Dual inhibition of the α-glucosidase and butyrylcholinesterase studied by Molecular Field Topology Analysis

Article abstract of DOI:10.1016/j.ejmech.2014.04.018

A striking dual inhibition of enzymes α-glucosidase and butyrylcholinesterase by small drug-like molecules, including 1,4-disubstituted-1,2,3-triazoles, chalcones, and benzothiazepines, was rationalized with the help of Molecular Field Topology Analysis, a 3D QSAR technique similar to CoMFA. A common pharmacophore supported the concept of a link existing between type-2 diabetes mellitus and Alzheimer's disease. These findings will be instrumental for rational design of drug candidates for both of these conditions.

Full text of DOI:10.1016/j.ejmech.2014.04.018

Accepted Manuscript
Dual inhibition of the α-glucosidase and butyrylcholinesterase studied by Molecular
Field Topology Analysis
Farukh Jabeen, Polina V. Oliferenko, Alexander A. Oliferenko, Girinath G. Pillai,
Farzana Latif Ansari, C. Dennis Hall, Alan R. Katritzky
PII:
S0223-5234(14)00337-7
10.1016/j.ejmech.2014.04.018
EJMECH 6889
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 12 August 2013
Revised Date: 3 April 2014
Accepted Date: 4 April 2014
Please cite this article as: F. Jabeen, P.V. Oliferenko, A.A. Oliferenko, G.G. Pillai, F.L. Ansari,
C.D. Hall, A.R. Katritzky, Dual inhibition of the α-glucosidase and butyrylcholinesterase studied by
Molecular Field Topology Analysis, European Journal of Medicinal Chemistry (2014), doi: 10.1016/
j.ejmech.2014.04.018.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Dual inhibition of the α-glucosidase and butyrylcholinesterase studied by
Molecular Field Topology Analysis
Farukh Jabeena,b, Polina V. Oliferenkoa, Alexander A. Oliferenkoa, Girinath G. Pillaia,c, Farzana
Latif Ansarib, C. Dennis Hall,a* the late Alan R. Katritzkya*
aDepartment of Chemistry, University of Florida, PO Box 117200, Gainesville, Florida 32611,
ph (352) 392-0554, fax (352) 392-9199, e-mail: charlesdennishall@gmail.com, Katritzky@chem.ufl.edu
bDepartment of Chemistry, Quaid i Azam University, Islamabad, Pakistan 45320
cDepartment of Chemistry, University of Tartu, Tartu, Estonia 50411
Abstract
A striking dual inhibition of enzymes α-glucosidase and butyrylcholinesterase by small drug-like
molecules, including 1,4-disubstituted-1,2,3-triazoles, chalcones, and benzothiazepines, was
rationalized with the help of Molecular Field Topology Analysis, a 3D QSAR technique similar
to CoMFA. A common pharmacophore supported the concept of a link existing between type-2
diabetes mellitus and Alzheimer’s disease. These findings will be instrumental for rational
design of drug candidates for both of these conditions.
I. Introduction
α-Glucosidase is an exo-acting carbohydrase which catalyzes release of α-D-glucopyranose
from the non-reducing ends of various carbohydrate substrates [1-5]. The inhibition of its
catalytic activity leads to the retardation of glucose absorption and a decrease in postprandial
blood glucose level. Effective inhibitors of α-glucosidase can possess a significant effect on
carbohydrate digestion and post-translational processing of glycoproteins [6,9] and are used in
the treatment of a broad spectrum of diseases, including diabetes, cancer, HIV, hepatitis, obesity,
certain forms of dyslipidemia, and cardiovascular diseases [2-5,7-9].
1

ACCEPTED MANUSCRIPT
The mechanism of action of α-glucosidase inhibitors is based on their ability to bind
competitively to the carbohydrate-binding region of α-glucosidase enzymes [10,11]. Suppression
of these enzymes in the brush border membrane of the small intestine slows digestion of
carbohydrates, which consequently suppresses the rise of glucose, insulin and triglyceride in
blood [12-14]. Currently there are four α-glucosidase inhibitor drugs in the market: acarbose,
miglitol, and voglibose (Fig. 1). Acarbose is a pseudotetrasaccharide of microbial origin,
consisting of a maltose unit and a nitrogen-containing acarviosine unit. The latter is responsible
for high affinity to the α-glucosidase carbohydrate-binding site [14,15]. Recently it was found
that acarbose is efficient in patients with impaired glucose tolerance and could prevent or delay
the development of type 2 diabetes mellitus (T2D) [16]. Voglibose is another commonly
prescribed antidiabetic drug which is known to decrease postprandial glucose with a very low
risk of hypoglycemia [17]. Miglitol, a synthetically derived 1-deoxynojirimycin derivative, was
reported to be particularly efficient in prevention of β-cell insulin secretion dysfunction,
compared to other α-glucosidase inhibitors [18].
Insert Figure 1
Similar to acetylcholinesterase (AChE), butyrylcholinesterase (BChE) belongs to the
class of cholinesterase enzymes that hydrolyze acetylcholine (ACh) within cholinergic synapses
of the brain and autonomic nervous system. It shares 50-55% sequence identity with AChE [19] ,
regulates 10-15% of brain neurons [20], and is known to hydrolyze several choline and non-
choline esters. It is considerably less active in ACh hydrolysis than AChE at low concentrations
of the substrate but is highly efficient at higher levels of ACh, when AChE becomes substrate-
inhibited. Suppression of the cholinergic transmission in synapses is observed in severe
neurodegenerative disorders such as Alzheimer’s disease (AD). AD is characterized by a loss of
neurons caused by the formation of β-amyloid plaques and neurofibrillary tangles in brain nerve
cells. Simultaneous depletion of AChE and some increase of BChE activity shifts the balance of
ACh regulation [21]. Based on this observation, inhibitors of AChE able to reduce the rate of
ACh cleavage, different cholinesrease inhibitors e.g. tacrine, rivastigmine, galantamine,
donepezil have traditionally been used for symptomatic treatment of AD. Since BChE functions
as a co-regulator of AChE and can partly compensate the loss of AChE activity [22], it becomes
2

ACCEPTED MANUSCRIPT
responsible for hydrolysis of already depleted AChE in the AD brain. As a result, selective
inhibitors of BChE are suggested to provide more therapeutic effects in patients with more
severe form of disease [23].
Among compounds identified to inhibit BChE many are based on molecular scaffolds of
classical AChE inhibitors, e.g. cymserine analogues [22], galanthamine derivatives [21],
thienothiazines [24], diarylimidazoles [25], N-(10)-substituted phenothiazine amide derivatives
[26], 1,5-benzothiazepines and isosorbide-2-carbamate esters [27]; the latter being most potent
and highly selective so far (Fig.2).
Insert Figure 2
According to epidemiological studies, patients with T2D have a higher occurrence of dementia
and AD and vice versa. Active research has been performed over the last decade on
understanding the mechanisms of AD and T2D. Thus, experimental evidence was found that
impairment of insulin might be a mechanistic link between both conditions: not only T2D is
caused by insulin deficiency, but also insulin (and leptin) has been shown to regulate neuronal
and synaptic functions in brain [28]. At the same time, BChE may be involved in
parthenogenesis of T2D through suppression of amyloid formation [29]. It was demonstrated
also that selective inhibition of BChE in aged rats improved cognitive navigation [22].
QSAR. Computational chemistry tools such as Quantitative Structure-Activity
Relationship (QSAR), virtual screening, and molecular docking have been instrumental since the
very inception of rational drug design. QSAR analysis is one of the techniques used to determine
the structural properties related to the inhibitory activity of enzymes. The aim of QSAR analysis
is to investigate the correlation between activity, biological activity and the physicochemical
properties of a set of molecules [30,31].
QSAR is an integrated segment of chemoinformatics and its advances broaden the scope
of rational drug design [32]. It is unequivocal that chemical and pharmacological effects of
compounds are related to their physiochemical properties. Structure-activity analysis is the
foundation for understanding the structural features of both inhibitors and the target receptors
involved in a particular biological process and thus helps to design potent therapeutic moieties
against progression of AD and diabetes [33]. QSAR studies are a robust technique for the design
of bioactive compounds [34-36]. QSAR models are pointers to design effective drugs.
3

ACCEPTED MANUSCRIPT
Correlation of pharmacological properties with 3D description of a drug molecule, and the
identification of drug features that contribute to its potency, represents the central problem in the
field of drug design. Numerous methods for 3D QSAR analysis have been developed from the
early 1980's [37-38]. QSAR methodology is useful in predicting the activity of novel molecules
by mathematical equations which deduce the relationship(s) between a chemical structure and its
biological activity [39-44]. QSAR has been widely used for years to provide quantitative analysis
of structure and activity relationships of compounds. Statistical methods are applied in QSAR
modeling to establish correlations between chemical structures and their biological activities.
Once validated, the findings can be used to predict activities of untested compounds. Recently,
computer-assisted drug design based on QSAR has been successfully employed to develop new
drugs for the treatment of cancer, AIDS, SARS, and other diseases [45-47].
Synthesis of a library of different heterocyclic compounds and some chalcones has been
reported by one of us, with the members of which exhibited a reasonable in vitro activity against
either α-glucosidase or BChE or, in some case, both. It remained unclear what structural features
were responsible for such dual action. In this work, we investigate this effect computationally in
terms of the polypharmacology concept [48], which is being actively pursued nowadays. Multi-
target inhibitors, active against both α-glucosidase and BChE, will be beneficial in the
prevention and treatment of both T2D mellitus and AD, conditions that are frequently associated
with each other.
II. Materials and Methods
Chemistry
1,4-Disubstituted-1,2,3-triazoles 1-26 and 42 have been synthesized by a Cu-catalyzed azide
alkyne cycloaddition (CuAAC) by the reaction of an azide with an alkyne under click conditions
[49]. Phenyl styryl ketones (commonly known as chalcones) 30, 34-41 and 43-51 were
synthesized by a Claisen-Schmidt condensation of different substituted acetophenones and
benzaldehydes. Synthesis of 2,3-dihydro-1,5-benzothiazepines 27-29, 31-33, 52-59, 61-63 and
65 was carried out in a two-step process. The first step involved the synthesis of different
chalcones. The second step involved the reaction of the synthesized chalcones with o-
aminithiophenol under acid catalysis as reported by one of us earlier [50]. 2,3,4,5-
Tetrahydrobenzothiazepines 60 and 64 were synthesized by the reduction of their corresponding
4

ACCEPTED MANUSCRIPT
2,3-dihydrobenzothiazepines with lithium aluminium hydride. A diastereomeric mixture was
obtained in each which was separated through fractional crystallization. All synthesized
compounds were characterized by their XRD, spectroscopic and analytical data.
In vitro α-glucosidase inhibition assay
The enzyme inhibition assay s reported by one of us earlier is based on the breakdown of the
substrate to produce a colored product, followed by measuring the absorbance over a period of
time. Briefly, α-glucosidase (Sigma, type III, from yeast) was dissolved in buffer A (0.1 mol/L
potassium phosphate, 3.2 mmol/L-MgCl2, pH6.8) (0.1 units/ml) and, p-nitrophenyl-α-D-
glucopyranoside dissolved in buffer A at 6 mmol/L was used as substrate. 102 µl buffer B
(0.5mol/L potassium phosphate, 16mmol/L-MgCl2, pH 6.8), 120 µl sample solution (0.6 mg/ml
in DMSO), 282 µl water and 200 µl substrate were mixed. This mixture was incubated at 37 ˚C
for 5 minutes and then 200 µl of enzyme solution was added and mixed. The enzyme reaction
was carried out at 37 ˚C for 30 minutes followed by the addition of 1.2 ml of 0.4 mol/L glycine
buffer (pH, 10) to terminate the reaction. Enzyme activity was quantified by measuring
absorbance at 410 nm. Deoxynojirimycin hydrochloride (DNJ) was used as the standard inhibitor
[51].
In vitro BChE inhibition assay
The BChE inhibitory potential of the synthesized compounds was carried out as reported by one
of us earlier in 2005. Briefly, BChE enzyme was prepared in phosphate buffer (pH 8.0) having
0.2U per well. 180µL of 100 mM sodium phosphate buffer in each well of plate, 10 µL of
buffered Ellman’s Reagent (DTNB, 5-5- dithio-bis [2-nitrobenzoic acid] 0.1 M NaHCO3, 17.85
mmol/L), 10 µL of test compound solution (of various concentrations (5 µM- 500µM) was added
in each well labeled as test, 20 µL of enzyme solution was then added in each well, mixed and
incubated for 15 min at 25°C. The reaction was then initiated by the addition of 10 µL substrate
solution (BCh iodide,10 mM) in each well and after 10 min absorbance was measured at 412
nm by microtiter plate reader (Spectramax plus 384 Molecular Device, USA). The percentage
(%) of inhibition of BChE was calculated as follows: (E – S)/E 100, where E is the activity of the
enzyme without test compound and S is the activity of enzyme with test compound. The
concentration of the test compound that inhibited BChE activity by 50 % (IC50) was determined
5

ACCEPTED MANUSCRIPT
by means of EZ-Fit, Enzyme kinetics Program (Perrella Scientific In., Amhherset, USA) [52].
All experiments were carried out in triplicate using galanthamine as standard drug.
III. Methodology
Molecular Field Topology Analysis. In MFTA modeling, hundreds of descriptors are calculated
and analyzed using a dedicated statistical technique called PLS (partial least squares), which
reduces the dimensionality of the compound-descriptor matrix up to a few sensible components,
or factors [53]. This is the reason of using the number of factors (NF) in PLS jobs rather than a
number of descriptors. In MFTA (as well as in CoMFA and other PLS-based methods), the
factor is a linear combination of all descriptors selected. Analysis starts with the construction of
the so-called molecular supergraph, which is obtained by the superimposition of all data set
structures onto each other. Each vertex of this supergraph is assigned the values of descriptors
such as effective atomic charge, van der Waals (vdW), hydrogen bond donor (HBD) and
hydrogen bond acceptor (HBA) abilities, local lipophilicity, and others. The analysis of
descriptors impacts as color-shaded supergraph maps is a helpful tool for understanding the
structure-activity relationship and for search of new promising candidates.
Validation technique is also different compared to the Codessa-type multiple regression. Here we
use Q2 which is a cross-validation coefficient calculated as a result of extensive combinatorial
cv
process. Normally Q2 of PLS is smaller than R2 of multiple regression, but the former is
cv
cv
statistically more reliable.
IV. Results and discussion
Data set. The compounds studied include assorted heterocyclic compounds: 27 alkyl and phenyl
substituted triazoles, 20 benzothiazepines, 18 phenyl steryl ketones (chalcones). The structures
and activities in terms of IC50 values (the substance concentration at which enzyme loses 50% of
its activity) are summarized in (Table 1). α-Glucosidase and BChE inhibitory activities were
obtained for 42 and 65 compounds, respectively, of which 30 compounds had overlapping data.
The structure of the data used is visualized with the Venn diagram as shown in (Fig. 3).
Additional 13 structurally related diarylimidazoles, recently identified by Karlsson et al [25], as
potent BChE inhibitors were also added to the data pool for the purposes of QSAR studies;
details are shown in (Scheme 1). The IC50 values for α-glucosidase vary from 11.9 to 6756.7 µM
6

ACCEPTED MANUSCRIPT
while those for BChE lie between 3.97 and 585 µM. The highest bi-target activity was found for
two two compounds 27 and 28, with IC50 values equal to 12 µM for α-glucosidase and 14 µM
for BChE, respectively.
Insert Figure 3
Insert Table 1
Insert Scheme 1
MFTA
α-glucosidase model. MFTA model was obtained for the data set of 42 compounds
using -logIC50 as directly related to the energy scale values. The correlation coefficient and the
squared cross-validation coefficient dynamics are shown in (Fig. 4a). As can be seen, the highest
values for both parameters are achieved when seven factors are taken into account. At the same
time the rise of the correlation coefficient slows down starting from the sixth parameter. Because
of this, and to avoid the overtraining, we chose the five-factor model as the optimal one. The
correlation plot for the five-factor model is given in (Fig. 4b). The scale of the training data set
stretched for almost three logarithmic units, and the outline of the point’s distribution in the fit
plot is smooth and even. The squared correlation coefficient and the squared cross-validation
coefficients were 0.937 and 0.66, respectively. The average error was 0.133 log units.
Insert Figure 4, 5, 6
The molecular supergraph with some examples of superimposed molecules is shown in (Fig. 5).
The color representations of the five molecular fields or components are given in Fig. 6. Red and
blue colors on some vertices of the supergraph indicate the increase and decrease in the property
value with the increase of descriptor, respectively. (Table 2) summarizes descriptor
characteristics of different atom types populating some selected vertices. The compounds where
the atom types are present are also exemplified. For example, positions 65 and 64 are marked red
and blue on the atomic charge field, respectively. This means that a higher charge on position 65
and a lower charge on position 64 lead to an increase of -logIC50 value, thereby increasing the
target activity. In terms of chemical structure, these positions are occupied by the substituents
such as -OCH3 and -NH2 as opposed to -NO2 group. Methoxy group was found to be the most
favorable, since it carried the most negative charge on position 64 (ether oxygen) and most
positive charge on position 65 (methyl group). In addition, Osp3 and Nsp3 are good HBAs and
their presence contributes to the target property since position 64 in the hydrogen bond acceptor
7

ACCEPTED MANUSCRIPT
field is highlighted red. As a result, compound 35 has the highest measured inhibition activity
among all three compounds. Another highly informative position was 8, which was found to be
populated by Nsp3, Csp3, Osp2, and Csp2 atoms. According to the color patterns, the presence of
Csp3 and Osp2 atoms in this position plays an important role in terms of charge and lipophilicity
for benzothiazepines and in terms of vdW radii and hydrogen bond acceptor potency for phenyl
styryl ketones, respectively. Following the same logic, a large polarizable atom possessing some
hydrogen bond donor ability fits better for position 53; sulfhydryl and sulfamide groups can be
good candidates for this position. Similar tendency was detected for the three-vertex group
34/41/49, where hydrophobic and polarizable chloro- and cyano- substituents in a benzene ring
enhance the target property compared to the acetyl and nitro groups.
Insert Table 2
MFTA BChE model. Modeling of 66 data points for BChE (53 originally measured and 13
from ref. 25) was done similar to that of α-glucosidase. As can be seen from the factor dynamic
plot in (Fig. 7a), the best statistics is obtained for the six-factor model. The corresponding fit plot
is shown in (Fig. 7b). The statistical quality of the model is as good as that for BChE with square
correlation and cross-correlation coefficients equal to 0.948 and 0.937, respectively, and an
average error of 0.174 log units. Molecular supergraph, shown in (Fig. 8), is similar to the one
shown in (Fig. 5) due to a considerable overlap of BChE and
α-glucosidase data sets.
Insert Figures 7, 8, 9
The most eye-catching features in descriptors representations as shown in Fig. 9 are: (i) the
unfavorable in terms of vdW radius alkyl chain with more than four atoms; (ii) a positive effect
from a strong electronegative HBA substituent in position 66; (iii) and a highly undesirable, in
terms of lipophilicity, the presence of a -NO2 group in positions 35-36-37 opposite to
substituents e.g. -Cl, -OCH3, and -SCH3.
The analysis of the supergraphs and the color schemes of (Fig. 6 and Fig. 9) enable one to
identify main structural characteristics responsible for the dual inhibition of both α-glucosidase
and BChE. First of all, it is the presence of two aromatic rings separated by a link of three atoms.
This link could be flexible, as in phenyl styryl ketones, or sterically hindered by incorporation of
8

ACCEPTED MANUSCRIPT
a ring, as in triazoles, diazoles, and benzothiazepines. The second feature is the beneficial
influence of the polarizable and electronegative substituents in the aryl rings such as -Cl, -Br, -
SCH3, and -OCH3. A combination of these two characteristics in a molecule is expected to
enhance its inhibitory effect for both molecular targets. The common pharmacophore structure
embodying these features is shown in (Fig. 10). Synthesis and testing of a set of compounds
which meet these requirements will be a second step in working out more effective inhibitors
with dual α-glucosidase and BChE inhibition activity.
Insert Figure 10
Conclusions
A library of synthesized heterocyclic compounds and some chalcones were tested for their
potential as α-glucosidase and BChE inhibitors and a striking dual inhibition of both enzymes
was found for some of them. This dual inhibitors subset predominantly consisted of 1,4-
disubstituted-1,2,3- triazoles, whose specific structural features responsible for the
polypharmacological activity were identified by Molecular Field Topology Analysis. As T2D
mellitus is a risk factor to AD, dual mode drugs acting on both of them are highly promising. The
quantitative structure-activity relationships and the common pharmacophore pattern identified in
this work will help design better drug candidates to counteract those two debilitating conditions.
Acknowledgements
Part of the work presented was supported by the EU European Regional Development Fund
through the Center of Excellence in Chemical Biology, Estonia and by targeted financing from
the Estonian Ministry of Education and Research (SF0140031As09).
References
[1] Y-P. Zhu, L-J. Yin, Y-Q. Cheng, K. Yamaki, Y. Mori, Y-C. Su, L-T. Li, Effects of sources
of carbon and nitrogen on production of α-glucosidase inhibitor by a newly isolated strain
of Bacillus subtilis B2, Food Chem. 109 (2008) 737–742.
9

ACCEPTED MANUSCRIPT
[2] Z.Y. Du, R.R. Liu, W.Y. Shao, X.P. Mao, L. Ma, L.Q. Gu, Z.S. Huang, A.S. Chan, Alpha-
glucosidase inhibition of natural curcuminoids and curcumin analogues, Eur. J. Med.
Chem. 14 (2006) 213–218.
[3] Y. Liu, L. Zou, L. Ma, W.H. Chen, B. Wang, Z.L. Xu, Synthesis and pharmacological
activities of xanthone derivatives as alpha-glucosidase inhibitors, Bioorg. Med. Chem. 14
(2006) 5683–5690.
[4] K. Tatsuta, Total synthesis and chemical design of useful glycosidase inhibitors, Pure &
Appl. Chem. 68 (1996) 1341-1346.
[5] W.F. Caspary, Sucrose malabsorption in man after ingestion of alpha-glucosidehydrolase
inhibitor, Lancet, 1 (1978) 1231–1233.
[6] D.K. McCulloch, A.B. Kurtz, R.B. Tattersall, A new approach to the treatment of nocturnal
hypoglycemia using α-glucosidase inhibition, Diabetes Care, 6 (1983) 483-487.
[7] J-H. Park, S. Ko, H. Park, Toward the virtual screening of α-glucosidase inhibitors with the
homology-modeled protein structure, Bull. Korean Chem. Soc. 29 (2008) 921-927.
[8] P.B. Fischer, G.B. Karlsson, T.D. Butters, R.A. Dwek, F.M. Platt, N-
butyldeoxynojirimycin-mediated inhibition of human immunodeficiency virus entry
correlates with changes in antibody recognition of the V1/V2 region of gp120, J. Virol. 70
(1996) 7143–7152.
[9] P.B. Fischer, G.B. Karlsson, R.A. Dwek, F.M. Platt, N-butyldeoxynojirimycin-mediated
inhibition of human immunodeficiency virus entry correlates with impaired gp120
shedding and gp41 exposure, J. Virol. 70 (1996) 7153–7160.
[10] B. Brazdova, N.S. Tan, N.M. Samoshina, V.V. Samoshin, Novel easily accessible
glucosidase
inhibitors:
4-hydroxy-5-alkoxy-1,2-cyclohexanedicarboxylic
acids,
Carbohydrate Res. 344 (2008) 311-321.
[11] K. Kraim, D. Khatmi, Y. Saihi, F. Ferkous, M. Brahimi, Quantitative structure activity
relationship for the computational prediction of α-glucosidase inhibitory, Chemometrics
and Intelligent Laboratory Systems, 97 (2009) 118–126.
[12] L. Xiancui, N. Rongli, F. Xiao, H. Lijun, Z. Lixin, Macroalgae as a source of α-glucosidase
inhibitors, Chin. J. Oceanol. Limnol. 23 (2005) 354-356.
10

ACCEPTED MANUSCRIPT
[13] S.D. Kim, H.J. Nho, Isolation and Characterization of α-glucosidase inhibitor from the
fungus ganoderma lucidum, J. Microbiol. 09 (2004) 223-227.
[14] H.E. Lebovitz, Alpha-glucosidase inhibitors, Endocrinol. Metab. Clin. North Am. 26
(1997) 539-551.
[15] R. Rabasa-Lhoret, J-L. Chiasson, Potential of α-glucosidase inhibitors in elderly patients
with diabetes mellitus and impaired glucose tolerance, Drug & Aging, 13 (1998) 131-143.
[16] F.A. Van de Laar, P.L.B.J. Lucassen, R.P. Akkermans, E.H. Van de Lisdonk, G.E.H.M.
Rutten, C. Van Weel, Alpha-glucosidase inhibitors for type 2 diabetes mellitus, Cochrain
Review, The Cochrain Library, John Wiley & Sons, Ltd. 2009.
[17] Y. Iwamoto, A. Kashiwagi, N. Yamada, S. Terao, N. Mimori, M. Suzuki, H. Tachibana,
Efficacy and safety of vildaglipin and voglibose in Japanese patients with type 2 diabetes: a
12-week, randomized, double-blind, active-controlled study, Diabetes Obes. Metab. 12
(2010) 700-708.
[18] N. Fakaya, K. Mochizuki, Y. Tanaka, T. Kumazawa, Z. Jiuxin, M. Fuchigami, T. Goda,
The α-glucosidase inhibitor miglitol delays the development of diabetes and dysfunctional
insulin secretion in pancreatic β-cells in OLETF rats, Eur. J. Pharm. 624 (2009) 51-57.
[19] G.R. Sridhar, T. Sekhar, P.V.N. Rao, A.A. Rao, Human butyrylcholinesterase knock-out
equivalent : potential to assess role in Alzheimer’s disease, Adv. Alz. Dis. 1 (2012) 1-11.
[20] S. Darvesh, D.L. Grantham, D.A. Hopkins, Distribution of cholinesterase in the human
amygdala and hippocampal formation, J. Compar. Neur. 393 (1998) 374-390.
[21] E. Giacobini, Selective inhibitors of butyrylcholinesterase, Drugs Aging 18 (2001) 891-
898.
[22] N.H. Greig, T. Utsuki, D.K. Ingram, Y. Wang, G. Pepeu, C. Scali, Q.S. Yu, et al. Selective
butyrylcholinesterase inhibition elevates brain acetylcholine, augments learning and lowers
Alzheimer β-amyloid peptide in rodent, PNAS, 102 (2005) 17213-17218.
[23] M. Decker, B. Kraus, J. Heilmann, Design, synthesis and pharmacological evaluation of
hybrid molecules out of quinazolinimines and lipoic acid lead to highly potent and selective
butyrylcholinesterase inhibitors with antioxidant properties, Bioorg. Med. Chem. 16 (2008)
4252-4261.
11

ACCEPTED MANUSCRIPT
[24] D. Karlsson, A. Fallarero, G. Brunhofer, C. Mayer, O. Prakash, C.G. Mohan, P. Vuorela, T.
Erker, The exploration of thienothiazines as selective butyrylcholinesterase inhibitors, Eur.
J. Pharm. Sci. 47 (2012) 190-205.
[25] D. Karlsson, A. Fallarero, G. Brunhofer, P. Guzik, M. Prinz, U. Holzbrabe, T. Erker, P.
Vuorela, Identification and characterization of diarylimidazoles as hybrid inhibitors of
butyrylcholinesterase and amyloid beta fibril formation, Eur. J. Parm. Sci. 45 (2012) 169-
183.
[26] S. Darvesh, R.S. McDonald, K.V. Darvesh, D. Mataija, S. Conrad, G. Gomez, R. Walsh, E.
Martin, Selective reversible inhibition of human butyrylcholinesterase by aryl amide
derivatives of phenothiazine, Bioorg. Med. Chem. 15 (2007) 6367-6378.
[27] C.G. Carolan, G.P. Dillon, J.M. Gaynor, S. Reidy, S.A. Ryder, D. Khan, J.F. Marquez, J.F.
Gilmer, Isosorbide-2-carbamate esters: potent and selective butyrylcholiesterase inhibitors,
J. Med. Chem. 51 (2008) 6400-6409.
[28] W. Han, C. Li, Linking type 2 diabetes and Alzheimer’s disease, PNAS, 107 (2010) 6557-
6558.
[29] A.R. Allam, G.R. Sridhar, H. Thota, C.S. Babu, A.S. Prasad, C. Divakar, Alzheimer’s
disease and type 2 diabetes mellitus: the cholinesterase connection? Lipids Health Dis. 5
(2006) 28.
[30] J. Xu, S. Huang, H. Luo, G. Li, J. Bao, S. Cai,; Y. Wang, QSAR Studies on
andrographolide derivatives as α-glucosidase inhibitors, Int. J. Mol. Sci. 11 (2010) 880-
895.
[31] R. Suganya, P.S.K. Lilly, M. Saleena, Pharmacophore Modeling, Atom based 3D-QSAR
and Docking Studies on ADAMTS-5 Inhibitors, 2011 International Conference on
Bioscience, Biochemistry and Bioinformatics IPCBEE, IACSIT Press, Singapore. 5 (2011).
[32] M.B. Workalemahu, G.G. Pillai, A.A. Oliferenko, A.R. Katritzky, Quantitative structure-
activity relationships: the ubiquitous links between cause and effect, ChemPlusChem, 77
(2012) 507-517.
[33] H. Zaheer Ul-Haq, R. Uddin, H. Yuan, P.A. Petukhov, M.I. Choudhary, J.D. Madura,
Receptor-based modeling and 3D-QSAR for a quantitative production of the
12

ACCEPTED MANUSCRIPT
butyrylcholinesterase inhibitors based on genetic algorithm, J. Chem. Inf. Model. 48
(2008) 1092–1103.
[34] M. Karelson, Molecular Descriptors in QSAR/QSPR. Wiley-Interscience, 2000.
[35] M. Fernández, J. Caballero, A.M. Helguera, E.A. Castro, M.P. González, Quantitative
structure-activity relationship to predict differential inhibition of aldose reductase by
flavonoid compounds, Bioorg. Med. Chem. 13 (2005) 3269–3277.
[36] L. Saiz-Urra, M.P. Gonzalez, Y. Fall, G. Gomez, Quantitative structure-activity
relationship studies of HIV-1 integrase inhibition, Eur. J. Med. Chem. 42 (2007) 64–70.
[37] M.D.M. AbdulHameed, J. Liu, Y. Pan, L. Fang, C. Silva-Rivera, C-G. Zhan, Microscopic
binding of butyrylcholinesterase with quinazolinimine derivatives and the structure–
activity correlation, Theor. Chem. Acc. 130 (2011) 69–82.
[38] S.D. de Souza, A.M. de Sousa, A.C. de Souza, A.C. Sodero, L.M. Cabral, M.G.
Albuquerque, H.C. Castro, C.R. Rodrigues, Hologram QSAR models of 4-
[(diethylamino)methyl]-phenol inhibitors of acetyl/butyrylcholinesterase enzymes as
potential anti-Alzheimer agents, Molecules, 17 (2012) 9529-9539.
[39] Zaheer Ul-Haq, U. Mahmood, B. Jehangir, Ligand-based 3D-QSAR studies of
physostigmine analogues as acetylcholinesterase inhibitors, Chem. Biol. Drug Des. 74
(2009) 571-581.
[40] A. Duardo-Sanchez, G. Patlewicz, A. Lopez-Diaz, Current topics on software use in
medicinal chemistry: Intellectual property, taxes, and regulatory issues, Curr. Top. Med.
Chem., 8 (2008) 1666-1675.
[41] H. Gonzalez-Diaz, S. Vilar, L. Santana, E. Uriarte, Medicinal chemistry and bioinformatics
- current trends in drugs discovery with networks topological indices, Curr. Top. Med.
Chem. 7 (2007) 1015-1029.
[42] H. Gonzalez-Diaz, F. Prado-Prafo, F.M. Uberia, Predicting antimicrobial drugs and targets
with the MARCH-INSIDE approach, Curr. Top. Med. Chem. 8 (2008) 1676-1690.
[43] J. Caballero, M. Fernandez, Artificial neural networks from MATLAB in medicinal
chemistry. Bayesian-regularized generic neural networks (BRGNN): application to the
prediction of the anatagonistic actitivity against human platelet thrombin receptor (PAR-1),
Curr. Top. Med. Chem. 8 (2008) 1580-1605.
13

ACCEPTED MANUSCRIPT
[44] M.P. Gonzalez, C. Teran, L. Saiz-Urra, M. Teijeria, Variable selection methods in QSAR:
An overview, Curr. Top. Med. Chem. 8 (2008) 1606-1627.
[45] S. Vilar, G. Cozza, S. Moro, Medicinal chemistry and the molecular operating environment
(MOE): Application of QSAR and molecular docking to drug discovery, Curr. Top Med.
Chem. 8 (2008) 1555-1572.
[46] N.S.H.N. Moorthy, M.J. Ramos, P.A. Fernandes, Structural analysis of α-glucosidase
inhibitors by validated QSAR models using topological and hydrophobicity based
descriptors, Chemometr. Intell. Lab. 109 (2011) 101-112.
[47] K. Kraim, D. Khatmi, Y. Saihi, F. Ferkous, M. Brahimi, Quantitative structure activity
relationship for the computational prediction of α-glucosidase inhibitory. Chemometr.
Intell. Lab. 97 (2009) 118–126.
[48] L. Xie, L. Xie, S.L. Kinnings, P.E. Bourne, Novel Computational Approaches to
Polypharmacology as a Means to Define Responses to Individual Drugs, Ann. Rev. Pharm.
Toxicol. 52(2012) 361-379.
[49] F. Jabeen, M.Q. Fatmi, S.A. Shehzadi, S. Shaheen, L. Iqbal, N. Afza, F.L. Ansari,
Regioselective synthesis of 1,4-disubstituted 1,2,3-triazoles as novel α-glucosidase
inhibitors and probing ligand-protein interactions through docking studies, Eur. J. Med.
Chem., submitted
[50] F.L. Ansari, S. Umbreen, L. Hussain, T. Makhmoor, S.A. Nawaz, M.A. Lodhi, S.N. Khan,
F. Shaheen, M.I. Choudhary, Atta-ur-Rehmam, Syntheses and biological activities of
chalcone and 1,5-benzothiazepine derivatives: promising new free-radical scavengers, and
esterase, urease, and alpha-glucosidase inhibitors, Chem. Biodivers. 2 (2005) 487–496
[51] S. Farheen, L. Iqbal, A. Rehman, N. Afza, A. Malik, M.A. Rasool, M. Irfan, R.B. Tareen,
Galinsoside A and B, Bioactive flavanone glucosides from Glinsoga parviflora, J. Enz. Inh.
Med. Chem. 24 (2008) 1128–1132.
[52] Z. Ahmed, S. Mehmood, R. Ifzal, A. Malik, N. Afza, L. Iqbal, Cholinesterase inhibitory
triterpens from Salsola bryosma, Polish J. Chem, 81 (2007) 1427-1432.
[53] V. A. Palyuli; E. V. Radchenko, N. S. Zefirov, Molecular Field Topology Analysis Method
in QSAR Studies of Organic Compounds, J. Chem. Inf. Comput. Sci. 40 (2000) 659-667.
14

ACCEPTED MANUSCRIPT
Figure 1. Examples of common α-glucosidase inhibitors.
Figure 2. Examples of common butyrylcholinesterase inhibitors.
Figure 3. Venn diagram showing the ovelap of the experimental data for α-glucosidase and
BChE inhibitors. The 30 dual mode inhibitors lie in the intersection area.
15

ACCEPTED MANUSCRIPT
Figure 4. (a) Factor dynamics and (b) correlation fit plot obtained for α-glucosidase inhibitory
activity model.
Figure 5. Molecular supergraph for the
α-glucosidase MFTA model and examples of
superimpositions of molecules on the MSG (below). A unique number is assigned to each vertex.
16

ACCEPTED MANUSCRIPT
Figure 6. Color schemes of the molecular fields for the glucosidase inhibition model.
Figure 7. Factor dynamics (a) and fit plot (b) for the BChE inhibitory activity model.
17

ACCEPTED MANUSCRIPT
Figure 8. Molecular supergraph for the BChE MFTA model.
Figure 9. Color schemes of the molecular fields for the BChE model.
18

ACCEPTED MANUSCRIPT
Figure 10. Common pharmacophore of dual inhibitors of α-glucosidase and BChE.
Scheme 1. Diarylimidazoles as potent BChE inhibitors [25].
IC50
(µM)
3.6
Entry R1
R2
-logIC50
2-OCH3
Ar
Ar
Ar
-0.556
-0.176
0.699
1a
1b
1c
3-OCH3
1.5
0.2
2,5-(OCH3)2
3-
thienyl
2-SC2H5
0.1
1.000
6c
2,3-(OCH3)2
3-OC2H5
Ar
12
-1.079
-0.253
0.167
0.066
9a
9b
9c
9d
Ar
Ar
Ar
1.8
3-OC3H7
3-OC4H9
3-
0.68
0.86
Ar
0.98
0.009
9e
OCH2CH(CH3)2
2-OC2H5
Ar
Ar
Ar
Ar
3.4
-0.531
0.745
-0.505
0.796
9f
3-SCH3
2-SCH3
3-SC2H5
0.18
3.2
0.16
9g
9h
9i
19

ACCEPTED MANUSCRIPT
Table 1. Structures and experimental α-glucosidase and BChE inhibitory activities (µM) of 42
heterocyclic compounds.
BChE (Exp.)
α-Glucosidase (Exp.)
IC50
Sl. No.
1
Structure
-logIC50 IC50
-logIC50.
1190.4
-3.076
-2.568
500.0
430.0
-2.699
-2.633
370.0
862.0
2
3
-2.936
585.0
-2.767
57.2
-1.757
-2.612
41.3
-1.616
-2.574
4
5
409.0
375.3
520.0
-2.716
499.0
-2.698
6
20

ACCEPTED MANUSCRIPT
235.3
49.7
-2.372
-1.696
250.0
42.4
-2.398
-1.627
7
8
HO
N
N
N
N
O2N
CN
HO
O
1086.0
657.8
-3.036
-2.818
248.0
43.6
-2.394
-1.639
9
N
O
N
HO
10
N
N
N
C8H17
1386.0
812.5
320.0
-3.142
-2.910
-2.505
294.0
568.0
409.0
-2.468
-2.754
-2.612
11
12
13
HO
O
O
N
N
N
21

ACCEPTED MANUSCRIPT
HO
366.0
-2.563
65.5
-1.816
14
N
N
N
N
Cl
HO
O
320.0
1388.0
325.0
245.0
265.0
313.0
-2.505
-3.142
-2.512
-2.389
-2.423
-2.496
452.0
512.0
297.6
83.3
-2.655
-2.709
-2.474
-1.921
-2.654
-2.470
15
16
17
18
19
20
N
N
HO
Cl
N
N
N
HO
O
N
N
N
451.0
295.0
22

ACCEPTED MANUSCRIPT
HO
1712.0
-3.234
333.3
-2.523
21
N
N
N
NO2
HO
255.0
520.0
-2.407
-2.716
362.3
384.6
-2.559
-2.585
22
23
N
N
N
O
O
378.0
370.0
-2.577
-2.568
555.5
454.5
-2.745
-2.658
24
25
1562.2
12.0
-3.194
-1.079
450.3
14.0
-2.654
-1.146
26
27
23

ACCEPTED MANUSCRIPT
12.0
-1.079
14.0
-1.146
28
S
N
OH
12.8
-1.107
-2.179
62.0
32.0
-1.792
-1.505
29
30
151.0
31
11.9
-1.076
-
-
S
S
N
N
OH
32
33
18.0
33.0
-1.255
-1.519
-
-
-
-
Cl
34
35
15.0
20.2
-1.176
-1.305
-
-
-
-
24

ACCEPTED MANUSCRIPT
O
36
37
49.4
93.0
-1.694
-1.968
-
-
-
-
HO
O
OH
O
H2N
38
130.0
-2.114
-
-
39
40
41
151.0
1870
3850
-2.179
-2.272
-2.585
-
-
-
-
-
-
42
6756.7
-3.830
-
-
N
N
N
NO2
43
44
45
-
-
-
-
-
-
13.0
26.0
27.0
-1.114
-1.415
-1.431
25

ACCEPTED MANUSCRIPT
46
47
-
-
-
-
32.0
42.0
-1.505
-1.623
48
49
50
51
-
-
-
-
-
-
-
-
87.0
-1.940
-1.924
-2.021
-2.076
84.0
105.0
119.0
52
-
-
3.97
-0.599
53
54
-
-
-
-
3.97
4.7
-0.599
-0.672
26

ACCEPTED MANUSCRIPT
55
56
-
-
-
-
4.65
9.1
-0.667
-0.959
S
N
57
58
59
-
-
-
-
-
-
19.7
29.4
34.4
-1.294
-1.468
-1.537
HO
60
61
-
-
-
-
57.0
60.0
-1.756
-1.778
27

ACCEPTED MANUSCRIPT
S
N
62
63
-
-
-
-
75.2
76.6
-1.876
-1.884
HO
64
65
-
-
-
-
187.0
142.1
-2.272
-2.153
e-serine
-
-
0.93
0.032
Deoxynojirimycin
3.49
-0.543
-
-
O
O
Galanthamine
8.00
-0.903
-
-
N
HO
28