Exploration of Fragment Binding Poses Leading to Efficient Discovery of Highly Potent and Orally Effective Inhibitors of FABP4 for Anti-inflammation

Article abstract of DOI:10.1021/acs.jmedchem.9b02107

Fatty-acid binding protein 4 (FABP4) is a promising therapeutic target for immunometabolic diseases, while its potential for systemic inflammatory response syndrome treatment has not been explored. Here, a series of 2-(phenylamino)benzoic acids as novel and potent FABP4 inhibitors are rationally designed based on an interesting fragment that adopts multiple binding poses within FABP4. A fusion of these binding poses leads to the design of compound 3 with an ?460-fold improvement in binding affinity compared to the initial fragment. A subsequent structure-aided optimization upon 3 results in a promising lead (17) with the highest binding affinity among all the inhibitors, exerting a significant anti-inflammatory effect in cells and effectively attenuating a systemic inflammatory damage in mice. Our work therefore presents a good example of lead compound discovery derived from the multiple binding poses of a fragment and provides a candidate for development of drugs against inflammation-related diseases.

Full text of DOI:10.1021/acs.jmedchem.9b02107

Subscriber access provided by University of Rochester | River Campus & Miner Libraries
Article
Exploration of Fragment Binding Poses Leading to Efficient Discovery of
Highly Potent and Orally Effective Inhibitors of FABP4 for Anti-inflammation
Haixia Su, Yi Zou, Guofeng Chen, Huixia Dou, Hang Xie, Xiaojing
Yuan, Xianglei Zhang, Naixia Zhang, Minjun Li, and Yechun Xu
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b02107 • Publication Date (Web): 23 Mar 2020
Downloaded from pubs.acs.org on March 23, 2020
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 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 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.
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 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Exploration of Fragment Binding Poses Leading to Efficient Discovery
of Highly Potent and Orally Effective Inhibitors of FABP4 for Anti-
inflammation
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Haixia SuΔ,§,#, Yi Zou , Guofeng Chen , Huixia Dou , Hang Xie , Xiaojing
Δ,#
Δ,§,#
Δ,#
Δ
Δ
Δ,§
Δ,§
ǂ
Δ,§,*
Yuan , Xianglei Zhang , Naixia Zhang , Minjun Li , Yechun Xu
ΔCAS Key Laboratory of Receptor Research, Drug Discovery and Design Center,
Department of Analytical Chemistry, Shanghai Institute of Materia Medica,
Chinese Academy of Sciences, Shanghai 201203, China
§University of Chinese Academy of Sciences, Beijing 100049
ǂShanghai Synchrotron Radiation Facility, Shanghai Advanced Research Institute,
Chinese Academy of Sciences, Shanghai 201210, China
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Fatty-acid binding protein 4 (FABP4) is a promising therapeutic target for
immunometabolic diseases, while its potential for systemic inflammatory response
syndrome treatment has not been explored. Here, a series of 2-(phenylamino)benzoic
acids as novel and potent FABP4 inhibitors are rationally designed based on an
interesting fragment which adopts multiple binding poses within FABP4. A fusion of
these binding poses leads to design of compound 3 with a ~460-fold improvement in
binding affinity compared to the initial fragment. A subsequent structure-aided
optimization upon 3 results in a promising lead (17) with the highest binding affinity
among all the inhibitors, exerting a significant anti-inflammatory effect in cells and
effectively attenuating a systemic inflammatory damage in mice. Our work therefore
presents a good example of lead compounds discovery derived from the multiple
binding poses of a fragment and provides a candidate for development of drugs against
inflammation-related diseases.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0

INTRODUCTION
Fatty-acid binding proteins (FABPs) are a class of lipid chaperones involved in
regulation of the uptake and intracellular transport of fatty acids (FAs). So far, 10
1
different FABPs with tissue-specific expression have been identified. Among them,
FABP4 is predominantly expressed in mature adipocytes and macrophages, acting as
2
,3
an important regulator in inflammatory-related signaling pathways. In macrophages,
FABP4 aggravates the lipopolysaccharide (LPS)-induced inflammatory response by the
formation of a positive feedback loop with c-Jun N-terminal Kinase (JNK) and activate
4
protein 1 (AP-1). In addition, FABP4 can disturb the eicosanoid homostasis by
regulating the activity of cyclooxygenase 2 (COX2) and the stability of leukotriene A4
(LTA4), ultimately exerting an impact on macrophages functions and inflammatory
5
,6
response. It has been reported that FABP4-deficient macrophages exhibit a dramatic
decrease in production of inflammatory cytokines including monocyte chemo-attractant
2
ACS Paragon Plus Environment

Page 3 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
protein 1 (MCP-1), interleukin 6 (IL-6), tumor necrosis factor α (TNF-α), and
7
,8
interleukin 1β (IL-1β). In adipocytes, a fat-specific deletion of FABP4 also decreases
9
the expression of inflammatory cytokines. FABP4 inhibition or knockout in mice has
acquired therapeutic benefits for inflammation-associated diseases, such as
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
atherosclerosis,8,10,11 renal interstitial fibrosis, non-alcoholic steatohepatitis,
12
13,14
and
1
5
acute lung injury. Accordingly, FABP4 inhibitors provide the therapeutic potential
for these diseases.16
It is noteworthy that FABP3-deficient mice suffered from a regional cardiac
1
7
hypertrophy and stress-intolerability. The selectivity of inhibitors for FABP4 over
FABP3 has to be considered in order to avoid the potential cardiac toxicity. In contrast,
FABP4/FABP5 double-knockout mice displayed a more significant protection from
insulin resistance, type 2 diabetes, and atherosclerosis compared to FABP4 or FABP5
1
8
single-knockout. FABP4/FABP5 dual inhibitors are thus expected to have great utility
in treating these diseases. Recently, Li et al. reported that it is FABP4 rather than
FABP5 that promoted the growth and metastasis of breast cancer.19 Accordingly,
FABP4 selective inhibitors may afford a better choice for the treatment of cancer. In
terms of anti-inflammation, there is no evidence to show that FABP4/FABP5 dual
inhibitors would have better efficacy than the FABP4 selective inhibitors do.
To date, several classes of FABP4 inhibitors have been identified for therapeutic
intervention in diseases like insulin resistance, type 2 diabetes and atherosclerosis.20-32
BMS309403, the prime inhibitor of FABP4 developed by Bristol-Myers Squibb, has
been used to explore its anti-inflammatory effect in vitro and in vivo. It was shown that
BMS309403 decreased the production of pro-inflammatory cytokines and alleviated
the LPS-induced acute lung injury, non-alcoholic steatohepatitis, and renal interstitial
fibrosis.12-15,33,34 However, compared to the knockout of FABP4, the anti-inflammatory
efficacy of BMS309403 was unsatisfied.7,8 It is noteworthy that binding affinities of
most available inhibitors are equivalent to that of FAs such as palmitic acid and linoleic
1
6
acid. As a result, it is difficult for these inhibitors to compete with FAs which are
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
often at a high level in vivo. Therefore, it is worth identifying more potent FABP4
inhibitors to systematically explore the effectiveness of FABP4 in anti-inflammation
and to provide novel compounds possessing more-drug-like properties.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Herein, with a fragment-based lead discovery strategy, we sought to design and
synthesize a new class of potent FABP4 inhibitors, and therapeutic potentials of the
inhibitors in treating systemic inflammatory response syndrome were explored. To find
a hit with a new scaffold, we screened our in-house 500-fragment library and identified
an interesting hit (fragment 1) with multiple binding poses shown in the ligand binding
pocket of FABP4. With the aid of the determined crystal structures and binding
affinities, a series of 2-(phenylamino)benzoic acids as novel and potent FABP4
inhibitors were successfully designed on basis of this weak fragment binder. In the
following, we comprehensively illustrate the hit discovery, a hit-to-lead optimization
and evaluation of potency, selectivity, pharmacokinetic properties, and anti-
inflammatory efficacy of the inhibitors.

RESULTS AND DISCUSSION
Hit Identification and Characterization. Given the advantage of fragment-
screening campaigns in hit/lead discovery,35,36 an in-house 500-fragment collection was
screened against human FABP4 using the 8-anilino1-naphthalene-sulfonic acid (1,8-
ANS) displacement assay. The concentration of fragments used for the testing is 1 mM.
Two analogs of benzoic acid, fragments 1 (MW = 214.2, K = 50.7 μM) and 2 (MW =
i
213.2, K > 500 μM), were identified as FABP4 binders. Encouraged by a high ligand-
i
binding efficiency of fragment 1 (LE = 0.36), we carried out a structure-based hit-to-
lead optimization on this diphenylamine scaffold in order to discover novel, highly
potent and selective FABP4 inhibitors. However, the endogenous FAs bound in the
ligand binding pocket of FABP4 and the relatively weak binding affinity of 1 impeded
the complex structure determination. In this scenario, we tried to remove the
endogenous FAs from FABP4 through a denaturation and refolding of the recombinant
4
ACS Paragon Plus Environment

Page 5 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
FABP4. To solve the structure of FABP4 in complex with fragment 1, crystals of apo
FABP4 obtained at pH 6.5 and 7.5 were soaked with the fragment. To our surprise, the
complex structures show that fragment 1 adopts more than one binding pose in the
ligand binding pocket of FABP4. According to the electron density map, three different
but partially overlapped binding poses of 1 could be fitted (Figures 1A and S1). At pH
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
.5, two poses (Pose 1 and Pose 2) are found in the pocket (Figure S1A). Pose 1 presents
a classical binding mode which is utilized by most FABP4 inhibitors. With such a pose,
the carboxylic group forms multiple polar interactions with R126 as well as Y128, and
the free phenyl ring mainly establishes face-to-edge π-stacking interactions with F16
(
Figure 1B). However, in Pose 2, the carboxylic group forms hydrogen bonds (H-bonds)
to S53 and T60 directly, and to S55, K58 and A75 through water molecules (Figure
C). Besides, the phenyl ring of the benzoic acid group interacts with F57 via a π-π
1
stacking, while the free phenyl ring occupies a position as that seen in Pose 1 to form
face-to-edge π-stacking interactions with F16. The difference electron-density map
suggests that Pose 2 (occupancy = 0.58) is more stable than Pose 1 (occupancy = 0.42)
in the crystal structure determined at pH 6.5 (Figure S1A). There are also two binding
poses of the fragment found in the complex structure determined at pH 7.5. One is Pose
2 although its occupancy (0.41) is relatively low. The second one is a new pose, Pose 3
(occupancy = 0.59) (Figure S1B). Similar to Pose 1, the carboxylic group of the benzoic
acid in Pose 3 forms polar interactions with R126 and Y128, while the free phenyl ring
flips up to form hydrophobic interactions with F57 and A75 (Figure 1D). Taken
together, two crystal structures of FABP4 in complex with fragment 1 reveal three
possible binding poses of the fragment, suggesting that this fragment in the ligand
binding pocket is mobile and may switch from one pose to another.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Crystal structures of FABP4 in complex with fragment 1 determined at pH
6
.5 and 7.5 (PDB codes: 6LJW and 6LJX). (A) An overlay of three binding poses of 1
shown in B-D. (B-D) A representation of Pose 1 (B), Pose 2 (C) and Pose3 (D).
Residues interacting with 1 are shown as gray sticks and H-bonds are represented by
black dashed lines.
Although the occupancy of Pose 1 in the crystal structure is low, it is similar to the
typical binding mode utilized by other FABP4 inhibitors, in which F16, R126 and Y128
1
6
are the key residues for the ligand binding. To further explore the probability of Pose
1
and investigate the dynamic behavior of fragment 1 inside FABP4, molecular
16
dynamic (MD) simulations were performed on Pose 1 in complex with FABP4. As
expected, the fragment is mobile during the simulation and its root mean square
deviations (RMSDs) fluctuate from 1.0 to 7.0 Å, while the backbone RMSDs of FABP4
have a relatively smooth fluctuation around 2.0 Å (Figure 2). After a cluster analysis
on conformations of the fragment extracted from the MD trajectory, four major
conformations (I to IV) are found and their representative snapshots superimposed to
three poses in the crystal structures are displayed in Figure 2. It seems that
conformations I and III are very close to Poses 1 and 3, with averaged RMSDs of 2.0
and 1.3 Å, respectively. Conformation II is similar to Pose 2 but with a relatively larger
RMSD (2.9 Å). Conformation IV of the fragment is distinctive to three poses (RMSDs
> 4.6 Å), although the benzoic acid moiety is similar to that of Pose 3. Accordingly,
starting with Pose 1, the fragment can move to conformations almost identical to Pose
2
and Pose 3 during the 600-ns MD simulation, demonstrating again that the binding of
fragment 1 with FABP4 is mobile and the multiple bound conformations could be
6
ACS Paragon Plus Environment

Page 7 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
exchanged one another.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. An overlay of representative conformations of fragment 1 extracted from the
MD trajectory (gray sticks) with three poses shown in the crystal structures (Pose 1:
cyan sticks; Pose 2: yellow sticks; Pose 3: pink sticks). RMSDs of backbone atoms of
FABP4 (black) and all atoms of fragment 1 (red) versus the simulation time were
calculated based on the MD trajectory.
Micromolar-to-Nanomolar Potency Improvement. Though fragment 1 employs
multiple conformations to interact with FABP4, the essential amino acids participating
in binding include F16, S53, F57, R126, and Y128. This gave us a valuable clue to
rationally design more potent inhibitors with full consideration of interactions with
these residues. In addition, two phenyl rings of the fragment are overlapped each other
in three binding poses, resulting in three location of the phenyl rings (Figure 1A and
3
A). In order to obtain a new compound containing three phenyl rings so as to
simultaneously occupy the overlapped position of phenyl rings in three poses, we added
a benzene ring to C13 of Pose 1 (Figure 3A). The resulting 2-([1,1'-biphenyl]-2-
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ylamino)benzoic acid (3) has a K of 109.3 nM, which is a ~460-fold increase in K
i
i
compared to fragment 1 (Table 1). It is also more potent than the well-studied
BMS309403 with a K of 253.4 nM. Attachment of an anilino (4) or a phenoxy group
i
(5) to C13 of 1 also shows a modest increase in the binding affinity, but they are much
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
less potent than compound 3. The complex structure of FABP4 bound with 3 reveals
that three phenyl rings take positions as they are in three binding poses of fragment 1,
and the binding pose of compound 3 is more like a hybrid of Pose 1 and Pose 2 (Figures
1
and 3B). The added phenyl ring extends to a position which was once occupied by
the benzoic acid moiety of Pose 2 or the free phenyl ring of Pose 3. This ring of 3
simultaneously participates in an edge-to-face π−π stacking interactions with F57 and
cation-π interactions with R126. The rest part of 3 adopts a conformation similar to that
in Pose 1. The carboxylic group makes H-bonds with R126 and Y128 directly, and with
S53 through a water molecule, while the middle phenyl ring forms face-to-edge π-
stacking interactions with F16 (Figures 3B,C). These interactions between three phenyl
rings as well as the carboxylic group of 3 and five key residues (F16, S53, F57, R126,
and Y128) account for the huge improvement of potency for compound 3 relative to
fragment 1, demonstrating the rational for design of compound 3. However, inserting
an additional NH and O between two phenyl rings in compounds 4 and 5, respectively,
might generate clashes with the neighboring residues and result in remarkably reduced
potency (Table 1). In summary, the hybrid of the binding poses of fragment 1 leads to
efficiently design a novel and potent inhibitor of FABP4 by only adding a phenyl ring
to the right position of fragment 1.
8
ACS Paragon Plus Environment

Page 9 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. (A) Design of new inhibitors based on an overlay of different binding poses
of fragment 1. (B) An overlay of the binding poses of compound 3 (purple sticks) and
fragment 1 by superimposing crystal structures of FABP4 in complex with 3 and 1. (C)
Interactions of 3 with the surrounding residues revealed by the crystal structure (PDB
code: 6LJS). Residues are shown as cyan sticks and H-bonds are represented by black
dashed lines.
Binding Affinities Determined by ITC. Given that the concentration of FABP4
used in the 1,8-ANS displacement assay is ~0.8 µM, the detection limit of IC with
5
0
this assay is over 0.4 µM. Therefore, this assay usually measures K values over 77.4
i
nM according to the equation K = IC /(1+[1,8-ANS]/K ), where [1,8-ANS] is the
i
50
d
concentration of 1,8-ANS used in the assay and K means the dissociating constant for
d
1,8-ANS with FABP4. We found that the K of compound 3 (109.3 nM) is close to the
i
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
detection limit of the 1,8-ANS displacement assay. This makes it difficult to use this
assay to precisely determine the binding affinity of more potent inhibitors derived from
compound 3. Isothermal titration calorimetry (ITC), widely known as an invaluable tool
to determine thermodynamic parameters of protein-ligand interactions, is also
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-3
-8
frequently used to obtain K values ranging from 10 M to 10 M. To gain the accurate
d
binding affinity of the inhibitors with FABP4, ITC measurements were employed to
determine the K values of the compounds, in particular for the binding affinities of
d
those compounds (IC < 0.5 µM, K < 96.8 nM) close to or even beyond the detection
5
0
i
limit of the 1,8-ANS displacement assay. The ITC measured K of compound 3 binding
d
to apo FABP4 is 110.6 nM, which is also stronger than that of BMS309403 (252.7 nM).
1
Table 1. Effects of X, Y and R substitutions on binding affinities of compounds
with FABP4 a
HO
O
R1
Y
X
Compound
X
Y
R1
K or K (nM)
MW
LE
i
d
50690.3 ± 4326.7 b
1
2
H
H
NH
O
H
H
213.2
214.2
0.36
> 500000 b
< 0.27
1
09.3 ± 0.6 b
0.43
0.43
3
H
NH
289.3
110.6 ± 8.8 c
4
H
NH
3741.9 ± 201.3 b
304.3
0.32
HN
O
1476.0 ± 189.5 b
5
6
7
H
Cl
Br
NH
NH
NH
305.3
323.8
368.2
0.34
0.43
0.39
34.4 ± 7.5 c
262.9 ± 11.0 b
1
0
ACS Paragon Plus Environment

Page 11 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
8
9
CF3
NH
NH
164.1 ± 4.5 b
357.3
303.4
0.35
0.40
0.41
CH3
182.4 ± 4.4 b
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
0
OCH3 NH
43.9 ± 5.1 c
319.4
474.5
253.4 ± 21.2 b
0.25
0.25
BMS309403
2
52.7 ± 30.7 c
a
The values of K or K were expressed as mean ± SD based on three independent experiments
i
d
and determined via the nonlinear regression analysis using GraphPad Prism software 8.0.
b
K measured by the 1,8-ANS displacement assay
i
c
K measured by ITC
d
Optimization Based on Compound 3. Inspired by the above results, we embarked
on an optimization around the [1,1'-biphenyl]-2-amine scaffold rather than the benzoic
acid moiety of compound 3. The crystal structure of FABP4-3 reveals that small
cavities appear near atoms C11 and C12. It looks like that substituents with a small size
are suitable to fill these cavities. To test this, introducing various substituents at C12
(
the X substituent) was conducted in order to find the right substituent matching the
hydrophobic cavity defined by M20, T29 and F57 (Site P1) (Figure 4A and Table 1).
Compared to the unsubstituted 3, introduction of a chloro (compound 6) or a methoxy
(
compound 10) group results in higher binding affinities. In particular, the chloro-
substituted compound (6, K = 34.4 nM) yields an approximately 3-fold increase in
d
affinity compared to compound 3 (Table 1). However, introduction of a bromo
(
compound 7), a trifluoromethyl (compound 8) or a methyl (compound 9) at the same
position decreases the binding affinities compared to compound 3. The determined
crystal structure of FABP4 in complex with 6 shows the protein-ligand interactions
seen in the FABP4-6 complex are almost identical to those in the FABP4-3 complex,
except that the chloro group in 6 forms favorable hydrophobic interactions with the
surround residues such as M20, T29, A30, and F57 (Figure 4B). Besides the size of the
substituent groups, the compounds attached with the electron-withdrawing substituents
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
tend to be more potent compared to their counterparts, for example, 8 versus 9.
Accordingly, the chloro group, with an appropriately size and a negative inductive
effect, was identified as the favorable substituent of X for a further optimization.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. X-ray crystal structures of FABP4 in complex with compounds 3 (purple, A;
PDB code: 6LJS), 6 (green, B; PDB code: 6LJT), 11 (magenta, C; PDB code: 6LJU),
and 17 (yellow, D; PDB code: 6LJV). Sites P1, P2 and P3 show the optimization
directions. The ligand binding pocket is presented by molecular surface. The dashed
lines indicate H-bonds between compounds and residues.
Optimization Based on Compound 6. The crystal structure of FABP4-6 also
reveals that a region defined by V23 and V25 (Site P2) could potentially accommodate
the substituents at C11 of 6 (Figure 4A,B). Considering the synthesis feasibility and the
selectivity for FABP4 over FABP3, several aliphatic amine substituted compounds
1
2
ACS Paragon Plus Environment

Page 13 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were designed, synthesized and evaluated (Table 2). The most potent compound is 11
(
K = 60.1 nM), which is slightly less potent than 6. The determined crystal structure of
d
FABP4 in complex with compound 11 suggests that this compound adopts a binding
pose almost identical to 6 and the introduced methyl group establishes hydrophobic
interactions with V23 and V25 (Figure 4C). The underlying mechanism accounting for
the lower potency of 11 over 6 may be ascribed to the electron-donating ability of the
aliphatic amine substituents, which are helpful to dense the π-electron density and
thereby are harmful to the π–π interactions between the attached benzene ring and F16.
However, compounds 12 and 13, decorated with a bulkier N,N-dimethylamino and
ethylamino group, respectively, show a remarkably reduced potency compared to 11.
It is thus suggested that the N,N-dimethylamino or ethylamino group is too large to
compatible with Site P2.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Another small region could be exploited by our inhibitors is Site P3 which is around
the free phenyl ring of 6 (Figure 4A). Aiming to engage more favorable interactions
with the neighboring residues and maintain the binding pose of 6, the free phenyl ring
of 6 was replaced by various bicyclic rings. Six synthesized compounds (14~19) retain
high binding affinities with FABP4. Among them, the most potent one, compound 17,
has a K of 37.4 nM which nearly equals to the K of 6 (Table 2). The binding mode of
d
d
1
7 with FABP4 was also determined by the X-ray protein crystallography (Figure 4D).
As expected, the bound pose of 17 is overlapped well with that of 6 and an additional
H-bond is observed between the hydroxyl group of S55 and the oxygen atom of the
dihydrofuran ring in 17. Increase of aromaticity of the bicyclic rings reduces the binding
affinities of compounds 14, 15, 18, and 19 with FABP4. Introducing the second oxygen
into the bicyclic ring also results in a lower binding affinity of 16 compared to 17 (136.2
nM versus 37.4 nM). Overall, small modification of the free phenyl ring of 6 so as to
fill the neighbouring space leads to discovery of a novel and potent inhibitor (17) with
a binding affinity comparable to that of 6.
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
Table 2. Effects of R and R substituents on binding affinities of compounds with
FABP4 a
HO
O
R1
H
N
Cl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
R2
Compound
6
R1
R2
K or K (nM)
MW
LE
i
d
H
34.4 ± 7.5 c
60.1 ± 9.2 c
323.8
0.43
H
N
11
12
352.8
366.8
366.8
0.39
0.32
0.30
826.4 ± 22.8 b
1525.0 ± 54.4 b
N
H
N
1
1
3
4
HN
N
H
96.2 ± 4.5 c
125.5 ± 22.5 c
136.2 ± 2.7 c
37.4 ± 7.4 c
363.8
365.8
367.8
365.8
363.8
0.36
0.36
0.36
0.39
0.37
N O
N
1
1
5
6
H
H
H
H
O
O
O
O
17
1
8
90.1 ± 18.3 c
1
4
ACS Paragon Plus Environment

Page 15 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
2
9
0
H
149.7 ± 28.7 b
58.4 ± 3.6 c
373.8
394.9
0.34
0.35
O
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H
N
a
The values of K or K were expressed as mean ± SD from three independent experiments and
i
d
determined via the nonlinear regression analysis using GraphPad Prism software 8.0.
b
K measured by 1,8-ANS displacement assay.
i
c
K measured by ITC
d
Selectivity. Considering the side effect caused by the inhibitory activity of
compounds against the cardiac form of FABP (FABP3), compounds that show a great
binding affinity with FABP4 were examined for their affinities with FABP3 by the 1,8-
ANS displacement assay. BMS309403 was used as a positive control. The data listed
in Table 3 shows that several compounds (3, 6, 10, 11, 17, and 20) have a good
selectivity for FABP4 over FABP3 and their selectivity indexes are superior or similar
to BMS309403. In particular, compound 11 exhibits a 238-fold selectivity for FABP4
over FABP3, which is significantly more selective than BMS309403 (29-fold). In order
to understand the structural basis for the different selectivity index of the compounds,
crystal structures of FABP4 and FABP3 were superimposed (Figure S2). Residue V23
is in close proximity to and forms hydrophobic interactions with the methylamino group
of 11 in the FABP4-11 complex while it is a leucine at the same position in FABP3. A
larger side-chain of L23 in FABP3 might create a steric hindrance for the binding of
the methylamino group, resulting in a weak binding of 11 with FABP3. In contrast,
without the methylamino group, compound 6 has a lower selective index, though it has
a stronger binding affinity with FABP4. Stimulated by the high selectivity index of 11,
the methylamino group was added to the C11 of compound 17, whose binding affinity
is similar to that of 6, to yield compound 20. As expected, 20 does exhibit a much higher
selectivity index for FABP4 over FABP3 compared to 17, while its binding affinity to
FABP4 is slightly less than that of 17. Therefore, the methylamino substitution on C11
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
in the middle phenyl ring plays a key role in improving the selectivity of this series of
compounds toward FABP4 over FABP3. It should be noted that other factors can also
affect the selectivity index. For example, compound 3 without such a methylamino
substitution also has a selectivity index of 117. Taken together, compounds 3, 6, 10, 11,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
7, and 20 show high affinity with FABP4 and favorable selectivity for FABP4 over
FABP3.
In addition, the binding affinities of these inhibitors with FABP5 were examined. It
shows that compounds 6, 14, 15, 17, 18, and 19 have a good binding affinity with
FABP5.
Table 3. Selectivity of the FABP4 Inhibitors toward FABP3 and FABP5.
a
a
FABP3/FABP4 b
Compound
FABP3 K (μM)
FABP5 K (μM)
i
i
BMS309403
7.45 ± 0.28
12.82 ± 1.91
1.00 ± 0.17
1.51 ± 0.02
14.29 ± 0.24
0.74 ± 0.02
1.50 ± 0.04
0.86 ± 0.08
1.40 ± 0.13
0.35 ± 0.01
1.42 ± 0.03
9.62 ± 0.63
9.24 ± 0.21
8.23 ± 0.42
0.88 ± 0.06
1.20 ± 0.12
17.75 ± 0.92
0.49 ± 0.01
0.60 ± 0.04
1.07 ± 0.04
0.82 ± 0.02
0.43 ± 0.01
0.60 ± 0.04
32.04 ± 1.63
29
117
29
34
238
8
3
6
10
11
14
15
12
6
1
1
6
7
37
4
1
1
8
9
9
2
0
165
a
The values of Ki were expressed as mean ± SD based on three independent experiments and
determined via the nonlinear regression analysis using GraphPad Prism software 8.0.
b
Selectivity index was resulted from the K of a compound with FABP3 divided by K or K of the
i
i
d
compound with FABP4.
Anti-inflammatory Effects on THP-1 Macrophages. Genetic ablation or
pharmacological inhibition of FABP4 has been demonstrated to attenuate inflammatory
1
6
ACS Paragon Plus Environment

Page 17 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
responses in macrophages.8,10 Therefore, we selected several compounds with high
affinity with FABP4 and favorable selectivity for FABP4 over FABP3 to explore their
anti-inflammatory effect on the LPS-stimulated THP-1 macrophages. Firstly, the
cytotoxicity was evaluated with a CCK-8 assay, and the result shows that all tested
compounds (3, 6, 10, 11, 17, and 20) at 20 μM and 50 μM have no effect on the cell
viability of THP-1 macrophages (Figure S3A). Next, THP-1 macrophages were
incubated with these compounds at 20 μM for 18 h followed by a LPS stimulation for
another 6 h. As shown in Figure S3B and S3C, the application of LPS resulted in a
dramatic increase in secretion of pro-inflammatory cytokines, MCP-1 and IL-6, and our
compounds as well as BMS309403 attenuated the inflammatory responses by reducing
the level of MCP-1 and IL-6. Among these compounds, 17 exhibits the best in vitro
anti-inflammatory efficacy in the LPS-stimulated THP-1 macrophages and it is thus
chosen for a further study. As illustrated in Figure 5, compound 17 inhibited both MCP-
1 and IL-6 release from the LPS- stimulated THP-1 macrophages in a dose-dependent
manner, and its efficacy is even superior to BMS309403. Taken together, these results
suggest that our compounds possess an anti-inflammatory effect on the LPS-stimulated
THP-1 macrophages and the efficacy of compound 17 is even better than that of
BMS309403.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Effects of BMS309403 and compound 17 on the secretion of pro-inflammatory cytokines,
MCP-1 (A) and IL-6 (B), in the LPS-stimulated THP-1 macrophages. MCP-1 and IL-6 were
measured by the Elisa Kits. Data are given as mean ± standard error of the mean (SEM) of three
independent experiments. Statistical analysis is performed by one-way analysis of variance
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
ANOVA). *P < 0.05, ***P < 0.001 vs LPS group.
Lipolysis Inhibition in 3T3-L1 Adipocytes. As previously reported, ablation or
chemical inhibition of FABP4 decreases lipolysis in adipocytes,27,37 hence, we also
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
investigated the effects of compound 17 on adipocyte lipolysis. Again, the CCK-8 assay
was first carried out to test the cytotoxicity of the compounds. It shows that both
BMS309403 and 17 have no significant cytotoxicity on 3T3-L1 adipocytes at any tested
concentration (Figure S4). The lipolysis experiment was then conducted at various
concentrations of tested compounds and the result is shown in Figure 6. The lipolytic
inhibition by compound 17 was more significant than that of the positive control,
BMS309403. These results further confirm the efficacy of compound 17 on the lipolysis
inhibition in 3T3-L1 adipocytes by binding to FABP4.
Figure 6. Effect of BMS309403 and compound 17 on the release of free glycerol from the forskolin-
stimulated mature 3T3-L1 adipocytes. Data are given as mean ± SEM of at least three independent
experiments, and each is performed in triplicate. Statistical analysis is performed by one-way
ANOVA. *P < 0.05, ***P < 0.001 vs Control group.
In Vitro ADME and in Vivo PK Properties of 17. Compound 17, with an excellent
inhibitory activity against FABP4 both at molecular and cellular level, was progressed
to an in vivo efficacy study. Before that, the metabolic stability of 17 were evaluated
using mouse and human liver microsomes. Table 4 shows that 17 is relatively stable in
1
8
ACS Paragon Plus Environment

Page 19 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the human and mouse liver microsomes. Furthermore, the Caco-2 cell permeability
experiment implies that 17 possesses a good cell permeability in vitro with a low efflux
ratio of 0.21.
The PK properties of compound 17 were preliminary evaluated in mice with an
intravenous (3 mg/kg) and oral (10 and 25 mg/kg) administration. The results are shown
in Table 5. After an oral delivery of 17 at a dose of 25 mg/kg, it exhibits an acceptable
PK profile with a T of 3.82 h, a favorable area under curve (AUC) of 43.77 μg·h/mL,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
/2
a good oral bioavailability of 57.0%, and a good plasma duration (MRT) of 3.1 h.
Together, these data imply that compound 17 is suitable for in vivo efficacy evaluation.
Table 4. Metabolic Stability in Liver Microsomes and Caco-2 Permeability of 17
Parameter
Value
T1/2 (min, Human)
T1/2 (min, Mouse)
86.5
112.2
24.3
18.7
6.23
0.21
Metabolic Stability in
Liver Microsomes
CL (mL/min/gprot, Human)
int
CL (mL/min/gprot, Mouse)
int
−6
P (A−B) (10 cm/s)
app
Caco-2 permeability
Efflux ratio
Table 5. PK Profile of 17 in Mice
p.o.
i.v.
Administration
(
10 mg/kg)
(25 mg/kg)
3.82 ± 0.29
0.25 ± 0.00
12.99 ± 1.78
43.77 ± 5.53
(3 mg/kg)
T1/2 (h)
4.77 ± 0.76
0.25 ± 0.00
3.43 ± 1.66
12.8 ± 3.51
5.76 ± 0.40
Tmax (h)
Cmax (μg/mL)
AUC0-24h (μg•h/mL)
CL (mL/min/kg)
9.38 ± 3.57
5.80 ± 1.86
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Vss (L/kg)
MRT (h)
F (%)
1.65 ± 0.46
4.83 ± 0.69
7.59 ± 1.33
40.9
6.55 ± 0.12
57.0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In Vivo Anti-Inflammatory Efficacy Study. As compound 17 exhibits excellent
anti-inflammatory effects in the LPS-stimulated THP-1 macrophages and a satisfying
performance in liver microsomes stability and PK properties, we next explored its anti-
inflammatory effect in vivo. C57BL/6J mice were intragastrically administrated with
BMS309403 or compound 17 at a dose of 50 mg/kg 1 h before a single intraperitoneal
injection of LPS at a dose of 5 mg/kg. Concentrations of MCP-1 and IL-6 in serum
were measured 5 h after LPS injection. As shown in Figure 7, LPS treatment
dramatically increases the level of inflammatory cytokines such as MCP-1 and IL-6 in
the serum of vehicle mice, while an oral administration of BMS309403 or compound
1
7 prominently downregulate the secretion of MCP-1 and IL-6. Consistent with the in
vitro results noted above, the anti-inflammatory efficacy of compound 17 in vivo is also
more significant than that of BMS309403, demonstrating the anti-inflammatory
potential of compound 17.
Figure 7. Effects of BMS309403 (50 mg/kg) and compound 17 (50 mg/kg) on the secretion of pro-
inflammatory cytokines, MCP-1 (A) and IL-6 (B), in the serum of LPS-treated C57BL/6J mice (n
= 8 mice/group). MCP-1 and IL-6 were measured by the Elisa Kits. Data are given as mean ± SEM
and statistical analysis is performed by one-way ANOVA. ** P < 0.01, ***P < 0.001 vs LPS group.
Efficacy of Compound 17 for Multi-Organ Protection. To explore the effect of
2
0
ACS Paragon Plus Environment

Page 21 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound 17 on organ damage, C57BL/6J mice were intragastrically administrated
with BMS309403 (25 mg/kg) or 17 (10, 25 and 50 mg/kg) once a day for consecutive
four days prior to a single LPS intraperitoneal injection at a dose of 10 mg/kg. As shown
in Figure 8A, the level of liver damage markers, serum aspartate aminotransferase (AST)
as well as alanine transaminase (ALT), and of a kidney damage marker, blood urea
nitrogen (BUN), are significantly increased after the injection of LPS compared with
the vehicle mice. In contrast, administration of compound 17 decreases the level of
these biomarkers in a dose-dependent manner, which is equivalent to or even better
than BMS309403, indicating a beneficial effect of compound 17 on hepatorenal
function. In addition, the histological staining results can clearly interpret organ damage
which is manifested as the infiltration of inflammatory cells and damaged cells.
Compared with the LPS-treated mice, BMS309403 and compound 17-treated mice
have a limited extent of damage in key organs such as lung, liver, kidney, and spleen,
which are close to those of healthy animals (Figure 8B). Taken together, these results
indicate that compound 17 is able to effectively attenuate a systemic inflammatory
damage and acts as a good candidate for the treatment of inflammation-related diseases.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 8. Compound 17 effectively protects mice from the LPS-induced multiorgan damage. (A)
BMS309403 (25 mg/kg) or compound 17 (10, 25 and 50 mg/kg) administration decreases the level
of tissue damage biomarkers (n = 8 mice/group). (B) BMS309403 and compound 17 show an
efficient protection against the LPS-induced injury in different organs. (H&E staining,
magnification: ×200). Each image is representative of four mice at least. Data are given as mean ±
SEM and statistical analysis is performed by one-way ANOVA. * P < 0.05, ** P < 0.01, ***P <
0
.001, each vs LPS group.
Chemistry. The preparation of compounds 3-20 is depicted in Schemes 1-3.
Commercially available compounds 1a or 1c were first coupled to the phenylboronic
acid with different substituents by the Suzuki-Miyaura coupling reaction to yield
compounds 3a and 3d-3r. In addition, compounds 3b and 3c were synthesized by a
nucleophilic substitution reaction of aniline and phenol with 1-fluoro-2-nitrobenzene
(1b). Compound 1j was prepared by a methylation of 2,3-dichloro-4-nitroaniline (1i)
using methyl iodide, and compound 1k was a byproduct of this reaction. Compound 1l
was synthesized by a procedure similar to 1j. Different nitrobenzene derivatives 3a-3s
2
2
ACS Paragon Plus Environment

Page 23 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were then reduced by iron to afford the corresponding amines 4a-4r. The target
compounds were synthesized from these amines (4a-4r) by a palladium-catalyzed
Buchwald-Hartwig amination followed by the hydrolysis of esters 6a-6r.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. Synthesis of Compounds 3~5.
O
O
Cl
Cl
B(OH)2
O
O
HO
O
5
a
H
N
H
N
O2N
2a
O2N
H2N
a
b
c
d
1a
3a
4a
6a
3
NH2
OH
O
O
2b
Cl
O
O
HO
O
F
X
X
X
X
H
5
a
H
O2N
2c
2
H N
N
N
O2N
b
c
d
e
1b
3b X = NH
3c X = O
4b X = NH
4c X = O
6b X = NH
6c X = O
4
5
X = NH
X = O
o
Reagents and conditions: (a) Pd(PPh ) , K CO , dioxane, H O, 15 h, 95 C; (b) Fe, HAc, NH Cl,
3
4
2
3
2
4
o
o
EtOH, 50 C; (c) Pd (dba) , BINAP, Cs CO , Toluene, 6 h, 90 C; (d) NaOH, MeOH, H O, 1 h, 70
2
3
2
3
2
o
o
C; (e) NaH, THF, 2 h, 75 C.
Scheme 2. Synthesis of Compounds 6~10 and 14~19.
R1
R1
R1
Cl
I
B(OH)2
H2N
X
O2N
X
O2N
X
O2N
Br
or
a
b
3d ~ 3n
4d ~ 4n
1c ~ 1g
1h
R1 = Ph; X = Cl
R1 = Ph; X = Br
4d R1 = Ph; X = Cl
3
d
e
1
1
1
c
d
e
X = F
3
4e
4f
R1 = Ph; X = Br
R1 = Ph; X = CF3
X = Cl
X = CF3
3f R1 = Ph; X = CF3
3g
3h
3i
R1 = Ph; X = CH3
R1 = Ph; X = OCH3
R1 = 1H-indazol-5-yl; X = Cl
4g
4h
4i
R1 = Ph; X = CH3
R1 = Ph; X = OCH3
1f X = CH3
1g
X = OCH3
R1 = 1H-indazol-5-yl; X = Cl
R1 = benzo[c][1,2,5]oxadiazol-5-yl; X = Cl
R1 = benzo[d][1,3]dioxol-5-yl; X = Cl
R1 = 2,3-dihydrobenzofuran-5-yl; X = Cl
1
3j R1 = benzo[c][1,2,5]oxadiazol-5-yl; X = Cl
4j
3
3
3
3
k
R1 = benzo[d][1,3]dioxol-5-yl; X = Cl
R1 = 2,3-dihydrobenzofuran-5-yl; X = Cl
R1 = benzofuran-5-yl; X = Cl
4k
4l
l
m
n
4m
4n
R
= benzofuran-5-yl; X = Cl
R1 = naphthalen-2-yl; X = Cl
R1 = naphthalen-2-yl; X = Cl
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O
O
Cl
O
O
HO
O
R1
R1
H
H
5
a
N
X
N
X
c
d
6d ~ 6n
6 ~ 10; 14 ~ 19
R1 = 1H-indazol-5-yl; X = Cl
14
15 R1 = benzo[c][1,2,5]oxadiazol-5-yl; X = Cl
16 R1 = benzo[d][1,3]dioxol-5-yl; X = Cl
17 R1 = 2,3-dihydrobenzofuran-5-yl; X = Cl
18 R1 = benzofuran-5-yl; X = Cl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
R1 = Ph; X = Cl
R1 = Ph; X = Br
R1 = Ph; X = CF3
6j R1 = benzo[c][1,2,5]oxadiazol-5-yl; X = Cl
R1 = Ph; X = Cl
R1 = Ph; X = Br
R1 = Ph; X = CF3
R1 = Ph; X = CH3
6
6
6
d
e
f
6
7
8
9
R1 = benzo[d][1,3]dioxol-5-yl; X = Cl
R1 = 2,3-dihydrobenzofuran-5-yl; X = Cl
6k
6l
6g R1 = Ph; X = CH3
6m R1 = benzofuran-5-yl; X = Cl
6n R1 = naphthalen-2-yl; X = Cl
R1 = Ph; X = OCH3
R1 = 1H-indazol-5-yl; X = Cl
10 R1 = Ph; X = OCH3
6
h
i
6
19 R1 = naphthalen-2-yl; X = Cl
o
Reagents and conditions: (a) Pd(PPh ) , K CO , dioxane, H O, 15 h, 95 C; (b) Fe, HAc, NH Cl,
3
4
2
3
2
4
o
o
EtOH, 50 C; (c) Pd (dba) , BINAP, Cs CO , Toluene, 6 h, 90 C; (d) NaOH, MeOH, H O, 1 h, 70
2
3
2
3
2
oC.
Scheme 3. Synthesis of Compounds 11~13 and 20
R1
R1
Cl
Cl
R1
B(OH)2
O2N
Cl
O2N
Cl
O N
Cl
H N
Cl
2
2
f
a
b
R2
R2
R2
NH2
1j ~ 1l
3
o ~ 3r
4o ~ 4r
1i
R1 = Ph; R2 = NHCH3
R1 = Ph; R2 = N(CH3)2
3q R1 = Ph; R2 = NHC2H5
R1 = 2,3-dihydrobenzofuran-5-yl; R2 = NHCH3
4o R1 = Ph; R2 = NHCH3
4p R1 = Ph; R2 = N(CH3)2
4q R1 = Ph; R2 = NHC H
3
o
p
2
1j
R
R
R
= NHCH3
= N(CH3)2
= NHC2H5
3
2
2
1
1
k
l
2
5
R
1
=
2
,
3
-
d
i
h
y
d
r
o
b
e
n
z
o
f
u
r
a
n
-
5
-
y
l
;
R
2
=
N
H
C
H
3
3r
4
r
O
O
Cl
O
O
HO
O
R1
R1
H
H
5a
N
Cl
N
Cl
d
c
R2
R2
6
o ~ 6r
11 ~ 13, 20
R1 = Ph; R2 = NHCH3
11
12 R1 = Ph; R2 = N(CH3)2
13 R1 = Ph; R2 = NHC2H5
R1 = Ph; R2 = NHCH3
R1 = Ph; R2 = N(CH3)2
R1 = Ph; R2 = NHC H
6
o
p
q
r
6
6
6
2
5
R1 = 2,3-dihydrobenzofuran-5-yl; R2 = NHCH3
20
R1 = 2,3-dihydrobenzofuran-5-yl; R2 = NHCH3
o
Reagents and conditions: (a) Pd(PPh ) , K CO , dioxane, H O, 15 h, 95 C; (b) Fe, HAc, NH Cl,
3
4
2
3
2
4
o
o
EtOH, 50 C; (c) Pd (dba) , BINAP, Cs CO , Toluene, 6 h, 90 C; (d) NaOH, MeOH, H O, 1 h, 70
2
3
2
3
2
o
o
C; (f) C H I or CH I, CH CN, 2 d, 85 C.
2
5
3
3
 CONCLUSIONS
Here we reported the efficient discovery of a new class of highly potent, selective
FABP4 inhibitors from an interesting fragment hit which displays multiple binding
poses in the ligand binding pocket of FABP4. Both crystal structures and MD
simulations revealed the multiple binding poses of the fragment. Based on three
distinctive but partially overlapped binding poses of this fragment, compound 3 was
2
4
ACS Paragon Plus Environment

Page 25 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
rationally designed, synthesized and evaluated, demonstrating a ~460-fold
improvement in the binding affinity compared to the initial fragment. The binding
affinity of compound 3 with FABP4 is superior to the well-known inhibitor,
BMS309403. The protein-ligand structures determination together with the binding
affinity precisely measured by ITC renders us to rapidly gain the sufficient knowledge
of SAR and to discover more potent and selective FABP4 inhibitors. As a result, the
representative compound (17) with a ~1300-fold improvement in the binding affinity
compared to the initial fragment (1), is able to potently inhibit the production of MCP-1
and IL-6 in THP-1 macrophages as well as the lipolysis in 3T3-L1 adipocytes, with a
better performance than BMS309403. Moreover, compound 17 displays favorable PK
profiling for the in vivo study, and exhibits a better anti-inflammatory effect and
effective protection against multi-organ damage in the LPS-induced inflammatory mice
model compared to BMS309403. Given the good efficacy in vivo, application of 17 in
other FABP4-associated indications such as breast cancer and non-alcoholic
steatohepatitis will be explored in the future. Overall, the present study demonstrates
that an in-depth exploration of fragment binding to the target protein plays a significant
role in efficient discovery of novel and potent ligands. It is noteworthy that removing
endogenous fatty acids from the purified FABP4 is crucial for elucidation of the
multiple binding poses of the fragment, a weak binder, and for accurate determination
of the binding affinity of highly potent FABP4 inhibitors by ITC.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0

EXPERIMENTAL SECTION
Materials and Methods for Chemistry. All chemical reagents and compounds
included in our in-house library for screening were used as supplied by standard
suppliers without further purification. All reactions were monitored by the thin-layer
chromatography (TLC) and visualization was achieved by using ultraviolet light (254
nm). The column chromatography was carried out using a silica gel (300−400 mesh) in
1
13
glass columns under proper pressure. H NMR and C NMR spectra were recorded on
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a BRUKER AVANCE NEO 500 at 500 and 126 MHz, respectively. Coupling constants
(J) are expressed in hertz. Chemical shifts (δ) of NMR spectra are reported in parts per
million (ppm) units. Mass spectra (MS) of compounds were measured using a Thermo
Fisher FINNIGAN LTQ spectrometer. Purity of all compounds used for biological
testing was determined to be > 95% by HPLC analysis. Analysis was performed on a
SHIMADZU LC-20AD HPLC system under the following analytical method: column,
SHIMADZU Shim-pack GIST C18 (5 μm, 4.6 mm × 250 mm); solvent A: water
containing 1‰ TFA; solvent B: acetonitrile; gradient, 60% B to 100% B over 15 min,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
00% B for 5 min, 100% B to 60% B over 5 min; flow rate, 1 mL/min; detective
wavelength, 254 nm; column temperature, 10 °C.
Synthesis Procedure for 2-([1,1'-biphenyl]-2-ylamino)benzoic acid (3). 2-Nitro-
1
,1'-biphenyl (3a). 1-chloro-2-nitrobenzene (1a) (0.1 g, 0.635 mmol), phenylboronic
acid (2a) (0.077 g, 0.635 mmol), Pd(PPh ) (0.059 g, 0.051 mmol) and K CO (0.263
3
4
2
3
g, 1.90 mmol) were taken up in 1,4-dioxane (8 mL) and water (2 mL) in a flask and
o
purged with N for 5 min. Then, the mixture was heat to 95 C for 15 h. After cooling
2
to room temperature, the mixture was filtered by celite and the filtrate was extracted
with ethyl acetate, washed with brine, and the organic phase was dried over Na SO .
2
4
The solvent was removed and the residue was purified by a flash column
chromatography on silica (80% PE in ethyl acetate ∼50% PE in ethyl acetate) to give a
1
yellow solid 3a (0.102 g, 80.67% yield). H NMR (500 MHz, Chloroform-d) δ 7.63 –
7
.58 (m, 4H), 7.44 (dd, J = 8.4, 7.0 Hz, 3H), 7.40 – 7.31 (m, 2H).
1,1'-Biphenyl]-2-amine (4a). Iron powder (561 mg, 10.04 mmol), NH Cl (53.7
[
4
mg, 1.004 mmol) and AcOH (0.115 mL, 2.008 mmol) were added into water (2 ml) and
stirred at 50 °C for 15 min. A solution of 3a (200 mg, 1.004 mmol) in MeOH (3 ml)
was added into the above solution quickly, and stirring was continued at 50 °C for 25
min. Then the reaction solution was alkalized to pH = 9 with an aqueous solution of
sodium carbonate. The reaction mixture was filtered, the formed solid was washed with
water and ethyl acetate. The combined filtrate was extracted with ethyl acetate. Then
the combined organic layers were washed with brine, dried over anhydrous Na SO ,
2
4
2
6
ACS Paragon Plus Environment

Page 27 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
filtered, and concentrated under reduced pressure to give the crude 4a (180 mg), which
was directly used to the next step without further purification.
Methyl 2-([1,1'-biphenyl]-2-ylamino)benzoate (6a). Under nitrogen a solution of
Pd (dba) (14.23 mg, 0.016 mmol), BINAP (15.48 mg, 0.025 mmol) and [1,1'-
2
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
biphenyl]-2-amine (4a) (57.8 mg, 0.342 mmol) derivative in dry toluene (5 mL) was
stirred for 5 min. To this deep red solution, the corresponding methyl 2-chlorobenzoate
(
53 mg, 0.311 mmol) and Cs CO (152 mg, 0.466 mmol) were added and stirring at 90
2 3
°
C. The reaction was continued until the TLC showed the reaction to be completed
(usually 12 hours). The solid was filtered off and the filtrate concentrated in vacuo to
afford a residue that was subjected to a chromatography (silica gel, PM/EtOAc: 100/1-
1
4
1
0/1) to yield 6a, methyl 2-([1,1'-biphenyl]-2-ylamino)benzoate (40 mg, 0.132 mmol,
1
2.4 % yield). H NMR (500 MHz, DMSO-d ) δ 8.03 – 7.97 (m, 1H), 7.90 (dd, J = 7.5,
6
.5 Hz, 1H), 7.65 (dd, J = 7.5, 1.6 Hz, 1H), 7.48 – 7.41 (m, 2H), 7.39 – 7.29 (m, 7H),
7.29 – 7.21 (m, 2H), 3.89 (s, 3H).
2-([1,1'-Biphenyl]-2-ylamino)benzoic acid (3). To a solution of 6a (50 mg, 0.165
mmol) in methanol (4 mL) an aqueous solution of sodium hydroxide (5%, 4 mL) was
o
added and the reaction mixture was stirred for 1 h at 70 C. After neutralization with
diluted hydrochloric acid (6 N), the solvents were removed and the product was washed
with water and dried to yield the corresponding acid (3, 44 mg, 0.152 mmol, 92% yield).
-
-
HPLC purity: 99.62% (RT = 12.11 min); ESI-MS: 288.5 [M-H] , 599.4 [2M-H+Na] ;
1
H NMR (500 MHz, DMSO-d ) δ 12.83 (s, 1H), 9.56 (s, 1H), 7.82 (dd, J = 7.9, 1.7 Hz,
6
1
H), 7.48 (dd, J = 8.1, 1.3 Hz, 1H), 7.41 – 7.30 (m, 8H), 7.22 (td, J = 7.4, 1.3 Hz, 1H),
1
3
7
.11 (dd, J = 8.5, 1.1 Hz, 1H), 6.73 – 6.70 (m, 1H); C NMR (126 MHz, Methanol-d )
4
δ 170.14, 148.36, 139.27, 137.80, 136.66, 133.58, 131.76, 130.80, 128.68, 127.98,
1
27.82, 126.87, 124.16, 123.94, 116.42, 113.28, 112.04.
Synthesis Procedure for 2-((2-(phenylamino)phenyl)amino)benzoic acid (4). 2-
nitro-N-phenylaniline (3b). To a solution of NaH (170 mg, 4.25 mmol) (60 percent
dispersion in mineral oil) in THF (8 mL) was slowly added aniline (198 mg, 2.126
mmol). After 15 minutes of stirring, 1-fluoro-2-nitrobenzene (300 mg, 2.126 mmol)
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
was added and the resulting solution was heated to 75 °C for 1 hour. The reaction
mixture was cooled to room temperature, quenched with water (5 mL) and extracted
with EtOAc. The combined organic phases were washed with brine (5 mL), dried over
Na SO and concentrated. The crude residue was purified by the flash chromatography
2
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(eluent: EtOAc/PE = 50/1) to give the desired product, 2-nitro-N-phenylaniline (3b,
1
4
30 mg, 2.007 mmol, 94 % yield). H NMR (500 MHz, DMSO-d ) δ 9.42 (s, 1H), 8.15
6
(dd, J = 7.5, 1.5 Hz, 1H), 7.39 – 7.35 (m, 1H), 7.32 – 7.25 (m, 2H), 7.24 – 7.21 (m, 2H),
7
.19 – 7.16 (m, 1H), 7.02 – 6.98 (m, 1H), 6.90 (dd, J = 7.5, 1.5 Hz, 1H).
Compounds 4b, 6b and 4 were prepared following the synthetic procedure similar
to
that
of
compounds
4a,
6a
and
3,
respectively.
2-((2-
(phenylamino)phenyl)amino)benzoic acid (4, white powder, 32 mg, 72 % yield for the
-
last step). HPLC purity: 95.13% (RT = 11.73 min); ESI-MS: 303.6 [M-H] , 629.4
-
1
[
2M-H+Na] ; H NMR (500 MHz, DMSO-d ) δ 12.85 (s, 1H), 9.43 (s, 1H), 7.86 (dd, J
6
= 8.0, 1.7 Hz, 1H), 7.57 (s, 1H), 7.38 – 7.33 (m, 2H), 7.27 – 7.25 (m, 1H), 7.13 (t, J =
1
3
7.8 Hz, 2H), 7.10 – 6.99 (m, 3H), 6.89 (d, J = 8.0 Hz, 2H), 6.76 – 6.71 (m, 2H); C
NMR (126 MHz, DMSO-d ) δ 170.16, 147.98, 145.01, 137.00, 134.38, 133.04, 132.17,
6
129.35, 124.60, 123.15, 122.75, 121.15, 119.63, 117.28, 116.83, 114.04, 113.15.
2-((2-Phenoxyphenyl)amino)benzoic acid (5). The title compound (white powder,
2
3 mg, 92% yield for the last step) was prepared analogous to compound 4. HPLC
-
- 1
purity: 99.03% (RT = 9.47 min); ESI-MS: 304.5 [M-H] , 631.4 [2M-H+Na] ; H NMR
(
500 MHz, DMSO-d ) δ 12.41 (s, 1H), 9.82 (s, 1H), 7.85 (dd, J = 7.9, 1.5 Hz, 1H), 7.59
6
–
7
7.57 (m, 1H), 7.43 – 7.40 (m, 1H), 7.33 – 7.29 (m, 3H), 7.23 – 7.19 (m, 1H), 7.11 –
1
3
.05 (m, 3H), 6.92 (d, J = 8.0 Hz, 2H), 6.79 (t, J = 7.5 Hz, 1H); C NMR (126 MHz,
DMSO-d ) δ 170.26, 157.53, 147.54, 146.59, 134.50, 133.19, 132.23, 130.34, 125.29,
6
1
23.98, 123.40, 121.64, 121.30, 118.19, 117.45, 114.32, 113.73.
-((6-chloro-[1,1'-biphenyl]-2-yl)amino)benzoic acid (6). The title compound
2
(white powder, 40 mg, 92% yield for the last step) was prepared analogous to
-
compound 3. HPLC purity: 97.68% (RT = 19.46 min); ESI-MS: 322.1 [M-H] , 667.1
- 1
[
2M-H+Na] ; H NMR (500 MHz, DMSO-d ) δ 10.40 (s, 1H), 7.81 (dd, J = 7.7, 1.7 Hz,
6
2
8
ACS Paragon Plus Environment

Page 29 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
H), 7.43 (t, J = 7.4 Hz, 3H), 7.37 – 7.35 (m, 1H), 7.29 (t, J = 8.1 Hz, 1H), 7.26 – 7.20
1
3
(m, 3H), 7.18 (d, J = 8.0 Hz, 1H), 7.13 (d, J = 8.0 Hz, 1H), 6.72 – 6.69 (m, 1H); C
NMR (126 MHz, DMSO-d ) δ 169.43, 145.98, 140.53, 135.47, 134.04, 133.43, 132.95,
6
1
31.64, 129.54, 129.25, 128.68, 127.99, 123.33, 118.86, 118.06, 114.19, 113.21.
-((6-bromo-[1,1'-biphenyl]-2-yl)amino)benzoic acid (7). The title compound
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
(white powder, 31 mg, 91% yield for the last step) was prepared analogous to
-
compound 3. HPLC purity: 96.18% (RT = 13.38 min); ESI-MS: 366.2 [M-H] , 734.2
- 1
[2M-H] ; H NMR (500 MHz, DMSO-d ) δ 9.21 (s, 1H), 7.78 (dd, J = 7.9, 1.8 Hz, 1H),
6
7
.53 (dd, J = 8.1, 0.8 Hz, 1H), 7.47 – 7.37 (m, 5H), 7.32 – 7.28 (m, 2H), 7.22 – 7.20
13
(m, 2H), 6.80 – 6.77 (m, 1H); C NMR (126 MHz, DMSO-d ) δ 169.95, 146.42, 141.03,
6
1
1
37.97, 135.31, 134.39, 132.12, 130.11, 129.90, 129.15, 128.45, 126.87, 124.52,
19.79, 118.53, 114.64, 113.99.
2
-((6-(trifluoromethyl)-[1,1'-biphenyl]-2-yl)amino)benzoic acid (8). The title
compound (white powder, 10 mg, 82% yield for the last step) was prepared analogous
to compound 3. HPLC purity: 96.62% (RT = 13.03 min); ESI-MS: 356.2 [M-H]-,
-
1
712.5 [2M-H] ; H NMR (500 MHz, DMSO-d ) δ 12.98 (s, 1H), 9.11 (s, 1H), 7.84 –
6
7.76 (m, 2H), 7.58 (t, J = 8.0 Hz, 1H), 7.51 – 7.49 (m, 1H), 7.46 – 7.40 (m, 4H), 7.30
1
3
(dd, J = 8.5, 1.1 Hz, 1H), 7.24 – 7.22 (m, 2H), 6.82 – 6.79 (m, 1H); C NMR (126
MHz, DMSO-d ) δ 169.87, 146.18, 141.17, 135.02, 134.53, 132.99, 132.89, 132.16,
6
1
29.98, 129.22, 128.81, 128.61, 124.11, 119.96, 119.92, 118.79, 114.55, 113.95.
-((6-methyl-[1,1'-biphenyl]-2-yl)amino)benzoic acid (9). The title compound
2
(white powder, 30 mg, 87% yield for the last step) was prepared analogous to
-
compound 3. HPLC purity: 99.45% (RT = 13.33 min); ESI-MS: 302.2 [M-H] , 604.6
-
1
[2M-H] ; H NMR (500 MHz, DMSO-d ) δ 9.11 (s, 1H), 7.76 (dd, J = 7.9, 1.8 Hz, 1H),
6
7
3
.42 (t, J = 7.5 Hz, 2H), 7.36 – 7.32 (m, 3H), 7.27 (t, J = 7.7 Hz, 1H), 7.21 – 7.16 (m,
1
3
H), 7.05 (d, J = 7.4 Hz, 1H), 6.71 – 6.68 (m, 1H), 2.02 (s, 3H); C NMR (126 MHz,
DMSO-d ) δ 169.99, 147.69, 138.77, 138.01, 137.53, 135.55, 134.44, 132.11, 129.73,
6
1
29.13, 128.14, 127.74, 125.31, 119.35, 117.42, 114.02, 112.72, 21.04.
-((6-methoxy-[1,1'-biphenyl]-2-yl)amino)benzoic acid (10). The title compound
2
2
9
ACS Paragon Plus Environment