. In this context, we have recently crystallized BMP, and x-ray diffraction data to a resolution better than 2.OA were also ~b t a i n e d .~ I).