Novel coumarin derivatives as potent acetylcholinesterase inhibitors: insight into efficacy, mode and site of inhibition

Article abstract of DOI:10.1080/07391102.2018.1465853

The inhibitory efficacy of two substituted coumarin derivatives on the activity of neurodegenerative enzyme acetylcholinesterase (AChE) was assessed in aqueous buffer as well as in the presence of human serum albumin (HSA) and compared against standard ch

Full text of DOI:10.1080/07391102.2018.1465853

Journal of Biomolecular Structure and Dynamics
ISSN: 0739-1102 (Print) 1538-0254 (Online) Journal homepage: http://www.tandfonline.com/loi/tbsd20
Novel Coumarin Derivatives as Potent
Acetylcholinesterase Inhibitors: Insight into
Efficacy, Mode and Site of Inhibition
Prayasee Baruah, Grace Basumatary, Semen O. Yesylevskyy, Kripamoy
Aguan, Ghanashyam Bez & Sivaprasad Mitra
To cite this article: Prayasee Baruah, Grace Basumatary, Semen O. Yesylevskyy, Kripamoy
Aguan, Ghanashyam Bez & Sivaprasad Mitra (2018): Novel Coumarin Derivatives as Potent
Acetylcholinesterase Inhibitors: Insight into Efficacy, Mode and Site of Inhibition, Journal of
Biomolecular Structure and Dynamics, DOI: 10.1080/07391102.2018.1465853
To link to this article: https://doi.org/10.1080/07391102.2018.1465853
Accepted author version posted online: 17
Apr 2018.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tbsd20

Publisher: Taylor & Francis
Journal: Journal of Biomolecular Structure and Dynamics
DOI: http://doi.org/10.1080/07391102.2018.1465853
Novel Coumarin Derivatives as Potent Acetylcholinesterase Inhibitors: Insight
into Efficacy, Mode and Site of Inhibition
1
a
1a
2
1b
Prayasee Baruah , Grace Basumatary, Semen O. Yesylevskyy, Kripamoy Aguan ,
1
a
1a*
Ghanashyam Bez, and Sivaprasad Mitra
1Centre for Advanced Studies in Chemistry and Department of Biotechnology & Bioinformatics,
North-Eastern Hill University, Shillong – 793 022, India
a
b
Centre for Advanced Studies in Chemistry; Department of Biotechnology & Bioinformatics.
2
Department of Physics of Biological Systems, Institute of Physics of the National Academy of
Sciences of Ukraine, Prospect Nauky 46, 03028 Kyiv, Ukraine
*
3
Author to whom all correspondence should be addressed. Phone: (91)-364-2722634. Fax: (91)-
64-2550076. E-mail: smitra@nehu.ac.in, smitranehu@gmail.com.
1

Abstract: The inhibitory efficacy of two substituted coumarin derivatives on the activity of
neurodegenerative enzyme acetylcholinesterase (AChE) was assessed in aqueous buffer as well
as in the presence of human serum albumin (HSA) and compared against standard cholinergic
AD drug, Donepezil (DON). The experimental data revealed the inhibition to be of non-
competitive type with both the systems showing substantial inhibitory activity on AChE. In fact,
one of the tested compounds Chromenyl Coumarate (CC) was found to be better inhibitor (IC =
5
0
4
8.49 ± 5.6 nM) than the reference drug DON (IC = 74.13 ± 8.3 nM), unequivocally amplifying
50
its importance. The structure of the compound was found to play a vital role in the inhibitory
efficiency, validating previous Structure Activity Relationship (SAR) reviews for coumarin. The
mechanism of inhibition remained impervious when the experimental medium was switched from
aqueous buffer to HSA, albeit noticeable change in the inhibition potency of the compound 3, 3'-
Methylene-bis (4-hydroxy coumarin) (MHC) (38%) and CC (35%). Both the coumarin
derivatives were observed to bind to the peripheral anionic site (PAS) of AChE and also found to
displace the fluorescence marker thioflavinT (ThT) from AChE binding pocket. All experimental
observations were seconded by molecular docking and MD simulation results. The inferences
drawn in this study form a foundation for further investigation on these compounds; magnifying
the probability of their usage as AD drugs and re-emphasizes the significance of drug delivery
media while considering the inhibition potencies of targeted drugs.
Key-words: coumarin derivatives; AD drugs; acetylcholinesterase inhibition; human serum
albumin; peripheral anionic site.
2

1
. Introduction
Alzheimer’s disease (AD), a chronic neurodegenerative disorder characterized by
progressive cognitive and neuropsychiatric dysfunction, is the most prevalent form of dementia,
accounting for roughly 60% of all types of dementia (Gard & Rusted, 2004; Kalaria et al., 2008).
The alarmingly large number of populace impaired by the disease and its escalating growth rate
have called for extensive research in discovering novel AD drugs with improved therapeutic
effectiveness that could ameliorate and/or slowdown the progression of AD process (Ager et al.,
2
015; Wimo et al., 2017; Wimo, Jönsson, Bond, Prince, & Winblad, 2013). Several approaches
have been undertaken to isolate/design novel drugs for the treatment of AD by exploring various
AD-forming pathways. These include inhibition of both the AChE and Amyloid-beta (A)
pathways experimentally (Azam et al., 2017; Ferreira Neto et al., 2017; Gurung, Aguan, Mitra,
&
Bhattacharjee, 2016; Hojati, Ghahghaei, & Lagzian, 2017; Islam & Pillay, 2018) and also
through pharmacophore screening (Iqbal, Anantha Krishnan, & Gunasekaran, 2017; Jafari et al.,
2
018; Malik, Choudhary, Srivastava, Mehta, & Sharma, 2017; Patel et al., 2017).
Acetylcholinesterase (AChE) is the enzyme involved in the termination of impulse
transmission through rapid hydrolysis of the neurotransmitter acetylcholine (ACh), a progressive
decline of which plays a key role in the pathogenesis of AD (Craig, Hong, & McDonald, 2011;
Francis, Palmer, Snape, & Wilcock, 1999; Peters et al., 2012). Several drugs used in the
treatment of AD are based on the established cholinergic hypothesis (Colovic, Krstic, Lazarevic-
Pasti, Bondzic, & Vasic, 2013; McGleenon, Dynan, & Passmore, 1999), where the objective is to
increase the concentration of ACh in the synaptic cleft by the inhibition of cholinesterase (ChE)
activity. AChE inhibitors (AChEIs) impede the breakdown of the cholinesterase enzyme,
augmenting the synaptic accessibility of ACh in the brain and subsequently boosting cholinergic
3

neurotransmission in forebrain regions that results in compensating the loss of functioning of
brain cells. Owing to this, AChE inhibition has been documented as a critical treatment route of
AD (Birks, 2006; McGleenon et al., 1999; Zemek et al., 2014).
AChEIs belong to myriad classes of structural and functional groups of compounds.
However, numerous side effects of the available AD drugs have spurred the designing of novel
AChEIs for pharmacological usage. The discovery of ensaculin (Hilgert, Nöldner, Chatterjee, &
Klein, 1999; Hoerr & Noeldner, 2002), which is a coumarin derivative with piperazine ring
having the ability to slow down the progression of AD led to the exploration of the efficacy of
varied natural and synthetic coumarin derivatives as AChEIs. Coumarins are naturally occurring
organic compounds which exhibit a variety of physiological and clinical applications with
diverse effects, out of which AChE inhibition is one of the well-reported events (Jain & Joshi,
2
012).
The crucial structural features of AChE consist of a catalytic active site (CAS) and a
peripheral active site (PAS) connected by a narrow, yet deep (20 Å) gorge (Martis et al., 2015).
The location of the CAS is near the bottom of the gorge; whereas, the PAS is situated near the
enzyme surface (Colletier et al., 2006). The knowledge of inhibitor binding site is of primary
importance for AChE, since the binding to the PAS has also been implicated in playing a role to
decrease the aggregating effects of the enzyme on the amyloid peptide (Alvarez et al., 1995). A
range of AChEIs have been developed for AD which can bind to the CAS, the PAS, or both
(Castro & Martinez, 2001; Chen et al., 2015; Gupta & Mohan, 2014).
Coumarins are known to bind with the Peripheral Anionic Site (PAS) of AChE (Fallarero
et al., 2008); but, functionalization of the coumarin heterocycles and substitution of the coumarin
template, particularly at positions 3 and 4 with amine functional moiety including benzylamino,
4

phenylpiperazine or aniline groups using appropriate spacer, renders it capable of greater
interaction on AChE, increasing the potency of inhibition (Anand, Singh, & Singh, 2012; Shen et
al., 2005). Further, recent studies have also revealed that a large number of coumarin derivatives
with a longer lateral chain promotes a good interaction with peripheral active site of AChE and
hence represents the best inhibitory activity (de Souza, Rennã, & Figueroa-Villar, 2016).
Nevertheless, irrespective of their binding site (CAS or PAS) with AChE, coumarins have shown
promising activity as AChEIs. Further, it has also been reported that 4-hydroxycoumarin
derivatives containing an amine functional group on alkyl side chain, particularly a tertiary
amino group, show promising AChEI activity (Razavi et al., 2013). The detection of crucial
structural features within the coumarin template and comprehensive reviews on SAR have
helped in designing as well as synthesizing new coumarin analogues with enhanced AChE
inhibitory activity.
Ellman’s method (Ellman, Courtney, Andres, & Featherstone, 1961) remains the most
popular avenue to assess the potency of newly developed AChEIs, where the enzymatic
hydrolysis is normally carried out in a buffer solution of pH 8.0. However, considering the
complex nature of the body fluid, experiments in aqueous buffer solution is considered to be an
extremely simplified and crude model that is incapable of correctly predicting the potency as
well as efficacy of an inhibitor over the existing drugs. Therefore, in-vitro experiments in
presence of human serum albumin (HSA), a highly soluble negatively charged protein occupying
almost 60% in blood plasma (Lexa, Dolghih, & Jacobson, 2014; Zunszain, Ghuman, McDonagh,
&
Curry, 2008) is considered to be more relevant and realistic. HSA plays the role of a transport
protein (Zergani et al., 2012) and further finds its use clinically as a drug delivery system owing
to the prospect of improved targeting and reducing the side effects (Yang, Zhang, & Liang,
5

2
014). In our previous work, we have already reported the effect of HSA on some of the well-
known AD drugs (Islam, Gurung, Bhattacharjee, Aguan, & Mitra, 2016).
In the current work, the inhibitory effect of a new 4-hydroxycoumarin derivative (6-
amino-5-(3-(4-hydroxy-2-oxo-2H-chrome-3-yl)-2-phenylpropyl)-1,3-dimethylpyrimidine-
2
,4(1H,3H)-dione) containing amine functional groups (Chromenyl Coumarate, CC) on AChE
activity is evaluated. The result is compared with another coumarin compound: 3, 3'- Methylene-
bis (4-hydroxy coumarin), MHC which is basically two hydroxyl coumarin skeletons linked by a
methylene group. The structures of both these systems are shown in Scheme - I.
Scheme - I. The structures of the investigated systems: CC (a) and MHC (b)
The mechanism of binding as well as the efficacies of inhibition for both the coumarin
derivatives have been evaluated against donepezil (DON), the most widely prescribed reversible
AChE inhibitor, which has been found to be better tolerated amongst other known AChEIs with
apparently less gastrointestinal side effects than rivastigmine or galanthamine (Tayeb, Yang,
Price, & Tarazi, 2012). The results inferred are consequences of both enzyme kinetics as well as
6

fluorescence experiments. The mode of inhibition has been investigated from kinetic parameters
in buffer solution and also in the presence of HSA under standard physiological condition. This
study puts forward the possibility of these coumarin derivatives as potent new AChEIs taking
into account of the substantial AChE inhibitory activities toward the treatment of AD through
cholinergic pathway and also reinstates the modulations caused by the change in delivery media
in determining their efficiency.
2
2
.
Experimental methods
.1.
Chemicals
Acetylcholinesterase (AChE) was purchased as type V-S, lyophilized powder from
Electrophorus electricus (electric eel) with activity 1000 units/mg of protein, from Sigma Aldrich
Chemical Company (cat. No. C2888). The reference inhibitor, Donepezil hydrochloride
monohydrate (DON, purity 98% HPLC, cat no. D6821), the substrate acetylthiocholine iodide
(purity 98% TLC powder or crystals cat no. A5751) and thioflavinT (ThT, cat. No. T3516) were
also received from Sigma and used as received without any additional purification. Dithiobis (2-
nitrobenzoic acid) (DTNB, the Ellman reagent) in the extra-pure form was procured from Sisco
Research Laboratories (SRL), India (product no. 054817, 044883 and 1944142 respectively.
Chromenyl Coumarate (CC) was synthesized following the protocol reported below; whereas, 3,
3
' – Methylene-bis(4-hydroxy coumarin), MHC was obtained from Sigma Aldrich Chemical
Company (cat. No.66762). Human Serum Albumin (HSA) was purchased as a fatty acid and
globulin free, >99% (agarose gel electrophoresis) lyophilized powder (USB Corp. USA, cat. no.
1
0878 and was used as received. The gelatin used for preparing the AChE aliquots was received
from Qualigens fine chemicals (a division of Glaxo Smithkline Pharmaceuticals Ltd., India). The
analytical grade type II water used in carrying out the experiments (resistivity ~10 MΩ cm at
7

room temperature) was acquired from Elix 10 water purification system (Millipore India Pvt.
Ltd.). Systronics µ-pH system 361 was used for checking the pH of the buffer solution.
2
.2. Synthesis and characterization of CC
The compound CC was synthesized following the reported procedure (Bharti & Parvin,
015) with slight modification. To an equimolar solution of 4-hydroxycoumarin (0.162 g, 1
2
mmol) and benzaldehyde (0.106 g, 1 mmol) in 1 mL ethanol, a solution of L-proline (0.023 g,
.2 mmol) in water (0.2 mL) was added and stirred under reflux for 15 min. Then 6-amino-1,3-
0
dimethyluracil (0.270 g, 1 mmol) was added to the reaction mixture and stirred for additional 5
h. The resulting mixture was cooled to room temperature, and the solid was filtered off and
washed with water and ethanol to obtain the pure white solid product in 81% yield (0.328 g, 0.81
°
 1
mmol). Melting point 215 C; IR (KBr):         cm  H
NMR (400 MHz, DMSO-d , δ ppm): 3.15 (s, 3H, -CH ), 3.39 (s, 3H, -CH ), 5.64 (s, 1H, -CH),
6
3
3
7
.15-7.39 (m, 8H, ArH, -NH ), 7.45 (d, J= 8 Hz, 1H, ArH), 7.66 (t, J= 8 Hz, 1H, ArH), 7.85 (d,
2
1
3
J = 8 Hz, 1H, ArH), 14.0 (s, 1H, -OH); C NMR (100 MHz, DMSO-d , δ ppm): 28.24, 30.58,
6
3
6.06, 74.91, 86.81, 104.65, 116.16, 116.95, 123.73, 124.33, 125.74, 126.38, 128.12, 132.44,
+
+
1
38.30, 150.07, 151.97, 155.17, 163.84, 164.10, 165.85. ESI-MS (m/z): 428 [M+Na] , 405 [M] .
1
13
H NMR, C NMR and ESI-MZ spectra are represented in Fig. 1S – 3S in the supplementary
section.
2
.3. Measurement of AChE enzyme activity
Kinetic characterization of AChE was performed by following the method of Ellman et al.
(1961) with slight modification in 0.1 M phosphate buffer of pH = 8.0at room temperature (298
K). The enzyme kinetics experiment in the absence and presence of HSA (concentration fixed at
2
50 µM) was carried out in synergy H1 hybrid meter plate reader instrument (BioTek) using 96
8

well plates. For both enzymatic and non-enzymatic studies, all the components of the reaction
mixture were incubated for 5 min on ice cold water bath. The reaction was followed
spectrophotometrically at 412 nm for 45 minutes inside the plate reader at the experimental
temperature to follow the kinetics. The substrate concentration was varied from 50 µM to 2.0
mM in the experiments for measuring the kinetic parameters (Michaelis Menten constant, Km
and maximum hydrolysis rate, Vmax) at two different inhibitor concentration. The effect of HSA
on the IC values (concentration of the inhibitor at half maximal inhibition) of the inhibitor was
5
0
checked by fixing the substrate concentration at the saturated condition of AChE catalysis
reaction (~1.5 mM); while varying the inhibitor concentrations from very low till the saturation
inhibition condition. The non-enzymatic hydrolysis reactions were performed by adding all the
components except enzyme and were taken as corresponding blanks for the reaction. In both the
non-enzymatic as well as enzymatic reactions, the concentration of DTNB was kept fixed at
~317 mM and the enzyme concentrations were ~0.079 units/ml in the enzymatic reactions. All
the experimental results reported here are averaged for at least three independent sets.
2
.4. Kinetic characterization for the enzyme catalysis and data analysis
The kinetic parameters were evaluated and data analysis was done using the Microcal Origin
2
017 software. The non-enzymatic hydrolysis curve was subtracted off from the corresponding
enzymatic hydrolysis curve and the initial linear portion of the progress curve considered for
calculating the initial velocity of the reaction, V by using Eq. (1).
0
Slope
Initial rate, V (in moles per litre per second) =
(1)
0
є × l
Where ε is the molar absorptivity of the yellow anion, (Pheifer & Briggs, 1995), l = 0.442 cm is
the path length and the slope is in the units of absorbance per second. The enzyme kinetics was
studied based on the kinetic scheme shown in Scheme - II.
9

Scheme - II. The Michaelis-Menten scheme for enzyme catalysis and inhibition
Michaelis-Menten (MM) constant, K and maximum hydrolysis rate, V which are the
m
max
standard measures for enzyme kinetics in biochemistry were obtained by non-linear regression
analysis of the velocity (V ) data against [S ] with the relations corresponding to normal and
0
0
inhibition cases given by Eq. 2a and 2b, respectively (Nelson & Cox, 2004).
Vmax[S0 ]
V =
0
(2a)
K + [S ]
m
0
Vmax'[S0 ]
V =
0
(2b)
'
K + [S ]
m
0
Where the various terms are defined as follows:
푉푚푎푥
훼
V '=
, K ' = × K , or, K ' = α ×K (for inhibition path A only);
max
m
m
m
m
훼'
훼'
[
퐸퐼]
[퐼퐸푆]
α = 1 +
and α' = 1+ [퐸푆] respectively (for both inhibition paths A and B)
[
퐸]
1
0

The IC values of all the inhibitors both in absence and presence of HSA were calculated from
5
0
the K0.5 values by non-linear regression analysis using the modified Hill relation given in Eq. (3)
(Copeland, 2005).
푛
퐻
∆
푉
[퐼]
=
(3)
푛
푛
∆
푉푚푎푥
퐻
퐻
퐾0 + [퐼]
.5
Where, ∆V represents the decrease in initial velocity in the presence of certain inhibitor
concentration [I], and ∆Vmax denotes the maximum decrease in initial velocity, K0.5 is the
concentration of the inhibitor that gives half-maximal initial velocity change and n is the Hill
H
coefficient. The IC values were further verified by a non-linear regression of four parameter
5
0
logistic plots (Volpe, Hamed, & Zhang, 2014) for individual drugs (not shown) and were found
to be consistent with the values obtained from analysis of the modified Hill equation.
2
.5. Fluorescence measurements
The medium used for the fluorescence measurements was 0.1 M phosphate buffer of pH
8
.0. The solutions were prepared afresh and the spectroscopic measurements were performed
after incubation of the solutions for sufficient time. Fluorescence spectral measurements were
performed in Quanta Master (QM-40) steady state apparatus obtained from Photon Technology
International (PTI) .The spectra were corrected for the instrument response function. Samples
were excited at 412 nm for detecting the ThT fluorescence emission. The results were averaged
from three separate experiments and analysed using Origin 2017 (Microcal Software, Inc., USA).
Despite the compounds not having significant absorption at the excitation wavelength, the
fluorescence spectra were corrected for any possible inner filter effect using the following
equation (Islam, Sonu, Gashnga, Moyon, & Mitra, 2016):
퐴(휆퐸)
퐶표푟푟
퐹
(
휆퐸, 휆퐹
)
= 퐹푂푏푠
(
휆퐸,휆퐹) × 퐴
(4)
(휆 )
푡표푡
퐸
1
1

Where A represents the absorbance of the free ThT, and A is the total absorbance of the
tot
solution at the excitation wavelength (λ ). After correction, the spectrum was considered to be
E
the contribution from only ThT.
2
.6. Molecular docking calculation
Molecular docking study was executed in two steps: the first step comprised of the protein
and ligand molecules preparation which was followed by docking analysis. The 3D structures of
the tested compounds were prepared using ChemDraw and optimized at the PM7 semi-empirical
level of theory in Gaussian16 (Frisch et al. 2016) following the protocol described in the
literature (Cazelli et al., 2017; Gonçalves et al., 2015). The 3D structures of AChE (PDB ID:
1
C2O) and HSA (PDB ID: 1AO6) were retrieved from Protein Data Bank (www.rcsb.org). The
protein preparation steps included the removal of all heteroatoms including ions and water
molecules, addition of polar hydrogens and assignment of Kollmann charges using AutoDock
tools-1.5.6 (ADT). The docking was executed using AutoDockVina (Trott & Olson, 2009)
following the protocol of Lamarckian Genetic Algorithm (LGA) (Morris et al., 1998) for
docking to protein targets. The detailed parameters for running the docking calculation are
incorporated in the supplementary section.
Pymol was used for molecular visualization (Seeliger & De Groot, 2010). The
conformational cluster analysis was done using maximum RMS tolerance of 2.00. Ten docking
rounds were carried out for both the systems. The lowest energy conformer was chosen for
modelling the docked images using Chimera (Pettersen et al, 2004) and Ligplot+ (Lakowski &
Swindells, 2011) systems. The theoretically estimated inhibition constant (K ) was calculated
i
using Eq. 5; where, ΔGbinding (binding free energy) represents the sum of the inter- and
1
2

intramolecular enthalpies, R is the universal gas constant, and T denotes the temperature in
Kelvin (Cazelli et al, 2017).
ΔG
푏푖푛푑푖푛푔
푅푇
퐾 = 푒
(5)
푖
2
.7.
Molecular dynamics (MD) simulations
The best docking conformers of CC and MHC in complexes with AChE were used as input
structures for MD simulations. Initial topologies of CC and MHC were generated by Acpype
topology generator (da Silva & Vranken, 2012). The AMBER99 force field was used. The
structures of both the ligands were optimized in Gaussian09 (Frisch et al, 2009) at the B3LYP/6-
3
1G level of theory. The ESP partial charges were computed and added to initial topology. The
charges of topologically equivalent atoms where averaged. All MD simulations were performed
in Gromacs 5.1.2 package (Abraham et al, 2015). The protein-ligand complexes were solvated in
+
truncated dodecahedron box containing ~13700 water molecules and 9 Na ions, which balance
the net negative charge of the protein. TIP3P water model was used. All simulations were
performed in NPT conditions with the temperature 310 K and the pressure of 1 bar maintained
by v-rescale thermostat and Berendsen barostat with isotropic pressure coupling respectively.
The time step of 2 fs was used with all bonds converted to rigid constraints.
The following simulation protocol was used. On the first stage all heavy atoms of the protein
-2
and the ligands were restrained by harmonic restraints with the force constant of 1000 kJ·nm .
The system was simulated for 2 ns to equilibrate water and ions. On the second stage only the
atoms of protein backbone were restrained for 10 ns to equilibrate protein side chains and
optimize position of the ligand accordingly. Finally, on the third stage free MD simulation
without restraints was performed for 10 ns.
1
3

The stages 2 and 3 were used for analysis. The RMSDs of the ligands and protein heavy
atoms and interaction energies between the ligands, protein and water were monitored. The
energies were computed within the cutoff of 0.8 nm. Analysis of MD trajectories was performed
with Pteros molecular modeling library (Yesylevskyy, 2012; Yesylevskyy, 2015).
3
3
. Results and Discussion
.1. Quantification of inhibitor potency in buffer medium
Both the investigated coumarin systems, viz. CC and MHC were found to exhibit inhibitory
effect on the activity of free enzyme. An analysis of the kinetic parameters as well as the nature
of Lineweaver-Burk (LB) plot point to the fact that the inhibition mechanism is analogous for all
the tested systems. The inhibition, in all the three cases, is associated with the reduction of the
Vmax values while the magnitude of K remains unaltered; a condition characteristic of a non-
m
competitive type of inhibition, which is known to be a special type of the mixed inhibition
mechanism. In the non-competitive pathway, an inhibitor binds to the enzyme irrespective of
whether the enzyme has bound to the substrate or not and hence involves both the inhibition
pathways (marked A and B in Scheme - II), both of which are known to proceed simultaneously
with comparable affinities. This process also renders identical values of α and αʹ. The MM as
well as the LB plots is shown in Fig. 1 & Fig. 2, respectively.
1
4

Fig. 1.Hydrolysis curve (scattered points) for AChE activity and its inhibition in presence of different
concentrations of CC (a) and MHC (b) in phosphate buffer solution of pH = 8.0. The solid line represents
non-linear regression of the experimental data points by Eq. (2). [AChE] = 0.079 u/ml.
Fig 2. Lineweaver-Burk (LB) plot for the AChE hydrolysis data in aqueous buffer (a) and HSA medium
(b) in absence and presence of CC and MHC.
1
5

DON was used as a reference compound and the results of kinetic analysis were found to be in
good agreement with the previously published reports (Islam, et al., 2016; Mehta, Adem, &
Sabbagh, 2012). The calculation of the IC values were done from the enzyme hydrolysis data
5
0
(Fig. 3) using modified Hill relation given in Eq. 3. The kinetic results (Table 1) indicate that CC
acts as the most potent AChE inhibitor with an IC value of 48.49±4.6 nM followed by MHC
5
0
(80.16 ± 7.1 nM). All the compounds have been found to lower the AChE activity substantially
with CC having an IC value even lower than DON (74.13 ± 8.3 nM), signifying its irrefutably
5
0
strong potency towards AChE inhibition.
Fig. 3. Modified Hill analysis of AChE inhibition in aqueous buffer (a) and HSA medium (b) of CC and
MHC.
1
6

The non-competitive mechanism is usually connected with binding at the PAS site of AChE,
with Ethidium Bromide (EB) and Propidium Iodide (PI) being known examples in this category
(
Tong et al., 2013). Also in the present investigation, the molecular docking calculation for the
action of the investigated coumarin derivatives on AChE confirms the binding site to be PAS
see molecular docking calculation results below). The PAS has been reported to be involved in
(
substrate inhibition characteristics of AChE and binds ACh and other quaternary ligands, which
act as uncompetitive inhibitors binding at a site separate from that occupied by the mono-
quaternary competitive inhibitors (Colovic, et al., 2013). The presence of nitrogen containing
moiety has been found to exhibit stronger interaction with AChE (Di Pietro et al., 2014; Pandolfi
et al., 2017) and can explain the highest inhibitory potency of CC which is consistent with the
experimental results reported here.
Table 1: Kinetic data of AChE hydrolysis and the effect of different concentration of inhibitors on
various parameters in aqueous buffer medium.a
Systems
Km/µM
Vmax/nM s-1
From Michaelis Menten equation
α
α'
IC50 (nM)
From modified Hill equation
nH
AChE in presence of CC
[
[
[
CC] = 0 nM
141 ± 21
149 ± 27
858 ± 42
498 ± 38
401 ± 22
1.0
1.7
2.4
1.0
1.7
2.1
48.49 ± 5.6
1.44 ± 0.1
CC] = 5 nM
CC] = 10 nM 167 ± 30
AChE in presence of MHC
[MHC] = 0 nM 170 ± 28
[MHC] = 5 nM 174 ± 24
[MHC] =10 nM 166 ± 31
808 ± 26
617 ± 24
527 ± 37
1.0
1.3
1.5
1.0
1.3
1.5
80.16 ± 7.1
1.51 ± 0.1
a
[AChE] = 0.079 units/ ml
1
7

3
.2. Quantification of inhibitor potency in presence of HSA
The modulation of drug efficiency in HSA, which is a highly versatile protein, is important
in clinical diagnosis and is widely studied (Ariga, Naik, Nandibewoor, & Chimatadar, 2016;
Jafari et al., 2018; Karthikeyan et al., 2017; Khan & Suresh Kumar, 2017; Wan et al., 2017)
since serum albumin has been observed to bind compactly with drugs, even at very low
concentration and it harbours multifunctional properties that are relevant to key pathologies
underlying AD, including inhibition of the kinetics of Aβ fibrillation (Ezra, Rabinovich-Nikitin,
Rabinovich-Toidman, & Solomon, 2015; Stanyon & Viles, 2012). It has been reported that low
serum albumin level group of AD patients performed better in cognition level compared to
higher serum level group (Llewellyn, Langa, Friedland, & Lang, 2010). To check the effect of
HSA on AChE inhibition activity of the investigated compounds, the enzymatic reaction was
monitored under the condition of fixed concentration of HSA (ca. 250 mM), which is close to the
reported level of physiological abundance of serum albumins (Taverna, Marie, Mira, & Guidet,
2
013). The control experiment was performed with free enzyme (without inhibitor) in presence
of same concentration of the protein. In general, the inhibitory activity of both the compounds
reduces significantly in presence of HSA with MHC being affected the most. However, LB
analysis of the kinetic data indicates that the inhibition mechanism remains unaltered even in
presence of HSA. The LB plots and the MM plots have been represented in Fig. 2 & Fig. 4,
respectively whereas the corresponding kinetic parameters have been depicted in table 2.
1
8

Fig. 4. Hydrolysis curve (scattered points) for AChE activity and its inhibition in presence of
different concentrations of CC (a) and MHC (b) in HSA. The solid line represents non-linear
regression of the experimental data points by Eq. (2). [AChE] = 0.079 u/ml.
1
9

Table 2: Kinetic data of AChE hydrolysis and the effect of different concentration of inhibitors on
various parameters in HSA medium.a
Systems
Km/µM
Vmax/nM s-1
α
α'
IC50 (nM)
nH
(From Michaelis Menten equation)
(From modified Hill equation)
AChE in presence of CC
[CC] = 0 nM
[CC] = 5 nM
[CC] = 10 nM
143 ± 17 784 ± 57
158 ± 17 545 ± 30
156 ± 38 486 ± 38
1.0
1.5
1.7
1.0
1.5
1.6
65.49 ± 8.0
1.40 ± 0.1
AChE in presence of MHC
[
[
[
MHC] = 0 nM
169 ± 23 825 ± 23
178 ± 17 667 ± 18
1.0
1.3
1.4
1.0
1.3
1.4
110.34 ± 11.6
1.51 ± 0.1
MHC] = 5 nM
MHC] = 10 nM 174 ± 22 619 ± 22
a
[AChE] = 0.079 units/ ml
The extent of reduction in the inhibition efficiency also varies significantly among different
compounds. The comparative plot showing the Hill analysis of various inhibitors in presence of
HSA is represented in Fig. 3. It is interesting to note that even in HSA medium, CC (IC = 65.49
5
0
±
8.0 nM) shows greater capacity of inhibiting AChE activity than standard drug DON in
aqueous medium (74.13 ± 8.32 nM). The kinetic parameters as well as the figures for analysis of
inhibition potency of the reference compound (DON) have been given in the supplementary
section (Table 1S & Figs. 4S – 6S, respectively).
The modulation of inhibition activity of the investigated coumarin systems in serum
albumin matrix can be attributed to be due to the sequestration of the compounds by HSA and
subsequent reduction in the availability of active fraction of the free inhibitor in solution (Islam,
et al., 2016). The quantitative information on the binding affinities of the tested compounds can
2
0

be calculated by molecular docking studies, discussed later. Thus MHC, which binds with HSA
more strongly, shows significant reduction of potency toward AChE inhibition in serum albumin,
in comparison with its inhibition activity in aqueous buffer solution. The relative change in the
IC values for all the compounds in HSA were calculated by using a protocol described earlier
5
0
(Islam, et al., 2016) and were found to be 35% and 38% for CC and MHC respectively, which
confirms the hypothesis proposed above.
3
.3. Fluorescence spectral results
Thioflavin T is a well-known probe for monitoring AChE fluorescence. In one of our
previous works, we have reported that the cholinergic drugs have a quenching effect on AChE
bound ThT fluorescence (Islam & Mitra, 2016). In conformity with this observation, here too the
two coumarin compounds exhibited quenching effect on ThT-AChE emission intensity. The
extent of quenching was found to be more in case of CC (Fig. 7S) and consistent with the
observations made from the kinetics experiments discussed in the previous section. The
fluorescence excitation spectra of ThT-AChE binary system also remained unaltered with the
addition of drug. The trends of the Stern-Volmer analysis using Eq. 5(a) are similar to those of
the earlier reported cholinergic drugs and exhibit a downward curvature. The experimental data
points did not show acceptable linear behaviour even when fitted with the modified Stern-
Volmer equation (Eq. 5(b)).
퐹0
=
1 + 퐾 [Q]
5 (a)
5 (b)
푆푉
퐹
퐹0
1
1
= 푓
+
퐹0 ‒ 퐹
푓 푘 [푄]
푎
푎 푎
Hence, these synthesized coumarin compounds could be envisaged to follow a similar activity
with those of standard AChEI drugs. The quenching of AChE bound ThT fluorescence in
presence of these compounds can be cogitated to be the result of the partial replacement of ThT
2
1

from the AChE gorge to the aqueous buffer phase. From the kinetics experiment as well as
molecular docking results to be discussed in the next section, CC and MHC have been found to
have the tendency to bind at the peripheral anionic site (PAS) of the AChE binding region, thus
hindering the approach of ACh enroute for its hydrolysis by the enzyme. Therefore, the coumarin
compounds compete towards the AChE binding pocket and addition of these compounds in
AChE – ThT system causes the substitution of a fraction of bound ThT from the AChE gorge.
The compound CC, which possesses a greater affinity for AChE than ThT manages to displace
ThT with ease, resulting in a sizeable decrease in fluorescence spectral intensity. On the other
hand, MHC with a relatively lesser inhibitory effectiveness, reveals lesser degree of curvature
(greater linearity) in the Stern-Volmer plots while evaluated against relatively strong inhibitor,
CC. These results re-emphasize the validity of fluorimetric assay protocol for assessing the
inhibition efficiency of the drugs. The results of the quenching experiments for the drugs with
HSA are found to have good correlation with the kinetic results. For example, MHC shows a
-1
4
greater quenching efficiency than CC with representative K (M ) values of (5.90.21) × 10
SV
4
and (4.60.09) × 10 respectively, while considering the linear portion of the graph for individual
systems. The emission spectra for the binding of MHC with ThT-AChE system and with HSA
are given in Fig. 5; while the corresponding graphs for CC can be found in the supplementary
section (Fig. 7S).
2
2

Fig 5. Quenching of emission intensity of ThT-AChE binary system (a) and HSA (b) in presence
of increasing concentrations of MHC. Inset shows the Stern Volmer analysis.
3
3
.4. Molecular Docking Results
.4.1. Docking with AChE
The active site of AChE (catalytic site, CAS) consists of Ser 200, Glu 327 and His 440;
whereas the peripheral anionic site (PAS) is characterized mainly by five amino acids—Tyr70,
Asp72, Tyr121, Trp279 and Tyr334 (Bajda et al., 2013). The PAS is located at the entry to the
active gorge and is responsible for extra activities, including the interaction with β-amyloid.
Since the PAS is located on the enzyme surface, it directly occludes the active-site gorge. Thus,
the binding of any ligand at this location would block the entry of substrates and the exit of
products from the foot of the active site. This blockade can exhibit purposeful significance, as
during ACh hydrolysis an incoming ACh molecule bound at PAS blocks the exit of produced
choline molecule in active site, thus generating substrate inhibition (Dvir, Silman, Harel,
Rosenberry & Sussman, 2010).
2
3

From the docking results, we have further validated our experimental findings as well as
correlated the literature reports of general binding mode of coumarins with AChE. The docking
images of both the compounds have been depicted in Fig 6.
Fig 6. Lowest energy docked poses of CC (a) and MHC (b) with AChE. The binding site has been
magnified for clarity.
From the docked pose, visual inspection conveys that neither of the ligands penetrates
very deeply with the AChE gorge. Both the compounds have been found to bind at the periphery
of the enzyme and in close contact with the residues characterizing the PAS (magnified in Fig 6),
albeit in not the exact conformations. While considering a distance of 3.80 A, it was found that
the ligands do not interact via hydrogen bonding or hydrophobic interactions with any residue of
the active site gorge, despite being in close enough proximity to cause steric hindrance to the
incoming substrate. The residues having closest contact are represented in table 3 and shown in
2
4

Fig. 8S. Atom O1 of CC is found to interact via H-bonding with His398 along with hydrophobic
interactions whereas all interactions in case of MHC were found to be hydrophobic. CC is found
to have the highest binding affinity with AChE (- 34.7 kJ/mol), followed by MHC (-33.0 kJ/mol)
indicating CC to have a stronger binding with AChE. From docking, the binding energy of ThT
was estimated to be 31.7 KJ/mol (Table 2S), which is very close to the binding affinities of CC
and MHC. This provides the thermodynamic justification of the replacement of ThT in presence
of inhibitors, as predicted by the fluorescence experimental results. The trend of inhibition
constants remain the same from docking as obtained from experimental results.
3
.4.2. Docking with HSA
The structure of HSA contains three homologous domains (named I, II, and III). Each
domain contains two separate helical subdomains (named A and B), connected by random coil.
The modular structural organization of HSA provides a variety of ligand binding site. Both CC
and MHC have been found to bind at Domain I of HSA. While CC binds at domain IA, MHC
has been found to bind at Domain IB of HSA with higher affinity. Table 3 and fig. 9S depict the
residues in closest contact with the coumarins. The results indicate relatively stronger interaction
of MHC with HSA (-38.51 kJ/mol) than CC (-37.52 kJ/mol). As in case of AChE, CC was found
to form H-bonding with Leu115 and Arg117 through atoms O1 and O2 respectively in addition
to hydrophobic interactions, whereas all interactions of MHC with its closest residues were
hydrophobic in nature. The negative total free energy change observed for the interaction of all
the tested compounds with AChE indicate a spontaneous binding and corroborate the
experimentally observed inhibition behaviour of these inhibitors. Further, the binding affinity
data given in Table 3 for the best pose structures of the inhibitors with AChE as well as with
HSA, also predicts CC to be the most active AChE inhibitor, in conformity with the
2
5

experimental results discussed before. The corresponding parameters for some of the other poses
are given in supplementary Table 3S.
Table 3. Binding affinities and other relevant parameters of the most stable docked pose of
different inhibitors with HSA and AChE obtained from molecular docking calculation.
System ΔGbinding
K (µM) Binding Residues
i
(kJ/mol)
H Bonding
partners
Hydrophobic
(distance, Å)
CC and
AChE
-34.748
-33.069
0.81
O1---His398
(3.17)
Trp524, Asn525, His406,
Glu306, Pro232, Pro 403,
Asn230
MHC
and
AChE
1.59
His398, Trp524, Pro232,
Pro403, Asn525, Asn230,
His406,Glu306, Leu528,
His362, Pro529, Leu532
Tyr138, Met123, Leu182,
Tyr161, Ile142, Gly189,
Arg186, lys190, Arg145,
Arg114
CC and
HSA
-37.255
-38.512
0.29
O1---Leu115
(3.24)
O2---Arg117
(3.27)
MHC
and
0.17
Phe70, Ala26, Leu251,
Leu66,
HSA
Leu250,Leu22,Ala254,
Pro152, Leu14, Arg10
3
.5.
Molecular Dynamics Simulations
Fig. 7(a) shows RMSD of the ligands and heavy atoms of AChE during MD simulations. It is
clearly seen that during first 10 ns of simulation, when the protein backbone is restrained, the
protein side chains evolve slightly from their positions in the crystal structure which results in
RMSD of ~0.1 nm. The last 10 ns of simulation, which were performed without restraints, show
increase of protein RMSD to ~0.2 nm. This value is very small for the protein of such significant
size and indicates that the AChE structure is stable in water environment during MD simulations
and remains close to its crystal structure.
2
6

Initial positions and conformations of the ligands obtained from docking evolve during the
course of MD simulations. The CC ligand moves a bit away from its docking position and
stabilizes at new conformation with average RMSD of ~0.8 nm from its initial structure. The
behavior of MHC ligand is more complex. It shows several metastable states with different
RMSD values but eventually also stabilize at conformation with RMSD of ~0.8 nm from its
initial structure. Despite rather large deviations from initial docking conformations the final
positions and orientations of the both ligands are remarkably stable.
Fig 7: RMSD of ligands and heavy atoms of AChE (a) and evolution of interaction energy between
ligands, AChE (protein) and water (b) during MD simulations. The curves on (b) are smoothed by
running average method for clarity.
Fig. 7(b) shows evolution of interaction energy between the ligands, protein (AChE) and
water during MD simulations. Both CC and MHC ligands initially have comparable interaction
energy with protein, which is in agreement with their similar docking scores. The interaction of
CC ligand with water becomes much stronger during the first 3 ns of simulation and then
stabilizes at ca. -250 kJ/mol. In contrast, its interaction with AChE decreases from -150 kJ/mol to
-100 kJ/mol. Several transient fluctuations are observed when the CC ligand binds stronger to
2
7

protein and lose contacts with water molecules. The MHC ligand behaves similar to CC during
the first 10 ns of simulations when the protein backbone is restrained. After the release of this
restraint, its interaction with protein increases dramatically and interaction with water decreases
to the same extent. This means that the conformational mobility of the AChE backbone allows
MHC to find more favorable binding position where it is better screened from water
environment. However, no such behavior is observed for CC.
Fig 8. Initial position after docking (left) and equilibrated position after MD (right) for CC ligand. Each
amino acid residue is shown in different color.
2
8

Fig 9. Initial position after docking (left) and equilibrated position after MD (right) for MHC ligand.
Each amino acid residue is shown in different color.
Comparison of the best docking structures of the ligands and their final equilibrated
conformations after MD simulations are shown in Figs 8 and 9. All-atom MD simulations in
water environment confirm that both CC and MHC bind strongly to the binding site of AChE
and remain stable there at the time scale of tens of nanoseconds. Equilibrated positions of both
ligands in MD simulations are different from the lowest energy docking poses. The RMSDs
between docking conformations and final MD structures are close to 0.8 nm for both ligands.
4
. Conclusion
The inhibitory efficacies of two coumarin derivatives towards AChE were checked following
Ellman’s protocol and fluorescence measurements in conjunction with molecular docking as well
as MD simulation results. Further, the modulatory effect of switching reaction medium from
simple aqueous buffer to physiologically compatible serum matrix was also investigated. All the
investigated systems were found to inhibit AChE in a non-competitive manner with CC showing
the highest inhibition potency. Sequestration of the inhibitors by HSA rendered the reduction in
AChE inhibition activity. Therefore, a potential AChEI is expected to have strong affinity to bind
with AChE, but with lower tendency to be confiscated in HSA. Interestingly, one of the
investigated systems (CC) showed greater inhibition potency even in presence HSA than
standard cholinergic drug DON in free aqueous medium. Docking as well as MD result reveals
that the binding of the coumarins with AChE to be strong and stable, which in turn could
attribute towards their superior inhibitory potencies. The compounds were also found to displace
ThT from AChE gorge, resulting in diminution of the emission intensity of ThT-AChE binary
system and certifying the application of a fluorescence assay. These conjectures envisage the
2
9