Identification of the Binding Site on Cytochrome P450 2B4 for Cytochrome b 5 and Cytochrome P450 Reductase*

A model of cytochrome P450 2B4, which was constructed by homology modeling with the four known crystal structures of the cytochromes P450 (Chang, B., A., Bridges, and Waskell, L. (1997) Protein Eng. 10, was used to select amino acids predicted, by computer docking studies and numerous previous biochemical and site-directed mutagenesis studies, to be involved in binding the heme domain of cytochrome b 5 . Twenty-four amino acid residues located on both the distal and the proximal surface of the molecule were chosen for mutagenesis. These 24 mutant proteins were expressed in Escherichia coli , purified, and characterized with re-spect to their ability to bind cytochrome b 5 and support substrate oxidation. Seven mutants, R122A, R126A, R133A, F135A, M137A, K139A, and K433A, all on the proximal surface of cytochrome P450 2B4 near the heme ligand, were identified that exhibited decreased ability to bind cytochrome b 5 . All of the mutants except K433A are located in either the C or C* helices or their termini. In addition, three original contact regions obtained using this method at low reso- lution (6.4–6.8 Å) have been described (1). In the studies reported here, this same method has been used for the CYP2B4-cyt b 5 pair but at the higher resolution of 3.4 Å. The results are more accurate than those from the previous 6.5-Å simulation but still can tolerate local inaccu- racies in atomic details. The 10 lowest energy docking positions were analyzed and the results used to select preferred binding regions on CYP2B4 for cyt b 5 .

where RH is the substrate and ROH is the oxidized product. The enzymatic cycle includes substrate binding, first electron transfer, oxygen binding, second electron transfer, substrate oxidation, and finally product dissociation. The redox partners for the microsomal P450s are cytochrome P450 reductase (P450 reductase) which contains both a FAD and FMN cofactor and cytochrome b 5 (cyt b 5 ). The crystal structure of P450 reductase has recently been published, and the two domains of the enzyme have been individually expressed and characterized (5,6). In contrast, the crystal structure of cyt b 5 has been known for many years but has just recently been refined (7,8). The first and second electrons are donated to P450 by P450 reductase. Because of its redox potential (Х ϩ 25 mV), cyt b 5 can only donate the second electron to P450 (9). In fact, it has been suggested that cyt b 5 may be able to transfer the second electron to selected P450s even faster than P450 reductase, thereby decreasing the amount of superoxide produced (10 -12).
In an attempt to understand more thoroughly the in vitro and in vivo regulation and mechanism of reduction of P450 by its redox partners, the functional binding site on cytochrome P450 2B4 (CYP2B4) for cyt b 5 and P450 reductase has been identified. A model of the microsomal CYP2B4 has been constructed based on the previously determined crystal structures of four bacterial P450s (1), and residues on both the proximal and distal face of CYP2B4 have been mutated and characterized. Herein, we report that, as predicted, the site for binding cyt b 5 and P450 reductase is located on the proximal surface of CYP2B4 where the heme comes closest to the surface (3,13,14). lose was purchased from Whatman (Maidstone, UK). DEAE-Sepharose Fast Flow resin was obtained from Amersham Pharmacia Biotech (Uppsala, Sweden). CYP2B4 B0 cDNA was kindly supplied by Dr. R. M. Philpot (NIEHS, Research Triangle Park, NC) (15). pCWOriϩ plasmid DNA was the gift of Professor M. Waterman (Vanderbilt University School of Medicine, Nashville, TN) (16).
Construction of the Plasmids Used to Express CYP2B4 and Its Mutants-Unless otherwise specified, all DNA manipulations were performed as described (17). All mutations were confirmed by doublestranded cycle sequencing using an Applied Biosystems 373 Stretch Sequencer and Ampli-Taq FS. Primer synthesis for site-directed mutagenesis and DNA sequencing was performed by the Biomolecular Resource Center at the University of California, San Francisco. The purpose of the following experiments (summarized in Figs. [1][2][3] was to insert the coding region of the CYP2B4 (B0) cDNA downstream of the T 7 promoter for overexpression in E. coli. CYP2B4 cDNA (15) containing an extra 500 bp at the C terminus of the P450 gene was digested with NcoI and EcoRI and cloned into pET-23d (low copy origin of replication from PBR-322) and pLWO1 (high copy origin of replication) to generate expression plasmids pET-P450 extratail and pLW-P450 extratail, respectively (Fig. 2). pLW01 ( Fig. 1) was generated by ligating NaeI/PvuII-digested pET-23d (Novagen, Madison, WI) with NaeI/PvuIIdigested pBluescript II KSϩ (Stratagene, La Jolla, CA). pLW01 contains the bacteriophage T 7 promoter, upstream of an NcoI (CCATGG) multi-cloning site coincident with the ATG start codon, and a high copy origin of replication (ColE1) from pBluescript II KS(ϩ). pLW-P450extratail contained 500 bp of noncoding DNA at the 3Ј end of the sequence, which was subsequently removed using PCR (Fig. 2). A 5Ј-oligonucleotide primer complementary to the unique RsrII restriction (1328 bp from the ATG start codon of the CYP2B4 gene) was synthesized (AAGGCATCGCGCGGACCGAG, RsrII site underlined). A 3Ј-oligonucleotide primer was synthesized which contained a unique HindIII site (GTCCCCGAAGCTTCATCAGCG, HindIII site underlined) a few base pairs downstream of the CYP2B4 stop codon. All PCR reactions were performed in a volume of 100 l and contained 2 units of Vent TM DNA polymerase in the recommended buffer, 1-2 M of each primer, and 200 M dNTPs. The DNA was melted at 95°C for 5 min in a Minicycler (MJ Research, Inc., Watertown, MA) and subsequently amplified using a temperature program of 45 s at 95°C, 45 s at 55°C, and 1.5 min at 72°C for 30 cycles. The product of the PCR reaction was digested with RsrII and HindIII and cloned into RsrII/HindIII-digested pLW-P450 extratail to generate the expression plasmid pLW-P450.
In order to express CYP2B4 protein with either a 4-, 6-, or 10histidine tag upstream of a Factor Xa cleavage site at the N terminus of wild-type P450, appropriate plasmids were constructed as follows. To generate the 10 His-tagged construct, it was necessary to introduce unique 5Ј-NdeI and 3Ј-HindIII restriction sites into the P450 gene. A 5Ј-oligonucleotide primer was synthesized encoding an NdeI site coincident with the ATG start codon of the CYP2B4 gene (GATATACATAT-GGAATTCAGCC, NdeI site underlined). The 3Ј primer contained an HindIII restriction endonuclease site a few base pairs downstream of the CYP2B4 stop codon (primer sequence described above). After annealing of the primers to the CYP2B4 cDNA, the gene was amplified by PCR, digested with NdeI and HindIII, and cloned into NdeI/HindIIIdigested pET-16b (Novagen, Madison, WI), which encodes a 10 His tag upstream of the factor Xa protease cleavage site and the multiple cloning site. This generated the expression plasmid pET16b-10HisP450. The high copy expression plasmid pLW-10HisP450 was then generated by digesting pET16b-10HisP450 with NcoI and HindIII and subcloning the NcoI/HindIII fragment containing the P450 sequence into NcoI/HindIII-digested pLW01. The 4 and 6 His-tagged N-terminal constructs were generated by PCR, using pLW-10HisP450 as the template and GTCCCCGAAGCTTCATCAGCG (HindIII site underlined) as the 3Ј primer. The 5Ј primers for the 6-His tag and 4-His tag were ATATACCATGGGCCATCATCATCATCATCACAGCAGC and ATATACCATGGGCCATCATCATCACAGCAGC, respectively (NcoI site underlined in both primers). After annealing of the primers to pLW-10HisP450, the P450 DNA was amplified, and the PCR products were digested with NcoI and HindIII. Following digestion, the DNA fragments were cloned into NcoI/HindIII-digested pLW01 (Fig. 1), generating the expression plasmids pLW-6HisP450 and pLW-4His P450.
To express a P450 with 4 histidines at its C terminus (not illustrated), a 3Ј-oligonucleotide primer was synthesized that encoded 4 His residues after the last amino acid of the CYP2B4 gene and upstream of the stop codon and the 3Ј-HindIII site (ATCGATAAGCTTCAATGGT-GATGGTGACGGGCCAGG, HindIII site underlined). The 5Ј primer was hybridized to a segment of the P450 sequence encoding the unique RsrII site (1348 bp from the ATG start codon, primer sequence described above). After annealing of the primers to pLW-P450, the CYP2B4 gene was amplified by PCR as described above. The PCR product was digested with RsrII and HindIII and cloned into RsrII/ HindIII digested pLW-P450 to generate the expression plasmid pLW-P450 -4His tail.
A second expression plasmid was generated containing the CYP2B4 DNA sequence and the vector pCWOriϩ, which contains a tandem tac-tac promoter (16). To clone the CYP2B4 gene into pCWOriϩ, it was necessary to introduce unique 5Ј-NdeI and 3Ј-HindIII restriction sites into the P450 gene which was in the plasmid, pLW-450. A 5Ј-oligonucleotide primer was synthesized encoding an NdeI site coincident with the ATG start codon (GGAGATATACATATGGAATTCAGCC, NdeI site underlined). The 3Ј-primer contained a HindIII site a few base pairs downstream of the CYP2B4 stop codon (primer sequence described above). After annealing of the primers to the CYP2B4 cDNA, the gene was amplified by PCR, digested with NdeI and HindIII, and cloned into NdeI/HindIII-digested pCWOriϩ to produce the expression vector pCW-P450.
Construction of N-terminal Signal Sequence Mutants of CYP2B4 -Mutations were introduced into the N-terminal signal sequence of the P450 DNA using PCR with the plasmid pLW-P450 as a template. The N-terminally mutated P450 genes were cloned into two vectors, pKK223-3 (tac promoter) and pLW01 (T 7 promoter) to determine whether there was any difference in expression from the different promoters. To facilitate cloning into both pKK223-3 and pLW01, both EcoRI (for pKK223-3) and NcoI (for pLW01) sites were encoded in the mutagenic primers. Two mutants were constructed as follows: in the first mutant a single amino acid (E2A) was changed and in the second mutant 5 amino acids (E2A, F3L, S4L, L6A, L7V) were changed. The 5Ј-oligonucleotide primers synthesized for mutant 1 (E2A) and for mutant 2 had the following sequences: CTCAGAATTCACCATGGCTTTC-AGCCTGCTCCTCCTC and CTGAGAATTCACCATGGCTCTGTTACT-GGCAGTTCTCCTGGCTTTCCTCGCAGGC, respectively (NcoI site underlined, EcoRI site in bold). The 3Ј primer encoded a HindIII restriction site downstream of the stop codon (primer sequence described above). After annealing of the primers to pLW-P450, the P450 DNA was amplified, and the PCR products were digested with EcoRI and HindIII. They were then cloned into the EcoRI/HindIII-digested vector pKK223-3 (Amersham Pharmacia Biotech, Uppsala, Sweden), which contains a tac promoter, to generate the plasmids pKK-P450Nmut1 and pKK-P450Nmut2. These plasmids were then digested with NcoI and HindIII, and the mutated CYP2B4 fragments were cloned into NcoI/ HindIII-digested pLW01 to generate the plasmids pLW-P450Nmut1 and pLW-P450Nmut2.
Site-directed Mutagenesis-To facilitate the generation of numerous mutants of CYP2B4, four unique restriction sites were introduced via silent mutations into the CYP2B4 cDNA (Fig. 3). The four unique restriction sites are as follows: KpnI (241 bp from ATG start codon), SpeI (522 bp from ATG start codon), EcoRI (843 bp from ATG start codon), and XbaI (1238 bp from ATG start codon). The sequences of the oligonucleotides used to introduce these restriction sites are as follows: for KpnI, CTCGCGGATGGCATCGGTACCGCACAGCACGACCAC-GGG (KpnI site underlined); for SpeI, AATGGAGCAGATGATGTTAC-TAGTGATTGAGTGGAAGAC (SpeI site underlined); for EcoRI, GTTC-TGGTGGTGGAATTCGCTGCTTGGGTGGA (EcoRI site underlined); and for XbaI, TGCCCCGTTGCCATCTAGAAAGTGGCCGGGG (XbaI site underlined). All four oligonucleotide primers were annealed to

Determination of Binding Site on CYP2B4 for Its Redox Partners
pLW-P450 in one reaction, and site-directed mutagenesis to produce the plasmid pLW-P450-EKSX was accomplished using a Muta-Gene Phagemid in vitro mutagenesis kit from Bio-Rad. Following mutagenesis to form the internal restriction sites, KpnI/SpeI-, SpeI/EcoRI-, EcoRI/XbaI-, and XbaI/HindIII-digested fragments of the CYP2B4 gene were excised from pLW-P450-EKSX and subcloned into the phagemid vectors pET-23a (Novagen, Madison, WI) or pTZ18U (Bio-Rad) for further mutagenesis. To generate the 24 alanine mutants of CYP2B4, 24 oligonucleotide primers were synthesized. Sequences of the oligonucleotides are shown in Table I. Site-directed mutagenesis was again accomplished using a Muta-Gene Phagemid in vitro mutagenesis kit from Bio-Rad. Following mutagenesis, correctly mutated P450 DNA fragments were excised from the mutagenesis vectors and subcloned into the expression vector pLW-P450-EKSX for expression in E. coli.
Expression of Wild-type and Mutant CYP2B4 in E. coli-The appropriate CYP2B4 (both wild-type and mutated) containing plasmid DNA was transformed into E. coli JM109 (DE3) cells containing pLysS (Novagen, Madison, WI) and grown overnight at 30°C on Luria-Bertani (LB) agar plates containing 1 mM thiamine, 100 g/ml ampicillin, and 74 g/ml chloramphenicol. 3 ml of LB medium containing 1 mM thiamine, 100 g/ml ampicillin, and 74 g/ml chloramphenicol were inoculated with a single transformed colony and grown overnight at 30°C, 200 rpm. 2 ml of this culture were then used to inoculate 100 ml of terrific broth (18) containing 1 mM thiamine, 0.5 mM ␦-aminolevulinic acid, 100 g/ml ampicillin, and 74 g/ml chloramphenicol in a 500-ml Erlenmeyer flask. The cultures were grown at room temperature (approximately 25°C), 120 rpm, until they reached an A 600 of at least 4, and then IPTG was added to a final concentration of 0.1 mM to induce expression. The cultures were then incubated for a further 72 h at room temperature (approximately 25°C), with shaking at 120 rpm. Overexpression of wild-type CYP2B4 was attempted on a larger scale (200 ml of medium in a 1000-ml Erlenmeyer flask and 750 ml in a 2800-ml Fernbach flask), but the yield of holo-P450 obtained was reduced by 50 -70%.
Purification of Wild-type CYP2B4 from E. coli--Unless otherwise specified, all operations were performed at 4°C. Cells expressing P450, from 4000 ml of cell culture medium, were harvested by centrifugation at 4000 ϫ g for 25 min, and the pellet was resuspended in 60 ml of buffer A (100 mM potassium phosphate, pH 7.4, 20% glycerol, 0.1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 0.3 M aprotinin, 130 M bestatin, 1 M leupeptin, 1 M pepstatin, 1 mg/ml lysozyme, 0.1 mg/ml RNase, 0.1 mg/ml DNase). The solution was stirred for 1 h at 4°C and then frozen at Ϫ80°C for 1 h. After thawing, the cycle of stirring, freezing, and thawing was repeated twice for a total of three times. The cells were lysed by sonication in an ice/salt bath using a Vibra Cell High Intensity Ultrasonic Processor (Sonics and Materials, Inc., Danbury, CT) for a total of 6 min at 40% power in 30-s pulses, with 5 min for cooling after each pulse. Unbroken cells were removed by centrifugation at 3000 ϫ g for 15 min, and the supernatant was ultracentrifuged at 105,000 ϫ g for 1 h to pellet the "membrane fraction." The supernatant was discarded, and the pellet was resuspended in 40 ml of buffer B (10 mM potassium phosphate, pH 7.4, 20% glycerol, 0.1 mM EDTA, 0.6% polyethoxyethylene 9 lauryl ether) by stirring for 3 h at 4°C. After the pellet had been completely resuspended, polyethoxyethylene 9 lauryl ether was added dropwise to the solution to a final ratio of 5:1, milligram of detergent:milligram of protein. The solution was stirred for 2 h at 4°C and subsequently centrifuged at 105,000 ϫ g for 1 h. The P450-containing supernatant was decanted and loaded directly onto a Reactive Red-agarose column (100 nmol of P450/ml of Reactive Red-agarose) pre-equilibrated with 10 volumes of buffer B. After loading the P450-containing supernatant, the column was washed with buffer C (10 mM potassium phosphate, pH 7.4, 20% glycerol, 0.1 mM EDTA, 0.3% Tergitol NP-10) at the same rate until the A 418 of the eluant was below 0.05. Subsequently, the column was washed with 4 column volumes of buffer C containing 100 mM NaCl to elute weakly binding proteins. CYP2B4 was eluted with buffer C containing 500 mM NaCl. Fractions containing P450 were combined, dialyzed overnight against 100 volumes of buffer C, and loaded onto a Bio-Gel HTP hydroxylapatite column (30 nmol of P450/ml of hydroxylapatite), preequilibrated with 10 volumes of buffer D (5 mM potassium phosphate, pH 7.4, 20% glycerol, 0.1 mM EDTA). After loading the CYP2B4 containing fractions, the hydroxylapatite column was washed with buffer D until the A 275 of the eluant was below 0.005. P450 was eluted using buffer E (500 mM potassium phosphate, pH 7.4, 20% glycerol, 0.1 mM EDTA), and the fractions containing P450 were combined and dialyzed overnight against 100 volumes of buffer F (50 mM potassium phosphate, pH 7.4, 20% glycerol, 0.1 mM EDTA). After dialysis, the fractions were concentrated to 100 -300 M P450 using an Amicon centricon apparatus and stored at Ϫ80°C. The amount of P450 in intact E. coli and solutions was determined from the CO minus the reduced difference spectrum (16), using a few grains of sodium dithionite as the electron donor and an extinction coefficient of 91 mM Ϫ1 cm Ϫ1 for the absorbance difference at 450 minus 490 nm (19).
Purification of Mutant CYP2B4 from E. coli-Mutant P450 proteins were purified as described for the wild-type P450, except for mutants R122A, R126A, Y190A, and F115A which had a specific content of less than 9 nmol/mg protein after the Reactive Red column. With these mutants, an additional purification step was required. Combined fractions from the Reactive Red columns were dialyzed overnight against 100 volumes of buffer C and loaded onto a DEAE-Sepharose Fast Flow column (100 nmol of P450/ml of DEAE Sepharose), pre-equilibrated with buffer C. The eluant containing the mutant P450s which did not bind to the DEAE-Sepharose column was then loaded onto a Bio-Gel HTP hydroxylapatite column, as described for the wild-type P450.

Determination of Binding Site on CYP2B4 for Its Redox Partners
Purification of CYP2B4, Cyt b 5 , and P450 Reductase from Rabbit Liver-Liver microsomes were prepared from phenobarbital-treated White New Zealand White rabbits (20). CYP2B4 was purified from rabbit liver microsomes as described previously (21) and had a specific content of 18 nmol of P450/mg of protein. NADPH P450 reductase was purified according to the method of Yasukochi and Masters (22) from rabbit liver microsomes. The effective concentration of the P450 reductase was calculated from its activity in the cytochrome c assay (10,23). It typically had a specific content of 48 nmol of P450 reductase/mg of protein. Cyt b 5 was purified from detergent-solubilized rabbit liver microsomes as described previously (24) and had a specific content of 45-52 nmol of cyt b 5 /mg of protein. The concentration of cyt b 5 in crude preparations was determined by measuring the difference spectrum (reduced minus oxidized) using an extinction coefficient of 190 mM Ϫ1 cm Ϫ1 (424 minus 409 nm) (25). In purified preparations an extinction coefficient of 117 mM Ϫ1 cm Ϫ1 at 413 nm was used (26). Apocytochrome b 5 was generated by acetone precipitation in the presence of HCl (27). Soluble cyt b 5 was produced by treating cyt b 5 with trypsin as described previously (28).
Protein Analysis-Protein concentration was determined using the BCA assay (Pierce) or by the Lowry method (29) after precipitation of the proteins in the presence of trichloroacetic acid and deoxycholate (30). Bovine serum albumin was used as a standard. SDS-polyacrylamide gels (9% for P450 and 15% for cyt b 5 ) were run as described (31). Immunoblot analysis of CYP2B4 was performed as described (17), using Immobilon P transfer membranes (Millipore Corp., Bedford, MA) and polyclonal goat anti-rabbit CYP2B4 antibody at a 1:1000 dilution as the primary antibody (Oxford Biomedical Research, Oxford, MI) and horseradish peroxidase-linked sheep anti-goat antibody (Sigma) at a 1:1000 dilution as the secondary antibody.
Determination of the Equilibrium Dissociation Constant (K d ) of the Cyt b 5 -CYP2B4 Complex-The binding of cyt b 5 to CYP2B4 was determined by measuring the type I spectral change (decrease in absorbance at 420 nm plus the increase in absorbance at 385 nm) occurring when cyt b 5 is added to a solution of P450, methoxyflurane, and DLPC. Suspensions of DLPC in water (2-5 mg/ml) were sonicated using a microtip probe on a Vibra Cell High Intensity Ultrasonic processor (Sonics and Materials, Inc., Danbury, CT) for a total of 2 min at 40% power in 30-s pulses, with 5 min cooling after each pulse, and then centrifuged at 13,000 ϫ g for 5 min. P450 (0.3 nmol) in 50 l of 100 mM potassium phosphate, pH 7.4, 20% glycerol, was mixed with DLPC (48 nmol) and preincubated at room temperature for 1 h. The solution was then adjusted to 1.0 ml using a solution of 100 mM KP i , pH 7.4, and 20% glycerol which was saturated with methoxyflurane. The P450 solution was placed in one compartment of a quartz dual chamber cuvette. The other chamber contained 1.0 ml of a solution of cyt b 5 (0.14 to 4.0 M) in 100 mM KP i , pH 7.4, and 20% glycerol which was saturated with methoxyflurane. The solutions were equilibrated to 25°C, and the absorbance at 420 and 385 nm was recorded. The solutions were then mixed by inverting the cuvette. After mixing, the absorbance at 420 and 385 nm was followed until it stabilized (typically 10 -15 min). The absorbance changes resulting from nonspecific protein binding to the cuvette were determined by preparing solutions in two additional cuvettes containing buffer in one chamber and either P450 or cyt b 5 in the other chamber. The absorbance change occurring on mixing in these cuvettes, presumably due to nonspecific adsorption of the proteins to the walls of the cuvettes, was determined at each protein concentration and subtracted from the observed absorbance changes to obtain the change resulting only from cyt 5 binding to P450. In preliminary experiments, wild-type and mutant P450s (0.15 M final concentration after mixing with cyt b 5 ) were assayed using at least three different concentrations of cyt b 5 within the range of 0.07 to 2.0 M. The experimentally observed absorbance changes ((A 420 before mixing Ϫ A 420 final) Ϫ (A 385 before mixing Ϫ A 385 final)) were fitted to a theoretical binding curve (as described under the "Appendix"), and the best fit between the experimental and theoretical data was determined by nonlinear regression. From the fitting procedure, estimates for the parameters K d and ⌬A max (the maximum absorbance change) were obtained. Those mutants found to have similar binding constants to wild-type P450 in the preliminary investigations (F115A, Y190A Determination of the Apparent Equilibrium Dissociation Constant for the P450 Reductase-CYP2B4 Complex-The apparent dissociation constant of the P450 reductase-CYP2B4 complex was determined by observing the rate of formaldehyde produced by the N-demethylation of benzphetamine at a constant P450 concentration and varying concentrations of P450 reductase (32,33). The final wild-type and mutant P450 protein concentration was 0.2 M, but the final concentration of P450 reductase varied (0.2, 0.3, 0.4, 0.5, 0.6, and 1.0 M). Appropriate concentrations of P450 and P450 reductase were preincubated with DLPC (32 M final concentration) in 50 mM potassium phosphate, pH 7.4, for 1 h at room temperature. The remainder of the reaction was performed as described for benzphetamine metabolism. The equilibrium dissociation constant for each P450 protein was calculated using nonlinear regression to fit the rate of formaldehyde production to a theoretical binding curve (described under the "Appendix"). From the fitting procedure, best fit estimates for the parameters K d (the apparent equilibrium dissociation constant) and V max (the maximum rate of benzphetamine metabolism) were obtained.
Measurement of Benzphetamine Metabolism-In order to measure the rate of N-demethylation of benzphetamine to formaldehyde, P450 (0.16 nmol) was mixed with P450 reductase (0.16 nmol) and DLPC (36 nmol) in 50 mM potassium phosphate buffer, pH 7.4, and incubated for 1 h at room temperature. The solution was then diluted to 0.8 ml with 50 mM potassium phosphate buffer, pH 7.4, containing 1 mM benzphetamine (final concentration of P450 and P450 reductase was 0.2 M). Stock solutions of benzphetamine were 10 mM in water. The solution was equilibrated to 30°C, and the reaction was initiated by the addition of NADPH (300 M final concentration). 200-l aliquots were removed after 1 and 6 min incubation at 30°C and quenched immediately with 22 l of 70% trichloroacetic acid. The quenched reaction mixtures were centrifuged at 13,000 ϫ g for 5 min, and 200 l of supernatant was assayed for formaldehyde (34). The activity of each mutant P450 protein was assayed in duplicate during four separate experiments for the 5 min between min 1 and 6 of the reaction as described previously (10).
Determination of the K d of the Benzphetamine-CYP2B4 Complex-The equilibrium dissociation constant for the benzphetamine-CYP2B4 complex (the spectral binding constant K s ) and its mutants was determined using a spectrophotometric assay (35). P450 (0.72 nmol) was mixed with DLPC (176 nmol) in 100 mM potassium phosphate buffer, pH 7.4, 20% glycerol and incubated for 1 h at room temperature. The solution was diluted to 360 l with 100 mM potassium phosphate buffer, pH 7.4, to give a final P450 concentration of 2 M and equilibrated to 30°C. The spectrum of the solution was recorded between 350 and 500 nm. Aliquots of a 10 mM benzphetamine hydrochloride solution were added, and the spectrum was recorded after each addition of benzphetamine. Spectral changes were complete within 1-2 min of addition of benzphetamine. Final concentrations of benzphetamine were 10, 30, 100, 300, and 1000 M. Spectra were corrected for dilution and difference spectra were calculated at each concentration. A double-reciprocal plot of benzphetamine concentration versus ⌬(A 420 -A 385 ) gave a straight line, which was used to calculate the K d (K d ϭ slope/intercept with y axis).
Determination of Methoxyflurane Metabolism-The products of methoxyflurane metabolism (methoxydifluoroacetic acid and dichloroacetic acid) were measured by a gas chromatographic-selected ion mass spectrometry assay, as described previously (10). The final concentration of P450, P450 reductase, and cyt b 5 in the reaction mixture of the wild-type and all mutant proteins was 0.2 M. Experiments were performed twice in duplicate with 0.5 ml of reaction mixture for 5 min. Aliquots (200 l) were removed after 1 and 6 min and were quenched with 140 l of 30% H 2 SO 4 . The amount of methoxyflurane metabolism occurring between 1 and 6 min was recorded. The internal standards containing 2 H 3 -methoxydifluoroacetic acid and 2,2-[2-13 C]dichloroacetic acid were added, and samples were stored at Ϫ80°C before extraction and analysis by mass spectrometry. In addition, assays with the mutant proteins R122A, R126A, R133A, F135A, M137A, K139A, R422A, K433A, and R443A were performed twice in duplicate in a final volume of 2 ml (800 l aliquots were quenched after 1 and 11 min with 560 l 30% H 2 SO 4 ) in order to obtain sufficient product to measure by gas chromatography selected ion mass spectrometry.
Identification of the Interprotein Surface Docking Regions of CYP2B4 and Cyt b 5 -The method used to find surface regions for binding is a surface complementarity algorithm for protein docking (36,37). The method, embodied in the program GRAMM, performs an exhaustive search for protein-protein surface structure complementarity. The three original contact regions obtained using this method at low resolution (6.4 -6.8 Å) have been described (1). In the studies reported here, this same method has been used for the CYP2B4-cyt b 5 pair but at the higher resolution of 3.4 Å. The results are more accurate than those from the previous 6.5-Å simulation but still can tolerate local inaccuracies in atomic details. The 10 lowest energy docking positions were analyzed and the results used to select preferred binding regions on CYP2B4 for cyt b 5 .

RESULTS AND DISCUSSION
Optimization of the Expression of CYP2B4 and Its Mutants in E. coli-Previous overexpression of mammalian P450s has been achieved, using either derivatives of pCWOriϩ (16) or the Amersham Pharmacia Biotech vector pKK223-3 (38), both of which use a tac promoter. CYP2B4 cDNA was initially cloned into both of these vectors, but when expression was attempted, CYP2B4 protein could not be identified in either Western blots of cell extracts or spectrophotometrically. A T 7 promoter containing vector, pET-23d, was then selected for further expression attempts. Initially, P450 was only observed on Western blots. However, reduction of the temperature to 28°C after induction with IPTG resulted in the production of small amounts of holo-CYP2B4. To optimize the expression, different host strains, concentrations of ␦-aminolevulinic acid, incubation temperatures, and incubation times were systematically tested. However, the maximum yield of holo-CYP2B4 obtained using this vector could not be increased above 200 nmol/liter. To increase the yield, the low copy pBR322-derived origin of replication in pET-23d was replaced with that from the high copy pBluescript II KS ϩ vector (ColE1 origin of replication), to generate the expression vector pLW01 (Fig. 1). The level of holo-CYP2B4 expressed using this vector was increased at least 2-fold over that obtained from pET-23d and reproducibly gave 400 -600 nmol of holo-P450/liter. To achieve optimal expression, it was necessary to grow the E. coli (both before and after induction with IPTG) at room temperature. The addition of 0.5 mM ␦-aminolevulinic to the growth medium and extending the length of the incubation after induction to 72 h both resulted in an increase in CYP2B4 expression. Continuing expression beyond 72 h after induction did not result in a further increase in holo-P450 expression.
In order to overexpress significant levels of a number of the P450s in E. coli, it has frequently been necessary to mutate the N terminus of the protein as first described by Waterman and co-workers (16,39,40). In an attempt to optimize the expression of CYP2B4, two mutants were constructed, one in which the Ala codon GGT (which replaced a glutamic acid residue) was placed in second position and a second mutant also with alanine in the second position E2A, and in addition F3L, S4L, L6A, L7V mutations which produced the N-terminal sequence of native bovine P450 17␣-hydroxylase. Ala is the preferred second codon for expression in E. coli (41). Expression of these mutant proteins in JM109 E. coli using the pKK 223-3 vector with a tac promoter yielded no immunoreactive or holoprotein, whereas expression using the pLW01 vector with a T 7 promoter yielded holo-P450 in quantities similar to those obtained with the wild-type CYP2B4 construct. Therefore, the wild-type CYP2B4 cDNA in vector pLW01 was used for all further experiments.
Four additional plasmids were generated containing His tags. DNA encoding 4, 6, and 10 consecutive His residues was cloned upstream of the P450 DNA so that N-terminally tagged proteins would be produced. Unfortunately, none of these Histagged P450 genes expressed holo-P450, although apoprotein could be identified. Expression of holo-CYP2B4 was obtained from a plasmid encoding four adjacent His residues at the C terminus of the CYP2B4 sequence; however, the wild-type construct lacking the C-terminal His tag reproducibly expressed higher levels of holo-P450, so subsequent experiments were performed without the His tag.
To facilitate the generation of numerous P450 mutant proteins, four silent mutations were introduced into the CYP2B4 sequence to create unique restriction endonuclease sites, which divided the P450 sequence into five 300-bp regions. Similar yields of native holo-P450 were obtained with the wild-type gene (pLW-P450) and the P450 gene containing the 4 silent mutations in pLW-P450-EKSX. pLW-P450-EKSX was, therefore, used as the template for the construction of the mutants. The level of expression of the wild-type and mutant P450s is summarized in Table II.
Purification of Wild-type CYP2B4 and Its Mutants-The purification of P450 was achieved by a procedure that permits the isolation of high specific content protein from the solubilized "membrane fraction" in a single step using Reactive Red 120agarose affinity chromatography (Table III).
The pelleted P450 containing the membrane fraction of E. Group E S128A (C) 4, 5 no P450 R140A (C*-D) 1, 5 no P450 a The numbers in this column correspond to the numbers in the text under "Rationale for the Mutation of Specific CYP2B4 Amino Acids" which describe the rationale for the construction of the mutants. The location of the mutated amino acids in the secondary structure of the cytochrome P450 2B4 model is given in parentheses in the first column. The secondary structure nomenclature is defined in Refs. 1 and 3. m-␤ indicates the meander to ␤-bulge. ␤ indicates the ␤ bulge. The capital letters indicate the helix in which the residue is located. Group A mutants bind cyt b 5 normally; group B mutants bind cyt b 5 poorly; groups C, D, and E mutants could not be characterized because the proteins were unstable, only apo protein was produced or protein was not produced, respectively. b Mutants that bind reductase poorly.
coli was used as the starting point of the purification. A number of different detergents were tested for their ability to solubilize the membrane pellet. The best results were obtained using polyethoxyethylene 9 lauryl ether, which extracted approximately 90% of the holo-CYP2B4. The first purification step after solubilization of P450 from the E. coli membrane fraction has conventionally involved DEAE-ion exchange chromatography (42)(43)(44)(45)(46)(47). However, with CYP2B4 only a small increase in specific content (e.g. from 1.0 to 2-3 nmol of P450/mg of protein) was achieved. The solubilized CYP2B4 was, therefore, purified using Reactive Red 120-agarose affinity chromatography. This single step resulted in extensive purification, with the specific content typically increasing from 1.0 to 11.4 nmol of holo-P450/mg of protein. Since cyt P420 did not bind to the resin, this column also separated cyt P420 from P450. The final step in the procedure was a hydroxylapatite column to remove detergent, which also resulted in a slight purification, yielding a specific content of 15.4 nmol of wild-type P450/mg of protein. Table III illustrates the overall recovery (15%) and purification achieved at each step in the P450 isolation procedure. The spectrum of wild-type CYP2B4 purified from E. coli is identical to that purified from rabbit liver microsomes (20). The purification of the majority of the mutant P450 proteins was performed as described for wild-type CYP2B4. However, the F115A, R122A, and R126A mutant P450s had specific contents of below 6.5 nmol of P450/mg of protein after the Reactive Red affinity chromatography, so an additional purification step using DEAE chromatography was performed. This second column produced protein with a specific content of 7.9 -12.1 nmol of P450/mg of protein. The overall yield from the cell culture (Х10 -15%) was not significantly affected by this additional DEAE column. Sufficient quantities of the P450 mutant proteins, F171A and R125A, could not be isolated due to their instability during the purification procedure (90% was lost during solubilization of the membrane fraction, Table II).
Rationale for the Mutation of Specific CYP2B4 Amino Acids- Table II lists the residues in CYP2B4 selected for mutation, the rationale for their mutation, and their location in the secondary structure of the model of CYP2B4. In an effort to simplify the presentation of the data we may appear to discuss the model as though it were a crystal structure, although clearly the model is not a crystal structure. It is simply a detailed hypothesis that has been used to design and tentatively interpret the experiments described in this article. All residues were mutated to the most common amino acid, alanine, in order to evaluate the function of the amino acid side chain distal to the ␤-carbon. The selection of these residues was guided by computer docking studies of the heme domain of cyt b 5 and a model of CYP2B4 (1), and the considerable amount of experimental evidence which indicated that the redox partners of P450 would bind on its proximal face where the heme is closest to the surface (3,13,14). Column 2 of Table II also tabulates the reason a specific mutant was constructed. The numbers in column 2 of Table II correspond to the seven reasons provided in the following paragraphs. The group A mutants in Table II correspond to those that bind cyt b 5 normally, and the group B mutants are those that bind cyt b 5 poorly. Group C mutants produced holo-P450 in quantities inadequate for protein characterization. Groups D and E mutants produced apoP450 or no P450, respectively. The two group A mutants R422A and R443A that bind cyt b 5 normally but bind P450 reductase poorly are indicated by a superscript b. Thus Table II summarizes and collates why the mutants were constructed, which mutants were expressed, and which mutants bound cyt b 5 and P450 reductase poorly. The detailed characterization of the mutant proteins is provided in later sections of this report.
1. The five proximal region mutants R125A, R133A, M137A, R140A, and R443A of CYP2B4, the five distal region mutants F171A, Y190A, H226A, P472A, and Y484A, and the two side surface mutants F115A and F283A were selected partly based on predictions from the computer simulations at low and high resolution of complexes of CYP2B4 with cyt b 5 . The x-ray structure of cyt b 5 and a model of CYP2B4 were used as input to the program GRAMM (37) that uses a surface complementarity algorithm to identify surface contact regions in the two protein partners of a protein-protein complex. The results of a high resolution study at 3.4 Å indicated a strong preference for a proximal binding region and eliminated the other two candidate regions on the distal and side surface of CYP2B4 found at low resolution (6.4 -8.6 Å). Specifically, of the 10 lowest energy docked complexes obtained from GRAMM at 3.4 Å, six of them formed a tight cluster at the proximal binding region of CYP2B4. Such a cluster is a strong indication of the real binding site (37). The remaining four were spread randomly in different surface regions without forming clusters and were discounted as false positives. The residues of CYP2B4 found to be in close contact with residues of cyt b 5 in the three lowest energy representative configurations of these six proximal region complexes are shown in Fig. 4, A-C, and Fig. 5 illustrates the surface location of the amino acids chosen for mutation.
2. Arg-125, Lys-421, and Arg-443 on the proximal face were selected because they are homologous to P450 camphor residues Arg-112, Lys-344, and Arg-364 which had been predicted to form salt bridges with cyt b 5 and putidaredoxin (14). Arg-112 has also been shown to be important in electron transfer (48).
4. Ser-128 was chosen for mutation because cyt b 5 had been demonstrated to be a competitive inhibitor of the phosphorylation of Ser-128 by cAMP-dependent protein kinase (50).
5. Arg-122, Arg-125, Arg-126, Ser-128, Arg-133, Phe-135, Met-137, Lys-139, and Arg-140 were chosen for mutation because of two previous studies. In one study, it was suggested that cytochrome P450 2B1 (CYP2B1) residues 116 -134 (residues 116 -134 in CYP2B4) were involved in binding cyt b 5 because a synthetic peptide composed of these CYP2B1 residues could inhibit the binding of cyt b 5 to CYP2B1 (51,52). In the second study, Davydov and co-workers (53) predicted that CYP2B4 residues 121-145 would be involved in binding cyt b 5 . This prediction was based on their sequence alignment studies that indicated that residues 68 -87 of cyt c which are involved in cyt b 5 binding were homologous to CYP2B4 residues 121-145. Residues 125-139 constitute the C and C* helices (see Refs. 1 and 3 for secondary structure nomenclature) which are on the proximal surface of CYP2B4 near the heme. 6. Arg-422, Lys-433, and Arg-434 were selected because they were basic residues near the heme that previous investigators had suggested might be involved in binding P450 reductase (54 -56).
7. The basic residues, Lys-276 and His-335, were controls since they are not predicted to be in any surface contact region or identified in biochemical studies.
Location in the CYP2B4 Model of Groups C, D, and E Mutants-Four of the alanine mutants (S128A, R140A, F283A, and R434A) did not express holo-P450 protein (Group D and E in Tables I and IV). The location of the mutated residues and the hydrogen bonding pattern of their side chains in the model was studied in an attempt to gain insight into why the mutant proteins may not have expressed holoprotein (Table IV). Phe-283 is located on the distal face of the model, in the loop before the I helix with no obvious important structural role unless it stabilizes loop formation. Ser-128, Arg-140, and R434A are all located on the proximal face of the model. R140A, which does not express either holo-or apoprotein, is located on a loop between the C* and D helices with its side chain completely exposed on the molecular surface to water with no discernable structural role.
The side chain hydroxyl of Ser-128 on the C helix is buried and forms a hydrogen bond to the backbone carbonyl oxygen of Leu-124, the residue immediately preceding the beginning of the C helix. This hydrogen bond is possibly important for the structural stability of the protein. However, mutation of the homologous residue Ser-129 to alanine or glycine in murine P450 2E1 and expression of the mutant proteins in COS cells resulted in holoprotein (57). A buried Ser-128 side chain is consistent with the most recent studies of Schenkman and co-workers (50) that indicate that Ser-128 in CYP2B1 can only be phosphorylated in the apoprotein.
In our model, the guanidinium group of Arg-434 forms a salt bridge to both oxygens of the heme D ring propionate and two hydrogen bonds to the backbone carbonyls of Tyr-111 and Ile-114, respectively, via a water molecule. The fifth possible hydrogen bond is formed with water. ApoP450 production by the R434A mutant is consistent with its hypothesized role in heme binding and stabilizing the protein structure via hydrogen bonds to the carbonyl oxygens of two amino acids (1,58).
The three mutants P472A, R125A, and F171A in group C in Tables II and IV could be expressed but could not be purified in sufficient quantity to characterize. The location of these mu-tated amino acids in the CYP2B4 model and the possible result of mutating the amino acid on the structure of CYP2B4 was examined in an attempt to understand why these mutants might be unstable (see Table IV). One of the side chain NH groups of Arg-125 (aligns with P450 camphor Arg-112) forms a hydrogen bond with the backbone carbonyl of Lys-433 and a buried water which, in turn, forms a hydrogen bond to the heme D ring propionate. The ⑀and second NH are exposed on the surface to solvent. The alanine mutant would be unable to form the hydrogen bonds with Lys-433 and the heme and the shorter alanine side chain would likely produce a cavity on the surface of the protein by which water could gain access to the heme and thereby decrease the heme-protein binding. A similar phenomenon was observed when Tyr-74 was mutated to lysine in cyt b 5 (59). At present, the poor expression and instability of P472A and F171A cannot be explained (Table II).
It is of interest that the remaining mutants gave rise to wild-type levels of protein expression and stability even though Group E S128 (C) C¢O of Leu-124 in loop before C helix Lose H bonds for C helix 6 9 Hydrophilic cavity created R140 (C*-D) Exposed to solvent None 69 72 a The location of each residue in the secondary structure of the P450 model is shown in parentheses. The secondary structure nomenclature is described in the footnotes to Table II and references 1  many of the side chains formed hydrogen bonds or salt bridges with other surface residues. The hydrogen bonds formed by selected mutated side chains and the possible structural consequences resulting from the mutations are summarized in Table IV. The CYP2B4 model indicates that a large number of basic amino acid residues is concentrated on the proximal surface of the molecule near the heme. These basic residues presumably function to neutralize the charge on the two buried heme propionate residues and establish a molecular dipole which may facilitate catalysis, in addition to being involved in redox partner binding (3).
Characterization of the Ability of Wild-type and Mutant P450s to Bind Cyt b 5 - Table V lists the K d values of the wild-type and mutant P450s with cyt b 5 . The wild-type protein has a K d of Х0.2 M in good agreement with a previously determined value of the K d of the cyt b 5 -CYP2B4 complex (35). Mutants Y190A, K276A, H335A, K421A, R422A, R443A, and Y484A had a K d of Х0.2 M similar to that of the wild-type protein and served as negative controls. The K d of His-226 was 0.45 M and was considered to be indistinguishable from wild type. The K d could not be determined for the F115A mutant protein because it precipitated in the cuvette. CYP2B4 residues K421A and R443A are homologous to cyt P450 camphor residues Lys-344 and Arg-364 which had previously been hypothesized to be involved in cyt b 5 and putidaredoxin binding (14,60). The mutants R122A, R126A, R133A, F135A, M137A, K139A, and K433A were observed in initial experiments to have a higher K d than wild-type proteins (Table V). Repeat experiments using higher P450 (0.2-0.5 M) and cyt b 5 concentrations (0.5-10 M) were conducted in an attempt to obtain more accurate estimates of the K d values by using protein concentrations that would elicit at least 50% of the predicted ⌬A max . The data obtained with R122A and R133A covered the middle to upper part of the binding curve; the spectral changes with F135A and M137A plotted to the middle of the binding curve, whereas the absorbance changes observed with R133A and R126A only covered the lower part of the binding curve even with a P450 concentration of 0.5 M. Therefore, the K d values of R133A and R126A are the most tentative. The ⌬A max for all mutants was estimated from curve fitting and normalized to 0.15 M P450. Within experimental error (data not shown) the ⌬A max for all mutants was similar.
The seven amino acids that partially define the binding site on CYP2B4 for cyt b 5 are all located on the proximal surface near the heme ligand cysteine (Fig. 6A). All of the mutants except K433A are either located in the C or C* helices or in loops at the termini of these helices. These results are in striking agreement with the predictions by Davydov and coworkers (53) that CYP2B4 residues 121-145 would be involved in binding cyt b 5 . The results also confirm the predictions of the surface complementarity algorithm which was able to localize the binding site to the proximal surface of CYP2B4 near the heme (37). However, the results do not substantiate the prediction that Lys-421 and Arg-443 would be involved in cyt b 5 binding because of their homology to residues Lys-344 and Arg-364 in P450 camphor which were hypothesized to participate in cyt b 5 binding (14).
Characterization of the Ability of Wild-type and Mutant P450s to Metabolize the Substrates Methoxyflurane and Benzphetamine-The results obtained with the cyt b 5 binding assay indicate that seven of the mutant proteins have decreased ability to bind cyt b 5 . If the active site and substrate access channel have the same conformation as the wild-type protein, the binding and metabolism of substrates which do not require cyt b 5 for their metabolism should be similar to that observed for wild-type P450. In order to determine whether the CYP2B4 mutant proteins functioned like wild type, three assays were performed. The K d (K s ) between the substrate, benzphetamine, and the mutant P450s was determined (see Table V). Except for Y190A which binds benzphetamine tighter than the wild-type protein, these results indicate that the benzphetamine-binding site and substrate access channel of the mutants are similar to that of the native protein.
The ability of the mutant proteins to catalyze the oxidation of methoxyflurane was investigated. Methoxyflurane, whose metabolism is markedly stimulated in the presence of cyt b 5 , is a poor substrate for CYP2B4 (only 2% coupled) (10). The ability of the mutant P450s to metabolize methoxyflurane both in the presence and absence of cyt b 5 was followed by measuring the production of the metabolites, methoxydifluoroacetic acid, and dichloroacetic acid. The results of these assays which were performed at a P450 and cyt b 5 concentration of 0.2 M are shown in Table V. The ratio of methoxyflurane metabolism in the presence and absence of cyt b 5 is also given. The data indicate that mutants with reduced ability to bind cyt b 5 in the spectrophotometric assay are stimulated to a lesser extent by cyt b 5 . An exception to this is mutant K433A; although the rate of methoxyflurane metabolism in the presence of cyt b 5 is low, there is a 4-fold stimulation which cannot be explained. Note that these experiments were performed at protein concentrations close to the K d of the P450-cyt b 5 complex.
The ability of the mutants to metabolize the model substrate, benzphetamine, is shown in Table V. The results were obtained from experiments where P450 and P450 reductase were both present at a final concentration of 0.16 M. All of the P450 mutants except M137A, with diminished ability to bind cyt b 5 and mutants R422A and R443A, have a 50 -85% reduced rate of benzphetamine metabolism compared with wild type (determined by t test analysis). If the mutants with a normal ability to bind cyt b 5 are compared as a group to the mutants showing a reduced ability to bind cyt b 5 , it can be demonstrated that the mutants with the diminished ability to bind cyt b 5 have a significantly lower rate of benzphetamine metabolism than the mutants with normal cyt b 5 binding. This finding prompted us to examine the ability of these mutants to bind P450 reductase.
Characterization of the Ability of Wild-type and Mutant P450s to Bind P450 Reductase-A number of previous investigators have identified basic residues on the proximal surface of CYP2B4 which were presumed to participate in binding acidic residues on P450 reductase. Strobel and co-workers (55,56) demonstrated that when eight lysine residues in CYP2B1, which is 85% identical to CYP2B4, were chemically modified by acetic anhydride, 95% of the activity was lost. Residues that were modified in this inactive protein and their predicted location in the secondary structure of P450 are Lys-384 (␤1-3; side chain hydrogen bond to carboxyl of Asp-374 in ␤2-1), Arg-422 (m-␤), Lys-433 (␤ bulge), and Arg-473 (substrate recognition site 6). These residues correspond to the same residues in CYP2B4. (See Table IV for location of the residues in the secondary structure of P450, as well as the hydrogen bonding pattern of the Arg-422 and Lys-433 side chains.) Fujii-Kuriyama and co-workers (54) mutated the conserved lysine residues in cytochrome P450 1A2 (P450 d ) by site-directed mutagenesis and showed that several had a decreased ability (2-4-fold) to bind and be reduced by P450 reductase. Some of these residues are predicted to be on the proximal surface of CYP2B4 and correspond to residues Arg-422, Lys-433, and Arg-443 in CYP2B4 which were mutated in the experiments described here. Cyt P450scc (CYP 11A1), a mitochrondrial P450, was also mutated at residues Arg-377 and Arg-381 which are predicted to be located in the K helix and correspond to CYP2B4 residues Asp-350 and His-354 (61). Arg-377 and Arg-381 were shown to play crucial roles in binding adrenodoxin. Thus, the consensus of a large body of experimental evidence is that basic residues on the proximal surface of P450 are involved in binding acidic residues on a redox partner. Until recently, our understanding of the result of such mutagenesis studies has been hindered by the unavailability of reliable models of microsomal P450s.
The P450-catalyzed oxidation of substrates requires the transfer of two electrons from P450 reductase. Electron transfer occurs within a transient intermolecular complex between P450 and P450 reductase (33). If complex formation between electron donor and acceptor molecules did not occur, efficient electron transfer would be impossible and electrons would be dissipated into the medium. It has been assumed that the rate of substrate oxidation is proportional to the concentration of the P450-P450 reductase complex, which is an approach similar to that used by Miwa et al. (33). The observation that oxyferrous cytochrome P450 accumulates during turnover (62,63) supports this assumption, since it indicates that product formation is limited by electron transfer. This relationship can be described by Equation 1. where rate refers to the rate of substrate oxidation, k is a proportionality constant, and [P450-P450 reductase] is the concentration of the interprotein complex. A general equation that relates the equilibrium dissociation constant of the P450-P450 reductase complex and the rate of substrate oxidation was derived (see "Appendix") and has been used to determine the K d , equilibrium dissociation constant, of the P450-P450 reductase complex. The rate of oxidation of the substrate, benzphetamine, was measured experimentally in the presence of varying but known concentrations of P450 and P450 reductase. These data were fit to Equation 21 under the "Appendix," and the K d values between the individual mutant P450s and the reductase were calculated using nonlinear regression.
The apparent K d for the wild-type CYP2B4-P450 reductase complex is 0.02 Ϯ 0.02 M, in agreement with previously published values (33). Consistent with the determination of the K d values of the individual complexes was the finding that substrate oxidation was proportional to reductase concentration with a fixed concentration of P450. As predicted the weaker binding P450 mutants exhibited a greater percentage increase in product formation with a given increment of the reductase. In the absence of P450 the control reactions exhibited no substrate oxidation. Nine mutant P450s exhibited an increased apparent K d in complex with P450 reductase (Table V). The seven mutants, R122A, F135A, M137A, K139A, K433A, R126A, and R133A, that bound cyt b 5 poorly also bound reductase poorly. In addition, the mutants R422A and R443A showed elevated K d values. The binding of these nine mutant P450s to P450 reductase was diminished 6 -47-fold (Table V). The location of the mutated amino acids on the surface of CYP2B4 that exhibited diminished binding to reductase is shown in Fig. 6B. As mentioned previously, six of the mutants with decreased ability to bind cyt b 5 are located on the proximal surface of CYP2B4 slightly below the heme in the C and C* helices. The location of the other amino acids in the secondary structure of the CYP2B4 model and their possible role in its structural stability follows. Lys-433 is in the highly conserved ␤-bulge three residues away from the heme ligand, cysteine 436. A nitrogen in its side chain forms a salt bridge with the carboxyl group of Asp-90 in the B helix; the remainder of the guanidinium group is exposed to solvent. R422A is in a loop in a conserved region between the meander and the ␤-bulge above the heme. One of the nitrogens in its side chain guanidinium group forms a salt bridge with the side chain carboxyl of Glu-424 also in the conserved region between the meander and ␤-bulge. The remainder of the Arg-422 guanidinium group is on the surface exposed to water. Arg-443 is in the L helix, and its side chain forms a salt bridge with the side chain carboxyl group of Glu-439 also in the L helix. Arg-443 is homologous to the P450 camphor residue, Arg-364, which forms a structurally important salt bridge to Glu-286 in the K helix (64) and was predicted to form a salt bridge with P450 camphor's redox partners, cyt b 5 and putidaredoxin (14). Table IV lists the hydrogen bonding pattern of the mutated amino acid side chains and the expected structural effects of mutating the amino acid to an alanine. In many instances, structural perturbations are expected, and cavities are created which will give water access to previously shielded surfaces. Our results unexpectedly reveal that the binding sites for cyt b 5 and P450 reductase have considerable overlap. These results are not in agreement with previous experiments with a purified, covalently cross-linked, functional CYP2B4-cyt b 5 complex (65). The previous studies indicated that there was no overlap between the cyt b 5 and P450 reductase-binding site on CYP2B4. At present, there is no explanation for the discrepancy.
Note that the initial assays for methoxyflurane and benz-phetamine metabolism were conducted at total P450 reductase concentrations of 0.2 M and 0.16 M, respectively. The reductase concentrations are 8-and 6.4-fold higher than the K d of the P450-P450 reductase complex. The presence of these relatively high concentrations of reductase during the assays of substrate metabolism presumably accounts for the fact that substrate metabolism was only minimally disturbed in those mutants exhibiting decreased ability to bind both P450 reductase and cyt b 5 .
Although the predicted V max values for benzphetamine metabolism had a large standard deviation, they were all indistinguishable from the wild-type protein, except for the R126A mutant with a Х66% decreased V max (Table IV), indicating that in the presence of elevated levels of P450 reductase the mutant P450s function normally.
Characterization of the Free Energy of Binding between the Mutant P450s and Their Redox Partners-The specific interactions between CYP2B4 and its redox partners cyt b 5 and P450 reductase are critical events during substrate oxidation in vitro and in vivo. In this article, seven cyt P450 residues that participate in the cyt b 5 -P450 interaction and nine cyt P450 residues that participate in the P450 reductase-P450 interaction have been identified, partially on the basis of the complementarity of the surfaces of cyt b 5 and a CYP2B4 model. In an effort to more thoroughly understand the function of the mutated amino acids in the protein-protein association, the difference in the free energy of binding between the wild-type and mutant P450s and cyt b 5 was calculated (Table V). The free energy of binding (calculated using the formula ⌬G ϭ RTlnK d ) between cyt b 5 and wild-type CYP2B4 is Ϫ9.2 kcal/mol. Of the seven mutant proteins with decreased ability to bind cyt b 5 , the R133A mutant protein had the greatest reduction in affinity (Ͼ2.2 kcal/mol). The sum of the difference in free energy of binding between the seven mutant P450s and cyt b 5 is Ϫ10.8 kcal/mol, which exceeds the known binding free energy for the entire complex by 1.6 kcal/mol (Table V) (66,67).
Modeling of the effects of deleting the hydrophobic side chains, hydrogen bonds, and salt bridges, formed by the mutated side chains in CYP2B4, indicates that the mutations are likely to lead to local structural and electrostatic perturbations that would contribute to the decrease in the binding of the mutant P450s to cyt b 5 (Tables IV and V). At least one of the hydrogens of the basic side chains of Arg-122, Arg-133, Lys-139, and Lys-433 interact with an oppositely charged residue in the CYP2B4 model and is likely to generate a local structural change when mutated.
Mutation of basic residues on the proximal surface of CYP2B4 may also disrupt the molecular dipole which has been suggested to function pre-collisionally, both to orient the interaction of P450 with its redox partners and to promote the electrochemical flow of the reactants involved in oxidation (3). The molecular dipole of P450 is oriented so that it will promote the proximal-to-distal flow of electrons from the redox partner and the distal-to-proximal flow of protons from the solvent.
Previous studies of the cyt b 5 -cyt c and the growth hormonegrowth hormone receptor interprotein-binding sites indicate that their surfaces are complementary and that a few hydrophobic residues near the center of the interprotein-binding site provide the majority of the binding energy. These core residues were surrounded by less energetically important, more hydrophilic and charged contact residues. Examination of the surface location of the mutated residues, which result in diminished binding of cyt b 5 , shows that Arg-133 (Table IV) appears to be located in the middle of the putative cyt b 5 -binding site, along with the uncharged amino acids Phe-135 and Met-137. Since a model of the binding site on CYP2B4 for cyt b 5 rather than a crystal structure of the interprotein complex is being examined, there is insufficient data at this time to allow us to conclude whether or not the CYP2B4-cyt b 5 -binding site has a crosssection similar to the two other interprotein complexes mentioned previously (66 -69).
The free energy of binding between wild-type CYP2B4 and P450 reductase is Ϫ10.5 kcal/mol. CYP2B4 binds reductase about 10-fold tighter than it binds cyt b 5 . The sum of the difference in the free energy of binding between P450 reductase and the nine mutant P450s is Ϫ14.4 kcal/mol which exceeds the known binding free energy for the P450-P450 reductase complex (Table V) (67)(68)(69). The side chains of Arg-422 (Arg-422 aligns one residue away from Arg-344 on P450 camphor) and Arg-443 (Arg-443 aligns with the structurally important Arg-364 in P450 camphor (64)), which are involved in reductase but not cyt b 5 binding, form salt bridges on the surface of the CYP2B4 model with glutamic acids residues 424 and 439, respectively. It should be noted that Sligar and co-workers (14) predicted that residues Arg-344 and Arg-364 in P450 camphor would be involved in binding cyt b 5 and putidaredoxin. The mutation of these charged residues may give rise to local structural perturbations which may account for some of the decrease in binding energy in a given mutant (Table IV).
In summary, the existence of a model of CYP2B4 has enabled us to formulate specific hypotheses about the structural and electrostatic consequences of mutations in basic and neutral amino acids on the proximal surface of CYP2B4. The heme has its closest approach to the surface of the molecule on its proximal face and is the logical place for docking of its redox partners and subsequent electron transfer. The studies reported herein demonstrate that the binding sites on the proximal surface of CYP2B4 for cyt b 5 and P450 reductase partially overlap. Basic and neutral residues in the C and C* helices of CYP2B4 play an especially prominent role in binding its redox partners.