![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 293, Issue 2, 585-591, May 2000
University of Texas Medical Branch, Department of Pharmacology and Toxicology, Galveston, Texas
 |
Abstract |
|---|
 Top
 Abstract
 Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
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 
-naphthoflavone, exhibiting a 2-fold decrease in the
Km value for progesterone 6
-hydroxylation
in the presence of 25 µM effector. Combining substitutions to yield the triple mutant L211F/D214E/F304W maintained the
Vmax and decreased the
Km for progesterone 6-hydroxylation,
minimized stimulation by -naphthoflavone, and decreased the rate of
-naphthoflavone oxidation to one-eighth of the wild type.
Interestingly, the 
Amax for spectral
binding of -naphthoflavone was unaltered in L211F/D214E/F304W. Overall, the results suggest that progesterone and -naphthoflavone are oxidized at separate locations within the P450 3A4 binding pocket,
although both substrates appear to have equal access to the reactive oxygen.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
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 17-estradiol (Ueng et al., 1997), as
well as with aflatoxin B1 (Ueng et al., 1997) and
amitriptyline (Ueng et al., 1997). The flavonoid -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 6-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
6-hydroxy- and 16-hydroxy testosterone formation by ANF (Harlow and Halpert, 1998). However, ANF still stimulated the formation of
6-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 6-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.
|
| |
Experimental Procedures |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
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 DH5
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).
). Solubilized membranes were prepared and the P450s were purified on Talon metal affinity columns (Clontech, Palo Alto, CA) under conditions described in Domanski 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 cytochrome
b5 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 cm1)
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 A = (AmaxSn)/(S50n
+ Sn).
| |
Results |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
Kinetics of Progesterone 6-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 and
Vmax 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 that
Vmax1 
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 the
Km1 value is much smaller than the
Km2 value (Table
1 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 the
Km2 value from 110.8 to 33.7 µM
compared with 37.0 µM when the analysis was performed with the
Michaelis-Menten equation.
|
|
). To elucidate the basis of the residual stimulation of 6-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 the
Km 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 6-OH progesterone production (Fig. 3C and Table 1). In fact, in the absence of ANF,
L211F/D214E/F304W showed a 2-fold decrease in the
Km compared with the
Km2 of 3A4 WT. The data in Fig. 2
demonstrate 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.
|
|
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).
|
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. Figure 5 demonstrates the data analyzed with the
Hill equation. The maximal change in absorbance was similar for 3A4 WT
(Amax=70.1 mM1, S.D. = 6), L211F/D214E (Amax=69.1
mM1, S.D. = 8), and L211F/D214E/F304W
(Amax= 58.1 mM1,
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.
|
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
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 6-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 in
Km2 in the presence of ANF could be
quantified. In addition, one interpretation of the conversion to
hyperbolic kinetics is a decrease in
Km1 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 6-hydroxylase activity and an altered ratio of 6-OH:16-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-16,17-[butylidenebis(oxy)]-6,9-difluoro-11-hydroxy-17-(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 respective
Km 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 6-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.
|
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
6-hydroxylation by L211F/D214E/F304W is not responsive to ANF,
16-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 |
|---|
1 This study was supported by National Research Award GM19058, National Institutes of Health Grant GM54995, and Center Grant ES06676.
Received for publication
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
| |
Abbreviations |
|---|
P450, cytochrome P450;
ANF, -naphthoflavone;
PCR, polymerase chain reaction;
DOPC, dioloeoylphosphatidylcholine;
CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid;
E, enzyme;
S, substrate;
WT, wild type.
| |
References |
|---|
|
Top
Abstract
Introduction
Experimental Procedures
Results
Discussion
References
|
|---|
-naphthoflavone stimulation.
Arch Biochem Biophys
350:
223-232[Medline].-naphthoflavone.
Mol Pharmacol
33:
493-499[Abstract].This article has been cited by other articles:
|
|
 |
|
  N. Torimoto, I. Ishii, K.-I. Toyama, M. Hata, K. Tanaka, H. Shimomura, H. Nakamura, N. Ariyoshi, S. Ohmori, and M. Kitada Helices F-G Are Important for the Substrate Specificities of CYP3A7 Drug Metab. Dispos., March 1, 2007; 35(3): 484 - 492. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S.-J. Lee, I. P. van der Heiden, J. A. Goldstein, and R. H. N. van Schaik A New CYP3A5 Variant, CYP3A5*11, Is Shown to Be Defective in Nifedipine Metabolism in a Recombinant cDNA Expression System Drug Metab. Dispos., January 1, 2007; 35(1): 67 - 71. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kumar, D. R. Davydov, and J. R. Halpert ROLE OF CYTOCHROME B5 IN MODULATING PEROXIDE-SUPPORTED CYP3A4 ACTIVITY: EVIDENCE FOR A CONFORMATIONAL TRANSITION AND CYTOCHROME P450 HETEROGENEITY Drug Metab. Dispos., August 1, 2005; 33(8): 1131 - 1136. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.-C. Egnell, J. B. Houston, and C. S. Boyer Predictive Models of CYP3A4 Heteroactivation: In Vitro-in Vivo Scaling and Pharmacophore Modeling J. Pharmacol. Exp. Ther., March 1, 2005; 312(3): 926 - 937. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. K. Yano, M. R. Wester, G. A. Schoch, K. J. Griffin, C. D. Stout, and E. F. Johnson The Structure of Human Microsomal Cytochrome P450 3A4 Determined by X-ray Crystallography to 2.05-A Resolution J. Biol. Chem., September 10, 2004; 279(37): 38091 - 38094. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kalbag, C. Elder, G. Scott, Y. Wang, S. Milosavljev, P. Leese, J. Caldwell, and C. Rordorf Concomitant Administration of Lumiracoxib and a Triphasic Oral Contraceptive Does Not Affect Contraceptive Activity or Pharmacokinetic Profile J. Clin. Pharmacol., June 1, 2004; 44(6): 646 - 654. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.-C. Egnell, C. Eriksson, N. Albertson, B. Houston, and S. Boyer Generation and Evaluation of a CYP2C9 Heteroactivation Pharmacophore J. Pharmacol. Exp. Ther., December 1, 2003; 307(3): 878 - 887. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. C. Patki, L. L. von Moltke, and D. J. Greenblatt IN VITRO METABOLISM OF MIDAZOLAM, TRIAZOLAM, NIFEDIPINE, AND TESTOSTERONE BY HUMAN LIVER MICROSOMES AND RECOMBINANT CYTOCHROMES P450: ROLE OF CYP3A4 AND CYP3A5 Drug Metab. Dispos., July 1, 2003; 31(7): 938 - 944. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.-C. Egnell, B. Houston, and S. Boyer In Vivo CYP3A4 Heteroactivation Is a Possible Mechanism for the Drug Interaction between Felbamate and Carbamazepine J. Pharmacol. Exp. Ther., June 1, 2003; 305(3): 1251 - 1262. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. K. Khan, H. Liu, and J. R. Halpert Homotropic Versus Heterotopic Cooperativity of Cytochrome P450eryF: A Substrate Oxidation and Spectral Titration Study Drug Metab. Dispos., April 1, 2003; 31(4): 356 - 359. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Nakamura, N. Torimoto, I. Ishii, N. Ariyoshi, H. Nakasa, S. Ohmori, and M. Kitada CYP3A4 and CYP3A7-Mediated Carbamazepine 10,11-Epoxidation Are Activated by Differential Endogenous Steroids Drug Metab. Dispos., April 1, 2003; 31(4): 432 - 438. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Galetin, S. E. Clarke, and J. B. Houston Quinidine and Haloperidol as Modifiers of CYP3A4 Activity: Multisite Kinetic Model Approach Drug Metab. Dispos., December 1, 2002; 30(12): 1512 - 1522. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Nakamura, H. Nakasa, I. Ishii, N. Ariyoshi, T. Igarashi, S. Ohmori, and M. Kitada Effects of Endogenous Steroids on CYP3A4-Mediated Drug Metabolism by Human Liver Microsomes Drug Metab. Dispos., May 1, 2002; 30(5): 534 - 540. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Hutzler and T. S. Tracy Atypical Kinetic Profiles in Drug Metabolism Reactions Drug Metab. Dispos., April 1, 2002; 30(4): 355 - 362. [Full Text] [PDF]
|
|
|
|
 |
|
 K. K. Khan, Y. Q. He, T. L. Domanski, and J. R. Halpert Midazolam Oxidation by Cytochrome P450 3A4 and Active-Site Mutants: an Evaluation of Multiple Binding Sites and of the Metabolic Pathway That Leads to Enzyme Inactivation Mol. Pharmacol., March 1, 2002; 61(3): 495 - 506. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. E. Kenworthy, S. E. Clarke, J. Andrews, and J. B. Houston Multisite Kinetic Models for CYP3A4: Simultaneous Activation and Inhibition of Diazepam and Testosterone Metabolism Drug Metab. Dispos., December 1, 2001; 29(12): 1644 - 1651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Lu, Y. Lin, A. D. Rodrigues, T. H. Rushmore, T. A. Baillie, and M. Shou Testosterone, 7-Benzyloxyquinoline, and 7-Benzyloxy-4-trifluoromethyl-coumarin Bind to Different Domains within the Active Site of Cytochrome P450 3A4 Drug Metab. Dispos., November 1, 2001; 29(11): 1473 - 1479. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
|
,
absence of ANF; 
, presence of 25 µM ANF. Data in the absence of
ANF were analyzed with the modified two-site equation described in
Experimental 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)].
, L211F/D214E; and 
,
L211F/D214E/F304W. The values shown represent the mean of two or three
assays.
