Synthesis of 2-(2-oxo-2H-chromen-4-yl)acetamides as potent acetylcholinesterase inhibitors and molecular insights into binding interactions

Article abstract of DOI:10.1002/ardp.201800310

Sixteen novel coumarin-based compounds are reported as potent acetylcholinesterase (AChE) inhibitors. The most active compound in this series, 5a (IC₅₀ 0.04 ± 0.01 μM), noncompetitively inhibited AChE with a higher potency than tacrine and gala

Full text of DOI:10.1002/ardp.201800310

Received: 23 October 2018
Revised: 30 March 2019
Accepted: 3 April 2019
|
|
DOI: 10.1002/ardp.201800310
F U L L P A P E R
Synthesis of 2‐(2‐oxo‐2H‐chromen‐4‐yl)acetamides as potent
acetylcholinesterase inhibitors and molecular insights into
binding interactions
Jiraporn Kara1
Teerapat Nualnoi3
Vannajan Sanghiran Lee5
|
Paptawan Suwanhom1
|
Chatchai Wattanapiromsakul2
Pasarat Khongkow4
Luelak Lomlim1
|
|
Jindaporn Puripattanavong2
|
|
|
Anand Gaurav6
|
1Department of Pharmaceutical Chemistry,
Faculty of Pharmaceutical Sciences, Prince of
Songkla University, Hat Yai, Songkhla,
Thailand
Abstract
Sixteen novel coumarin‐based compounds are reported as potent acetylcholinesterase
(AChE) inhibitors. The most active compound in this series, 5a (IC50 0.04 0.01 µM),
noncompetitively inhibited AChE with a higher potency than tacrine and galantamine.
Compounds 5d, 5j, and 5 m showed a moderate antilipid peroxidation activity. The
compounds showed cytotoxicity in the same range as the standard drugs in HEK‐293
cells. Molecular docking demonstrated that 5a acted as a dual binding site inhibitor. The
coumarin moiety occupied the peripheral anionic site and showed π‐π interaction with
Trp278. The tertiary amino group displayed significant cation‐π interaction with Phe329.
The aromatic group showed π‐π interaction with Trp83 at the catalytic anionic site. The
long chain of methylene lay along the gorge interacting with Phe330 via hydrophobic
interaction. Molecular docking was applied to postulate the selectivity toward AChE of
5a in comparison with donepezil and tacrine. Structural insights into the selectivity of the
coumarin derivatives toward huAChE were explored by molecular docking and 3D QSAR
and molecular dynamics simulation for 20 ns. ADMET analysis suggested that the 2‐(2‐
oxo‐2H‐chromen‐4‐yl)acetamides showed a good pharmacokinetic profile and no
hepatotoxicity. These coumarin derivatives showed high potential for further develop-
ment as anti‐Alzheimer agents.
2Department of Pharmacognosy and
Pharmaceutical Botany, Faculty of
Pharmaceutical Sciences, Prince of Songkla
University, Songkhla, Thailand
3Department of Pharmaceutical Technology,
Faculty of Pharmaceutical Sciences, Prince of
Songkla University, Hat Yai, Songkhla,
Thailand
4Faculty of Medicine, Institute of Biomedical
Engineering, Prince of Songkla University,
Songkhla, Thailand
5Department of Chemistry, Faculty of Science,
University of Malaya, Kuala Lumpur, Malaysia
6Department of Pharmaceutical Chemistry,
Faculty of Pharmaceutical Sciences, UCSI
University, Kuala Lumpur, Malaysia
Correspondence
Luelak Lomlim, Department of Pharmaceutical
Chemistry, Faculty of Pharmaceutical
Sciences, Prince of Songkla University, 15
Karnjanavanich Rd., Hat Yai, Songkhla 90112,
Thailand.
Email: luelak.l@psu.ac.th
K E Y W O R D S
3D QSAR, acetylcholinesterase inhibitor, coumarin, lipid peroxidation, molecular docking
Funding information
Prince of Songkla University, Grant/Award
Number: PHA570404S; University of Malaya,
Grant/Award Numbers: RP020C‐14AFR,
GPF063B‐2018
|
1
INTRODUCTION
complicated and still unclear. β‐Amyloid aggregation and neurofi-
brillary tangles are major abnormalities found in the brain of patients
Alzheimerʼs disease (AD) is the most common form of dementia. The
symptoms start with difficulty in remembering new information and
recent events, followed by apathy, depression, subsequent impaired
judgment, confusion, and behavior change. Etiology of AD is
with AD, which eventually lead to neuronal damage. This results in
the decrease of acetylcholine (ACh), the neurotransmitter respon-
sible for memory and learning. Acetylcholinesterase inhibitors
(AChEIs), such as tacrine, donepezil, galantamine, and rivastigmine,
|
Arch Pharm Chem Life Sci. 2019;e1800310
wileyonlinelibrary.com/journal/ardp © 2019 Deutsche Pharmazeutische Gesellschaft
1 of 11
https://doi.org/10.1002/ardp.201800310

2 of 11
KARA ET AL.
|
are the major drugs prescribed for patients with AD. These agents
reversibly inhibit acetylcholinesterase that degrades ACh, and in turn
increase the ACh level in the brain.[1]
active compound to determine the mode of AChE inhibition. Insights in
binding interaction between the coumarin derivative and AChE and
BChE were predicted by molecular docking and atom‐based 3D‐QSAR.
Drug‐likeness properties of compounds in this series were anticipated via
ADMET analysis. The molecular dynamics simulations were performed
on the selected compound to confirm the stability in the binding pockets.
AChEIs have been used in clinics for several decades. However, the
3D structure of the enzyme was unknown until the 1990s when the
crystal structure of Torpedo californica AChE was first reported.[2] It has
been found that the active site of AChE appears as a deep and narrow
gorge with aromatic amino acid residues lining the entrance of the gorge
forming the peripheral anionic site (PAS). The catalytic anionic site (CAS)
is located underneath the PAS. This is the area where Trp84 binds to the
ammonium group of ACh and, subsequently, the acetyl group of ACh
orientated to the the acetylation site (Ser200) at the bottom of the gorge.
Hydrolysis of ACh by the Ser200 residue then occurs, resulting in
acetylated Ser200 residue (inactive AChE) and choline as products.
Butyrylcholinesterase (BChE) is another enzyme that hydrolyzes
ACh. Both cholinesterase enzymes differ in substrate specificity,
enzyme kinetics, expression and activity in different brain regions. In
|
2
RESULTS
|
2.1
Acetyl‐ and butyrylcholinesterase inhibitory
activities
Acetyl‐ and butyrylcholinesterase inhibitory activities of the synthe-
sized compounds were evaluated by comparing to the drugs
donepezil, tacrine, and galantamine (Table 1). Compound 5a (IC50
=
0.04 μM, huAChE) was the most potent AChEI in this series. The
compound exhibited inhibitory activity against AChE in the micro-
molar scale (0.04‐27 μM) and showed a higher potency than tacrine
(IC50 = 0.10 μM) and galantamine (IC50 = 0.75 μM). The selectivity
index (SI) of the coumarins ranged from 0.06 to 17.
a
normal brain, the activity of BChE is relatively low but
progressively increases as AD advances while AChE activity declines.
Therefore, inhibition of BChE may provide an additional benefit in
AD treatment.[3] BChE contains 524 amino acids. The active site of
the enzyme is similar to AChE.[4] AChEI molecules vary in their
selectivity for AChE vs BChE. Tacrine and galantamine mainly bind
with the CAS and inhibit AChE and BChE with close IC50 values.[5]
Donepezil binds with both CAS and PAS, acting as a dual binding site
inhibitor and showing higher selectivity toward AChE.[6,7]
|
2.2
Kinetic study
Compound 5a was chosen as a representative for this assay because
it shows the highest AChE inhibitory activity. Model fittings for
evaluating the mode of inhibition were carried out with a global
nonlinear regression mode of the GraphPad Prism software and are
presented in a double reciprocal Lineweaver‐Burk plot (Figure 2).
The coumarin ring has been of high interest as a scaffold for
design of several potent AChEIs.[8–24] In previous studies, coumarin
moiety showed affinity to the PAS of AChE due to its aromatic
character that facilitates the π‐π interaction with the amino acid
residues in the region. Amino group‐containing moiety of the
compounds occupied the CAS of the active site gorge showing
cation‐π interaction. Although most reported coumarin derivatives
were substituted with amino group‐containing moiety on position 3
or 7 of the coumarin ring system, potent AChEIs with phenyl
piperazinyl moiety substituted on position 4 of the coumarin ring
were also reported.[9,14] Recently, isoindoline‐1,3‐dione and related
derivatives of 1‐benzylpiperazine, which were developed by our
group and others, showed relatively high acetylcholinesterase
inhibitory activity and the benzyl‐piperazinyl moiety was suggested
to be responsible for interaction with CAS.[25–27] In this study,
coumarin derivatives with N‐substituted benzylpiperazinyl moiety
replaced on position 4 of the coumarin ring were designed as shown
in Figure 1.
|
2.3
Antilipid peroxidation activity
As shown in Table 1, compounds 5d, 5j, and 5 m showed a moderate
antioxidant activity while other compounds exhibited only weak
antilipid peroxidation activity.
Research proposes that oxidative stress plays a role in initiating
β‐amyloid aggregation and tau protein hyperphosphorylation.[28–31]
Thus, prevention of oxidative stress‐induced neuronal cell damage
has become another approach to AD therapy. The antioxidative
activity of numerous coumarin derivatives has been reported.[16,19]
Therefore, the lipid peroxidation inhibitory effects of the designed
compounds have also been investigated.
Furthermore, the synthesized compounds were evaluated for their
cytotoxicity. An enzyme kinetic study was performed with the most
FIGURE 1 Design of coumarin derivatives

KARA ET AL.
3 of 11
|
TABLE 1 Inhibitory activity against AChE and BChE, antilipid peroxidation and in vitro cytotoxic activities of the synthesized derivatives
against HEK‐293 cell lines
IC50 (µM SEM)
Linoleic acid peroxidation inhibition
assayd % inhibition SD
Cytotoxicity
LC50 (µM)
Cpd.
5a
5b
5c
5d
5e
5f
R1
R2
R3
n
2
2
2
2
3
3
3
3
2
2
2
2
3
3
3
3
huAChEa
huBChEb
SIc
OH
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
H
H
0.04 0.01
1.33 0.21
9.05 0.49
26.09 0.44
7.02 0.06
2.92 0.20
12.94 0.44
9.45 0.27
0.43 0.03
16.56 0.40
1.20 0.03
5.23 0.13
3.47 0.12
10.46 0.70
12.61 0.32
27.29 1.16
0.004 0.0001
0.10 0.003
0.75 0.03
…
0.68 0.07
1.30 0.12
1.94 0.25
1.52 0.19
14.62 0.85
17.63 0.62
11.61 0.58
19.61 1.06
2.72 0.14
17.24 0.27
5.05 0.33
68.44 2.66
19.03 0.91
23.49 1.74
21.22 0.58
11.55 0.73
1.90 0.02
0.02 0.003
2.70 0.09
…
17
25.18 0.06
5.76 0.05
42.52 0.04
65.60 0.01
29.57 0.01
32.45 0.01
n.d.
56.388
62.921
85.621
67.545
57.108
>100
OH
F
0.98
0.21
0.06
2.08
6.04
0.90
2.08
6.33
1.04
4.21
13.09
5.48
2.25
1.68
0.42
475
0.2
OH
CH3
NO2
H
OH
OH
OH
F
5g
5h
5i
OH
CH3
NO2
H
n.d.
OH
n.d.
n.d.
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
OCH3
23.74 0.04
55.81 0.03
34.60 0.02
39.64 0.04
50.10 0.07
n.d.
76.701
86.454
>100
5j
H
F
5k
5l
H
CH3
NO2
H
H
>100
5m
5n
5o
5p
H
70.131
n.d.
H
F
H
CH3
NO2
42.45 0.01
43.24 0.04
n.d.
>100
H
82.146
69.751
72.247
83.182
n.d.
Donepezil
Tacrine
n.d.
Galantamine
BHT
3.6
n.d.
…
95.11 0.04
Note. BHT: butylated hydroxytoluene; nd: not determined; SD: standard deviation; SEM: standard error of mean.
ahuman acetylcholinesterase.
bhuman butyrylcholinestrase.
cSI = selectivity index = IC50BChE/IC50AChE.
dData are reported as mean standard deviation (n = 3).
|
2.4
Cytotoxicity assessment
kcal/mol (Table 2) were retrieved. Compound 5a was analyzed in
comparison to donepezil. Donepezil (binding energy −11.9 kcal/mol)
was bound with huAChE with a slightly higher affinity than 5a
(binding energy −11.6 kcal/mol). This result corresponded with the
findings in the enzyme inhibition assay. The structures of the
donepezil, tacrine, and 5a complexes were minimized and the residue
interaction energy with huAChE and huBChE in 4 Å region from the
ligands are reported in Table 3. The docking model of donepezil, 5a
and tacrine with huAChE and huBChE are illustrated in Figure 3.
Compound 5a was stably bound in the binding pocket throughout
20 ns MDs. The complexed structure at 20 ns, potential energy and
the root mean square displacement data are reported in the
Supporting Information.
The results from the SRB assay are shown in Table 1. Generally, all of
the tested compounds displayed low cytotoxicity (LC50 > 50 µM).
Compound 5a, the most potent acetylcholinesterase inhibitor in this
series, showed LC50 at 56.388 µM. This value is more than 1000‐fold
higher than the concentration for human AChE inhibition.
|
2.5
Molecular docking and molecular dynamics
simulations
Several conformers of structure with favorable binding affinity as
indicated by the low negative values in the range from −11.9 to −7.8

4 of 11
KARA ET AL.
|
substituents on the inhibitory activity was not observed, the larger
size of the substituents seemed to reduce activity. When the
ethylene linker was changed to propylene linker (n = 3) (5e–h), the
corresponding derivatives showed lower potency, with the exception
of 5h which exerted higher potency than 5d.
For 7‐methoxyl coumarin derivatives with ethylene linker (n = 2)
(5i–l), substitution on R3 also decreased huAChE inhibitory activity.
An increase in size of the substituent resulted in a more potent
inhibitor. An extension of the linker to propylene group resulted in
increased activity of the corresponding derivatives (5m–p) with the
exception of 5n. Substitution on 3‐position of the benzyl moiety of
7‐hydroxyl‐8‐methyl coumarin derivatives with ethylene linker (n = 2)
(5a–d) resulted in a slight decrease in BChE inhibitory activity.
Change from ethylene linker to propylene linker led to reduced
activity (5e–h). The same trend was observed in the 7‐methoxyl
coumarin series. Additionally, all the compounds evaluated in this
study showed no selectivity toward AChE as indicated by the low
selectivity index (SI).
FIGURE 2 Kinetic assay for AChE inhibition by 5a. The kinetic
assay of AChE inhibition was carried out at 0, 10, and 25 µM of
compound 5a. The concentration of ATCI substrate was varied
between 10 to 125 µM. The resulting over‐layered double reciprocal
Lineweaver‐Burk plot generated by the GraphPad Prism software
demonstrated that compound 5a inhibited the enzyme by
noncompetitive mode
|
3.2
Kinetic study
|
2.6
3D QSAR
Kinetic analysis of AChE inhibition was performed to determine the
inhibition mode of the compounds in this series. In the Lineweaver‐
Burk plot (Figure 2), Vmax decreased but Km was not affected as the
increase of concentration of the inhibitor. The result suggested that
The atom‐based 3D QSAR model was built using the 3D QSAR model
module of Discovery Studio 3.1. Default settings were used for model
development. Validation of the developed model was performed
based on the internal predictions of the data set and leave‐one‐out
cross‐validation method (five‐fold). PLS analyses of the data sets
compound 5a is
a noncompetitive inhibitor with an inhibition
using three principal components showed a high cross‐validated r2
constant (Ki) of 15.29 µM.
cv
value of 0.788, RMS residual error (cross‐validation) of 0.311,
noncross‐validated r2 value of 0.898, RMS residual error of 0.201,
r2adj of 0.873. pIC50 values of the synthesized compounds for
huAChE and huBChE and the AChE selectivity index (SI) are listed in
Supporting Information. In a plot of actual vs predicted SI, predicted
value was obtained using the developed 3D QSAR model. the van der
Waals and electrostatic contour plots were obtained using the atom
based 3D QSAR methods.
|
3.3
Antilipid peroxidation activity
Since the brain tissues are rich in phospholipids, free‐radical induced
lipid peroxidation can be another factor that causes brain damage.
Several synthetic and natural coumarin derivatives have potential
antioxidant activity.[16,19,32–34] The compounds in this series showed
only low to moderate anti‐lipid peroxidation activity in comparison to
butylated hydroxytoluene (BHT).
|
2.7
ADMET analysis
|
3.4
Cytotoxicity assessment
All compounds conformed to Lipinskiʼs rule of five (MW < 500;
AlogP < 5; donorHB ≤ 5; acceptorHB ≤ 10) (see Supporting Information).
Selected compounds in this series were evaluated for their
cytotoxicity against human embryonic kidney (HEK‐293) cells
compared with the reference drugs. HEK‐293 has characteristics of
immature neurons and is used in neuroscience research. 5a showed
LC50 at 56.388 µM (Table 1). This value is more than 1000‐fold
higher than the concentration for human AChE inhibition.
|
3
DISCUSSION
|
3.1
Acetyl‐ and butyrylcholinesterase inhibitory
activities
|
3.5
Molecular docking and molecular dynamics
Cholinesterase inhibitory activity of the synthesized compounds is
shown in Table 1. Among 7‐hydroxyl‐8‐methyl coumarin derivatives
with ethylene linker (n = 2) (5a–d), the most potent huAChE inhibitor
in this series was 5a (0.04 0.01 µM). Substitution on 3‐position of
the benzyl moiety with F, CH3, and NO2 groups led to decreased
activity. While clear electron donating/withdrawing effects of the
simulations
The binding mode between the commercial drugs (donepezil and
tacrine) and 5a were compared. Two docking centers were set at the
center of the PAS and CAS binding sites. The huAChE and the
huBChE crystal structures have very high similarity of amino acid

KARA ET AL.
5 of 11
|
TABLE 2 Binding affinity in kcal/mol from AutoDock Vina. The structures in mode A are in the same direction as the binding mode of the
drugs in the X‐ray structure (‐benzyl in donepezil toward the inner gorge and ‐NH2 in tacrine lied in the same side as in X‐ray) whereas those in
mode B are also the low energy structure but aligning in the opposite direction. The docked structures of ligands are in a ball and stick model;
nonpolar hydrogens are not shown and the line structures of donepezil and tacrine from the X‐ray structure are superposed in blue and green.
(
M)
)
affinity (k
A)
B)
A)
B)
A)
B)
A)
B)
A)
B)
A)
B)
A)
B)
A)
B)
A)
B)
A)
B)
A)
B)
A)
B)
sequence homology with similar molecular forms and active center
structure; however, the binding cavity volume for the huAChE is
larger to accommodate donepezil (about 347.63 Å3), whereas the
huBChE is about 246.75 Å3 in binding with tacrine.
has shifted from 475 (donepezil) to 17 (5a). 5a had slightly better
activity for huAChE than tacrine; however, tacrine had the best
affinity to huBChE among all the compounds. The AutoDock Vina
results correlated well with the huAChE activity results where the
best binding of donepezil > 5a > tacrine was found, but the
correlation was not in agreement with huBChE which theoretically
indicated better binding affinity of 5a than tacrine where the
experimental result observed better activity of tacrine > 5a >
donepezil. 5a fit well in the AChE pocket and due to its large
molecular size, could not fit into the BChE cavity and ended up with a
curl structure (Table 2). The residue interaction at the binding
From IC50 values, donepezil is specific and preferably bound to a
larger cavity of huAChE than huBChE whereas tacrine, which has a
smaller size, can bind to both structures. Compound 5a is similar in
structure to donepezil; however, a slight reduction of the binding
affinity from −11.9 to −11.6 kcal/mol in the PAS pocket was
observed, reflecting the lower selectivity of 5a against AChE
compared to donepezil. The selectivity index (IC50 BChE/ IC50 AChE)

6 of 11
KARA ET AL.
|
TABLE 3 Contribution of van der Waals (VDW) and electrostatic (Elect) interaction energy (IE, kcal/mol) of huAChE and huBChE residues in
4 Å region from donepezil, tacrine, and 5a. The residues in the inner gorge are in bold
Donepezil
IE
Tacrine
IE
5a
IE
Residue
VDW
Elect
Residue
VDW
Elect
Residue
VDW
Elect
huAChE residues
A_TYR72
1.26
−4.34
1.33
−2.33
−2.55
−0.5
3.59
A_ASP74
−2.02
−5.41
−2.79
−2.02
−2.36
−1.28
−0.09
2.25
−1.1
−0.92
A_TYR72
−1.53
−1.78
−2.23
−0.48
−4.95
−1.43
−1.39
−3.39
−0.57
−1.26
−0.65
−7.91
−1.69
−0.41
−2.2
0.25
A_ASP74
−1.79
1.83
A_TRP86
−4.6
−0.82
−2.44
0.23
A_ASP74
−6.85
1.2
−4.62
1.68
A_THR83
A_TRP86
A_GLY120
A_GLY121
A_TYR124
A_TYR133
A_GLU202
A_SER203
A_TRP286
A_LEU289
A_GLU292
A_SER293
A_VAL294
A_PHE295
A_ARG296
A_PHE297
A_TYR337
A_PHE338
A_TYR341
A_GLY342
A_HIS447
A_GLY448
A_ILE451
IE in 4 Å
A_TYR119
A_GLY120
A_GLY121
A_TYR124
A_SER125
A_TYR133
A_GLU202
A_SER203
A_TYR337
A_PHE338
A_TYR341
A_HIS447
A_GLY448
A_ILE451
−0.35
−2.25
−1.82
−1.1
A_THR83
A_TRP86
−3.06
−1.15
−3.35
−4.24
−0.75
−3.66
−3.74
−5.43
0.79
−4.57
−1.21
−1.19
−2.71
−0.5
1.51
−0.36
−0.96
−8.59
5.06
4.6
0.06
−0.54
−0.18
0.43
A_GLY120
A_GLY121
A_TYR124
A_TYR133
A_GLU202
A_SER203
A_TRP286
A_LEU289
A_GLU292
A_SER293
A_VAL294
A_PHE295
A_ARG296
A_PHE297
A_TYR337
A_PHE338
A_TYR341
A_HIS447
A_GLY448
A_ILE451
0.47
−2.16
−1.53
−0.25
−2.5
−7.19
8.46
−0.52
−1.01
−1.33
−1.24
−3.08
−2.23
−1.73
−2.8
3.26
0.99
1.56
−1.16
−0.75
−5.72
−0.39
−0.53
−1.49
−2.18
−1.26
−0.58
−1.41
−4.57
−4.42
−7.96
−0.37
−2.63
−1.03
−0.35
−52.39
−17.96
−57.49
−2.1
−0.77
−3.83
−0.88
1.08
−4.55
−1.94
−8.47
−4.6
−3.29
−1.29
−0.56
−2.91
3.66
−2.99
0.29
−5.07
−3.96
−1.15
−1.95
−7.82
0.07
1.19
8.43
8.97
−0.22
−5.02
1.26
3.25
−6.26
−9.29
−9.73
−0.28
−1.09
−3.56
−7.10
−7.11
0.62
−4.77
−7.11
−8.47
0.3
1.03
3.22
−1.19
−0.6
−4.7
−1.64
−1.42
−1.87
−2.85
−4.18
−3.32
−6.33
−2.21
−0.85
−0.65
−3.06
3.47
3.32
3.92
2.05
−2.37
−2.29
−11.89
−4.62
−8.02
−0.86
−3.68
−2.17
−0.5
0.32
0.56
1.01
−7.7
−2.68
0.86
−1.3
−1.69
1.35
0.99
−6.63
−1.46
−1.38
−71.17
−28.74
−72.76
−3.99
−0.43
−1.03
−18.78
−10.77
−15.27
−2.84
−1.53
−32.38
−32.38
−44.81
−26.94
−26.94
−30.95
−5.44
−5.44
−64.85
−34.01
−85.41
−55.65
−18.14
−61.76
−9.2
IE inner
−15.86
−23.66
IE
−13.87
huBChE residues
A_ASP70
−7.46
−5.54
0.96
−2.81
−0.79
−0.40
−5.15
−1.63
−1.55
−1.61
−0.75
−0.61
−1.65
−1.50
−0.48
−2.25
−4.75
−4.65
−4.75
1.37
A_SER79
−1.27
−6.94
−2.04
−1.32
0.01
−0.78
−7.28
−1.54
−1.27
−0.37
−0.71
−1.41
−1.30
−1.21
−1.34
−1.63
−0.91
−3.02
−1.49
−0.49
0.34
A_ASP70
−6.19
−1.38
1.19
−2.97
−0.56
−2.18
−6.07
−0.70
v2.30
−4.39
−2.99
−2.22
−0.70
−1.60
−2.99
−3.46
−0.99
−3.22
−0.82
3.37
A_SER72
A_TRP82
A_GLY78
A_SER79
A_SER79
A_GLY115
A_GLY116
A_THR120
A_TYR128
A_GLU197
A_SER198
A_ALA328
A_TYR332
A_TRP430
A_MET437
A_HIS438
A_GLY439
−0.50
−0.04
0.37
A_TRP82
−3.95
−0.76
−1.80
−2.40
1.63
1.20
A_TRP82
A_ASN83
A_GLY115
A_GLY116
A_GLY117
A_THR120
A_TYR128
A_GLU197
A_SER198
A_TRP231
A_PRO285
−7.27
−0.84
−4.56
−1.58
−6.52
3.46
−1.20
−0.14
−2.26
2.82
A_GLY115
A_GLY116
A_GLN119
A_THR120
A_TYR128
A_GLU197
A_SER198
A_GLY283
A_THR284
A_PRO285
0.87
−0.26
−0.79
2.38
1.23
1.94
−2.01
−2.79
−1.34
−1.98
−3.56
−0.26
−5.04
−1.32
−0.60
−1.49
−0.13
−0.63
−1.94
0.66
−3.53
5.68
−2.21
8.98
−1.60
10.63
−8.07
−3.00
−1.65
1.31
−0.42
0.85
0.29
−9.58
−3.48
−3.91
−3.44
2.45
−0.69
−0.51
−0.97
2.30
−2.02
0.17
2.94
0.02
(Continues)

KARA ET AL.
7 of 11
|
TABLE 3 (Continued)
Donepezil
IE
Tacrine
IE
5a
IE
Residue
VDW
−1.58
−3.83
−0.57
−0.70
−1.00
−2.60
−4.64
−3.23
−1.06
−0.32
Elect
−0.74
Residue
VDW
−1.40
−0.33
Elect
−3.94
Residue
VDW
−4.34
−1.52
−1.77
−0.94
−5.10
−3.50
−0.37
−1.66
−0.53
−4.53
−1.70
−0.86
−1.02
−61.97
−25.30
−57.02
Elect
A_LEU286
A_SER287
A_VAL288
A_ASN289
A_ALA328
A_PHE329
A_TYR332
A_HIS438
A_GLY439
A_ILE442
−2.32
−2.41
−0.60
−1.71
−5.79
−2.51
−8.39
−5.31
−1.80
0.79
A_TYR440
A_ILE442
−5.34
−0.32
A_LEU286
A_SER287
A_VAL288
A_ALA328
A_PHE329
A_TYR332
A_ASN397
A_PHE398
A_TRP430
A_HIS438
A_GLY439
A_TYR440
A_ILE442
−6.90
−2.56
6.24
1.41
−0.03
−1.01
−4.78
0.09
0.01
4.72
−10.44
−5.39
−6.31
−3.21
−1.04
−1.03
0.85
−8.67
−4.45
−1.21
0.29
−3.75
−2.08
−0.74
1.11
−0.67
0.63
1.38
−5.23
−4.78
−8.01
−3.94
−76.14
−27.62
−59.36
−0.70
−3.08
−7.14
−2.92
−14.17
−2.30
−2.34
IE in 4 Å
IE inner
IE
−63.02
−15.64
−52.00
−45.49
−16.69
−39.77
−17.53
1.05
−34.29
−34.29
−33.94
−26.00
−26.00
−23.81
−8.29
−8.29
−12.23
−10.12
Note. CAS for huAChE: Ser203, Glu334, and His447; PAS for huAChE: Tyr72, Asp74, Tyr124, Trp286; anionic site for huAChE: Trp86. CAS for huBChE:
Ser198, Glu325 and His438; PAS for huBChE: Tyr341 and Asp70 and Tyr332; anionic site for huBChE: Trp82.
interface explained the activity of the compounds in agreement with
the experimental investigation where the activity of donepezil (−71
kcal/mol) > 5a (−65 kcal/mol) > tacrine (−32 kcal/mol). High affinity of
tacrine toward huAChE and huBChE is from the interaction with
inner gorge residues in the range of −32 to −34 kcal/mol, where
donepezil and 5a contributed to both inner and outer gorge.
Gly115 (−0.76), Gly116 (−1.80), Tyr128 (−2.21), Glu197 (8.98),
Ser198 (−9.58), His438 (−5.31), Gly439 (−1.80), Ile442 (0.79) (IE
inner = −15.64 kcal/mol). Hydrophobic interaction with Tyr332 and
Phe329 at midgorge of active site and π‐π interaction between the
benzyl group and Trp82 (distance of 3.82 Å) were exhibited.
Compound 5a showed interaction via Val288 (−10.44), Tyr440
(−8.01), Trp82 (−7.27), Leu286 (−6.90), Gly117 (−6.52), Phe329
(−6.31), and Asp70 (−6.19) while interaction of the inner gorge of
enzyme (IE inner) are Trp82 (−7.27), Gly115 (−4.56), Gly116 (−1.58),
Tyr128 (−0.42), Glu197 (0.85), Ser198 (−0.69), His438 (−5.23),
Gly439 (−4.78), Ile442 (−3.94) (IE inner = −27.62 kcal/mol). 5a cannot
fit into huBChE cavity observed from the curl structure, but it can
still interact with Tyr332 via the hydrophobic interaction and π‐π
interaction between the benzyl group and Trp82 (distance of 4.07 Å).
Residues that showed interaction with tacrine were Trp82 (−6.94),
Tyr440 (−5.34), His438 (−5.04). This compound bound only with CAS
of the active site of enzyme as same as interaction with huAChE (IE
inner = −34.29 kcal/mol) and showed stacking interaction with Trp82
(distance of 3.59 Å) and Trp430 of the active site of huBChE.
Donepezil, 5a and tacrine interactions with huBChE are shown in
Figure 3.
The strong favorable interactions toward huAChE (negative
value, kcal/mol) are summarized in Table 3. For compound 5a,
huAChE residues that showed interaction are Tyr337 (−11.89),
Gly121 (−8.59), Trp286 (−8.47), Tyr341 (−8.02) and Asp74 (−6.85)
while interaction energy of the inner gorge of enzyme (IE inner) are
Trp86 (−0.36), Gly120 (−0.96), Gly121 (−8.59), Tyr133 (0.99), Glu202
(−4.05), Ser203 (−1.94), Tyr337 (−11.89), His447 (−0.86), Gly448
(−3.68) and Ile451 (−2.17) (IE inner = −34.01 kcal/mol). 5a was able to
bind both CAS and PAS of huAChE. At the PAS, the coumarin ring
interacted with the indole ring of Trp286 via π‐π interaction
(distance of 3.67 Å). At the mid‐gorge, the long chain of methylene
interacted with Phe338 and Tyr341 via the hydrophobic interaction.
At the CAS, the benzyl group of compound can interact with π‐π
interaction by Trp86 (distance of 3.80 Å). These findings correspond
with our rationale in designing the coumarin derivatives.
For huBChE, the higher activity of tacrine > 5a > donepezil can be
explained in Table 3 where the interaction in the inner gorge was
mainly interacting with the interaction energy (IE inner) of −34.29,
−27.62, −15.64 kcal/mol. The favorable interaction toward huBChE
(‐negative value) for donepezil are Ser198 (−9.58), Tyr332 (−8.39),
Asp70 (−7.46), Ala328 (−5.79), Ser72 (−5.54) and His438 (−5.31)
while the interaction of inner gorge of enzyme are Trp82 (−3.95),
The docked structures of 5a in binding with AChE and BChE
were selected for MDs simulations. During the 20 ns MDs, 5a
remained stable inside the binding pocket of both enzymes. The
MD final structure of 5a inside AChE and BChE and potential
energy and the RMSD of AChE and BChE with 5a which indicated
the stability of the complex are reported in the Supporting
Information.

8 of 11
KARA ET AL.
|
FIGURE 3 The crystal structure and the binding cavity (in gray) of huAChE in complex with donepezil (a) and of huBChE in complex with
tacrine (b). Docking model of donepezil (c), 5a (d) and tacrine (e)‐huAChE complex. Docking model of donepezil (f), 5a (g), and tacrine (h)‐huBChE
complex

KARA ET AL.
9 of 11
|
FIGURE 4 Compound 5a mapped to (a) the van der Waals contour map; (b) the electrostatic contour map
|
|
ADMET analysis
3.6
3D QSAR
3.7
The atom‐based 3D QSAR models were developed from the set of
the 16 inhibitors using substructure based molecular overlay
alignment. From the predicted huAChE selectivity (SI) at 3rd PLS
factor for the training and test set, it is quite evident that almost all
compounds yielded a good predicted value (see Supporting Informa-
tion). All the parameters of the QSAR models confirmed their
reliability and predictability which can be used in the design of new
and high selectivity huAChE inhibitors.
In general, the coumarin derivatives showed good absorption and
medium blood‐brain barrier penetration. All compounds were
predicted to be water soluble. Hepatotoxicity, which is the key
problem that limits clinical use of tacrine, was not defined in all
synthesized compounds.
|
4
EXPERIMENTAL
Chemistry
From the contour maps (Figure 4), the van der Waals plot (Figure
4a) shows green and yellow contours indicating regions favoring
bulky and lighter groups respectively. The mostly yellow colored
contours surrounding the most selective molecule 5a suggests that
steric bulk in general is not favorable for huAChE selectivity. Higher
selectivity of compounds 5a, 5i and 5l supports this finding.
|
4.1
The synthesis pathway of the coumarin derivatives (5a–p) is illustrated
in Scheme 1. Derivatives of 2‐(2‐oxo‐2H‐chromen‐4‐yl)acetic acid (2a
and 2b) were prepared via Pechmann condensation between 3‐acetone
dicarboxylic acid and the corresponding phenol. Isoindoline‐1,3‐dione
derivatives of 1‐benzylpiperazine or 1‐(3‐substituted)‐benzylpiperazine
(3a–h) were synthesized as described in our previous work.[19] Amines
4a–h were obtained from the treatment of the isoindoline‐1,3‐diones
(3a–h) with hydrazine hydrate in ethanol at 80oC.[35] Coupling of the
The electrostatic plot (Figure 4B) shows red and blue contours
indicating regions favoring electronegative and electropositive
groups respectively. Red colored contours can be seen surrounding
the central part of the most selective molecule 5a indicating that
electronegative substituents in the central part of the molecule will
be favorable for huAChE selectivity. Both ends of the compound 5a
are surrounded by blue contours, suggesting that these parts can be
substituted by electropositive groups for better huAChE selectivity.
2‐(2‐oxo‐2H‐chromen‐4‐yl)acetic acid (2a and 2b) with amines 4a–h was
[30,36]
performed using coupling reagents
to yield the amides 5a–p.
Structures of synthesized compounds were determined by IR, 1H‐NMR,
13C‐NMR and high resolution mass spectrometry.
SCHEME 1 Synthesis of the coumarin derivatives. (i) 1,3‐Acetone‐dicarboxylic acid, H2SO4, MeOH; (ii) hydrazine hydrate, EtOH, 80°C; (iii)
a. DCC, DMAP, DIEA, CH2Cl2, room temperature; b. EDCI, DMAP, DIEA,DMF, room temperature, c. HATU, DIEA, NMP, room temperature

10 of 11
KARA ET AL.
|
The InChI codes of the investigated compounds together with
some biological activity data are provided as Supporting Information.
bath for 10 min. The generated TBA reactive substances were
measured by their absorbance at 532 nm.
|
4.2
Biological assays
|
4.2.4
In vitro cytotoxicity assay
|
4.2.1
Cholinesterase inhibition assay
The cytotoxicity assay of the coumarins against HEK‐293 cells was
measured using a sulforhodamine B (SRB) assay.[39] The cells were
cultured in Dulbeccoʼs modified Eagleʼs medium supplemented with
10% fetal bovine serum (Gibco, Paisley, UK) and grown at 37°C in a
humid atmosphere containing 5% CO2. Cells (5 × 103cells/well) were
seeded in a 96‐well plate and incubated at 37°C, 5% CO2 for 24 hr.
The cells were then treated with or without various concentrations
(0, 3.125, 6.25, 12.5, 25, 50, and 100 µM) of each compound for
48 hours. After an incubation period, 40% (w/v) trichloroacetic acid
(TCA) was added to the cells and the cells were then incubated at 4°C
for 1 hr. The medium was removed and the cells were rinsed with
slow running tap water. A total of 0.4% (w/v) SRB solution (100 μl)
was added to each well and the cells were incubated for 1 hr at room
temperature. The SRB solution was removed and then the cells were
washed three times with 1% (v/v) acetic acid and they were allowed
to dry at room temperature. The protein‐bound dye was dissolved
with 10 mM Tris base solution and the absorbance was measured at
492 nm using a microplate reader.
The 2‐(2‐oxo‐2H‐chromen‐4‐yl)acetamides were evaluated with the
[37]
Ellmanʼs method
using AChE from humans (huAChE) and human
BChE (huBChE). Donepezil, tacrine, and galantamine were used as
positive controls. Experimental protocol was as follows: 125 µl of
3 mM 5,5‐disthiobis(2‐nitrobenzoic acid; DTNB); 25 µl of 1.5 mM
acetylthiocholine iodide (ACTI); 1.5 mM butylthiocholine iodide
(BuTI) were used for the butylcholinesterase inhibitory assay. A
total of 50 µl of 50 mM Tris‐HCl buffer pH 8.0; and 25 µl of sample
dissolved in ethanol were added to the 96‐well microplate followed
by 25 µl of AChE enzyme. The absorbance was then measured at
405 nm every 11 seconds for 2 minutes using a microplate reader. A
total of 50% inhibition of the AChE and BChE activity (IC50) studies
using nine concentrations of compounds 5a–p (from 0.000398 to
1000 µM) were performed, and the obtained data were analyzed
using the Software Prism. Each sample was analyzed in triplicate
(n = 3).
|
4.2.2
Kinetic study
|
4.3
Molecular docking and molecular dynamics
simulations
Compound 5a was selected for the kinetic study because of its
highest potency in this series. The velocities of the reaction in the
presence and absence of the inhibitor were measured by the
spectrophotometric method as described above. Three concentra-
tions of the inhibitor, 0, 10, and 25 μl, were chosen for the assay. For
each concentration of the inhibitor, the concentration of the
substrate ACTI was varied from 10 to 125 μM. The mode of
inhibition of the compound 5a was evaluated by a global nonlinear
regression fit using the kinetic inhibition analysis mode of
the GraphPad Prism software and the result was displayed in a
Lineweaver‐Burk plot.
The initial structures of recombinant human acetylcholinesterase
(huAChE) (PDB ID: 4EY7) and human butyrylcholinesterase (huBChE)
(PDB ID:4BDS) were retrieved from the protein data bank and
minimized using AMBER software. The protonation state of ionizable
protein side chains was assigned at pH 7. The structures of inhibitors
were optimized using the density functional theory. Molecular
docking was performed using AutoDock Vina where the binding site
was covered the PAS and CAS of the enzyme. The box dimension of
30 × 30 × 30 Å was allowed for the ligand to explore for the best
binding interaction. The catalytic binding sites (CAS) include three
catalytic triads (huAChE: Ser203, Glu334 and His447 and huBChE:
Ser198, Glu325, and His438) and the peripheral site, PAS (huAChE:
Tyr72, Asp74, Tyr124, Trp286 and huBChE: Tyr341 and Asp70 and
Tyr332). The best docked pose in agreement with the X‐ray structure
with the lowest binding affinity was selected and its respective
residue binding interaction energy was calculated. Selected com-
plexes were further investigated with MDs using AMBER14 force
field for 20 ns. More method details are included in the Supporting
Information.
|
4.2.3
Linoleic acid peroxidation inhibition assay
Anti‐linoleic acid peroxidation activity was measured by the method
of Sekiwa et al.[38] Briefly, the tested compound (final concentration
100 µg/ml) was added to a 2.5% linoleic acid solution, 50 mM
phosphate buffer (pH 7.0), and purified water in a screw‐cap vial. A
solution without the sample was used as a control. BHT was used as
the reference compound. Each sample was analyzed in duplicate.
Each vial was kept in darkness at 45°C for 20 days and measured
every 5 days. The hydroperoxide and aldehyde produced from
linoleic acid were determined by the thiocyanate and thiobarbituric
acid (TBA) methods, respectively. Sample solutions after the first day
and last day of incubation were each subjected to the TBA method.
The reaction mixture was added with 20% trichloroacetic acid and
0.67% TBA/99.5% ethanol and the mixture heated in a boiling‐water
|
4.3.1
3D QSAR modeling
A data set comprising 16 compounds with their huAChE and huBChE
IC50 (nM) data was used for the development of the atom‐based 3D
QSAR model for huAChE selectivity. The huAChE IC50 and huBChE
IC50 values were first transformed to the respective pIC50

KARA ET AL.
11 of 11
|
[12] M. Catto, L. Pisani, F. Leonetti et al., Bioorg. Med. Chem. 2013, 21, 146.
[13] J. I. da Silva, M. C. de Moraes, L. C. C. Vieira, A. G. Corrêa, Q. B. Cass,
C. L. Cardoso, J. Pharm. Biomed. Anal. 2013, 73, 44.
[14] S. F. Razavi, M. Khoobi, H. Nadri et al., Eur. J. Med. Chem. 2013, 64,
252.
(–log IC50). The huAChE selectivity (SI = pIC50huAChE−pIC50huBChE
)
was then calculated and used as the dependent variable for the 3D
QSAR modeling. The 3D QSAR studies were performed using
Discovery Studio 3.1. The ligands were pre‐aligned using the
substructure based molecular overlay method and placed in a 3D
grid space. The grid spacing was 1.5 Å. The energy potentials on
every grid point were then calculated using a CHARMm force field
which used the electrostatic potential and the van der Waals
potential and treated them as separate terms. A + 1e point charge
is used as the electrostatic potential probe and distance‐dependent
dielectric constant is used to mimic the solvation effect. For the van
der Waals potential a carbon atom with a 1.73 Å radius was used as a
probe. The energy grid potentials can be used as independent
variables to create partial least‐squares. Validation was performed
using leave‐one‐out cross validation method (five folds). The 3D
QSAR was evaluated by using r2 and cross validated r2.
[15] S. O. Nam, D. H. Park, Y. H. Lee, J. H. Ryu, Y. S. Lee, Bioorg. Med.
Chem. 2014, 22, 1262.
[16] J. B. Shaik, B. K. Palaka, M. Penumala et al., Eur. J. Med. Chem. 2015,
107, 219.
[17] S. Ghanei‐Nasab, M. Khoobi, F. Hadizadeh et al., Eur. J. Med. Chem.
2016, 121, 40.
[18] L. G. de Souza, M. N. Renn, J. D. Figueroa‐Villar, Chem. Biol. Interact.
2016, 254, 11.
[19] J. B. Shaik, B. K. Palaka, M. P. et al., Eur. J. Med. Chem. 2016, 107, 219.
[20] F. Sonmez, B. Z. Zengin kurt, I. Gazioglu et al., J. Enzyme Inhib. Med.
Chem. 2017, 32, 285.
[21] M. Alipour, M. Khoobi, H. Nadri et al., Arch. Pharm. Chem. Life Sci.
2013, 346, 577.
[22] M. Alipour, M. Khoobi, A. Moradi et al., Eur. J. Med. Chem. 2014, 82,
536.
[23] D. Yao, J. Wang, G. Wang et al., Bioorg. Chem. 2016, 68, 112.
[24] L. Pisani, R. Farina, M. Catto et al., J. Med. Chem. 2016, 59, 6791.
[25] J. Kara, C. Wattanapiromsakul, L. Lomlim, SDU Res. J. 2013, 6, 141.
[26] W. Si, T. Zhang, L. Zhang et al., Bioorg. Med. Chem. Lett. 2016, 26, 2380.
[27] D. Panek, A. W. Ckowska, T. Wichur et al., Eur. J. Med. Chem. 2017,
125, 676.
|
4.3.2
ADMET analysis
Drug‐likeness of the designed compounds was predicted using the
Discovery Studio 2.5 program.
[28] M. A. Ansari, S. W. Scheff, J. Neuropathol. Exp. Neurol. 2010, 69, 155.
[29] W. ‐J. Huang, X. Zhang, W. ‐W. Chen, Biomed. Rep. 2016, 4, 519.
[30] S. Sharma, S. Verma, M. Kapoor, A. Saini, B. Nehru, Neurol. Res. 2016,
38, 838.
ACKNOWLEDGMENTS
The authors would like to thank Prince of Songkla University (grant
no. PHA570404S) and University of Malaya (Faculty Research Grant:
GPF063B‐2018) for financial support. The authors would like to
thank Miss Maria Mullet for providing linguistic proofreading for the
manuscript.
[31] C. Cheignon, M. Tomas, D. Bonnefont‐Rousselot, P. Faller, C. Hureau,
F. Collin, Redox Biol. 2018, 14, 450.
[32] I. Kostova, S. Bhatia, P. Grigorov et al., Curr. Med. Chem. 2011, 18, 3929.
[33] G. B. Bubols, R. Vianna Dda, A. Medina‐Remon et al., Mini Rev. Med.
Chem. 2013, 13, 318.
[34] Y. Al‐Majedy, A. Al‐Amiery, A. A. Kadhum, A. Bakar Mohamad, Sys.
Rev. Pharm. 2017, 8, 24.
[35] S. N. Haydar, C. Ghiron, L. Bettinetti et al., Bioorg. Med. Chem. 2009,
17, 5247.
CONFLICT OF INTERESTS
[36] L. Lomlim, J. Einsiedel, F. W. Heinemann, K. Meyer, P. Gmeiner,
J. Org. Chem. 2008, 73, 3608.
The authors declare that there are no conflict of interests.
[37] G. L. Ellman, K. D. Courtney, J. R. V. Andres, V. Featherstone, M.
Robert, Biochem. Pharmacol. 1961, 7, 88.
ORCID
[38] Y. Sekiwa, K. Kubota, A. Kobayashi, J. Agric. Food. Chem. 2000, 48,
373.
Luelak Lomlim
http://orcid.org/0000-0003-1229-2189
[39] P. Skehan, R. Storeng, D. Scudiero et al., .J. Natl. Cancer Inst. 1990, 82,
1107.
REFERENCES
SUPPORTING INFORMATION
[1] Alzheimer's Association, J. Alzheimer's Assoc. 2017, 13(4), 325.
[2] J. L. Sussman, M. Harel, F. Frolow et al., Science 1991, 253, 872.
[3] R. M. Lane, S. G. Potkin, A. Enz, Int. J. Neuropsychopharmacol. 2006,
9, 101.
Additional supporting information may be found online in the
Supporting Information section at the end of the article.
[4] A. N. Cokugras, J. Biochem. 2003, 28, 54.
[5] U. Holzgrabe, P. Kapková, V. Alptüzün, J. Scheiber, E. Kugelmann,
Expert Opin. Ther. Targets 2007, 11, 161.
How to cite this article: Kara J, Suwanhom P,
Wattanapiromsakul C, et al. Synthesis of 2‐(2‐oxo‐2H‐
chromen‐4‐yl)acetamides as potent acetylcholinesterase
inhibitors and molecular insights into binding interactions.
Arch. Pharm. Chem. Life Sci. 2019;e1800310.
[6] G. Kryger, I. Silman, J. L. Sussman, Structure 2003, 7, 297.
[7] D. Muñoz‐Torrero, P. Camps, Curr. Med. Chem. 2006, 13, 399.
[8] L. Piazzi, A. Rampa, A. Bisi et al., J. Med. Chem. 2003, 46, 2279.
[9] X. Zhou, X. B. Wang, T. Wang, L. Y. Kong, Bioorg. Med. Chem. 2008,
16, 8011.
[10] B. Tasso, M. Catto, O. Nicolotti et al., Eur. J. Med. Chem. 2011, 46, 2170.
[11] M. Alipour, M. Khoobi, A. Foroumadi et al., Bioorg. Med. Chem. 2012,
20, 7214.
https://doi.org/10.1002/ardp.201800310