Dual targeting of cholinesterase and amyloid beta with pyridinium/isoquinolium derivatives

Article abstract of DOI:10.1002/ddr.21631

With the surge in the cases of Alzheimer's disease (AD) over the years, several targets have been explored to curb the disease. Cholinesterases, namely acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE), remain to be the available targets that are amendable to currently approved treatments. In this study, a series of novel compounds based on tramiprosate, a highly specific amyloid beta (Aβ) inhibitor, was designed to inhibit AChE, BuChE, and Aβ aggregation. In particular, the addition of a pyridinium/isoquinolinium ring to the tramiprosate moiety (to give compounds 3a–j) led to an increase in the binding affinity for the catalytic active site of cholinesterase, which was hampered by the presence of sulfonic acid. Exclusion of the sulfonic acid moiety led to a novel but effective class of cholinesterase inhibitors (9a–w). in vitro Aβ aggregation inhibition assay indicated that compounds 3a–j, 9e–f, 9i–l, 9q, 9r, 9u–w, and 12 could inhibit over 10% Aβ aggregation at 1 mM concentration. Cholinesterase inhibition assay suggested that compounds 9g, 9h, 9o, and 9q–t exhibit over 70% inhibition on both AChE and BuChE at a concentration of 100 μM. Amongst the designed molecules, compound 9r (ca 18% at 1 mM) showed comparable inhibitory effect on the inhibition of Aβ aggregation with tramiprosate (ca 20% at 1 mM), along with impressive cholinesterase inhibitory potential (AChE IC₅₀ = 13 μM and BuChE IC₅₀ = 12 μM), acceptable toxicity and ability to pass through blood brain barrier, which could be used to ameliorate the phenotypes of AD in preclinical models.

Full text of DOI:10.1002/ddr.21631

Received: 19 June 2019
DOI: 10.1002/ddr.21631
Revised: 8 September 2019
Accepted: 3 December 2019
R E S E A R C H A R T I C L E
Dual targeting of cholinesterase and amyloid beta
with pyridinium/isoquinolium derivatives
Harapriya Chakravarty1 | Yaojun Ju1 | Wen-Hua Chen2,3 | Kin Y. Tam1
1Faculty of Health Sciences, University of
Abstract
Macau, Macau, China
2School of Biotechnology and Health Sciences,
Wuyi University, Jiangmen, China
With the surge in the cases of Alzheimer's disease (AD) over the years, several
targets have been explored to curb the disease. Cholinesterases, namely acetylcholin-
esterase (AChE) and butyrylcholinesterase (BuChE), remain to be the available targets
that are amendable to currently approved treatments. In this study, a series of novel
compounds based on tramiprosate, a highly specific amyloid beta (Aβ) inhibitor, was
designed to inhibit AChE, BuChE, and Aβ aggregation. In particular, the addition of a
pyridinium/isoquinolinium ring to the tramiprosate moiety (to give compounds 3a–j)
led to an increase in the binding affinity for the catalytic active site of cholinesterase,
which was hampered by the presence of sulfonic acid. Exclusion of the sulfonic acid
moiety led to a novel but effective class of cholinesterase inhibitors (9a–w). in vitro
Aβ aggregation inhibition assay indicated that compounds 3a–j, 9e–f, 9i–l, 9q, 9r,
9u–w, and 12 could inhibit over 10% Aβ aggregation at 1 mM concentration. Cholin-
esterase inhibition assay suggested that compounds 9g, 9h, 9o, and 9q–t exhibit over
70% inhibition on both AChE and BuChE at a concentration of 100 μM. Amongst the
designed molecules, compound 9r (ca 18% at 1 mM) showed comparable inhibitory
effect on the inhibition of Aβ aggregation with tramiprosate (ca 20% at 1 mM), along
with impressive cholinesterase inhibitory potential (AChE IC50 = 13 μM and BuChE
IC50 = 12 μM), acceptable toxicity and ability to pass through blood brain barrier,
which could be used to ameliorate the phenotypes of AD in preclinical models.
3International Healthcare Innovation Institute
(Jiangmen), Jiangmen, China
Correspondence
Faculty of Health Sciences, University of
Macau SAR, Avenida de Universidade, Taipa,
Macau, China.
Email: kintam@um.edu.mo
K E Y W O R D S
acetylcholinesterase, Alzheimer's disease, amyloid beta, butyrylcholinesterase, dual target
inhibitor, isoquinolinium, pyridinium
1
|
INTRODUCTION
lifestyle, the innate human desire to live longer has been achieved,
however, with the price of a steep increase in the number of AD
Even after more than hundred years since the term Alzheimer's
disease (AD) was coined (Zilka & Novak, 2006), its lethal effects grasp
more than 30 million debilitating mostly the elderly population
(Wilkinson, Charoenkitkarn, O'Neill, Kanik, & Chignell, 2017). With
the advent of contemporary medical facilities and well-equipped
cases. The complexity in the AD prognosis arises due to the inade-
quate understanding of its etiology and numerous risk factors (Leon,
Garcia, & Marco-Contelles, 2013). These cholinergic dysfunctions are
often observed in the aged and demented central nervous system
(Bartus, Dean 3rd, Beer, & Lippa, 1982). Accordingly, loss of presynap-
tic cholinergic function should be corrected with the increase in ace-
tylcholine levels in cholinergic synapses. However, acetylcholine itself
cannot be administered since it is rapidly hydrolyzed by cholinester-
ases. The hydrolysis can be catalyzed either by acetylcholinesterase
Abbreviations: AChE, acetylcholinesterase; AD, Alzheimer's disease; Aβ, amyloid beta; BBB,
blood brain barrier; BuChE, butyrylcholinesterase.
Harapriya Chakravarty and Yaojun Ju contributed equally to this study.
Drug Dev Res. 2019;1–14.
wileyonlinelibrary.com/journal/ddr
© 2019 Wiley Periodicals, Inc.
1

2
CHAKRAVARTY ET AL.
(AChE), a serine hydrolase enzyme, which metabolizes acetylcho-
line within a submillisecond time frame (English & Webster, 2012)
or by a related, nonspecific, enzyme butyrylcholinesterase (BuChE)
(Soreq & Seidman, 2001). Cholinesterase inhibitors can impede the
metabolism of acetylcholine and prolong its action at the synapses
(Nordberg & Svensson, 1998).
2011). Tramiprosate, is a very simple molecule which selectively
binds to Aβ (Nie et al., 2011), and is highly efficacious as demon-
strated in preclinical trials. It is a homolog of the amino acid taurine,
3-aminopropanesulfonic acid (Gervais et al., 2007) a mimic of gly-
cosaminoglycans (Figure 1). Owing to its high efficacy several
derivatives based on this scaffold have been synthesized
(Abushakra et al., 2017; Kong, Szarek, & Gervais, 2007) (Figure 1).
The modifications include: (a) optimization of the alkyl chain length
Understanding the binding site of cholinesterases facilitates the
structure-based design of inhibitors. Cholinesterases are composed of
a 20 Å deep “aromatic gorge” within which lies the catalytic active
site, the choline binding site, and the peripheral anionic site. (Harel
et al., 1993) The groove is essentially composed of aromatic amino
acids with only a few acidic residues. The catalytic active site or the
esteric site is present just 4 Å above the base, and is composed of a
catalytic triad: serine, histidine, and an acidic residue (Gupta, 2015).
The choline binding site is mainly composed of hydrophobic trypto-
phan residues (Soreq & Seidman, 2001) and the peripheral anionic site
is identified at the entry of the aromatic gorge. The electrostatic map-
ping suggests that charged substrates pave towards the gorge into
the active site (Gupta, 2015).
(Kisilevsky, Szarek,
& Weaver, 1998; Kong et al., 2007),
(b) introduction of side chains and thereby chirality to the aliphatic
carbons, (c) derivatization of the sulfonic acid and/or amine
(Bamborough et al., 2007), and (d) altering the amino functionality.
Interestingly, an amide derivative, valine-conjugated prodrug of
tramiprosate (ALZ-801), a small molecule prodrug of tramiprosate,
currently in Phase III clinical trials, shows remarkable potential as
an Aβ inhibitor (Hey et al., 2018).
Veritably, targeting AD with multi-target molecules is on a surge.
This includes: (a) GSK-3β and tau aggregation inhibitors (Gandini et al.,
2018), (b) AChE and monoamine oxidase B inhibitors, (Xu et al., 2019)
(c) beta-secretase 1 (BACE-1) and glycogen synthase kinase 3 (GSK-
3β), (Prati et al., 2015) (d) the hybridization of AChE inhibitor and sero-
tonergic subtype 4 receptor (5-HT4R) activator (Rochais et al., 2015),
(e) the introduction of antioxidant selenium into the Aβ-aggregation
inhibitor (Wang et al., 2014), and (f) the modifications of
multifunctional anti-AD curcumin scaffold (Di Martino et al., 2016).
Despite these prior efforts, Aβ and cholinesterases remain to be of
paramount significance (Mohamed & Rao, 2017; Panek, Wichur,
In AD patients cholinesterases have been found to be associated
with Aβ plaques, the major hallmark of the disease (Darvesh, 2016).
The formation of Aβ isomers (40/42) (Lue et al., 1999) is a result of
amyloidogenic processing of amyloid precursor protein (APP), which
is accumulated in the sub cellular chambers, triggering inflammation
and hindering neuronal transmission. These plaques are formed by a
series of Aβ oligomers aggregating together into clumps. Conse-
quently, restricting the oligomer formation is a promising approach
for ameliorating the phenotypes of AD (Mohamed, Shakeri, &
Rao, 2016).
ꢀ
Godyn, Pasieka, & Malawska, 2017; Tonelli et al., 2015). Presently,
the only pharmacological therapy for AD apart from N-methyl D-
aspartate receptor antagonist memantine, are the cholinesterase
inhibitors, that is, donepezil, rivastigmine, and galantamine (Figure 2).
Herein, we applied structure based molecular computer-aided ligand-
based drug design to develop inhibitors against Aβ aggregation and
Several molecular classes target different regions of the Aβ
monomer (Nie, Du, & Geng, 2011). Sycllo inositol binds at the C-
terminus, Aβ antibody at the N-terminus, and tramiprosate at the
histidine13–histidine–glutamine–lysine16(HHQK) region (Nie et al.,
FIGURE 1 Chemical structures of a known Aβ inhibitor, tramiprosate (E), and some of its derivatives (F–J) synthesized previously

CHAKRAVARTY ET AL.
3
swissdock.ch/docking).The minimum energy conformer hence
obtained was chosen for analyzing the binding mode of the molecules.
Further, the Aβ ligand complex was equilibrated using standard
dynamics cascade protocol with the default parameters of DS2017.
2.2
|
Chemistry
2.2.1
|
Synthesis of compounds 3a–j
Sulfonated pyridinium derivatives 3a–j were synthesized by
stirring the corresponding pyridine derivatives (0.11 mol) and
1,3-propanesultone 2 (0.11 mol) in toluene (30 ml) at 40–60ꢀC for
6–24 hr under nitrogen atmosphere (Rafiee
& Eavani, 2013)
(Scheme 1). The progress of the reaction was monitored by TLC,
which was also indicated by the formation of the precipitate.
The precipitate was filtered and washed three times by dic-
hloromethane (DCM, 5 ml), to remove any trace of amine salt
formed during the course of reaction. This was further washed
with acetone (10 ml) and diethyl ether (20 ml). The obtained solids
were dried under vacuum to obtain compounds 3a-j.
FIGURE 2 Approved drugs for Alzheimer's disease.
Cholinesterase inhibitors: Donepezil, Rivastigmine, and Galantamine;
N-methyl D-aspartate receptor antagonist: memantine
cholinesterases (AChE and BuChE). We reported the synthesis and
biological evaluation of some pyridinium/isoquinolinium derivatives as
dual-target inhibitors.
2.2.2
|
Synthesis of intermediates 6a–d
N-phenylnicotinamides 6a–b were prepared by dropwise addition
of respective aniline 5a–b (5.61 mmol) to nicotinyl chloride hydro-
chloride 4 (22.46.mmol) with rigorous stirring at 0ꢀC under nitrogen
atmosphere for 30 min. The reaction mixture was brought to room
temperature and stirred for another hour. Then it was quenched by
adding saturated sodium bicarbonate solution (Al-Sha'er & Taha,
2011). The white solid formed was filtered, dissolved in dic-
hloromethane, and dried using sodium sulfate to obtain the product.
The product was used without any further purification.
2
|
MATERIALS AND METHODS
| Molecular docking
2.1
Molecular docking and synthesis were performed iteratively to create
the suitable modifications in the designed molecules. Molecular dock-
ing simulations were performed using CDOCKER module of Discov-
ery Studio 2017 (DS 2017; Accelrys Inc., San Diego, CA). CDOCKER
uses CHARMM based molecular dynamics to find the optimized con-
formation for the ligand in the protein. The crystal structure of the
human AChE in Complex with Donepezil (code ID: 4EY7), human
BuChE in complex with tacrine (code ID: 4BDS), and the nuclear mag-
netic resonance (NMR) ensemble of Aβ40 (code ID: 2LMN) were
obtained from the Protein Data Bank. The inhibitor molecule and
water molecules were deleted since no information was available
regarding the conserved water molecules. Molecular docking studies
were performed using a dimer structure of Aβ, rest of the peptide
chains, from the hexapeptide, were deleted. The 3D structure of the
ligand molecules was built and geometry optimization was performed
by molecular mechanics. The protein structures were prepared; pro-
tonation states corrected, the missing side chains were added and
minimized, using DS 2017. The ligand molecules were docked into the
active site of AChE (x = 135.01, y = 114.18, z = 39.81, and radius 8.1)
and BuChE (x = 133.20, y = 116.24, z = 40.89, and radius 5.90). In
absence of an understanding regarding the binding site of the Aβ we
performed blind docking twice using DS and Swiss dock (http://www.
1-aminoanthracene-9,10-dione 5c (4.5 mmol) was dissolved in
50 ml of toluene, nicotinyl chloride hydrochloride 4 (13.3 mmol) was
added, and refluxed for 24 hr. (Barasch et al., 1999) With the comple-
tion of the reaction, solvent was removed under reduced pressure.
The product was dissolved in ethyl acetate and washed with 0.1 N
NaOH solution, followed by 1 N HCl and finally with brine. Ethyl ace-
tate was removed under reduced pressure and the product hence
obtained, (9,10-dioxo-9,10-dihydroanthracen-1-yl)nicotinamide 6c
was used without further purification.
To a stirred solution of 2-amino-5-fluorobenzoic acid 7 (19.33 mmol)
in pyridine (15 ml) was added nicotinoyl chloride hydrochloride
4 (40 mmol). The reaction was stirred at 5ꢀC for 2 hr, and thereaf-
ter moved to room temperature and poured onto ice and neutral-
ized with 2 N HCl. After partitioning, the organic layer was washed
with saturated sodium bicarbonate, saturated sodium chloride, and
water. The organic layer was dried (Na2SO4) and concentrated in
vacuum. The product, 6-fluoro-2-(pyridin-3-yl)-4H-benzo[d][1,3]
oxazin-4-one 6d, was obtained after purification by column
chromatography.

4
CHAKRAVARTY ET AL.
SCHEME 1 Reagents and conditions.
(i) Toluene, 40–60ꢀC, 6–24 hr.
(ii) 0ꢀC ! rt, 30 min. (iii) Pyridine 12 hr
2.2.3
|
Synthesis of compounds 9a–w
and concentrated under reduced pressure; this was combined with
the solid precipitate to obtain 8a-i.
Pyridine/Isoquinoline (excess ~25.549 mmol) was heated to 160ꢀC in
a sealed tube, respective monochloro acetamides 8a–i (2.450 mmol)
was added to it and stirred for another 15–30 min. TLC was
checked to monitor the completion of the reaction. The reaction
mixture was allowed to cool to room temperature then triturated
with cold acetone. The solid precipitated was washed twice with
diethyl ether (10 ml) and DCM (5 ml) to yield the final products
9a–w.
2.2.5
|
Synthesis of compound 12
Pyridine (excess ~12.580 mmol) was heated to 160ꢀC in a sealed tube,
4-chloronaphthalen-1-yl carbonochloridate 11 (0.635 mmol) was added
to it and stirred for another 30 min. Thin-layer chromatography (TLC) was
checked to monitor the completion of the reaction. The reaction mixture
was allowed to cool to room temperature then triturated with cold ace-
tone. The solid precipitated was washed twice with diethyl ether (5 ml)
and DCM (1 ml) to yield the final product 12.
2.2.4
|
Synthesis of intermediates 8a–i
Mono chloroacetamides 8a–i was prepared from the respective
aromatic/aliphatic amines 5a–i. Amine (53.68 mmol) was dissolved
in DCM (150 ml) and stirred at 0ꢀC, while chloroacetyl chloride
(80.53 mmol) was added dropwise over a period of 30 min. (Qiu
et al., 2017) The reaction mixture was stirred at 0ꢀC for 2 hr and
thereafter stirred at room temperature until the completion of the
reaction. Ice water was slowly added to the mixture to quench the
reaction. Solid hence obtained was filtered. Organic layer was
washed with NaHCO3 and water three times, dried over Na2SO4
2.2.6
|
Synthesis of intermediate 11
4-chloronaphthalen-1-yl carbonochloridate 11 was prepared from
4-chloronaphthalen-1-ol 10. 10 (13.4 mmol) was dissolved in DCM
(50 ml) and stirred at 0ꢀC, while chloroacetyl chloride (20.12 mmol) was
added dropwise over a period of 30 min. (Qiu et al., 2017) The reaction
mixture was stirred at 0ꢀC for 2 hr and thereafter stirred at room tem-
perature until the completion of the reaction. Ice water was slowly

CHAKRAVARTY ET AL.
5
added to the mixture to quench the reaction. Solid hence obtained was
filtered. Organic layer was washed with NaHCO3 and water three
times, dried over Na2SO4 and concentrated under reduced pressure;
this was combined with the solid precipitate to obtain 11.
acid (CHCA, Sigma-Aldrich) in 60% acetonitrile and 0.3%
trifluoroacetic acid (TFA) solution spiked with 0.25 μM insulin as inter-
nal standard. The mixtures were then spotted onto the target polished
steel and detected by the Bruker matrix-assisted laser desorption/ion-
ization mass spectrometry, matrix-assisted laser desorption/ionizatio
(MALDI MS) system. Intensity of Aβ42 to insulin is correlated to the
content of Aβ from which the aggregation percentages were calcu-
lated by comparing the extent aggregation in the presence or absence
of inhibitors.
2.3
|
Biological evaluation
2.3.1
|
In vitro AChE and BuChE inhibition assays
The inhibition on AChE (recombinant human acetylcholinesterase
expressed in human embryonic kidney (HEK) 293 cells from Sigma-
Aldrich) and BuChE (butyrylcholinesterase from equine serum from
Sigma-Aldrich) was colorimetrically determined by modified Ellman's
method (Ellman, Courtney, Andres, & Featherstone, 1961). Briefly, the
reaction system and all reaction reagents were prepared in 0.1 M
phosphate-buffer solution containing 1.6 mM MgCl2 and 100 mM
NaCl at pH 8.0 on 96-well plates. In 200 μl AChE catalytic reaction
system, 0.1 U/ml AChE, 1 mM acetylthiocholine (ATC; Sigma-Aldrich,
St. Louis, MO 63103, USA) and 1.2 mM 5,50-dithiobis(2-nitrobenzoic
acid) (DTNB; Sigma-Aldrich) were incubated with different concentra-
tions of testing compounds. In 200 μl BuChE catalytic reaction sys-
tem, 0.05 U/ml BuChE, 5 mM butyrylthiocholine (BTC; Sigma-Aldrich)
and 6 mM dithiobis (2-nitrobenzoic acid) (DTNB) were tested with dif-
ferent concentrations of compounds. The prewarmed ATC and BTC in
37ꢀC were finally added to initiate the reaction and monitored at
412 nm in 37ꢀC for 120 min with a SpectraMax microplate reader.
The absorbance-time curves were then plotted, and the K values were
calculated to represent the enzymatic activity. All the reactions were
detected in duplicate and one duplicate sample without any inhibitors
was always set to yield 100% of AChE or BuChE activities. Data from
concentration − inhibition experiments of the inhibitors were calcu-
lated by nonlinear regression analysis, using the GraphPad Prism pro-
gram package (GraphPad software; San Diego, CA) to predict the IC50
(concentration of drug producing 50% of enzyme activity inhibition).
Repeated assay was performed for statistic analysis.
2.3.3
|
Cytotoxicity test
The human neuroblastoma SH-SY5Y cell was a gift from Prof. Duncan
Leung, Institute of Chinese Medical Sciences, University of Macau.
The viability of cells incubated with inhibitors was determined using
the MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bro-
mide) assay. The methodological details were included in Supporting
Information.
2.3.4
|
Brain exposure of compound 9r in
Alzheimer’s disease (AD) mouse model
The transgenic (Tg) AD mice (B6; 129-Psen1tm1MpmTg [APPSwe,
tauP301L] 1Lfa/Mmjax) were obtained from the Jackson Laboratory
(Bar Harbor, ME) and maintained by the Animal Facility of Faculty of
Health Sciences (University of Macau). The animal study protocol
(#UMARE—032-2018) was approved by the Animal Research Ethics
Committee of the University of Macau. All the animal experiments
were performed strictly under the supervision and recommendation
of the committee. Six 8-month-old male mice were fasted but free
access to water for 24 hr before dosing. Compound 9r was dissolved
in phosphate-buffered saline (PBS) at a concentration of 1 mg/ml and
orally administered to mice at a dosage of 10 mg/kg. The animals
were then sacrificed by cervical dislocation and brain collected. Brain
and blood samples were prepared using the protocols previously
described (Wei Zhou, Hu, & Tam, 2017) and analyzed by liquid Chro-
matography with tandem mass spectrometry (LC–MS/MS) (Waters,
TQD) as previously described (W. Zhou, Wu, Li, Hu, & Tam, 2016).
Briefly, under positive mode of Waters Xevo TQD, compound 9r (m/z
343 >130 as quantifier ion and m/z 343 >142 as qualifier ion) and ber-
berine (as internal standard, m/z 336 >320 as quantifier ion and m/z
336 >278 as qualifier ion) were simultaneously monitored.
2.3.2
|
In vitro Aβ aggregation inhibition assay
The lypophilized artificial synthetic human Aβ42 peptide (>95%) was
purchased from GL Biochem (Shanghai, China). The MALDI-MS was
applied for rapid detection of the residual Aβ content in aggregation.
In brief, 50μm Aβ42 in aggregation buffer above was incubated in
37ꢀC incubator (New Brunswick S41i; Eppendorf, Hamburg, Ger-
many), shaking at a speed of 80 rpm for 2 days. Aβ fibril seeds were
then generated by sonicating the 50 μM aggregate mixture for 30 min
and stored at −20ꢀC for further usage. The aggregation of 20 μM
Aβ42 peptide was induced by addition of 1 μM Aβ42 fibril seeds in
20 mM phosphate buffer, 0.1 M NaCl at pH 7.4 with or without
1 mM testing compounds and incubated quiescently in the incubator
at 37ꢀC for 12 hr. Twenty microliters of sample was mixed thoroughly
with 50 μl matrix containing 10 mg/ml α-Cyano-4-hydroxycinnamic
3
|
RESULTS AND DISCUSSIONS
| Molecular design
3.1
Structure based ligand design was performed in conjunction with syn-
thesis and biological evaluation, to sculpt the molecule towards our
desired results (Figure 3 and Table 1). Lipinski's rules and Veber's rules

6
CHAKRAVARTY ET AL.
FIGURE 3 Schematic of molecule design from Tramiprosate to its pyridinium derivative 3a–f. Addition of the pyridinium moiety helps
stabilize the ligands into the aromatic gorge of cholinesterase. Inspired by ALZ-801, addition of an amide linker 3g–j promotes the binding to both
cholinesterase and Aβ, delimited by the unfavorable donor-donor interaction in cholinesterase and poor Blood Brain Barrier (BBB) penetration.
Exclusion of the alkyl group with the sulfonic acid 9a–w enhances the binding to both of the targets
were taken into account, such that none of our designed analogs vio-
lated the same. The structural simplicity of tramiprosate makes it
highly specific to Aβ (Nie et al., 2011). Interestingly, we found that
addition of a pyridinium group enhances the affinity for cholinester-
ases as suggested by the decrease in CDOCKER interaction energy
from −16.22 to −30.94 and −16.02 to −27.16 for AChE and BuChE,
respectively. Stabilizing it into the hydrophobic gorge, by π–π interac-
tion, the presence of sulfonic acid however, creates an unfavorable
bump between the acidic residues in the peripheral anionic site of
the cholinesterases. Addition of an amino group to the pyridinium
ring fetched an additional hydrogen bond, only to be limited by
unfavorable donor–donor interaction with the sulfonic acid. Incor-
poration of an amide linker helped to form a new H-bonding but
did not overcome the unfavorable bump. Moreover, it was impor-
tant to have a precise knowledge regarding the blood brain pene-
trating ability of the designed molecules when targeting the
cognitive processes in the brain. Complete exclusion of the sulfonic
acid with the alkyl chain lowered the ADMET_BBB level massively
(Figure S1, Table S1). Interestingly this change enhanced the bind-
ing affinity of the molecules to the cholinesterases. The catalytic
active site of the cholinesterases is characterized by the presence
of several nonpolar residues. Indeed addition of fused aromatic sys-
tem, naphthyl, or anthraquinone moiety, led to multifold enhance-
ment of binding affinity. Conversely, this alteration destabilized the
H-bonding network with the Aβ. Replacing the fused ring system
by substituting pyridinium with isoquinolinium not only favored the
binding in cholinesterases but also showed a remarkable blood
brain penetrating ability.
3.2
|
Chemistry
The synthetic strategy employed for the synthesis of sulfonate
derivatives 3a–j is depicted in Scheme 1. Compounds 3a–j were
synthesized by reacting 1,3 propane sultone with the respective
pyridine derivatives (Rafiee & Eavani, 2013). The pyridine deriva-
tives 6a–e, were prepared by reacting suitable aromatic amines
with nicotinyl chloride. The synthesis of 9a–w and 12 was per-
formed according to Scheme 2 (Al-Sha'er & Taha, 2011). Mono-
chloroacetamides 8a–i and monochloroacetate 11 were prepared
from the appropriate amines with chloroacetyl chlorides. Further,
the treatment of 8a–i with pyridine/isoquinoline yielded the
designed derivatives 9a–w. The structures of the final compounds
were elucidated on the basis of 1H NMR, 13C NMR, and Mass spec-
trometry. Purity of the selected compounds was verified using high
performance liquid chromatography (HPLC) (see Supporting
Information).
3.3
|
Biological evaluations
3.3.1
aggregation
|
In vitro inhibition of self-induced Aβ
The MALDI-MS detection was applied for the determination of the
residual Aβ content in aggregate system. Tramiprosate, which was
previously in clinical trial for Aβ aggregation inhibition, was used
as the positive control. As shown in Table 1, tramiprosate at

CHAKRAVARTY ET AL.
7

8
CHAKRAVARTY ET AL.

CHAKRAVARTY ET AL.
9

10
CHAKRAVARTY ET AL.

CHAKRAVARTY ET AL.
11
SCHEME 2 Reagents and conditions.
(i) Chloroacetyl chloride, DCM, 2 hr,
0 ꢀC ! rt. (ii) Pyridine/isoquinoline 160ꢀC
15–20 min. (iii) Chloroacetyl chloride,
DCM, 2 hr, 0ꢀC ! rt. (ii) Isoquinoline
160ꢀC, 30 min. Structures for (9a–w) in
Table 1
FIGURE 4 Binding mode of compound 9r in (a) AChE (protein data bank (PDB) code 4EY7) and (b) BuChE (PDB code 4BDS). The
isoquinolinium motif is projected towards the catalytic active site and forms strong π–π interaction Trp86 and Trp82, of AChE and BuChe,
respectively. Isoquinolinium nitrogen forms attractive charged interactions with Asp 74 of AChE and Asp70 of BuChE. The phenyl ring is
stabilized at peripheral anion site forming multiple π–π interactions with the aromatic amino acids
concentration of 1 mM inhibited the aggregation by 19.68% as
compared with the buffer control. In contrast, compounds 3a–h,
3j, 9e, 9f, 9i–l, 9q, 9r, 9u–w, and 12 at a concentration of 1 mM
exhibited 10% inhibition on Aβ aggregation suggesting that our
designed compounds are comparable with tramiprosate in
inhibiting Aβ aggregation.
3.3.3
|
Cytotoxicity of the synthesized compounds
The SH-SY5Y cell line was used to test the compounds toxicity with
MTT assay. Amongst all the molecules synthesized in this study only
9n, 9o, 9s, 9t, 12 exhibited cytoxicity with the IC50 values being
27.79, 59.40, 88.83, 116.2, and 186.25 μM, respectively. As for the
other compounds, particularly the best multifunctional compounds 9f,
9q, and 9r no cytotoxicity was observed in SH-SY5Y cells even at a
concentration of 200 μM.
3.3.2
|
In vitro inhibition of AChE and BuChE
The Ellman's method was used to evaluate the cholinesterase inhi-
bition of our compounds. Donepezil, the Food and Drug Adminis-
tration (FDA) approved ChE inhibitor for AD treatment, was used
as the positive control. At the concentration of 100 μM, donepezil
exhibited 100% inhibition on the activity of AChE and 89.62%
inhibited the activity of BuChE. As shown in Table 1, compounds
9g, 9h, 9o, 9q–t exhibit dual inhibition (>70% at 100 μM) on both
AChE and BuChE. Some of these compounds even outperformed
donepezil, in BuChE inhibition. Our results suggested that by
excluding the alkyl group with sulfonic acid, the potency of ChE
inhibition by compounds 9a–w was improved considerably.
3.3.4
|
Brain exposure of the multifunctional
candidates in AD mouse
The compound 9r exhibited the most promising inhibition potency
on both Aβ aggregation and cholinesterase. We thus tested whether
compound 9r exhibited any brain exposure in AD mice. Compound
9r was orally administered to 6 AD mice at a dose of 10 mg/kg. The
blood samples and brains were collected at time point 2 hr and
detected by triple quadrupole detector (TQD) MS/MS. Compound
9r was detected 4.89 ng/g in brain and 18.14 ng/ml in plasma in the

12
CHAKRAVARTY ET AL.
FIGURE 5 Binding of compound 9r in Aβ dimer (from PDB code ID: 2LMN). (a) Aβ dimer prior to molecular docking of 9r. (b) 9r bound to Aβ.
The monomer–monomer interaction is diminished upon binding of 9r to the dimer, as indicated by the loss of the secondary structure. 9r
interacts with KLVFFA stretch of Aβ; it forms an H-bonding with Phe20 and forms pi–pi and Vander Walls interactions with Phe19 and Val21
mice samples. This result demonstrated that the multifunctional
agent 9r could reach the brain of the AD mouse and pass through
the BBB.
and establishes Vander Walls interaction with Val20 and Leu21. The
amide group holds the molecule across the monomers by H-bonding
with Phe20.
3.3.5
|
Binding mode of 9r
4
|
CONCLUSIONS
Compound 9r was found to exhibit substantial inhibitory effect on
both the cholinesterases and Aβ. We performed computational dock-
ing to analyze and quantify the macromolecule–ligand interaction.
Analysis of the binding pose of compound 9r with cholinesterases
indicates a strong interaction in the aromatic gorge. Compound 9r
presents an analogous binding mode for both AChE and BuChE
(Figure 4). The molecule aligns itself from the peripheral anionic site
towards the catalytic active site. The isoquinoline motif is projected
towards the catalytic active site. It establishes van der Walls interac-
tions with His447, Glu202, Phe295, and Ser203, a residue from the
catalytic triad in AChE. Similarly, in BuChE, the isoquinoline moiety
forms van der Walls interactions with the residues of the catalytic
triad, Ser198, His438, and Phe329. The isoquinolinium moiety is sta-
bilized by π–π interaction with Trp86 and Trp82, of AChE and BuChE,
respectively. There are only a very few acidic residues in the aromatic
grove but it appears to play a critical role in the binding of our
designed analogs. Isoquinolinium nitrogen forms an attractive charged
interaction with Asp 74 of AChE and Asp70 of BuChE. These acidic
residues are further involved in the formation of H-bonding with the
amide nitrogen in each case. The phenyl ring is stabilized by π–π inter-
actions with multiple aromatic amino acids in peripheral anionic site.
Cl–π interaction with these aromatic residues further aids the binding.
9r binds to the KLVFFA stretch of the Aβ (Figure 5). Inhibitors
targeting the key nucleation site (KLVFFA region), which facilitates
monomer–monomer interactions forming a steric zipper model, disrupt
this interaction by destabilizing the beta sheet in the Aβ oligomer.
(Mohamed et al., 2016). Protein-ligand complex obtained after, molecu-
lar docking followed by two subsequent minimizations, indicated that
hydrogen bonding network between the monomers was suggestively
reduced. The ligand forms a π-π interaction with Phe19 of both chains
In this study, we have developed dual-target molecules to inhibit cho-
linesterase and Aβ aggregation. Our initial molecular design was based
on tramiprosate, an investigate compound for anti-amyloid therapy, to
engineer the binding affinity towards cholinesterases. It was interesting
to observe that a highly selective tramiprosate molecule upon addition
of just a pyridinium ring (3a–j) could provide binding affinity for cho-
linesterases. However, the presence of sulfonic acid creates an unfa-
vorable bump between the acidic residues in the peripheral anionic
site and sulfonic acid of the cholinesterases. Exclusion of the sulfonic
acid along with the alkyl chain changed the core completely (9a–w).
Among them isoquinolinium derivatives (9m–t) showed decent in sil-
ico BBB permeability and substantial binding affinity for cholinester-
ases. Aβ peptide aggregation inhibition test revealed that
compounds 3a–h, 3j, 9e, 9f, 9i–l, 9q, 9r, 9u–w, and 12 give over
10% inhibitions, which is comparable to tramiprosate at the same
concentration. From the cholinesterase inhibition assay, we found
that that 9f–h, 9n, 9o, 9q–t exhibited dual inhibition (>70% at
100 μM) on both AChE and BuChE. Most of the compounds were
nontoxic to SH-SY5Y cells even at a concentration of 200 μM. Com-
pounds 9f, 9q, and 9r exhibited low toxicity with good dual inhibitory
effect. Compound 9r was the best of the class with >85% inhibition, at
100 μM, for both cholinesterase and 18% inhibition for Aβ aggregation.
Moreover, compound 9r showed good brain exposure in AD mice
(4.89 ng/g) suggesting that this compound could cross blood brain bar-
rier and elicit potential therapeutic effects in vivo.
ACKNOWLEDGMENTS
This work was financially supported by the University of Macau (grant
no. MYRG2016-00102-FHS). We thank Prof. Duncan Leung (ICMS,
University of Macau) for samples of SH-SY5Y cells.

CHAKRAVARTY ET AL.
13
Alzheimer's disease. Clinical Pharmacokinetics, 57(3), 315–333. https://
doi.org/10.1007/s40262-017-0608-3
CONFLICT OF INTEREST
The authors declare no competing interests.
Kisilevsky, R., Szarek, W., & Weaver, D. (1998). Method for treating amy-
loidosis. In Google Patents. US patent application US20070265334A1.
https://patents.google.com/patent/US20070265334
ORCID
Kong, X., Szarek, W. A., & Gervais, F. (2007). Methods and compositions
for treating amyloid-related diseases. In Google Patents. PCT
WO2006085149A3. https://patents.google.com/patent/WO2006
085149A3
Kin Y. Tam
https://orcid.org/0000-0001-5507-8524
REFERENCES
Leon, R., Garcia, A. G., & Marco-Contelles, J. (2013). Recent advances in
the multitarget-directed ligands approach for the treatment of
Alzheimer's disease. Medicinal Research Reviews, 33(1), 139–189.
https://doi.org/10.1002/med.20248
Abushakra, S., Porsteinsson, A., Scheltens, P., Sadowsky, C., Vellas, B.,
Cummings, J., … Tolar, M. (2017). Clinical effects of tramiprosate in
APOE4/4 homozygous patients with mild Alzheimer's disease suggest
disease modification potential. The Journal of Prevention of Alzheimer's
Disease, 4(3), 149–156. https://doi.org/10.14283/jpad.2017.26
Al-Sha'er, M. A., & Taha, M. O. (2011). Rational exploration of new
pyridinium-based HSP90α inhibitors tailored to thiamine structure.
Medicinal Chemistry Research, 21(4), 487–510. https://doi.org/10.
1007/s00044-011-9557-9
Lue, L.-F., Kuo, Y.-M., Roher, A. E., Brachova, L., Shen, Y., Sue, L., … Rogers, J.
(1999). Soluble amyloid β peptide concentration as a predictor of synaptic
change in Alzheimer's disease. The American Journal of Pathology, 155(3),
853–862. https://doi.org/10.1016/s0002-9440(10)65184-x
Mohamed, T., & Rao, P. P. (2017). 2, 4-Disubstituted quinazolines as
amyloid-β aggregation inhibitors with dual cholinesterase inhibition and
antioxidant properties: Development and structure-activity relationship
(SAR) studies. European Journal of Medicinal Chemistry, 126, 823–843.
Mohamed, T., Shakeri, A., & Rao, P. P. (2016). Amyloid cascade in
Alzheimer's disease: Recent advances in medicinal chemistry. European
Journal of Medicinal Chemistry, 113, 258–272.
Nie, Q., Du, X.-g., & Geng, M.-y. (2011). Small molecule inhibitors of amy-
loid β peptide aggregation as a potential therapeutic strategy for
Alzheimer's disease. Acta Pharmacologica Sinica, 32(5), 545–551.
Nordberg, A., & Svensson, A.-L. (1998). Cholinesterase inhibitors in the
treatment of Alzheimer's disease. Drug Safety, 19(6), 465–480.
Bamborough, P., Barker, M. D., Campos, S. A., Cousins, R. P. C., Faulder, P.,
Hobbs, H.,
… Payne, J. J. (2007). Pyrrolo [2, 3-b] pyridin-4-yl-
benzenesulfonamide compounds as 1kk2 inhibitors. In Google Patents.
US patent application US20100035917A1. https://patents.google.
com/patent/US20100035917
Barasch, D., Zipori, O., Ringel, I., Ginsburg, I., Samuni, A., & Katzhendler, J.
(1999). Novel anthraquinone derivatives with redox-active functional
groups capable of producing free radicals by metabolism: Are free radi-
cals essential for cytotoxicity? European Journal of Medicinal Chemistry,
34(7–8), 597–615.
Bartus, R. T., Dean, R. L., 3rd, Beer, B., & Lippa, A. S. (1982). The choliner-
gic hypothesis of geriatric memory dysfunction. Science, 217(4558),
408–414.
Darvesh, S. (2016). Butyrylcholinesterase as a diagnostic and therapeutic
target for Alzheimer's disease. Current Alzheimer Research, 13(10),
1173–1177.
Di Martino, R. M. C., De Simone, A., Andrisano, V., Bisignano, P., Bisi, A.,
Gobbi, S., … Belluti, F. (2016). Versatility of the curcumin scaffold: Dis-
covery of potent and balanced dual BACE-1 and GSK-3 beta inhibi-
tors. Journal of Medicinal Chemistry, 59(2), 531–544. https://doi.org/
10.1021/acs.jmedchem.5b00894
ꢀ
Panek, D., Wichur, T., Godyn, J., Pasieka, A., & Malawska, B. (2017).
Advances toward multifunctional cholinesterase and β-amyloid aggre-
gation inhibitors. Future Medicinal Chemistry, 9(15), 1835–1854.
Prati, F., De Simone, A., Bisignano, P., Armirotti, A., Summa, M.,
Pizzirani, D., … Perez-Castillo, A. (2015). Multitarget drug discovery for
Alzheimer's disease: Triazinones as BACE-1 and GSK-3β inhibitors.
Angewandte Chemie International Edition, 54(5), 1578–1582.
Qiu, Q., Liu, B., Cui, J., Li, Z., Deng, X., Qiang, H., … Qian, H. (2017). Design,
synthesis, and pharmacological characterization of N-(4-(2 (6,-
7-dimethoxy-3,4-dihydroisoquinolin-2[1H]yl)ethyl)phenyl)quinazolin-
4-amine derivatives: Novel inhibitors reversing P-glycoprotein-
mediated multidrug resistance. Journal of Medicinal Chemistry, 60(8),
3289–3302. https://doi.org/10.1021/acs.jmedchem.6b01787
Rafiee, E., & Eavani, S. (2013). Organic–inorganic polyoxometalate based
salts as thermoregulated phase-separable catalysts for selective oxida-
tion of thioethers and thiophenes and deep desulfurization of model
fuels. Journal of Molecular Catalysis A: Chemical, 380, 18–27. https://
doi.org/10.1016/j.molcata.2013.09.009
Ellman, G. L., Courtney, K. D., Andres, V., & Featherstone, R. M. (1961). A
new and rapid colorimetric determination of acetylcholinesterase
activity. Biochemical Pharmacology, 7(2), 88–95. https://doi.org/10.
1016/0006-2952(61)90145-9
English, B. A., & Webster, A. A. (2012). Chapter 132—Acetylcholinesterase
and its inhibitors A2—Robertson, David. In I. Biaggioni, G. Burnstock,
P. A. Low, & J. F. R. Paton (Eds.), Primer on the autonomic nervous sys-
tem (3rd ed., pp. 631–633). San Diego, CA: Academic Press.
Gandini, A., Bartolini, M., Tedesco, D., Martinez-Gonzalez, L., Roca, C.,
Campillo, N. E., … Bolognesi, M. L. (2018). Tau-centric multitarget
approach for Alzheimer's disease: Development of first-in-class dual
glycogen synthase kinase 3beta and tau-aggregation inhibitors. Journal
of Medicinal Chemistry, 61(17), 7640–7656. https://doi.org/10.1021/
acs.jmedchem.8b00610
Rochais, C., Lecoutey, C., Gaven, F., Giannoni, P., Hamidouche, K., Hedou, D.,
… Dallemagne, P. (2015). Novel multitarget-directed ligands (MTDLs) with
acetylcholinesterase (AChE) inhibitory and serotonergic subtype 4 recep-
tor (5-HT4R) agonist activities as potential agents against Alzheimer's dis-
ease: The design of donecopride. Journal of Medicinal Chemistry, 58(7),
3172–3187. https://doi.org/10.1021/acs.jmedchem.5b00115
Soreq, H., & Seidman, S. (2001). Acetylcholinesterase—New roles for an
old actor. Nature Reviews Neuroscience, 2(4), 294–302.
Gervais, F., Paquette, J., Morissette, C., Krzywkowski, P., Yu, M., Azzi, M., …
Laurin, J. (2007). Targeting soluble Aβ peptide with Tramiprosate for the
treatment of brain amyloidosis. Neurobiology of Aging, 28(4), 537–547.
Gupta, R. C. (2015). Handbook of toxicology of chemical warfare agents.
Amsteram: Academic Press.
Tonelli, M., Catto, M., Tasso, B., Novelli, F., Canu, C., Iusco, G., …
Sparatore, A. (2015). Multitarget therapeutic leads for Alzheimer's dis-
ease: Quinolizidinyl derivatives of bi-and tricyclic systems as dual
inhibitors of cholinesterases and β-amyloid (Aβ) aggregation. Chem-
MedChem, 10(6), 1040–1053.
Wang, Z. R., Wang, Y. L., Li, W. R., Mao, F., Sun, Y., Huang, L., & Li, X. S.
(2014). Design, synthesis, and evaluation of multitarget-directed
selenium-containing clioquinol derivatives for the treatment of Alzheimer's
disease. ACS Chemical Neuroscience, 5(10), 952–962. https://doi.org/10.
1021/cn500119g
Harel, M., Schalk, I., Ehret-Sabatier, L., Bouet, F., Goeldner, M., Hirth, C., …
Sussman, J. (1993). Quaternary ligand binding to aromatic residues in
the active-site gorge of acetylcholinesterase. Proceedings of the
National Academy of Sciences, 90(19), 9031–9035.
Hey, J. A., Yu, J. Y., Versavel, M., Abushakra, S., Kocis, P., Power, A., …
Tolar, M. (2018). Clinical pharmacokinetics and safety of ALZ-801, a
novel Prodrug of Tramiprosate in development for the treatment of

14
CHAKRAVARTY ET AL.
Wilkinson, A., Charoenkitkarn, V., O'Neill, J., Kanik, M., & Chignell, M.
(2017). Journeys to engagement: Ambient activity technologies for people
living with dementia. Paper presented at the Proceedings of the 26th
International Conference on World Wide Web Companion.
Xu, Y., Zhang, J., Wang, H., Mao, F., Bao, K., Liu, W., … Li, J. (2019). Ratio-
nal design of novel selective dual-target inhibitors of acetylcholines-
terase and monoamine oxidase B as potential anti-Alzheimer's disease
agents. ACS Chemical Neuroscience, 10(1), 482–496. https://doi.org/
10.1021/acschemneuro.8b00357
Zilka, N., & Novak, M. (2006). The tangled story of Alois Alzheimer.
Bratislavske Lekarske Listy, 107(9/10), 343.
SUPPORTING INFORMATION
Additional supporting information may be found online in the
Supporting Information section at the end of this article.
Zhou, W., Hu, X., & Tam, K. Y. (2017). Systemic clearance and brain distri-
bution of carbazole-based cyanine compounds as Alzheimer's disease
drug candidates. Scientific Reports, 7(1), 16368. https://doi.org/10.
1038/s41598-017-16635-4
How to cite this article: Chakravarty H, Ju Y, Chen W-H,
Tam KY. Dual targeting of cholinesterase and amyloid beta
with pyridinium/isoquinolium derivatives. Drug Dev Res. 2019;
1–14. https://doi.org/10.1002/ddr.21631
Zhou, W., Wu, X., Li, J., Hu, X., & Tam, K. Y. (2016). Quantification of per-
manent positively charged compounds in plasma using one-step dilu-
tion to reduce matrix effect in MS. Bioanalysis, 8(6), 497–509. https://
doi.org/10.4155/bio-2015-0014