Advertisement
Research ArticleABSORPTION, DISTRIBUTION, METABOLISM, AND EXCRETION

Dual Role of Human Cytochrome P450 3A4 Residue Phe-304 in Substrate Specificity and Cooperativity

Tammy L. Domanski, You-Ai He, Greg R. Harlow and James R. Halpert
Journal of Pharmacology and Experimental Therapeutics May 2000, 293 (2) 585-591;

Abstract

The structural basis of cooperativity of progesterone hydroxylation catalyzed by human cytochrome P450 3A4 has been investigated. A recent study suggested that substitution of larger side chains at positions Leu-211 and Asp-214 partially mimics the action of effector by reducing the size of the active site. Based on predictions from molecular modeling that Phe-304 in the highly conserved I helix is involved in both effector and substrate binding, a tryptophan residue was substituted at this position. The purified F304W mutant displayed hyperbolic progesterone hydroxylase kinetics, indicating a lack of homotropic cooperativity. However, the mutant remained responsive to stimulation by α-naphthoflavone, exhibiting a 2-fold decrease in theKm value for progesterone 6β-hydroxylation in the presence of 25 μM effector. Combining substitutions to yield the triple mutant L211F/D214E/F304W maintained theVmax and decreased theKm for progesterone 6β-hydroxylation, minimized stimulation by α-naphthoflavone, and decreased the rate of α-naphthoflavone oxidation to one-eighth of the wild type. Interestingly, the ΔAmax for spectral binding of α-naphthoflavone was unaltered in L211F/D214E/F304W. Overall, the results suggest that progesterone and α-naphthoflavone are oxidized at separate locations within the P450 3A4 binding pocket, although both substrates appear to have equal access to the reactive oxygen.

Cytochrome P450 (P450) 3A4 is the most abundant P450 in human liver (Guengerich, 1990) and is involved in the metabolism of ∼50% of clinically used drugs (Guengerich, 1995). The large binding pocket in P450 3A4 allows this enzyme to accommodate substrates of very diverse size and structure while maintaining strict regioselectivity (Kronbach et al., 1989; Waxman et al., 1991). The size of the active site also may be responsible for the phenomenon of cooperativity that is observed with members of the 3A family. Homotropic cooperativity has been observed with steroids such as progesterone (Schwab et al., 1988; He et al., 1995; Harlow and Halpert, 1998), testosterone (Ueng et al., 1997;Harlow and Halpert, 1998), and 17β-estradiol (Ueng et al., 1997), as well as with aflatoxin B1 (Ueng et al., 1997) and amitriptyline (Ueng et al., 1997). The flavonoid α-naphthoflavone (ANF) heterotropically stimulates the oxidation of progesterone and testosterone (Schwab et al., 1988; He et al., 1995; Harlow and Halpert, 1998) and various other 3A substrates (Shimada and Guengerich, 1989;Kerlan et al., 1992; Kerr et al., 1994; Li et al., 1994; Shou et al., 1994; Ueng et al., 1995). In addition, diazepam displays sigmoidal kinetics in vitro (Shou et al., 1999), and it has recently been found that meloxicam metabolism is activated by quinidine and hydroquinidine in vitro (Ludwig et al., 1999). Consequently, cooperativity may be clinically significant due to the role it can play in enhancing drug-drug interactions (Lasker et al., 1984).

Previously, the lack of a P450 3A4 crystal structure and allelic variants that specifically affect cooperativity has hindered a clear understanding of its catalytic mechanism. To overcome this obstacle, a combination of molecular modeling (Szklarz and Halpert, 1997) and site-directed mutagenesis has been used. These approaches have previously been successful in identifying specific 3A4 residues that are involved in substrate specificity and have provided further evidence for the large size of the binding pocket (Harlow and Halpert, 1997; He et al., 1997; Domanski et al., 1998). An initial study with alanine-scanning mutagenesis identified a role for residue Leu-211 in ANF stimulation of progesterone 6β-hydroxylation (Harlow and Halpert, 1997). Further modeling suggested that Asp-214 also was located in the effector binding site, and predicted that increasing the size of residues 211 and 214 would decrease the size of the effector site and mimic the presence of bound effector (Harlow and Halpert, 1998). The mutant L211F/D214E lost homotropic cooperativity of progesterone and testosterone hydroxylation as well as responsiveness to stimulation of 6β-hydroxy- and 16α-hydroxy testosterone formation by ANF (Harlow and Halpert, 1998). However, ANF still stimulated the formation of 6β-OH progesterone) by L211F/D214E, although to a lesser degree than observed with 3A4 WT. The residual stimulation of progesterone hydroxylase activity indicated that additional residues could be involved in effector action. Furthermore, questions remained concerning the location of residues Leu-211 and Asp-214 within the 3A4 structure. An amino acid sequence alignment of P450 3A4 with bacterial sequences of known crystal structure indicated that these residues are located in the F helix (Szklarz and Halpert, 1997). This region is variable in both length and sequence and in P450BM-3 shifts considerably on substrate binding (Modi et al., 1996). Therefore, the possibility remained that the substitutions at positions 211 and 214 altered the conformation of the enzyme in a similar fashion to the effector, without actually altering the effector-binding site.

Examination of our molecular model for additional sites with the potential to affect P450 3A4 cooperativity predicted that residue Phe-304 lies within both the proposed effector-binding site and the substrate-binding site (Fig. 1). Phe-304 is predicted to be located in the highly conserved I helix (Hasemann et al., 1995). This region in 3A4 has been shown to contain a number of residues that are important substrate contact points (Domanski et al., 1998). Mutant F304A exhibited increased progesterone 6β-hydroxylase activity and an altered metabolite profile but unaltered ANF stimulation, indicating a role for this residue in substrate binding (Domanski et al., 1998). In the current study, we found that converting Phe-304 to a larger side chain caused a decrease in stimulation by ANF, a change that was augmented when combined with L211F and D214E. This mutant, L211F/D214E/F304W, effectively mimics the ANF-bound wild-type enzyme. These data reveal a role for residue Phe-304 in cooperativity and provide further evidence that the 3A4 substrate and effector-binding sites are separate, but lie close to each other and comprise several common residues.

Figure 1
Figure 1

Molecular model of P450 3A4 in the I helix region. The model developed by Szklarz and Halpert (1997) was used to illustrate the putative location of P450 3A4 residue Phe-304. The side chains of Leu-211 and Asp-214 also are labeled and shown in gray. The ribbons represent portions of substrate recognition sites 2 and 4.

Experimental Procedures

Materials.

Primers for polymerase chain reaction (PCR) amplification were obtained from the University of Arizona Macromolecular Structure Facility, Tucson, and National Biosciences, Inc. (Plymouth, MN). Restriction endonucleases and bacterial growth media were purchased from Life Technologies (Grand Island, NY), and the Expand PCR kit was purchased from Boehringer Mannheim (Indianapolis, IN). Progesterone, ANF, dioloeoylphosphatidylcholine (DOPC), 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid (CHAPS), and NADPH were obtained from Sigma Chemical Co. (St. Louis, MO). 4-[14C]Progesterone was purchased from DuPont NEN (Boston, MA). HEPES was purchased from Calbiochem (La Jolla, CA), and thin-layer chromatography plates [silica gel, 250 mm, Si 250 PA (19C)] were purchased from Baker (Phillipsburg, NJ). All other reagents and supplies were obtained from standard sources.

Cloning and Expression of P450 3A4 Mutants F304W and L211F/D214E/F304W.

Plasmid pSE3A4His, described previously (Domanski et al., 1998), was used as the template for amplification reactions with the Expand PCR kit according to the manufacturer's directions. The F304W forward primer (5′GCTGGCTATGAAACCACGAGCAGTGTTCTCTCC) and the reverse primer (5′GCTCGTGGTTTCATAGCCAGCCCAAATAAAGAT) were designed to contain a 21-base pair overlap and to amplify the entire pSE3A4His plasmid. DpnI digested DNA was transformed into DH5α cells, and DNA from several of the resulting colonies was isolated. The sequence of the 3A4 cDNA was checked for the presence of the desired F304W mutation and the absence of extraneous mutations (University of Arizona Sequencing Facility). The triple mutant L211F/D214E/F304W was produced and analyzed as described for F304W, except that the template used was the double mutant-containing plasmid pSE3A4His (L211F/D214E) (Harlow and Halpert, 1998).

Growth and induction of Escherichia coli were performed as described previously (Domanski et al., 1998). Solubilized membranes were prepared and the P450s were purified on Talon metal affinity columns (Clontech, Palo Alto, CA) under conditions described inDomanski et al. (1998). P450 content was determined by reduced carbon monoxide difference spectra with the addition of 1% Triton X-100 to the protein sample before dilution in microsome solubilization buffer (100 mM potassium phosphate, pH 7.3; 20% glycerol; 0.5% sodium cholate; 0.4% Renex; and 1.0 mM EDTA).

Progesterone Hydroxylase Assays and Kinetic Analysis of Data.

Purified 3A4 enzyme (5 pmol) was reconstituted in the presence of 0.4% CHAPS, 0.1 mg DOPC, 20 pmol of rat NADPH-P450 reductase, and 10 pmol of cytochromeb5 in 10 μl for 10 min at room temperature. Assay mixtures contained 50 mM HEPES (pH 7.6), 15 mM MgCl2, 0.1 mM EDTA, and 25 μM [14C]progesterone, with or without 25 μM ANF (unless otherwise stated). To minimize error due to adherence of progesterone and/or reaction contents to the glass reaction vials or pipet tips, siliconized disposable glass tubes and pipet tips were used. Each 100-μl reaction contained a final concentration of 1% methanol (v/v) and 0.04% CHAPS. The reactions were started by the addition of 1 mM NADPH and carried out for 5 min at 37°C before being stopped by the addition of 50 μl of tetrahydrofuran. A portion of each reaction was aliquoted into scintillation vials and measured in a Beckman LS6500 multipurpose scintillation counter (Fullerton, CA) to determine the actual concentration of progesterone in each reaction. Metabolites were resolved by three cycles of thin-layer chromatography in benzene/ethyl acetate/acetone (10:1:1). Metabolites were visualized by autoradiography. Data analysis was performed with Sigma Plot (Jandel Scientific, San Rafael, CA) for data analyzed with the Michaelis-Menten equation v =VmaxS/(Km+ S), the Hill equation v = (VmaxSn)/(S50n+ Sn) (Ueng et al., 1997), or the modified two-site equation (Vmax1= 0) v = (Vmax2S2/Km1Km2)/(1 + S/Km1 + S2/Km1Km2) (Korzekwa et al., 1998).

ANF Oxidation Assays.

The reconstitution of proteins for ANF oxidation studies was performed as described above for steroid hydroxylase assays, except that 10 pmol of P450 was reconstituted in a final volume of 20 μl with 40 pmol of reductase and 20 pmol of cytochrome b5. The ANF oxidation assays were carried out in 50 mM HEPES (pH 7.6), 15 mM MgCl2, 0.1 mM EDTA, and 25 μM ANF (unless otherwise stated). Each 100-μl reaction contained a final concentration of 1% methanol (v/v) and 0.04% CHAPS. The reactions were started by the addition of 1 mM NADPH and carried out for 5 min at 37°C before being stopped by the addition of 300 μl of methylene chloride. Naringenin, used as an internal standard, was added at a final concentration of 2.5 μM. The reactions were centrifuged at low speed for 3 min and the aqueous phase was discarded. The samples were extracted two additional times with methylene chloride before being dried under N2. When progesterone was added to the reactions, it was desiccated before being resuspended in an ANF/methanol solution to maintain the final methanol concentration at 1%.

Metabolites were separated with a Beckman ODS 5-μm column (4.6 × 25 mm) in 70% methanol at 1.0 ml/min and detected at 280 nm. ANF 5,6-oxide formation was estimated with the extinction coefficient for ANF (23.7 mM−1 cm−1) because a standard for the metabolite was not available (Ueng et al., 1997). Although the extinction coefficients for the parent compound and the metabolite might not be identical, any variation would not alter the comparison between samples. The identity of the major metabolite ANF 5,6-oxide was confirmed by liquid chromatography-mass spectrometry. Reactions were run as described above before being submitted to this analysis. HPLC was performed on a Hewlett Packard Series 1050 apparatus. The metabolites were separated in 70% methanol as described above. Mass spectrometry was carried out with a Finnegan MAT TSQ700 at 70 eV.

ANF Spectral Binding Studies.

Spectral binding assays were carried out as described in Harlow and Halpert (1998). P450 samples were diluted to 0.5 μM in 0.1 mg/ml DOPC, 0.05% CHAPS, and 50 mM HEPES (pH 7.6) and divided into 7 or 8 equal aliquots. A 0.01 volume of ANF, at various concentrations, was added to each aliquot with a final methanol concentration of 1%. Difference spectra were recorded on a Beckman DU-7 spectrophotometer from 500 to 340 nm, with a protein sample containing methanol alone as a reference. Because ANF absorbs in the range used, the absorbance of ANF at each concentration studied was also determined and subtracted from the absorbance change for each sample. The absorbance difference between 388 and 420 nm for each sample was calculated, and the data were analyzed with the Hill equation ΔA = (ΔAmaxSn)/(S50n+ Sn).

Results

Kinetics of Progesterone 6β-Hydroxylation.

As shown previously, in the absence of ANF, P450 3A4 wild type (WT) activity was sigmoidal over a range of progesterone concentrations from 5 to 150 μM (Fig. 2). Historically, the Hill equation has been used to analyze P450 3A4-catalyzed reactions that display positive cooperativity. However, the information that can be obtained from this equation is limited. The n value obtained from the Hill equation has no quantitative value, and the equation assumes that all substrate-binding sites are equivalent. Recently,Korzekwa et al. (1998) proposed a two-site equation as an alternative to the Hill equation for analyzing sigmoidal 3A4 kinetics. The two-site equation can provide Km andVmax values for two potential binding sites. However, it is very difficult to derive a unique solution due to the inherent errors in the experimental data (Korzekwa et al., 1998). Our initial attempts to use the full two-site model with a number of initial parameter estimates suggested thatVmax1 ≪Vmax2 (data not shown). Consequently, we applied a modified two-site equation, in which it is assumed that the enzyme and substrate can form an enzyme-substrate (ES) or an ESS complex, but that only the ESS complex results in product formation. Therefore, Vmax1 is set to zero. In this way, separate kinetic constants can be determined for two sites with a reasonable number of data points. This equation was very useful for analyzing 3A4 WT data and indicated that theKm1 value is much smaller than theKm2 value (Table1 and Fig. 2). In the presence of 25 μM ANF, however, 3A4 exhibited hyperbolic kinetics (Table 1 and Fig. 2). Analysis with the modified two-site equation showed that the addition of ANF to the reactions decreased theKm2 value from 110.8 to 33.7 μM compared with 37.0 μM when the analysis was performed with the Michaelis-Menten equation.

Figure 2
Figure 2

Kinetic analysis of 6β-OH progesterone production by wild-type 3A4 in the absence and presence of ANF. Assays were performed as described in Experimental Procedures. ●, absence of ANF; ■, presence of 25 μM ANF. Data in the absence of ANF were analyzed with the modified two-site equation described inExperimental Procedures and used to create the ideal curve shown in the figure [Km1=19.9 μM (S.D. = 9.5), Km2=89.6 (S.D. = 27.2),Vmax2=45.4 nmol/min/nmol (S.D. = 5.2)]. The Michaelis-Menten equation was used to analyze the data collected from samples with ANF to generate the ideal curve [Km=39.7 (S.D. = 2.9),Vmax=40.4 (S.D. = 1.0)].

View this table:
Table 1

Kinetic constants of progesterone 6β-hydroxylation catalyzed by P450 3A4 and mutants in the absence and presence of 25 μM ANF

In a previous study, mutant L211F/D214E was refractory to ANF stimulation of testosterone hydroxylase activity (Harlow and Halpert, 1998). To elucidate the basis of the residual stimulation of 6β-OH progesterone formation observed with this double mutant, we performed a kinetic analysis of L211F/D214E progesterone hydroxylation (Fig.3A and Table 1). The double mutant displayed hyperbolic kinetics in the absence of ANF and a 2-fold decrease in the Km value when 25 μM ANF was included in the reactions. The single mutant F304W also demonstrated hyperbolic kinetics in the absence of ANF, and like the double mutant L211F/D214E, a 2-fold decrease in theKm value for F304W progesterone hydroxylation in the presence of 25 μM ANF (Fig. 3B and Table 1). However, when these mutations were combined in the triple mutant L211F/D214E/F304W, the addition of 25 μM ANF only slightly altered the Km for 6β-OH progesterone production (Fig. 3C and Table 1). In fact, in the absence of ANF, L211F/D214E/F304W showed a 2-fold decrease in theKm compared with theKm2 of 3A4 WT. The data in Fig. 2demonstrate that for individual experiments, the data fit the Michaelis-Menten equation well, and Table 1 illustrates that despite variance among separate experiments, meaningful comparisons between the kinetic constants of the individual mutants can be made. The data in Table 1 also show that the Vmax values for L211F/D214E, F304W, and L211F/D214E/F304W were all approximately the same as the value for the wild type. P450 3A4 WT and L211F/D214E/F304W progesterone hydroxylase assays were performed in the presence of 0 μM to 50 μM ANF, the concentration of which was limited by its solubility in 1% methanol, the final concentration in the reactions. Data in Table 2 illustrate that ANF stimulated wild-type 3A4 and neither stimulated nor inhibited the progesterone hydroxylase activity of the triple mutant significantly.

Figure 3
Figure 3

Kinetic analysis of P450 3A4 mutants for progesterone 6βhydroxylation. ●, absence of ANF; ■, presence of 25 μM ANF. Data were analyzed with the Michaelis-Menten equation as described in Experimental Procedures to create the ideal curves shown. A, mutant L211F/D214E, values in the absence of ANF wereKm=79.7 (S.D. = 9.1) andVmax= 52.0 (S.D. = 2.8), and values in the presence of ANF were Km=27.1 (S.D. = 4.0) and Vmax= 38.4 (S.D. = 1.7). B, F304W, values in the absence of ANF were Km=78.8 (S.D. = 13.7) and Vmax= 49.1 (S.D. = 4.0), and values in the presence of ANF wereKm=34.0 (S.D. = 7.2) andVmax= 34.3 (S.D. = 2.5). C, L211F/D214E/F304W, values in the absence of ANF wereKm=48.4 (S.D. = 4.9) andVmax= 45.8 (S.D. = 1.8), and values in the presence of ANF were Km=38.9 (S.D. = 2.9) and Vmax= 40.6 (S.D. = 1.1).

View this table:
Table 2

Effect of increasing concentrations of ANF on progesterone 6β-hydroxylation

ANF Oxidation Assays.

Because ANF is also a substrate of P450 3A4 (Ueng et al., 1997), it was important to study the effect of the substitutions at positions 211, 214, and 304 on ANF oxidation. Compared with wild type, L211F/D214E showed only a slight decrease in the rate of ANF 5,6-oxide formation (Fig. 4). However, F304W displayed a 4.6-fold and L211F/D214E/F304W displayed an 8.2-fold loss of ANF oxidation. Higher concentrations of ANF did not significantly alter the oxidation rates (data not shown), indicating that the concentration of 25 μM was essentially saturating. In addition, the ability of progesterone to inhibit or activate ANF oxidation was studied. When progesterone was added to reactions at concentrations ranging up to 100 μM, there was no effect on the rate of ANF oxidation for 3A4 WT or the triple mutant L211F/D214E/F304W (data not shown).

Figure 4
Figure 4

Production of ANF 5,6-oxide by 3A4 and mutants. The assays were performed as described in Experimental Procedures using 25 μM ANF. The values represent the results of three to seven assays and are as follows: wild type, 2.86 nmol/min/nmol P450; L211F/D214E, 2.22 nmol/min/nmol P450; F304W, 0.62 nmol/min/nmol P450; and L211F/D214E/F304W, 0.35 nmol/min/nmol P450.

ANF Spectral Binding Assays.

After establishing the effect of the triple mutation on ANF oxidation, the binding of ANF to 3A4 WT, L211F/D214E, and L211F/D214E/F304W was studied in an attempt to identify the step at which ANF oxidation was altered. Increasing concentrations of ANF were added to each enzyme preparation and absorbance changes were monitored. Figure5 demonstrates the data analyzed with the Hill equation. The maximal change in absorbance was similar for 3A4 WT (ΔAmax=70.1 mM−1, S.D. = 6), L211F/D214E (ΔAmax=69.1 mM−1, S.D. = 8), and L211F/D214E/F304W (ΔAmax= 58.1 mM−1, S.D. = 2), suggesting that binding is not significantly affected by the amino acid alterations. It is noteworthy that the triple mutant appears to bind ANF more tightly, S50=5.9 μM (S.D. = 0.4), compared with wild type and the double mutant, S50=17.0 μM (S.D. = 2) and 15.0 μM (S.D. = 3), respectively (Fig. 5). All of the samples showed sigmoidal kinetics with n values of 1.9 for the wild type and L211F/D214E and 1.6 for L211F/D214E/F304W.

Figure 5
Figure 5

ANF spectral binding studies with 3A4 WT and mutants. The reactions were performed as described in Experimental Procedures. ●, WT 3A4; ▵, L211F/D214E; and ▪, L211F/D214E/F304W. The values shown represent the mean of two or three assays.

Discussion

In this study, the structural basis of P450 3A4 cooperativity was examined with a combination of molecular modeling and site-directed mutagenesis. The goals were to localize the effector-binding site and to elucidate the relationship between ANF oxidation and enzyme activation. Previous work suggested that Leu-211 and Asp-214 as well as additional residues comprise an effector-binding site (Harlow and Halpert, 1998), and molecular modeling (Fig. 1) predicted that the central location of Phe-304 in the large binding pocket of P450 3A4 would allow this residue to interact with both substrate and effector. Mutant F304W showed hyperbolic progesterone 6β-hydroxylation kinetics in the absence of ANF and diminished ANF stimulation compared with wild-type 3A4. L11F/D214E/F304W also displayed hyperbolic kinetics, but unlike the single F304W mutant or L211F/D214E, the triple mutant showed little change in the Km for progesterone in the presence of ANF. Mutants F304W and L211F/D214E/F304W showed decreased ANF 5,6-oxide production compared with 3A4 WT. Interestingly, the ability of the triple mutant to bind ANF, as shown by the magnitude of the type I spectral change, was unaffected.

The discovery of a pivotal role of residue Phe-304 in homotropic and heterotropic cooperativity is an important advance relative to the information gained from the double mutant L211F/D214E. In addition, the use of the modified two-site equation clarified the effects exerted by ANF. Although previous use of the Hill equation provided clues that ANF has two effects on steroid hydroxylation by 3A4 WT, a decrease in the S50 and in the n value (Domanski et al., 1998; Harlow and Halpert, 1998), the S50 is not equivalent to a Km. With the modified two-site equation the decrease inKm2 in the presence of ANF could be quantified. In addition, one interpretation of the conversion to hyperbolic kinetics is a decrease inKm1 to an undetectable level. L211F/D214E and F304W were indistinguishable kinetically, and neither was able to fully mimic effector-bound wild-type enzyme, as illustrated by the decrease in the Km of both mutants on addition of ANF. The location of Phe-304 in the highly conserved I helix (Hasemann et al., 1995) greatly reinforces the contention that this residue along with Leu-211 and Asp-214 in the variable F helix comprise an effector-binding site. In support of this interpretation, recent crystallographic findings by Cupp-Vickery demonstrate that two molecules of androstenedione are present in the P450eryF active site (Anderson and Cupp-Vickery, 1999), and that the second molecule is <5 Å from residues corresponding to P450 3A4 residues Phe-304, Leu-211, and Asp-214 (J. Cupp-Vickery, personal communication). Docking studies with our own 3A4 model also suggest that a second molecule of ANF or progesterone could contact these three residues.

The results of this and other recent studies provide further evidence for the overlapping nature of the effector and substrate-binding sites in 3A4. Recently, we reported on the effect of a Phe-304 → Ala substitution in P450 3A4 (Domanski et al., 1998). Mutant F304A displayed increased progesterone 6β-hydroxylase activity and an altered ratio of 6β-OH:16α-OH products in the absence of ANF, but retained responsiveness to ANF stimulation similar to 3A4 WT. This finding, in conjunction with the results reported herein on mutant F304W, suggests that residue Phe-304 acts as a contact point for both substrate and effector. A similar dual role for residue Leu-211 has been noted because substitutions alter not only cooperativity of progesterone and testosterone hydroxylation (Harlow and Halpert, 1998) but also stereospecificity of oxidation of the airway-specific steroid 20R-16α,17α-[butylidenebis(oxy)]-6α,9α-difluoro-11β-hydroxy-17β-(methylthio)androsta-4-en-3-one (Stevens et al., 1999). Consequently, it is becoming clear that the substrate and effector sites are closely linked and contain residues that can be important to either substrate and/or effector binding, depending on the molecule present.

Thus far, three models to explain P450 3A cooperativity have been proposed that involve double occupancy of the binding pocket. Korzekwa et al. (1998) put forth the idea of a two substrate-bound active site, in which both substrates would need to have access to the reactive oxygen through translations and rotations that occur within the time frame of the oxidation step. The relative orientation of the substrates was not defined. The model of Shou et al. (1999) is an extension of the model of Korzekwa et al. (1998) and proposes two distinct substrate-binding sites, with each substrate having a preferred orientation. Our laboratory recently suggested the presence of separate effector- and substrate-binding sites with only the latter having access to the reactive oxygen (Harlow and Halpert, 1998). In light of the results from this study, we have had difficulty reconciling all of our data to any of these three models. For example, although two substrate-bound or two-site models can explain heterotropic cooperativity and lack of competitive inhibition between two substrates that show hyperbolic kinetics, the situation becomes more complicated with individual substrates such as ANF and progesterone that show cooperative binding and/or kinetics. To explain the lack of inhibition observed between these two compounds, one would have to assume that either substrate can bind at two locations when alone, but when combined each gravitates toward a single, preferred location. An attractive alternative is that each substrate occupies a preferred location in the vicinity of the active oxygen and that the substrates compete for a more distant effector site. Such a possibility was initially proposed by Ueng et al. (1997) to account for lack of inhibition between aflatoxin B1 and ANF, and is shown pictorially in Fig. 6. Triple occupancy also was entertained by Shou et al. (1994) and most recently by Hosea and Guengerich (1999). The key distinction between models involving double as opposed to triple occupancy of the binding pocket is whether effectors such as ANF are binding at the same location when serving as activators as when serving as substrates. A crucial observation of Korzekwa et al. (1998) was that phenanthrene and ANF showed binding constants as effectors similar to their respectiveKm values as substrates, suggesting that effectors bind at the same site when acting as substrates. However, in that experimental system, neither substrate showed homotropic cooperativity of binding or kinetics. Our mutagenesis data provide some evidence that activation by ANF and ANF oxidation can be dissociated. Thus, although L211F/D214E and F304W showed a similar, diminished response to stimulation of progesterone 6β-hydroxylation by ANF, the mutants differed in their ability to oxidize ANF, with only F304W demonstrating a large decrease in 5,6-oxide production. This result suggests that ANF may bind at different locations, depending on its role as an effector or as a substrate.

Figure 6
Figure 6

Model of P450 3A4 binding pocket.

In addition, the studies of ANF oxidation suggest that ANF is able to bind to L211F/D214E/F304W but is prevented from completing the P450 catalytic cycle and forming product. This apparent discrepancy may result from differences in how ANF binds to different redox states of the enzyme. Studies with P450BM-3 have shown that initial binding of substrate to the ferric form of the enzyme occurs at a significant distance from the heme group, although a spin shift occurs (Modi et al., 1996). Only after reduction and a shift of the substrate of several angstroms closer to the heme, can oxidation occur. In the triple mutant, ANF may bind initially but be inhibited by the altered side chains from getting close enough to the active oxygen to complete the reaction cycle. Alternatively, a more conservative explanation is that there is only a slight alteration in the orientation of ANF within the binding pocket of the triple mutant that is, however, sufficient to prevent product formation. This interpretation is consistent with previous work in our laboratory on 2B enzymes, showing uncoupling of product formation from NADPH and oxygen consumption in certain site-directed mutants (Fang et al., 1997). In addition, it appears that in the ferric state the triple mutant binds ANF more tightly than the wild type, although the implications of this are still unclear.

In conclusion, the data from this study and the recent findings with P450eryF (Anderson and Cupp-Vickery, 1999) strongly suggest that progesterone and ANF occupy different positions within the 3A4 active site but would each have access to the reactive oxygen. Our data and the recent findings with P450eryF (Anderson and Cupp-Vickery, 1999) also provide further evidence that 3A4 cooperativity represents the ability of the large binding pocket to accommodate multiple ligands, although the possibility cannot be ruled out that progesterone and ANF are oxidized by different conformers of 3A4 (Koley et al., 1996). Studies are ongoing to better understand the relationship between the role of ANF as an activator and as a substrate and to map additional residues involved in 3A4 cooperativity. For example, although progesterone 6β-hydroxylation by L211F/D214E/F304W is not responsive to ANF, 16α-hydroxylation by this mutant remains partially responsive to ANF, whereas kinetic analysis revealed hyperbolic behavior (data not shown). Other substrate/effector pairs also must be analyzed. Ultimately, thorough delineation of the structural features of the effector and substrate oxidation sites should allow rational predictions of drug-drug interactions involving P450 3A4.

Acknowledgments

We thank Dr. Grazyna Szklarz for the use of the P450 3A4 model.

Footnotes

  • Send reprint requests to: Dr. Tammy L. Domanski, Department of Pharmacology and Toxicology, University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555-1031. E-mail: tadomans{at}utmb.edu

  • ↵1 This study was supported by National Research Award GM19058, National Institutes of Health Grant GM54995, and Center Grant ES06676.

  • Received for publication

  • Abbreviations:
    P450
    cytochrome P450
    ANF
    α-naphthoflavone
    PCR
    polymerase chain reaction
    DOPC
    dioloeoylphosphatidylcholine
    CHAPS
    3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid
    E
    enzyme
    S
    substrate
    WT
    wild type
    • Accepted December 23, 1999.

References

  1. ↵
    1. Anderson R,
    2. Cupp-Vickery J
    (1999) Crystallographic evidence for the simultaneous binding of two substrates to cytochrome P450eryF. FASEB J 13:A1345.
    OpenUrl
  2. ↵
    1. Domanski TL,
    2. Liu J,
    3. Harlow GR,
    4. Halpert JR
    (1998) Analysis of four residues within substrate recognition site 4 of human cytochrome P450 3A4: Role in steroid hydroxylase activity and α-naphthoflavone stimulation. Arch Biochem Biophys 350:223–232.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Fang X,
    2. Kobayashi Y,
    3. Halpert JR
    (1997) Stoichiometry of 7-ethoxycoumarin metabolism by cytochrome P450 2B1 wildtype and five active-site mutants. FEBS Lett 416:77–80.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Guengerich FP
    (1990) Mechanism-based inactivation of human liver microsomal cytochrome P450 IIIA4 by gestodene. Chem Res Toxicol 3:363–371.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Ortiz de Montellano
    1. Guengerich FP
    (1995) Human cytochrome P450 enzymes. in Cytochromes P450, ed Ortiz de Montellano (Plenum Press, New York), pp 473–535.
  6. ↵
    1. Harlow GR,
    2. Halpert JR
    (1997) Alanine-scanning mutagenesis of a putative substrate recognition site in human cytochrome P450 3A4. J Biol Chem 272:5396–5402.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Harlow GR,
    2. Halpert JR
    (1998) Analysis of human cytochrome P450 3A4 cooperativity: Construction and characterization of a site-directed mutant that displays hyperbolic steroid hydroxylation kinetics. Proc Natl Acad Sci USA 95:6636–6641.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Hasemann CA,
    2. Kurumbail RG,
    3. Boddupalli SS,
    4. Peterson JA,
    5. Deisenhofer J
    (1995) Structure and function of cytochromes P450: A comparative analysis of three crystal structures. Structure 3:41–62.
    OpenUrlCrossRefPubMed
  9. ↵
    1. He YA,
    2. He YQ,
    3. Szklarz GD,
    4. Halpert JR
    (1997) Identification of three key residues in substrate recognition site 5 of human cytochrome P450 3A4 by cassette and site-directed mutagenesis. Biochemistry 36:8831–8839.
    OpenUrlCrossRefPubMed
  10. ↵
    1. He YQ,
    2. He YA,
    3. Halpert JR
    (1995) Escherichia coli expression of site-directed mutants of cytochrome P450 2B1 from six substrate recognition sites: Substrate specificity and inhibitor selectivity studies. Chem Res Toxicol 8:574–579.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Hosea NA,
    2. Guengerich FP
    (1999) Characterization of ligand binding sites in P450 3A4 with YPFP-NH2. Ninth North American ISSX Proceedings: 126.
  12. ↵
    1. Kerlan V,
    2. Dreano Y,
    3. Bercovici JP,
    4. Beaune PH,
    5. Floch HH,
    6. Berthou F
    (1992) Nature of cytochromes P450 involved in the 2-/4-hydroxylations of estradiol in human liver microsomes. Biochem Pharmacol 44:1745–1756.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Kerr BM,
    2. Thummel KE,
    3. Wurden CJ,
    4. Klein SM,
    5. Kroetz DL,
    6. Gonzalez FJ,
    7. Levy RH
    (1994) Human liver carbamazepine metabolism. Role of CYP3A4 and CYP2C8 in 10,11-epoxide formation. Biochem Pharmacol 47:1969–1979.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Koley AP,
    2. Robinson RC,
    3. Friedman FK
    (1996) Cytochrome P450 conformation and substrate interactions as probed by binding kinetics. Biochimie 78:9706–9713.
    OpenUrl
  15. ↵
    1. Korzekwa KR,
    2. Krishnamachary N,
    3. Shou M,
    4. Ogai A,
    5. Parise RA,
    6. Rettie AE,
    7. Gonzalez FJ,
    8. Tracy TS
    (1998) Evaluation of atypical cytochrome P450 kinetics with two-substrate models: Evidence that multiple substrates can simultaneously bind to cytochrome P450 active sites. Biochemistry 37:4137–4147.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Kronbach T,
    2. Mathys D,
    3. Umeno M,
    4. Gonzalez FJ,
    5. Meyer UA
    (1989) Oxidation of midazolam and triazolam by human liver cytochrome P450 IIIA4. Mol Pharmacol 36:89–96.
    OpenUrlAbstract
  17. ↵
    1. Lasker JM,
    2. Huang MT,
    3. Conney AH
    (1984) In vitro and in vivo activation of oxidative drug metabolism by flavonoids. J Pharmacol Exp Ther 229:162–170.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Li Y,
    2. Wang E,
    3. Patten CJ,
    4. Chen L,
    5. Yang C
    (1994) Effects of flavonoids on cytochrome P450-dependent acetaminophen metabolism in rats and human liver microsomes. Drug Metab Dispos 22:566–571.
    OpenUrlAbstract
  19. ↵
    1. Ludwig E,
    2. Schmid J,
    3. Beschke K,
    4. Ebner T
    (1999) Activation of human cytochrome P450 3A4 catalyzed meloxicam 5′-methylhydroxylation by quinidine and hydroquinidine in vitro. J Pharmacol Exp Ther 290:1–8.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Modi S,
    2. Sutcliffe MJ,
    3. Primrose WU,
    4. Lian LY,
    5. Roberts GCK
    (1996) The catalytic mechanism of cytochrome P450 BM3 involves a 6 angstrom movement of the bound substrate on reduction. Nat Struct Biol 3:414–417.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Schwab GE,
    2. Raucy JL,
    3. Johnson EF
    (1988) Modulation of rabbit and human hepatic cytochrome P450 catalyzed steroid hydroxylations by α-naphthoflavone. Mol Pharmacol 33:493–499.
    OpenUrlAbstract
  22. ↵
    1. Shimada T,
    2. Guengerich FP
    (1989) Evidence for cytochrome P450NF, the nifedipine oxidase, being the principal enzyme involved in the bioactivation of aflatoxins in human liver. Proc Natl Acad Sci USA 86:462–465.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Shou M,
    2. Ettore MW,
    3. Dai R,
    4. Baillie TA,
    5. Rushmore TH
    (1999) Sigmoidal kinetic model for two cooperative substrate-binding sites in a cytochrome P450 3A4 active site: An example of the metabolism of diazepam and its derivatives. Biochem J 340:845–853.
  24. ↵
    1. Shou M,
    2. Grogan J,
    3. Mancewica JA,
    4. Krausz KW,
    5. Gonzalez FJ,
    6. Gelboin HV,
    7. Korzekwa KR
    (1994) Activation of CYP 3A4: Evidence for the simultaneous binding of two substrates in a cytochrome P450 active site. Biochemistry 33:6450–6455.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Stevens JC,
    2. Domanski TL,
    3. Harlow GR,
    4. White RB,
    5. Orton E,
    6. Halpert JR
    (1999) Use of the steroid RPR 106541 in combination with site-directed mutagensis for enhanced CYP3A4 structure/function analysis. J Pharmacol Exp Ther 290:594–602.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Szklarz GD,
    2. Halpert JR
    (1997) Molecular modeling of cytochrome P450 3A4. J Comput Aided Mol Des 11:265–272.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Ueng Y,
    2. Shimada T,
    3. Yamazaki H,
    4. Guengerich FP
    (1995) Oxidation of aflatoxin B1 by bacterial recombinant human cytochrome P450 enzymes. Chem Res Toxicol 8:218–225.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Ueng YF,
    2. Kuwabara T,
    3. Chun YJ,
    4. Guengerich FP
    (1997) Cooperativity in oxidations catalyzed by cytochrome P450 3A4. Biochemistry 36:370–381.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Waxman DJ,
    2. Lapenson DP,
    3. Aoyama T,
    4. Gelboin HV,
    5. Gonzalez FJ,
    6. Korzekwa K
    (1991) Steroid hormone hydroxylase specificities of eleven cDNA-expressed human cytochrome P450s. Arch Biochem Biophys 290:160–166.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleABSORPTION, DISTRIBUTION, METABOLISM, AND EXCRETION

Dual Role of Human Cytochrome P450 3A4 Residue Phe-304 in Substrate Specificity and Cooperativity

Tammy L. Domanski, You-Ai He, Greg R. Harlow and James R. Halpert
Journal of Pharmacology and Experimental Therapeutics May 1, 2000, 293 (2) 585-591;
Reddit logo Twitter logo Facebook logo Mendeley logo

Jump to section

Advertisement