Structure-Based Inhibitor Design for Evaluation of a CYP3A4 Pharmacophore Model

Article abstract of DOI:10.1021/acs.jmedchem.5b01146

Human cytochrome P450 3A4 (CYP3A4) is a key xenobiotic-metabolizing enzyme that oxidizes and clears the majority of drugs. CYP3A4 inhibition may lead to drug-drug interactions, toxicity, and other adverse effects but, in some cases, could be beneficial and enhance therapeutic efficiency of coadministered pharmaceuticals that are metabolized by CYP3A4. On the basis of our investigations of analogs of ritonavir, a potent CYP3A4 inactivator and pharmacoenhancer, we have built a pharmacophore model for a CYP3A4-specific inhibitor. This study is the first attempt to test this model using a set of rationally designed compounds. The functional and structural data presented here agree well with the proposed pharmacophore. In particular, we confirmed the importance of a flexible backbone, the H-bond donor/acceptor moiety, and aromaticity of the side group analogous to Phe-2 of ritonavir and demonstrated the leading role of hydrophobic interactions at the sites adjacent to the heme and phenylalanine cluster in the ligand binding process. The X-ray structures of CYP3A4 bound to the rationally designed inhibitors provide deeper insights into the mechanism of the CYP3A4-ligand interaction. Most importantly, two of our compounds (15a and 15b) that are less complex than ritonavir have comparable submicromolar affinity and inhibitory potency for CYP3A4 and, thus, could serve as templates for synthesis of second generation inhibitors for further evaluation and optimization of the pharmacophore model.

Full text of DOI:10.1021/acs.jmedchem.5b01146

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN
Article
Structure-based inhibitor design for
evaluation of a CYP3A4 pharmacophore model
Parminder Kaur, Richard Chamberlin, Thomas L. Poulos, and Irina F. Sevrioukova
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01146 • Publication Date (Web): 15 Sep 2015
Downloaded from http://pubs.acs.org on September 16, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Structure-based inhibitor design for evaluation of a CYP3A4 pharmacophore
model
Parminder Kaur†§, A. Richard Chamberlin†‡, Thomas L. Poulos†,‡,¶ and Irina F.
Sevrioukova¶*
¶
‡
†Departments of Pharmaceutical Sciences, Chemistry and Molecular Biology and
Biochemistry, University of California Irvine, California, 92697, United States
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
2
1
2
3
4
Abstract
5
6
7
8
Human cytochrome P450 3A4 (CYP3A4) is a key xenobioticꢀmetabolizing enzyme that
oxidizes and clears the majority of drugs. CYP3A4 inhibition may lead to drugꢀdrug
interactions, toxicity and other adverse effects but, in some cases, could be beneficial
and enhance therapeutic efficiency of coꢀadministered pharmaceuticals that are
metabolized by CYP3A4. Based on our investigations of analogs of ritonavir, a potent
CYP3A4 inactivator and pharmacoenhancer, we have built a pharmacophore model for
a CYP3A4ꢀspecific inhibitor. This study is the first attempt to test this model using a set
of rationally designed compounds. The functional and structural data presented here
agree well with the proposed pharmacophore. In particular, we confirmed the
importance of a flexible backbone, the Hꢀbond donor/acceptor moiety and aromaticity of
the side group analogous to Pheꢀ2 of ritonavir, and demonstrated the leading role of
hydrophobic interactions at the sites adjacent to the heme and phenylalanine cluster in
the ligand binding process. The Xꢀray structures of CYP3A4 bound to the rationally
designed inhibitors provide deeper insights into the mechanism of the CYP3A4ꢀligand
interaction. Most importantly, two of our compounds (15a and 15b) that are less
complex than ritonavir have comparable subꢀmicromolar affinity and inhibitory potency
for CYP3A4 and, thus, could serve as templates for synthesis of second generation
inhibitors for further evaluation and optimization of the pharmacophore model.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 3 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
3
1
2
3
4
Introduction
5
6
7
8
Cytochrome P450 3A4 (CYP3A4) is the most abundant and clinically relevant
xenobioticꢀmetabolizing enzyme in humans. CYP3A4 catalyzes diverse oxidative
reactions and is involved in the metabolism of endogenous steroids, the majority of
administered drugs and other foreign molecules, such as environmental pollutants,
carcinogens, pesticides and insecticides.1ꢀ8 CYP3A4 has a large and malleable active
site that accommodates a wide variety of molecules, some of which could inhibit or
enhance metabolism. In vivo, CYP3A4 inhibition has a greater impact on human health,
as it can lead to xenobioticꢀinduced toxicity and drugꢀdrug interactions, one of the
reasons for lateꢀstage clinical trial failures and withdrawal of marketed pharmaceuticals.
However, carefully controlled CYP3A4 inhibition can be beneficial and is
currently exploited in the treatment of HIV infection. To date, two CYP3A4 inhibitors,
ritonavir and cobicistat (Fig. 1), are marketed as pharmacoenhancers for the HIVꢀI
protease inhibitors that serve as substrates for CYP3A4.9ꢀ13 Both drugs were developed
based on chemical structure/activity relationships studies rather than the CYP3A4
crystal structure. Moreover, ritonavir was originally designed to inhibit the HIVꢀ1
protease,14 whereas its ability to potently inactivate CYP3A4 was coincidental and
discovered later.9 Cobicistat is a derivative of ritonavir that lacks the backbone hydroxyl
group, critical for binding to catalytic Asp25 and Asp225 of HIVꢀ1 protease, and has a
morpholine ring instead of the valine side group (Fig. 1). As a result, cobicistat is not an
antiꢀHIV protease inhibitor so its prolonged usage does not promote the development of
drugꢀresistant HIV strains. Also, this pharmaceutical has fewer side effects and better
physicoꢀchemical properties, yet is comparable but not superior to ritonavir in terms of
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
4
1
2
3
4
inhibitory potency for CYP3A4.12
5
6
7
8
There is no general agreement on the CYP3A4 inhibitory mechanism of ritonavir.
Several groups suggested that ritonavir is a mechanismꢀbased inactivator,15ꢀ19 whereas
others argued against the mechanismꢀbased CYP3A4 inhibition20 and concluded that
ritonavir acts as a competitive21 or a mixed competitiveꢀnoncompetitive CYP3A4
inactivator.22ꢀ24 Our data on highly purified recombinant CYP3A4 do not support the
mechanismꢀbased or competitive type of inhibition as a predominant inactivation
pathway and rather suggest that ritonavir inactivates CYP3A4 via strong thiazole
nitrogen coordination to both ferric and ferrous CYP3A4 which decreases the heme
redox potential and impedes electron transfer from the redox partner, cytochrome P450
reductase.25 This conclusion is well supported by the Xꢀray structures of CYP3A4
complexed with ritonavir and 10 ritonavir analogs.25ꢀ28 Most importantly, the
accumulated functional and structural data enabled us to identify several important
trends in the ligand binding process and define strategies for rational, structureꢀbased
inhibitor design.29 In particular, we found that: (i) strong binding through the hemeꢀ
ligating moiety is a prerequisite for potent CYP3A4 inhibition, and pyridine possesses
more favorable stereoelectronic properties for the heme ligation than any other chemical
groups tested; (ii) the flexible backbone enables a better fit into the CYP3A4 active site;
(iii) the side group at a Pheꢀ2 position (indicated in Fig. 1) is critically important because
it not only fills the hydrophobic cavity (P2 site) but also has a potential to stabilize the
proteinꢀligand complex via different types of interactions with the hemeꢀligating group
and nearby Arg105; (iv) occupation of the pocket hosting the Pheꢀ1 group (P1 site) is
also important and greatly improves inhibitory potency; (v) polar interactions with
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 5 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
5
1
2
3
4
5
6
7
8
9
Ser119 increase the ligand affinity and association rate, regardless of whether the
hydrogen bond with the ligand is established or not; and (vi) the terminal group
modulates the CYP3A4ꢀinhibitor complex affinity and stability which, in turn, correlate
with the ligandꢀinduced changes in the melting temperature and the amplitude of the
442 nm absorption of the ferrous ligandꢀbound form.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Previous 3Dꢀpharmacophores for drug metabolism prediction were built based on
P450 homology and quantitative structureꢀactivity relationship (QSAR) modeling
because the Xꢀray structure of CYP3A4 was not available at that time.30ꢀ33 In addition,
there is no structureꢀbased inhibitor pharmacophore that could assist in identification
and early elimination of potential CYP3A4 inactivators during development of drugs and
other chemicals relevant to public health. Having mapped the inhibitory determinants
within the CYP3A4 active site, we were able to build a pharmacophore model for a
CYP3A4ꢀspecific inhibitor (Fig. 2).29 Our strategy for the model evaluation was to
synthesize and analyze ritonavirꢀlike molecules using a general scaffold (Fig. 3) where
the pyridine ring serves as a hemeꢀligating moiety and the remainder of the inhibitor is
stitched together with units containing various side and terminal groups (R1ꢀR3). By
systematically changing R1ꢀR3 substituents, it can be probed how each chemical moiety
contributes to the binding affinity and inhibitory potency, and how CYP3A4 adapts to
changes in the ligand’s structure.
Here we present structural and functional results on the first set of commercially
available and rationally designed pyridineꢀcontaining compounds that differ in the
backbone length and R2 substituents (Fig. 4). With this initial set, we were able to
confirm the importance of a flexible backbone, aromaticity of the R2 group and the Hꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
6
1
2
3
4
5
6
7
8
9
bond donor/acceptor moiety (pharmacophoric determinants II, III and V), as well as the
leading role of hydrophobic interactions at the P1 and P2 sites in the ligand binding
process.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and Discussion
Inhibitor Design – As mentioned in the Introduction, both CYP3A4 inhibitors currently
marketed as pharmacoenhancers, ritonavir and cobicistat (Fig. 1), were developed
based on chemical structure/activity relationships studies rather than the CYP3A4
crystal structure. Ritonavir is a large peptidomimetic drug designed to inhibit an HIVꢀI
protease,9 whereas cobicistat was developed through ritonavir derivatization.13,
34
Synthesis of both compounds is a complex process requiring production of specific
stereoisomers. To evaluate our pharmacophore model, we decided to build CYP3A4
inhibitors from scratch using a general simple scaffold (Fig. 3) where the modules
containing the side and terminal groups (R1, R2 and R3) are linked to the hemeꢀligating
pyridine via a flexible backbone. Pyridine and aminoethylpyridine (Fig. 4) were obtained
from SigmaꢀAlrdrich. The rationale for designing compounds 4, 5, 11 and 15a-b was to
determine how elongation of the backbone, introduction of the Hꢀbond donor/acceptor
(peptide bond), and substitution of R2 affect affinity, inhibitory potency and the ligand
binding mode. Mass spectrometry and NMR data for compounds 4ꢀ15a/b are included
in the Supporting Information section.
Binding Affinity and Association Kinetics – Addition of pyridine, aminoethylpyridine and
compound 5 to CYP3A4 led to a decrease and a small red shift (~2 nm) in the Soret
ACS Paragon Plus Environment

Page 7 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
7
1
2
3
4
5
6
7
8
9
band (Fig. 5). Spectral dissociation constants (Ks) for pyridine and aminoethylpyridine
were in the millimolar range (Table 1), whereas 5 had two binding sites with Ks of 10 ꢀM
and 4.4 mM. Compounds 4, 11 and 15a-b, on the other hand, bind to a single site and
induce a notable shift in the Soret band, typical for type II ligands (Fig. 6). Introduction of
the hexane or phenyl side group increased the binding affinity by 10ꢀ and 100ꢀfold,
respectively, relative to 4. The phenylꢀtoꢀindole substitution led to a further decline in Ks
(by ~2ꢀfold; Table 1). Another important feature is a pronounced 442 nm absorption of
the ferrous 15aꢀ and 15bꢀbound forms of CYP3A4. The corresponding 442 nm peak
was considerably smaller for the CYP3A4ꢀ4/11 complexes and undetectable for the
pyridineꢀ, aminoethylpyridineꢀ and 5ꢀbound forms (Figs. 5 and 6). In our previous
studies,25ꢀ28 we found that the amplitude of the 442 nm band of the ferrous species
correlates with the inhibitory potency of the ligand, in that compounds whose binding led
to a higher 442 nm absorption peak were acting as more potent CYP3A4 inactivators
(e.g., ritonavir and pyridineꢀsubstituted desoxyritonavir (GS3; Fig. 1)). Therefore, a
significant but not complete conversion to the 442 nm absorbing species taking place
upon association of CYP3A4 with 15a and 15b was the first indication that these
compounds may have a comparable although not superior inhibitory potency relative to
GS3 and ritonavir.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The binding reactions of CYP3A4 with pyridine, aminoethylpyridine and 5 were
monophasic, with rate constants of 0.5ꢀ0.8 sꢀ1. In contrast, association with 4, 11 and
15a-b proceeded in two phases. The rate constant for the fast phase (kfast) was the
highest for 4 and the lowest for 15a (20 and 1.7 sꢀ1, respectively; Table 1). Notably, kfast
for 15a was close to that of ritonavir (1.4 sꢀ1)25 and nearly 5ꢀfold lower than the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
8
1
2
3
4
5
6
7
8
9
respective value for 15b and GS3. Together, the spectral and kinetic data suggest that
(i) the backbone elongation and addition of a side group complicate the binding
reaction, (ii) the presence of a bulky, hydrophobic R2 group increases the ligand affinity
but slows down the association rate, and (iii) indole as a R2 substituent promotes
complex formation with CYP3A4 to a higher degree than the hexane or phenyl rings.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Melting Temperature and Inhibitory Potency – None of the investigated compounds
increased the thermal stability of CYP3A4, as evidenced by the lack of feasible ligandꢀ
dependent changes in melting temperature (ꢁTm; Table 1). CYP3A4 Tm was slightly
increased upon complex formation with 15b but decreased in the presence of other
compounds, most notably for the 11ꢀ and 15aꢀbound forms (by 1ꢀ2°C). This is in
contrast to ritonavir and its analogs, whose ligation to CYP3A4 increases Tm by 2ꢀ
28
7°C.27,
Nevertheless, a common trend revealed by thermal denaturation is an
increase in thermostability of CYP3A4 upon association with ritonavirꢀlike compounds.
The heme ligation and occupation of the P1 and P2 cavities by hydrophobic moieties
likely stabilize CYP3A4 by decreasing conformational flexibility.
As expected, the inhibitory potency of the investigated compounds on the BFC
debenzylation activity of CYP3A4 correlated with the binding affinity (Table 1). The IC50
values were the highest for pyridine, aminoethylpyridine and 5 (1ꢀ4 mM), intermediate
for 4 and 11 (75 and 30 ꢀM, respectively), and the lowest for 15a and 15b (0.5 and 0.2
ꢀM, respectively). Importantly, IC50s for 15a and 15b were close to those for ritonavir
and GS3 (0.55 and 0.13 ꢁM, respectively).27 Thus, addition of only one aromatic side
ACS Paragon Plus Environment

Page 9 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
9
1
2
3
4
group and tertꢀbutyloxycarbonyl (Boc) as a terminal moiety transforms compound 5, a
5
6
very weak ligand, into a potent CYP3A4 inactivator.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Crystal Structures of the 15a- and 15b-bound CYP3A4 – We succeeded in coꢀ
crystallization of CYP3A4 with 4, 15a and 15b. Compound 11 dissociated from CYP3A4
during crystallization. Crystal structures of CYP3A4 ligated to 15a and 15b were
determined to 2.60 and 2.76 Å, respectively (Supporting Information Table 1S). Both
compounds bind to the active site and ligate to the heme iron through the pyridine
nitrogen (Fig. 7). The GS3ꢀbound model (PDB ID 4I4H) was used for comparative
analysis because this analog is most similar to the investigated compounds.
Superposition of the 15aꢀ, 15bꢀ and GS3ꢀbound structures reveals that 15a and 15b
rotate around the pyridine nitrogen by ca. 180° (Fig. 7C) to place the phenyl/indole rings
between the hemeꢀligating pyridine and Arg105 guanidinium group (P2 site), right where
the Pheꢀ2 side group of GS3 is positioned Most strikingly, the terminal Boc groups are
in a similar position and displace Phe304 and the Iꢀhelix to the same extent as Pheꢀ1 of
GS3 does (Fig. 7D). Since the phenyl and indole moieties are oriented differently than
GS3 Pheꢀ2, which is nearly parallel to the pyridine and Arg105 guanidine groups and
closer to the heme plane, the 15aꢀb conformations are less optimal for hydrophobic, πꢀπ
and πꢀcation interactions with the heme, pyridine and the Arg105 guanidine,
respectively. Another important dissimilarity that may decrease affinity of 15aꢀb is that
their binding mode disallows Hꢀbonding with Ser119 (Fig. 7C). Nevertheless, the
comparable Ks and IC50 values for 15b and GS3 imply that the larger and more electronꢀ
rich indole ring may partially compensate for the less favorable orientation by
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
10
1
2
3
4
5
6
establishing new or more extensive Van der Waals contacts with the neighboring
Arg105, Ala370, heme and pyridine.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Modification of Surface Residues in CYP3A4 for Improvement of Crystal Diffraction –
Wild type CYP3A4 willingly coꢀcrystallizes with 4 but, similar to other ligandꢀbound
forms, diffraction data of this complex did not exceed 2.5 Å. At this resolution, electron
density for 4 was discontinuous and the ligand association mode could not be
accurately determined. To overcome this problem, we decided to introduce surface
mutations that help form intermolecular contacts in the crystal lattice and improve Xꢀray
diffraction power without affecting the CYP3A4 ligand binding and metabolism. Using a
Surface
Entropy
Reduction
prediction
(SERp)
server
(http://services.mbi.ucla.edu/SER/),35 several surface peptides containing clusters of
bulky charged residues (Arg, Lys and/or Glu) were identified (Table 2). One cluster with
a high score, Lys282/Glu283/Glu285, is part of the surface loop that is disordered in all
crystal structures reported to date. We mutated this site and found a variant,
K282A/E285A, that was better expressed in E. coli, associated with 4 similarly to the
wild type, and produced crystals with lower mosaicity and higher diffraction power. This
mutant was used for determination of the CYP3A4ꢀ4 complex structure.
Crystal Structures of the CYP3A4-4 Complexes – Coꢀcrystals of CYP3A4 K282A/E285A
with 4 were grown under two different conditions that led to two distinct modes of ligand
binding. The first complex was crystallized in the presence of imidazole and contained
both compound 4 and imidazole in the active site (Fig. 8A). In this structure, determined
ACS Paragon Plus Environment

Page 11 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
11
1
2
3
4
5
6
7
8
9
to 1.93 Å, imidazole ligates to the heme iron and 4 binds nearby. Such a ligand
arrangement is stabilized by two hydrogen bonds formed between the carbonyl oxygen
and amide nitrogen of 4 and the imidazole nitrogen and the Ser119 hydroxyl group,
respectively. The pyridine moiety of 4 is adjacent but not embedded into the P1 pocket.
Through hydrophobic interactions with Phe108, Phe213, Phe215 and Phe340 (Fig. 8B),
the pyridine ring reinforces the phenylalanine cluster, a unique feature of CYP3A4.36
The aliphatic linker of 4 forms Van der Waals contacts with the Arg105 side chain,
whereas the Boc group is placed at the P2 site and forms hydrophobic and Van der
Waals interactions with the Ile369ꢀAla370 peptide and the Arg212 side chain. Thus,
when the space near the heme iron is occupied, the flexible compound 4 adopts a
conformation that allows Hꢀbonding to Ser119 and optimizes hydrophobic interactions
mediated by the Boc and pyridine moieties. It is not clear though why the pyridine of 4
does not insert into the P1 pocket as Boc of 15a-b (Fig. 7D) and Pheꢀ2 of GS3 and
other ritonavir analogs do (Fig. 8C).25, 27, 28 The observed orientation may be preferable
because it prevents steric clashing with Phe304 and the Iꢀhelix displacement.
Alternatively, insertion into the P1 cavity may require some force (e.g., a push), which
the flexible core of nonꢀligated 4 cannot provide.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The Xꢀray structure of the CYP3A4ꢀ4 complex formed in the absence of
imidazole was determined to 2.25 Å. In this structure, 4 is directly ligated to the heme
iron via the pyridine nitrogen and, similar to 15a-b, rotates by ~180° relative to GS3
(Fig. 9). The flexible linker bends to allow the end portion occupy the P2 site. This and
the lack of Hꢀbonds between 4 and the protein suggest once again that hydrophobic
interactions, especially at the P2 site, are dominant and define the ligand binding mode.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
12
1
2
3
4
5
6
Conclusions
7
8
9
The data obtained with the first set of rationally designed compounds support the
proposed pharmacophore model for a CYP3A4ꢀspecific inhibitor and emphasize the
importance of pharmacophoric determinants II, III and V (Fig. 3), as well as the
dominant role of hydrophobic forces at the P1 and P2 sites during the ligandꢀbinding
process. In particular, with this initial set, we were able to confirm that (i) the flexible
backbone enables ligands to better fit into the CYP3A4 active site and optimize
interactions mediated by the side and terminal moieties; (ii) the P2 site is occupied by
ligands first, and interactions at this site are dominant and define the overall ligand
binding mode; (iii) the P1 cavity is the second preferable site that ligands tend to
occupy, even if it leads to the Iꢀhelix distortion; and (iv) Hꢀbonding interactions with
Ser119 are established when possible to stabilize the CYP3A4ꢀligand complex.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
There were also new findings that provided additional insights into the ligand
binding process. First, pyridine and its shortꢀchain aliphatic derivative,
aminoethylpyridine (Fig. 4), were found to be very weak CYP3A4 inactivators (Table 1).
This means that ligation of the pyridine nitrogen to the heme iron is weak and
dissociable but becomes strong and virtually irreversible when the pyridine ring
becomes part of the larger ligands, especially those capable of occupying the P2 and
P1 sites. Second, based on the Ks and IC50 values for 11, 15a and 15b (Table 1) and
the inability of 11 to coꢀcrystallize with CYP3A4, we conclude that aromaticity rather
than hydrophobicity of the R2 substituent is the key feature that stabilizes the proteinꢀ
ligand complex and drastically improves affinity and inhibitory potency. Third, the
ACS Paragon Plus Environment

Page 13 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
13
1
2
3
4
5
6
7
8
9
structural data demonstrate that the P1 site can be occupied by the terminal group if R1
is absent. That none of the R1ꢀlacking compounds markedly affected CYP3A4 Tm
suggests, however, that Bocꢀmediated interactions at the P1 site are subꢀoptimal and
cannot provide the binding energy on the same scale as phenyl substituents of R1 do
(e.g., in ritonavir and GS3). Fourth, there was no correlation between ꢁTm and IC50 for
the investigated compounds (Table 1) but, as observed for other ritonavir analogs,27, 28
the inhibitory potency was proportional to the 442 nm absorption of the ferrous ligandꢀ
bound species. Thus, during ligand screening, the latter parameter may predict strong
CYP3A4 inactivators more reliably than ꢁTm. Finally, two of the rationally designed
compounds, 15a and 15b, are less structurally complex than GS3 and ritonavir but
inhibit CYP3A4 with a comparable, subꢀmicromolar potency. This raises the possibility
that simpler and more potent CYP3A4 inactivators than ritonavir can be designed
through 15aꢀb optimization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Procedures
Synthesis of analogs
Synthesis of compound 2. Diꢀtertꢀbutyloxycarbonyl, (Boc)2O (1.99g, 9.14mmol) was added
to a solution of 6ꢀaminohexanoic acid (1g, 7.61mmol) and NaOH (10% aqueous solution) in
1,4ꢀdioxane. The reaction mixture was then stirred overnight at room temperature. The
solvent was evaporated and the mixture was acidified using 1N HCl. The acidified mixture
was then extracted with dichloromethane (DCM). The organic layer was washed with brine,
dried over Na2SO4, filtered and concentrated. The crude material was then purified using
column chromatography to give the pure product 2 (1.57g, 89%).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
14
1
2
3
4
5
6
7
8
Synthesis of compounds 4 and 5 (Scheme 1). Compound 2 (1g, 4.23mmol) was added to
a solution of EDAC (1.67g, 10.80mmol) in dimethylformamide (DMF). The mixture was stirred
at room temperature for 30 min after which DIPEA (3.76 mL, 21.61mmol) and 3ꢀ
(aminomethyl)pyridine, 3 (0.70g, 6.34mmol) were added to the reaction mixture. The reaction
was allowed to stir for 16 h. On completion, the solvent was evaporated and the reaction
mixture was diluted with ethylacetate. The organic layer was then washed with saturated
NaHCO3, water and brine. The combined organic layer was dried over Na2SO4 and
concentrated to give the crude product which was then purified using column
chromatography (ethylacetate:methanol). The pure product 4 was obtained as white solid in
65% yield (0.91g). 1H NMR (CDCl3, 500 MHz) δ 8.53 (s, 2H), 7.64 (d, J = 7.82 Hz, 1H), 7.27
(t, J = 4.80 Hz, 1H), 5.95 (bs, 1H), 4.56 (bs, 1H), 4.46 (d, J = 5.91 Hz, 2H), 3.10 (q, J = 6.45
Hz, 2H), 2.23 (t, J = 7.45 Hz, 2H), 1.68 (quin, J = 7.5 Hz, 2H), 1.50ꢀ1.46 (m, 2H), 1.42 (s,
9H), 1.34 (m, 2H); 13C NMR (CDCl3, 125 MHz) δ 149.1, 148.8, 135.8, 123.7, 41.0, 36.4,
29.8, 28.4, 26.3, 25.2; HRMS m/z calculated for C17H27N3O3 [M+Na]+: 344.21, found: 344.2.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound 4 (0.6g, 1.86mmol) was treated with trifluoroacetic acid in DCM for 14 h at room
temperature in order to remove the Bocꢀgroup. The completion of the reaction was monitored
through TLC. The reaction solvent was evaporated and the crude product was passed
through celite to obtain the pure amine 5 (0.47g, 78yield%). 1HꢀNMR (500 MHz, CD3OD) δ 8.81
(1H, s), 8.79ꢀ8.78 (2H, d, J = 5.5 Hz), 8.53ꢀ8.51 (1H, d, J = 8.3 Hz), 8.06ꢀ8.03 (1H, m), 2.97ꢀ2.91 (2H,
m), 2.37ꢀ2.34 (2H, t, J = 7.4 Hz), 1.72ꢀ1.68 (4H, m), 1.48ꢀ1.39 (4H, m); 13C NMR (CD3OD, 125 MHz)
δ 175.2, 161.1, 144.5, 142.4, 139.2, 126.4, 39.7, 38.9, 34.3, 29.8, 26.6, 24.3. HRMS m/z calculated
for C12H19N3O [M+H]+: 222.23, found: 222.2.
ACS Paragon Plus Environment

Page 15 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
15
1
2
3
4
5
6
7
8
After optimizing different conditions and routes, we started synthesis of the first generation
analogs using (S)ꢀBocꢀphenylalaninol as shown in Schemes 2 and 3. (S)ꢀBocꢀphenylalaninol
or other corresponding amino alcohols (shown in the schemes) were converted to a tosylate
using pꢀtoluenesulphonyl chloride, which was then reacted with 3ꢀmercaptopropionic acid in
the presence of NaOH at 50oC to give the corresponding acid. The acid was then coupled to
the 2ꢀ(3ꢀpyridyl)ethylamine using the known usual EDAC coupling to give the corresponding
final product.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of compound 7. SꢀBocꢀphenylalaninol (2g, 7.96mmol) was dissolved in
methanol and treated with rhodium on charcoal (5%; 100mg) under H2 atmosphere
overnight to reduce the phenyl ring. The solution was then passed through celite plug, and
methanol was evaporated to give the pure product 7 in quantitative yield (2.01g, quant).
Synthesis of compound 8. SꢀBocꢀcyclohexylalaninol (2g, 7.78mmol) was dissolved in
DCM. To this solution, triethylamine (3.25mL, 23.34mmol) and pꢀtoluenesulphonyl chloride
(2.23g, 11.67mmol) were added slowly at 0oC. The reaction was then allowed to stir at room
temperature overnight. After reaction completion, DCM was evaporated and the crude
reaction mixture was purified using column chromatography. The pure product 8 was
obtained as white fluffy solid (2.3g, 71% yield). LCMS m/z calculated for C21H33NO5S
[M+Na]+: 434.21, found 434.23.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
16
1
2
3
4
5
6
7
8
9
Synthesis of compound 10. To solution of compound 8 (1g, 2.43mmol) in DMF, 3ꢀ
mercaptopropanoic acid, 9 (0.309g, 2.91mmol) was added. To this mixture, 1N NaOH (2
mL) was added and reaction was allowed to stir for 5 h at 50oC. The crude product 10 was
obtained by evaporating the solvent and was used for the next step without any further
purification due to instability of the product on silica column. LCMS m/z calculated for
C17H31NO4S [M+2Na]+: 368.20, found 368.17.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of compound 11. Compound 10 (0.1g, 0.3mmol) was added to a solution of
EDAC (0.12g, 0.72mmol) in DMF. The mixture was stirred at room temperature for 30 min
after which DIPEA (0.23mL, 1.45mmol) and 3ꢀ(aminomethyl)pyridine, 3 (0.042g, 0.43mmol)
were added. The reaction was allowed to stir for 18 h. On completion, the solvent was
evaporated and the reaction mixture was diluted with ethylacetate. The organic layer was
then washed with saturated NaHCO3, water and brine. The combined organic layers were
dried over Na2SO4 and concentrated to give the crude product which was then purified using
column chromatography (ethylacetate:methanol). The pure product 11 was obtained as
yellow oil in 22% yield (0.005g). 1HꢀNMR (400 MHz, CD2Cl2) δ 9.31ꢀ9.07 (1H, m), 8.56 (2H, bs),
7.70ꢀ7.57 (2H, m), 7.33 (1H, bs), 4.54 (1H, s), 3.74ꢀ3.20 (5H, m), 3.14ꢀ3.04 (6H, m), 2.43ꢀ2.26 (1H,
13
m), 1.67ꢀ1.51 (4H, m), 1.44ꢀ1.33 (2H, m), 1.26 (8H, s), 1.17ꢀ1.11 (3H, m), 1.04 (3H, m). C
NMR (100 MHz, CD2Cl2) δ 132.6, 130.5, 64.4, 32.0, 31.6, 29.9, 29.3, 27.0, 24.7, 23.4, 22.9, 22.5,
15.4, 15.2, 14.6, 13.8, 12.8, 12.0. LCMS m/z calculated for C23H37N3O3S [M+Na]+: 458.22,
found 458.21.
General procedure for synthesis of compounds 15a and 15b
ACS Paragon Plus Environment

Page 17 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
17
1
2
3
4
5
6
7
8
9
Synthesis of compound 13a. SꢀBocꢀphenylalaninol (1g, 3.98mmol) was dissolved in DCM.
To this solution, triethylamine and pꢀtoluenesulphonyl chloride were added slowly at 0oC.
The reaction was then allowed to stir at room temperature overnight. After reaction
completion, DCM was evaporated and the crude reaction mixture was purified using column
chromatography. The pure product 13a was obtained as white fluffy solid in 78% yield
(1.26g). The pure product 13b was obtained as white solid in 81% yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of compound 14a. To the solution of compound 13a (0.5g, 1.23mmol) in DMF,
3ꢀmercaptopropanoic acid, 9 (0.157g, 1.48mmol) was added. To this mixture, 1N NaOH (2
mL) was added and reaction was allowed to stir at 50oC for 5 h. The crude product 14a was
obtained by evaporating the solvent and was used for the next step without any further
purification. LCMS m/z calculated for 14a: C17H25NO4S [M+Na]+: 362.15, found 362.17.
LCMS m/z calculated for 14b: C19H26N2O4S [MꢀH+]: 378.20, found 377.14.
Synthesis of compound 15a. Compound 14a (0.5g, 1.47mmol) was added to a solution of
EDAC (0.57g, 3.67mmol) in DMF. The mixture was stirred at room temperature for 30 min
after which DIPEA (1.28mL, 7.35mmol) and 3ꢀ(aminomethyl)pyridine, 3 (0.24g, 2.21mmol)
were added. The reaction was allowed to stir for 18 h. On completion, the solvent was
evaporated and the reaction mixture was diluted with ethylacetate. The organic layer was
then washed with saturated NaHCO3, water and brine. The combined organic layers were
dried over Na2SO4 and concentrated to give the crude product which was then purified using
column chromatography (ethylacetate:methanol). The pure product 15a was obtained as
white solid in 62% yield (0.39g). 1H NMR (CDCl3, 500 MHz) δ 8.54 (d, J = 14.4 Hz, 2H), 7.66
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
18
1
2
3
4
(d, J = 9.4 Hz, 1H), 7.31ꢀ7.16 (m, 6H), 6.69 (bs, 1H), 4.72 (d, J = 10.3 Hz, 1H), 4.52ꢀ4.40 (m,
5
6
13
2H), 2.87 (t, J = 9.3 Hz, 3H), 2.63ꢀ2.51 (m, 4H), 1.40 (d, J = 4.15 Hz, 2H), 1.36 (s, 9H); C
7
8
9
NMR (CDCl3, 125 MHz) δ 149.2, 148.8, 135.6, 129.4, 129.3, 128.8, 128.6, 126.7, 123.6,
41.0, 36.6, 28.6, 28.3; HRMS m/z calculated for C23H31N3O3S [M+H]+: 430.231, found
430.21.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The pure product 15b was obtained as yellow oil in 56% yield. 1H NMR (CDCl3, 500 MHz) δ
8.51 (bs, 2H), 8.25 (s, 1H), 7.64 (d, J = 7.59 Hz, 2H), 7.35 (d, J = 8.07 Hz, 1H), 5.95 (bs,
1H), 7.23 (t, J = 4.85 Hz, 1H), 7.18 (t, J = 7.58 Hz, 1H), 7.11 (t, J = 7.28 Hz, 1H), 7.02 (s,
1H), 6.54 (s, 1H), 4.76 (bs, 1H), 4.41 (t, J = 5.91 Hz, 2H), 4.05 (m, 1H), 3.01 (bs, 2H), 2.86
(bm, 2H), 2.62 (bm, 2H), 2.49 (bm, 2H), 1.36 (s, 9H); 13C NMR (CDCl3, 125 MHz) δ 155.7,
149.1, 148.7, 136.3, 135.7, 134.1, 129.4, 126.6, 123.6, 122.9, 122.2, 119.7, 119.0, 111.3,
111.2, 50.7, 41.0, 36.5, 35.8, 28.8, 28.4.; HRMS m/z calculated for C17H27N3O3 [M+Na]+:
491.21, found 491.2.
The purity of compounds 4, 5, 15a and 15b was >95% as determined by 1H NMR
analysis. The purity of compound 11, however, did not exceed 90% even after repeated
attempts of column chromatography purification.
Protein Expression and Purification – The K282A/E285A mutation was introduced to the
CYP3A4ꢁ3ꢀ22 expression plasmid using a QuikChange mutagenesis kit (Stratagene).
The Cꢀterminally 4ꢀhistidine tagged wild type and mutant CYP3A4 were produced,
purified and quantified as reported previously.25
ACS Paragon Plus Environment

Page 19 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
19
1
2
3
4
5
6
7
8
9
Spectral Binding Titrations - Ligand binding to CYP3A4 was monitored in 50 mM
phosphate, pH 7.4, containing 20% glycerol and 1 mM dithiothreitol (buffer A) in a Cary
300 spectrophotometer. The protein was titrated with small aliquots of the investigated
compounds dissolved in dimethyl sulfoxide, with the final solvent concentration <2%.
Spectral dissociation constants (Ks) were determined from the titration curves using
oneꢀ or twoꢀsite saturation fitting.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Kinetics of Ligand Binding - Kinetics of ligand binding to CYP3A4 was monitored at
room temperature in 50 mM phosphate, pH 7.4, in a SX.18MV stopped flow apparatus
(Applied Photophysics, UK) after mixing 2 ꢀM CYP3A4 with various concentrations of
ligands. Absorbance changes were followed at 409ꢀ411 nm for pyridine,
aminoethylpyridine and compound 5, and at 427ꢀ428 nm for 4, 11, and 15a-b. Kinetic
data were analyzed using the program IgorPro (WaveMetrics, Inc).
Thermal Denaturation and Inhibitory Potency Assays – Thermal denaturation
experiments were conducted as previously described27 to compare ligandꢀdependent
changes in CYP3A4 stability. The inhibitory potency of the investigated compounds on
the 7ꢀbenzyloxyꢀ4ꢀ(trifluoromethyl)coumarin (BFC) Oꢀdebenzylation activity of CYP3A4
was evaluated fluorometrically in a reconstituted system with rat cytochrome P450
reductase (CPR). The reaction was carried out at room temperature in 100 mM
phosphate buffer, pH 7.4, containing catalase and superoxide dismutase (2 Units/ml
each). A mixture of 1 µM CYP3A4 and 1 µM CPR was preincubated for 1 hour at room
temperature and diluted by 20ꢀfold with the assay buffer immediately before
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
20
1
2
3
4
5
6
7
8
9
measurements. BFC (50 µM) and various concentrations of analogs were added to the
protein mixture 2 minutes prior to the reaction initiation with 100 µM NADPH. Formation
of 7ꢀhydroxyꢀ4ꢀtrifluoromethylcoumarin was followed in a Hitachi F100 fluorometer using
λex = 430 nm and λem = 500 nm. IC50 values were derived from the [% activity] vs.
[inhibitor] plots.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Crystallization and Determination of the X-ray Structures of Ligand-bound CYP3A4 -
CYP3A4 K282A/E285A was coꢀcrystallized with 4 by a sitting drop vapor diffusion
method. Protein (80 mg/ml) in buffer A was mixed with a 10ꢀfold excess of 4 and
centrifuged to remove the precipitate. CYP3A4 K282A/E285A directly ligated to 4 was
crystallized against the 80% solution E7 from the Morpheus crystallization screen
(Molecular Dimensions). The imidazoleꢀcontaining solution A3 from the same screening
kit resulted in coꢀcrystallization of the mutant with both imidazole and 4 in the active site.
Crystals of the wild type CYP3A4 bound to compounds 15a and 15b were grown by a
microbatch method under oil. A half of a microliter of ligandꢀbound CYP3A4 was mixed
with 0.5 ꢀl of 6ꢀ10% polyethylene glycol 3,350 and 60ꢀ90 mM sodium malonate, pH 5.0,
and covered with paraffin oil. All crystals were cryoprotected with ParatoneꢀN and
frozen in liquid nitrogen. Xꢀray diffraction data were collected at the Stanford
Synchrotron Radiation Lightsource beamline 7ꢀ1 and the Advanced Light Source
beamline 8.2.1. Crystal structures were solved by molecular replacement with
PHASER37 and the ligandꢀfree 1TQN structure as a search model. The initial models
were rebuilt and refined with COOT38, PHENIX39 and REFMAC.37 Data collection and
refinement statistics are summarized in the Supporting Information Table 1S. The
ACS Paragon Plus Environment

Page 21 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
21
1
2
3
4
5
6
7
8
9
atomic coordinates and structure factors for 4ꢀ, imidazole/4ꢀ, 15aꢀ and 15bꢀbound
CYP3A4 have been deposited in the Protein Data Bank with the ID codes 4D75, 4D6Z,
4D78 and 4D7D, respectively.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Supporting Information
The Xꢀray data collection and refinement statistics, and mass spectrometry and NMR
data for compounds 4ꢀ15a/b. This material is available free of charge via the Internet at
http://pubd.acs.org.
PDB ID codes: 4D75, 4D6Z, 4D78 and 4D7D.
Author Information
Corresponding Author:
*Email: sevrioui@uci.edu. Phone: (949) 824ꢀ1953.
Present address
§William Paterson University, 300 Pompton Road, Wayne NJ 07470.
Notes
The authors declare no competing financial interest.
Acknowledgement
This work was supported by the N.I.H. National Institute of General Medical Sciences
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
22
1
2
3
4
5
6
7
8
9
(GM57353, T.L.P.) and the University of California’s Center for Antiviral Drug Discovery
(MRPI 143226, A.R.C.), and involves research carried out at the Advanced Light Source
and the Stanford Synchrotron Radiation Lightsource, a national user facility operated by
Stanford University on behalf of the U.S. Department of Energy, Office of Basic Energy
Sciences. The Advanced Light Source is supported by the Director, Office of Science,
Office of Basic Energy Sciences, of the U.S. Department of Energy under Contract No.
DEꢀAC0205CH11231. The SSRL Structural Molecular Biology Program is supported by
the Department of Energy, Office of Biological and Environmental Research, and by
National Institutes of Health, National Center for Research Resources, Biomedical and
Technology Program, and the National Institute of General Medical Sciences. The
authors also thank J. Milligan and M. Vasquez for technical assistance.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Abbreviations
CYP3A4, 3A4 isoform of cytochrome P450; CPR, cytochrome P450 reductase; BFC, 7ꢀ
benzyloxyꢀ4ꢀ(trifluoromethyl)coumarin.
References
(1)
human cytochrome Pꢀ450 enzymes. Chem. Res. Toxicol. 1991, 4, 391ꢀ407.
(2) Li, A. P.; Kaminski, D. L.; Rasmussen, A. Substrates of human hepatic
cytochrome P450 3A4. Toxicology 1995, 104, 1ꢀ8.
(3) Guengerich, F. P. Cytochrome Pꢀ450 3A4: regulation and role in drug
Guengerich, F. P.; Shimada, T. Oxidation of toxic and carcinogenic chemicals by
metabolism. Annu. Rev. Pharmacol. Toxicol. 1999, 39, 1ꢀ17.
ACS Paragon Plus Environment

Page 23 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
23
1
2
3
4
5
6
7
8
9
(4)
metabolism of 2,4ꢀdichlorophenol. Chemosphere 1997, 34, 2281ꢀ2291.
(5) Chae, Y. H.; Yun, C. H.; Guengerich, F. P.; Kadlubar, F. F.; elꢀBayoumy, K.
Mehmood, Z.; Kelly, D. E.; Kelly, S. L. Cytochrome P450 3A4 mediated
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Roles of human hepatic and pulmonary cytochrome P450 enzymes in the metabolism of
the environmental carcinogen 6ꢀnitrochrysene. Cancer. Res. 1993, 53, 2028ꢀ2034.
(6)
Hodgson, E. In vitro human phase I metabolism of xenobiotics I: pesticides and
related compounds used in agriculture and public health. J. Biochem. Mol. Toxicol.
2003, 17, 201ꢀ206.
(7)
Mehmood, Z.; Williamson, M. P.; Kelly, D. E.; Kelly, S. L. Human cytochrome
P450 3A4 is involved in the biotransformation of the herbicide 2,4ꢀ
dichlorophenoxyacetic acid. Environ. Toxicol. Pharmacol. 1996, 2, 397ꢀ401.
(8)
Mehmood, Z.; Williamson, M. P.; Kelly, D. E.; Kelly, S. L. Metabolism of
organochlorine pesticides: the role of human cytochrome P450 3A4. Chemosphere
1996, 33, 759ꢀ769.
(9)
Kempf, D. J.; Marsh, K. C.; Kumar, G.; Rodrigues, A. D.; Denissen, J. F.;
McDonald, E.; Kukulka, M. J.; Hsu, A.; Granneman, G. R.; Baroldi, P. A.; Sun, E.;
Pizzuti, D.; Plattner, J. J.; Norbeck, D. W.; Leonard, J. M. Pharmacokinetic
enhancement of inhibitors of the human immunodeficiency virus protease by
coadministration with ritonavir. Antimicrob. Agents Chemother. 1997, 41, 654ꢀ660.
(10) Xu, L.; Desai, M. C. Pharmacokinetic enhancers for HIV drugs. Curr. Opin.
Investig. Drugs. 2009, 10, 775ꢀ786.
(11) Mathias, A. A.; German, P.; Murray, B. P.; Wei, L.; Jain, A.; West, S.; Warren, D.;
Hui, J.; Kearney, B. P. Pharmacokinetics and pharmacodynamics of GSꢀ9350: a novel
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
24
1
2
3
4
5
6
pharmacokinetic enhancer without antiꢀHIV activity. Clin. Pharmacol. Ther. 2010, 87,
322ꢀ329.
7
8
9
(12) Xu, L.; Liu, H.; Murray, B.; Callebaut, C.; Lee, M. S.; Hong, A.; Strickley, R. G.;
Tsai, L., K. ; Stray, K. M.; Wang, Y.; Rhodes, G. R.; Desai, M. C. Cobicistat (GSꢀ9350):
A potent and selective inhibitor of human CYP3A as a novel pharmacoenhancer. ACS
Med. Chem. Lett. 2010, 1, 209ꢀ213.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(13) Xu, L.; Liu, H.; Hong, A.; Vivian, R.; Murray, B. P.; Callebaut, C.; Choi, Y. C.; Lee,
M. S.; Chau, J.; Tsai, L. K.; Stray, K. M.; Strickley, R. G.; Wang, J.; Tong, L.;
Swaminathan, S.; Rhodes, G. R.; Desai, M. C. Structureꢀactivity relationships of
diamine inhibitors of cytochrome P450 (CYP) 3A as novel pharmacoenhancers. Part II:
P2/P3 region and discovery of cobicistat (GSꢀ9350). Bioorg. Med. Chem. Lett. 2014, 24,
995ꢀ999.
(14) Kempf, D. J.; Marsh, K. C.; Denissen, J. F.; McDonald, E.; Vasavanonda, S.;
Flentge, C. A.; Green, B. E.; Fino, L.; Park, C. H.; Kong, X. P.; et al. ABTꢀ538 is a potent
inhibitor of human immunodeficiency virus protease and has high oral bioavailability in
humans. Proc. Natl. Acad. Sci. U S A 1995, 92, 2484ꢀ2488.
(15) Koudriakova, T.; Iatsimirskaia, E.; Utkin, I.; Gangl, E.; Vouros, P.; Storozhuk, E.;
Orza, D.; Marinina, J.; Gerber, N. Metabolism of the human immunodeficiency virus
protease inhibitors indinavir and ritonavir by human intestinal microsomes and
expressed cytochrome P4503A4/3A5: mechanismꢀbased inactivation of cytochrome
P4503A by ritonavir. Drug Metab. Dispos. 1998, 26, 552ꢀ561.
ACS Paragon Plus Environment

Page 25 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
25
1
2
3
4
5
6
7
8
9
(16) von Moltke, L. L.; Durol, A. L.; Duan, S. X.; Greenblatt, D. J. Potent mechanismꢀ
based inhibition of human CYP3A in vitro by amprenavir and ritonavir: comparison with
ketoconazole. Eur. J. Clin. Pharmacol. 2000, 56, 259ꢀ261.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(17) Ernest, C. S., 2nd; Hall, S. D.; Jones, D. R. Mechanismꢀbased inactivation of
CYP3A by HIV protease inhibitors. J. Pharmacol. Exp. Ther. 2005, 312, 583ꢀ591.
(18) Lin, H. L.; D'Agostino, J.; Kenaan, C.; Calinski, D.; Hollenberg, P. F. The effect of
ritonavir on human CYP2B6 catalytic activity: heme modification contributes to the
mechanismꢀbased inactivation of CYP2B6 and CYP3A4 by ritonavir. Drug Metab.
Dispos. 2013, 41, 1813ꢀ1824.
(19) Rock, B. M.; Hengel, S. M.; Rock, D. A.; Wienkers, L. C.; Kunze, K. L.
Characterization of RitonavirꢀMediated Inactivation of Cytochrome P450 3A4. Mol.
Pharmacol. 2014, 86, 665ꢀ674.
(20) Sekiguchi, N.; Higashida, A.; Kato, M.; Nabuchi, Y.; Mitsui, T.; Takanashi, K.;
Aso, Y.; Ishigai, M. Prediction of drugꢀdrug interactions based on timeꢀdependent
inhibition from high throughput screening of cytochrome P450 3A4 inhibition. Drug
Metab. Pharmacokinet. 2009, 24, 500ꢀ510.
(21) Iribarne, C.; Berthou, F.; Carlhant, D.; Dreano, Y.; Picart, D.; Lohezic, F.; Riche,
C. Inhibition of methadone and buprenorphine Nꢀdealkylations by three HIVꢀ1 protease
inhibitors. Drug Metab. Dispos. 1998, 26, 257ꢀ260.
(22) Eagling, V. A.; Back, D. J.; Barry, M. G. Differential inhibition of cytochrome P450
isoforms by the protease inhibitors, ritonavir, saquinavir and indinavir. Br. J. Clin.
Pharmacol. 1997, 44, 190ꢀ194.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
26
1
2
3
4
5
6
7
8
9
(23) Kumar, G. N.; Rodrigues, A. D.; Buko, A. M.; Denissen, J. F. Cytochrome P450ꢀ
mediated metabolism of the HIVꢀ1 protease inhibitor ritonavir (ABTꢀ538) in human liver
microsomes. J. Pharmacol. Exp. Ther. 1996, 277, 423ꢀ431.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(24) Zalma, A.; von Moltke, L. L.; Granda, B. W.; Harmatz, J. S.; Shader, R. I.;
Greenblatt, D. J. In vitro metabolism of trazodone by CYP3A: inhibition by ketoconazole
and human immunodeficiency viral protease inhibitors. Biol. Psychiatry 2000, 47, 655ꢀ
661.
(25) Sevrioukova, I. F.; Poulos, T. L. Structure and mechanism of the complex
between cytochrome P4503A4 and ritonavir. Proc. Natl. Acad. Sci. U S A 2010, 107,
18422ꢀ18427.
(26) Sevrioukova, I. F.; Poulos, T. L. Interaction of human cytochrome P4503A4 with
ritonavir analogs. Arch. Biochem. Biophys. 2012, 520, 108ꢀ116.
(27) Sevrioukova, I. F.; Poulos, T. L. Pyridineꢀsubstituted desoxyritonavir is a more
potent cytochrome P450 3A4 inhibitor than ritonavir. J. Med. Chem. 2013, 56, 3733ꢀ
3741.
(28) Sevrioukova, I. F.; Poulos, T. L. Dissecting cytochrome P450 3A4ꢀligand
interactions using ritonavir analogues. Biochemistry 2013, 52, 4474ꢀ4481.
(29) Sevrioukova, I. F.; Poulos, T. L. Ritonavir analogues as a probe for deciphering
the cytochrome P450 3A4 inhibitory mechanism. Curr. Top. Med. Chem. 2014, 14,
1348ꢀ1355.
(30) Ekins, S.; Bravi, G.; Binkley, S.; Gillespie, J. S.; Ring, B. J.; Wikel, J. H.;
Wrighton, S. A. Threeꢀ and fourꢀdimensional quantitative structure activity relationship
ACS Paragon Plus Environment

Page 27 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
27
1
2
3
4
5
6
analyses of cytochrome Pꢀ450 3A4 inhibitors. J. Pharmacol. Exp. Ther. 1999, 290, 429ꢀ
438.
7
8
9
(31) Ekins, S.; Bravi, G.; Wikel, J. H.; Wrighton, S. A. Threeꢀdimensionalꢀquantitative
structure activity relationship analysis of cytochrome Pꢀ450 3A4 substrates. J.
Pharmacol. Exp. Ther. 1999, 291, 424ꢀ433.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(32) Ekins, S.; Stresser, D. M.; Williams, J. A. In vitro and pharmacophore insights
into CYP3A enzymes. Trends Pharmacol. Sci. 2003, 24, 161ꢀ166.
(33) Riley, R. J.; Parker, A. J.; Trigg, S.; Manners, C. N. Development of a
generalized, quantitative physicochemical model of CYP3A4 inhibition for use in early
drug discovery. Pharm. Res. 2001, 18, 652ꢀ655.
(34) Liu, H.; Xu, L.; Hui, H.; Vivian, R.; Callebaut, C.; Murray, B. P.; Hong, A.; Lee, M.
S.; Tsai, L. K.; Chau, J. K.; Stray, K. M.; Cannizzaro, C.; Choi, Y. C.; Rhodes, G. R.;
Desai, M. C. Structureꢀactivity relationships of diamine inhibitors of cytochrome P450
(CYP) 3A as novel pharmacoenhancers, part I: core region. Bioorg. Med. Chem. Lett.
2014, 24, 989ꢀ994.
(35) Goldschmidt, L.; Cooper, D. R.; Derewenda, Z. S.; Eisenberg, D. Toward rational
protein crystallization: A Web server for the design of crystallizable protein variants.
Protein Sci. 2007, 16, 1569ꢀ1576.
(36) Williams, P. A.; Cosme, J.; Vinkovic, D. M.; Ward, A.; Angove, H. C.; Day, P. J.;
Vonrhein, C.; Tickle, I. J.; Jhoti, H. Crystal structures of human cytochrome P450 3A4
bound to metyrapone and progesterone. Science 2004, 305, 683ꢀ686.
(37) CCP4. Collaborative computational project number 4. The CCP4 suite programs
for protein crystallography. Acta Crystallogr. Section D 1994, 50, 760ꢀ763.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 46
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
28
1
2
3
4
5
6
(38) Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Features and development of
Coot. Acta Crystallogr. Section D 2010, 66, 486ꢀ501.
7
8
9
(39) Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis, I. W.; Echols, N.;
Headd, J. J.; Hung, L. W.; Kapral, G. J.; GrosseꢀKunstleve, R. W.; McCoy, A. J.;
Moriarty, N. W.; Oeffner, R.; Read, R. J.; Richardson, D. C.; Richardson, J. S.;
Terwilliger, T. C.; Zwart, P. H. PHENIX: a comprehensive Pythonꢀbased system for
macromolecular structure solution. Acta Crystallogr Section D 2010, 66, 213ꢀ221.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 29 of 46
Journal of Medicinal Chemistry
Kaur et al. Evaluation of the CYP3A4 inhibitor pharmacophore model
29
1
2
3
4
Table 1. Properties of pyridine-containing compounds
5
6
7
8
_________________________________________________________________________________________
b
c
d
e
442 nm banda
(Fe2+)
Ks
(ꢀM)
kfast
ꢁTm
(°C)
IC50
(ꢀM)
(sꢀ1)
9
_________________________________________________________________________________________
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
pyridine
ꢀ
ꢀ
ꢀ
4500±400
2500±200
0.5±0.1
0.6±0.1
0.8±0.1
0
4000±500
5000±600
1000±150
aminoethylpyridine
ꢀ0.1
ꢀ0.2
5
10±1 f
4400±500 f
4
+
+
105±10
10±1
20±2
4.0±0.2
1.7±0.2
8.0±1.5
7.0
ꢀ0.2
ꢀ2±0.4
ꢀ1±0.2
0.2
75±5
11
30±3
15a
++
0.9±0.1
0.5±0.1
0.025
0.52±0.05
0.21±0.04
0.13
15b
++
GS3g
+++
5.1
ritonavirg
+++
0.051
1.4
2.7
0.55
_________________________________________________________________________________________
a Appearance and magnitude of the 442 nm peak in the absorbance spectrum of ligandꢀbound ferrous CYP3A4.
b Spectral dissociation constant.
c Rate constant for the fast phase of the binding reaction measured at saturating ligand concentrations.
d Ligandꢀdependent changes in the melting temperature of CYP3A4.
e Concentration required for halfꢀmaximal inactivation of recombinant CYP3A4.
f Values derived from the two site saturation fit to a titration curve.
g Determined previously. 25, 27
ACS Paragon Plus Environment