Download topological changes in the cyp3a4 active site probed with

0090-9556/04/3201-155–161$20.00
DRUG METABOLISM AND DISPOSITION
Copyright © 2004 by The American Society for Pharmacology and Experimental Therapeutics
DMD 32:155–161, 2004
Vol. 32, No. 1
1220/1116439
Printed in U.S.A.
TOPOLOGICAL CHANGES IN THE CYP3A4 ACTIVE SITE PROBED WITH
PHENYLDIAZENE: EFFECT OF INTERACTION WITH NADPH-CYTOCHROME P450
REDUCTASE AND CYTOCHROME B5 AND OF SITE-DIRECTED MUTAGENESIS
Yoshitaka Yamaguchi, Kishore K. Khan,1 You Ai He, You Qun He, and James R. Halpert
Department of Pharmacology and Toxicology, University of Texas Medical Branch, Galveston, Texas
(Received June 30, 2003; accepted September 22, 2003)
This article is available online at http://dmd.aspetjournals.org
ABSTRACT:
pared in the absence of redox partners and in the presence of CPR,
b5, or both. Formation of all four regioisomers in CYP3A4 wild type,
particularly the minor ones, was reduced in the presence of b5.
CPR also greatly decreased the three minor isomers but increased
the major isomer significantly. The presence of b5 and CPR restored minor isomer formation and suppressed the enhancement
of NA formation caused by CPR alone. Interestingly, the effects of
the redox partners differed among representative active site mutants. In particular, the increase in NC upon substitution of Ala-370
with Phe was significantly reversed in the presence of redox partners, strongly suggesting that a conformational change occurs
around pyrrole ring C due to protein-protein interactions between
CYP3A4 and CPR or b5.
Previous studies have shown that aryldiazenes are useful reagents
to obtain topological information about the active site of cytochromes
P450 (P4502) (Ortiz de Montellano, 1995). Phenyldiazene binds initially to the heme iron in P450 to form a ␴-bonded phenyl-iron
complex. Upon subsequent oxidation, the phenyl group shifts to an
available nitrogen atom of the four pyrrole rings (A, B, C, and D). The
ratio of these four N-protoporphyrin IX regioisomers (NA, NB, NC, and
ND) reflects the available space above each pyrrole ring. In the case of
several bacterial P450 enzymes, topological information was consistent with the active site structure determined by X-ray crystallography
(Ortiz de Montellano, 1995). Therefore, this method has also been
applied to selected human P450 enzymes, including 2D6, 2E1, 3A4,
and 4A11 (Dierks et al., 1998; Mackman et al., 1996a,b; Schrag and
Wienkers, 2000).
CYP3A4 is the most abundant P450 enzyme in the human liver and
plays a significant role in the metabolism of a wide variety of drugs
(Guengerich, 1999; Nebert and Russell, 2002). Because of its pharmacological significance and the potential for drug-drug interactions,
CYP3A4 has been the subject of a large number of studies focused on
elucidating the relationship between structure and function and the
basis for the atypical kinetics often exhibited (Shou et al., 1994;
Korzekwa et al., 1998). In our laboratory, a CYP3A4 molecular model
was constructed using sequence alignment with four bacterial P450s
of known structure, which provided insight into the basis of substrate
recognition (Szklarz and Halpert, 1997). Functional analysis using
bacterially expressed CYP3A4 substrate recognition site (SRS) mutants resulted in the identification of a number of amino acid residues
involved in substrate or effector binding (Domanski and Halpert,
2001; Domanski et al., 2001; Harlow and Halpert, 1998; Khan et al.,
2002b,c; He et al., 2003). The results strongly support the inference
from other studies for the crucial role of multiple substrate binding
sites in CYP3A4 in atypical kinetics.
Another characteristic feature of CYP3A4 is a drastic enhancement
of its catalytic activities by cytochrome b5 (b5), the precise mechanism
of which remains controversial. For many years, the role of b5 was
thought to be enhancement of introduction of the second electron
required for oxygen activation (Schenkman and Jansson, 2003) However, a series of studies using CYP3A4 and apo-b5 strongly suggested
This work was supported by National Institutes of Health Grant GM54995 and
Center Grant ES06676.
1
Current address: In Vitro Drug Metabolism, Cedra Corporation, 8609 Cross
Park Drive, Austin, TX 78754.
2
Abbreviations used are: P450, cytochrome P450; SRS, substrate recognition
site; b5, cytochrome b5; CPR, NADPH-cytochrome P450 reductase; 7-BFC,
7-benzyloxy-4-(trifluoromethyl)coumarin; HPLC, high-performance liquid chromatography.
Address correspondence to: Dr. Yoshitaka Yamaguchi, Department of Drug
Metabolism and Pharmacokinetics, Development Research Laboratories, Shionogi &
Co., LTD., 3-1-1, Futaba-cho, Toyonaka, Osaka, 561-0825, Japan. E-mail:
[email protected]
155
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 14, 2017
The active site topology of heterologously expressed CYP3A4 purified from an Escherichia coli expression system was examined
using phenyldiazene. Incubation of CYP3A4 with phenyldiazene
and subsequent oxidation yielded all four potential N-phenylprotoporphyrin IX regioisomers derived from attack on an available
nitrogen atom in pyrrole rings B, A, C, or D (NB:NA:NC:ND ⴝ 6:73:7:
13). Further study using 28 active site mutants showed that substitution of residues closer to the heme, Ala-305, Thr-309, or Ala370, with a larger residue caused the most drastic changes in
regioisomer formation, which reflected the location of each amino
acid residue replaced in a CYP3A4 homology model. Previous
studies have suggested a conformational change in CYP3A4 upon
binding of NADPH-cytochrome P450 reductase (CPR) or cytochrome b5 (b5). Therefore, regioisomer formation was also com-
156
YAMAGUCHI ET AL.
Materials and Methods
Materials. Methyl phenyldiazene carboxylate azo ester was purchased from
Research Organics (Cleveland, OH). Imidazole, potassium ferricyanide, diazepam, midazolam, mifepristone, and horse skeleton myoglobin were purchased
from Sigma-Aldrich (St. Louis, MO), and 7-benzyloxy-4-(trifluoromethyl)coumarin (7-BFC) was obtained from BD Gentest (Woburn, MA). Recombinant CPR and b5 from rat liver were prepared as described previously (Harlow
and Halpert, 1997). All other chemicals were of the highest grade available
from standard commercial sources.
Expression and Purification of CYP3A4 and Mutants. CYP3A4 wild
type and mutants were expressed as His-tagged proteins in Escherichia coli
TOPP3 and purified using Talon metal affinity resin (BD Biosciences Clontech, Palo Alto, CA), as described previously (Domanski and Halpert, 2001;
Domanski et al., 2001; Harlow and Halpert, 1998; Khan et al., 2002b,c; He et
al., 2003). P450 contents were determined by measuring reduced carbon
monoxide difference spectra. Protein concentration was determined with the
bicinchoninic acid protein assay kit (Pierce, Rockford, IL) and bovine serum
albumin as a standard. The specific contents of CYP3A4 wild type and mutants
were 6 to 15 nmol of P450 per mg protein except for L373F (specific
content ⫽ 3).
Spectral Binding Studies. Binding spectra were recorded on a Shimazu2600 spectrophotometer fitted with a temperature controller. The solution in
the sample cuvette contained 1 nmol of CYP3A4 wild type in 1 ml of 100 mM
phosphate buffer (pH 7.4). Absolute spectra were measured between 350 and
600 nm using 1 ml of 100 mM phosphate buffer (pH 7.4) as a reference. Then,
spectral changes were monitored by adding aliquots of 65 mM methyl phenyldiazene carboxylate azo ester in 1 N KOH up to a final concentration of
0.13 mM to both sample and reference cuvettes.
Formation and Determination of N-Phenylprotoporphyrin IX Regioisomers. N-Phenylprotoporphyrin IX regioisomers were formed and determined as described previously (Swanson and Ortiz de Montellano, 1991; Tuck
et al., 1992; Khan et al., 2002a). CYP3A4 wild type or mutant (1 nmol) in 0.5
ml of 0.1 M MOPS buffer (pH 7.3) containing 10% glycerol, 0.1 mM
dithiothreitol, and 1 mM EDTA was treated with 10 ␮l of 65 mM methyl
phenyldiazene carboxylate azo ester solution in 1 N KOH. After 1 h, 2 ␮l of
62.5 mM potassium ferricyanide solution in the same buffer was added four
times at 5-min intervals, and the reaction was completed by allowing the
mixture to stand for 10 min after the final addition. All of these procedures
were performed at room temperature.
FIG. 1. Absolute spectral changes in CYP3A4 wild type (1 ␮M) with increasing
concentrations of phenyldiazene (a ⫽ 0, b ⫽ 0.13, c ⫽ 0.26, d ⫽ 0.52, and e ⫽
1.3 mM).
Then, protein was denatured by incubation for 2 h with 5 ml of freshly
prepared 5% (v/v) sulfuric acid dissolved in acetonitrile. The reaction mixtures
were concentrated to 1 to 2 ml under reduced pressure, and 2 ml of 5% (v/v)
aqueous sulfuric acid was added. N-Phenylprotoporphyrin IX regioisomers
were extracted three times with 1 ml of CH2Cl2. The extracts were washed
with 1 ml of water and dried under reduced pressure. The dried sample was
resuspended in 100 ␮l of solvent A for HPLC analysis.
The regioisomers were separated using a Partisil ODS-3 column (5 ␮m ⫻
250 mm ⫻ 4.6 mm; Alltech Associates, Deerfield, IL) by isocratic elution with
65:35 (v/v) solvent A (methanol/H2O/acetic acid, 6:4:1, v/v) and solvent B
(methanol) for 35 min at room temperature. The flow rate was 1.0 ml/min, and
regioisomers were monitored at 416 nm. Under these conditions, Nprotoporphyrin regioisomers NB, NA, NC, and ND eluted at 18 min, 20 min, 22
min, and 24.5 min, respectively, as confirmed by comparison with the standards formed using horse heart myoglobin (5 nmol).
Results
N-Phenylprotoporphyrin IX Regioisomer Formation from
CYP3A4 Wild Type. The addition of phenyldiazene to CYP3A4 wild
type yielded a typical peak at 478 nm with a decrease at 418 nm, as
expected for a phenyl-iron complex (Fig. 1). After oxidation using
ferricyanide, the products were analyzed by HPLC, revealing four
N-phenylprotoporphyrin IX regioisomers that were matched with the
standard products from myoglobin (Fig. 2). NA was the main product
and represented 73 ⫾ 2% (mean ⫾ S.D. of six individual determinations) of total regioisomer formation. NB, NC, and ND were minor and
constituted 6 ⫾ 2%, 7 ⫾ 1%, and 13 ⫾ 2%, respectively, of the total.
The amounts of these regioisomers were proportional to the P450
added in a range from 0.5 to 2.5 nmol/0.5 ml (data not shown).
Regioisomer formation was also determined in the presence of
MgCl2 (10 mM) and the following CYP3A4 substrates: midazolam
(25 or 250 ␮M), 7-BFC (100 ␮M), mifepristone (100 ␮M), and
diazepam (250 ␮M) (Table 1). There was little effect on any regioisomer formation of MgCl2 or 7-BFC. However, midazolam, mifepristone, and diazepam enhanced total regioisomer formation 1.4- to
1.8-fold with little change in the ratios.
N-Phenylprotoporphyrin IX Regioisomer Formation by
CYP3A4 SRS Mutants. In previous studies from our laboratory, a
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 14, 2017
that b5 also caused a structural change in CYP3A4, which contributed
to stimulation of monooxygenase activities (Yamazaki et al., 1996,
2001). More recently, b5 has been shown to alleviate substrate inhibition of CYP3A4 by triazolam (Schrag and Wienkers, 2001). In
addition, a similar structural change in CYP3A4 may occur upon
interaction with NADPH-cytochrome P450 reductase (CPR), because
basic amino acid residues responsible for interaction of CYP2B4 with
b5 or CPR are located on the proximal surface near the heme and
mostly overlap (Bridges et al., 1998). A conformational change in
P450 caused by CPR binding is also supported by a previous observation that the Km value of rat CYP2B1 or CYP1A1 was changed by
chemical modification of acidic amino acid residues on the surface of
CPR that are involved in interaction with the P450 (Strobel et al.,
1989).
In the present study, amino acid residues at 14 SRS positions in the
CYP3A4 active site were selected and substituted with a smaller or
larger side chain. The mutants were purified and tested to validate the
use of phenyldiazene as a topological probe. Major changes in Nprotoporphyrin regioisomer formation were observed in some SRS
mutants, which were largely consistent with the location of the residues relative to the heme in the CYP3A4 model. Subsequent comparison of N-protoporphyrin regioisomer formation in CYP3A4 wild
type and representative SRS mutants in the absence and presence of
CPR, b5, or both supplied compelling evidence for a structural change
in the CYP3A4 active site upon interaction with redox partners.
157
TOPOLOGICAL CHANGES IN CYP3A4 ACTIVE SITE
combination of molecular modeling and site-directed mutagenesis
was successful in allowing us to identify a number of amino acid
residues in the CYP3A4 active site that play a significant role in
substrate or effector binding and cooperativity (Harlow and Halpert,
1998; Domanski and Halpert, 2001; Domanski et al., 2001; Khan et
al., 2002b,c; He et al., 2003). In this study, 14 SRS residues were
selected based on our previous studies and substituted with a smaller
or larger amino acid to maximize the effect of the substitution. The
formation of each N-phenylprotoporphyrin IX regioisomer by the
mutants was always compared with wild type and is presented as a
percentage of total regioisomer formation by wild type. In the
CYP3A4 molecular model, SRS-1 is in the B⬘-C loop, which is a short
distance from the heme and located between the I-helix and ␤-sheet
6-1 (Fig. 3). Three amino acid residues, Phe-108, Ser-119, and Leu120, were selected in SRS-1. As shown in Fig. 4A, the most drastic
change was observed in the Ser-119 mutants. The substitution of
Ser-119 with Ala decreased formation of all regioisomers, and Trp
substitution increased NA, NC, and ND significantly. In contrast, in the
Ile-120 mutants, NA was enhanced by substitution with Ala and
unchanged by Trp. Little change was observed in the Phe-108 mutants.
FIG. 3. Molecular model of the CYP3A4 active site showing amino acid residues
studied.
Residues located in SRS-1, SRS-2, SRS-4, SRS-5, and SRS-6 are shown in pale
blue, green, yellow, purple and orange, respectively. Letters (A–D) represent the
corresponding pyrrole rings in the heme (colored red).
SRS-2 is in the F-helix, which is at a greater distance from the heme
than any other SRS and crosses above the I-helix (Fig. 3). Substitution
of Leu-210 with Phe enhanced the formation of NA, but Ala substitution had little effect (Fig. 4B). As reported previously, Leu-211 and
Asp-214, along with Phe-304, play a major role in cooperativity
(Harlow and Halpert, 1998; Domanski et al., 2001). Interestingly, a
double mutant, L211F/D214E, which loses cooperativity of testosterone hydroxylation, showed reduction of the three minor regioisomers
with enhancement of NA formation.
SRS-4 is in the I-helix, crossing above pyrrole ring B (Fig. 3), and
significant changes were observed in some mutants in this SRS, as
shown in Fig. 4C. In the CYP3A4 molecular model, Ile-301 is close
to Ser-119 in the B⬘-C loop. The Ile-301 Ala and Trp mutants showed
changes similar to those in the Ser-119 mutants, consistent with close
proximity between these amino acids. Phe-304 is on the opposite side
of the I-helix toward the heme and is also involved in cooperativity.
Replacement with Trp reduced the formation of the minor regioisomers, as in L211F/D214E. In contrast, the replacement of Phe-304
with Ala enhanced regioisomer formation. Ala-305 and Thr-309 are
TABLE 1
Effects of MgCl2 and substrates on N-phenylprotoporphyrin regioisomer formation from CYP3A4
CYP3A4 substrates were dissolved in methanol, and the final concentration of methanol was 1% in all conditions.
Percentage of the Total under Each Condition
Total (Percentage of Control)
Control
⫹ MgCl2 (10 mM)
⫹ Midazolam (25 ␮M)
⫹ Midazolam (250 ␮M)
⫹ 7-BFC (100 ␮M)
⫹ Mifepristone (100 ␮M)
⫹ Diazepam (250 ␮M)
a
NB
NA
NC
ND
7 (10,4)a
7 (5,8)
5 (5,6)
8 (6,11)
5 (6,5)
10 (9,11)
9 (12,7)
72 (68,76)
72 (74,69)
74 (72,76)
72 (75,70)
72 (67,76)
66 (67,66)
63 (60,66)
5 (6,5)
6 (8,5)
7 (8,7)
8 (8,7)
9 (13,5)
9 (10,8)
8 (7,8)
16 (17,15)
15 (13,17)
13 (15,12)
11 (10,13)
14 (14,14)
15 (14,15)
20 (21,18)
All values are the mean of duplicate determinations, which are shown in parentheses.
100 (106,94)
95 (92,97)
136 (128,143)
161 (152,169)
103 (100,106)
182 (190,173)
140 (144,136)
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 14, 2017
FIG. 2. HPLC profiles of N-phenylprotoporphyrin IX regioisomers (NA, NB, NC,
and ND) formed by reaction of phenyldiazene with myoglobin (A) and CYP3A4
(B).
158
YAMAGUCHI ET AL.
located just above the heme, and replacement with a larger amino acid
residue caused a significant decrease in all regioisomers. In contrast,
the substitution of Thr-309 with Ala caused an increase in NA formation, whereas the replacement of Ala-305 with Gly showed little effect
on any regioisomer.
SRS-5 is in ␤-sheets 6-1 and 1-4, which are located just opposite
the I-helix (Fig. 3). The most significant changes in regioisomer ratios
were observed in these SRS mutants (Fig. 4D). Ile-369 is away from
the heme, and replacement with a smaller or larger amino acid residue
reduced NA formation. In contrast, Ala-370 is located above the
middle of pyrrole rings A and D in the CYP3A4 molecular model.
Replacement with Phe increased NC and decreased NA. Leu-373 is
relatively near pyrrole rings C and D (Fig. 3). Substitution with Phe
decreased the three minor regioisomers and increased NA, whereas no
significant change was observed in L373A. SRS-6 is in ␤-sheets 6-2
and 4-2 below the F-helix and between SRS-4 and SRS-5 (Fig. 3).
Leu-479 was the only amino acid residue replaced in SRS-6. The
mutants showed the most significant change in NB, with Ala yielding
a decrease in NB and Phe an increase.
Topological Changes in the CYP3A4 Active Site upon Interac-
tion with CPR and b5. Previous reports indicated a conformational
change in the P450 active site following interaction with CPR or b5
(Strobel et al., 1989; Schenkman and Jansson, 2003). To obtain
experimental evidence for such structural changes, N-phenylprotoporphyrin IX regioisomers formed from reaction of phenyldiazene with
the CYP3A4 wild type were determined in the absence of redox
partners and in the presence of two equivalents of CPR, one equivalent of b5, or both. As shown in Table 2, formation of all regioisomers,
particularly the minor ones, was reduced in the presence of b5. A large
decrease in the three minor regioisomers was also observed in the
presence of CPR, but the main regioisomer was significantly enhanced. Interestingly, NC and ND were restored slightly in the presence of both CPR and b5, and the enhancement of NA formation by
CPR alone was suppressed in the presence of CPR and b5. These
observations strongly suggested structural changes in the CYP3A4
active site caused by protein-protein interactions between CYP3A4
and CPR or b5.
Structural changes in the CYP3A4 active site upon interaction with
redox partners should be influenced by substitution of key amino acid
residues. Therefore, a representative mutant in each SRS was selected
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 14, 2017
FIG. 4. Effects on N-phenylprotoporphyrin IX regioisomer formation of amino acid substitutions in SRS-1 (A), SRS-2 (B), SRS-4 (C), SRS-5 (D), and SRS-6 (E).
The regioisomer profile of SRS mutants was always determined in parallel with wild type (WT). The formation of each regioisomer in SRS mutants is represented as
a percentage of the total regioisomer formation by WT. All values are the mean of duplicate determinations.
159
TOPOLOGICAL CHANGES IN CYP3A4 ACTIVE SITE
TABLE 2
Effects of redox partners on N-phenylprotoporphyrin regioisomer formation from CYP3A4 wild type
The control contained 1 nmol of CYP3A4 wild type. Also, 1 nmol of b5, 2 nmol of CPR, or both were present under each condition. No regioisomers were formed in the reaction of
phenyldiazene with b5 or CPR.
Percentage of the Total under Each Condition
Total (Percentage of Control)
Control
⫹ b5
⫹ CPR
⫹ b5, CPR
a
NB
NA
NC
ND
10 (11, 9)a
5 (4,5)
2 (1,2)
2 (3,2)
74 (74,75)
87 (92,81)
95 (94,95)
91 (89,93)
5 (4,5)
3 (0,7)
0 (0,0)
2 (2,2)
11 (11, 11)
6 (4,7)
3 (4,3)
5 (6,4)
100 (101,99)
62 (65,59)
137 (139,136)
118 (117,118)
All values are the mean of duplicate determinations, which are shown in parentheses. Values below 0.5% are represented as zero.
Discussion
The reaction of phenyldiazene with bacterially expressed CYP3A4
formed one major and three minor regioisomers, and the profile was
changed by substitution of SRS residues and interaction with redox
partners. Some CYP3A4 substrates, including midazolam, also enhanced total regioisomer formation with little change in the ratio. The
fact that midazolam induces a prominent type I spectral change in
CYP3A4 (Khan et al., 2002c) excludes spin-state changes as the basis
for the altered regioisomer patterns caused by active site substitutions
or redox partner binding.
In the CYP3A4 model based on the structures of four bacterial
P450 enzymes (Szklarz and Halpert, 1997), SRS-4 and SRS-5 are near
the heme (Fig. 3), and the regioisomer formation changes caused by
amino acid substitutions in these SRSs were consistent with the
location of each amino acid residue in the CYP3A4 active site. As
shown in the model, Ala-305 and Thr-309 are the closest amino acid
residues to the heme iron, and their replacement with a larger amino
acid residue caused the most drastic decrease in all of the regioisomers
(Fig. 4C), presumably due to interference with initial phenyldiazene
coordination to the heme iron. In addition, Ala-370 is located above
the middle of pyrrole rings A and D, and replacement with Phe
reduced NA but enhanced NC formation (Figs. 3 and 4D). In contrast,
Leu-373 is a short distance from pyrrole ring C, and substitution with
Phe enhanced NA with reduction of NC (Figs. 3 and 4D). These
observations reveal that Phe replacement could interrupt adduct formation on the closest pyrrole ring, while enhancing that on the
opposite side.
In contrast, substitution of amino acid residues more distant from
the heme also caused large changes in regioisomer formation. Ser-119
(SRS-1) substitution with Ala, Phe, or Trp caused changes in regioisomer profiles similar to the corresponding changes at Ile-301
(SRS-4) (Fig. 4, A and C). These data support the inference from a
CYP3A4 homology model that the B⬘-C loop and I-helix interact at
these sites, as suggested for the corresponding residues in CYP2D6
and CYP2C5 (Williams et al., 2000; Kirton et al., 2002). Leu-211,
Asp-214, and Phe-304 are known to play a major role in cooperativity,
which was lost following the substitution with larger amino acid
residues (Harlow and Halpert, 1998; Domanski and Halpert, 2001;
Domanski et al., 2001). The altered kinetics of L211F/D214E strongly
suggested a structural change in the substrate oxidation site by these
substitutions at a more distal effector site. Interestingly, both L211F/
D214E and F304W formed decreased amounts of NB, NC, and ND
(Fig. 4, B and C). Because these residues are too distant from pyrrole
rings B, C, and D to interrupt phenyl group migration to the respective
nitrogen atoms (Fig. 3), the substitutions may change the location of
the other SRSs in the active site.
The regioisomer profile was also changed in the presence of redox
partners (Table 2), strongly suggesting that protein-protein interactions between CYP3A4 and CPR or b5 caused a conformational
change surrounding the heme. Previous studies revealed that basic
amino acids on the proximal surface near the heme were involved in
the binding of CYP2B4 and 1A1 with b5 or CPR (Bridges et al., 1998;
Cvrk and Strobel, 2001). Our results indicate that each redox partner
affects the binding of the other partner with CYP3A4, which may be
due to the overlapping binding sites. Previous work using the apoprotein of b5 strongly suggested that protein-protein interactions, not
enhanced second electron transfer, are responsible for enhanced
CYP3A4-catalyzed activities (Yamazaki et al., 1996, 2001). Moreover, a conformational change in the CYP3A4 active site upon interaction with redox partners, as suggested by this study, was supported
by the observation that the Km value for testosterone 6␤-hydroxylation decreased in the presence of the apo-b5 (Yamazaki et al., 1996).
Topological information on CYP3A4 was also reported using phenyldiazene with a human lymphoblast expression system (Schrag and
Wienkers, 2000). In that report, MgCl2 dramatically changed the
regioisomer ratio, unlike our study (Table 1). However, the profiles of
Schrag and Wienkers (2000) in the presence of MgCl2 were very close
to our findings in the presence of CPR and b5 (Table 2), indicating that
both CYP3A4 preparations have a similar conformation of the active
site under this condition.
Changes in the regioisomer profile upon interaction with redox
partners were different among SRS mutants (Fig. 5). Substitution of
Ser-119 with Trp significantly enhanced all regioisomers (Fig. 4), and
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 14, 2017
for study of the effects of redox partners on regioisomer formation
(Fig. 5). S119F, which showed an increase in all regioisomers, was
selected from SRS-1 (Fig. 5A). As with wild type, all regioisomers
were reduced in the presence of b5. Surprisingly, the greatest decrease,
including that of NA, was observed in the presence of CPR alone. All
regioisomers were partially restored in the presence of both proteins,
but the total regioisomer formation was only about one-half that under
control conditions. In the case of the SRS-2 mutant L211F/D214E,
very little of the minor regioisomers was formed under any conditions
(Fig. 5B), and the effect of redox partners was less than with the wild
type. In the case of T309A, an SRS-4 mutant, no minor regioisomer
was detected in the presence of b5, CPR, or both (Fig. 5C). As shown
in Fig. 5D, A370F, an SRS-5 mutant, formed almost equal amounts of
NA, NC, and ND. NC was strongly reduced in the presence of redox
partners, especially CPR, and NA was enhanced in the presence of b5
or both proteins but decreased by CPR alone. In contrast, ND formation changed little under any conditions. L479F, an SRS-6 mutant,
showed a similar profile of regioisomers to wild type in the presence
of redox partners (Fig. 5E). The only difference from the wild type
was a slightly increased restoration of minor regioisomers in the
presence of both CRP and b5. Overall, the studies with the mutants
confirmed that topological changes occur upon interaction of
CYP3A4 with redox partners and suggested regions of the active site
that are most sensitive to such changes.
160
YAMAGUCHI ET AL.
The formation of each regioisomer is represented as a percentage of the total regioisomer formation in the absence of redox partners (control). All values are the mean
of duplicate determinations.
the increases were suppressed in the presence of CPR and b5 (Fig.
5A). The drastic change observed in this SRS-1 mutant strongly
indicated a large conformational change in the B⬘-C loop upon redox
partner binding, which is supported by a previous report that some
basic amino acid residues in the C- and C⬘-helix are involved in
binding between P450 and CPR or b5 (Bridges et al., 1998). Studies
of A370F revealed that redox partners caused the most drastic change
in adduct formation with pyrrole ring C, a space surrounded by the
B⬘-C loop, the I-helix, and ␤-sheets 6-1 and 1-4 in the CYP3A4 model
(Figs. 3 and 5D). Another interesting observation in A370F was the
very small change of ND caused by redox partners (Fig. 5D), in
contrast to the decrease in ND in the case of the other mutants and wild
type (Table 2; Fig. 5). The difference revealed that Phe substitution
for Ala-370 might impede a conformational change that covers the
space above pyrrole ring D upon redox partner binding. In some SRS
mutants, effects on regioisomer profile were quite different between
CPR and b5. Bridges et al. (1998) reported that two additional basic
amino acids outside the C- or C⬘-helix were involved in the binding of
CYP2B4 to CPR but not b5. These residues are located in a conserved
region between the meander and L-helix (Bridges et al., 1998). Since
this conserved region also include the Cys coordinating the heme and
two amino acids forming hydrogen bonds with the propionate side
chains of the two pyrrole rings, as previously reported in CYP2C5
crystal structure (Williams et al., 2000), CPR but not b5 binding to
CYP3A4 might affect the accessibility of the heme in the active site.
In conclusion, the data obtained using CYP3A4 SRS mutants
validated use of the N-phenylprotoporphyrin IX regioisomer profile to
assess topological changes in the CYP3A4 active site. Subsequent
experiments in the presence of redox partners supplied crucial evidence that conformational changes in the CYP3A4 active site occur
upon interaction with CPR and b5. The present study implies that
interaction with redox partners can also change the regioselectivity or
stereoselectivity of substrate oxidation. In fact, substrate inhibition of
CYP3A4 triazolam 1⬘-hydroxylation but not 4-hydroxylation has been
reported, and the response was altered in the presence of b5, leading
to a change in regioselectivity (Schrag and Wienkers, 2001). Studies
of representative SRS mutants strongly suggested that redox partner
binding to CYP3A4 alters the relative location of the B⬘-C loop
toward the heme, which is consistent with the interaction site of
CYP2B4 with redox partners previously reported (Bridges et al.,
1998). Our results also suggest that the conformation of CYP3A4 is
different in the presence of both redox partners than in the presence of
either one alone (Table 2; Fig. 5). Thus, stimulation of CYP3A4-
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 14, 2017
FIG. 5. Effects on N-phenylprotoporphyrin IX regioisomer formation of interaction with redox partners using S119W (A), L211F/D214E (B), T309A (C), A370F (D),
and L479F (E).
TOPOLOGICAL CHANGES IN CYP3A4 ACTIVE SITE
catalyzed activities by b5 might also involve improved electron transfer following a change in orientation of CPR toward CYP3A4.
References
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.
Mackman R, Guo Z, Guengerich FP, and Ortiz de Montellano PR (1996a) Active site topology
of human cytochrome P450 2E1. Chem Res Toxicol 9:223–226.
Mackman R, Tschirret-Guth RA, Smith G, Hayhurst GP, Ellis SW, Lennard MS, Tucker GT,
Wolf CR, and Ortiz de Montellano PR (1996b) Active-site topologies of human CYP2D6 and
its aspartate-301 3 glutamate, asparagine and glycine mutants. Arch Biochem Biophys
331:134 –140.
Nebert DW and Russell DW (2002) Clinical importance of the cytochromes P450. Lancet
360:1155–1162.
Ortiz de Montellano PR (1995) Arylhydrazines as probes of hemoprotein structure and function.
Biochimie 77:581–593.
Schenkman JB and Jansson I (2003) The many roles of cytochrome b5. Pharmacol Ther
97:139 –152.
Schrag ML and Wienkers LC (2000) Topological alteration of the CYP3A4 active site by the
divalent cation Mg2⫹. Drug Metab Dispos 28:1198 –1201.
Schrag ML and Wienkers LC (2001) Triazolam substrate inhibition: evidence of competition for
heme-bound reactive oxygen within the CYP3A4 active site. Drug Metab Dispos 29:70 –75.
Shou M, Grogan J, Mancewicz JA, Krausz KW, Gonzalez FJ, Gelboin HV, and Korzekwa KR
(1994) Activation of CYP3A4: evidence for the simultaneous binding of two substrates in a
cytochrome P450 active site. Biochemistry 33:6450 – 6455.
Strobel HW, Nadler SG, and Nelson DR (1989) Cytochrome P-450: cytochrome P-450 reductase
interactions. Drug Metab Rev 20:519 –533.
Swanson BA and Ortiz de Montellano PR (1991) Structure determination and absolute stereochemistry of the four N-phenylprotoporphyrin IX regioisomers. J Am Chem Soc 113:8146 –
8153.
Szklarz GD and Halpert JR (1997) Molecular modeling of cytochrome P450 3A4. J ComputAided Mol Des 11:265–272.
Tuck SF, Peterson JA, and Ortiz de Montellano PR (1992) Active site topologies of bacterial
cytochromes P450101 (P450cam), P450108 (P450terp) and P450102 (P450BM-3). In situ
rearrangement of their phenyl-iron complexes. J Biol Chem 267:5614 –5620.
Williams PA, Cosme J, Sridhar V, Johnson EF, and McRee DE (2000) Mammalian microsomal
cytochrome P450 monooxygenase: structural adaptations for membrane binding and functional diversity. Mol Cell 5:121–131.
Yamazaki H, Johnson WW, Ueng YF, Shimada T, and Guengerich FP (1996) Lack of electron
transfer from cytochrome b5 in stimulation of catalytic activities of cytochrome P450 3A4.
Characterization of a reconstituted cytochrome P450 3A4/NADPH-cytochrome P450 reductase system and studies with apo-cytochrome b5. J Biol Chem 271:27438 –27444.
Yamazaki H, Shimada T, Martin MV, and Guengerich FP (2001) Stimulation of cytochrome
P450 reactions by apo-cytochrome b5: evidence against transfer of heme from cytochrome
P450 3A4 to apo-cytochrome b5 or heme oxygenase. J Biol Chem 276:30885–30891.
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 14, 2017
Bridges A, Gruenke L, Chang YT, Vakser IA, Loew G, and Waskell L (1998) Identification of
the binding site on cytochrome P450 2B4 for cytochrome b5 and cytochrome P450 reductase.
J Biol Chem 273:17036 –17049.
Cvrk T and Strobel HW (2001) Role of LYS271 and LYS279 residues in the interaction of
cytochrome P4501A1 with NADPH-cytochrome P450 reductase. Arch Biochem Biophys
385:290 –300.
Dierks EA, Zhang Z, Johnson EF, and de Montellano PR (1998) The catalytic site of cytochrome
P4504A11 (CYP4A11) and its L131F mutant. J Biol Chem 273:23055–23061.
Domanski TL and Halpert JR (2001) Analysis of mammalian cytochrome P450 structure and
function by site-directed mutagenesis. Curr Drug Metab 2:117–137.
Domanski TL, He YA, Khan KK, Roussel F, Wang Q, and Halpert JR (2001) Phenylalanine and
tryptophan scanning mutagenesis of CYP3A4 substrate recognition site residues and effect on
substrate oxidation and cooperativity. Biochemistry 40:10150 –10160.
Guengerich FP (1999) Cytochrome P-450 3A4: regulation and role in drug metabolism. Annu
Rev Pharmacol Toxicol 39:1–17.
Harlow GR and Halpert JR (1997) Alanine-scanning mutagenesis of a putative substrate
recognition site in human cytochrome P450 3A4. Role of residues 210 and 211 in flavonoid
activation and substrate specificity. J Biol Chem 272:5396 –5402.
Harlow GR and 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.
He YA, Roussel F, and Halpert JR (2003) Analysis of homotropic and heterotropic cooperativity
of diazepam oxidation by CYP3A4 using site-directed mutagenesis and kinetic modeling. Arch
Biochem Biophys 409:92–101.
Khan KK, He YA, He YQ, and Halpert JR (2002a) Site-directed mutagenesis of cytochrome
P450eryF: implications for substrate oxidation, cooperativity and topology of the active site.
Chem Res Toxicol 15:843– 853.
Khan KK, He YQ, Correia MA, and Halpert JR (2002b) Differential oxidation of mifepristone
by cytochromes P450 3A4 and 3A5: selective inactivation of P450 3A4. Drug Metab Dispos
30:985–990.
Khan KK, He YQ, Domanski TL, and Halpert JR (2002c) 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 61:495–506.
Kirton SB, Kemp CA, Tomkinson NP, St-Gallay S, and Sutcliffe MJ (2002) Impact of incorporating the 2C5 crystal structure into comparative models of cytochrome P450 2D6. Proteins
49:216 –231.
Korzekwa KR, Krishnamachary N, Shou M, Ogai A, Parise RA, Rettie AE, Gonzalez FJ, and
161