Structure-Guided Discovery of Novel, Potent, and Orally Bioavailable Inhibitors of Lipoprotein-Associated Phospholipase A2

Article abstract of DOI:10.1021/acs.jmedchem.7b01530

Lipoprotein-associated phospholipase A2 (Lp-PLA2) is a promising therapeutic target for atherosclerosis, Alzheimer's disease, and diabetic macular edema. Here we report the identification of novel sulfonamide scaffold Lp-PLA2 inhibitors derived from a relatively weak fragment. Similarity searching on this fragment followed by molecular docking leads to the discovery of a micromolar inhibitor with a 300-fold potency improvement. Subsequently, by the application of a structure-guided design strategy, a successful hit-to-lead optimization was achieved and a number of Lp-PLA2 inhibitors with single-digit nanomolar potency were obtained. After preliminary evaluation of the properties of drug-likeness in vitro and in vivo, compound 37 stands out from this congeneric series of inhibitors for good inhibitory activity and favorable oral bioavailability in male Sprague-Dawley rats, providing a quality candidate for further development. The present study thus clearly demonstrates the power and advantage of integrally employing fragment screening, crystal structures determination, virtual screening, and medicinal chemistry in an efficient lead discovery project, providing a good example for structure-based drug design.

Full text of DOI:10.1021/acs.jmedchem.7b01530

Subscriber access provided by READING UNIV
Article
Structure-guided Discovery of Novel, Potent and Orally
Bioavailable Inhibitors of Lipoprotein-associated Phospholipase A2
Qiufeng Liu, Fubao Huang, Xiao-Jing Yuan, Kai Wang, Yi Zou, Jianhua Shen, and Yechun Xu
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01530 • Publication Date (Web): 01 Dec 2017
Downloaded from http://pubs.acs.org on December 1, 2017
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 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structure-guided Discovery of Novel, Potent and Orally
Bioavailable Inhibitors of Lipoprotein-associated Phospholipase A2
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
§
§
§‡
Qiufeng Liu ‡, Fubao Huang ‡, Xiaojing Yuan , Kai Wang ，Yi Zou , Jianhua Shen , and


*
Yechun Xu*
CAS Key Laboratory of Receptor Research, 
State Key Laboratory of Drug Research, Drug
Discovery and Design Center, Shanghai Institute of Materia Medica, Chinese Academy of Sciences
(CAS), Shanghai 201203, China.
School of Life Science and Technology, ShanghaiTech University, Shanghai 200031, China
§
University of Chinese Academy of Sciences, Beijing, 100049, China
KEYWORDS Lipoprotein-associated phospholipase A2 Inhibitor, Sulfonamide scaffold,
Fragment-based
lead
discovery,
Structure-based
hit-to-lead
optimization,
Pharmacokinetics
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 57
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
Lipoprotein-associated phospholipase A2 (Lp-PLA2) is a promising therapeutic target for
atherosclerosis, Alzheimer’s disease and diabetic macular edema. Here we report the
identification of novel sulfonamide scaffold Lp-PLA2 inhibitors derived from a relatively
weak fragment. Similarity searching on this fragment followed by molecular docking
leads to the discovery of a micromolar inhibitor with 300-fold potency improvement. By
the application of a structure-guided design strategy, a successful hit-to-lead optimization
was achieved and a number of Lp-PLA2 inhibitors with single-digit nanomolar activity
were obtained. After preliminary evaluation of the properties of drug-likeness in vitro and
in vivo, compound 37 outstands from this congeneric series of inhibitors for good
inhibitory activity and favorable oral bioavailability in male Sprague-Dawley rats,
providing a quality candidate for further development. Our current study thus clearly
demonstrates the power and advantage of integrally employing fragment screening,
complex crystal structures, virtual screening, and medicinal chemistry in an efficient lead
discovery project, providing a good example for structure-based drug design.
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
ACS Paragon Plus Environment

Page 3 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Lipoprotein-associated phospholipase A2 (Lp-PLA2) belongs to the phospholipase A2
1
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
superfamily. The 3D structure of Lp-PLA2 adopts a canonical lipase α/β-hydrolase
conformation and contains a large substrate binding pocket starting from the catalytic
2
triad (Ser273, His351 and Asp296) to lipoprotein binding helices. In human plasma,
Lp-PLA2 mainly associates with low density lipoproteins (LDL) and hydrolyzes the sn-2
3
4
ester bond in platelet activating factor (PAF) and other recognized phospholipids.
Specifically, Lp-PLA2 can hydrolyze oxidized modified phospholipids into several
pro-inflammatory stimuli, such as lysophosphatidylcholine (lyso-PC) and oxidized
non-esterified free fatty acids (ox-NEFA), both of which trigger a series of inflammatory
responses. Both Lp-PLA2 activity and mass have shown recurrent association with risk
5
of coronary heart disease (CHD) in a meta-analysis. And Lp-PLA2 was demonstrated to
6
be an independent biomarker of poor prognosis for coronary artery disease. Selective
inhibition of this enzyme showed protective effects on blood-brain barrier (BBB)
dysfunction in a pig model.7 Moreover, Lp-PLA2 was involved in retinal
8
vasopermeability and the pathogenesis of diabetic macular edema (DME). Accordingly,
inhibition of Lp-PLA2 may be a potential therapy for clinical intervention on
atherosclerosis inflammation, Alzheimer’s disease and DME, and a promising approach
to address the underlying pathological mechanisms.
Owing to the interest in Lp-PLA2 as a promising therapeutic target, several classes of
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 57
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
9
Lp-PLA2 inhibitors have been reported (Figure 1). The majority of them occupied the
1
0
11
catalytic site of Lp-PLA2, such as lactams (1) , oximes (2) , amides of xanthurenic acid
12
13
14
(
3) , carbamates (4) , pyrimidone derivatives (5 and 7) , or imidazo[1,2-a]pyrimidine
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
5
derivatives (6) . Darapladib (5) is the most advanced Lp-PLA2 inhibitor in the clinic,
though it failed to meet the primary end points in two phase III clinical trials focusing on
coronary heart disease.16 This raises the concern about Lp-PLA2 if it is a target for
development of therapeutic agents or it may just serve as a biomarker of vascular
inflammation.9, 17 However, in a 3-month placebo-controlled study, darapladib showed
modest improvement in vision and macular edema in DME patients.18 In addition,
Rilapladib, another Lp-PLA2 inhibitor, is in the phase II clinical trial for the treatment of
1
9
Alzheimer’s disease. Therefore, inhibitors with different scaffolds should be explored to
better investigate the clinical intervention on Lp-PLA2.
Figure 1. Selected examples of reported Lp-PLA2 inhibitors.
Previously, we have reported imidazo[1,2-a]pyrimidine derivatives as potent and
1
5
orally bioavailable Lp-PLA2 inhibitors. A further study led to the identification of the
4
ACS Paragon Plus Environment

Page 5 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pyrimidone inhibitors, which significantly degraded the retinal thickening in a
Sprague-Dawley rat model of DME.20 We also disclosed the crystal structures of
Lp-PLA2 in complex with two inhibitors, darapladib and a pyrimidone derivative, and
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
b
shed lights on the binding mechanisms of different inhibitors with Lp-PLA2 , providing
an important prerequisite for the accomplishment of a successful fragment-based lead
discovery (FBLD) project. FBLD has matured over the last decade to a reliable and
efficient method, and has been generating promising leads for subsequent medicinal
21
chemistry optimization. Recently, Astex together with Glaxosmithkline (GSK) applied
2
2
23
the FBLD strategy to identify novel Lp-PLA2 inhibitors (compound 8 and 9 ).
Aiming to identify potent Lp-PLA2 inhibitors with new scaffolds, we screened our
in-house 500-fragment library with the PAF enzymatic assay at a high concentration. A
sulfonamide fragment 10 with millimolar potency was obtained, and the binding mode
was validated by its co-crystal structure with recombinant human Lp-PLA2 (rhLp-PLA2)
(Figure 2). To the best of our knowledge, no sulfonamide scaffold Lp-PLA2 inhibitors
were known at the time we initiated our work. With such a novel binding moiety in hand,
we conducted the fragment evolution in order to obtain highly potent Lp-PLA2 inhibitors.
With an invaluable aid from the crystal structure of rhLp-PLA2 in complex with fragment
10, a virtual screening strategy was employed. Firstly, a similarity search was performed,
and compounds containing fragment 10 were subsequently subjected to a docking study.
The selected hits were purchased and assessed with the PAF enzymatic assay, resulting in
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 57
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 micromolar inhibitor 11. Guided by the crystal structures of rhLp-PLA2 bound with
these novel compounds, several low nanomolar inhibitors were obtained through a
hit-to-lead optimization. Herein, we will clearly illustrate how this efficient lead
discovery, from a rather weak binder to a series of nanomolar inhibitors, is accomplished.
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
RESULTS AND DISCUSSION
Fragment screening and validation. Fragments characterized by small size and
simple structure could provide much broader chemical diversity, afford novel binding
moieties of interested targets, and possess good ligand efficiency. After screening our
in-house 500-fragment library, a small sulfonamide fragment 10 (MW= 207; IC50
=
~1mM, Figure 2) was obtained and it afforded efficient binding to rhLp-PLA2 (LE ≈
24
0
.30 ). The crystal structure shows that it, exhibiting with unambiguous electron density
(Figure S1A), occupies the center of the ligand binding pocket and bridges residues
2
a
25
comprising the oxyanion hole to those near the lipoprotein binding helices (Figure 2),
thereby exerting an influence on the full binding pocket of Lp-PLA2. The
methylsulfonamide tail of fragment 10 interacts with the catalytic sub-pocket by forming
multiple direct and/or water-bridged hydrogen-bonds (H-bonds) with surrounding
residues. The amide nitrogen in the side-chain of Gln352 serves as a H-bond donor to one
of the sulfonamide oxygen (O1). Meanwhile carbonyl oxygen in this side chain is linked
to residues (Phe357 and Thr358) near the lipoprotein binding helices via a water
molecule (W2). The second sulfonamide oxygen (O2) bonds the hydroxyl group of the
6
ACS Paragon Plus Environment

Page 7 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
catalytic residue Ser273 and another crucial water molecule (W1) located in the oxyanion
2
a
hole . The last H-bond is established between the main-chain carbonyl oxygen of
Leu153 and the NH group linked to the sulfonamide tail. In addition, the
3,4-difluorophenyl group of fragment 10 is buried into a hydrophobic sub-pocket
consisting of residues Leu107, Phe110, Gly154, Ala155, and Leu159, contributing to the
efficient binding of this fragment to Lp-PLA2.
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 The crystal structure of rhLp-PLA2 in complex with fragment 10 (PDB ID:
5
YE8). (A) An overall view of the complex structure. Lp-PLA2 is shown as gray cartoon.
Two lipoprotein binding helices (residues 114-126 and 362-369) are colored in orange. (B)
A close-up view of fragment 10 within the binding pocket which is presented by
molecular surface. H-bonds are represented by black dashed lines. Fragment 10 (yellow)
and the surrounding residues (cyan) are shown as sticks. Two water molecules (W1 and
W2) are shown as red spheres.
Virtual screening and hit identification. Inspired by the elegant binding of
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 57
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
fragment 10 with the protein, we carried out fragment optimization to enhance the
binding affinity to Lp-PLA2. A similarity search was first performed with the SPECS
database and subsequently a molecular docking of the compounds containing the
N-phenylsulfonamide core into the ligand binding pocket of Lp-PLA2 was conducted.
The 500 highest-ranking compounds resulted from the virtual screening were visually
inspected and those with distinct binding modes compared to fragment 10 were discarded.
The survivors were then imported into the Schrödinger program Canvas for clustering
analysis using the Leader-Follower method. Ultimately, 100 compounds with diverse
structures were purchased and tested with the PAF enzymatic assay. Compound 11
representing low micromolar inhibitory activity (IC50 = 3431 nM, Table 1) was identified
as a promising hit.
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
To elucidate the significant inhibitory activity improvement and guide further
optimization, we determined the X-ray crystal structure of compound 11 binding with
rhLp-PLA2. Compound 11 is superimposed perfectly on fragment 10 and almost all the
aforementioned protein-ligand interactions are conserved (Figure 3A). Furthermore, the
N-phenylacetamide group and the diaryl ether moiety of 11 expand along the canyon-like
groove in two opposite directions. The N-phenylacetamide group projects into the
solvent-accessible region nearby the catalytic site and forms hydrophobic interactions
with Phe110. In this case, the side-chain of Gln352 is slightly changed compared to its
conformation in the Lp-PLA2-10 structure, disrupting the aforementioned
water-mediated network among Gln352, Phe357 and Thr358. The naphthyl group of
8
ACS Paragon Plus Environment

Page 9 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound 11 sits above the lipoprotein binding helices at the other side and it is
surrounded by hydrophobic residues Leu111, Phe357, Leu369, and Leu371. It is noticed
that the electron density of the terminal benzene ring of the naphthyl group is partially
missing (Figure S1B), indicating that it may not perfectly fit into this sub-pocket.
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) Superimposed crystal structures of compound 11 (green sticks) and 10
(yellow sticks) in complex with rhLp-PLA2. Residues interacting with compound 11 are
shown as gray sticks, and H-bonds are represented by black dashed lines. (B-D) A surface
representation of 10 (B, PDB ID: 5YE8), 11 (C, PDB ID: 5YE7) and 14a (D, PDB ID:
5YE9) from the opposite view (180° rotation) of (A). The ligand binding pockets are
shown with gray meshes. The sub-pocket A is marked in (C) with black dashed lines and
the purple arrow highlights the growing direction at the 3-position of the central benzene
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 57
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
ring of compound 11.
Structure optimization and SAR. Elaboration of hit 11 was initiated by exploring the
N-phenylsulfonamide binding core derived from fragment 10. After a comparative
investigation on complex structures of Lp-PLA2 bound with 10 and 11, we deduced that
a small sub-pocket (sub-pocket A) proximal to the middle benzene ring of 11 could be
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
utilized and different substitutes (R ) were thereby explored by increasing the size from
1
the smallest fluorine to a methoxyl group (Figure 3 and Table 1). The synthesized
compounds, at concentrations of 25 μM and 2.5 μM, were tested with the PAF enzymatic
assay, and compounds with stronger potency compared to 11 were subjected to IC50
measurement. It turns out that a fluorine substitution of R (compound 12) displays
1
two-fold potency improvement, verifying the rationale of our optimization strategy.
Compounds 13a and 15a decorated with a bulkier chlorine and methyl group,
respectively, alleviate the inhibition of the enzyme compared to 11. While a
trifluoromethyl (16a) or methoxyl (17a) substituent is unlike to be tolerated at the
sub-pocket A as they almost completely lost their inhibitory activity at 25 μM.
Remarkably, when a linear cyano group was attached to R
0-fold potency improvement compared to 11. It is thus suggested that the sub-pocket A
could only accommodate a narrow and short group like fluorine or cyano. In addition,
compatible with the complex structure, compounds with different R substitutions all lose
their potencies against rhLp-PLA2 (Table S1), since there is no space to accommodate
1
, compound 14a displays a
1
2
10
ACS Paragon Plus Environment

Page 11 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
these groups (Figure 3B-D). Accordingly, introduction of substituents to fill the narrow
space between the inhibitor and the binding pocket should be very careful.
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
Table 1. Effects of R on inhibitory activity of rhLp-PLA2.
1
%
inhibition
IC50
Compd
R
1
R
2
X
rhLp-PLA2
(nM)
2
5 μM 2.5 μM
1
1
H
F
H
H
H
H
H
H
H
H
H
H
O
O
85
93
82
100
75
13
0
38
59
8
3431
1418
1
2
1
3a
Cl
CN
O
14a
O
100
25
0
355
1
1
1
5a
6a
7a
CH
3
O
CF
OCH
CN
3
O
3
O
0
1
8
S
99
87
45
30
53
0
19
CN
NH
NCH
1915
2
0
CN
3
Additionally, a linker between the N-phenylsulfonamide core and the terminal
naphthalene ring was investigated based on compound 14a. Replacement of the oxygen
atom with a sulfur (18) or nitrogen (19) both resulted in a remarkably reduced potency.
When a methyl group was attached to the nitrogen linker, compound 20 also loses its
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 57
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
inhibition of rhLp-PLA2. Therefore, the oxygen may be the best option for this linker.
Table 2. Effects of R
3
and R on inhibitory activity of rhLp-PLA2.
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
Compd
14a
R
3
R
4
IC50 (nM)
355
2477
1070
322
613
222
45
2
2
2
2
2
1
2
3
4
5
H
F
H
H
Cl
CN
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
F
H
H
F
26
27
Cl
CN
152
100
17
2
2
3
3
8
9
0
1
CH
3
CF
3
OCH
3
924
902
1027
45
benzyl
phenyl
31a
32
CF
3
3
3
3
CN
CF
23
Based on the results above, compound 14a was selected for further optimization, and
the crystal structure of 14a in complex with rhLp-PLA2 was determined. Similar to the
fluorine at the 3-position of fragment 10, the cyano group of 14a fits well into the
12
ACS Paragon Plus Environment

Page 13 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
sub-pocket A, forming a H-bond with the main-chain nitrogen of Phe357 and
hydrophobic interactions with Ala355 and Phe357 (Figure 3D), both of which render it
more potent than compound 11. Like compound 11, the electron density of the terminal
benzene ring of the naphthyl group in the complex structure is not completed (Figure
S1C), leading us to believe that this ring is not comfortable in the binding pocket.
Unexpectedly, converting the naphthyl to a phenyl group results in compound 21 with an
IC50 of 2477 nM, ~8-fold weaker in the potency, which indicates that this terminal
benzene ring is indeed dedicated to the ligand binding to Lp-PLA2 despite of the
uncompleted electron density.
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
In order to refill this binding site, substituent exploration on terminal benzene ring of
21 was therefore conducted (Table 2). Deciphered from the complex structure of 14a
bound with rhLp-PLA2, the two ortho-positions of the ether bond in the naphthyl group
are in close proximity with Gln352, Phe357 and Leu107 (Figure S2), suggesting that
these positions are not suitable for substituent exploration. Learning the experience from
R
2
substitution, we only investigated the para- and meta-substituents (R
3
and R
4
) based
) in 21
on compound 21. Empirically, the para-position of the terminal benzene ring (R
3
tends to be oxidatively metabolized into a benzoquinone. We primarily introduced a
fluorine, chlorine or cyano group (22-24) to block the para-position. The chlorine
substituent (23) restores the activity of 14a, and the cyano (24) or fluorine (22)
substitution also enhance the potency compared to 21. Encouraged by these data, we
evaluated the substitution in R based on compound 23. As shown in Table 2, the
4
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 57
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
meta-position of the terminal benzene ring prefers lipophilic substituents with an
appropriate size. Larger groups like methoxy (30), phenyl (31a), or benzyl (31) ones
decrease the inhibitory activity by three folds. Smaller substituents such as the fluorine
(25), chlorine (26), cyano (27), methyl (28) and the trifluoromethyl (29) are propitious to
affinity improvement. Especially the trifluoromethyl substitute 29 displays 20-fold
potency improvement compared to 14a. In addition, a fluorine (32) or cyano (33)
substitution at the para-position renders the compound less potent compared to its
chlorine analogue (29). In summary, after two rounds of optimization, two-digital
nanomolar Lp-PLA2 inhibitors (26, 29, 32, and 33) were identified.
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
Lastly, substitutions on the solvent-exposed benzene ring were carried out (Table 3).
Inhibitor 29, displaying the most potent inhibitory activity to rhLp-PLA2, was selected as
the starting point for further optimization. Neither meta- nor ortho-position analogue
shows better activity compared to 29 (Table S2), thus we focused on the exploration of
the para-position substituents on the right benzene ring. Removing the 4-position
substituent leads to compound 34 with a lipophilic phenyl group facing to the solvent,
showing ~4-fold potency drop compared to 29. Moreover, replacement of the N-acetyl
amino group with a hydrophobic methyl (35) or nitro (36) group further weakens their
binding affinities to rhLp-PLA2. All of these stated above indicate that the lipophilic
moieties are not favorable in this solvent-exposed region. Unsubstituted (37),
monosubstituted (29), and disubstituted (38) amino groups on R
5
are generally acceptable
as they exhibit comparable activities. In addition, the linker between the N-acetyl amino
14
ACS Paragon Plus Environment

Page 15 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
group and the right benzene ring of 29 was also explored. A single methylene elongation
analogue 39 displays slight improvement in the inhibition of rhLp-PLA2, and further
elongation of this linker with an ethylidene or n-propylidene does not enhance the
potency compared to 39 (Table S2). A heterocyclic substituent such as morphine (42) or
piperazine (43) results in unfavorable binding to the enzyme. Replacing the N-acetyl
group in 29 with a morphine-4-carbamoyl group (40) hardly has effect on the potency.
Based on these results, we surmise that the catenulate linker for extending the hydrophilic
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
substituents of R
5
is preferable. To our surprise, the carboxyl analogue 41 outstands from
this series and performs the best inhibitory activity (IC50=5 nM) against rhLp-PLA2.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 57
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
Table 3. Effects of R on inhibitory activity of rhLp-PLA2.
5
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
rhLp-PLA2
Compd
29
R
5
IC50(nM)
17
65
232
236
14
12
9
3
3
3
3
4
5
6
7
H
CH
3
NO
2
2
NH
38
3
4
4
4
9
0
1
2
13
5
COOH
42
57
0.6
43
Darapladib
/
The crystal structure of compound 41 in complex with rhLp-PLA2 was determined to
elucidate the structure basis underlying its high potency. This time the electron density of
the bounded compound is clear and completed as well (Figure S1D). In addition, the
16
ACS Paragon Plus Environment

Page 17 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ortho-positions of the ether bond in the terminal benzene ring are close to Lp-PLA2
binding pocket and indeed have no space to be decorated. Analogously, the sulfonamide
oxygen forms H-bonding interactions with Phe274, Ser273 and Leu153 (Figure 4). The
additional H-bond is established between the CN group of 41 and the main-chain of
Phe357. Noteworthy, in the complex structure of Lp-PLA2-41, the side-chain of Gln352
rotated 90° from that in the complex structure with fragment 10, resulting in another
ordered conformation mediated by two H-bonds with Lys370 and a water molecule
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 4). In contrast, in the fragment 10 bound complex structure, the side-chain of
Gln352 is oriented by the sulfonamide oxygen (O1) and a water molecule (W2)
connecting the main-chains of Gln352, Phe357 and Thr358 (Figure 2B). Upon binding
with compounds 11 and 14a, this originally ordered conformation of Gln352 is disrupted,
while the binding of compound 41 further induces Gln352 to form a new ordered
conformation.
Hydrophobic
interactions,
mainly
formed
between
the
4-chloro-3-(trifluoromethyl) phenyl moiety and residues Phe357, Leu371, Leu111, and
Leu121, also contribute to the binding affinity. Accordingly, multiple H-bonds as well as
hydrophobic interactions of compound 41 with the surrounding residues account for its
high inhibition of rhLp-PLA2.
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 57
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 4. The detailed interactions of compound 41 with rhLp-PLA2 revealed by the
crystal structure (PDB ID: 5YEA). H-bonds between residues and 41 are presented by
black dashed lines.
Selectivity, plasma enzyme inhibition, rat hepatocytes stability and
permeability assessment. PLA2VIIB is an intracellular phospholipase and shares 62%
2
6
sequence similarity with Lp-PLA2. The inhibitory activity of PLA2VIIB was measured
to assess the selectivity of Lp-PLA2 inhibitors.22-23 Overall, our Lp-PLA2 inhibitors show
good selectivity (more than 1000 folds) over PLA2VIIB (Table 4).
Subsequently, the inhibitory activity of the compounds against human and rat
plasma Lp-PLA2 was tested in order to select compounds for inhibition assessment in
vivo. Compounds 37, 39, 40, and 43 exhibit more than 50% inhibition of Lp-PLA2 in
human plasma at a concentration of 50 nM (Table 4). For the rat plasma Lp-PLA2,
compounds 39, 37, 41, and 43 displays relatively good inhibition (>30% at 50 nM).
18
ACS Paragon Plus Environment

Page 19 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The Caco-2 cell permeability measurement implies that selected compounds possess
good permeability in vitro (Table 4). However, the high efflux ratio of 39 (11.7) was
27
found. In the rat hepatocytes stability assay , compound 39 with a single methylene
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
spacer displays a relatively poor stability, though it is the most potent inhibitor of the
plasma Lp-PLA2. Given the high stability of 29, the high CLint values of 38 (157.33
mL/min/kg) and 39 (186.36 mL/min/kg) might be due to the N-dealkylation of these two
compounds. Other inhibitors (29, 32, 33, 37, 40, and 41) show good stability in rat
hepatocytes.
Table 4. Selectivity, plasma enzyme inhibition, stability and permeability of selected
compounds.
Human plasma
% @ nM)
Rat plasma
(% @ nM)
Rat Hepatocytes
Stability
Caco-2
Permeability
PLA2VIIB
(
Compd
CLint
mL/min/kg)
NCa
A To B
(10 cm/s) Ratio
Efflux
IC50(μM)
250
50
250
50
-6
(
2
9
>25
>25
>25
>25
>25
>25
>25
>25
>25
>25
2
56
43
37
85
63
85
82
65
41
80
100
23
14
8
64
55
42
67
45
80
64
67
48
68
97
32
23
11
33
15
51
27
31
17
31
98
9.5
14.9
9.0
5.7
4.4
1.1
3.8
0.5
0.8
1.0
1.9
0.4
2.4
11.7
1.9
1.9
32
5.1
0.9
14.9
157.3
186.3
NCa
3
3
3
7
62
29
64
55
28
14
57
93
38
3
4
9
0
41
5.1
4
4
2
3
Darapladib
aNC= cannot be calculated: CLint value was too low to be calculated.
In vivo assessment of inhibitors. Compounds 37 and 41, with acceptable inhibitory
activities against the plasma enzyme and good stability as well as permeability in vitro,
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 57
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
were selected to assess their pharmacokinetics and potencies in the male SD rats. To
determine the bioavailability, 37 and 41 were dosed at 1 mg/kg and 3 mg/kg, respectively,
for intravenous and oral administration (Table 5). Though 41 displays a poor oral
bioavailability (F=8.8%), compound 37 obtains excellent clearance (4.9 mL/min/kg),
favorable AUC value (3.4 μg∙h/mL), and good oral bioavailability (F = 35.5%).
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
Table 5 Pharmacokinetic profiles of 37 and 41 in male SD Ratsa
Dose
CL
V
ss
T
max
C
max
T
1/2
AUC0-24h
F
Compd Route
(mg/kg) (ml/min/kg) (L/kg) (h)
(μg /ml) (h) (μg∙h/ml) (%)
IV
1
3
1
3
4.9
3.1
2.9
0.3
7.1
7.7
2.7
2.7
3.2
3.4
6.2
1.6
3
7
PO
IV
4
0.27
0.49
35.5
8.8
41
PO
0.25
an = 5 animals/ group, data are the mean values.
Compounds 37, 41 and darapladib were orally administrated at doses of 3 mg/kg and
5 mg/kg to evaluate the inhibitory potency in vivo. The inhibitory activity and plasma
2
concentration of 37 and 41 are dose-dependent (Figures S3-S4) and the results of 25
mg/kg dose are presented in Figure 5A. Although 41 displays better inhibitory activity
compared to darapladib in the first four hours, its potency dramatically decreases during
the next four hours. However, compound 37 maintains the inhibitory activity for 24 hours
after oral administration, which is superior to that of darapladib. The concentration-time
curves of 37 and 41 in vivo (Figure 5B) may account for their distinct activities, as the
20
ACS Paragon Plus Environment

Page 21 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
concentration of 41 reduces almost 10-fold from four hours to eight hours after being
orally dosed, while compound 37 owns high concentrations in plasma for twenty-four
hours. And the quick inhibition of Lp-PLA2 by 41 may benefit from its fast absorption
(Tmax=0.25 h). In addition, compound 37 displays better inhibition of the enzyme
compared to darapladib at 3 mg/kg dose concentration (Figure S4). Overall, 37 is
identified as a potential candidate for further development.
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. (A) The relative activity of Lp-PLA2 in the plasma of SD rats after an oral dose
of 37, 41 and darapladib at 25 mg/kg (n=5). For clearly representing the plasma
concentration of darapladib, two different scales are displayed in (B) (bottom: 200
intervals; top: 2000 intervals), concentration-time curves of 37, 41 and darapladib at 25
mg/kg (n=5), and all data are represented. Error bars represent the SD.
Chemistry. Synthetic route for the target compounds (11-33) is shown in Scheme 1. The
nucleophilic substitution was firstly employed to get intermediates 44a-44o and 46a-46n,
followed by reduction with iron powder to yield the corresponding aniline derivatives
45a-45o and 47a-47n, respectively. Then these intermediates reacted with
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 57
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
4-acetamidobenzenesulfonyl chloride in pyridine to give the desirable compounds.
Scheme 1. Synthesis of compounds 11-33a.
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
aReagents and Conditions : (a) K
CO
, DMF, 80°C, 2h; (b) EtOH/H O/AcOH= 10:10:1,
2
3
2
iron powder, 75°C, 2h; (c) pyridine, rt, overnight.
The target compounds 34-43 and 49 were prepared from 47i under similar conditions as
that for preparation of 11-33 (Scheme 2). Among these, compounds 34-41 and 49 were
obtained through an electrophilic aromatic substitution of chlorosulfonic acid (detailed
procedures shown in Supporting Information). In addition, 36 was reduced to give 37,
and the fluorine in 49 was substituted by two heterocyclic groups to generate 42 and 43,
respectively.
Scheme 2. Synthesis of compounds 34-43 and 49a.
22
ACS Paragon Plus Environment

Page 23 of 57
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
a
Reagents and Conditions : (b) EtOH/H O/AcOH= 10:10:1, iron powder, 75°C, 2h; (c)
2
pyridine, rt, overnight. (d) DMSO, 120°C, 10h.
CONCLUSIONS
Here we report the efficient discovery of novel and potent Lp-PLA2 inhibitors
containing a sulfonamide scaffold with the structure-guided strategy. A ~200,000-fold
affinity improvement was achieved from the initial sulfonamide fragment to the most
potent compound 41. Structure-based virtual screening (gaining ~300-fold potency
improvement) together with the subsequent structure-based ligand optimization
(contributing ~700-fold potency improvement) significantly accelerated the hit-to-lead
optimization. Most of the resulting inhibitors display good Caco-2 cell permeability and
excellent rat hepatocytes stability in vitro. Gratifyingly, compound 37 represents good
enzyme-inhibition activity and oral bioavailability in SD rats, and further optimization of
this compound is now under way.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 57
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
EXPERIMENTAL SECTION
PAF enzymatic assay. The inhibitory activity against the human recombinant Lp-PLA2,
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
0
PLA2-VIIB or plasma Lp-PLA2 was performed according to the protocol of Chen et al .
The substrate 2-thio-PAF (Cayman Chemical) and 5,5′-dithiobis(2-nitrobenzoic acid),
(DNTB, Sigma-Aldrich) were used to produce the absorbance at 412 nm after enzymatic
reaction. The purification of recombinant Lp-PLA2 and PLA2-VIIB was presented in the
Supporting Information. The recombinant and plasma Lp-PLA2 were added to the
enzymatic system directly. 5 µL compound, 10 µL 2 mM DNTB and 10 µL protein/
plasma were incubated for 30 min at room temperature. Finally, 175 µL substrate solution
(50 mM Tris pH 7.2, 1 mM EGTA, 50 µM 2-S-PAF) was added to the enzymatic system.
Enzymatic kinetics were measured every minute of 10-min reaction time and Vmax was
obtained to measure the inhibition. The IC50 curves were generated using GraphPad
28
Prism and each IC50 value was replicated at least three times.
This study has been approved and supervised by Institutional Animal Care and Use
Committee (IACUC), Shanghai Institute of Materia Medica, Chinese Academy of
Sciences (IACUC approval no.:2017-02-YY-05), and the Ethics Committee of Shanghai
Blood Center.
Crystal structure determination. The purified protein was concentrated to 4 mg/ml for
crystallization. Crystallization of rhLp-PLA2 was carried out by mixing a solution of the
protein with an equal volume of the precipitant solution (0.1 M MOPS pH 6.6, 0.4 M
24
ACS Paragon Plus Environment

Page 25 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Li
2
SO
4
, 27% (w/v) (NH4)
2
SO , 1 M Na-Ac, 1.4% (v/v) 1,4-butanediol). Crystals were
4
obtained by the vapour-diffusion method in hanging drops at 20C. The protein-inhibitor
complex crystals were prepared by soaking crystals of apo rhLp-PLA2 into the reservoir
solution containing 10 mM inhibitors or 50 mM fragments overnight. Subsequently,
crystals were directly flash frozen in liquid nitrogen.
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
Data were collected at 100 K at the Shanghai Synchrotron Radiation Facility
2
9
30
(SSRF) , and were processed with the HKL software packages . The structure was
31
solved by molecular replacement, using the program PHASER with the search model of
2
b
PDB ID: 5I8P . The structure was refined with the simulated-annealing protocol
32
33
implemented in the program PHENIX . With the aid of the program Coot , compounds
and water molecules were fitted into the initial Fo-Fc map. The complete statistics, as
well as the quality of four solved structures, are shown in Table S3.
Metabolic stability in rat hepatocytes. The procedures of rat hepatocytes isolation were
performed in accordance with Guide for the Care and Use of Laboratory Animals at
Shanghai Institute of Materia Medica, Chinese Academy of Sciences. The details for
3
4
preparation of rat hepatocytes followed the classic protocol . Compounds were dissolved
in DMSO to a final concentration of 1.0 μM (DMSO% was less than 0.1%, v/v) and
6
incubated with the rat hepatocytes (1.0×10 cells/mL) in Williams’ E medium at 37C.
Aliquots were removed at 0, 0.25, 0.5, 1, 1.5, and 2 h, and the reactions were quenched
with equal volume of ice-cold acetonitrile. Samples were injected to HPLC-MS/MS for
analysis and each experiment was performed in duplicate.
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 57
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
Half-live of the compound in hepatocytes was calculated using the following equation:
T1/2 =0.693/k
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
(-k) was the slope of the linear regression from log [substrate] versus time plot.
Intrinsic clearance was calculated using the following equation:
CLint = 0.693×Scaling factor/ T1/2 =2.990/ T1/2 (L/min∙kg)
Scaling factors employed in this calculation:
Scaling factor = (Rat liver weight/rat body weight) × (hepatocellularity/liver weight) /
(hepatocellular density in reaction system)
Rat liver weight/rat body weight = 40 g/kg; hepatocellularity/liver weight = 117×106
6
cells/g; hepatocellular density in reaction system= 1.0×10 cells/mL.
Caco-2 Permeability Assay. The Caco-2 monolayer assays were performed by using
20
standard procedure as previously reported for pyrimidone derivatives . Briefly, drug
transport from the apical side to the basolateral side (A–B) and from the basolateral side
to the apical side (B–A) was measured simultaneously under the same condition.
Propranolol and nadolol were used as the hypertonic and hypotonic control, respectively.
Digoxin was used as the positive control for Pgp-mediated drug efflux. After washing the
monolayer with HBSS (Hanks’ Balanced Salt solution, Sigma-Aldrich) three times, the
compounds were diluted and added to the appropriate well (pH 6.8 for the apical side and
pH 7.4 for the basolateral side). The plate was incubated at 37 °C for 95 min. Samples
were collected from the donor side at 5 and 95 min and from the receiver side at 35 and
26
ACS Paragon Plus Environment

Page 27 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
95 min post-incubation. The concentration of samples was measured by LC-MS/MS. The
P
app was calculated from the following equation:
[drug]acceptoꢀ
푉A
Papp
=
×
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
푆A × 푇 [drug]initial ꢁonoꢀ
is the volume of the acceptor well, S is the surface area of the membrane, T is
Where V
A
A
the total transport time, [drug]acceptor is the drug level at the acceptor side, and [drug]initial
donor is the drug level at the donor side at T = 0.
Animals. Male Sprague-Dawley (SD) rats (180- 220g for compound 37 and darapladib,
2
00- 250g for 41, supplied by Shanghai experimental animal center, CAS) with five
animals in each group were used for the in vivo studies. Animal experiments were
approved and supervised by Institutional Animal Care and Use Committee (IACUC),
Shanghai Institute of Materia Medica, Chinese Academy of Sciences (IACUC approval
no.:2017-02-YY-05).
In vivo study of selected compounds. Compounds dissolved in DMSO/0.25%
carboxymethylcellulose sodium (5/95, v/v) were subjected to oral administration at doses
of 3 mg/kg and 25 mg/kg. For the IV route, a group of five male SD rats were injected
with a single dose of 1 mg/kg with a dose volume of 5 mL/kg using PEG300/EtOH/saline
(40/10/50, v/v/v) as the vehicle. For both routes, serial specimens were collected predose
as well as 0.25, 0.5, 1, 2, 4, 8, and 24 h after administration. For pharmacokinetic analysis,
samples (except for the predose specimens) were precipitated with MeOH/acetonitrile
(1:1, v/v) and quantified by HPLC-MS. Pharmacokinetic parameters were calculated
from the mean serum concentration by non-compartmental analyses (MassLynx V4.1,
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 57
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
Waters).
For the in vivo activity analysis, blood samples, drawn predose as well as 0.25, 0.5,
1
, 2, 4, 8, and 24 h after administration, were applied to measure the Lp-PLA2 activity in
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
plasma. The Lp-PLA2 activity in the plasma of SD rats was measured by using the same
protocol as that used for measurement of Lp-PLA2 inhibitory activity in vitro.
Synthetic Chemistry. All reagents were purchased from commercial suppliers and used
without further drying or purification unless otherwise stated. Yields were not optimized.
1
13
H NMR and C NMR spectra were recorded on a Bruker AC400 or a Bruker AC500
NMR spectrometer using tetramethylsilane as an internal reference. Low-resolution mass
spectra were determined on an Agilent liquid-chromatography mass spectrometer system
that consisted of an Agilent 1260 infinity LC coupled to Agilent 6120 Quadrupole mass
spectrometer (ESI). High-resolution mass spectra were conducted on a triple TOF 5600+
MS/MS system (AB Sciex, Concord, Ontario, Canada) in the negative or positive ESI
mode. The purity of test compounds were determined by HPLC (Agilent ChemStation,
Agilent Eclipse XDB-C18, 5 μM, 4.6×150 mm, 30 °C, UV 240 nM, flow rate = 1.0
mL/min) with aqueous CH CN (30%−90%, 0.1% FA only for compound 43) for 12 min.
3
All the assayed compounds possess ≥95% purity. Column chromatography was
performed on silica gel (200−300 mesh) or with pre-packed silica cartridges (4-40g) from
Bonna-Agela Technologies Inc. (Tianjin, China) and eluted with a CombiFlash@ Rf 200
from Teledyne Isco, and preparative TLC was performed on HSGF 254 (0.4−0.5 mm
thickness; Yantai Jiangyou Company, Yantai, Shangdong, China).
2
8
ACS Paragon Plus Environment

Page 29 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
General Procedure 1. Step a: Dissolving the p-fluoronitrobenzene or its derivative (0.1
mM) in DMF, then K
2
CO (0.13 mM) and corresponding nucleophilic reagent (0.1 mM)
3
were added, and the mixture was heated at 80°C for 2h. The reaction was quenched with
water and the product was extracted with EtOAc (×3). The organic layer was washed
with saturated NaCl solution (×3) and dried with anhydrous magnesium sulfate. The
extracting solvent was evaporated in vacuo, and the residue was directly for the next step
without purification.
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
Step b: The residue obtained (0.1 mM) from above step was dissolved in
EtOH:H O:AcOH (v:v:v = 10:10:1). After being added with reduced iron powder (0.5
2
mM), the solution was heated to 75°C for 2h. Then it was diluted with EtOAc and the
water layer was extracted with EtOAc for another 2 times. The organic layer was
combined, washed with brine, dried with anhydrous MgSO , filtered, and evaporated in
4
vacuo. The residue was purified by column chromatography (PE/EA= 5:1).
General Procedure 2: Selected aniline derivative (0.1 mM) was previously dissolved in
pyridine, and the corresponding sulfonyl chloride (0.11 mM) was added slowly. If the
sulfonyl chloride was oil, beforehand dissolving it in THF was necessary. The mixture
was stirred overnight at room temperature, concentrated to small volume in vacuo, and
neutralized with 2M HCl. Then it was extracted by EtOAc for three times. The combined
organic layers were washed by a saturated aqueous NaCl solution (×3), dried over
MgSO
4
, filtered, and concentrated under reduced pressure. Purification by column
chromatography on silica gel (DCM/MeOH=50:1) afforded the desirable compound.
29
ACS Paragon Plus Environment