Comparative study of the affinity and metabolism of type i and type II binding quinoline carboxamide analogues by cytochrome P450 3A4

Article abstract of DOI:10.1021/jm201207h

Compounds that coordinate to the heme-iron of cytochrome P450 (CYP) enzymes are assumed to increase metabolic stability. However, recently we observed that the type II binding quinoline carboxamide (QCA) compounds were metabolically less stable. To test if the higher intrinsic clearance of type II binding compounds relative to type I binding compounds is general for other metabolic transformations, we synthesized a library of QCA compounds that could undergo N-dealkylation, O-dealkylation, benzylic hydroxylation, and aromatic hydroxylation. The results demonstrated that type II binding QCA analogues were metabolically less stable (2- to 12-fold) at subsaturating concentration compared to type I binding counterparts for all the transformations. When the rates of different metabolic transformations between type I and type II binding compounds were compared, they were found to be in the order of N-demethylation > benzylic hydroxylation> O-demethylation > aromatic hydroxylation. Finally, for the QCA analogues with aza-heteroaromatic rings, we did not detect metabolism in aza-aromatic rings (pyridine, pyrazine, pyrimidine), indicating that electronegativity of the nitrogen can change regioselectivity in CYP metabolism.

Full text of DOI:10.1021/jm201207h

Article
pubs.acs.org/jmc
Comparative Study of the Affinity and Metabolism of Type I and
Type II Binding Quinoline Carboxamide Analogues by Cytochrome
P450 3A4
Upendra P. Dahal, Carolyn Joswig-Jones, and Jeffrey P. Jones*
Department of Chemistry, Washington State University, P.O. Box 644630, Pullman, Washington 99164-4630, United States
S
* Supporting Information
ABSTRACT: Compounds that coordinate to the heme-iron
of cytochrome P450 (CYP) enzymes are assumed to increase
metabolic stability. However, recently we observed that the
type II binding quinoline carboxamide (QCA) compounds
were metabolically less stable. To test if the higher intrinsic
clearance of type II binding compounds relative to type I
binding compounds is general for other metabolic trans-
formations, we synthesized a library of QCA compounds that
could undergo N-dealkylation, O-dealkylation, benzylic
hydroxylation, and aromatic hydroxylation. The results demonstrated that type II binding QCA analogues were metabolically
less stable (2- to 12-fold) at subsaturating concentration compared to type I binding counterparts for all the transformations.
When the rates of different metabolic transformations between type I and type II binding compounds were compared, they were
found to be in the order of N-demethylation > benzylic hydroxylation> O-demethylation > aromatic hydroxylation. Finally, for
the QCA analogues with aza-heteroaromatic rings, we did not detect metabolism in aza-aromatic rings (pyridine, pyrazine,
pyrimidine), indicating that electronegativity of the nitrogen can change regioselectivity in CYP metabolism.
nated high spin state (see Figure 1). Reduction of the iron(III)
into iron(II) by cytochrome P450 reductase with the cofactor
NADPH is followed by binding of a molecular oxygen into
heme-iron(II). It has been hypothesized that formation of the
high-spin pentacoordinated heme-iron(III) is required to
initiate the catalytic cycle.1,8 Binding of a compound leading
to an increase in formation of the high-spin penta-coordinated
iron(III) is observable in the UV spectra and is called type I
binding (Figure 2). If a substrate has a sterically unhindered
aromatic nitrogen, the molecule can bind with heme-iron which
will prevent formation of the high-spin pentacoordinated heme-
iron. This binding mode, which leads to an increase in low-spin
hexa-coordinated iron(III), is known as type II binding (Figure
2). The two binding modes, namely type I and type II, can be
easily differentiated spectrophotometrically and were discov-
ered more than 4 decades ago.9−11
Since type II binding compounds have a higher affinity for
CYP enzymes compared to similar type I binding compounds,
type II binding compounds are extensively used as inhibitors of
CYP enzymes.12−14 It has been assumed that type II binding of
a molecule would lead to a dead-end complex and that the drug
cannot be metabolized or that metabolism would be slowed.
This leads to the hypothesis that type II binding compounds
are metabolically more stable than the type I binding
compounds.15,16 Contrary to this hypothesis, Peng et al.17
revealed that CYP3A4 can metabolize a series of type II binding
INTRODUCTION
■
Cytochrome P45O (CYP) enzymes are a superfamily of heme
thiolate proteins that can insert an atom of oxygen from
molecular oxygen into many organic compounds. More than
80% of the clinically used drugs are metabolized by CYP
enzymes.1 In humans, 57 genes were identified that code for
various CYP enzymes. Cytochrome P450 3A4 (CYP3A4) is the
most abundantly expressed isoform in liver and comprises as
much as 60% of total CYP enzymes.2,3 CYP3A4 contributes to
the metabolism of more than 50% of marketed drugs,4,5 and it
is often involved in unwanted drug−drug interactions. Because
of the ubiquitous nature of CYP enzymes in body tissues and
their ability to metabolize multiple organic compounds
containing wide variety of functional groups,6 metabolic
stabilization of drugs is a challenge in drug optimization
process. Drugs need to be metabolized so that they can be
excreted; however, drugs should be stable enough so that they
can reach the site of the action. Hence, optimization of drug
metabolism by CYP enzymes is always a goal during drug
design and development.
CYP enzymes contain heme as a cofactor. The heme is
tethered to the enzyme by a thiolate bond with cysteine. In
resting state of the enzyme, iron of the heme is hexa-
coordinated (coordinated with four nitrogen of porphyrin
system, with the sulfur of cysteine proximally and with water
distally) and is in low-spin state. According to the consensus
mechanism for CYP,1,7 the catalytic cycle of CYP reaction
begins with displacement of water by a substrate that converts
hexa-coordinated low spin heme-iron(III) into pentacoordi-
Received: September 12, 2011
Published: November 16, 2011
© 2011 American Chemical Society
280
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
regioselectivity of various metabolic transformations between
type I and type II binding compounds, we synthesized a library
of QCA analogues. The library contained multiple series of type
I and type II binding compounds with different functional
groups on the same core structure. The binding mode of each
compound was determined using UV/vis difference spectrum,
and saturation kinetics was carried out to determine the kinetic
parameters.
RESULTS
■
Design of the Substrates and Metabolite Quantifica-
tion. In an attempt to systematically compare affinity and
kinetic parameters of type I and type II binding compounds, a
QCA core was derivatized to get a library of compounds
(Figure 3). To probe the different types of oxidation carried out
by the CYP enzymes, four different functional groups were
coupled in the amide position of the QCA, namely, benzene,
toluene, anisole, and N,N-dimethylaniline. Similarly the 2-
position of the quinoline ring of QCA was incorporated with
benzene, pyridine, pyrazine, or pyrimidine to change the
binding mode as well as to investigate the role of various aza-
heteroaromatic rings in binding mode and in metabolism. All of
the substrate were synthesized by the method developed in our
lab17 with a few modifications as shown in Scheme 1. Series 1
compounds are designed without aza-aromatic ring so that they
can acts as type I substrates. Series 2 compounds are designed
with pyrid-4-yl at position 2 of the quinoline ring; these
compounds are type II substrate because of the presence of
accessible aromatic nitrogen. Series 3 compounds are designed
with pyrid-2-yl at position 2 of the quinoline ring. These
compounds are type I substrates, as they do not have accessible
aromatic nitrogen but are electronically equivalent to series 2
type II substrates. Series 4 and series 5 compounds are designed
to compare the effect of multiple nitrogens in a single aromatic
ring, and they contain pyrazine and pyrimidine in position 2 of
the quinoline ring, respectively.
To quantify the metabolites formed enzymatically, at least
one metabolite per substrate was synthesized. The standard
curve was prepared using an internal standard (phenacetin)
versus the standard synthesized metabolite. Although multiple
metabolites were observed for most of these compounds,
ionization potentials of all of the metabolites from a substrate
were assumed to be same with the synthesized metabolite for
the purpose of quantification. Hence, the same standard curve
was used for all the metabolites of a substrate. Structures of the
substrates and the synthesized metabolites are presented in
Table 1.
Figure 1. Consensus mechanism of cytochrome P450 enzymes. Fe3+
represents the pentacoordinated heme-iron in resting state of CYP
enzymes.
Figure 2. Binding modes of CYP enzymes. Fe3+ represents the
pentacoordinated heme-iron: (a) low-spin heme-iron in wild type
enzyme; (b) high-spin heme-iron in type I binding mode; (c) low-spin
heme-iron in type II binding mode.
compounds significantly at subsaturating concentrations. More
recently we18 reported that two type II binding quinoline
carboxamide (QCA) compounds were metabolized more
quickly than a similar type I binding QCA compound. We
observed aromatic hydroxylation on the naphthalene ring of the
QCA analogs. On the basis of the results from kinetics assay,
electrochemical measurements, and surface plasmon resonance
analysis, we proposed a direct reduction kinetic model for the
metabolism of those type II binding compounds. Although it
has been confirmed that the type II binding mode contributes
to the observed high affinity,12,19−22 it is still a subject of debate
whether type II binding increases the metabolic stability of
these compounds. The answer can only be assessed by a
systematic comparison of metabolic rates between type I and
type II binding compounds.
Binding Spectra. The type I binding mode has a
difference spectrum with a Soret maximum at 385−390 nm
and a trough at around 420 nm, whereas type II binding mode
has a difference spectrum with a Soret maximum at 420−435
and a trough at 390−405 nm.10,11 The binding mode of the
substrates was determined in purified enzyme using a split
CYP enzymes catalyze a number of transformations, for
example, aliphatic hydroxylation, aromatic hydroxylation, N-
dealkylation, O-dealkylation, epoxidation, S-oxidation, and N-
oxidation. To compare affinity, metabolic stability, and
Figure 3. Quinoline carboxamide (QCA) core derivatization to build the library of compounds with different functional groups at the amide position
and different aromatic rings at position 2.
281
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
Scheme 1. Synthesis of Quinoline Carboxamide Model Compounds
Table 1. Structures of the Substrates Used for the Study and
Structures of the Metabolites Synthesized To Quantify the
b
Metabolites Formed during the Enzymatic Assay
Figure 4. Schematic picture of two chambered cuvettes used to correct
absorbance of substrate during binding mode determination.
Table 2. UV−Visible Absorbance (Difference Spectra) To
a
Determine Binding Mode of the Compounds
b
series compd Soret peak (nm) Soret trough (nm) binding mode
1
2
3
4
5
1
388
385
392
393
427
427
427
428
388
386
389
391
423
425
427
422
425
422
424
425
415
416
417
417
391
392
394
392
417
416
418
416
396
393
397
391
391
392
390
390
type I
type I
type I
type I
type II
type II
type II
type II
type I
type I
type I
type I
type II
type II
type II
type II
type II
type II
type II
type II
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
a
b
See Experimental Section for procedure of synthesis. The
synthesized metabolite was a major metabolite formed during the
bioassay for the corresponding substrate.
a
A representative spectrum of each series is presented in Supporting
beam Olis spectrophotometer. Obtaining type II binding
spectra is challenging,23 and in addition the QCA analogues
showed absorbance in the visible range, making determination
of difference spectra particularly difficult. To overcome the
problems, we used two chambered cuvettes, as shown in Figure
4, to determine the binding mode. Absorbance of the
compounds along with binding mode is presented in Table 2.
As expected, 2-phenyl QCA analogues and 2-pyridin-2-yl QCA
analogues produced type I binding difference spectra whereas
2-pyridin-4-yl QCA, 2-pyrimidin-4-yl QCA, and 2-pyrazin-2-yl
QCA analogues produced type II binding difference spectra.
The difference spectra of a representative compound from each
series are presented in Supporting Information (Figure S1).
b
Information. Type I binding mode has a difference spectrum with a
maximum at 385−390 nm and a trough at 415−420 nm. Type II
binding mode has a difference spectrum with a maximum at 420−435
and a trough at 390−405 nm.
Affinity Differences between Type I and Type II QCA
Analogues. Affinities of the QCA compounds with CYP3A4
were measured in terms of inhibition constants (Ki).
Testosterone was used as a substrate to determine the
inhibition constants. The results of inhibition assays are
presented in Table 3. The Ki values are an average of triplicate
assays. It was observed that the affinity of the type II binding
compounds is higher than the affinity of their type I binding
282
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
Table 3. Inhibition Constants (Ki), Turnover Number (kcat), Michalis−Menten Constants (KM), and in Vitro Intrinsic
a
Clearance (V/K) of the Quinoline Carboxamide Analogues
a
Kinetic parameter values represent the mean of triplicate assays, and values after “ ” represent standard deviation.
counterparts (specifically 7- to 21-fold higher affinity for type II
binding compounds in series 2 compared to type I binding
compounds in series 3). 2-Pyridin-2-yl QCA analogues (series
2, type II binders) showed strong affinity for CYP3A4, and all
of the compounds of series 2 have Ki values in the nanomolar
range. Type I binding compounds (series 1 and series 3) have
Ki values in the range 3−33 μM. Type I binding compounds
without nitrogen (series 1) have up to 4-fold lower Ki values
compared to type I binding compounds with sterically hindered
nitrogen (series 3) except compounds with tolyl group
(compound 4 vs compound 12), whose Ki values are close to
each other. Similarly, insertion of the second nitrogen in the
pyridine ring of the type II binding compounds increased the Ki
values from 3- to 37-fold (series 2 to series 4 and 5).
Saturation Kinetics. To determine the kinetic parameters,
we carried out a saturation kinetic assay. Five concentrations of
substrate ranging from 0.2KM to 5KM were used to determine
the Michaelis−Menten constant (KM) and turnover number
(kcat). The kinetic parameters KM, kcat, and in vitro metabolic
intrinsic clearance (V/K) are presented in Table 3. The kcat
values of type I binding compounds (series 1 and 3) are larger
than type II binding compounds in series 2 except compounds
6 and 7. The Kcat value of type II binding compound 6 is more
than 2-fold smaller than type I binding compound 2 of series 1,
but kcat value of compound 6 is 1.2 times larger than type I
binding compound 10 of series 3. Type II binding compound 7
has about 1.5 times higher kcat value than the type I binding
compounds 3 and 11. This indicates that the rate of
metabolism of type II binder is slower (with few exceptions)
than the type I binding counterpart when the enzyme is
saturated with substrate. Insertion of a second nitrogen in the
pyridine ring generally increased the turnover number of type II
binding compounds (series 2 to series 4 and series 5) with few
exceptions.
In vitro intrinsic clearance (V/K) is used in drug industry as
an important predictor for the metabolic stability of a drug in
vivo. V/K is a measure of the rate of clearance of a drug in
subsaturating concentration of the substrate. V/K values of the
type II binding compounds in series 2 were found to be 2- to
12-fold higher than electronically similar type I binding
compounds in series 3. The V/K values of the type I binding
compounds in series 1 were 1.3- to 4.7-fold lower than the type
II binding compounds in series 2. Overall, the type II binding
QCAs were metabolically less stable than the type I binding
283
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
Table 4. Comparison of O-Demethylation, N-Demethylation, Aromatic Hydroxylation, and Benzylic Hydroxylation in Type I
d
Binding Quinoline Carboxamide Analogues
a
b
c
Turnover number. Michalies−Menten constant. Ratio of product formation from functional group oxidation at ring A to aromatic hydroxylation
d
at ring B. Kinetic parameter values represent the mean of triplicate assays, and values after “ ” represent standard deviation.
Table 5. Comparison of O-Demethylation, N-Demethylation, Aromatic Hydroxylation, and Benzylic Hydroxylation in Type II
d
Binding Quinoline Carboxamide Analogues
a
b
c
Turnover number. Michalies−Menten constant. Ratio of product formation from functional group oxidation at ring A to aromatic hydroxylation
d
at ring A. Kinetic parameter values represent the mean of triplicate assays, and values after “ ” represent standard deviation.
284
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
Scheme 2. Metabolic Transformations by CYP3A4
counterparts. Insertion of nitrogen in type I binding
compounds (series 1 to series3) as well as type II binding
compounds (series 2 to series 4 and 5) decreased the intrinsic
clearance (Table 4).
reduction potential was considered to be a major contributor
to the metabolic stability12,15,16 of type II binding compounds
compared to type I binding compounds, and hence type II
binding compounds are assumed to be inhibitors of CYP
enzymes. Inhibition of CYP enzymes can lead to a high risk of
drug−drug interactions with potentially serious clinical
implications.24 Therefore, it is an important goal in drug
industry to get optimum metabolic stability and inhibitory
potential of the drug candidate during the drug development
and design. Recently we reported that the affinity of a
compound can be increased by more than 4200-fold just by
changing the position of nitrogen in QCA.21 Our earlier study
involving electrochemistry, surface plasmon resonance, and
saturation kinetics revealed the kinetic pathway for the
metabolism of type II binding QCA compounds via direct
reduction of heme-iron in type II binding mode.18 In the past
study we used QCA analogues with naphthalene moiety and
observed aromatic hydroxylated metabolites. In this study we
synthesized five series of compounds with four different
functional groups so that we can compare the four trans-
formations, namely, N-demethylation, O-demethylation, ben-
zylic hydroxylation, and aromatic hydroxylation. We coupled
the different binding motifs, namely, phenyl, pyridinyl,
pyrazinyl, pyrimidyl at the 2 position of quinoline to compare
the metabolic stability and regioselectivity of type II binding
QCA compounds to that of structurally similar type I binding
QCA compounds.
The library of compounds we chose to use in the in vitro
assay allows for the systematic comparison of kinetic
parameters between type I and type II binding compounds
and was based on the hypothesis that the sterically accessible
aromatic nitrogen is required for type II binding compounds.21
To accomplish this goal, we synthesized the QCA analogues
with 2-pyridin-4-yl (series 2), which can type II bind with
CYP3A4 (Figure 3). For a type I binding counterpart, we
synthesized 2-phenyl QCA analogues (series 1) and 2-pyridin-
2-yl QCA analogues (series 3). The series 3 compounds are
electronically similar to the type II binding series 2 compounds;
however, series 1 compounds are deprived of nitrogen in the
aromatic ring at position 2 of QCA. To investigate the effect of
insertion of a second nitrogen in pyridine ring, 2-pyrimid-4-yl
Comparison of O-Demethylation, N-Demethylation,
Aromatic Hydroxylation, and Benzylic Hydroxylation
between Type I and Type II Binding Compounds. The
kinetic parameters for O-demethylation, N-demethylation,
aromatic hydroxylation, and benzylic hydroxylation for type I
binding 2-phenyl QCA analogues (series 1) are presented in
Table 4 and those of type II binding 2-pyridin-4-yl QCA
analogues (series 2) are presented in Table 5. Functional group
oxidation as well as hydroxylation of the benzene ring at
position 2 of the 2-phenyl QCA (series 1) was observed except
in compound 2 with N,N-dimethyl functional group (Scheme
2). Metabolism in the 2-aza-heteroaromatic rings (pyridine,
pyrimidine, and pyrazine) of the QCA analogues (series 2, 3, 4,
5) was not detected within the detection limit of the
instrument. The rates of functional group oxidation were
compared in reference to the aromatic hydroxylation in ring B
of the same compound (see Table 4 for the ratio of the rate of
functional group oxidation to the rate of aromatic oxidation; the
bigger ratio indicates faster rate of metabolism for the
functional group) for type I binding compounds (series 1),
and the regioselectivity of metabolism was found to be in the
order of N-demethylation > O-demethylation ∼ benzylic
hydroxylation > aromatic hydroxylation. The order of rates of
functional group transformation for the type II binding
compounds (series 2) was determined in reference to the
aromatic hydroxylation in the ring A (see Table 5 for the ratio
of the rate of functional group oxidation to the rate of aromatic
oxidation), and the order of regioselectivity of metabolism was
found to be N-demethylation > benzylic hydroxylation > O-
demethylation ∼ aromatic hydroxylation.
DISCUSSION
■
Aromatic nitrogen can coordinate with the heme-iron of CYP
enzymes to maintain a low-spin state in type II binding mode.
The reduction potential of the type II bound heme-iron is
higher than the type I bound heme-iron.18 This higher
285
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
QCA analogues (series 4) and 2-pyrazin-2-yl QCA analogues
(series 5) were synthesized. As expected, ligands in series 1 and
series 3 produced type I binding difference spectra because of
the absence of accessible nitrogen to coordinate with the heme-
iron. The sterically hindered ortho-nitrogen of pyridine in series
3 analogues could not coordinate with heme-iron21 and
produced type I binding difference spectrum. Both meta- and
para-nitrogen of aza-heteroaromatic rings at position 2 of the
QCA analogues (series 2, 4, and 5) coordinated with heme-iron
to give the type II binding difference spectra.
mode and kcat is not perfect, it most likely reflects a slower rate
of reduction combined with differences in binding orientation.
For all the compounds in series 2 the reported off rate for type
II binding QCA analogue18 (0.006−0.03 s−1) is slower than the
kcat of type II binding compounds 0.05 −0.5 s−1 (Table 3).
Since kcat, the overall rate of the metabolism for all the steps in
the catalytic cycle, is faster than the debinding rate of type II
binding compounds, it is unlikely that the substrate debinds
before reduction. Hence, the type II binding QCA compounds
are expected to follow the direct reduction pathway as
described by Pearson et al.18
The affinity of the type II ligands is higher because of the
coordination of the electron rich nitrogen with the electron
deficient heme-iron.17 Replacing the 2-phenyl ring of the QCA
analogues with the 2-pyridin-4-yl ring increased the affinity
from 6- to 18-fold (series 1 to series 2), indicating that the type
II ligands contributed to the higher affinity with the heme-iron.
The 2-pyridin-2-yl QCA analogues (series 3) cannot coordinate
with the heme-iron because of steric hindrance but has the
same electronic core as type II ligands of series 2. The series 3
compounds have lower affinity (7- to 41-fold) with CYP3A4
compared to series 2 compounds, demonstrating that direct
coordination to heme-iron by an aromatic nitrogen is
contributing to the higher affinity. QCA analogues with
sterically hindered nitrogen in an aromatic ring (series 3),
which cannot bind in type II binding mode, have lower affinity
than the QCA analogues without nitrogen in the aromatic ring
(series 1). Both series 1 and series 3 compounds are type I
binding compounds, but series 3 has a nitrogen in an aromatic
ring which did not contribute to type II binding. Series 3
compounds had about 4-fold lower affinity than series 1
compounds except compound 4 and compound 12, which
showed about same affinity. Affinity of type II binding
compounds with two nitrogens in an aromatic ring is lower
than that of the type II binding compounds with one nitrogen
in aromatic ring. Affinity of the type II ligands containing two
nitrogens in a ring (2-pyrazin-2-yl and 2-pyrimidin-4-yl) of
series 4 and series 5 was lower than that of series 2 which has
only one nitrogen, possibly because of (a) a decrease in the
electron density of the coordinating nitrogen25 by electro-
negativity effect of the inserted second nitrogen and/or (b) an
unfavorable interaction of the second nitrogen with amino acids
in the active site of the enzyme and/or (c) an increased water
solubility of the substrates on inserting a nitrogen which can
change the partition equilibrium between polar aqueous phase
and lipophilic active site. These possibilities are under
investigation in our lab.
It has been speculated that the higher reduction potential of
the type II ligand−enzyme complex is a reason16 for higher
inhibitory potential of the type II binding compounds.
However, the electrochemical measurement of the reduction
potential in type I and type II ligands showed only a 2-fold
difference18 in the rate of reduction, indicating the possibility
for the metabolism of the type II ligands. We observed that the
kcat of the type II binding compounds in series 2 is generally
lower or comparable to the kcat of the type I binding
compounds in series 1 and series 3, in agreement with the
lower reduction rate for the type II ligand−enzyme complex
compared to the type I ligand−enzyme complex. In series 2
(which are all type II bindering compounds) 5 has a lower kcat
than compounds 1 and 9, 8 has a lower kcat than compounds 4
and 12, while 6 has a lower kcat than compounds 2 but not
compound 10, and compound 7 has about the same kcat as
compounds 3 and 11. While the agreement between binding
The in vitro intrinsic metabolic clearance (V/K) is the
determinant of the metabolic stability of the compounds in
subsaturating concentration of the enzyme. It is worth noting
that clinical drugs are normally administered at subsaturating
concentration of substrate. The V/K values for the series 2
compounds are up to 3-fold higher than the series 1
compounds and up to 12-fold higher than the series 3
compounds, indicating that type II binding compounds get
metabolized more quickly in subsaturating concentration of the
substrate, in agreement with previously reported results.18 The
intrinsic clearance of the type II binding compounds with
multiple nitrogens in an aromatic ring is lower than the single
nitrogen containing counterpart, indicating that tightly bound
type II ligands get metabolized more quickly than the loosely
bound type II ligands. In spite of higher reduction potential and
lower kcat, the type II binding QCA analogues get metabolized
more quickly in subsaturating concentration because of their
higher affinity for the enzyme.
We did not detect aza-aromatic ring hydroxylation for any of
the four series of heteroaromatic rings at position 2 of QCA
analogues (Figure 3) within our instrument’s detection limit.
Electronegative nitrogen in an aza-aromatic ring can reduce
electron density of carbons of the aromatic ring. The oxy-iron
species of the CYP enzyme is electrophilic in nature; hence, the
decrease in electron density in the aromatic ring is likely the
reason for the metabolic stability of the aza-aromatic ring. Thus,
incorporation of nitrogen in the aromatic ring results in
branching to alternative sites of metabolism.
To investigate if regioselectivity is altered between type I and
type II ligands for different types of metabolic transformations,
we compared QCA analogues that could undergo O-
demethylation, N-demethylation, aromatic hydroxylation, and
benzylic hydroxylation. The comparison of the different
transformations was based on the ratio of the functional
group transformation of interest to aromatic hydroxylation in
the same compound. In type I binding compounds of series 1,
we observed aromatic hydroxylation in the 2-phenyl ring of
QCA; hence, for the comparison the ratio of functional group
transformation to that of the 2-phenyl ring hydroxylation
(Table 4) was considered. In type II binding series 2
compounds, the pyridine ring was not metabolized but we
observed metabolism in the benzene ring A (Table 5) which
also contains a functional group of interest for the trans-
formation; hence, the regioselectivity was determined with
reference to the aromatic hydroxylation in ring A. For the
compounds undergoing N-demethylation, we did not observed
aromatic hydroxylation (Scheme 2), indicating N-demethyla-
tion outcompetes the aromatic hydroxylation in both type I and
type II ligands. In the case of type I binding compounds (series
1), O-demethylation and benzylic hydroxylation have about
same relative rate of metabolism and the aromatic hydrox-
ylation is the slowest of the transformations under the
286
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
Binding Spectra. Difference binding spectra were collected on a
double-beam Olis upgraded Aminco DW-2000 spectrometer (Olis
Inc., Bogart, GA). A baseline was recorded with protein in both
reference and sample cuvettes before collecting the difference spectra.
To correct the absorbance from the QCA substrates, two chambered
cuvettes (NSG Precision Cells Inc., Farmingdale, NY) were used.
Purified CYP3A4 (0.5 μM) was added to the sample and reference
cuvettes, and an equal volume of 100 mM potassium phosphate buffer,
pH 7.4, was added in the other chamber of the both cuvettes. Substrate
(20 μM final concentration) was added in the protein chamber of the
sample cuvette and the buffer chamber of the reference cuvette. To
correct for dilution, an equal volume of solvent used to add the
substrate was also added in the protein chamber of the reference
cuvette and the buffer chamber of the sample cuvette. Absorbance was
recorded from 350 to 500 nm to determine the mode of binding.
Inhibition Study. Testosterone was used as a CYP3A4 substrate,
and Ki values were determined from the formation rate of 6-β-
hydroxytestosterone. Four different final concentrations (30, 60, 90,
and 120 μM) of testosterone in 100 mM potassium phosphate buffer
at pH 7.4 were used. Three different inhibitor concentrations were
added to each different concentration of testosterone, and one set was
incubated without inhibitor. CYP3A4 baculosomes (4 pmol) were
added to the incubation mixtures containing substrate and inhibitor.
After preincubation of the mixture for 5 min at 37 °C, NADPH was
added to a final concentration of 1 mM to initiate the reaction. The
final volume of all the incubations was 0.5 mL. At the end of the 10
min incubation the reactions were quenched by adding 200 μL of
acetonitrile containing internal standard phenacetin (25 μM). The
enzyme was removed by centrifugation (Eppendorf centrifuge 5415D)
for 10 min at 16000g, and the formation of product was analyzed by
HPLC with UV detection. The HPLC system used reversed phase
chromatography (Alltech, Altima C18 5 μm, 150 mm length, 3.2 mm
i.d.), beginning with 95% mobile phase A (0.1% trifluoroacetic acid in
water) and 5% mobile phase B (0.1% trifluoroacetic acid in
acetonitrile) with linear gradient to 65% mobile phase B over 25
min. All the compounds were found to be competitive inhibitors based
on graphical analysis using reciprocal plots. Each compound’s Ki was
determined based on the best fit to the Michaelis−Menten equation
for competitive inhibition using GraphPad Prism 4 (San Diego, CA)
graphing software.
Saturation Kinetics. At least five final concentrations (0.2KM to
5KM) of substrates were used to determine the kinetic parameters for
the metabolism by CYP3A4. CYP3A4 baculosomes (4 pmol) were
added to the incubation mixtures containing a substrate in 100 mM
potassium phosphate buffer at pH 7.4 followed by preincubation for 5
min at 37 °C. NADPH was added to a final concentration of 1 mM to
initiate the reaction. The final volume of all incubations was 0.5 mL.
Reactions were quenched after 10 min by adding 200 μL of
acetonitrile containing phenacetin (25 μM) as an internal standard.
The enzyme was removed by centrifugation (Eppendorf centrifuge
5415D) for 10 min at 16000g, and metabolites were analyzed by LC/
MS (Thermo Quest HPLC series coupled with Thermo Finnigan
LCQ Advantage). The HPLC system used reverse phase chromatog-
raphy (Agilent, Eclipse plus C18 5 μm, 150 mm length, 3.2 mm i.d.)
beginning with 85% of mobile phase A (0.1% acetic acid in water) and
15% mobile phase B (0.1% acetic acid in acetonitrile) with a linear
gradient to 85% of mobile phase B over 25 min. The m/z was
monitored for the internal standard and metabolites of the substrates.
The kinetic constants were determined based on the best fit to the
Michaelis−Menten equation using GraphPad Prism 4 (San Diego,
CA) graphing software.
investigation. Rates obtained for each transformation were
similar to that of the reported activation energy for the
transformations using in silico methods.26−29 In the case of type
II binding compounds, the trend of transformation is same as
that of the type I binding compounds except that the benzylic
hydroxylation in compound 8 is about 3-fold faster compared
to O-demethylation of compound 5. In all the transformations,
the KM values stay the same within experimental error for the
two metabolites observed, indicating the common binding
site30 for both transformations by the CYP3A4.
Thus, for this series of compounds, metabolic stability is not
increased by nitrogen substitution in a single aromatic ring,
since this substitution results in branching to a different
metabolite. To decrease metabolism and still have a strong
inhibitor, a second alteration would be required at the site of
the branched metabolism. For example, benzylic and aromatic
hydroxylation could be stopped by introducing fluorine at the
site of metabolism. Another alternative or complementary
strategy would be to introduce a nitrogen atom in into the “A
ring” (see Table 4), which would slow aromatic hydroxylation
on that ring. If a decrease in inhibitory potency is desired with
increased metabolic stability, then nitrogen substitution ortho
to the quinoline linkage seems to work best. Such a substitution
has the electronic effect of slowing metabolism, but since it
does not coordinate to the iron, these compounds have much
lower inhibitory potency.
In conclusion, a comparative study between type I and type
II binding compounds using a library of QCA analogues was
undertaken with compounds having the same structural core
but differing in number and position of nitrogen. The type II
binding compounds showed higher affinity than the type I
binding counterparts. We observed that the turnover numbers
(kcat) of the type II binding compounds were lower than the
type I binding compounds. More importantly we observed that
the in vitro intrinsic metabolic clearance, as measured by V/K,
for type II binding compounds was up to 12-fold higher than
the structurally similar type I binding counterparts. The
insertion of a second nitrogen in heteroaromatic ring of the
type II ligands decreased the affinity and the intrinsic clearance.
The order of the metabolic rates for four studied metabolic
transformations by CYP3A4 was found to be N-demethylation
> benzylic hydroxylation > O-demethylation > aromatic
hydroxylation in type both I and type II binding compounds.
EXPERIMENTAL SECTION
■
Materials and Instruments. Solvents and chemicals were
purchased from Aldrich (St. Louis, MO), Fisher Scientific (Pittsburgh,
PA), EMD (Gibbstown, NJ), and Mallinckrodt Baker (Phillipsburg,
NJ) and used without purification. CYP3A4 baculosomes were a
1
generous gift from Tom Rushmore. H NMR spectra were obtained
using a 300 MHz spectrometer equipped with a quad-detection probe
(1H, 13C, 31P, and 19F). 1H-decoupled 13C NMR results were obtained
at 75 MHz. Mass spectrometry was performed on a ThermoQuest
Surveyor coupled to Thermo-Finnegan LCQ Advantage ESI-MS.
Absorbance spectra were collected on a double-beam Olis upgraded
Aminco DW-2000 spectrometer.
Metabolite Identification and Quantification. To identify the
position of metabolism, LC/MS/MS analysis was carried out using a
Thermo Quest HPLC coupled with Thermo Finnigan LCQ
Advantage. The HPLC system used reverse phase chromatography
(Agilent, Eclipse plus C18 5 μm, 150 mm length, 3.2 mm i.d.)
beginning with 85% of mobile phase A (0.1% acetic acid in water) and
15% mobile phase B (0.1% acetic acid in acetonitrile) with a linear
gradient to 85% of mobile phase B over 25 min. The fragments were
monitored in a range of 100−450 m/z for the metabolites using the
Enzyme Expression and Purification. The CYP3A4 NF14
construct was purified and expressed from Escherichia coli as described
by Roberts.31 The purity was >95% as determined by SDS−PAGE
analysis. The CYP3A4 concentration was determined by absorbance at
32
450 nm using an extinction coefficient of 91 mM−1 cm−1 for the
carbon monoxide bound protein. The purified CYP3A4 was stored in
100 mM potassium phosphate (pH 7.4) buffer with 20% glycerol at
−80 °C.
287
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
Scheme 3. Synthesis of the Metabolites
electrospray ionization (ESI) mode with collision energy of 40 V. The
synthesized standard metabolites have the same retention time and
fragmentation pattern as one of the metabolites obtained by CYP3A4
metabolism. The standard curve of the standard synthesized
metabolite with the internal standard was used to quantify the
metabolites formed.
Synthesis of Substrates. Substrates and metabolites were
synthesized as described by Peng et al.21 with modifications as
needed. The general synthetic methodology is described below in
terms of synthesis of compound 1. Synthesis of the other substrates
and characterization data are presented in Supporting Information.
Purity of the substrates was determined by reversed phase HPLC
system. The HPLC was recorded on an Agilent 1100 series instrument
equipped with an Alltech Altima C18 column (5 μm, 150 mm length,
3.2 mm i.d.) and a UV detector. The solvent system starts with 95%
mobile phase A (0.1% trifluoroacetic acid in water) and 5% mobile
phase B (0.1% trifluoroacetic acid in acetonitrile) with linear gradient
to 65% mobile phase B over 25 min. An indicated purity of >99%
indicates that no other peaks in the chromatogram occur. All
substrates reported in this publication were determined to have
purities of >95%.
N-(4-Methoxyphenyl)-2-phenylquinoline-4-carboxamide
(1). To a 100 mL round-bottom flask equipped with a water
condenser and a stir bar were combined potassium hydroxide (842 mg,
15 mmol) and ethanol (5 mL). The reaction mixture was stirred at 80
°C to dissolve the potassium hydroxide. Isatin (736 mg, 5 mmol) was
added to the reaction mixture followed by dropwise addition of
acetophenone (661 mg, 5.5 mmol). The mixture was refluxed at 80 °C
for 2 days. Then the solvent was evaporated using a rotary evaporator
and the residue was dissolved in 50 mL of water. The aqueous phase
was neutralized with dropwise addition of 1 N HCl to pH 6. The
resulting solid was collected by vacuum filtration to get crude 2-
phenylquinoline-4-carboxylic acid. The crude 2-phenylquinoline-4-
carboxylic acid was used directly to synthesize acyl chloride. To the 50
mL round-bottom flask equipped with a water condenser and a stir bar
were mixed the crude acid obtained above and 5 mL of neat thionyl
chloride, and the mixture was refluxed at 80 °C. After 2 h excess
thionyl chloride was evaporated using a stream of argon to get 2-
phenylquinoline-4-carbonyl chloride. The resulting solid was dissolved
in 5:1 mixture of dichloromethane/triethylamine (10 mL) in a 50 mL
round-bottom flask, and p-anisidine (677 mg, 5.5 mmol) was added.
The reaction mixture was stirred at room temperature. After 2 h the
solvent was evaporated using a rotary evaporator and the crude
product was purified using flash chromatography (30 g silica gel 60,
0.063−0.200 mm) with 10% ethyl acetate in dichloromethane. The
product obtained was crystallized in dichloromethane/ethanol to get
pure N-(4-methoxyphenyl)-2-phenylquinoline-4-carboxamide in
1
38.8% overall isolated yield. H NMR (CDCl3) δ 3.85 (s, 3H), 6.96
(dt, J = 9.2, 2.2 Hz, 2H), 7.45−7.5 (m, 4H), 7.64−7.75 (m, 3H), 7.82
(s, 1H), 8.03−8.15 (m, 5H); 13C NMR (CDCl3) δ 55.79, 114.61,
116.54, 122.18, 123.34, 125.17, 127.62, 127.69, 129.18, 130.05, 130.12,
130.59, 130.94, 138.65,143.09, 148.78, 156.90, 157.21, 165.63; ESI-MS
[M + H]+ = 355.4. Purity, ≥99%.
Synthesis of Metabolites. Metabolites were synthesized as
shown in Scheme 3.
O-Demethylation of the Substrate. The O-demethylation of
the substrate was carried out as described elsewhere33 with
modification as needed (Scheme 3a). The general synthetic method-
ology for O-demethylation is described below in terms of synthesis of
1M. Synthesis of the other metabolites and characterization data are
presented in Supporting Information.
N-(4-Hydroxyphenyl)-2-phenylquinoline-4-carboxamide
(1M). To the 10 mL solution of 1 (127 mg, 0.36 mmol) in CH2Cl2,
boron tribromide (1 M in CH2Cl2, 1.1 mL, 1.1 mmol) was added over
2 min with stirring. The reaction mixture was stirred for 45 min at
room temperature. The reaction was quenched by saturated sodium
hydrogen carbonate, and the mixture was adjusted to pH 8 and was
extracted with CH2Cl2 (3× 30 mL). The combined organic phase was
washed with water. The solvent was removed under reduced pressure,
and the residue was flash chromatographed using 10% EtOAc/CH2Cl2
followed by 30% MeOH/CH2Cl2. The isolated overall yield was 83%.
1H NMR (DMSO) δ 6.77 (dt, J = 8.8, 2.2 Hz, 2H), 7.52−7.68 (m,
6H), 7.83 (ddd, J = 7.7, 7.0, 1.3 Hz, 1H), 8.13 (m, 1H), 8.29−8.37 (m,
3H), 9.35 (s, 1H), 10.57 (s, 1H); 13C NMR (DMSO) δ 115.85,
117.44, 122.41, 124.02, 125.90, 128.03, 129.62, 130.29, 130.63, 130.96,
131.19, 138.87, 144.00, 148.60, 154.70, 156.50, 165.35; ESI-MS [M +
H]+ = 341.4.
288
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
N-Demethylation. A few methods were reported in literature for
N-demethylation. We attempted N-demethylation using N-iodosucci-
namide34 and a selective N-demethylation method using a column
chromatography-like setup by Rosenau et al.,35 but they did not work
in our hands. The N-demethylation method developed by Acosta et
al.36 using calcium oxide and iodine in the presence of methanol
worked for all five series of our compounds (Scheme 3b).
The general synthetic methodology for N-demethylation is
described below in terms of synthesis of 2M. Syntheses of the other
metabolites and characterization data are presented in Supporting
Information.
were mixed. The mixture was refluxed at 80 °C for 2 h. After 2 h,
excess thionyl chloride was evaporated using a stream of argon to get
2-phenylquinoline-4-carbonyl chloride. The resulting solid was
dissolved in a 5:1 mixture of dichloromethane/triethylamine (5 mL),
and 4-((tert-butyldimethylsilyloxy)methyl)aniline (473 mg, 2 mmol)
in 1 mL of dichloromethane was added. The reaction mixture was
stirred at room temperature for 2 h. Then the solvent was evaporated
using a rotary evaporator and the crude product was purified by flash
chromatography (30 g silica gel 60, 0.063−0.200 mm) using 25% ethyl
acetate in hexanes to get 701 mg of N-(4-((tert-butyldimethylsilyloxy)-
methyl)phenyl)-2-phenylquinoline-4-carboxamide with 75% isolated
1
overall yield in four steps. H NMR (CDCl3) δ 0.00 (s, 6H), 0.83 (s,
N-(4-(Methylamino)phenyl)-2-phenylquinoline-4-carboxa-
mide (2M). To an ice chilled mixture of 2 (50 mg, 0.14 mmol) and
CaO (61 mg, 1.1 mmol) in tetrahydrofuran (1.6 mL) and methanol
(1.2 mL) was added iodine (69 mg, 0.27 mmol). The mixture was
stirred at 0 °C until the reaction was completed. The reaction was
monitored by TLC, and in 4 h all the reactant was consumed. The
reaction mixture was filtered, and the filtrate was sequentially washed
with 15% sodium thiosulfate solution, water, and brine solution. The
organic phase was dried over magnesium sulfate, and solvent was
evaporated in a rotary evaporator. The product was separated by flash
chromatography using 4% acetone in dichloromethane. The product
(2M) was crystallized in CH2Cl2/hexanes system to get 25 mg of pure
product with an isolated yield of 55%. 1H NMR (DMSO) δ 2.67 (d, J
= 5.0 Hz, 3H), 5.58 (q, J = 5.3 Hz, 1H), 6.54 (d, J = 8.9 Hz, 2H),
7.51−7.59, (m, 5H), 7.64 (t, J = 8.2 Hz, 1H), 7.83 (t, J = 8.2 Hz, 1H),
8.16 (t, J = 9.1 Hz, 2H), 8.27 (s, 1H), 8.35 (d, J = 6.5 Hz, 2H), 10.43
(s, 1H); ESI-MS [M + H]+ = 354.3.
Benzylic Hydroxylation. Attempts to carry out direct benzylic
hydroxylation were not successful. We attempted to convert the benzyl
group into benzyl halide using N-bromosuccinimide/benzoyl peroxide
system as performed by Dauben and McCoy,37 using iodine/hydrogen
peroxide system as performed by Barluenga et al.38 and using
bromine/dimethyl sulfoxide system as performed by Ghaffarzadeh et
al.39 We also tried to convert the benzyl group into benzaldehyde
using ferric chloride/tert-butyl hydroperoxide system as performed by
Nakanishi and Bolm40 and using 2-iodoxybenzoic acid as performed by
Nicolaou et al.41 As all these attempts did not work in our hands, we
synthesized the benzylic metabolite by protecting 4-aminobenzyl
alcohol and reacting protected benzylic alcohol with appropriate acid
chlorides to get the protected benzylic alcohols which on deprotection
yielded the desired benzylic hydroxylated metabolites (Scheme 3c).
The general synthetic methodology of benzylic hydroxylated
metabolites is described below in terms of the synthesis of 4M.
Synthesis of the other metabolites and characterization data are
presented in Supporting Information.
4-((tert-butyldimethylsilyloxy)methyl)aniline. To a solution
of 4-aminobenzyl alcohol (1.5 g, 12.2 mmol) in dichloromethane
(20 mL) were added tert-butyldimethylsilyl chloride (TBDMSCl, 1.8
g, 12.2 mol) and imidazole (0.91 g, 13.4 mmol). After being stirred for
1 h at room temperature, the reaction mixture was diluted with brine
solution and then extracted with ethyl acetate. The organic phase was
dried over sodium sulfate and filtered. The solvent was evaporated in a
rotary evaporator to get the oil (2.8 g, 11.8 mmol) with 97% yield. 1H
NMR (CDCl3) δ 0.01 (s, 6H), 0.85 (s, 9H), 3.58 (s, 2H), 4.55 (s, 2H),
6.56 (d, J = 8.6 Hz, 2H), 7.03 (d, J = 8.5 Hz, 2H).
N-(4-(Hydroxymethyl)phenyl)-2-phenylquinoline-4-carbox-
amide (4M). To a 100 mL round-bottom flask equipped with a water
condenser and a stir bar, potassium hydroxide (337 mg, 6 mmol) in
ethanol (5 mL) was dissolved at 80 °C. Isatin (294 mg, 2 mmol) was
added to the reaction mixture followed by dropwise addition of
acetophenone (264 mg, 2.2 mmol). The mixture was refluxed at 80 °C
for 3 days. Then the solvent was evaporated out using a rotary
evaporator and the residue was dissolved in 50 mL of water. The
aqueous phase was neutralized with dropwise addition of 1 N HCl to
pH 6. The resulting solid was collected by vacuum filtration to get
crude 2-phenylquinoline-4-carboxylic acid. The crude 2-phenylquino-
line-4-carboxylic acid was used directly to get acyl chloride. To a 50
mL round-bottom flask equipped with a water condenser and a stir
bar, the crude acid obtained above and 5 mL of neat thionyl chloride
9H), 4.64 (s, 2H), 7.26 (d, J = 8.5 Hz, 2H), 7.31−7.42 (m, 4H), 7.58−
7.64 (m, 3H), 7.77 (s, 1H), 7.94−8.06 (m, 5H); ESI-MS [M + H]+ =
369.4. N-(4-((tert-Butyldimethylsilyloxy)methyl)phenyl)-2-phenylqui-
noline-4-carboxamide (701 mg) was dissolved in 20 mL of
tetrahydrofuran, and 2 mL of 1 M tetrabutylammonium fluoride
solution in tetrahydrofuran was added. The reaction mixture was
stirred, and the progress of the reaction was monitored by TLC. After
the completion of reaction in 4 h, the solvent was evaporated and flash
chromatography (30 g silica gel 60, 0.063−0.200 mm) was carried out
to purify the product using hexane, dichloromethane, and methanol in
9:10:1 ratio. The product was crystallized in methanol/dichloro-
1
methane to get the title compound with 67% isolated yield. H NMR
(DMSO) δ 4.48 (d, J = 5.6 Hz, 2H), 5.18 (t, J = 5.6 Hz, 1H), 7.33 (d, J
= 8.2 Hz, 2H), 7.52−7.61 (m, 3H), 7.68 (t, J = 8.2 Hz, 1H), 7.75 (d, J
= 8.2 Hz, 2H), 7.84 (t, J = 8.2 Hz, 1H), 8.15 (d, J = 8.5 Hz, 2H),
8.33−8.37 (m, 3H), 10.79 (s, 1H); 13C NMR (DMSO) δ 65.58,
119.81, 122.76, 126.23, 128.11, 129.96, 130.36, 131.95, 132.65, 132.99,
133.34, 140.42, 141.14, 141.37, 146.09, 150.92, 158.84, 168.19, 225.16;
ESI-MS [M + H]+ = 355.4.
ASSOCIATED CONTENT
* Supporting Information
■
S
Representative difference binding spectra for each series and
synthesis and characterization data of substrates/metabolites.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Phone: (509) 335-5983. Fax: (509) 335-8867. E-mail: jpj@
wsu.edu.
■
ACKNOWLEDGMENTS
■
This work was supported by NIH Grant GM84546. Special
thanks are extended to John T. Barr for valuable discussions
during experiments and manuscript development.
ABBREVIATIONS USED
■
QCA, quinoline carboxamide; CYP, cytochrome P450;
CYP3A4, cytochrome P450 3A4; V/K, in vitro metabolic
intrinsic clearance; KM, Michaelis−Menten constant; kcat,
turnover number
REFERENCES
■
(1) Ortiz de Montellano, P. R. Cytochrome P450: Structure,
Mechanism, and Biochemistry, 3rd ed.; Kluwer Academic/Plenum
Publishers: New York, 2005.
(2) Shimada, T.; Yamazaki, H.; Mimura, M.; Inui, Y.; Guengerich, F.
P. Interindividual variations in human liver cytochrome P-450 enzymes
involved in the oxidation of drugs, carcinogens and toxic chemicals:
studies with liver icrosomes of 30 Japanese and 30 Caucasians. J.
Pharmacol. Exp. Ther. 1994, 270, 414−23.
(3) Yamazaki, H.; Nakajima, M.; Nakamura, M.; Asahi, S.; Shimada,
N.; Gillam, E. M.; Guengerich, F. P.; Shimada, T.; Yokoi, T.
289
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290

Journal of Medicinal Chemistry
Article
Enhancement of cytochrome P-450 3A4 catalytic activities by
cytochrome b(5) in bacterial membranes. Drug Metab. Dispos. 1999,
27, 999−1004.
3A4 in nanodiscs by localized surface plasmon resonance
spectroscopy. Anal. Chem. 2009, 81, 3754−3759.
(24) Fleishaker, J. C.; Herman, B. D.; Carel, B. J.; Azie, N. E.
Interaction between ketoconazole and almotriptan in healthy
volunteers. J. Clin. Pharmacol. 2003, 43, 423−427.
(4) Evans, W. E.; Relling, M. V. Pharmacogenomics: translating
functional genomics into rational therapeutics. Science 1999, 286,
487−491.
(25) Vinh, T. K.; Ahmadi, M.; Delgado, P. O.; Perez, S. F.; Walters,
H. M.; Smith, H. J.; Nicholls, P. J.; Simons, C. 1-[(Benzofuran-2-
yl)phenylmethyl]-triazoles and -tetrazoles: potent competitive
inhibitors of aromatase. Bioorg. Med. Chem. Lett. 1999, 9, 2105−2108.
(26) Korzekwa, K. R.; Jones, J. P.; Gillette, J. R. Theoretical studies
on cytochrome P-450 mediated hydroxylation: a predictive model for
hydrogen atom abstractions. J. Am. Chem. Soc. 1990, 112, 7042−7046.
(27) Korzekwa, K. R.; Jones, J. P. Predicting the cytochrome P450
mediated metabolism of xenobiotics. [Review]. Pharmacogenetics 1993,
3, 1−18.
(28) Jones, J. P.; Mysinger, M.; Korzekwa, K. R. Computational
models for cytochrome P450: a predictive electronic model for
aromatic oxidation and hydrogen atom abstraction. Drug Metab.
Dispos. 2002, 30, 7−12.
(29) Olsen, L.; Rydberg, P.; Rod, T. H.; Ryde, U. Prediction of
activation energies for hydrogen abstraction by cytochrome p450. J.
Med. Chem. 2006, 49, 6489−6499.
(30) Porter, W. R.; Branchflower, R. V.; Trager, W. F. A kinetic
method for the determination of multiple forms of microsomal
cytochrome P-450. Biochem. Pharmacol. 1977, 26, 549−550.
(31) Roberts, A. G.; Campbell, A. P.; Atkins, W. M. The
thermodynamic landscape of testosterone binding to cytochrome P
450 3A4: ligand binding and spin state equilibria. Biochemistry 2005,
44, 1353−1366.
(32) Omura, T.; Sato, R. The carbon monoxide binding pigment of
liver microsomes. I. Evidence for its hemoprotein nature. J. Biol. Chem.
1964, 239, 2370−2378.
(33) McOmie, J. F. W.; Warrs, M. L.; West, D. E. Demethylation of
aryl methyl ethers by boron tribromide. Tetrahedron 1968, 24, 2289−
2292.
(34) Stenmark, H. G.; Brazzale, A.; Ma, Z. Biomimetic synthesis of
macrolide/ketolide metabolites through a selective N-demethylation
reaction. J. Org. Chem. 2000, 65, 3875−3876.
(35) Rosenau, T.; Hofinger, A.; Potthast, A.; Kosma, P. A general,
selective, high-yield N-demethylation procedure for tertiary amines by
solid reagents in a convenient column chromatography-like setup. Org.
Lett. 2004, 6, 541−544.
(36) Acosta, K.; Cessac, J. W.; Rao, P. N.; Kim, H. K. Oxidative
demethylation of 4-substituted N,N-dimethylanilines with iodine and
calcium oxide in the presence of methanol. Chem. Commun. 1994,
1985−1986.
(37) Dauben, H. J.; McCoy, L. L. N-Bromosuccinimide. III.
Stereochemical course of benzylic bromination. J. Am. Chem. Soc.
1959, 81, 5404−5409.
(38) Barluenga, J.; Campos-Gomez, E.; Rodriguez, D.; Gonzalez-
Bobes, F.; Gonzalez, J. M. New iodination reactions of saturated
hydrocarbons. Angew. Chem., Int. Ed. 2005, 44, 5851−5854.
(39) Ghaffarzadeh, M.; Bolourtchiana, M.; Gholamhossenia, M.;
Mohsenzadeha, F. Synthesis of arylaldehydes: Br2/DMSO catalytic
system for the chemoselective oxidation of methylarenes. Appl. Catal.
A 2007, 333, 131−135.
(40) Nakanishi, M.; Bolm, C. Iron-catalyzed benzylic oxidation with
aqueous tert-butyl hydroperoxide. Adv. Synth. Catal. 2007, 349, 861−
864.
(5) Li, A. P.; Kaminski, D. L.; Rasmussen, A. Substrates of human
hepatic cytochrome P450 3A4. Toxicology 1995, 104, 1−8.
(6) Dowers, T. S.; Rock, D. A.; Rock, D. A.; Perkins, B. N. S.; Jones,
J. P. An analysis of the regioselectivity of aromatic hydroxylation and
N-oxygenation by cytochrome P450 enzymes. Drug Metab. Dispos.
2004, 32, 328−32.
(7) Cooper, H. L.; Groves, J. T. Molecular probes of the mechanism
of cytochrome P450. Oxygen traps a substrate radical intermediate.
Arch. Biochem. Biophys. 2010, 507, 111−118.
(8) Das, A.; Sligar, S. G. Modulation of the cytochrome P450
reductase redox potential by the phospholipid bilayer. Biochemistry
2009, 48, 12104−12112.
(9) Jefcoate, C. R. Measurement of substrate and inhibitor binding to
microsomal cytochrome P-450 by optical-difference spectroscopy.
Methods Enzymol. 1978, 52.
(10) Schenkman, J. B. Studies on the nature of the type I and type II
spectral changes in liver microsomes. Biochemistry 1970, 9, 2081−
2091.
(11) Schenkman, J. B.; Remmer, H.; Estabrook, R. W. Spectral
studies of drug interaction with hepatic microsomal cytochrome. Mol.
Pharmacol. 1967, 3, 113−123.
(12) Ahlstrom, M. M.; Zamora, I. Characterization of type II ligands
in CYP2C9 and CYP3A4. J. Med. Chem. 2008, 51, 1755−1763.
(13) Ballard, S. A.; Lodola, A.; Tarbit, M. H. A comparative study of
1-substituted imidazole and 1,2,4-triazole antifungal compounds as
inhibitors of testosterone hydroxylations catalysed by mouse hepatic
microsomal cytochromes P-450. Biochem. Pharmacol. 1988, 37, 4643−
4651.
(14) Testa, B.; Jenner, P. Inhibitors of cytochrome P-450s and their
mechanism of action. Drug Metab. Rev. 1981, 12, 1−117.
(15) Chiba, M.; Tang, C.; Neway, W. E.; Williams, T. M.; Desolms,
S. J.; Dinsmore, C. J.; Wai, J. S.; Lin, J. H. P450 interaction with
farnesyl-protein transferase inhibitors. Metabolic stability, inhibitory
potency, and P450 binding spectra in human liver microsomes.
Biochem. Pharmacol. 2001, 62, 773−776.
(16) Chiba, M.; Jin, L.; Neway, W.; Vacca, J. P.; Tata, J. R.; Chapman,
K.; Lin, J. H. P450 interaction with HIV protease inhibitors:
relationship between metabolic stability, inhibitory potency, and
P450 binding spectra. Drug Metab. Dispos. 2001, 29, 1−3.
(17) Peng, C. C.; Pearson, J. T.; Rock, D. A.; Joswig-Jones, C. A.;
Jones, J. P. The effects of type II binding on metabolic stability and
binding affinity in cytochrome P450 CYP3A4. Arch. Biochem. Biophys.
2010, 497, 68−81.
(18) Pearson, J.; Dahal, U. P.; Rock, D.; Peng, C. C.; Schenk, J. O.;
Joswig-Jones, C.; Jones, J. P. The kinetic mechanism for cytochrome
P450 metabolism of type II binding compounds: evidence supporting
direct reduction. Arch. Biochem. Biophys. 2011, 511, 69−79.
(19) Peng, C. C.; Rushmore, T.; Crouch, G. J.; Jones, J. P. Modeling
and synthesis of novel tight-binding inhibitors of cytochrome P450
2C9. Bioorg. Med. Chem. 2008, 16, 4064−4074.
(20) Jones, J. P.; Joswig-Jones, C. A.; Hebner, M.; Chu, Y.; Koop, D.
R. The effects of nitrogen-heme-iron coordination on substrate
affinities for cytochrome P450 2E1. Chem.-Biol. Interact. 2011, 193,
50−56.
(41) Nicolaou, K. C.; Baran, P. S.; Zhong, Y. L. Selective oxidation at
carbon adjacent to aromatic systems with IBX. J. Am. Chem. Soc. 2001,
123, 3183−3185.
(21) Peng, C. C.; Cape, J. L.; Rushmore, T.; Crouch, G. J.; Jones, J. P.
Cytochrome P450 2C9 type II binding studies on quinoline-4-
carboxamide analogues. J. Med. Chem. 2008, 51, 8000−8011.
(22) Yano, J. K.; Denton, T. T.; Cerny, M. A.; Zhang, X.; Johnson, E.
F.; Cashman, J. R. Synthetic inhibitors of cytochrome P-450 2A6:
inhibitory activity, difference spectra, mechanism of inhibition, and
protein cocrystallization. J. Med. Chem. 2006, 49, 6987−7001.
(23) Das, A.; Zhao, J.; Schatz, G. C.; Sligar, S. G.; Van Duyne, R. P.
Screening of type I and II drug binding to human cytochrome P450-
290
dx.doi.org/10.1021/jm201207h | J. Med. Chem. 2012, 55, 280−290