Isoindolines/isoindoline-1,3-diones as AChE inhibitors against Alzheimer’s disease, evaluated by an improved ultra-micro assay

Article abstract of DOI:10.1007/s00044-018-2226-5

Alzheimer’s disease (AD), the most common form of dementia, is characterized by a progressive degeneration of the brain that leads to loss of memory and deterioration of others cognitive functions. The only drugs currently approved for treating AD are AChE inhibitors (AChEIs). We previously tested a novel isoindoline-1,3-dione, finding potent inhibition of AChE, in part because the two carbonyl groups of phthalimide facilitate hydrogen bonds with the enzyme. The aims of the present study were: (A) To achieve a faster and cheaper technique with a reduced quantity of reactive, without significant difference in the validation of the results, by modifying the version of the method described by Bonting and Featherstone. (B) To test new isoindolines and dioxoisoindolines as AChEIs and see if the carbonyl group is really important for affinity. Both families of compounds (isoindolines and dioxoisoindolines) had an inhibitory effect. The enzymatic inhibitions produced by isoindolines were uncompetitive, whereas that evoked by dioxoisoindolines were competitive. One of the isoindoline derivatives (IsoB with a Ki of 88–160μM) showed about 5-fold greater inhibition of AChE than its corresponding dioxoisoindoline. According to molecular docking performed, dioxoisoindolines apparently interact with the catalytic active site, the peripheral anionic site, and the aromatic patch, which can explain the kind of inhibition observed. Due to the uncompetitive inhibition of isoindolines, their inhibitory behavior could not be explored in silico. We afforded a faster and more efficient method, while yielding similar results than Bonting and Featherstone method. Additionally, we demonstrated that carbonyl group affects the kind of inhibition and the affinity.

Full text of DOI:10.1007/s00044-018-2226-5

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
MEDICINAL
CHEMISTRY
RESEARCH
Medicinal Chemistry Research
https://doi.org/10.1007/s00044-018-2226-5
ORIGINAL RESEARCH
Isoindolines/isoindoline-1,3-diones as AChE inhibitors against
Alzheimer’s disease, evaluated by an improved ultra-micro assay
Erik Andrade-Jorge1 Luis A. Sánchez-Labastida1 Marvin A. Soriano-Ursúa2 Juan A. Guevara-Salazar3
José G. Trujillo-Ferrara1
●
●
●
●
Received: 9 May 2018 / Accepted: 20 July 2018
© Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
Alzheimer’s disease (AD), the most common form of dementia, is characterized by a progressive degeneration of the brain
that leads to loss of memory and deterioration of others cognitive functions. The only drugs currently approved for treating
AD are AChE inhibitors (AChEIs). We previously tested a novel isoindoline-1,3-dione, finding potent inhibition of AChE,
in part because the two carbonyl groups of phthalimide facilitate hydrogen bonds with the enzyme. The aims of the present
study were: (A) To achieve a faster and cheaper technique with a reduced quantity of reactive, without significant difference
in the validation of the results, by modifying the version of the method described by Bonting and Featherstone. (B) To test
new isoindolines and dioxoisoindolines as AChEIs and see if the carbonyl group is really important for affinity. Both
families of compounds (isoindolines and dioxoisoindolines) had an inhibitory effect. The enzymatic inhibitions produced by
isoindolines were uncompetitive, whereas that evoked by dioxoisoindolines were competitive. One of the isoindoline
derivatives (IsoB with a Ki of 88–160µM) showed about 5-fold greater inhibition of AChE than its corresponding
dioxoisoindoline. According to molecular docking performed, dioxoisoindolines apparently interact with the catalytic active
site, the peripheral anionic site, and the aromatic patch, which can explain the kind of inhibition observed. Due to the
uncompetitive inhibition of isoindolines, their inhibitory behavior could not be explored in silico. We afforded a faster and
more efficient method, while yielding similar results than Bonting and Featherstone method. Additionally, we demonstrated
that carbonyl group affects the kind of inhibition and the affinity.
●
●
●
●
Keywords Alzheimer’s disease AChE 2,3-Dihydro-1H-isoindoles N-substituted phthalimide Molecular docking
Introduction
Alzheimer’s disease (AD) is the most common form of
dementia, constituting up to 80% of all cases (Duthey 2013;
Korolev 2014). It is expected to affect around 66 million
people by 2030. Today, the prevalence of AD is reported to
* Juan A. Guevara-Salazar
qfb_beto@hotmail.com
be about 3.5% of the world population (>20 million peo-
* José G. Trujillo-Ferrara
ple), with some authors estimating 6% (~35 million people).
This disease mostly afflicts the elderly, including 10% of
people over 65 and nearly 50% of those over 85. Since life
expectancy is on the rise, prevalence is expected to increase
(Musiał et al. 2007; Aliabadi et al. 2013; Anand and Singh
2013; Guzior et al. 2014). In 2010, 58% of all AD patients
had low or middle incomes and this is projected to reach
63% in 2030 (Prince et al. 2013).
AD is characterized by a progressive degeneration of the
brain that leads to loss of memory, disorientation, anxiety,
delusion, depression, insomnia, wandering, learning
impairment, and deterioration of cognitive functions.
jtrujillo@ipn.mx
1
2
3
Laboratorio de Investigación en Bioquímica, Sección de Estudios
de Posgrado e Investigación, Escuela Superior de Medicina del
Instituto Politécnico Nacional, Plan de San Luis y Díaz Mirón s/n
Casco de Santo Tomás, 11340 Mexico City, Mexico
Laboratorio de Investigación en Fisiología, Sección de Estudios de
Posgrado e Investigación, Escuela Superior de Medicina del
Instituto Politécnico Nacional, Plan de San Luis y Díaz Mirón s/n
Casco de Santo Tomás, 11340 Mexico City, Mexico
Academia de Farmacología, Escuela Superior de Medicina del
Instituto Politécnico Nacional, Plan de San Luis y Díaz Mirón s/n
Casco de Santo Tomás, 11340 Mexico City, Mexico

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Medicinal Chemistry Research
Although the etiology of AD is not well understood, various
factors contribute to the development of this disease, such
as an increase in β-amyloid plaques, an accumulation of
neurofibrillary tangles, hyperphosphorylation of tau protein,
chronic neuroinflammation, oxidative stress, inflammatory
processes, dyshomeostasis of biometals, and decreased
levels of acetylcholine caused by a depletion of the choli-
nergic system (Mandelkow and Mandelkow 1994;
Grathwohl et al. 2009; Murphy and LeVine 2010; Iqbal
et al. 2010; Heneka et al. 2015; Czarnecka et al. 2017).
All the manifestations of this disorder are linked to a
reduction in acetylcholine levels in the brain, resulting from
the degeneration of the cholinergic neurons (Ignasik et al.
2012) and the deposition of the β-amyloid peptide as plaque
and neurofibrillary tangles in the brain (Lahiri et al. 2002;
Wenk 2003; Musiał et al. 2007; Aliabadi et al. 2013; Anand
and Singh 2013; Guzior et al. 2014). It is thought that these
peptides and protein accumulations are toxic to neurons and
contribute to their death, producing cerebral atrophy
(Nwidu et al. 2017). Fourteen years ago, the degeneration of
cholinergic nuclei was detected in the basal forebrain during
the course of AD (Ibach and Haen 2004). Several theories
now exist to explain the origin of AD, including the cho-
linergic theory, the β-amyloid theory (extracellular deposits
of β-amyloid in brain parenchyma and neurofibrillary
intracellular tangles, which is the most studied hypothesis),
the tau hypothesis, and the inflammation theory (Kumar
et al. 2015).
Since the 1970s, the stimulation of cholinergic neuro-
transmission has been the principal strategy of many
researchers (Ozadali-Sari et al. 2017), mainly carried out by
inhibiting acetylcholinesterase (AChE, EC 3.1.1.7), an
enzyme that hydrolyzes the neurotransmitter acetylcholine
at cholinergic synapses and thus terminates nerve trans-
mission (Barnard 1974; Kamkwalala and Newhouse 2016;
Li et al. 2017). Another non-specific cholinesterase enzyme,
butyrylcholinesterase (BuChE), is also a target for treating
AD. In the late phases of the disease, antagonizing NMDA
forms part of this same strategy (Musiał et al. 2007; Bajda
et al. 2013; Anand and Singh 2013; Guzior et al. 2014).
Currently, AChEIs are the only drugs approved by the
FDA for the treatment of AD (Musiał et al. 2007). Of these
drugs, the most widely used are tacrine, donepezil, rivas-
tigmine, galantamine, and memantine (Musiał et al. 2007;
Anand and Singh 2013; Guzior et al. 2014; Kumar et al.
2015). Others such as phenserine are currently in clinical
trials (Lahiri et al. 2002; Ozadali-Sari et al. 2017). The most
promising agents for AD treatment are dual-binding
AChEIs (Ignasik et al. 2012). Diverse studies indicate that
the inhibition of AChE in AD patients can improve cog-
nitive function, delay the progression of mental deteriora-
tion, and reduce neuropsychiatric symptoms (Lahiri et al.
2002; Ozadali-Sari et al. 2017; Nwidu et al. 2017; Li et al.
2017). Certainly, AChEIs are still the best established
treatment for mild to moderate AD, although they do not
represent a cure.
Since the AChEIs currently on the market have adverse
effects (Birks 2006), there are ongoing efforts to seek new
drugs with greater potency and less toxicity. As demon-
strated by different crystals, AChE is characterized by a
deep gorge with the catalytic active site inside. This site is
surrounded by numerous other ligand recognition sites, the
nearest oxyanionic hole, an aromatic patch, an anionic site,
and a peripheral anionic site (Dvir et al. 2010; Shazi 2012;
Kiametis et al. 2017). The peripheral anionic site mediates
the aggregation rate of β-amyloid. Hence, a moiety or
molecule that binds to the latter site could possibly modify
the physiopathology of AD (Gupta and Mohan 2014).
The catalytic site for Torpedo californica AChE has been
described as having a gorge 20 Å deep and 5 Å wide. The
most important part (the stearic site) is surrounded by three
essential aminoacids, Ser200, His440, and Gly327, creating
the catalytic triad. The anionic site is composed of Trp84,
Tyr130, Phe330, and Phe331. The stabilizing transition
complex is formed by Gly118, Gly119, and Ala201, while
the peripheral anionic site consists of five aminoacids:
Tyr70, Asp72, Tyr121, Trp279, and Tyr334 (Bajda et al.
2013).
Isoindolines and isoindolines-1,3-diones, potent phar-
macophore groups found in natural and non-natural pro-
ducts, have shown anticonvulsant, anti-inflammatory,
analgesic, anti-hypertensive, and antibacterial activity
(Kukkola et al. 2001; Kim et al. 2007; Van Goethem et al.
2011; Çizmecioğlu et al. 2011; Davood et al. 2012; Shakir
et al. 2012; Raveendra et al. 2014; Barrio et al. 2015;
Achary et al. 2017). According to various research teams,
isoindoline-1,3-diones are potent and novel inhibitors of
AChE with two carbonyl groups of phthalimide that facil-
itate hydrogen bonds with the catalytic active site and the
peripheral anionic site simultaneously. Consequently, this
molecule can inhibit both the formation of β-amyloid fibrils
and the hydrolytic activity of AChE on acetylcholine. The
main interactions with AChE carried out by isoindolines-
1,3-diones and their analogs are π–π and π–cation
(Mohammadi-Farani et al. 2017). Additionally, iso-
indolines-1,3-diones are considered to provide neuropro-
tection (Mary et al. 1998; Zhao et al. 2009; Alipour et al.
2012; Mohammadi-Farani et al. 2013; Guzior et al. 2015;
Hebda et al. 2016; Si et al. 2016).
Some researchers have examined the AChE inhibitory
capacity of phthalimide derivatives, identifying inhibitory
concentrations 50 (IC50) at a micromolar scale (16.42µM)
(Mohammadi-Farani et al. 2013). Michalina et al. designed
and tested a series of 2-(diethylaminoalkyl)-isoindoline-1,3-
dione derivatives, establishing that they bind to the catalytic
active site and the peripheral anionic site. The IC50 of the

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Medicinal Chemistry Research
best compound ranged from 0.9 to 19.5µM (Ignasik et al.
2012). Bajda et al. (2013) evaluated different structures of
AChEIs by performing fragment-based design, finding 2-(5-
(2-fluorobenzylamino)pentyl)isoindoline-1,3-dione to be
the most active compound, with an IC50 of 87 nm.
were used as received. The reactions were carried out in a
flask at the melting point of the starting material and agi-
tated with a stirring bar for a few minutes in solventless
conditions (green chemistry), and then the products were
purified as described (Andrade-Jorge et al. 2017). The
reactions were monitored by TLC (0.25 mm thick silica gel,
60 F254 plates) and the products were characterized by
nuclear magnetic resonance (NMR) spectra, recorded on a
The in silico studies on the latter compounds revealed a
CH–π interaction with Trp84, a π–cation interaction with
Phe330, an H-bond with Tyr121, hydrophobic interactions in
the middle of the active gorge with Phe290, Phe331, and
Tyr334, π–π interactions with Trp279, and a CH–π with
Tyr70 by the phthalimide moiety residues (Bajda et al. 2013).
Others studies with phthalimide have shown that this kind of
compound can interact by π–π stacking and CH–π interactions
with aromatic amino acids at the anionic subsite (Phe330,
Phe331, and Trp84), and/or with residues at the peripheral
anionic site (Trp279, Tyr334) of AChE (Ignasik et al. 2012).
Due to the importance of measuring ACh levels, during the
last few decades, many techniques have been elaborated to
measure AChE activity in experiments. In 1955, Bonting and
Featherstone presented their ultramicro assay of cholinester-
ase, which was an adaptation of the hydroxamic acid method
for the determination of choline esters (Bonting and Feath-
erstone 1956). The method is time-consuming and costly in
materials. In 1961, Ellman et al. developed the colorimetric
analysis of AChE activity, measuring the enzymatic activity
based on the intensification of the color yellow produced by
thiocholine upon reacting with the dithiobisnitrobenzoate ion
(Ellman et al. 1961). Though sensitive, this method can
generate errors if the compounds tested are α,β-unsaturated
carbonyls or good electrophiles, because these compounds
tend to react with the analytical substance.
For the present study, it was decided to improve the 1956
technique by Bonting and Featherstone because of its
accuracy and easiness AChE catalytic activity detection.
Hence, it was modified to make it inexpensive and
straightforward. In a previous investigation based on this
method, we crystallized and tested an isoindoline-1,3-dione
as an AChEI (paper in peer review), finding good inhibitory
activity (inhibitory constant Ki = 330–930 µM; p < 0.05).
Therefore, new isoindolines and dioxoisoindolines with
were herein evaluated in terms of their inhibition of AChE.
Both sets of compounds were compared in order to elucidate
how carbonyl groups influence the inhibition of the enzyme.
1
Varian Mercury 300 spectrometer at 300MHz for H and
75MHz for 13C, with tetramethylsilane as internal reference.
Solutions
For the phosphate buffer (0.1 M, pH 8), 7.48 g of K2HPO4
were weighed and added to 430 mL of distilled water, while
0.916 g of KH2PO4 were mixed with 70 mL of distilled
water. Then both solutions were combined to provide the
buffer solution, which had a final volume of 500 mL and a
pH of 8 (monitoring pH with a potentiometer). For the
sodium hydroxide solution (4.2 M NaOH), 16.96 g NaOH
were dissolved in distilled water to attain a final volume of
100 mL. For the hydroxylamine solution (2.4 M NH2OH),
16.84 g NH2OH·HCl were dissolved in distilled water to
obtain a final volume of 100 mL. For the alkaline hydro-
xylamine solution, equal volumes of sodium hydroxide and
hydroxylamine solution were mixed immediately before use.
For the ferric chloride (FeCl3) solution, 20.31 g FeCl3·6H2O
were slowly mixed (to avoid any damage) with an HCl
solution (61.95%) to reach a final volume of 100 mL. For the
stock acetylcholine solution, 50 mg ACh iodide were dis-
solved in 0.695 mL phosphate buffer to afford a solution of
256 mM. Subsequently, different dilutions were prepared for
the kinetic experiment (Scheme 1a). For the stock AChE
solution, 10 U/mL lyophilized powder of AChE from Elec-
tric electricus (EE-AChE, Sigma Chemical Co. C1682) were
prepared from the original source, and the solutions were
kept at −20 °C. The three synthesized compounds and
neostigmine were prepared at concentrations of 2, 4, and
8 mM, with DMSO (<1%) or distilled water as the dis-
solution medium. The dilutions were stored at −20 °C.
Docking
The ligands were modeled on GaussView 5.0.9 software
and the protonation state was considered at physiological
conditions (pH 7.4). Conformational analysis was per-
formed on Gaussian 09 (Frisch et al. 2009) with a semi-
empirical method (PM3). The rotating bonds, torsional
degree of freedom, atomic partial charges, and non-polar H-
bonds were assigned with AutoDock tools 1.5.4. (Morris
et al. 2009). One of the crystal structures of AChE of
Electrophorus electricus was downloaded from the Protein
Data Bank (PDB code: 1C2O) (Bourne et al. 1999).
Materials and methods
Synthesis and characterization of 2,3-dihydro-1H-
isoindoles and N-substituted phthalimide
Our previously reported synthetic methodology was fol-
lowed without any modifications (Andrade-Jorge et al.
2017). In brief, reagents and solvents from Sigma-Aldrich

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Medicinal Chemistry Research
A
600 µl Sol.
(256 mM) +
600 µl buffer
400 µl Sol.
(96 mM) +
200 µl buffer
450 µl Sol.
(128 mM) +
150 µl buffer
300 µl Sol.
(64 mM) +
300 µl buffer
300 µl Sol.
(16 mM) +
300 µl buffer
300 µl Sol.
(32 mM) +
300 µl buffer
150 µl Sol.
(08 mM) +
450 µl buffer
50 mg ACh +
694 µl buffer
[2 mM]
ACh
[8 mM]
ACh
[16 mM]
ACh
[256 mM]
ACh
[128 mM]
ACh
[96 mM]
ACh
[64 mM]
ACh
[32 mM]
ACh
9
10
2
3
4
5
1
6
Blank
Buffer 160 µL
ACh curve
AChE curve
Inhibition curve
A
B
C
D
E
Buffer 140 µL
ACh 20 µL*
Buffer 120 µL
ACh 20 µL*
Buffer 100 µL
ACh 20 µL*
Test compound
13
20 µL
AChE 20 µL
AChE 20 µL
B
Incubate 20 min at 37 °C
F
(NH2OH + NaOH) 40 µL
G
H
FeCl3 100 µL
* With the corresponding concentrations of ACh: 128, 96, 64, 48, 32, 16, 8 and 2 mM.
With the corresponding concentrations of test compound: 8, 4 and 2 mM.
Scheme 1 a Dilutions performed for ACh curve. b Representation of the procedure of the improved method in a 96-well microplate
Docking conditions were programmed with AutoDock
tools 1.5.4 and Raccoon (Forli et al. 2016), by utilizing a
hybrid Lamarckian Genetic Algorithm (Morris et al. 1998), an
initial population of 100 randomly placed individuals, Koll-
mann partial charges for all protein atoms, and Gasteiger
charges for ligands. Subsequently, the grid parameter file was
established, using a 60 × 60 × 60 Å gridbox with the coordi-
nates X = 42.095, Y = 66.809, and Z = −81.47 and a mesh
spacing of 0.375 Å. The calculations were run on
AutoDock4 software in a Linux operative system (Fedora 22).
Finally, the lowest energy state, expressed as Gibbs free
energy (ΔG), was obtained for each compound. The dis-
sociation constant (Kd), the –log(10) dissociation constant
(pKd), the number of interactions, and the distance and type
of binding were determined with the Visual Molecular
Dynamics program (VMD v.1.8.6) (Humphrey et al. 1996).
96, 64, 48, 32, 16, 8, and 2mM) were placed in each well of
a 96-well microplate, following the pattern described in
Scheme 1b. The microplate was allowed to incubate for
20min at 37°C in a water bath (Felisa). Completing this
period, the reaction was stopped with 40μL alkaline
hydroxylamine solution and then 100μL of FeCl3 solution
was added to each well and the optical density was read at
540nm in a microplate reader (Benchmark BIO-RAD), with
a pre-mixed of 40 s. The standard curve of ACh was con-
structed with these values.
The same dilutions as well as incubation and reading
procedures were employed to construct the standard curve
for AChE, with 20μL of AChE in the starting mixture
(utilized to make subsequent dilutions, see Scheme 1b). The
standard and test compounds were used at 2, 4, and 8mM.
The process for evaluating AChE activity was repeated, this
time including 20μL of the test compounds before adding
AChE. All experiments were performed in triplicate and
data are expressed as the average of three values. The AChE
Anticholinesterase activity assay
Dilutions of ACh were made to prepare different con-
centrations, which were mixed in an Analog Vortex Mixer
(Thomas Scientific). Twenty μL of each solution (at 128,
activity/inhibition
Michaelis–Menten plot and non-linear regression to find the
characteristic parameters. To compare the modified method
analysis
was
made
with
a