N-Substituted pyrimidinethione and acetophenone derivatives as a new therapeutic approach in diabetes

Article abstract of DOI:10.1002/ardp.202000075

In this study, compounds with 4-hydroxybutyl, 4-phenyl, 5-carboxylate, and pyrimidine moieties were determined as α-glycosidase inhibitors. N-Substituted pyrimidinethione and acetophenone derivatives (A1–A5, B1–B11, and C1–C11) were good inhibitors of the α-glycosidase enzyme, with K_i values in the range of 104.27 ± 15.75 to 1,004.25 ± 100.43 nM. Among them, compound B7 was recorded as the best inhibitor, with a K_i of 104.27 ± 15.75 nM against α-glycosidase. In silico studies were carried out to clarify the binding affinity and interaction mode of the compounds with the best inhibition score against α-glycosidase from Saccharomyces cerevisiae. Compounds B7 (S) and B11 (R) exhibited a good binding affinity with docking scores of ?8.608 and 8.582 kcal/mol, respectively. The docking results also showed that the 4-hydroxybutyl and pyrimidinethione moieties play a key role in S. cerevisiae and human α-glycosidase inhibition.

Full text of DOI:10.1002/ardp.202000075

Received: 10 March 2020
Revised: 17 May 2020
Accepted: 22 May 2020
|
|
DOI: 10.1002/ardp.202000075
F U L L P A P E R
N‐Substituted pyrimidinethione and acetophenone
derivatives as a new therapeutic approach in diabetes
Parham Taslimi1
| Afsun Sujayev2 | Muhammet Karaman3
|
Gunel Maharramova2 | Nastaran Sadeghian4 | Sabiya Osmanova2
|
Sabira Sardarova2 | Nargiz Majdi2 | Handan U. Ozel5 | İlhami Gulcin4
1Department of Biotechnology, Faculty of
Science, Bartin University, Bartin, Turkey
Abstract
2Laboratory of Theoretical Bases of Synthesis
In this study, compounds with 4‐hydroxybutyl, 4‐phenyl, 5‐carboxylate, and pyr-
and Action Mechanism of Additives, Institute
of Chemistry of Additives, Azerbaijan National
Academy of Sciences, Baku, Azerbaijan
imidine moieties were determined as α‐glycosidase inhibitors. N‐Substituted pyr-
imidinethione and acetophenone derivatives (A1–A5, B1–B11, and C1–C11) were
good inhibitors of the α‐glycosidase enzyme, with Ki values in the range of
104.27 15.75 to 1,004.25 100.43 nM. Among them, compound B7 was recorded
as the best inhibitor, with a Ki of 104.27 15.75 nM against α‐glycosidase. In silico
studies were carried out to clarify the binding affinity and interaction mode of the
compounds with the best inhibition score against α‐glycosidase from Saccharomyces
cerevisiae. Compounds B7 (S) and B11 (R) exhibited a good binding affinity with
docking scores of −8.608 and 8.582 kcal/mol, respectively. The docking results also
showed that the 4‐hydroxybutyl and pyrimidinethione moieties play a key role in
S. cerevisiae and human α‐glycosidase inhibition.
3Department of Molecular Biology and
Genetics, Faculty of Arts and Science, Kilis 7
Aralik University, Kilis, Turkey
4Department of Chemistry, Faculty of
Sciences, Ataturk University, Erzurum, Turkey
5Department of Environmental Engineering,
Faculty of Engineering, Bartin University,
Bartin, Turkey
Correspondence
Parham Taslimi, Department of Biotechnology,
Faculty of Science, Bartin University, 74100
Bartin, Turkey.
Email: parham_taslimi_un@yahoo.com;
ptaslimi@bartin.edu.tr
K E Y W O R D S
acetophenone, diabetes mellitus, molecular docking, N‐substituted pyrimidinethione,
α‐glycosidase
|
1
INTRODUCTION
acids are used as plasticizers of industrial and intermediate products
in the synthesis of dyes, as well as sulfa drugs in medicine.[9]
Heterocyclic compounds are the main class combinations with a large
spectrum of pharmacological and biological activities.[1] Recently, it
has been observed that dihydropyrimidines with functional values
are the compounds that have biological activities including antiviral
and antibacterial properties.[2,3] However, while analyzing the most
recent literature, it becomes clear that the formulation of hetero-
cyclic compound of thiourea by three‐component condensation re-
actions mainly covers the synthesis of dihydropyrimidinethiones.[4,5]
The reaction products of aldehydes and ketones with thioglycolic
acid are compounds containing functional groups such as carboxyl
and hydroxyl groups, which are receiving great attention. These
compounds can be valuable synthons for the synthesis of different
esters,[6,7] amides, and heterocyclic compounds.[8] Amides of organic
α‐Glycosidase is a digestive enzyme that hydrolyzes carbohy-
drate molecules including disaccharides and starches to generate
more metabolically available sugars in the course of catabolic me-
tabolism.[10,11] Indeed, it can functionally hydrolyze carbohydrate
molecules. The α‐glycosidase enzyme is distinct from β‐glycosidase,
which cleaves glycosidic bonds. It is generally well recorded that α‐
glycosidase is associated with type 2 diabetes mellitus (T2DM) due to
the fact that the high activity of this enzyme enhances plasma glu-
cose levels, which, in turn, affect glucose absorption in T2DM.[12]
Thus, several studies regarding α‐glycosidase inhibition and the de-
velopment of its inhibitors have been reported and also conducted
due to interest in the treatment of T2DM via downregulation of α‐
glycosidase activity (Figure 1). Thus, various α‐glycosidase inhibitors
|
Arch Pharm. 2020;e2000075.
wileyonlinelibrary.com/journal/ardp
© 2020 Deutsche Pharmazeutische Gesellschaft
1 of 11
https://doi.org/10.1002/ardp.202000075

2 of 11
TASLIMI ET AL.
|
FIGURE 1 The mechanism of action of
α‐glycosidase enzyme inhibitors
|
such as acarbose, voglibose, and miglitol have been highlighted and
mostly addressed in the clinical setting.[13,14]
2
RESULTS AND DISCUSSION
|
Chemistry
This study aimed at investigating the α‐glycosidase enzyme in-
hibition of N‐substituted pyrimidinethione and acetophenone deri-
vatives (A1–A5, B1–B11, and C1–C11) on α‐glycosidase enzyme.
Also, the study aimed to understand a possible inhibition mechanism
of the most active compounds against the enzyme using molecular
docking simulation.
2.1
First, in the presence of trifluoroacetic acid (TFA), an efficient method for
the synthesis of tetra(hexa)hydropyrimidinethione‐carboxylates (A1–A5)
has been used on the basis of three‐component condensation of thiourea
with different aldehydes and β‐diketones derivatives[15] (Scheme 1).
A4.
p-
SCHEME 1 The synthesis of cyclic thioureas (A1–A5)

3 of 11
TASLIMI ET AL.
|
|
Enzyme results
Various aldehydes and active methylene compounds with
2.2
thiourea have been elaborated in the effective method of synthesis
of dihydropyrimidinethiones, and their conversions have been carried
out. Nowadays, many methods have been reported for the prepara-
tion of tetra(hexa)hydropyrimidinethiones. Taking this into con-
sideration, tetrahydropyrimidinethiones were synthesized by
Biginelli reaction. The nucleophilic displacement reaction occurred
between 1,2‐epoxypropane and 1,2‐epoxybutane, and pyr-
imidinethiones and new compounds were obtained (B1–B6). By this
known method, the mutual influence reaction was conducted be-
tween tetrahydropyrimidinethiones and 4‐chlorobutanol‐1 as a nu-
cleophile. Triethylamine is used as a catalyst. As a result of the
reaction, compounds B7–B11 were synthesized with 65–70%
yield[16] (Scheme 2).
For α‐glycosidase, N‐substituted pyrimidinethione and acetophenone
derivatives (A1–A5, B1–B11, and C1–C11) have IC50 values in the
range of 94.16–1,005.24 nM and Ki values in the range of
104.27 15.75 to 1,004.25 100.43 nM (Table 1 and Figure 2). The
results clearly show that all of the derivatives have inhibitory effects
on α‐glycosidase as efficient as acarbose (IC50: 22.8 µM), a standard
α‐glycosidase inhibitor.[18] In fact, the most effective Ki values were
observed for B7 and B11, with Ki values of 104.27 15.75 and
236.32 24.74 nM, respectively. However, the weakest inhibition
was obtained with compounds A5 and A4, with Ki values of
1,004.25 100.43 and 989.95 135.21 nM, respectively. Group B
showed better results than Groups A and C. Compounds B1–B11
have IC50 values in the range of 94.16–504.28 nM and Ki values of
104.27 15.75 to 605.23 93.08 nM. It was known that heterocyclic
compounds have α‐glycosidase inhibition profiles. In this context,
Natori and colleagues found that 1‐C‐butyl‐L‐arabino‐iminofuranose
had potent inhibitory effects against α‐glycosidase enzymes.[19] In a
recent study, it was reported that a novel Ag‐N‐heterocyclic carbene
complex had α‐glycosidase inhibition properties.[20] Some inhibitors
like acarbose, miglitol, and voglibose are commercial α‐glycosidase
inhibitors, which are considered as the first‐line therapy for diabetic
individuals with postprandial hyperglycemia. Although these com-
pounds are effective in attenuating the rise in the blood sugar level in
The thiolation reaction was performed in a benzene solution at
80°C with p‐substituted ketones and mercaptoacetic acid in a molar
ratio of 1:4 in the presence of a catalytic amount of toluene sulfonic
acids. According to the reaction of mutilation of acetophenone and
some of its p‐substituted derivatives 1,1‐bis‐(carboxymethylthio)‐1‐
arylethanes were obtained. It was found that during the reaction, 1‐
hydroxy‐1‐(carboxymethylthio)‐1‐arylethanes spontaneously cy-
clized to form 2‐aryl‐2‐methyl‐1,3‐oxathiolane‐5‐ones. A direct ami-
dation reaction of 1,1‐bis‐(carboxymethylthio)‐1‐arylethanes with
primary amines was conducted[17] (Scheme 3).
SCHEME 2 The synthesis of derivatives of new pyrimidinethiones (B1–B11)

4 of 11
TASLIMI ET AL.
|
SCHEME 3 The synthesis of amides of 1,1‐bis‐(carboxymethylthio)‐1‐arylethanes (C1–C11)
patients, unfortunately, their continuous use may lead to undesirable
side effects, for example, adverse gastrointestinal symptoms and liver
toxicity.[21,22]
compounds. After QikProb calculation, we considered the number of
reactive functional groups, molecular weight, hydrogen bond donor
and acceptor, octanol/water partition coefficient, IC50 value for
blockage of HERG K+ channels, Caco‐2 cell permeability, brain/blood
partition coefficient, Madin–Darby canine kidney (MDCK) cell per-
meability, and human oral absorption rate of the compounds for
evaluation of druglikeness properties. Druglikeness results of com-
pounds B7 and B11 have also been shown in Table 2. The results
have indicated that both compounds are capable of displaying good
drug properties, because they were nontoxic, highly permeable
across membrane, and well adsorbable. A and B compounds have
exhibited similar physical properties. However, unlike others, most of
C compounds did not have good physical properties on the basis of
blockage of HERG K+ channels, Caco‐2, and MDCK cells membrane
permeability, and human oral absorption.
In this study, the enzyme inhibition of N‐substituted pyr-
imidinethione and acetophenone derivatives (A1–A5, B1–B11, and
C1–C11) against α‐glycosidase enzyme was analyzed, and also the
results were calculated and evaluated. The α‐glycosidase enzyme is
involved in many biological functions associated with several dis-
eases: (a) It participates in tumor metastasis via cellular interaction
with collagen type I, II, and IV; (b) α‐glycosidase enzyme inhibition
effectively decreases the risk of colorectal cancer and cere-
brovascular events in T2DM; and (c) an autosomal recessive dis-
turbance affecting α‐glycosidase causes Pompe disease. Thus, α‐
glycosidase has an extensive range of potential effects, which include
aiding in the extension of an inhibitor of α‐glycosidase for treating
T2DM.[23,24] The α‐glycosidase inhibitors act against these metabolic
enzymes in the gut tissue, slowing down the liberation of D‐glucose
from dietary complex carbohydrate molecules that lower available
glucose for absorption. Indeed, they have beneficial effects, as they
delay the progression of diabetes and decrease postprandial blood
glucose in treating prediabetic conditions.[22,25]
We identified the binding site of S. cerevisiae α‐glycosidase and
human lysosomal α‐glycosidase receptors before docking process of
B7 and B11, which exhibited best inhibitory effect during in vitro
experiment. We evaluated catalytic active site and druggable site
properties of the binding sites on the basis of SiteScore and DScore.
The binding sites of S. cerevisiae α‐glycosidase and human lysosomal
α‐glycosidase receptors have SiteScore values of 1.136 and 0.907,
respectively, and DScore values of 1.024 and 0.802, respectively. The
scores have shown quite good catalytic active site and druggable site
properties of the predicted binding site. The predicted catalytic ac-
tive sites have been used for evaluating the best poses of B7 and B11
compounds, as seen in Figure 3.
|
2.3
In silico studies
In silico studies were carried out for clarifying binding affinity and
interaction mode of the compounds with the best inhibition score
against the α‐glycosidase enzyme from Saccharomyces cerevisiae. To
understand binding affinity and interaction mode of the compounds,
the compounds were additionally docked into the catalytic active site
of the S. cerevisiae and human lysosomal α‐glycosidase enzyme. The
inhibitory effect of stereoisomers of the compounds on α‐glycosidase
enzyme was evaluated with both α‐glycosidase from S. cerevisiae and
human lysosomal α‐glycosidase receptors. Before performing docking
studies, we tested the druglikeness properties of synthesized
We first tested the reliability of induced‐fit docking technique.
For this purpose, we docked cocrystallized ligands, which are D‐
glucose and acarbose, into predicted catalytic active site of S. cere-
visiae α‐glycosidase and human lysosomal α‐glycosidase receptors,
and subsequently analyzed the binding position and root mean
square deviation (RMSD) value of redocking ligands. RMSD values
were calculated as 1.651 and 1.772 Å for S. cerevisiae α‐glycosidase
and human lysosomal α‐glycosidase receptors, respectively, with

5 of 11
TASLIMI ET AL.
|
TABLE 1 The inhibition results of N‐substituted pyrimidinethione
and acetophenone derivatives (A1–A5, B1–B11, and C1–C11)
against the α‐glycosidase enzyme
slightly weak, compared with acarbose, which is a standard inhibitor.
The best binding affinity score for compound B7 was obtained with
(S)‐stereoisomer for both S. cerevisiae α‐glycosidase and human ly-
sosomal α‐glycosidase receptors. The best binding affinity score for
compound B11 was contrarily obtained with (R)‐isomer. To clarify the
possible inhibition mechanism of the inhibitors against S. cerevisiae
and human α‐glycosidase receptors, 2D and 3D detailed binding
modes of best‐posed ligands were analyzed and presented in
Figure 5.
Compounds
IC50 values (nM)
R2
Ki values (nM)
A1
A2
A3
A4
A5
945.03
902.63
734.88
1,005.24
941.62
0.9345
0.9813
0.9104
0.9893
0.9424
988.23 94.34
974.11 105.50
805.25 84.55
989.95 135.21
1,004.25 100.43
4‐Hydroxybutyl moiety directly formed hydrogen bonds with
Asp69 and Arh442 residues and formed a hydrogen bond with
Arg446 residue via a water bridge. The hydroxyphenyl moiety
formed hydrogen bonds with Arg315 and Glu411 residues and also
interacted with Tyr158 residue through π–π stacking. Also, the tet-
rahydropyrimidine moiety formed a hydrogen bond with Gln353
residue of S. cerevisiae α‐glycosidase receptor. Besides, compound B7
(S) formed an aromatic hydrogen bond with Gln279 residue, as seen
in Figure 5A‐a. The (R)‐stereoisomer of the compound formed hy-
drogen bonds with Tyr158 and Asn455 residues via 4‐hydroxybutyl
moiety and with Arg213 and Asp352 residues via hydroxyphenyl
moiety (Figure S2A‐a). The 4‐hydroxybutyl, 4‐phenyl, 5‐carboxylate,
and pyrimidine moieties of compound B11 (R) were responsible for
inhibition of the enzyme due to many interactions with residues of S.
cerevisiae α‐glycosidase enzyme. Whereas the 4‐phenyl moiety
formed a π–π stacking interaction with Phe303 residue, the other
moieties formed hydrogen bonds with catalytic active site residues.
The 4‐hydroxybutyl moiety interacted with Arg213 and Asp352,
5‐carboxylate interacted with Arg315, and pyrimidine moiety inter-
acted with Glu411 residues of the receptor. Moreover, the
compound formed an aromatic hydrogen bond with Hie280
(Hie: histidine with hydrogen on the epsilon nitrogen), Phe314,
Asp352, and Glu411 residues, as seen in Figure 5A‐b. The compounds
also interacted with hydrophobic residue via the 4‐phenyl moiety, as
seen in Figure 3a,b. The (S)‐stereoisomer of the compound has in-
teracted with more active site residues through a hydrogen bond.
Nevertheless, the interaction is not with key residues in the active
site. Also, it has lost interaction with gatekeeper residues, Phe303
(Figure S2A‐b).
B1
B2
B3
B4
B5
B6
B7
B8
B9
B10
B11
473.16
402.70
283.04
489.11
504.28
416.10
94.16
0.9821
0.9683
0.9611
0.9904
0.9732
0.9808
0.9527
0.9623
0.9842
0.9011
0.9384
506.95 57.32
465.13 35.94
325.82 52.63
605.23 93.08
483.27 52.05
402.54 24.88
104.27 15.75
458.43 47.13
495.14 42.30
546.83 100.32
236.32 24.74
394.37
400.24
473.68
198.15
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
602.10
638.16
724.24
624.03
538.41
688.93
396.16
605.30
695.14
734.22
615.83
0.9931
0.9642
0.9730
0.9882
0.9297
0.9688
0.9180
0.9724
0.9821
0.9903
0.9068
705.30 40.82
623.11 74.03
811.42 125.32
680.33 91.26
598.12 41.17
745.33 103.21
361.27 37.96
734.21 78.36
750.30 95.03
730.73 38.94
657.11 153.78
ACRa
22,800
–
12,600 780
aAcarbose (ACR) was used as a standard inhibitor for the α‐glycosidase
enzyme.[18]
atom pair method. The RMSD values have indicated that the docking
technique was quite reliable. The best pose of cocrystallized and
redocked ligands is shown in Figure 4.
After the reliability of the docking technique has been identified,
we have carried out docking process of compounds B7 and B11 with
best inhibition score. To evaluate binding affinity and the possible
inhibition mechanism of the compounds, docking results have been
evaluated on the basis of docking score and binding mode. We also
compared docking scores and binding modes of the compounds and
standard inhibitors of S. cerevisiae α‐glycosidase and human lysosomal
α‐glycosidase receptors. Ligands with highest glide Emodel score in
the negative direction have been selected as best‐posed ligands, and
their binding affinity against the receptors is presented in Table 3.
According to the binding affinity scores, the compounds very well
interacted with S. cerevisiae α‐glycosidase receptor. Besides, the
compounds have more high docking and Glide score in the negative
direction from the standard inhibitor. The binding affinity of the
compounds against the human lysosomal α‐glycosidase receptor was
The 4‐hydroxybutyl and hydroxyphenyl moieties of the compound
B7 (S) interacted with catalytic active site residues of the human
lysosomal α‐glycosidase receptor. The 4‐hydroxybutyl moiety formed
three hydrogen bonds with Ser676, Leu677, and Leu678 residues.
The hydroxyphenyl moiety formed a hydrogen bond with Asp616
residue. Besides, 6‐methyl‐2‐thioxo‐1,2,3,4‐tetrahydropyrimidine
moiety interacted with Phe649 residue through π–π stacking, as
seen in Figure 5B‐a. The (R)‐stereoisomer of the compound formed
hydrogen bonds with Asp518 and Gly651 residues and π–π stacking
interactions with Trp376 and Phe649 residues. Moreover, the com-
pound formed an aromatic hydrogen bond with Asp404 and Trp481
residues (Figure S2B‐a).
The 4‐hydroxybutyl, thione and pyrimidine moieties of com-
pound B11 (R) formed a hydrogen bond with Asp518, Arg600, and
Asp616, respectively. The 4‐hydroxybutyl moiety also formed a

6 of 11
TASLIMI ET AL.
|
(a)
(b)
FIGURE 2 Determination of Lineweaver‐Burk graphs for excellent inhibitors of α‐glycosidase enzyme (a: B7 and b: B11)
TABLE 2 Pharmaceutical properties of the N‐substituted pyrimidinethione and acetophenone derivatives
Compound
rtvFG
MW
DHB
AHB
logPo/w
logHERG
Caco
logBB
PMDCK
% Hum. oral abs.
A1
A2
A3
A4
A5
1
0
1
0
0
306.379
260.353
276.353
306.422
214.325
0.000
0.000
0.000
0.000
0.000
2.500
2.500
2.500
2.500
2.500
4.470
3.792
3.992
4.655
2.792
−4.667
−4.461
−4.801
−4.047
−2.859
2,071.235
2,257.208
1,727.031
3,364.389
3,345.171
−0.137
0.083
−0.105
0.187
0.244
3,010.318
3,311.506
2,479.503
4,520.368
4,953.326
100.000
100.000
100.000
100.000
100.000
B1
B2
B3
B4
B5
B6
B7
B8
B9
B10
B11
0
1
0
1
0
1
0
0
0
1
2
320.406
334.432
304.406
389.512
318.433
376.513
334.432
318.433
348.459
446.560
432.534
2.000
1.000
1.000
1.000
1.000
1.000
2.000
1.000
1.000
1.000
1.000
5.950
5.200
5.200
6.200
5.200
5.200
5.950
5.200
5.950
7.200
7.200
2.748
3.592
3.032
4.308
3.585
4.799
2.938
3.334
3.390
4.563
4.471
−4.177
−4.849
−4.139
−4.933
−4.561
−5.130
−4.890
−4.596
−4.570
−5.640
−5.837
1,213.220
1,243.202
1,364.591
1,289.362
1,819.753
1,558.831
658.875
−0.404
−0.494
−0.301
−0.638
−0.279
−0.574
−0.885
−0.629
−0.786
−1.619
−1.361
1,360.712
1,266.003
1,514.714
1,376.302
2,128.634
1,721.984
683.585
100.000
100.000
100.000
100.000
100.000
100.000
94.597
937.555
999.946
100.000
100.000
100.000
100.000
811.860
854.668
314.450
305.570
488.697
492.520
C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
0
0
0
0
0
0
0
0
0
0
0
302.360
316.386
331.358
365.256
454.561
468.588
483.559
517.457
396.605
464.639
464.639
3.000
2.000
2.000
2.000
2.000
2.000
2.000
2.000
2.000
2.000
2.000
5.750
5.750
6.000
5.000
8.750
8.750
9.000
8.000
6.000
6.000
6.000
1.664
2.474
1.694
2.927
3.598
3.954
3.617
4.915
3.178
6.154
5.004
−0.684
−0.673
−0.763
−0.772
−7.874
−6.225
−7.801
−7.796
−2.015
−7.903
−5.454
1.558
5.132
−2.285
−1.812
−2.730
−1.574
−2.083
−0.975
−2.589
−1.131
−0.964
−1.057
−1.178
1.252
4.557
40.131
54.144
33.142
56.860
88.662
100.000
81.582
93.319
94.298
100.000
89.302
0.620
0.463
5.176
12.207
186.683
974.410
74.005
133.922
829.040
49.261
667.738
529.141
1,183.134
372.246
1,456.136
1,062.773
1,005.316
934.025
Abbreviations: AHB, number of hydrogen bond acceptors; Caco, Caco‐2 cell membrane permeability; DHB, number of hydrogen bond donors; logBB,
brain/blood partition coefficient; logHERG, IC50 value for blockage of HERG K+ channels; logPo/w, octanol/water partition coefficient; MW, molecular
weight; PMDCK, MDCK cell permeability in nm/s; rtvFG, reactive group (tox); % Hum. oral abs., qualitative human oral absorption.
hydrogen bond with a water molecule in the catalytic active site. The
4‐phenyl moiety interacted with Arg600 residue through π–cation.
The 4‐phenyl moiety also formed aromatic hydrogen bonds with
Trp481 via water molecule and Asp616 residues, as seen in
Figure 5B‐b. Figure 3c,d has shown that the compounds less inter-
acted with hydrophobic residues in comparison with S. cerevisiae
α‐glycosidase receptor. In contrast to the (R)‐stereoisomer, the (S)‐
stereoisomer of the compound interacted with the hydrophobic
residues Trp376, Leu677, and lLeu678 (Figure S2B‐b).
Tetrahydropyrimidine, 4‐hydroxybutyl, and hydroxyphenyl moieties
of compound B7 (S) play a key role in the inhibition of S. cerevisiae
α‐glycosidase receptors. These binding modes have shown that the

7 of 11
TASLIMI ET AL.
|
FIGURE 3 The predicted catalytic active sites of the α‐glycosidase receptors. Saccharomyces cerevisiae α‐glycosidase is presented in the top
panel: (a) B7 (S) and (b) B11 (R). Human α‐glycosidase is presented in the bottom panel: (c) B7 (S) and (d) B11 (R). The catalytic active site has
been represented as a gray mesh surface, the hydrophobic site has been represented as a yellow bubble, and the hydrophilic site has been
represented as a green bubble
compounds located in the catalytic active site of the receptors exhibited a
similar interaction with acarbose, as seen in Figure S1‐a. Yamamoto
et al.[26] have expressed that Asp69 and Asp352 residues interacted with
substrate, and Arg213, Asp352, and Arg446 residues interacted with
water molecules in the catalytic active site, and they have played a crucial
role. Other publications reported by Yamamoto et al.[27] indicated that
Tyr158, Phe303, Glu411, and Asp442 residues were important for the
enzyme activity. Besides, it has been reported that tetrazolo[1,5‐a]‐
pyrimidine pyrimidine‐2,4,6‐trione
derivatives,[28] derivatives,[29]
coumarin–iminothiazolidinone hybrids,[30] and diarylimidazole‐1,2,3‐
triazole hybrids[31] inhibited the enzyme by interacting with catalytic
active site residue. However, the inhibitors mostly interacted with
hydrophobic residues in catalytic active sites of the enzyme due to
their aromatic ring moieties. According to our docking results, the
FIGURE 4 Induced‐fit docking technique was tested calculating RMSD value between cocrystallized ligand and best‐pose of redocked
ligand. (a) D‐Glucose was redocked for Saccharomyces cerevisiae α‐glycosidase receptor and (b) acarbose was redocked for human α‐glycosidase
receptor. Cocrystalized D‐glucose was represented with yellow ball–stick, cocrystallized acarbose was represented with faded‐red ball‐stick and
re‐docked ligands were represented with grey ball‐stick

8 of 11
TASLIMI ET AL.
|
TABLE 3 Binding affinity scores (kcal/
mol) of B7 and B11 stereoisomers in the
catalytic sites of the α‐glycosidase receptors
Saccharomyces cerevisiae α‐glycosidase
Human α‐glycosidase
Com-
Docking
score
Glide
score
Glide
Docking
score
Glide
score
Glide
pounds
Emodel
Emodel
B7 (S)
B7 (R)
B11 (S)
B11 (R)
ACRa
−8.608
−7.378
−8.368
−8.582
−7.962
−8.610
−7.379
−8.368
−8.582
−8.290
−88.520
−74.148
−88.086
−95.083
−140.803
−6.065
−5.559
−5.861
−6.163
−7.334
−6.065
−5.561
−5.861
−6.163
−7.662
−56.393
−60.102
−62.267
−70.314
−112.995
aAcarbose (ACR) was used as a standard inhibitor for the α‐glycosidase enzyme.[18]
(S)‐stereoisomer of compound B7 and the (R)‐stereoisomer of compound
B11 have mainly interacted with mentioned residues, especially Tyr158,
Phe303, and Glu411. However, other stereoisomers of the compounds
interacted with less important active site residues due to conformational
changing. The differences lead to losing π–π stacking interaction with
gatekeeper residues, Tyr158 and Phe303.
hydrophobic contact between Trp376, Ile441, Trp516, Met519, Trp613,
and Phe649 residues results in the stabilization.[32] Taj et al.[33] have
reported that pyrazolobenzothiazine 5,5‐dioxide derivatives inhibit
Asp203, Asp542, Asp327, His600, and Arg526 residues of human
maltase‐glucoamylase enzyme that is related to the human α‐glycosidase.
In our previous study, we have explained that imidazolinium chloride salt
potently inhibited the human maltase‐glucoamylase enzyme by interact-
ing with Asp542 and Gln603 residues through a salt bridge and hydrogen
bond, respectively.[34] According to the binding modes of the (S)‐
stereoisomer of compound B7 and the (R)‐stereoisomer of compound
B11, they interacted with Arg600, Asp616, and Asp512 residues, which
On the other side, interactions between the compounds and residues
of human lysosomal α‐glycosidase receptor were less than interactions
between the compounds and residues S. cerevisiae α‐glycosidase enzyme.
The human receptor's active site is stabilized with H‐bonds between side
chains of Asp404, Asp518, Arg600, Asp616, and His674 residues. A
FIGURE 5 (A) 2D and 3D interaction diagrams of the compounds complexed with Saccharomyces cerevisiae α‐glycosidase receptor. The
receptor structure was represented as a yellow ribbon model, amino acid residues were represented as a thick tube model, (a) compound B7 (S)
was represented as a turquoise ball–stick model, and (b) compound B11 (S) was represented as a green ball–stick model. (B) 2D and 3D
interaction diagrams of the compounds complexed with human α‐glucosidase receptor. The receptor structure was represented as a faded red
ribbon model, amino acid residues were represented as a thick tube model, (a) compound B7 (R) was represented as a turquoise ball–stick
model, and (b) compound B11 (R) was represented as a green ball–stick model

9 of 11
TASLIMI ET AL.
|
are critical for enzyme catalytic activity. However, other stereoisomers of
the compounds have lost interaction with the critical residues as with the
other enzyme. The 4‐hydroxybutyl and pyrimidinethiones moieties
played a key role in S. cerevisiae and human α‐glycosidase inhibition.
was confirmed by spectral and physicochemical methods, such as 1H
and 13C nuclear magnetic resonance spectroscopy.[15–17]
The InChI keys of the investigated compounds, together with some
biological activity data, are provided as Supporting Information Data.
|
|
3
CONCLUSION
4.2
α‐Glycosidase inhibition assay
These synthesized novel compounds, B7 and B11, have shown effi-
cient inhibitory actions; hence, they can be used as a leading candi-
date for initiation of effective α‐glycosidase inhibitors. The molecular
docking studies on the enzymes have also shown that the most active
compounds B7 (S) and B11 (R) have strongly bound to the catalytic
active site of the enzyme.
The inhibitory effect of the N‐substituted pyrimidinethione and
acetophenone derivatives (A1–A5, B1–B11, and C1–C11) on the
α‐glycosidase enzyme activity was assayed using p‐nitrophenyl‐D‐
glycopyranoside (p‐NPG) substrate, according to the assay of Tao et al.[18]
as described previously.[35–37] First, 200 µl of phosphate buffer was mixed
with 40 µl of the homogenate solution in a phosphate‐buffered solution
(PBS; 0.15 U/ml, pH 7.4). Also, 50 µl of p‐NPG in a PBS (5 mM, pH 7.4)
after preincubation was added and again incubated at 30°C. The absor-
bances were spectrophotometrically measured at 405 nm, according to
previous studies.[18,38–40]
|
4
EXPERIMENTAL
|
4.1
Chemistry
|
|
In silico studies
4.1.1
Synthesis of cyclic thioureas (A1–A5)
4.3
The TFA,
a
new effective synthesis method of tetra(hexa)
Druglikeness properties of all novel compounds and inhibition me-
chanism of compounds, which have the best inhibition score against
α‐glycosidase receptor, were calculated and specified with in silico
methods. The Small Drug Discovery Suites package (Schrödinger
2019‐3 LLC) was used in performing in silico studies.[41–43]
hydropyrimidinethione‐carboxylates (A1–A5), has been employed in
this study. The three‐component condensation reactions were per-
formed for 2.5–3.0 hr at 60–75°C.[15]
|
4.1.2
Synthesis of derivatives of new
|
Ligand preparation and pharmacokinetic
pyrimidinethiones (B1–B11)
4.3.1
properties prediction
Pyrimidinethiones (0.02 mol) were dissolved in 2:1 ratio of acet-
ylacetone (12 ml) and ethyl alcohol (5 ml); also, 1,2‐epoxypropane (1,2‐
epoxybutane) (0.02 mol) was added to this mixture drop by drop. After
being dissolved, it was stirred for 30 min, and then 20 mg AlCl3 cata-
lyst was added to it. Then, it was mixed and heated at 60–65°C. The
synthesis of these compounds was described in detail in the previous
study.[16] After the completion of the reaction, the mixture is cooled,
purified by means of recrystallization with ethanol, and dried.
For the QikProb and induced‐fit docking studies, 2D structure of the
compounds with best inhibition score was sketched and their 3D
structure was produced with LigPrep module of Schrodinger Maestro
12.0. Correct molecular geometries and protonation state at pH
7.0 2.0 of ligand were prepared using Epik module and OPLS‐2005
force field.[44,45] The pharmacokinetic properties of prepared ligands
were calculated with QikProp module of Schrodinger Maestro 12.0.
Briefly, prepared ligands were opened and the calculation was per-
formed using the default parameter by selecting the five most similar
drug molecule options on QikProp module.[46,47]
|
4.1.3
The amides of 1,1‐bis‐(carboxymethylthio)‐1‐
arylethanes derivatives (C1–C11)
|
Protein preparation and binding site
prediction
The synthesis of these compounds was recorded in detail in the pre-
vious study.[17] All the synthesized compounds (A1–A5, B1–B11, and
C1–C11) were filtered and dried in dichloromethane crystallized from
ethyl alcohol. All reagents were purchased and used without further
purification. Glassware was dried before use. Compounds were purified
by dry flash chromatography using silica 60, <0063 mm, and water
pump vacuum or by flash chromatography using silica 60 Å, 230–400
mesh as stationary phases. Thin‐layer chromatography plates (silica gel
60 F254) were visualized either with a UV lamp or iodine. Their structure
4.3.2
X‐ray crystal structures of S. cerevisiae α‐glycosidase receptor (PDB
code: 3A4A) and human lysosomal α‐glycosidase (PDB code: 5NN8)
have been acquired from RCSB Protein Data Bank. They have been
selected due to their best resolution and good percentile ranks.
Moreover, the structures have a ligand which can be used for docking
validation test in the catalytic active site. The typical crystal structure

10 of 11
TASLIMI ET AL.
|
in the PDB format is not suitable for immediate use in molecular
modeling calculations. Hence, the crystal structure has been repaired
and prepared with protein preparation wizard module of Schrodinger
Maestro 12.0 before using in binding site prediction and induced‐fit
docking studies. The workflow that includes a detailed description of
previous studies[48,49] was described in the overview. Bond order and
charges were assigned and then missing hydrogen atoms were added
to crystal structure. Missing side chains were filled using Prime module
of the program. Amino acids were ionized by setting physiological pH
with the help of Propka software. Water molecules that formed less
than three contacts with the protein or ligand were removed. Finally,
energy minimization and geometry optimization have also been per-
formed using OPLS force field.
REFERENCES
[1] B. Jank, T. Pernat, C. O. Kappe, Molecules 2000, 5, 227.
[2] I. S. Zorkun, S. Sarac, S. Celebi, K. Erol, Bioorg. Med. Chem. 2006, 16,
4479.
[3] J. J. V. Eynde, N. Audiort, V. Canonne, S. Michel, Y. V. Haverbeke, C.
O. Kappe, Heterocycles 1997, 45, 1967.
[4] D. Dalinger, A. Stadler, C. O, Pure Appl. Chem. 2004, 76, 1017.
[5] A. Sujayev, E. Garibov, Organic Chemistry, Lambert Academic Pub-
lishing, Germany 2016, p. 55.
[6] S. A. Sardarova, S. F. Osmanova, F. A. Mamedov, Sh. Y. Gamidova,
Chem. Probl. 2015, 3, 319.
[7] P. P. Ashvini, K. P. Tejasvi, R. P. Ankita, World J. Pharm. Pharm. Sci.
2015, 4, 1780.
[8] H. K. Nisreen, H. T. Jumbad, H. Ammar, Al Ibn Al‐Haitham, J. Pure
Appl. Sci. 2013, 26, 296.
[9] A. Ricci, Angew. Chem., Int. Ed. Engl. 2007, 5, 22.
[10] I. Gulcin, A. Z. Tel, A. C. Gören, P. Taslimi, S. Alwasel, J. Food Meas.
Charact. 2019, 13, 2062.
After protein preparation, the catalytic site of the receptors
was predicted with SiteMap module of Schrodinger Maestro 12.0
for target selection and docking hits' evaluation. The prepared
receptor was imported into Maestro 12.0, and the binding site
was predicted using the default parameter of top‐ranked potential
protein binding site. The analysis of SiteScore and DScore de-
termined whether the binding site had catalytic active site
characteristics.[50,51]
[11] P. Taslimi, F. M. Kandemir, Y. Demir, M. İleritürk, Y. Temel, C. Cağ-
layan, I. Gulcin, J. Biochem. Mol. Toxicol. 2019, 33, e22313.
[12] H. Genc Bilgicli, A. Kestane, P. Taslimi, O. Karabay, A. Bytyqi‐Damoni,
M. Zengin, I. Gulcin, Bioorg. Chem. 2019, 88, 102931.
[13] R. Kaya, P. Taslimi, M. E. Naldan, I. Gulcin, Lett. Drug Des. Discov 2019,
14, 573.
[14] E. Bursal, A. Aras, Ö. Kılıç, P. Taslimi, A. C. Gören, I. Gulcin, J. Food
Biochem. 2019, 43, e12776.
[15] V. Farzaliyev, P. Taslimi, A. Sujayev, N. Nazarov, Z. Huyut, S. H. Al-
wasel, I. Gulcin, J. Biochem. Mol. Toxicol. 2017, 31, e21897. https://doi.
org/10.1002/jbt.21897
|
4.3.3
Induced‐fit docking
[16] P. Taslimi, A. Sujayev, F. Turkan, E. Garibov, Z. Huyut, V. Farzaliyev, S.
Mamedova, I. Gulcin, J. Biochem. Mol. Toxicol. 2018, 32, e22019.
https://doi.org/10.1002/jbt.22019
To determine binding affinities and inhibition mechanisms of the
compounds with the best inhibition score, molecular docking stu-
dies were carried out. Induced‐fit docking technique is one of the
best ways to calculate binding affinity between ligand and receptor,
because both ligand and receptor are completely flexible in this
docking technique. The technique has been performed with the
induced‐fit docking module of Schrodinger Maestro 12.0 according
to proposed previous studies.[52–54] The centroid of the residues
was generated around the selected ligand or residues in the cata-
lytic site of the receptor. Then, side chains were automatically
trimmed based on B‐factor; the closest residues to the ligand were
refined within 3.4 Å of ligand pose in prime refinement. Before
calculating binding affinity between ligands and the receptors,
docking validation test was carried out for understanding accuracy
of induced‐fit docking technique. The test has been performed with
redocking procedure by evicting inhibitor complexed in the crystal
structure of the receptors. The RMSD value between cocrystallized
ligand and redocked ligand has been calculated with the atom pair
method in Superposition panel of Schrodinger Maestro 12.0. Fol-
lowing the docking process, best‐scored compounds have been
determined by analyzing docking scores.
[17] P. Taslimi, S. Osmanova, C. Caglayan, F. Turkan, S. Sardarova, V.
Farzaliyev, A. Sujayev, N. Sadeghian, I. Gulcin, J. Biochem. Mol. Toxicol.
2018, 32, e22191. https://doi.org/10.1002/jbt.22191
[18] Y. Tao, Y. Zhang, Y. Cheng, Y. Wang, Biomed. Chromatogr. 2013, 27, 148.
[19] Y. Natori, T. Imahori, Y. Yoshimura, J. Synth. Org. Chem., Jpn. 2016, 74,
335.
[20] A. Aktas, D. Barut Celepci, Y. Gok, P. Taslimi, H. Akıncioglu, I. Gulcin,
Crystals 2020, 10, 171.
[21] A. Maharramov, R. Kaya, P. Taslimi, M. Kurbanova, A. Sadigova, V.
Farzaliyev, A. Sujayev, I. Gulcin, Arch. Pharm. 2019, 352, e1800317.
[22] P. Taslimi, H. E. Aslan, Y. Demir, N. Oztaskin, A. Maraş, I. Gulcin, S.
Beydemir, S. Goksu, Int. J. Biol. Macromol. 2018, 119, 857.
[23] M. Huseynova, P. Taslimi, A. Medjidov, V. Farzaliyev, M. Aliyeva, G.
Gondolova, O. Şahin, B. Yalçın, A. Sujayev, E. B. Orman, A. R. Özkaya,
I. Gulcin, Polyhedron 2018, 155, 25.
[24] M. Zengin, H. Genc, P. Taslimi, A. Kestane, E. Guclu, A. Ogutlu, O.
Karabay, I. Gulcin, Bioorg. Chem. 2018, 81, 119.
[25] I. Gulcin, P. Taslimi, A. Aygün, N. Sadeghian, E. Bastem, O. I. Ku-
frevioglu, F. Turkan, F. Şen, Int. J. Biol. Macromol. 2018, 119, 741.
[26] K. Yamamoto, H. Miyake, M. Kusunoki, S. Osaki, FEBS J. 2010, 277,
4205.
[27] K. Yamamoto, H. Miyake, M. Kusunoki, S. Osaki, J. Biosci. Bioeng.
2011, 112, 545.
[28] L. Suresh, P. Onkara, P. S. V. Kumar, Y. Pydisetty, G. V. P. Chan-
dramouli, Bioorg. Med. Chem. Lett. 2016, 26, 4007.
[29] A. Barakat, S. M. Soliman, A. M. Al‐Majid, G. Lotfy, H. A. Ghabbour, H.
Fun, S. Yousuf, M. I. Choudhary, A. Wadood, J. Mol. Struct. 2015,
1098, 365.
CONFLICTS OF INTEREST
The authors declare that there are no conflicts of interests.
[30] A. Ibrar, S. Zaib, I. Khan, Z. Shafique, A. Saeed, J. Iqbal, J. Taiwan Inst.
Chem. Eng. 2017, 81, 119.
ORCID
[31] M. Saeedi, M. Mohammadi‐Khanposhtani, M. S. Asgari, N. Eghbal-
nejad, S. Imanparast, M. A. Faramarzi, B. Larijani, M. Mahdavi, T.
Akbarzadeh, Bioorg. Med. Chem. Lett. 2019, 27, 115148.
Parham Taslimi
http://orcid.org/0000-0002-3171-0633
http://orcid.org/0000-0002-0155-3390
Muhammet Karaman

11 of 11
TASLIMI ET AL.
|
[32] V. Roig‐Zamboni, B. Cobucci‐Ponzano, R. Iacono, M. C. Ferrara, S.
Germany, Y. Bourne, G. Parenti, M. Moracci, G. Sulzenbacher, Nat.
Commun. 2017, 8, 1111.
[48] U. Atmaca, R. Kaya, H. S. Kahraman, M. Celik, I. Gulcin, Bioorg. Chem.
2019, 88, 102980.
[49] P. Taslimi, F. Turkan, A. Cetin, H. Burhan, M. Karaman, I. Bildirici, I.
Gulcin, F. Sen, Bioorg. Chem. 2019, 92, 103213.
[50] SiteMap, Schrödinger, LLC, New York, NY 2019.
[51] S. Bayindir, C. Caglayan, M. Karaman, I. Gulcin, Bioorg. Chem. 2019,
90, 103096.
[33] S. Taj, U. A. Ashfaq, S. Aslam, M. Ahmad, S. H. Bhatti, Biologia 2019,
74, 1523.
[34] B. Yiğit, R. Kaya, P. Taslimi, Y. Işık, M. Karaman, M. Yiğit, İ. Özdemir, I.
Gulcin, J. Mol. Struct. 2018, 1179, 709.
[35] P. Taslimi, E. Koksal, A. C. Gören, E. Bursal, A. Aras, O. Kılıç, S. Al-
wasel, I. Gulcin, Arab. J. Chem. 2020, 13, 4528.
[52] (a) Induced Fit Docking protocol, Glide, Schrödinger, LLC,
New York, NY 2016. (b) Prime, Schrödinger, LLC, New York,
NY 2019.
[36] S. Burmaoğlu, E. Akın Kazancioglu, R. Kaya, M. Kazancioglu, M.
Karaman, O. Algül, I. Gulcin, J. Mol. Struct. 2020, 1208, 127868.
[37] P. Taslimi, K. Turhan, F. Turkan, H. S. Karaman, Z. Turgut, I. Gulcin,
Bioorg. Chem. 2020, 97, 103647.
[53] F. Turkan, A. Cetin, P. Taslimi, M. Karaman, I. Gulcin, Bioorg. Chem.
2019, 86, 420.
[54] R. Kalin, Z. Koksal, P. Kalin, M. Karaman, I. Gulcin, H. Ozdemir,
J. Biochem. Mol. Toxicol. 2020, 34, e22421.
[38] C. Çaglayan, Y. Demir, S. Kucukler, P. Taslimi, F. M. Kandemir, I.
Gulcin, J. Food Biochem. 2019, 43, e12720.
[39] N. Eruygur, M. Ataş, M. Tekin, P. Taslimi, U. M. Koçyiğit, I. Gulcin,
S. Afr. J. Bot. 2019, 120, 141.
SUPPORTING INFORMATION
[40] Y. Demir, P. Taslimi, M. S. Ozaslan, N. Oztaskin, Y. Çetinkaya, I.
Gulcin, Ş. Beydemir, S. Goksu, Arch. Pharm. 2018, 351, 1800263.
[41] Z. Koksal, R. Kalın, P. Kalın, M. Karaman, I. Gulcin, H. Ozdemir, J. Food
Biochem. 2020, 44, e13132.
Additional supporting information may be found online in the
Supporting Information section.
[42] M. Yigit, B. Yigit, P. Taslimi, I. Ozdemir, M. Karaman, I. Gulcin, J. Mol.
Struct. 2020, 1207, 127802.
[43] S. Bal, R. Kaya, P. Taslimi, A. Aktas, M. Karaman, I. Gulcin, Bioorg.
Chem. 2020, 94, 103468.
How to cite this article: Taslimi P, Sujayev A, Karaman M,
et al. N‐Substituted pyrimidinethione and acetophenone
derivatives as a new therapeutic approach in diabetes. Arch
Pharm. 2020;e2000075.
[44] C. Bayrak, P. Taslimi, H. S. Kahraman, I. Gulcin, A. Menzek, Bioorg.
Chem. 2019, 85, 128.
[45] LigPrep, Schrödinger, LLC, New York, NY 2019.
[46] QikProp, Schrödinger, LLC, New York, NY 2019.
[47] F. Turkan, A. Cetin, P. Taslimi, H. S. Karaman, I. Gulcin, Arch. Pharm.
2019, 352, e1800359.
https://doi.org/10.1002/ardp.202000075