Inhibition of cytochrome P450 3A4 by a pyrimidineimidazole: Evidence for complex heme interactions

Article abstract of DOI:10.1021/tx060198m

PH-302 inhibits the inducible form of nitric oxide synthase (iNOS) by coordinating with the heme of the monomeric form and preventing formation of the active dimer. Inherent with the mechanism of pharmacology for this compound was the inhibition of cytochrome P450 3A4 (P450 3A4), observed from early ADME screening. Further investigation showed that PH-302 inhibited P450 3A4 competitively with a K_i of ～2.0 μM against both midazolam and testosterone hydroxylation in human liver microsomes. As expected, spectral binding analysis demonstrated that inhibition was a result of type II coordination to the P450 heme with the imidazole moiety of PH-302, although only 72% of the maximal absorbance difference was achievable with PH-302 compared to that of the smaller ligand imidazole. Time-dependent inhibition of P450 3A4 by PH-302 was also observed because of metabolite-inhibitory (MI) complex formation via metabolism of the methylenedioxyphenyl group. The profile for time-dependent inhibition in recombinant P450 3A4 was biphasic, and was kinetically characterized by a k_inact of 0.08 min^-1 and a K_i of 1.2 μM for the first phase (0-1.5 min) and a k _inact of 0.06 min^-1 and a K_i of 23.8 μM for the second phase (1.5-10 min). Spectral characterization of the PH-302 MI complex demonstrated that formation began to plateau within 3 min, consistent with the kinetic profile of inactivation by PH-302. Meanwhile, spectral evidence for the imidazole-derived type II difference spectrum of PH-302 was captured simultaneously with the formation of the MI complex. The presence of simultaneously operable type II coordination and rapidly saturable MI complex formation with heme by PH-302 indicates the presence of complex heme interactions with this unique molecule. Information from these mechanistic studies adds to our understanding of the nature of P450 3A4 inhibition by PH-302 and provides a potentially useful tool compound for future studies investigating binding interactions in this important drug-metabolizing enzyme.

Full text of DOI:10.1021/tx060198m

1650
Chem. Res. Toxicol. 2006, 19, 1650-1659
Inhibition of Cytochrome P450 3A4 by a Pyrimidineimidazole:
Evidence for Complex Heme Interactions
J. Matthew Hutzler,* Roger J. Melton, Jeanne M. Rumsey, Mark E. Schnute,†
Charles W. Locuson,‡ and Larry C. Wienkers§
Pharmacokinetics, Dynamics and Metabolism (PDM), Department of Medicinal Chemistry, Pfizer Global
Research and DeVelopment, 700 Chesterfield Parkway West T3A, Chesterfield, Missouri 63017, and Department
of Experimental and Clinical Pharmacology, UniVersity of Minnesota, Minneapolis, Minnesota 55455
ReceiVed August 18, 2006
PH-302 inhibits the inducible form of nitric oxide synthase (iNOS) by coordinating with the heme of
the monomeric form and preventing formation of the active dimer. Inherent with the mechanism of
pharmacology for this compound was the inhibition of cytochrome P450 3A4 (P450 3A4), observed
from early ADME screening. Further investigation showed that PH-302 inhibited P450 3A4 competitively
with a Ki of ∼2.0 µM against both midazolam and testosterone hydroxylation in human liver microsomes.
As expected, spectral binding analysis demonstrated that inhibition was a result of type II coordination
to the P450 heme with the imidazole moiety of PH-302, although only 72% of the maximal absorbance
difference was achievable with PH-302 compared to that of the smaller ligand imidazole. Time-dependent
inhibition of P450 3A4 by PH-302 was also observed because of metabolite-inhibitory (MI) complex
formation via metabolism of the methylenedioxyphenyl group. The profile for time-dependent inhibition
in recombinant P450 3A4 was biphasic, and was kinetically characterized by a kinact of 0.08 min-1 and
a Ki of 1.2 µM for the first phase (0-1.5 min) and a kinact of 0.06 min-1 and a Ki of 23.8 µM for the
second phase (1.5-10 min). Spectral characterization of the PH-302 MI complex demonstrated that
formation began to plateau within 3 min, consistent with the kinetic profile of inactivation by PH-302.
Meanwhile, spectral evidence for the imidazole-derived type II difference spectrum of PH-302 was captured
simultaneously with the formation of the MI complex. The presence of simultaneously operable type II
coordination and rapidly saturable MI complex formation with heme by PH-302 indicates the presence
of complex heme interactions with this unique molecule. Information from these mechanistic studies
adds to our understanding of the nature of P450 3A4 inhibition by PH-302 and provides a potentially
useful tool compound for future studies investigating binding interactions in this important drug-
metabolizing enzyme.
Introduction
PH-302 is a pyrimidineimidazole molecule (Figure 1) that
was investigated in preclinical studies for potency and selectivity
for inhibiting the inducible form of nitric oxide synthase
(iNOS1). This class of compounds have been shown to possess
a novel mechanism, inhibiting the iNOS enzyme by direct
interaction of the imidazole with the heme, and thus preventing
the dimerization of iNOS monomers to the active dimer (1).
Similar to azole antifungals, such as ketoconazole and itracona-
zole, which target the sterol 14R-demethylase P450 (CYP51)
by heme coordination, it was determined upon early ADME
screening that PH-302 was also a potent inhibitor of cytochrome
P450 3A4 (P450 3A4). P450 3A4 represents the most abundant
and important P450 in the human liver and intestine and is quite
Figure 1. Structure of PH-302, containing both an imidazole and a
methylenedioxyphenyl (MDP) functional group in its chemical structure.
unique from other major P450 enzymes in that it metabolizes
molecules with a broad range of shapes and molecular weights
(2, 3). As a result of this promiscuous substrate selectivity, P450
3A4 is responsible for metabolizing over 50% of marketed drugs
(4). Thus, inhibition of this drug-metabolizing enzyme has led
to a high risk of drug-drug interactions with potentially serious
clinical implications (5-7).
In addition to containing the pharmacologically required
imidazole ring, PH-302 also contains a methylenedioxyphenyl
(MDP) moiety (Figure 1). Both of these functionalities are
known to cause the potent inhibition of P450 enzymes, although
via distinct mechanisms. Imidazole-containing compounds are
known to coordinate with the heme of P450, inducing a type II
spectral shift (8, 9). Well-known azole antifungal compounds
* Corresponding author. Tel: 314-274-0261. Fax: 314-274-4426. E-
mail: J.Matt.Hutzler@Pfizer.com.
† Pfizer Global Research and Development.
‡ University of Minnesota.
§ Current address: Amgen, Inc., Pharmacokinetics & Drug Metabolism,
AW2-D / 2381, 1201 Amgen Court West, Seattle, WA 98119-3105.
1 Abbreviations: iNOS, inducible nitric oxide synthase; P450 3A4,
cytochrome P450 3A4; ADME, absorption, distribution, metabolism,
excretion; SSRI, selective serotonin reuptake inhibitor; MIC, metabolite-
inhibitory complex; MDP, methylenedioxyphenyl; 7-BFC, 7-benzyloxy-4-
(trifluoromethyl)-coumarin; 7-MFC, 7-methoxy-4-(trifluoromethyl)-cou-
marin; AMMC, 3-[2-(N,N-diethyl-N-methylammonium)ethyl]-7-methoxy-
4-methylcoumarin.
10.1021/tx060198m CCC: $33.50 © 2006 American Chemical Society
Published on Web 11/24/2006

Complex P450 3A4 Heme Interactions
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1651
hydrochloric acid (40 mL). The mixture was heated to 50 °C,
allowed to cool to room temperature, filtered, and the resulting
filtrate was concentrated to afford 485 mg (91%) of the product as
a white solid (1H NMR characterization in Supporting Information).
[3H]-PH-302 was synthesized with a uniform tritium-radiolabel by
an outside vendor and was received as a total of 300 mCi in 0.5
mL of MeOH and then purified internally at Pfizer (St. Louis, MO)
to 99.7% radiochemical purity (specific activity: 34.3 µCi/nmol).
[3H]-PH-302 was separated using a YMC ODS-AQ S-3 120 A 5
µm 4.6 × 150 mm column and a long gradient of 90% mobile
phase A (0.1% TFA)/10% B (acetonitrile) initially flowing at 1.5
mL/min, followed by slow gradient to 100% B over 15 min and
holding until 19 min, then returning to initial conditions, and
detected by UV (254 nm-retention time 3.06 min) and radiomatic
(retention time 3.40 min) methods.
P450 Inhibition Screening. PH-302 (3 µM) was screened against
P450s 3A4, 2C9, 2D6, 1A2, and 2C19 according to methods
previously reported (24) using the fluorescent probe substrates
7-BFC, 7-MFC, AMMC, and Vivid Blue (CYP1A2 and 2C19),
respectively, with some minor modifications. Briefly, fluorogenic
assays were conducted in 96-well clear-bottom Costar white
polystyrene plates, and fluorescence readings were recorded on a
TECAN (Durham, NC) Spectra Fluor plate reader. In a typical
incubation, the test compound (PH-302) was diluted two times from
a 10 mM DMSO stock solution into 40% methanol and then into
15% methanol for a final organic concentration of <1% in a 100
µL incubation volume. After spotting the plate with 5 µL of test
compound, 85 µL of an incubation mixture containing recombinant
P450 Baculosomes, a potassium phosphate buffer at pH 7.4, and a
fluorescent probe substrate were added to the plate. The plate was
then allowed to preincubate at 37 °C for 5 min in the plate reader,
and 10 µL of the NADPH generating system (10 mM MgCl2, 0.5
mM NADP+, 6.2 mM DL-isocitric acid, 0.5 units/mL isocitric
dehydrogenase) was then added to initiate the reactions. For P450
2D6 incubations, the NADP+ concentration was reduced to 0.03
mM because of the fluorescence interference that resulted from the
slow metabolism of AMMC. Plates were then placed back into the
plate reader, and fluorescence readings were recorded every minute
for 20-30 min (according to the linear formation of the fluorescent
metabolite). The fluorescent plate reader wavelength settings were
as follows: 7-BFC (P450 3A4) and MFC (P450 2C9), excitation
at 405 nm and emission at 535 nm; AMMC (P450 2D6), excitation
at 390 nm and emission at 460 nm; and Vivid Blue (P450 1A2
and 2C19), excitation at 405 and emission at 460 nm. Percent
inhibition was calculated comparing slopes of fluorescent metabolite
formation in the presence of PH-302 relative to a solvent control
well.
that potently inhibit P450 3A4 via this mechanism include
ketoconazole (IC50 ) 0.006 µM), itraconazole (IC50 0.15 µM),
and clotrimazole (IC50 ) 0.003 µM) (10). In addition, meth-
ylenedioxyphenyl (MDP) functional groups have been reported
to be metabolized by P450 at the methylenic carbon, leading to
ring-opening and the formation of a catechol metabolite (11,
12). The most often cited intermediate in this metabolic pathway
is a carbene, containing a free pair of electrons that can form a
pseudo-irreversible complex with the heme of P450, a species
known as a metabolite-inhibitory complex (MIC) (13, 14). The
formation of this complex is characterized spectrophotometri-
cally by an increase in absorbance at 455 nm and typically
results in time-dependent inhibition of P450 enzymatic activity.
This has been observed with many MDP-containing compounds,
including isosafrole (15) and the selective serotonin re-uptake
inhibitor (SSRI) paroxetine (16), as well as extracts from the
herbal medication goldenseal (17).
Cytochrome P450 3A4 is characterized by an especially
voluminous active site pocket, reported to be >1300 Å3 (18,
19), with recent crystal structure data showing that upon the
binding of larger ligands, the active site volume may approach
2000 Å3 (20). This often complicates the clear interpretation
and prediction of substrate/inhibitor binding characteristics with
this enzyme. The majority of evidence has suggested that the
observed complex binding properties of P450 3A4 ligands derive
from multiple binding regions within the spacious active site
(21, 22). Current observations of P450 3A4 inhibition by a
unique molecule containing two structural moieties known to
coordinate with the heme of P450 presented the opportunity to
probe the binding interactions directly using spectrophotometric
techniques. To this end, studies were carried out in an effort to
investigate these potential heme interactions and determine
whether both mechanisms were operable. Thus, an understand-
ing of the binding characteristics and mechanisms by which
PH-302 inhibits P450 3A4 will be gained.
Experimental Procedures
Materials. Potassium phosphate buffer, midazolam hydrochlo-
ride, 1′-hydroxy midazolam, testosterone, 6â-hydroxy testosterone,
NADPH, NADP+, isocitric acid, isocitric dehydrogenase, magne-
sium chloride, sodium dithionite, tolbutamide, fluoxetine, piperonyl
butoxide, and ketoconazole were purchased from Sigma-Aldrich
(St. Louis, MO). Purified cytochrome P450 3A4 (P450 3A4), P450
3A4, 2C9, 2D6, 1A2, and 2C19 Baculosomes and Vivid Blue P450
fluorescent probes were purchased from Invitrogen (Carlsbad, CA),
whereas P450 3A4 Supersomes (with coexpressed cytochrome b5)
and pooled human liver microsomes (0.33 nmol/mg of protein),
along with the fluorescent P450 probes 7-benzyloxy-4-(trifluorom-
ethyl)-coumarin (7-BFC), 7-methoxy-4-(trifluoromethyl)-coumarin
(MFC), and 3-[2-(N,N-diethyl-N-methylammonium)ethyl]-7-meth-
oxy-4-methylcoumarin (AMMC), were purchased from BD Bio-
sciences (San Jose, CA). All other chemicals were obtained from
commercial sources and were of the highest purity available.
Inhibition Constant (Ki) Determination. Incubations were
performed in triplicate in a 96-well plate format, with a final
incubation volume of 150 µL. PH-302 was incubated at concentra-
tions ranging from 0 to 5 µM (1% total organic v/v), whereas
midazolam and testosterone were incubated at three different
concentrations at their apparent Km values, shown in parentheses,
determined from previous experiments (1, 5, and 10 µM (Km,app
)
2.5 µM) and 10, 30, and 50 µM (Km,app ) 30 µM), respectively)
with human liver microsomes (0.2 mg/mL) in a 100 mM potassium
phosphate buffer at pH 7.4. Incubations were initiated by the
addition of 1 mM NADPH, allowed to proceed for 5 min in a 37
°C incubator, and quenched with 100 µL of acetonitrile containing
tolbutamide (1 µM) as the internal standard. The plate was then
centrifuged, and 1′-hydroxy midazolam and 6â-hydroxy testosterone
formation were quantitated by LC/MS-MS analysis. Inhibition
constants (Ki) were estimated from a visual inspection of Dixon
Plots, generated using GraphPad PRISM (San Diego, CA) graphing
software.
Synthesis. PH-302 was synthesized according to methods
reported in McMillan et al. (23) (Patent # US6,432,947) and was
provided to us by David Davey at Berlex Biosciences (Richmond,
CA). The catechol of PH-302 (N-(3-((3,4-dihydroxybenzyl)(methyl)-
amino) propyl)-N-(2-(1H-imidazol-1-yl)-6-methylpyrimidin-4-yl)-
glycinamide hydrochloride) was synthesized internally by Pfizer
(St. Louis, MO). Briefly, PH-302 (437 mg, 1.0 mmol) was dissolved
in CH2Cl2 (40 mL) and cooled to -70 °C. A solution of BCl3 (1.0
M in xylenes, 4.0 mL, 4.0 mmol) was slowly added. The mixture
was allowed to slowly warm to room temperature over 8 h and
was then stirred an additional 14 h. The reaction was quenched
with water (20 mL) and methanol (20 mL). The reaction mixture
was concentrated, and the residue was suspended in 1 M aqueous
Spectrophotometric Studies. Spectral binding titration studies
were carried out with purified P450 3A4 (1 µM) in a 100 mM
potassium phosphate buffer at pH 7.4. Incubation mixtures were
evenly divided into 1.5 mL 10 mm black-walled cuvettes, and
experiments were performed at room temperature by titrating in 1
µL aliquots of PH-302 (DMSO solution) into a sample cuvette

1652 Chem. Res. Toxicol., Vol. 19, No. 12, 2006
Hutzler et al.
(0.02-20 µM), with an equal aliquot of DMSO in the reference
cuvette (DMSO <1% v/v). After each titration, cuvettes were mixed
and allowed to sit for 1 min to allow the samples to reach
equilibrium. Difference spectra were recorded between 350 and 500
nm using a Hitachi (Danbury, CT) U3300 dual-beam spectropho-
tometer, and data were analyzed using UV Solutions 1.2 software.
The difference in the absorbance between the peak (∼430 nm) and
trough (∼405 nm) for type II spectra were then plotted versus titrant
concentration and fit to a hyperbolic curve function (eq 1) for the
estimation of Ks values using Graph Pad PRISM 4.0 software (San
Diego, CA).
DL-isocitric acid, and 0.5 units/mL isocitric dehydrogenase), 10 µL
aliquots were taken from the primary incubation at 0, 0.5, 1.5, 3,
5, and 10 min and placed into 190 µL of a secondary incubation
(200 µL final volume), containing 10 µM midazolam as the reporter
substrate, and NADPH regenerating system (similar to concentra-
tions in the primary assay) in a 100 mM potassium phosphate buffer
at pH 7.4. The 20-fold dilution of the primary incubation mixture
minimized any contribution of competitive inhibition in the
secondary midazolam activity assay. The secondary midazolam
assay was allowed to proceed for 5 min and was then quenched
with ice-cold acetonitrile containing tolbutamide as the internal
standard. The 96-well plates were centrifuged, and 1′-hydroxy
midazolam formation was analyzed by LC/MS-MS. Observed rates
of inactivation (kobs) were calculated by linear regression of the
log % remaining activity as a function of preincubation time and
subsequently re-plotted in a double-reciprocal Kitz-Wilson (25)
plot (1/inactivation rate (Kobs) versus 1/[PH-302]) to determine
inactivation rate constants Ki and kinact using GraphPad PRISM (San
Diego, CA) software. Concentrations of PH-302 were monitored
from parallel incubations equal to the primary incubation conditions
by LC/MS-MS to test for inactivator depletion.
MI Complex. P450 3A4 Supersomes (with cytochrome b5) were
used to characterize the potential MI complex (MIC) formation
associated with the metabolism of PH-302. MI complex formation
was measured at various times (0, 0.5, 1.5, 3, 5, 7, and 10 min)
with incubations containing PH-302 (10 µM) in a microsomal buffer
(100 mM potassium phosphate buffer and 3 mM MgCl2 at pH 7.4).
Two milliliters of incubation mixture was prepared and split
between two cuvettes (1 mL, 1 cm path length). For each study,
the sample cuvette contained 0.2 pmol/µL of P450 3A4, the
potential inactivator, and the NADPH regenerating system (3.3 mM
MgCl2, 1 mM NADP+, 12.8 mM DL-isocitric acid, and 0.5 units/
mL isocitric dehydrogenase), whereas the reference cuvette con-
tained the same matrix with only the inactivator vehicle (DMSO
1% v/v). In a separate experiment, the enzyme was reduced with
sodium dithionite prior to the addition of PH-302. MI complex
formation was monitored as described above with a Hitachi
(Danbury, CT) U-3300 dual-beam UV/vis spectrophotometer by
scanning at the indicated times from 390 to 500 nm to monitor the
simultaneous type II spectrum and the formation of an absorbance
maximum at ∼455 nm. Both fluoxetine and piperonyl butoxide
(100 µM) were used as positive controls for MI complex formation
by P450 3A4. Following a 15 min incubation under similar
incubation conditions, ketoconazole (10 µM) was added to the
sample cuvette to compare the magnitude of ketoconazole-derived
type II difference absorbance with the fluoxetine-derived MI
complex absorbance at 455 nm. All MI complex formation
experiments were maintained at 37 °C and reactions were initiated
by the addition of the inactivator after a 3 min preincubation period.
Data were analyzed using UV Solutions 1.2 software.
LC/MS-MS Analysis. Samples for measuring 1′-hydroxy mi-
dazolam (m/z 342) and 6â-hydroxy testosterone (m/z 305) were
analyzed on a Sciex 3000 triple-quadrupole instrument using
multiple reaction monitoring (MRM). The mass spectrometer was
equipped with an electrospray ionization (ESI) interface in positive
ion mode and was connected to an Agilent 1100 HPLC pump and
a Leap Technologies CTC PAL (Carrboro, NC) auto-sampler.
Analytes were separated using a Zorbax 5 µm Eclipse XDB C18
2.1 × 50 mm column with a gradient elution profile. Briefly, the
mobile phase was flowing at 0.4 mL/min, and the gradient was
initiated and held for the first 0.5 min at 95%A/5%B (A: 0.1%
formic acid in H2O; B: 0.1% formic acid in acetonitrile), and was
then ramped linearly to 5%A/95%B over the next 0.25 min and
held for 3 min. The profile was then immediately returned to initial
conditions and allowed to re-equilibrate for 2 min. The MRM
transitions were m/z 342/203 for 1′-hydroxy midazolam, m/z 305/
269 for 6â-hydroxy testosterone, and m/z 271/155 for the internal
standard (tolbutamide), with the source temperature set at 400 °C.
An analysis of PH-302 and its catechol metabolite was performed
using an Applied Biosystems Sciex API-4000 connected in line
with a Shimadzu LC10AD pump and a Leap Technologies
∆A ) BmaxS/Ks + S
(1)
The binding of PH-302 to P450 3A4 was also assessed in separate
spectral studies using an Aminco DW-2000 dual beam spectro-
photometer (OLIS Inc., Bogart, GA) operating in split beam mode.
An equal amount of purified P450 3A4 (0.5 µM) was added to
both a sample and reference micro-cuvette containing 100 mM
potassium phosphate at pH 7.4 and 20% glycerol (v/v). A baseline
spectrum was then recorded. PH-302, dissolved in DMSO and
diluted to desired concentrations in buffer, was then titrated into
the sample cuvette parallel (with equal volumes of buffer titrated
into the reference cuvette) until a change in absorbance could no
longer be detected (35 µM). A characteristic peak (430 nm) and
trough (410 nm), indicative of the coordination of PH-302 to the
heme iron, were observed (i.e., type-II interaction). Equilibration
was achieved <1 min after adding each aliquot of PH-302. The
observed difference between peak and trough absorbances was then
compared to that observed with the control imidazole, selected
because of its similarity to the nitrogen-bonded imidazole in PH-
302. Stock solutions of imidazole were prepared in water and their
pH adjusted to 7.4 using HCl. The maximum absorbance change
of P450 3A4 induced by imidazole was found to occur at a
concentration of g25 mM, and at this point, the majority of the
P450 3A4 enzyme was assumed to be in the low spin form (i.e.,
type II).
Metabolism and Reaction Phenotyping. [3H]-PH-302 was
preincubated at a concentration of 10 µM (3 µCi) with human liver
microsomes (1 mg/mL) in a 100 mM potassium phosphate buffer
at pH 7.4 and 3 mM MgCl2 for 5 min at 37 °C, followed by the
addition of 1 mM NADPH to initiate the reaction (200 µL total
volume). This incubation was allowed to proceed for 30 min, and
100 µL of cold acetonitrile was added to precipitate protein. Samples
were centrifuged for 10 min at 10,000 rpm, and the supernatant
was analyzed by LC-radiomatic detection methods. Separation of
analytes was performed using an HP Series 1100 LC system, a
Zorbax SB-C8 5 µm 4.6 × 250 mm column, and a mobile phase (1
mL/min) consisting of A (0.1% formic acid in water) and B (0.1%
formic acid in acetonitrile) run initially for 3 min at 100%A, then
to 50%B over the next 10 min, and then returned to initial conditions
for re-equilibration. The Ultima Gold Scintillation cocktail was
introduced postcolumn at a rate of 3 mL/min, and the quantitation
of metabolites was performed using a FLO-ONE Series A500
radiomatic detector (Perkin-Elmer Life Sciences). Peak areas were
integrated with the Windows-based Radio-HPLC Workstation
software. Incubation conditions for reaction phenotyping studies
were similar to those described above, but non-labeled PH-302 was
incubated (10 µM) with human liver microsomes (1 mg/mL) in
the absence and presence of 1 µM ketoconazole, a specific P450
3A4 inhibitor. Control incubations did not include NADPH.
Samples from these studies were quenched by the addition of ice-
cold acetonitrile, centrifuged at 3000 rpm for 10 min, and then
analyzed by LC/MS-MS.
Time-Dependent Inhibition Experiments. Primary incubations
were carried out in 96-well plate format, with mixtures containing
multiple concentrations of PH-302 (0-20 µM) and recombinant
P450 3A4 Supersomes (with cytochrome b5) (0.2 pmol/µL) in a
100 mM potassium phosphate buffer pH 7.4. After initiation of
the primary inactivation assay by the addition of the NADPH
regenerating system (3.3 mM MgCl2, 1 mM NADP+, 12.8 mM

Complex P450 3A4 Heme Interactions
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1653
Figure 3. (A) Spectral binding titration data with PH-302 and purified
P450 3A4, indicating a type II binding interaction due to imidazole-
heme coordination. A spectral binding affinity constant (Ks) of 5.5 µM
was calculated using nonlinear regression analysis (eq 1). (B) Com-
parison of maximal peak-to-trough absorbance difference for imidazole,
which saturated the type II signal at 25 mM, and PH-302, which
saturated at 35 µM. Titration of PH-302 to its saturation point could
Figure 2. Dixon plot analysis of PH-302 inhibition of P450 3A4-
mediated (A) midazolam 1′-hydroxylation and (B) testosterone 6â-
hydroxylation. Each data point is the mean of triplicate determinations.
Similar apparent Ki values (∼2 µM) were estimated with both P450
3A4 probe substrates.
only produce ∼72% of the absorbance difference maximum (∆Amax
)
achievable with smaller heme-coordinating ligand imidazole.
(Carrboro, NC) CTC PAL auto-sampler. A Thermo Aquasil 5 µm
C18 DASH 20 × 2.1 mm was used with mobile phase A (0.1%
heptafluorobutyric acid in water) and mobile phase B (0.1%
heptafluorobutyric acid in acetonitrile) delivered at 0.5 mL/min
using fast gradient elution starting at 0%B for 0.05 min, ramped to
100%B over 0.55 min, and then held for 0.9 min. The system was
returned to initial conditions over 0.1 min and allowed to equilibrate
for 2 min. Mass spectral analysis was performed using multiple
reaction monitoring (MRM) for PH-302 (transition m/z 438.0/135.1)
and the catechol (transition m/z 426.0/304.2) utilizing a turbo ion
spray source in positive ionization mode. The spray voltage was
set to 5 kV, source temperature set to 425 °C, collision energy
(CE) at 23 V, and collision exit potential (CXP) at 24 V. All data
were analyzed using PE Sciex Analyst 1.4.1 software.
the x-axis, suggesting a competitive inhibition mechanism for
PH-302 (26).
Spectral Binding. The titration of PH-302 with purified P450
3A4 resulted in a type II difference spectrum, characterized by
an asymmetrical and broad absorbance trough that extended
from ∼390-405 nm and a sharper peak at ∼430 nm, consistent
with the imidazole moiety of PH-302 coordinating to the heme
iron of P450 3A4 (Figure 3A). Such binding interactions
typically result in the potent inhibition of the metabolism of
substrates for that particular P450. When plotting the absorbance
difference (Abs430 nm - Abs405 nm) against PH-302 concentra-
tion, a spectral binding constant (Ks) of 5.5 µM was estimated
using nonlinear regression (Figure 3A, inset). In an attempt to
decipher the magnitude of the type II binding of PH-302
observed in Figure 3A, more extensive spectral analysis was
performed. The type II ligand aniline has been used to effectively
convert all of the P450 in a sample to the nitrogen-liganded
low spin iron state, thus acting as a reference for heme
absorbance changes induced by type I substrates of interest in
absolute binding spectra (27). However, monitoring the Soret
band of a predominantly low spin P450 population shift upon
the addition of different compounds is difficult in absolute
spectra when type II binding ligands are of interest because the
shift of the low spin Soret band is slight (417 to 420 nm). Hence,
we have determined an extinction coefficient for P450 3A4
under saturating levels of a type II ligand based on difference
spectra. Because imidazole is structurally similar to the nitrogen-
bonded imidazole in PH-302, but is likely not hindered sterically
from binding to the heme iron, the type II spectrum of imidazole
Results
P450 Inhibition Screening. P450 inhibition screening studies
using recombinant cytochrome P450 enzymes and fluorescent
probe substrates indicated that 3 µM PH-302 inhibited P450s
2D6, 2C9, and 3A4 by 32, 17, and 60%, respectively (N ) 5),
whereas P450 1A2 and 2C19 activity was unaffected by PH-
302 (Figure S1, Supporting Information).
Inhibition Constant (Ki) Determination. To more clearly
define the inhibition potency of PH-302 against P450 3A4 in
human liver microsomes, a Dixon Plot analysis was performed.
PH-302 had an apparent Ki value of 2 µM without preincubation
for P450 3A4 when both midazolam and testosterone were used
as marker substrates (Figure 2A and B, respectively). Linear
regressions from the three concentrations of the probe substrate
(midazolam or testosterone) in the Dixon Plot intersected above

1654 Chem. Res. Toxicol., Vol. 19, No. 12, 2006
Hutzler et al.
Figure 4. (A) Radio-chromatogram illustrating a single NADPH-dependent metabolite upon incubation of PH-302 with human liver microsomes.
The formation of the catechol (retention time 12.5 min) was confirmed by coelution with an authentic standard using in-line diode array detection
(254 nm); DPM, disintegrations per minute. (B) Inhibition of PH-302 metabolism to the single catechol metabolite (retention time 0.61 min) by
co-incubation of the specific P450 3A4 inhibitor ketoconazole.
should represent the signal (∆A(432-410nm)) given by P450 3A4
when the majority of the enzyme is both saturated and bound
in a type II fashion. In these studies, an extinction coefficient
for imidazole, ꢀ(432-410), was determined to be 58.2 mM-1 (55.8
and 60.6 mM-1 on separate days). When PH-302 was titrated
to saturating concentrations with the same concentration of
purified P450 3A4, it could only produce ∼72% of the spectral
shift relative to that achieved with imidazole (Figure 3B),
suggesting the likelihood of a second, albeit less preferred,
binding orientation for PH-302.
Metabolism and Reaction Phenotyping. The metabolism
of PH-302 in human liver microsomes was investigated using
radiolabeled material ([3H]-PH-302). Following a 30 min
incubation with human liver microsomes, only one NADPH-
dependent peak was observed with a retention time of 12.5 min
(Figure 4A). This peak was identified as the catechol metabolite
on the basis of coelution with the catechol authentic standard
upon UV detection (Abs 254 nm) under the same chromato-
graphic conditions (data not shown) as well as retention time
using mass spectrometric techniques (Figure 4B). Subsequent
phenotyping studies to identified the primary CYP responsible
for the metabolism of PH-302 to the catechol product and
suggested that P450 3A4 played a dominant role because the
formation of the catechol was substantially diminished when
PH-302 was co-incubated with 1 µM ketoconazole (Figure 4B).
Time-Dependent Inhibition. Concerns about the methyl-
enedioxyphenyl functional group prompted the investigation of
time-dependent inhibition via the use of an IC50-shift experiment
(Figure S2, Supporting Information). It was observed upon a
10 min preincubation that the IC50 decreased from 5.2 to 4.4
µM. For kinetic characterization, PH-302 was preincubated with
recombinant P450 3A4 Supersomes for up to 10 min, and time-
dependent inactivation of P450 3A4-mediated midazolam 1′-
hydroxylation was observed in a concentration- and NADPH-
dependent fashion (Figure 5A), whereas no inactivation of P450
3A4 was observed in the absence of PH-302. The inactivation
of P450 3A4 was characterized by a biphasic profile, with a
rapid and short initial phase (0-1.5 min), characterized by a

Complex P450 3A4 Heme Interactions
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1655
Figure 5. (A) Log % remaining activity vs preincubation time (0-10
min), illustrating time-dependent inhibition of midazolam 1′-hydroxyl-
ation by PH-302 (0-20 µM). Kinetically, inactivation was characterized
by a biphasic profile (0-1.5 min, and 1.5 to 10 min). (B) Kitz-Wilson
analysis showing time-dependent inhibition of P450 3A4-mediated
midazolam 1′-hydroxylation by PH-302. A kinact of 0.08 min-1 and a
Ki of 1.2 µM was determined for the first phase (9 0-1.5 min
preincubation), and a kinact of 0.06 min-1 and a Ki of 23.8 µM was
determined for the slower second phase (2 1.5-10 min preincubation).
Because the second phase of the 1 µM PH-302 concentration had a
zero slope, four data points were used to define inactivation kinetic
parameters for this phase.
Figure 6. (A) Spectrophotometric characterization of metabolite-
inhibitory (MI) complex formation with the positive control fluoxetine
(100 µM). After 15 min of incubation with P450 3A4 Supersomes, a
saturating concentration of ketoconazole (10 µM) was titrated into the
sample cuvette, inducing a type II shift, in addition to the fluoxetine
MI complex already present. Data suggests similar absorbance proper-
ties of the type II and MI complex heme species. (B) Spectrophotometric
characterization of metabolite-inhibitory (MI) complex formation
(increase absorbance at 455 nm) via the metabolism of PH-302 by P450
3A4 Supersomes containing cytochrome b5. Noticeable are the rapid
plateau of MI complex formation and the simultaneous existence of a
type II peak and trough, indicating simultaneous, yet distinct, heme-
coordinating interactions within the same molecule.
kinact of 0.08 min-1 and a Ki of 1.2 µM, and a slower second
phase (1.5-10 min), characterized by a kinact of 0.06 min-1 and
a Ki of 23.8 µM (Figure 5B). Studies investigating the
consumption of PH-302 under the same incubation conditions
as those in the inactivation studies showed that ∼85% of PH-
302 was remaining after the first 1.5 min of incubation, and a
biphasic profile was observed that was consistent with the
inactivation profile (Figure S3, Supporting Information).
MI Complex Characterization. Experiments with fluoxetine,
a positive control for MI complex formation with P450 3A4,
and ketoconazole, a known type II binding inhibitor, were
performed to evaluate the potential differences in absorbances
with these respective heme complexes. What was observed after
incubation with piperonyl butoxide was a second peak at ∼425-
430 nm that was NADPH-independent, reported previously by
Beumel et al. (28). This prevented a direct comparison of the
type II signal from ketoconazole with the MI complex (data
not shown). Thus, fluoxetine, despite forming an MI complex
via a distinct nitroso intermediate, was used for this exercise.
After a 15 min incubation of fluoxetine with P450 3A4
Supersomes, a saturating concentration of ketoconazole (10 µM)
was added to the sample cuvette, which then induced a type II
absorbance shift that had an Absmax at 430 nm, and was only
slightly higher than the peak absorbance of the fluoxetine MI
complex species at 455 nm (Figure 6A). This indicated to us
that there likely would not be substantially different extinction
coefficients for type II and MI complex heme coordinated
species. Incubation of PH-302 yielded a time-dependent increase
in UV absorbance at 455 nm, consistent with formation of a
metabolite-inhibitory (MI) complex, presumably because of
coordination of the carbene intermediate to the heme. The MI
complex-associated increase in absorbance at 455 nm was
observed in addition to a characteristic type II spectrum from
the imidazole coordination with the heme (Figure 6B). The
presence of a type II spectrum was verified with P450 3A4 in
this system after the addition of excess dithionite (Figure S4,
Supporting Information), and thus, we believe that the peak
observed at 425 nm is part of the imidiazole-derived type II
spectra and not a result of the MI complex, as suggested by
Beumel et al., with other carbene intermediates (28). Interest-
ingly, the rate of increase in absorbance at 455 nm appeared to
slow within 3 min with the P450 3A4 Supersomes, whereas
the type II spectrum due to imidazole coordination to the heme
was captured in the same wavelength scan. Interestingly, the
time course of the initial inactivation phase from Figure 5A

1656 Chem. Res. Toxicol., Vol. 19, No. 12, 2006
Hutzler et al.
Scheme 1. Metabolism of PH-302 by P450 3A4 to its Catechol Metabolite, with the Carbene Intermediate Proposed to
Coordinate with the Heme of P450 3A4 and Cause Time-Dependent Inhibitiona
a This is in addition to the imidazole coordinating with the heme of P450, causing type II binding spectra.
closely agrees with this spectral information, with the slower
second phase of inactivation occurring after the MI complex
formation rate appears to slow down to ∼3 min.
generally had improved metabolic stability relative to that of
type I binding substrates (32). Closer inspection of the type II
binding spectra of PH-302 showed that the absorbance trough
was asymmetrical and broad (Figure 3A), which suggests that
there may be a type I binding component co-incident with the
preferred type II heme coordination (33). Upon further spectral
characterization, it was determined that saturating concentrations
of PH-302 could only induce a type II spectral shift that gave
72% of the trough-to-peak absorbance difference relative to that
obtained with the heme-coordinating fragment imidazole when
exposed to the same amount of the P450 3A4 protein (Figure
3B). This to us provided evidence that a distinct binding mode
more amenable to metabolism may coexist in the overall
population of the P450 3A4 enzyme. One assumption from this
interpretation is that the spectral signal (ꢀ) observed for
imidazole and PH-302 are primarily derived from heme
coordination. Interestingly, the larger P450 3A4 inhibitor
itraconazole has been shown to be metabolized at various
positions, despite its known type II binding interaction with P450
3A4 (34, 35). Thus, we decided to investigate the metabolism
of PH-302 by obtaining [3H]-PH-302 and incubating with human
liver microsomes. From these studies, a single metabolite was
observed in the radio-chromatogram (Figure 4A), which we
determined to be the catechol metabolite, consistent with the
metabolism of the methylenedioxyphenyl functional group (12,
36). Furthermore, the formation of the catechol metabolite was
substantially inhibited when PH-302 was co-incubated with
ketoconazole (Figure 4B), suggesting that P450 3A4 was the
major P450 involved in the metabolism of PH-302 to the
catechol metabolite.
Discussion
PH-302 is a pyrimidineimidazole that potently and selectively
inhibits iNOS (1), an enzyme involved in the excess production
of nitric oxide (NO) and thus implicated in the pathogenesis of
inflammation and rheumatoid arthritis (29). Because of the
mechanism by which PH-302 inhibits the iNOS enzyme (heme
coordination), it seemed likely that the inhibition of the heme-
containing cytochrome P450 super family of drug-metabolizing
enzymes was a potential liability. PH-302 possesses within its
chemical structure an imidazole as well as a methylenediox-
yphenyl (MDP) functionality, both known to interact with the
heme of P450 enzymes and cause potent inhibition of P450
enzymatic activity, albeit via distinct mechanisms. In these
studies, a mechanistic evaluation of the observed inhibition of
P450 3A4 suggested that two distinct mechanisms were
simultaneously operable: reversible heme coordination (type
II binding) by the imidazole and time-dependent pseudo-
irreversible metabolite-inhibitory complex formation from the
metabolism of the methylenedioxyphenyl functional group.
Initial studies investigating the mechanism by which PH-302
inhibits P450 3A4 were performed using spectral binding and
kinetic analysis. Spectral data clearly showed the coordination
of the imidazole of PH-302 with the P450 heme, characterized
spectrally by a broad absorbance trough at 390-405 nm and a
sharper absorbance peak at 430 nm (Figure 3A), suggesting this
to be the preferred binding orientation. This interaction via heme
coordination was characterized by a spectral affinity binding
constant (Ks) of 5.5 µM. Interestingly, this affinity constant
closely resembled the apparent Ki values (∼2 µM) for P450
3A4 in human liver microsomes when using both midazolam
and testosterone as probe substrates (Figure 2A and B),
indicating that spectral binding affinity constants may be useful
for evaluating the potency of certain P450 3A4 inhibitors.
Although there are numerous examples in the literature of
substrate-dependent inhibition profiles of P450 3A4 (30, 31),
other investigators have shown similar inhibition potencies in
the presence of heme-coordinating type II inhibitors, such as
ketoconazole and itraconazole, when different probe substrates
are used (10), consistent with observations from our studies with
PH-302.
The observation of a catechol being the major metabolite
carries with it a secondary concern: time-dependent inhibition
of P450 3A4, which has led to a number of reported drug-
drug interactions, including those observed for calcium channel
blockers (37). In concert with catechol metabolite formation, a
carbene intermediate is proposed to be formed (Scheme 1) and
is the most frequently cited mechanism for the time-dependent
inhibition of P450 enzymes because of the metabolism of MDP
groups (14). The formation of a carbene intermediate is reported
to arise by the abstraction of a hydrogen atom from the
methylenic carbon or by the elimination of water from a
hydroxy-methylene intermediate (15). When evaluating the time-
dependent inhibition of P450 3A4 by PH-302, the magnitude
of inhibition increased (IC50 left-shift) upon a 10 min preincu-
bation with PH-302 (Figure S2, Supporting Information),
suggesting that the metabolism of the methylenedioxyphenyl
group, despite the imidazole coordinating to the heme, was a
viable metabolic pathway. We are confident that neither further
oxidation of the catechol to the ortho-quinone, nor reactive
oxygen species formation was involved in the time-dependent
inhibition of P450 3A4 in these studies on the basis of a
On the basis of results from spectral studies in Figure 3A,
suggesting that the imidazole coordinating with heme is the
preferred binding orientation, a reasonable assumption would
seem to be that PH-302 may not be metabolized to a substantial
degree in human liver microsomes. Similar to this notion, Chiba
et al. suggested that within a chemical series of HIV protease
inhibitors, analogues that bound P450 3A4 in a type II fashion

Complex P450 3A4 Heme Interactions
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1657
preliminary assessment of the in vitro stability of the catechol
metabolite (unpublished work). The kinetics for the observed
time-dependent inhibition of P450 3A4 were then determined
using the recombinant P450 3A4 enzyme. It was observed that
the inactivation of midazolam 1′-hydroxylation (Figure 5A) was
characterized by a biphasic profile with an initial rapid phase
of inactivation, characterized by a kinact of 0.08 min-1 and a Ki
of 1.2 µM, followed by a slower second phase, characterized
by a kinact of 0.06 min-1 and a Ki of 23.8 µM (Figure 5B). The
kinact of 0.08 min-1 and a Ki of 1.2 µM describing the initial
phase of inactivation is consistent with that reported for the
N-desmethyl metabolite of the anticancer drug tamoxifen (38),
the anti-HIV agent amprenavir (39), and the antihypertensive
drug verapamil (37), whereas the kinetic parameters describing
the second phase were consistent with those reported for
irinotecan (40), indicating the potential clinical relevance of
P450 3A4 inactivation by PH-302. It is important to note here
that without the collection of early preincubation time points
(pre-2 min), the earlier phase of inactivation would likely have
been missed, potentially affecting the accuracy of in vitro-in
vivo correlations that may be performed with this type of data.
with interpreting the binding orientations of this molecule with
the large P450 3A4 active site and potentially makes the PH-
302 an especially useful tool for future investigations of P450
3A4 binding and/or cooperativity.
An intriguing possible contributor to our observations is the
argument of kinetically distinct conformers of P450 3A4,
initially proposed by Koley et al. (46, 47). Additional evidence
in support of P450 3A4 conformational heterogeneity has
recently been published because Davydov et al. applied high-
pressure spectroscopy and demonstrated that only 70% of the
P450 3A4 enzyme pool in solution was susceptible to a
hydrostatic pressure-induced P450 to P420 transition (48). In
addition, multiexponential dithionite-dependent reduction kinet-
ics of P450 3A4 in the absence and presence of substrates has
been reported (49), further developing the potential of P450 3A4
oligomerization. It seems plausible to suggest that multiple
conformers of P450 3A4 may be involved in the observed
interactions with PH-302, where the first kinetic phase of
inactivation (0-1.5 min) in Figure 5A and the rapid saturation
of MI complex formation in Figure 6B may result from the
rapid inactivation of a small pool of the P450 3A4 enzyme.
Meanwhile, a larger pool of enzyme may predominately bind
PH-302 in a type II orientation. However, this hypothesis cannot
be confirmed with the current studies. Other possibilities include
distinct on/off rates for the type II versus productive binding
orientations, perhaps in a dynamic equilibrium, and the pos-
sibility that the initial MI complex may be displaced by other
molecules of PH-302, as has been demonstrated with isosafrole
and lipophilic type I binding compounds (50). If the PH-302-
derived MI complex can be dissociated, a previously complexed
P450 3A4 molecule would then be available to go through
another productive cycle and then potentially become inactivated
again, perhaps describing a slower phase of inactivation. More
extensive studies evaluating this dissociation hypothesis are
under way.
In summary, PH-302, a molecule possessing two structural
moieties known to cause the inhibition of P450 enzymes, was
found to inhibit P450 3A4 via both type II binding and time-
dependent MI complex formation. This observation is consistent
with previous reports describing the complex binding kinetics
within P450 3A4, and data from spectral studies appears to
support previous reports proposing kinetically distinct conform-
ers of P450 3A4. Because of its distinct heme interactions with
P450 3A4, PH-302 could prove to be a useful tool for further
evaluation of ligand-binding interactions of P450 3A4. Finally,
these studies underwrite the risks of disregarding the possibility
of time-dependent inhibition when type II binding is observed
in spectral studies with purified non-functional enzymes because
this may lead to an underprediction of P450 3A4-mediated
drug-drug interactions (DDI).
Of particular interest from these studies is a clearer under-
standing of the distinct heme-coordinating mechanisms by which
PH-302 apparently inhibits P450 3A4, consistent with the widely
documented complex binding kinetics/interactions characteristic
of this enzyme (21, 41, 42). To dissect these distinct heme-
coordinating mechanisms, the positive control fluoxetine was
used to generate an MI complex with P450 3A4 (43). After 15
min of incubation with fluoxetine, a saturating concentration
(10 µM) of the known type II binding inhibitor ketoconazole
was added to see how the absorbance pattern would be
perturbed. As observed in Figure 6A, both the MI complex from
fluoxetine (Absmax 455 nm) and the type II difference spectra
from ketoconazole could be captured in a single UV scan,
indicating that both heme-coordinated species were present
simultaneously. The question waiting to be answered, Could
this pattern be observed with a single molecule to demonstrate
spectral evidence for simultaneous distinct binding orientations?
Spectral analysis of PH-302 under active metabolic conditions
confirmed that a metabolite-inhibitory complex was formed,
characterized by a time-dependent increase in absorbance at 455
nm, in addition to type II spectra derived from the imidazole
end of the PH-302 molecule (Figure 6B). The observation that
MI complex formation by PH-302 in P450 3A4 Supersomes
(with cytochrome b5) appears to begin to saturate within the
first 3 min of incubation (Figure 6B) is especially interesting
because the timing is consistent with the initial phase of the
biphasic inactivation kinetic profile observed in Figure 5A. This
rapid saturation of MI complex formation with a methylene-
dioxyphenyl-containing compound has been noted in a separate
report with the herbal supplement hydrastine and human liver
microsomes (17), although the underlying mechanism was not
investigated. In addition, although the possibility of multiple
PH-302 molecules in the P450 3A4 active site (21, 44) and/or
the allosteric binding (45) of PH-302 contributing to the
observed interactions cannot be completely ruled out, it does
not appear to be consistent with our findings because both
binding interactions monitored in these studies are due to direct
coordination with the P450 heme. The apparent concomitant
type II and MI complex heme-iron coordinating interactions
captured in Figure 6B cannot theoretically occur simultaneously
within the same active site of a P450 3A4 molecule. This, along
with the fact that only the catechol metabolite is formed via
the MI complex pathway, helps reduce the ambiguity associated
Acknowledgment. We acknowledge David Davey from
Berlex Biosciences for supplying the original PH-302 material
and Babu Subramanyam from Berlex for initial discussions
during in-licensing. We also thank Brad McKinnis for the
purification of the PH-302 radiolabeled material and Chante
Bolden for her assistance and analytical expertise. Patent # US6,-
432,947.
Supporting Information Available: Initial P450 inhibition
screening data showing potent inhibition of P450 3A4, IC50 data
showing a shift in IC50 from 5.2 to 4.4 µM for inhibition of
midazolam 1′-hydroxylation following a 10 min preincubation of
PH-302 in human liver microsomes, data demonstrating lack of
significant depletion of PH-302 when incubated with P450 3A4

1658 Chem. Res. Toxicol., Vol. 19, No. 12, 2006
Hutzler et al.
(22) Hosea, N. A., Miller, G. P., and Guengerich, F. P. (2000) Elucidation
of distinct ligand binding sites for cytochrome P450 3A4. Biochemistry
39, 5929-5939.
Supersomes under the same incubation conditions as those of the
inactivation experiments, and verification of type II spectra with
PH-302 with reduced Fe-II. This material is available free of charge
via the Internet at http://pubs.acs.org.
(23) McMillan, K., Adler, M., Auld, D. S., Baldwin, J. J., Blasko, E.,
Browne, L. J., Chelsky, D., Davey, D., Dolle, R. E., Eagen, K. A.,
Erickson, S., Feldman, R. I., Glaser, C. B., Mallari, C., Morrissey,
M. M., Ohlmeyer, M. H., Pan, G., Parkinson, J. F., Phillips, G. B.,
Polokoff, M. A., Sigal, N. H., Vergona, R., Whitlow, M., Young, T.
A., and Devlin, J. J. (2000) Allosteric inhibitors of inducible nitric
oxide synthase dimerization discovered via combinatorial chemistry.
Proc. Natl. Acad. Sci. U.S.A. 97, 1506-1511.
(24) Cohen, L. H., Remley, M. J., Raunig, D., and Vaz, A. D. (2003) In
vitro drug interactions of cytochrome p450: an evaluation of fluoro-
genic to conventional substrates. Drug Metab. Dispos. 31, 1005-1015.
(25) Kitz, R. J., Ginsburg, S., and Wilson, I. B. (1967) The reaction of
acetylcholinesterase with diethylphosphoryl esters of quaternary and
tertiary aminophenols. Mol. Pharmacol. 3, 225-232.
(26) Segel, I. H. (1975) Enzyme Kinetics, p 110, John Wiley & Sons, New
York.
(27) Roberts, A. G., Campbell, A. P., and Atkins, W. M. (2005) The
thermodynamic landscape of testosterone binding to cytochrome P450
3A4: ligand binding and spin state equilibria. Biochemistry 44, 1353-
1366.
References
(1) Blasko, E., Glaser, C. B., Devlin, J. J., Xia, W., Feldman, R. I.,
Polokoff, M. A., Phillips, G. B., Whitlow, M., Auld, D. S., McMillan,
K., Ghosh, S., Stuehr, D. J., and Parkinson, J. F. (2002) Mechanistic
studies with potent and selective inducible nitric-oxide synthase
dimerization inhibitors. J. Biol. Chem. 277, 295-302.
(2) Wrighton, S. A., and Stevens, J. C. (1992) The human hepatic
cytochromes P450 involved in drug metabolism. Crit. ReV. Toxicol.
22, 1-21.
(3) Guengerich, F. P. (2002) Update information on human P450s. Drug
Metab. ReV. 34, 7-15.
(4) Shimada, T., Yamazaki, H., Mimura, M., Inui, Y., and Guengerich,
F. P. (1994) Interindividual variations in human liver cytochrome
P-450 enzymes involved in the oxidation of drugs, carcinogens and
toxic chemicals: studies with liver microsomes of 30 Japanese and
30 Caucasians. J. Pharmacol. Exp. Ther. 270, 414-423.
(5) Lomaestro, B. M., and Piatek, M. A. (1998) Update on drug
interactions with azole antifungal agents. Ann. Pharmacother. 32, 915-
928.
(6) Fleishaker, J. C., Herman, B. D., Carel, B. J., and Azie, N. E. (2003)
Interaction between ketoconazole and almotriptan in healthy volunteers.
J. Clin. Pharmacol. 43, 423-427.
(7) Moody, D. E., Walsh, S. L., Rollins, D. E., Neff, J. A., and Huang,
W. (2004) Ketoconazole, a cytochrome P450 3A4 inhibitor, markedly
increases concentrations of levo-acetyl-alpha-methadol in opioid-naive
individuals. Clin. Pharmacol. Ther. 76, 154-166.
(8) Schenkman, J. B., Sligar, S. G., and Cinti, D. L. (1981) Substrate
interaction with cytochrome P-450. Pharmacol. Ther. 12, 43-71.
(9) Schenkman, J. B., and Jansson, I. (1998) Spectral analyses of
cytochromes P450. Methods Mol. Biol. 107, 25-33.
(10) Racha, J. K., Zhao, Z. S., Olejnik, N., Warner, N., Chan, R., Moore,
D., and Satoh, H. (2003) Substrate dependent inhibition profiles of
fourteen drugs on CYP3A4 activity measured by a high throughput
LCMS/MS method with four probe drugs, midazolam, testosterone,
nifedipine and terfenadine. Drug Metab. Pharmacokinet. 18, 128-
138.
(11) Amacher, D. E., Fasulo, L. M., Charuel, C., Comby, P., and Beaumont,
K. (1998) In vitro toxicity of zamifenacin (UK-76,654) and metabolites
in primary hepatocyte cultures. Xenobiotica 28, 895-908.
(12) Beaumont, K. C., Causey, A. G., Coates, P. E., and Smith, D. A. (1996)
Pharmacokinetics and metabolism of zamifenacin in mouse, rat, dog
and man. Xenobiotica 26, 459-471.
(13) Mansuy, D., Carlier, M., Bertrand, J. C., and Azoulay, E. (1980)
Spectral characterization of cytochrome P-450 of a strain of Candida
tropicalis grown on tetradecane. Eur. J. Biochem. 109, 103-108.
(14) Ortiz de Montellano, P. R. (1995) Inhibition of Cytochrome P450
Enzymes. In Cytochrome P450: Structure, Mechanism, and Biochem-
istry (Ortiz de Montellano, P. R., Ed.) pp 305-364, Plenum Press,
New York.
(28) Beumel, G. A., Levi, P. E., and Hodgson, E. (1985) Spectral
interactions of piperonyl butoxide and isocyanides with purified hepatic
cytochrome P-450 from uninduced mice. Gen. Pharmacol. 16, 193-
197.
(29) Salerno, L., Sorrenti, V., Di Giacomo, C., Romeo, G., and Siracusa,
M. A. (2002) Progress in the development of selective nitric oxide
synthase (NOS) inhibitors. Curr. Pharm. Des. 8, 177-200.
(30) Wang, R. W., Newton, D. J., Liu, N., Atkins, W. M., and Lu, A. Y.
(2000) Human cytochrome P-450 3A4: in vitro drug-drug interaction
patterns are substrate-dependent. Drug Metab. Dispos. 28, 360-366.
(31) Stresser, D. M., Blanchard, A. P., Turner, S. D., Erve, J. C., Dandeneau,
A. A., Miller, V. P., and Crespi, C. L. (2000) Substrate-dependent
modulation of CYP3A4 catalytic activity: analysis of 27 test
compounds with four fluorometric substrates. Drug Metab. Dispos.
28, 1440-1448.
(32) Chiba, M., Jin, L., Neway, W., Vacca, J. P., Tata, J. R., Chapman,
K., and Lin, J. H. (2001) P450 interaction with HIV protease
inhibitors: relationship between metabolic stability, inhibitory potency,
and P450 binding spectra. Drug Metab. Dispos. 29, 1-3.
(33) Schenkman, J. B. (1970) Studies on the nature of the type I and type
II spectral changes in liver microsomes. Biochemistry 9, 2081-2091.
(34) Isoherranen, N., Kunze, K. L., Allen, K. E., Nelson, W. L., and
Thummel, K. E. (2004) Role of itraconazole metabolites in CYP3A4
inhibition. Drug Metab. Dispos. 32, 1121-1131.
(35) Kunze, K. L., Nelson, W. L., Kharasch, E. D., Thummel, K. E., and
Isoherranen, N. (2006) Stereochemical aspects of itraconazole me-
tabolism in vitro and in vivo. Drug Metab. Dispos. 34, 583-590.
(36) Klungsoyr, J., and Scheline, R. R. (1982) Metabolism of isosafrole
and dihydrosafrole in the rat. Biomed. Mass Spectrom. 9, 323-329.
(37) Ma, B., Prueksaritanont, T., and Lin, J. H. (2000) Drug interactions
with calcium channel blockers: possible involvement of metabolite-
intermediate complexation with CYP3A. Drug Metab. Dispos. 28,
125-130.
(38) Zhao, X. J., Jones, D. R., Wang, Y. H., Grimm, S. W., and Hall, S.
D. (2002) Reversible and irreversible inhibition of CYP3A enzymes
by tamoxifen and metabolites. Xenobiotica 32, 863-878.
(39) von Moltke, L. L., Durol, A. L., Duan, S. X., and Greenblatt, D. J.
(2000) Potent mechanism-based inhibition of human CYP3A in vitro
by amprenavir and ritonavir: comparison with ketoconazole. Eur. J.
Clin. Pharmacol. 56, 259-261.
(40) Hanioka, N., Ozawa, S., Jinno, H., Tanaka-Kagawa, T., Nishimura,
T., Ando, M., and Sawada, Ji, J. (2002) Interaction of irinotecan (CPT-
11) and its active metabolite 7-ethyl-10-hydroxycamptothecin (SN-
38) with human cytochrome P450 enzymes. Drug Metab. Dispos. 30,
391-396.
(41) Atkins, W. M. (2005) Non-Michaelis-Menten kinetics in cytochrome
P450-catalyzed reactions. Annu. ReV. Pharmacol. Toxicol. 45, 291-
310.
(42) Shou, M., Grogan, J., Mancewicz, J. A., Krausz, K. W., Gonzalez, F.
J., Gelboin, H. V., and Korzekwa, K. R. (1994) Activation of
CYP3A4: evidence for the simultaneous binding of two substrates in
a cytochrome P450 active site. Biochemistry 33, 6450-6455.
(43) Mayhew, B. S., Jones, D. R., and Hall, S. D. (2000) An in vitro model
for predicting in vivo inhibition of cytochrome P450 3A4 by metabolic
intermediate complex formation. Drug Metab. Dispos. 28, 1031-1037.
(44) Dabrowski, M. J., Schrag, M. L., Wienkers, L. C., and Atkins, W. M.
(2002) Pyrene‚pyrene complexes at the active site of cytochrome P450
3A4: evidence for a multiple substrate binding site. J. Am. Chem.
Soc. 124, 11866-11867.
(15) Murray, M., and Reidy, G. F. (1989) In vitro formation of an inhibitory
complex between an isosafrole metabolite and rat hepatic cytochrome
P-450 PB-B. Drug Metab. Dispos. 17, 449-454.
(16) Bertelsen, K. M., Venkatakrishnan, K., Von, Moltke, L. L., Obach,
R. S., and Greenblatt, D. J. (2003) Apparent mechanism-based
inhibition of human CYP2D6 in vitro by paroxetine: comparison with
fluoxetine and quinidine. Drug Metab. Dispos. 31, 289-293.
(17) Chatterjee, P., and Franklin, M. R. (2003) Human cytochrome p450
inhibition and metabolic-intermediate complex formation by goldenseal
extract and its methylenedioxyphenyl components. Drug Metab.
Dispos. 31, 1391-1397.
(18) Yano, J. K., Wester, M. R., Schoch, G. A., Griffin, K. J., Stout, C.
D., and Johnson, E. F. (2004) The structure of human microsomal
cytochrome P450 3A4 determined by X-ray crystallography to 2.05-A
resolution. J. Biol. Chem. 279, 38091-38094.
(19) Scott, E. E., and Halpert, J. R. (2005) Structures of cytochrome P450
3A4. Trends Biochem. Sci. 30, 5-7.
(20) Ekroos, M., and Sjogren, T. (2006) Structural basis for ligand
promiscuity in cytochrome P450 3A4. Proc. Natl. Acad. Sci. U.S.A.
103, 13682-13687.
(21) Korzekwa, K. R., Krishnamachary, N., Shou, M., Ogai, A., Parise, R.
A., Rettie, A. E., Gonzalez, F. J., and Tracy, T. S. (1998) Evaluation
of atypical cytochrome P450 kinetics with two-substrate models:
evidence that multiple substrates can simultaneously bind to cyto-
chrome P450 active sites. Biochemistry 37, 4137-4147.

Complex P450 3A4 Heme Interactions
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1659
(45) Atkins, W. M. (2004) Implications of the allosteric kinetics of
revealed by high pressure spectroscopy. Biochem. Biophys. Res.
Commun. 312, 121-130.
cytochrome P450s. Drug DiscoVery Today 9, 478-484.
(49) Davydov, D. R., Fernando, H., Baas, B. J., Sligar, S. G., and Halpert,
J. R. (2005) Kinetics of dithionite-dependent reduction of cytochrome
P450 3A4: heterogeneity of the enzyme caused by its oligomerization.
Biochemistry 44, 13902-13913.
(50) Dickins, M., Elcombe, C. R., Moloney, S. J., Netter, K. J., and Bridges,
J. W. (1979) Further studies on the dissociation of the isosafrole
metabolite-cytochrome P-450 complex. Biochem. Pharmacol. 28,
231-238.
(46) Koley, A. P., Buters, J. T., Robinson, R. C., Markowitz, A., and
Friedman, F. K. (1995) CO binding kinetics of human cytochrome
P450 3A4. Specific interaction of substrates with kinetically distin-
guishable conformers. J. Biol. Chem. 270, 5014-5018.
(47) Koley, A. P., Robinson, R. C., and Friedman, F. K. (1996) Cytochrome
P450 conformation and substrate interactions as probed by CO binding
kinetics. Biochimie 78, 706-713.
(48) Davydov, D. R., Halpert, J. R., Renaud, J. P., and Hui Bon Hoa, G.
(2003) Conformational heterogeneity of cytochrome P450 3A4
TX060198M