Synthesis, pharmacology and molecular docking on multifunctional tacrine-ferulic acid hybrids as cholinesterase inhibitors against Alzheimer’s disease

Article abstract of DOI:10.1080/14756366.2018.1430691

The cholinergic hypothesis has long been a “polar star” in drug discovery for Alzheimer’s disease (AD), resulting in many small molecules and biological drug candidates. Most of the drugs marketed for AD are cholinergic. Herein, we report our efforts in the discovery of cholinesterases inhibitors (ChEIs) as multi-target-directed ligands. A series of tacrine-ferulic acid hybrids have been designed and synthesised. All these compounds showed potent acetyl-(AChE) and butyryl cholinesterase(BuChE) inhibition. Among them, the optimal compound 10g, was the most potent inhibitor against AChE (electrophorus electricus (eeAChE) half maximal inhibitory concentration (IC₅₀) = 37.02 nM), it was also a strong inhibitor against BuChE (equine serum (eqBuChE) IC₅₀ = 101.40 nM). Besides, it inhibited amyloid β-protein self-aggregation by 65.49% at 25 μM. In subsequent in vivo scopolamine-induced AD models, compound 10g obviously ameliorated the cognition impairment and showed preliminary safety in hepatotoxicity evaluation. These data suggest compound 10g as a promising multifunctional agent in the drug discovery process against AD.

Full text of DOI:10.1080/14756366.2018.1430691

Journal of Enzyme Inhibition and Medicinal Chemistry
ISSN: 1475-6366 (Print) 1475-6374 (Online) Journal homepage: http://www.tandfonline.com/loi/ienz20
Synthesis, pharmacology and molecular docking
on multifunctional tacrine-ferulic acid hybrids
as cholinesterase inhibitors against Alzheimer’s
disease
Jie Zhu, Hongyu Yang, Yao Chen, Hongzhi Lin, Qi Li, Jun Mo, Yaoyao Bian,
Yuqiong Pei & Haopeng Sun
To cite this article: Jie Zhu, Hongyu Yang, Yao Chen, Hongzhi Lin, Qi Li, Jun Mo, Yaoyao
Bian, Yuqiong Pei & Haopeng Sun (2018) Synthesis, pharmacology and molecular docking
on multifunctional tacrine-ferulic acid hybrids as cholinesterase inhibitors against Alzheimer’s
disease, Journal of Enzyme Inhibition and Medicinal Chemistry, 33:1, 496-506, DOI:
10.1080/14756366.2018.1430691
To link to this article: https://doi.org/10.1080/14756366.2018.1430691
©
2018 The Author(s). Published by Informa
View supplementary material
Submit your article to this journal
View Crossmark data
UK Limited, trading as Taylor & Francis
Group.
Published online: 06 Feb 2018.
View related articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ienz20

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY, 2018
VOL. 33, NO. 1, 496–506
https://doi.org/10.1080/14756366.2018.1430691
RESEARCH PAPER
Synthesis, pharmacology and molecular docking on multifunctional tacrine-ferulic
acid hybrids as cholinesterase inhibitors against Alzheimer’s disease
a
a
b
a
a
a
c
b
a
Jie Zhu , Hongyu Yang , Yao Chen , Hongzhi Lin , Qi Li , Jun Mo , Yaoyao Bian , Yuqiong Pei and Haopeng Sun
a
b
Department of Medicinal Chemistry, China Pharmaceutical University, Nanjing, China; School of Pharmacy, Nanjing University of Chinese
c
Medicine, Nanjing, China; School of Nursing, Nanjing University of Chinese Medicine, Nanjing, China
ABSTRACT
ARTICLE HISTORY
The cholinergic hypothesis has long been a “polar star” in drug discovery for Alzheimer’s disease (AD),
resulting in many small molecules and biological drug candidates. Most of the drugs marketed for AD are
cholinergic. Herein, we report our efforts in the discovery of cholinesterases inhibitors (ChEIs) as multi-tar-
get-directed ligands. A series of tacrine-ferulic acid hybrids have been designed and synthesised. All these
compounds showed potent acetyl-(AChE) and butyryl cholinesterase(BuChE) inhibition. Among them, the
optimal compound 10g, was the most potent inhibitor against AChE (electrophorus electricus (eeAChE)
half maximal inhibitory concentration (IC50) ¼ 37.02 nM), it was also a strong inhibitor against BuChE
Received 27 November 2017
Revised 17 January 2018
Accepted 17 January 2018
KEYWORDS
Alzheimer’s disease;
cholinesterase inhibitor;
multi-target ligand; tacrine-
ferulic hybrid; molecular
docking
(
equine serum (eqBuChE) IC50 ¼ 101.40 nM). Besides, it inhibited amyloid b-protein self-aggregation by
6
5.49% at 25 lM. In subsequent in vivo scopolamine-induced AD models, compound 10g obviously ame-
liorated the cognition impairment and showed preliminary safety in hepatotoxicity evaluation. These data
suggest compound 10g as a promising multifunctional agent in the drug discovery process against AD.
eeAChE Inhibition
IC50 = 37.03 nM
Binding mode analysis
eqBuChE Inhibition
IC50 = 101.40 nM
In vivo
Hepatotoxicity study
Ab1-42 aggregation Inhibition
6
5.49% Inhibitory rate
Morris water maze
Behavioral studies
4
–7
Introduction
value in this day and age . Multi-target-directed ligands (MTDLs),
aiming to simultaneously regulate multiple pathological processes
in the development of neurodegenerative cascade, were devel-
Alzheimer’s disease (AD) is one of the major threats to the whole
world. Every 3 s one new case is diagnosed and the number will
almost double every 20 years . AD is estimated to cost the world
8
8
–11
oped as a promising therapeutic approach
.
1
Cholinergic system, which plays an important role in the regu-
1
2
18 billion dollars in 2015 alone and by 2030 the cost will rise to lation of learning, cognition, and memory processes , has been
2
13,14
two trillion . These numbers are struggling. Currently, the number extensively studied for the design of anti-Alzheimer’s drugs
.
3
of drugs against AD is five in all . Commercial drugs for AD treat- Currently, three of the drugs approved for AD treatment target on
ment are unable to prevent or halt Alzheimer’s disease, so it is cholinesterase, including donepezil, rivastigmine, and galantamine.
necessary to develop novel compounds with potential therapeutic The most promising approach for the symptomatic treatment of
CONTACT Haopeng Sun
10009, China
Supplemental data for this article can be accessed here.
sunhaopeng@163.com
Department of Medicinal Chemistry, China Pharmaceutical University, Tong Jia Xiang Road 24#, Nanjing
2
ß 2018 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distri-
bution, and reproduction in any medium, provided the original work is properly cited.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
497
AD is to increase the synaptic levels of acetylcholinesterase (AChE) compound 1 (10 g) was added and stirred at room temperature
1
5,16
in the brain
. At the neuronal level, AChE can be hydrolysed by overnight. The mixture was acidified with concentrated hydro-
two types of cholinesterases (ChEs): acetylcholinesterases (AChEs) chloric acid (HCl) until pH ¼ 4–5. The precipitate was collected by
1
7
and butyrylcholinesterases (BuChEs) . During the early stage of filtration, washed with cold water, and dried over an infrared
AD, AChE plays a dominant role in AChE hydrolysis while BuChE lamp, resulting in compound 2 as a white solid and used in the
1
8,19
plays only a supportive role
. So, inhibiting the activity of AChE next step without further purification. The total yield of compound
is an effective way to prevent AD progression. In advanced AD, 2 obtained was 8.9 g (97.4%).
the activity of AChE decreases to 10–15% of normal values in cer-
General procedure for the synthesis of 9-chloro-1,2,3,4-tetrahy-
tain brain regions, while the amount of BuChE remains the same droacridine (3). To a mixture of 2-aminobenzoic acid (compound 2;
2
0
or even increases up to two-fold . Hence, inhibiting the activity 5 g, 36.2 mmol) and cyclohexanone (3.8 ml, 36.2 mmol) in a three-
of AChE encounters ineffectiveness during later stages of AD. necked round bottom flask equipped with mechanical stirrer, add-
Evidence showed that BuChEs can rescue the cholinesterase func- itionally a funnel and thermometer was placed and 15 ml of
tion in the absence of AChE. Therefore, a balance between AChE Phosphoryl chloride (POCl
) was added placing the flask on ice
bath. The mixture was allowed to reflux for 3 h and then was
Previously, our group has disclosed a series of tacrine-cinnamic poured onto ice. The resulting mixture was filtered through a
3
2
1–23
and BuChE inhibition will be more beneficial
.
2
4
hybrids . The structure-activity relationship (SAR) analysis on cin- Celite pad and the filtrate was extracted with Dichloromethane
namic acid revealed that the introduction of 3-OMe-4-OBn to the (CH Cl
; 3 ꢀ 15 ml) and the organic layers were washed with brine,
benzene ring of cinnamic acid (4-Bn ferulic acid, compound 10a) dried over anhydrous sodium sulphate (Na SO ). After evaporation
2
2
2
4
was the best. So, structural modification on the cinnamic acid moi- in vacuo, the resulting residue was purified on silica gel chromato-
ety, which is replaced by ferulic acid with different substitutions graph (PE/EA ¼ 8:1) to furnish a pale brown solid compound 3.
deserved further research. In the present study, we designed a ser- The total yield of compound 3 obtained was 5.99 g (76.0%).
1
ies of tacrine-ferulic acid hybrids with structural modification on
General procedure for the synthesis of N –(1,2,3,4-tetrahydroacri-
benzyl moiety against both AChEs and BuChEs for the treatment din-9-yl)ethane-1,2-diamine (4). Ethylenediamine (3 ml, 45.94 mmol)
of AD, in which MTDLs approach was applied. The target com- and sodium iodide (catalytic amount) were added to 10 ml of 1-
ꢁ
pounds were synthesised and evaluated for their ChEs inhibitory pentanol and heated to 160 C. Then, a solution of Compound 3
activities with optimal compounds tested for their inhibitory (2 g, 9.19 mmol) in 30 ml 1-pentanol was added dropwise via an
ꢁ
effects on self-induced Ab1–42 aggregation. The SAR of synthe- additional funnel to the above mixture at 160 C. After being
ꢁ
sized compounds was discussed. Furthermore, two optimal com- stirred at 160 C for 18 h, the resulting mixture was quenched by
pounds 10d and 10g were tested in AD mice model for the in the addition of water, later the solution was acidified to pH ¼ 4
vivo behavioral and hepatotoxic evaluations.
with concentrated HCl. The mixture was stirred at room tempera-
ture for 30 min. The aqueous phase was separated and basified
with solid NaOH until pH ¼ 13–14 and extracted with CH
2 2
Cl
Methods
(3 ꢀ 15 ml). The CH Cl layer was then washed with brine and
2
2
dried over anhydrous Na
2
SO
4
. After concentration, the crude prod-
2
Cl /
Chemistry
uct was purified by silica gel column chromatograph (CH
2
General experimental
methanol(MeOH)/triethylamine(Et N) ¼ 60:1:0.3) to give compound
3
All chemicals, reagents, and solvents were purchased from com- 4 as a brown oil. The total yield of compound 4 obtained was
mercial companies. When necessary, solvents were used with fur- 0.6750 g (30.4%).
ther purification and dryness. Reactions were monitored by
General procedure for the synthesis of benzyl (E)-3–(4-(benzyloxy)-
analytical thin layer chromatography (TLC) on silica gel 60 F254 3-methoxyphenyl)acrylate (7a–7m). Compound 5 (0.5 g, 2.57 mmol)
precoated plates (purchased from Qingdao Haiyang Inc., China). and potassium carbonate (K CO ; 1.42 g, 10.30 mmol) were added
2
3
Spots were visualized by ultraviolet light at 254 and 365 nm. to 15 ml of DMF and stirred at room temperature for 15 min.
Column chromatography was performed on silica gel (200–300 Compound 6 was added dropwise to the above mixture solution.
ꢁ
mesh) for the purification of intermediates and final compounds. After being stirred at 82 C for 4 h, the reaction mixture was
Melting points were determined using a Mel-TEMP II melting point quenched with water. The precipitate was filtrated and the filter
1
13
apparatus. H NMR and C NMR spectra were recorded on a cake was washed with water to give the crude product which
1
13
Bruker Avance (300 MHz for H; 500 MHz for C, Billerica, MA) at could be used in the next step without further purification.
00 K dissolved in deuterated dimethyl sulfoxide(DMSO-d ) or deu- General procedure for the synthesis of (E)-3–(4-(benzyloxy)-3-
3
6
terated chloroform (CDCl ) with tetramethylsilane (TMS) as an methoxyphenyl)acrylic acid (8a–8m). To a mixture solution of
3
internal standard. NMR data were analysed by MestReNova soft- 2 N x NaOH (30 ml) and MeOH (30 ml) compound 7a–7 m was
ware (Mestrelab Research, S.L., Spain). Chemical shifts were added. The reaction mixture was heated to reflux for 3 h. Then,
reported in ppm (d), Coupling constants (J) were given in hertz, MeOH in the solution was removed and the pH was adjusted to
and peak multiplicities were reported as singlet (s), doublet (d), around 2 by adding concentrated HCl. The precipitate was fil-
triplet (t), and multiplet (m). High-resolution mass spectrometry trated, washed with cold water, and dried over an infrared lamp
(
HRMS) was performed on a Mariner Mass Spectrum (ESI) or an to get compound 8a–8m.
LC/MSD TOF HR-MS Spectrum. Final compounds were named fol-
General procedure for the synthesis of (E)-3–(4-(benzyloxy)-3-
lowing IUPAC rules as applied by ChemDraw Professional (version methoxyphenyl)acryloyl chloride (9a–9m). Thionyl chloride (SOCl
5.0, Darmstadt, Germany).
3 ml, 27.57 mmol) was added to a solution of compound 8a–8m
2
;
1
(
0.93 mmol) in 5 ml of anhydrous CH Cl . After being refluxed for
2 2
3
2
h, the reaction mixture was evaporated to remove excess SOCl .
General procedures for the synthesis and spectral data of the The residue was diluted with anhydrous CH Cl for next step.
2
2
synthesised compounds
General procedure for the synthesis of (E)-3–(4-(benzyloxy)-3-
General procedure for the synthesis of 2-aminobenzoic acid (2). To a methoxyphenyl)-N-(2-((1,2,3,4-tetrahydroacridin-9-yl)amino)ethyl)acry-
solution of Sodium hydroxide of 2N normality (2 N ꢀ NaOH; 40 ml), lamide (10a–10m). To a mixture solution of compound 4 (0.2 g,

4
98
J. ZHU ET AL.
1
3
0
.84 mmol) and K
2
CO
3
(0.26 g, 1.86 mmol) in anhydrous CH
2
Cl
2
3.05 (s, 2H), 2.75 (s, 2H), 2.40 (s, 3H), and 1.90 (s, 4H). C NMR
): d 167.58, 158.35, 150.74, 150.04, 149.79, 141.50,
ꢁ
cooled to 0 C compound 9a–9m was added dropwise. The reac- (500 MHz, CDCl
3
tion mixture was stirred at room temperature overnight and 136.44, 134.37, 130.42, 128.49, 128.30, 128.26, 127.98, 126.09,
quenched by the addition of water. The organic layer was sepa- 126.05, 123.84, 122.68, 121.89, 118.09, 116.12, 113.46, 110.38,
rated, washed with brine, and dried over anhydrous Na
solution was evaporated to afford the crude product. Then, the (ESI) m/z calculated for C33
2
SO
4
. The 69.48, 55.97, 49.80, 40.72, 25.07, 23.02, 22.66, and 18.93. HRMS
þ
H
35
N
3
O
3
[M þ H] 522.2751 was found
crude product was chromatographed on silica gel (CH Cl /MeOH/ to be 522.2753.
2
2
Et
3
N ¼ 120:1:0.6) to get compound 10a–10 m.
(E)-3–(3-methoxy-4-((3-methylbenzyl)oxy)phenyl)-N-(2-((1,2,3,4-tet-
(
E)-3–(4-(benzyloxy)-3-methoxyphenyl)-N-(2-((1,2,3,4-tetrahydroacri- rahydroacridin-9-yl)amino)ethyl)acrylamide (10f). Yellow powder,
ꢁ
1
din-9-yl)amino)ethyl)acrylamide (10a). Yellow powder, yield: 28%, yield: 33%, m.p. 98–100 C. H NMR (300 MHz, CDCl ): d 7.98 (d,
melting point (m.p.) 158–160 C. H NMR (300 MHz, CDCl
3
ꢁ
1
3
): d 7.93 J ¼ 8.4 Hz, 1H), 7.90 (d, J ¼ 8.3 Hz, 1H), 7.55 (t, J ¼ 12.9 Hz, 2H),
(
d, J ¼ 8.0 Hz, 1H), 7.84 (d, J ¼ 8.0 Hz, 1H), 7.67 (s, 1H), 7.58 (d, 7.40–7.23 (m, 5H), 7.14 (d, J ¼ 6.6 Hz, 1H), 7.02 (d, J ¼ 5.9 Hz, 2H),
J ¼ 15.5 Hz, 1H), 7.48–7.23 (m, 7H), 6.96 (s, 2H), 6.78 (d, J ¼ 8.0 Hz, 6.88 (d, J ¼ 8.5 Hz, 1H), 6.26 (d, J ¼ 15.5 Hz, 1H), 5.16 (s, 2H), 3.91 (s,
1
2
H), 6.43 (d, J ¼ 15.4 Hz, 1H), 5.11 (s, 2H), 3.78 (s, 3H), 3.64 (s, 4H), 3H), 3.70 (s, 4H), 3.04 (s, 2H), 2.74 (s, 2H), 2.37 (s, 3H), and 1.89 (s,
13 13
.95 (s, 2H), 2.58 (s, 2H), and 1.76 (s, 4H). C NMR (500 MHz, 4H). C NMR (500 MHz, CDCl
3
): d 167.58, 158.28, 150.75, 149.95,
CDCl
3
): d 167.63, 158.27, 150.76, 149.83, 149.62, 146.93, 141.40, 149.62, 146.93, 141.48, 138.35, 136.48, 128.81, 128.54, 128.49,
1
1
4
36.56, 128.64, 128.48, 128.12, 128.05, 127.96, 127.25, 123.79, 128.19, 127.95, 127.87, 124.34, 123.83, 122.68, 121.88, 119.85,
22.71, 121.81, 119.83, 118.17, 116.02, 113.44, 110.36, 70.85, 55.96, 118.05, 116.09, 113.40, 110.27, 70.95, 55.96, 49.81, 40.70, 33.70,
9.82, 40.68, 33.69, 25.05, 22.99, and 22.63. HRMS (ESI) m/z calcu- 25.06, 23.01, 22.65, and 21.44. HRMS (ESI) m/z calculated for
þ
þ
lated for C H N O [M þ H] 508.2595 was found to be 508.2592. C H N O [M þ H] 522.2751 was found to be 522.2755.
3
2
33
3
3
33 35 3 3
(E)-3–(4-((4-fluorobenzyl)oxy)-3-methoxyphenyl)-N-(2-((1,2,3,4-tetra-
(E)-3–(4-((3,4-dimethylbenzyl)oxy)-3-methoxyphenyl)-N-(2-((1,2,3,4-
hydroacridin-9-yl)amino)ethyl)acrylamide (10b). Yellow powder, tetrahydroacridin-9-yl)amino)ethyl)acrylamide (10g). Yellow powder,
ꢁ
1
ꢁ
1
yield: 46%, m.p. 75–77 C. H NMR (300 MHz, CDCl ): d 7.95 (d, yield: 45%, m.p. 78–80 C. H NMR (300 MHz, CDCl ): d 7.96 (d,
3
3
J ¼ 8.4 Hz, 1H), 7.86 (d, J ¼ 8.4 Hz, 1H), 7.57 (d, J ¼ 15.5 Hz, 1H), J ¼ 8.3 Hz, 1H), 7.88 (d, J ¼ 8.3 Hz, 1H), 7.58 (d, J ¼ 15.6 Hz, 1H), 7.50
7
7
5
1
1
1
1
1
.52–7.44 (m, 1H), 7.43–7.33 (m, 2H), 7.28 (d, J ¼ 9.0 Hz, 1H), (t, J ¼ 7.4 Hz, 1H), 7.31 (d, J ¼ 7.4 Hz, 1H), 7.20 (s, 1H), 7.18–7.09 (m,
.11–6.90 (m, 4H), 6.80 (d, J ¼ 8.7 Hz, 2H), 6.33 (d, J ¼ 15.6 Hz, 1H), 2H), 6.99 (d, J ¼ 5.2 Hz, 2H), 6.85 (d, J ¼ 8.8 Hz, 1H), 6.67 (s, 1H),
.08 (s, 2H), 3.82 (s, 3H), 3.66 (s, 4H), 2.98 (s, 2H), 2.66 (s, 2H), and 6.30 (d, J ¼ 15.5 Hz, 1H), 5.11 (s, 2H), 3.84 (s, 3H), 3.67 (s, 4H), 3.00
1
3
13
.82 (s, 4H). C NMR (500 MHz, CDCl
3
): d 163.24, 158.61, 156.97, (s, 2H), 2.68 (s, 2H), 2.25 (d, J ¼ 1.6 Hz, 6H), and 1.84 (s, 4H).
C
3
51.36, 147.18, 145.00, 144.95, 136.67, 127.59, 127.57, 124.53, NMR (500 MHz, CDCl ): d 168.27, 155.15, 152.48, 150.05, 149.64,
24.49, 124.44, 123.42, 119.22, 118.41, 117.14, 113.59, 110.89, 143.62, 141.43, 136.88, 136.47, 133.90, 129.82, 129.62, 129.30,
10.75, 108.80, 105.63, 65.51, 51.27, 45.42, 35.58, 27.52, 19.92, 128.68, 127.80, 124.87, 124.00, 123.40, 122.02, 118.18, 118.07,
8.26, 17.94, and 17.30. HRMS (ESI) m/z calculated for C32
H
32FN
3
O
3
114.01, 113.35, 110.27, 70.86, 56.02, 53.46, 50.39, 40.17, 24.54,
þ
[M þ H] 526.25 was found to be 526.2519.
22.58, 21.84, 19.81, and 19.54. HRMS (ESI) m/z calculated for
þ
(E)-3–(4-((4-chlorobenzyl)oxy)-3-methoxyphenyl)-N-(2-((1,2,3,4-tet-
C
34
H
37
3
N O
3
[M þ H] 536.2908 was found to be 536.2907.
rahydroacridin-9-yl)amino)ethyl)acrylamide (10c). Yellow powder, (E)-3–(3-methoxy-4-((4-methylbenzyl)oxy)phenyl)-N-(2-((1,2,3,4-tet-
yield: 28%, m.p. 70–72 C. H NMR (300 MHz, CDCl ): d 7.98 (d, rahydroacridin-9-yl)amino)ethyl)acrylamide (10h). Yellow powder,
ꢁ
1
3
ꢁ
1
J ¼ 8.4 Hz, 1H), 7.90 (d, J ¼ 8.3 Hz, 1H), 7.60–7.51 (m, 2H), 7.43–7.29 yield: 38%, m.p. 92–94 C. H NMR (300 MHz, CDCl
3
): d 7.95 (d,
(
m, 6H), 7.02 (d, J ¼ 7.2 Hz, 2H), 6.84 (d, J ¼ 8.4 Hz, 1H), 6.28 (d, J ¼ 8.4 Hz, 1H), 7.87 (d, J ¼ 8.3 Hz, 1H), 7.56 (d, J ¼ 15.6 Hz, 1H), 7.48
J ¼ 15.6 Hz, 1H), 5.15 (s, 2H), 3.91 (s, 3H), 3.71 (s, 4H), 3.04 (s, 2H), (t, J ¼ 7.5 Hz, 1H), 7.30 (d, J ¼ 7.7 Hz, 3H), 7.16 (d, J ¼ 7.8 Hz, 2H),
1
3
2
1
1
1
3
.73 (s, 2H), and 1.88 (s, 4H). C NMR (500 MHz, CDCl ): d 167.64, 7.02–6.94 (m, 2H), 6.82 (d, J ¼ 8.8 Hz, 1H), 6.69 (s, 1H), 6.30 (d,
3
58.03, 150.91, 149.65, 149.50, 146.68, 141.28, 135.08, 133.82, J ¼ 15.6 Hz, 1H), 5.12 (s, 2H), 3.83 (s, 3H), 3.67 (s, 4H), 2.99 (s, 2H),
1
3
28.81, 128.62, 128.57, 128.24, 127.86, 123.81, 122.77, 121.71, 2.66 (s, 2H), 2.33 (s, 3H), and 1.83 (s, 4H). C NMR (500 MHz,
19.70, 118.38, 115.87, 113.52, 110.40, 70.12, 55.94, 49.87, 40.68, CDCl ): d 168.11, 156.02, 152.00, 149.95, 149.66, 144.58, 141.47,
3.53, 25.02, 22.96, and 22.57. HRMS (ESI) m/z calculated for 141.45, 137.81, 133.51, 129.31, 127.85, 127.35, 125.79, 123.94,
3
þ
C
32
H
(
32ClN
3
O
3
[M þ H] 542.2205 was found to be 542.2206.
123.20, 121.97, 118.55, 118.17, 114.55, 113.42, 110.33, 70.78, 56.03,
E)-3–(4-((4-bromobenzyl)oxy)-3-methoxyphenyl)-N-(2-((1,2,3,4-tet-
50.21, 40.30, 32.31, 24.67, 22.68, 22.05, and 21.22. HRMS (ESI) m/z
þ
rahydroacridin-9-yl)amino)ethyl)acrylamide (10d). Yellow powder, calculated for C33
H
35
N
3
O
3
[M þ H] 522.2751 was found to be
ꢁ
1
yield: 31%, m.p. 73–75 C. H NMR (300 MHz, CDCl
3
): d 7.97 (d, 522.2749.
(E)-3–(4-((4-(tert-butyl)benzyl)oxy)-3-methoxyphenyl)-N-(2-((1,2,3,4-
m, 4H), 7.01 (d, J ¼ 6.5 Hz, 2H), 6.83 (d, J ¼ 8.4 Hz, 1H), 6.29 (s, 1H), tetrahydroacridin-9-yl)amino)ethyl)acrylamide (10i). Yellow powder,
J ¼ 8.4 Hz, 1H), 7.90 (d, J ¼ 8.4 Hz, 1H), 7.63–7.47 (m, 4H), 7.36–7.29
(
ꢁ
1
5
1
1
1
1
2
.12 (s, 2H), 3.89 (s, 3H), 3.69 (s, 4H), 3.04 (s, 2H), 2.73 (s, 2H), and yield: 15%, m.p. 78–80 C. H NMR (300 MHz, CDCl ): d 7.95 (d,
3
1
3
.88 (s, 4H). C NMR (500 MHz, CDCl
3
): d 167.50, 158.45, 150.66, J ¼ 8.3 Hz, 1H), 7.87 (d, J ¼ 8.3 Hz, 1H), 7.56 (d, J ¼ 15.6 Hz, 1H), 7.49
49.66, 149.49, 147.10, 141.35, 135.62, 131.77, 128.90, 128.44, (d, J ¼ 7.4 Hz, 1H), 7.41–7.28 (m, 5H), 6.99 (d, J ¼ 7.1 Hz, 2H), 6.86
28.33, 128.23, 123.82, 122.65, 121.96, 121.71, 119.93, 118.30, (d, J ¼ 8.7 Hz, 1H), 6.48 (s, 1H), 6.27 (d, J ¼ 15.5 Hz, 1H), 5.12 (s, 2H),
16.18, 113.53, 110.39, 70.15, 55.94, 49.78, 40.75, 33.81, 25.09, 3.84 (s, 3H), 3.66 (s, 4H), 3.00 (s, 2H), 2.69 (s, 2H), 1.84 (s, 4H), and
1
3
3.02, and 22.68. HRMS (ESI) m/z calculated for C H BrN O
1.30 (d, J ¼ 3.1 Hz, 9H).
C NMR (500 MHz, CDCl ): d 167.52,
3
2
32
3
3
3
þ
[M þ H] 586.17 was found to be 586.1696.
151.08, 150.70, 150.08, 149.64, 141.59, 137.66, 133.51, 128.49,
127.76, 127.23, 127.20, 125.58, 125.52, 123.85, 122.64, 121.92,
(
E)-3–(3-methoxy-4-((2-methylbenzyl)oxy)phenyl)-N-(2-((1,2,3,4-tet-
rahydroacridin-9-yl)amino)ethyl)acrylamide (10e). Yellow powder, 119.86, 117.91, 116.16, 113.34, 110.27, 70.75, 63.69, 55.99, 49.76,
ꢁ
1
yield: 13%, m.p. 93–95 C. H NMR (300 MHz, CDCl
3
): d 7.98 (d, 40.72, 34.60, 31.34, 25.07, 23.01, and 22.66. HRMS (ESI) m/z calcu-
þ
J ¼ 8.5 Hz, 1H), 7.91 (d, J ¼ 8.4 Hz, 1H), 7.61–7.52 (m, 2H), 7.41 (d, lated for C36
H
41
N
3
O
3
[M þ H] 564.3221 was found to be 564.3216.
J ¼ 6.6 Hz, 1H), 7.35 (t, J ¼ 7.5 Hz, 1H), 7.28 (s, 1H), 7.23 (s, 2H), 7.14
(E)-3–(4-((4-cyanobenzyl)oxy)-3-methoxyphenyl)-N-(2-((1,2,3,4-tetra-
(
(
d, J ¼ 10.2 Hz, 1H), 7.10–7.01 (m, 2H), 6.90 (d, J ¼ 8.2 Hz, 1H), 6.26 hydroacridin-9-yl)amino)ethyl)acrylamide (10j). Yellow powder, yield:
ꢁ
1
d, J ¼ 15.5 Hz, 1H), 5.16 (d, J ¼ 9.0 Hz, 2H), 3.89 (s, 3H), 3.70 (s, 4H), 22%, m.p. 88–90 C. H NMR (300 MHz, CDCl ): d 7.97 (d, J ¼ 8.6 Hz,
3

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
499
1
H), 7.87 (d, J ¼ 8.2 Hz, 1H), 7.64 (t, J ¼ 7.3 Hz, 2H), 7.54 (t, equine serum (eqBuChE)). ATC/BTC iodide solution (0.075 M) was
J ¼ 7.5 Hz, 2H), 7.49–7.41 (m, 1H), 7.31–7.17 (m, 2H), 7.01 (d, dissolved in deionized water. DTNB solution (0.01 M) was prepared
J ¼ 10.8 Hz, 2H), 6.76 (d, J ¼ 8.2 Hz, 1H), 6.45 (d, J ¼ 15.6 Hz, 1H), using water containing 0.15% (w/v) sodium bicarbonate. The assay
5
1
1
1
1
.19 (s, 2H), 3.87 (s, 3H), 3.73 (s, 4H), 2.98 (s, 2H), 2.63 (s, 2H), and solution was prepared as follows: potassium dihydrogen phos-
.81 (s, 4H). C NMR (500 MHz, CDCl
3
): d 168.02, 152.26, 149.76, phate (1.36 g, 10 mmol) was dissolved in 100 ml of water. The pH
49.16, 142.15, 141.18, 132.48, 129.51, 128.70, 127.47, 125.22, of the solution was then adjusted to 8.0 ± 0.1 with KOH. Stock sol-
24.01, 123.29, 121.78, 118.76, 118.68, 118.25, 114.27, 113.61, utions of the test samples were dissolved in ethanol to give a final
1
3
11.78, 110.52, 69.88, 56.03, 50.25, 40.28, 24.57, 22.62, and 21.92. concentration of 10 ꢂ 4M when diluted to the final volume of
þ
HRMS (ESI) m/z calculated for C33
H N
32 4
O
3
[M þ H] 533.2547 was 3.32 ml. For each compound, a dilution series of five different con-
9
found to be 533.2542.
centrations 10 ꢂ 5–10 ꢂ 9 M were prepared.
(
E)-3–(3-methoxy-4-((4-(trifluoromethyl)benzyl)oxy)phenyl)-N-(2-
For measurement, a cuvette containing 3 mL of phosphate buf-
(
(1,2,3,4-tetrahydroacridin-9-yl)amino)ethyl)acrylamide (10k). Yellow fer, 100 lL of eeAChE or eqBuChE, 100 lL of DTNB, and 100 lL of
powder, yield: 51%, m.p. 60–62 C. H NMR (300 MHz, CDCl
3
): d the test compound solution was added in sequence. The reaction
ꢁ
1
7
2
7
2
4
1
1
1
2
5
.98 (d, J ¼ 8.5 Hz, 1H), 7.90 (d, J ¼ 8.4 Hz, 1H), 7.65 (d, J ¼ 8.0 Hz, was initiated by adding 20 lL of ATC or BTC and the solution was
H), 7.59 (d, J ¼ 7.7 Hz, 2H), 7.57–7.47 (m, 3H), 7.39–7.30 (m, 1H), mixed immediately. Two minutes after ATC or BTC addition, the
ꢁ
.03 (d, J ¼ 8.9 Hz, 2H), 6.84 (d, J ¼ 8.0 Hz, 1H), 6.32 (s, 1H), 5.24 (s, absorption was measured at 25 C at 412 nm, using a Shimadzu
H), 3.92 (s, 3H), 3.71 (s, 4H), 3.04 (s, 2H), 2.73 (s, 2H), and 1.88 (s, 160 spectrophotometer. For the reference value, 100 lL of water
1
3
H). C NMR (500 MHz, CDCl
3
): d 167.57, 158.03, 150.89, 149.68, was replaced for the test compound solution. For the blank con-
49.37, 141.27, 140.70, 128.58, 128.42, 127.89, 127.22, 125.61, trol, additional 100 lL of water was also replaced for the enzyme
25.59, 125.56, 123.83, 122.74, 121.70, 119.71, 118.46, 115.91, solution. The measurement of each concentration was assayed in
13.51, 110.44, 70.04, 55.96, 49.84, 40.70, 33.53, 25.02, 22.96, and triplicate. GraphPad Prism 5 (GraphPad Software, Inc., La Jolla, CA)
þ
2.57. HRMS (ESI) m/z calculated for C H F N O [M þ H]
was used for data processing. The inhibition curve was fitted by
plotting percentage enzyme activity (100% for the reference) ver-
sus the logarithm of test compound concentration. The half max-
3
3 32 3 3 3
76.2469 was found to be 576.2457.
E)-3–(3-methoxy-4-((3-methoxybenzyl)oxy)phenyl)-N-(2-((1,2,3,4-
tetrahydroacridin-9-yl)amino)ethyl)acrylamide (10l). Yellow powder,
(
imal inhibitory concentration (IC50) values were calculated
ꢁ
1
according to the curve and the data were shown in mean ± SEM.
yield: 13%, m.p. 63–65 C. H NMR (300 MHz, CDCl
3
): d 7.96 (d,
J ¼ 8.4 Hz, 1H), 7.87 (d, J ¼ 8.5 Hz, 1H), 7.58 (d, J ¼ 15.5 Hz, 1H), 7.48
(
6
3
4
1
1
1
2
t, J ¼ 7.5 Hz, 1H), 7.29 (d, J ¼ 8.2 Hz, 2H), 6.99 (d, J ¼ 5.7 Hz, 4H),
Inhibition of self-induced Ab1–42 aggregation
.83 (t, J ¼ 7.0 Hz, 3H), 6.34 (d, J ¼ 15.6 Hz, 1H), 5.15 (s, 2H), 3.86 (s,
Inhibitory effects of the compounds on self-induced Ab1–42
aggregation were tested through a Thioflavin T (ThT)-(T3516,
Sigma-Aldrich) binding assay. Firstly, aliquots of 2 lL of Ab1–42
H), 3.80 (s, 3H), 3.69 (s, 4H), 2.99 (s, 2H), 2.66 (s, 2H), and 1.83 (s,
1
3
3
H). C NMR (500 MHz, CDCl ): d 168.59, 159.88, 153.53, 151.87,
49.97, 149.77, 141.64, 138.26, 130.43, 129.94, 129.90, 129.72,
27.95, 124.20, 123.72, 122.22, 119.36, 118.20, 113.49, 113.47,
12.68, 110.24, 70.75, 56.12, 55.27, 39.70, 29.34, 27.23, 22.29, and
(AS-64129–05 Anaspec Inc.) containing 2 mg/mL HFIP (1,1,1,3,3,3-
hexafluoro-2-propanol, 52517( Sigma-Aldrich) were stocked in
DMSO. Then, they were diluted with 0.215 M sodium phosphate
buffer (pH 8.0) to the final concentration of 500 lM. Test com-
pounds were dissolved in DMSO and then prepared at a concen-
tration of 25 lM by the buffer. Resveratrol was used as a positive
control. The Ab1–42 and the test sample solutions were incubated
in a 96-well plate for 24 h at the room temperature. After the incu-
bation, the tested compounds were diluted to a final volume of
þ
1.29. HRMS (ESI) m/z calculated for C H N O [M þ H] 538.27
3
3 35 3 4
was found to be 538.2697.
(E)-3–(3-methoxy-4-((4-methoxybenzyl)oxy)phenyl)-N-(2-((1,2,3,4-
tetrahydroacridin-9-yl)amino)ethyl)acrylamide (10m). Yellow powder,
yield: 38%, m.p. 67–69 C. H NMR (300 MHz, CDCl
J ¼ 8.3 Hz, 1H), 7.99 (d, J ¼ 8.4 Hz, 1H), 7.66 (d, J ¼ 15.4 Hz, 1H), 7.61
ꢁ
1
3
): d 8.06 (d,
(
d, J ¼ 7.6 Hz, 1H), 7.43 (d, J ¼ 8.8 Hz, 1H), 7.37 (t, J ¼ 3.9 Hz, 2H),
150 lL with 50 mM glycine-NaOH buffer (pH 8.5) containing 5 mM
7
1
2
1
1
1
4
.10 (d, J ¼ 3.7 Hz, 4H), 6.94 (d, J ¼ 7.6 Hz, 2H), 6.36 (d, J ¼ 15.6 Hz,
ThT. Fluorescence intensity was read (excitation wavelength
450 nm, emission wavelength 485 nm) on a SpectraMax Paradigm
Multimode Reader (Molecular Device, Sunnyvale, CA).
H), 5.26 (s, 2H), 3.98 (s, 3H), 3.90 (s, 3H), 3.78 (s, 4H), 3.12 (s, 2H),
1
3
3
.81 (s, 2H), and 1.96 (s, 4H). C NMR (500 MHz, CDCl ): d 168.22,
59.84, 155.64, 152.26, 149.78, 149.63, 144.21, 141.26, 138.23,
29.70, 129.42, 128.06, 125.39, 123.95, 123.34, 121.87, 119.40,
18.43, 114.36, 113.43, 112.74, 110.40, 70.72, 56.00, 55.25, 50.27,
0.31, 32.00, 24.66, 22.65, and 21.97. HRMS (ESI) m/z calculated for
The calculation of the inhibitory rate of Ab1–42 self-induced
aggregation was performed as the following equation: (1 ꢂ IFi
/
I ) ꢀ 100%. I and I were the fluorescence intensities measured
Fc
Fi
Fc
[M þ H]þ 538.27 was found to be 538.2698.
in the presence and absence of inhibitors, respectively, after sub-
tracting the background fluorescence of the 5 mM ThT solution.
Each compound was measured in triplicate.
33 35 3 4
C H N O
Biological activity
In vitro inhibitory evaluations on ChEs
AChE (EC 3.1.1.7, Type VI-S, from electric eel, C3389) and BuChE Behavioral studies were performed using the Morris water maze
Behavioral testing
2
5
2
6
(EC 3.1.1.8, from equine serum, C0663), 5,5’-dithiobis (2-nitroben- according to our method reported previously . Adult male
zoic acid) (DTNB, D218200), acetylthiocholine iodide (ATC, A5751) imprinting control region (ICR) mice (8–10 weeks old, weight
and butyrylthiocholine iodide (BTC, B3253) used as substrates 20–25 g) were obtained from the Yangzhou University Medical
were obtained from Sigma-Aldrich (St. Louis, MO).
Center (Yangzhou, China). All animal handling and experimental
The inhibitory capacity of the synthesised compounds on AChE protocols were approved by the Institutional Animal Care and Use
and BuChE biological activity in this paper were evaluated accord- Committee of the China Pharmaceutical University. Scopolamine
2
1
ing to our previously reported method . Briefly, AChE/BuChE hydrobromide was purchased from Aladdin Reagents (S107418,
stock solution was diluted with 0.2 M phosphate buffer pH 8.0 to Shanghai, China). Tacrine that was used as the positive control
give 2.5 units/mL (for electrophorus electricus (eeAChE) and was synthesized by our lab (purity >95%).

5
00
J. ZHU ET AL.
Thirty mice were randomly separated into five subgroups (six EG1150H, Germany). Subsequently, 5 lm sections were prepared
mice for each group): (i) vehicle as a blank control group, (ii) sco- from these paraffin blocks. They were deparaffinated and stained
polamine as a model group, (iii) tacrine plus scopolamine as a with hematoxylin and eosin (H&E) for histopathological
positive control, (iv) compound 10d plus scopolamine as a test examinations.
group, and (v) compound 10g plus scopolamine as a test group.
The mice for model, tacrine, compounds 10d, and 10g groups
Molecular docking studies
received intraperitoneal (i.p). injections of scopolamine (1 mg/kg),
while the blank control group was injected with saline. Tacrine,
compounds 10d, and 10g were orally administered (30 mg/kg) to
mice in groups (iii), (iv), and (v) 30 min before the i.p. administra-
tion of scopolamine or saline.
Molecular docking studies were completed by CDOCKER module
implemented in Discovery Studio (DS; version 3.0, BIOVIA, San
Diego, CA). The X-ray crystal structures of huAChE and huBuChE
with small molecular ligands were downloaded from protein data
bank (PDB, ID: 4EY7, 4TPK)
essed by “prepare protein” module in DS to give the structures
suitable for docking. Missed sidechains of the proteins were added
and the water molecules were removed, later the structures were
protonated at pH 7.4. “Prepare ligands” module in DS was applied
for the structural preparation of the test compounds, which were
then protonated at pH 7.4. The resulted molecules were subse-
quently minimized by “minimize ligands” module. The “full mini-
misation” algorithm was used to carry out the minimisation, with
max steps set to 2000, the root mean square (RMS) gradient set to
A circular pool (120 cm in diameter, 60 cm height) which was
fixed with an escape platform (10 cm diameter) and filled to a
27,28
. The structures were initially proc-
ꢁ
depth of 40 cm fresh water (kept at 25 C) composed the water
maze and this water maze placed in a lit room was used to evalu-
ate the cognition function and memory capacity of mice according
2
4
to our previously described method . After five days of learning
and memory training, a probe trial was performed on day 6. For
the evaluation of cognitive function, each mouse was individually
trained on a visible-platform (labeled by a small flag, 5 cm tall) for
two days and was followed by a hidden-platform version (placed
1
3
cm below the surface of the water) of the water maze from day
to day 5. All mice were subjected to two training trials daily,
0.01. Other parameters were set as default.
For molecular docking, the binding site was defined as a site
each of which lasted for 90 s. The time for each mouse to find the
platform (a successful escape) was recorded. If a mouse failed to
reach the platform within 90 s, the test was terminated and the
mouse was carefully navigated to the platform by hands. Each
mouse was kept on the platform for 30 s whether it succeeded or
failed to reach the platform. On the last day (day 6), a probe trial
was given to the mice in which the platform was removed from
the pool and each mouse was allowed 90 s to search for the plat-
form. The time the animals spent to reach the missing platform
location and the number of times they crossed that location were
recorded.
sphere (in 10 Å radius) around the original ligands in the co-crystal
structures. The simulated annealing parameters were set as fol-
lows: heating steps and cooling steps were set to 2000 and 5000,
respectively, while heating and cooling temperature was set to
700 and 300, respectively. Other parameters were kept as default.
Ten top-ranked conformations for each docked compound were
retained and visually inspected for binding pattern analysis, which
2
9
was visualised and depicted in PyMOL software.
Results
Data for the time of escape latency, the swimming trajectory,
and the number of visual platform crossings were recorded by
Panlab SMART 3.0 (Panlab, S.L.U, Spain) and GraphPad Prism 5 was
used to evaluate the level of memory retention.
Compound design and chemistry
Compound 10a, which was previously reported by our group, was
24
used as the lead compound for structural modification .
Differently substituted benzyl was introduced to the ferulic acid
moiety of 10a (Table 1).
Hepatotoxicity studies
Hepatotoxicity evaluation was performed on adult male ICR mice
The preparation of the target compounds started from
1
methyl 2-aminobenzoate through seven steps as described in
(8–10 weeks old, weighing 20–25 g) purchased from the Yangzhou
1
Scheme 1 methyl 2-aminobenzoate was hydrolysed to yield 2-
University Medicine Centre (Yangzhou, China) according to our
2
2
6
aminobenzoic acid , which is condensed with cyclohexanone to
method previously described . Tacrine, compounds 10d, and 10g
were dissolved in a sodium carboxymethyl cellulose (CMC-Na)
solution (0.5 g CMC-Na in 100 ml distilled water) corresponding to
3
1
give 9-chloro-1,2,3,4-tetrahydroacridine . N –(1,2,3,4-tetrahydro-
4
acridin-9-yl)ethane-1,2-diamine was obtained by aminolysis of
9-chloro-1,2,3,4-tetrahydroacridine
with
ethane-1,2-diamine.
3
mg/100 g body weight, concentration of 151.5 lmol/kg body
Compound 5 was substituted with benzyl bromide or benzyl
chloride, followed by the hydrolysis of ester group to form
compound 8a–8m. Then, the hydrolysates were treated with
thionyl chloride to prepare the corresponding chlorides, which
were condensed with compound 4 to afford target compounds
weight were administered intragastrically (i.g.). Heparinised serum
was obtained from the retrobulbar plexus 8, 22, and 36 h after the
administration and were subjected to hepatotoxicity evaluation.
The activity of aspartate aminotransferase (AST) and alanine ami-
notransferase (ALT), two indicators of liver damage, was deter-
mined by a commercial assay kit (EF551 and EF550 for ALT, EH027
and EF548 for AST, Wako, Japan). The data were analysed by
Biochemical Analyzer (HITACHI 7020, Japan).
10a–10m.
Cholinesterase inhibitory activity and SAR analysis
Mice were sacrificed 1 h after the collection of retrobulbar
blood and livers were removed for morphological studies by using All the synthesised compounds were assayed against AChE from
the immunohistochemical method. Two 3 mm sections of each eeAChE and BuChE from eqBuChE according to the modified
3
0
liver extending from the hilus to the margin of the left lateral lobe method of Ellman et al. , using tacrine as a reference standard.
were isolated using ultra-thin semiautomatic microtome (Leica Inhibitory activities were presented as IC50 (nM). The activity
RM2245, Germany) and placed in 10% buffered formaldehyde results and selectivity index (SI) are summarized in Table 1. All the
immediately, fixed for two days, and embedded together in one compounds displayed potent inhibitory activities, with IC50 values
paraffin block by using paraffin embedding station (Leica in the two or three-digit nanomolar range.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
501
Firstly, we introduced methyl group to the benzyloxyl moiety. Interestingly, for the multi-substituted compound, the 3,4-diCH
3
Compared to compound 10a, methyl substitution at ortho- or (10g) was the most potent AChE inhibitor (IC50 ¼ 37.02 nM).
meta- position of benzyloxyl (10e and 10f, respectively) showed
Next, we explored the inductive effects of substituents on ChEs
enhanced activities while the para-substituted compound led to inhibitory activities. The electron-withdrawing groups (EWGs), such
(10k) showed reduced AChE inhibitory
reduced inhibitory activity and lower selectivity against AChE. as ꢂCN (10j) and ꢂCF
3
activity while they were favorable for BuChE inhibition.
Then, we designed compounds bearing halogen substitution.
4-F (10b) and 4-Cl (10c) showed comparable activity to compound
Table 1. Structures, eeAChE and eqBuChE inhibitory activities of target
compounds.
10a, but the 4-Br substituted compound (10d) displayed obvious
improvement. The impact was consistent with the results men-
tioned above. The stronger electron-withdrawing ability of the
substituent was, the poorer the inhibition activity of the com-
pound was.
Next, we synthesised compounds with electron-donating
groups. para-substituted –OCH (10m) showed comparable activity
3
a
to 10a. However, replacement with 3-OCH (10l) and 4-t-Bu (10i)
IC50 (nM) ± SEM
3
was not tolerated.
b
c
d
Compound
R
AChE
BuChE
SI
1
1
1
1
1
1
1
1
1
1
1
1
1
0a
0b
0c
0d
0e
0f
0g
0h
0i
H
92.2 ± 18.3
114.8 ± 31.2
100.8 ± 14.6
49.5 ± 3.8
64.5 ± 7.4
63.2 ± 13.5
37.0 ± 6.4
139.9 ± 34.4
284.1 ± 59.2
107.1 ± 29.9
128.6 ± 16.5
143.0 ± 40.1
75.5 ± 12.8
14.5 ± 2.6
215.4 ± 39.5
163.1 ± 16.6
62.5 ± 9.9
69.4 ± 13.4
98.6 ± 19.1
256.7 ± 3.7
101.4 ± 15.5
147.5 ± 22.4
158.3 ± 22.8
79.8 ± 8.9
0.4
0.7
1.6
0.7
0.7
0.2
0.4
4-F
4-Cl
4-Br
2-CH
3-CH
3,4-diCH
4-CH
4-t-Bu
4-CN
4-CF
Binding mode analysis by molecular modeling
To further explore the binding interactions of the synthesised
compounds with ChEs, we next carried out molecular docking
studies by using DS. Binding mode of huAChE-10g and huBuChE-
3
3
3
0.9 10k were analysed. The two compounds were selected as repre-
1.8
1.3
2.4
3
sentatives for their most potent activity on AChE or BuChE. As
shown in Figure 1, compound 10g simultaneously occupied both
the CAS and PAS of AChE. The 1,2,3,4-tetrahydroacridin core was
0j
0k
0l
3
52.7 ± 6.3
3-OCH
4-OCH
–
3
3
195.2 ± 16.5
134.0 ± 11.1
4.5 ± 0.4
0.7
0m
0.6 located at the CAS through a p-p stacking interaction with Trp86.
3.2
Tacrine
The NH- group on 1,2,3,4-tetrahydroacridin ring interacted with
a
Concentration required for 50% inhibition of ChEs, data were shown in the backbone of Tyr337 through a hydrogen bond. The carbonyl
mean ± SEM of triplicate independent experiments.
group of the ferulic acid moiety formed a hydrogen bond with
the side chain of Tyr341. The phenyl core of the ferulic acid
moiety formed p-p stacking interactions with Trp286 and Tyr341.
b
AChE (EC 3.1.1.7) from electric eel.
BuChE (EC 3.1.1.8) from horse serum.
c
d
Selectivity index (SI) is defined as AChE IC50/BuChE IC50
.
ꢁ
Scheme 1. Reagents and conditions: (a) NaOH, H
2
O, room temperature (r.t), 12 h; (b) POCl
3
, reflux, 3 h; (c) 1-pentanol, ethylenediamine, 160 C, 18 h; (d) K
4 2 2
2 3
CO , DMF,
ꢁ
ꢁ
8
2
2 C, 4 h; (e) NaOH, MeOH, H O, 82 C, 3 h; (f) SOCl , CH Cl , reflux, 3 h; and (g) K CO , Na SO , CH Cl , r.t, 24 h.
2
2
2
2
3
2

5
02
J. ZHU ET AL.
Figure 1. Predicted binding mode for Compound 10g with AChE. (A) View from the top of the binding gorge, showing Compound 10g as cyan sticks, bound to the
surface of huAChE (PDB code: 4EY7). (B) Profile view of Compound 10g as magenta sticks bound in the gorge pocket of huAChE (grey surface). Key residues for the
binding of Compound 10g are shown as yellow sticks. (C) Two dimensional schematic diagram of docking model of Compound 10g with AChE. Intermolecular interac-
tions were described as dashed lines in different colors: green: hydrogen bond; purple: p–p stacking; and red: p–p T-shaped.
3
1
Such polar interactions stabilised the occupation of the ferulic acid neuroprotective agents . Furthermore, one single molecule simul-
moiety in the PAS site. The phenyl rings interacted with multiple taneously inhibits Ab1–42 aggregation and ChEs activity resulting
residues through van der Waals interactions, such as Leu289, in an anti-AD agent that is beneficial in the early and advanced
3
2,33
Ser293, and Tyr124.
stages of AD
.
Compared to the narrow and long binding site of huAChE,
Compounds 10d, 10 g, and 10j that showed potent ChEs
huBuChE contains a much broader and larger binding cavity. inhibition activity were selected for further evaluation for their
As a result, compound 10k exhibited a U-shaped conformation inhibitory capacity on self-induced Ab1–42 aggregation. A thioflavin
3
4
(
Figure 2), which was obviously different from the linear conform- T-based fluorometric assay was performed . Results are summar-
ation of 10g. Therefore, the U-shaped conformation of 10k can ised in Table 2. Three compounds displayed significant inhibition
better occupy the active site of BuChE. For benzyloxyl moiety, the on the aggregation of Ab1–42 with the inhibitory rate 66.84, 65.49,
phenyl ring interacted with Trp 82 through p–p stacking. The and 49.51%, respectively compared to 45.72% with resveratrol as
fluorine atoms formed a hydrogen bond with Tyr128 and two reference standard under the concentration of 25 lM. These
fluorine bonds with Gly115 and Glu197, which better occupied the results suggested that our compounds were MTDLs. Because of
bottom of the binding pocket. The phenyl ring of the ferulic acid these promising in vitro data, the in vivo activities of compounds
interacted with Tyr332 and Ala328 through p–p stacking and 10d and 10g were further tested.
p-Alkyl contact, respectively. The nitrogen atom of amide moiety
formed a hydrogen bond with Pro285. Multiple van der Waals
contacts were observed between compound 10k and different res-
Behavioral studies
idues such as Thr284 and Ala328, providing stronger binding
affinity.
To further evaluate the ability of the synthesised compound to
improve the cognitive ability, the optimal compounds 10d and
1
0g, were selected for in vivo behavioral analysis in scopolam-
ine-induced cognition-impaired adult ICR mice (male mice,
–10 weeks old, weight 20–25 g) using a Morris water maze with
tacrine at 30 mg/kg as a positive control. Thirty mice were ran-
in the brain lead to domly allocated into five subgroups (n ¼ 6 for each group): con-
Inhibition of self-induced Ab1–42 aggregation by selected
compounds
8
The formation and accumulation of Ab
1
–42
3
0
neurotoxicity in AD . Thus, targeting self-induced Ab1–42 aggrega- trol, model, tacrine, compounds 10d, and 10g. The mean
tion represents promising strategy to discover novel escape latency values, searching distances and swimming speed
a

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
503
Figure 2. Predicted binding mode for Compound 10k with BuChE. (A) View from the top of the binding cave, showing Compound 10k as magenta sticks, bound to
the surface of huBuChE (PDB code: 4TPK). (B) Profile view of Compound 10k as magenta sticks bound in the pocket of huBuChE (grey surface). Key residues for the
binding of Compound 10k are shown as yellow sticks. (C) Two dimensional schematic diagram of docking model of Compound 10k with BuChE. Intermolecular interac-
tions were described as dashed lines in different colors: green: hydrogen bond; purple: p–p stacking; red: p-alkyl contact.
The time of escape latency for mice that were administered tac-
Table 2. Inhibition of Ab1–42 self-induced aggregation of the synthesised
rine (25.85 ± 5.52 s, p > 0.05), compounds 10d (19.08 ± 2.56 s,
compounds.
p < 0.01), and 10g (24.73 ± 3.54 s, p < 0.05) were remarkably
reduced compared to the model group. Interestingly, the mice
a
Inhibitory rate of compounds
(
25 lM) on self-induced
Ab1-42 aggregation (%)
treated with compounds 10d (19.08 ± 2.56 s, p < 0.01), and 10g
24.73 ± 3.54 s, p < 0.05) both demonstrated a more favorable
Compound
(
1
1
1
0d
0g
0j
66.84
65.49
49.51
45.72
amelioration of the cognitive and memory functions compared
to the tacrine (25.85 ± 5.52 s, p > 0.05) group (Figure 3(A)).
Similarity, tacrine group (725.64 ± 165.15 cm, p > 0.05), com-
pounds 10d group (628.95 ± 87.55 cm, p < 0.05), and 10g group
Resveratrol
a
Inhibitory rate ¼ (1 ꢂ IFi/IFc) ꢀ 100%, IFi and IFc were the fluorescence measured
in the presence and absence of the tested compound.
(612.85 ± 81.54 cm, p < 0.05) demonstrated a significantly short-
ened distance to the target than the model
group (1146.00 ± 158.78 cm). Besides, the distance to the target
for compounds 10d (628.95 ± 87.55 cm, p < 0.05) and 10g
Table 3. The mean escape latency, searching distances, and swimming speed of
five subgroups.
(612.85 ± 81.54 cm, p < 0.05) group were obviously shorter than
Latency to
target (s)
Distance to
target (cm)
Speed in virtual
platform (cm/s)
a
tacrine group (725.64 ± 165.15 cm, p > 0.05), which further sug-
gested that compounds 10d and 10 g were more favorable
(Figure 3(B)). The results presented in Figure 3(C) showed that
the swimming speed for each group was almost equivalent,
which demonstrated that long-term use of compounds 10d and
Group
Control
Model
Compound 10d
Compound 10g
Tacrine
8.61 ± 1.53
40.51 ± 3.83
19.08 ± 2.56
24.73 ± 3.54
25.85 ± 5.52
218.50 ± 30.09
1146.00 ± 158.78
628.95 ± 87.55
612.85 ± 81.54
725.64 ± 165.15
19.42 ± 2.30
15.16 ± 1.18
18.82 ± 5.36
17.27 ± 3.39
15.78 ± 1.03
10g are safe at 30 mg/kg/day. Besides, compounds 10d
a
Values are expressed as the mean ± SEM of six independent experiments.
(18.82 ± 5.36 cm/s) and 10 g (17.27 ± 3.39 cm/s) did not injure the
moving and exploring activities of mice, which are safer than
tacrine (15.78 ± 1.03 cm/s). Meanwhile, the data of escape latency
of all the groups are shown in Table 3. It was clear that admin- and searching distance were also supported by the analysis of
istration of scopolamine led to a remarkable delay of the escape the swimming trajectories of the mice in each group. For the
latency (40.51 ± 3.83 vs. 8.61 ± 1.53 s, p < 0.001) as compared to mice in scopolamine model group (Figure 3(D)), the trajectory
the control group, which suggested that administration of was longer and more disordered than the control group. The
scopolamine led to a spatial memory deficiency in the mice. performance of tacrine group was slightly improved. Mice

5
04
J. ZHU ET AL.
Figure 3. Effects of memory retention determined by the Morris water maze test. (A) The escape latency values to target. (B) The swimming distance to target. (C) The
swimming speed in virtual platform. (D) The representative tracks of mice in each group in the Morris water maze test. Data are presented as the mean ±SEM (n ¼ 6;
ꢃ
ꢃꢃ
ꢃꢃꢃ
ns p > 0.05, p < 0.05, p < 0.01, and p < 0.0001 versus the model group).
accordance with the ALT and AST data. Treatment of com-
pounds 10d (Figure 5(B)) and 10g (Figure 5(C)) did not cause
adverse morphologic changes in liver compared to the control
group (Figure 5(A)). However, when we investigated the antipro-
liferative activity of compounds 10d and 10g on HepG2 cells
by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
Table 4. The ALT and AST levels (U/L) at the three-time points of five
subgroups.
a
ALT (U/L)
Group
8 h
22 h
36 h
Control
Model
Tacrine
Compound 10d
35.4 ± 3.4
23.7 ± 3.3
31.8 ± 2.0
26.0 ± 2.7
36.4 ± 4.4
39.0 ± 3.8
35.7 ± 2.6
25.4 ± 2.8
32.9 ± 2.3
38.5 ± 4.8
30.4 ± 3.0
35.9 ± 4.4
25.1 ± 2.7
33.9 ± 2.3
31.8 ± 4.7
(MTT) assays, we surprisingly observed that the two compounds
showed relatively strong toxic effects (Supplementary Table S1).
The two compounds showed high antiproliferative potency at
Compound 10g
5
lM, however, we did not observe obvious hepatoxicity in the
AST (U/L)
22 h
in vivo assays, even at a much higher concentration. In sum-
mary, the results indicated that these compounds may still have
potential hepatoxicity especially in long-term use, but consider-
Group
8 h
36 h
Control
Model
Tacrine
Compound 10d
Compound 10g
a
118.9 ± 8.1
114.1 ± 10.8
133.7 ± 9.0
130.9 ± 14.5
122.1 ± 11.6
124.6 ± 8.0
134.6 ± 9.0
124.3 ± 12.1
129.6 ± 9.5
128.7 ± 11.2
118.3 ± 5.6
111.4 ± 11.4
112.4 ± 11.4 ing the much lower IC50 values of the two compounds on ChEs
126.8 ± 8.7
137.3 ± 9.0
compared to the MTT assays, such potential hepatoxicity can be
controlled by rational administration of the molecules.
Additionally, the results also inspired us to further improve the
safety of compounds in the subsequent structural modification
Values are expressed as the mean ± SEM of six independent experiments.
treated with compounds 10d and 10g showed much-shortened of these hybrids.
distances compared to tacrine group. Taken together, these
behavioral performances demonstrated that compounds 10d
and 10g can markedly improve the cognitive function of mice.
Conclusion
The effectivity of 10d and 10g for symptom amelioration in AD
mice model identifies them as potential anti-AD agents, espe-
cially for the treatment of cholinergic deficiency.
In summary, our study involved the design, synthesis and bio-
logical evaluation of a series of tacrine-ferulic acid hybrids with
the aim to identify MTDLs for the treatment of AD. All of the
target compounds simultaneously inhibited AChE and BuChE
with the IC50 values in the nanomolar range. The binding man-
ner of compound 10g to huAChE and compound 10k to
Hepatotoxicity studies
To determine the possible drug-induced hepatotoxicity, 10d and huBuChE were also analysed by molecular docking studies. The
0g were selected for preliminary hepatotoxicity evaluation. The representatives, compounds 10d, 10g, and 10j, significantly
1
ALT and AST levels were measured, as shown in Table 4 and inhibited the Ab1–42 self-induced aggregation. In subsequent in
Figure 4. Heparinised serum was obtained after the treatment vivo evaluation, the most promising compound 10 g showed a
of the compounds at 8, 22, and 36 h, respectively. The levels of remarkable improvement of memory in the scopolamine-induced
ALT and AST were comparable to those from the control group cognitive impairment in the Morris water maze test.
and model group (p > 0.05) at the three-time points, which sug- Additionally, compound 10g exhibited preliminary safety in in
gested that compounds 10d and 10g had preliminary safety. vivo hepatotoxicity assays, without improving the level of ALT
Additionally, immunohistochemical staining was performed on and AST. Overall, our findings provide a valuable lead com-
the liver tissues for the histopathological study. The morpho- pound for further development of novel multifunctional agents
logic results of our tested compounds 10d and 10g were in for Alzheimer’s disease.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
505
pF i >g u 0r .e0 5 4) .. The ALT and AST levels. (A) The ALT levels of five subgroups. (B) The AST levels of five subgroups. Data are presented as the mean ± SEM (n ¼ 6; ns
Figure 5. Histopathological study of livers of male mice after treatment with the solvent only (control, A), or 36 h after administration of Compound 10d (B),
Compound 10g (C), and tacrine (D). HE staining, original magnificationꢀ200.
Disclosure statement
References
The authors report no conflicts of interest.
1.
Prince M, Ali GC, Guerchet M, et al. Recent global
trends in the prevalence and incidence of dementia,
and survival with dementia. Alzheimers Res Ther 2016;
Funding
8:23.
This work was supported by the National Natural Science
Foundation of China, under Grants [81402851, 81573281, and
2. Alzheimer’s Disease International. World Alzheimer Report
2015: The global impact of dementia: an analysis of preva-
lence, incidence, costs and trends. London: Alzheimer’s
Disease International (ADI); 2015. Available at: https://www.
alz.co.uk/research/world-report-2015
3. Kumar A, Singh A. Ekavali A review on Alzheimer’s disease
pathophysiology and its management: an update. Pharmacol
Rep 2015;67:195–203.
8
1603529]; Natural Science Foundation of Jiangsu Province,
under Grant [BK20140957]; Fundamental Research Funds for
the Central Universities, under Grant 2015ZD009, and Jiangsu
Qing Lan Project, Top-notch Academic Programs Project of
Jiangsu Higher Education Institutions, under Grant [TAPP-
PPZY2015A070].

5
06
J. ZHU ET AL.
4
.
.
Berk C, Paul G, Sabbagh M. Investigational drugs in 19. Greig NH, Lahiri DK, Sambamurti K. Butyrylcholinesterase: an
Alzheimer’s disease: current progress. Expert Opin Investig
Drugs 2014;23:837–46.
important new target in Alzheimer’s disease therapy. Int
Psychogeriatr 2002; 14:77–91.
5
Ozgeris B, Goksu S, Polat Kose L, et al. Acetylcholine- 20. Greig NH, Utsuki T, Yu Q, et al. A new therapeutic target in
sterase and carbonic anhydrase inhibitory properties of
novel urea and sulfamide derivatives incorporating dopa-
Alzheimer’s disease treatment: attention to butyrylcholines-
terase. Curr Med Res Opin 2001;17:159–65.
minergic 2-aminotetralin scaffolds. Bioorg Med Chem 21. Ballard CG. Advances in the treatment of Alzheimer’s dis-
016;24:2318–29.
2
ease: benefits of dual cholinesterase inhibition. Eur Neurol
002;47:64–70.
2. Lane RM, Potkin SG, Enz A. Targeting acetylcholinesterase
and butyrylcholinesterase in dementia. Int J Neuropsycho-
pharmacol 2006;9:101–24.
3. Eskander MF, Nagykery NG, Leung EY, et al. Rivastigmine is
a potent inhibitor of acetyl- and butyrylcholinesterase in
Alzheimer’s plaques and tangles. Brain Res 2005;1060:
6
.
.
Akincioglu A, Akincioglu H, Gulcin I, et al. Discovery of
potent carbonic anhydrase and acetylcholine esterase inhibi-
tors: novel sulfamoylcarbamates and sulfamides derived
from acetophenones. Bioorg Med Chem 2015;23:3592–602.
Gocer H, Akincioglu A, Goksu S, et al. Carbonic anhydrase
and acetylcholinesterase inhibitory effects of carbamates
2
2
2
7
and sulfamoylcarbamates.
015;30:316–20.
J Enzyme Inhib Med Chem
2
144–52.
8
9
.
.
Bajda M, Guzior N, Ignasik M, Malawska B. Multi-target-
directed ligands in Alzheimer’s disease treatment. Curr Med
Chem 2011;18:4949–75.
Bolognesi ML, Rosini M, Andrisano V, et al. MTDL design
strategy in the context of Alzheimer’s disease: from lipocrine
to memoquin and beyond. Curr Pharm Design 2009;15:
2
4. Chen Y, Lin H, Zhu J, et al. Design, synthesis, in vitro and in
vivo evaluation of tacrine-cinnamic acid hybrids as multi-tar-
get acetyl- and butyrylcholinesterase inhibitors against
Alzheimer’s disease. RSC Adv 2017;7:33851–67.
5. Morris R. Developments of a water maze procedure for
studying spatial learning in the rat. J Neurosci Methods
2
601–13.
1984;11:47–60.
1
1
0. Zemek F, Drtinova L, Nepovimova E, et al. Outcomes of
Alzheimer’s disease therapy with acetylcholinesterase inhib-
itors and memantine. Expert Opin Drug Saf 2014;13:
2
6. Chen Y, Sun J, Huang Z, et al. Design, synthesis and evalu-
ation of tacrine-flurbiprofen-nitrate trihybrids as novel anti-
Alzheimer’s disease agents. Bioorg Med Chem 2013;21:
759–74.
2462–70.
1. Wieckowska A, Kolaczkowski M, Bucki A, et al. Novel multi-
target-directed ligands for Alzheimer’s disease: Combining
cholinesterase inhibitors and 5-HT6 receptor antagonists.
Design, synthesis and biological evaluation. Eur J Med Chem
2
2
7. Cheung J, Rudolph MJ, Burshteyn F, et al. Structures of
human acetylcholinesterase in complex with pharmacologic-
ally important ligands. J Med Chem 2012;55:10282–6.
8. Brus B, Kosak U, Turk S, et al. Discovery, biological evalu-
ation, and crystal structure of a novel nanomolar selective
2016;124:63–81.
1
1
2. Ferreira-Vieira TH, Guimaraes IM, Silva FR, Ribeiro FM.
Alzheimer’s disease: targeting the cholinergic System. Curr
Neuropharmacol 2016;14:101–15.
3. Xie SS, Wang XB, Li JY, et al. Design, synthesis and evalu-
ation of novel tacrine-coumarin hybrids as multifunctional
cholinesterase inhibitors against Alzheimer’s disease. Eur J
Med Chem 2013; 64:540–53.
butyrylcholinesterase inhibitor.
167–79.
J
Med Chem 2014;57:
8
2
3
3
9. DeLano WL. The PyMOL molecular graphics system. San
Carlos (CA): DeLano Scientific; 2002. Available at: http://
www.pymol.org
0. Ellman GL, Courtney KD, Andres Jr V, Feather-Stone RM. A
new and rapid colorimetric determination of acetylcholin-
esterase activity. Biochem Pharmacol 1961;7:88–95.
1. Kam TI, Gwon Y, Jung YK. Amyloid beta receptors respon-
sible for neurotoxicity and cellular defects in Alzheimer’s dis-
ease. Cell Mol Life Sci 2014;71:4803–13.
1
4. Hosoi M, Hori K, Konishi K, et al. Plasma cholinesterase
activity in Alzheimer’s disease. Neurodegener Dis 2015;15:
188–90.
1
5. Garcia-Ayllon MS, Riba-Llena I, Serra-Basante C, et al. Altered
levels of acetylcholinesterase in Alzheimer plasma. PloS One
3
3
2. Sohma Y. Medicinal chemistry focusing on aggregation of
amyloid-b. Chem Pharm Bull 2016;64:1–7.
3. Inestrosa NC, Alvarez A, Perez CA, et al. Acetylcholinesterase
accelerates assembly of amyloid-beta-peptides into
Alzheimer’s fibrils: possible role of the peripheral site of the
enzyme. Neuron 1996;16:881–91.
2
010;5:e8701.
1
6. Giacobini E. Cholinesterases: new roles in brain function and
in Alzheimer’s disease. Neurochem Res 2003;28:515–22.
7. Ballard CG, Greig NH, Guillozet-Bongaarts AL, et al.
Cholinesterases: roles in the brain during health and disease.
Curr Alzheimer Res 2005;2:307–18.
1
1
8. Holzgrabe U, Kapkova P, Alptuzun V, et al. Targeting acetyl- 34. LeVine H. Thioflavine
T
interaction with synthetic
cholinesterase to treat neurodegeneration. Expert Opin Ther
Alzheimer’s disease beta-amyloid peptides: detection of
Tar 2007;11:161–79.
amyloid aggregation in solution. Protein Sci 1993;2:404–10.