in the brush border membrane of mature enterocytes in

2
the
Hindawi Publishing Corporation
International Journal of Medicinal
International
small intestine, Journal where
of Medicinal Chemistry
it absorbs dietary glucose and Chemistry
Volume 2014, Article ID 739646, 15 pages
galactose
http://dx.doi.org/10.1155/2014/739646 relationships between structural modifcations and
from the gut lumen [13]. SGLT2, expressed exclusively activities
in against hSGLT2 inhibitors with the help of QSAR
the kidney, is located in the S1 segment of the proximal modeling.
convoluted tubule of the kidney. It is a low-afnity, high- 2.2. Computational Details. All the computational
capacity cotransporter and is responsible for 90% of studies
renal were performed by V-life MDS (Molecular Design Suite)
Research Article
glucose reabsorption [14]. Several therapeutic agents 3.5
are software supplied by V-life Sciences Technologies Pvt.
Molecular Modeling Studies of Thiophenyl C-Aryl Glucoside
available for monotherapy or for combination therapy Ltd.,
SGLT2 Inhibitors as Potential Antidiabetic Agents
with
different mechanisms to treat diabetics, such as
Pune, India [69]. The sketched structures were used for
the
metformin, calculation of 2D molecular descriptors using QSAR
rosiglitazone, sitagliptin, acarbose, and glimepiride [15]. module
The obvious need for new approaches to treat patients of Molecular Design Suite software. Each compound was
with 1 2 subjected to energy minimization and batch
uncontrolledMukesh T2DM C.has Sharma
prompted andcontinuous
Smita Sharma optimization
exploration 1
School of Pharmacy, Devi Ahilya University, Takshila Campus, using Merck
Khandwa Molecular
Road, IndoreForce452 Field (MMFF), fxing Root
of alternative
001, Indiatargets involved in maintenance of ˚
glucose 2 Mean Square Gradients (RMS) to 0.01 kcal/mol A [70].
homeostasis. Department
Several of Chemistry,
SGLT2 Chodhary Dilip Singh Kanya Mahavidyalya,
The sphere Bhind 477001,
exclusion India [71] was adopted total
method
Correspondence shouldinhibitors
be addressed have to been
Mukesh C. Sharma;
reportedmukesh2206@rediffmail.com
as set
undergoing clinical
Received 14 Maytrials. Phlorizin
2014; Revised [16], 3-
29 October of inhibitors
2014; Accepted 29 October was2014;
divided randomly
Published 10 into a training set
December 2014
(benzo[b]furan- (26
5-yl)-2󸀠 ,6 Academic Editor: 󸀠Maria
󸀠 -dihydroxy-4 Cristina Breschi
-methylpropiophenone-2 󸀠 -O -�- compounds) for generation of QSAR models and a test
D- set
2.3. Calculating 2D Descriptors. In the current approach,
Copyright [17], © 2014 M. C. Sharma and S. Sharma. This isan
ancompounds)
(7 open access hasarticle
for distributed
validation under the Creative
glucopyranoside
Commons Attribution
sergliflozin [18], and remogliflozin attempt
(1) the maximum been taken
value to ofof the developed
understand
pIC model.
the structural
50 of test set should be
[19] are O-glycosides and show strong inhibition of Thisless
and physicochemical requirements of a set of hSGLT2
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
SGLT2. properly cited. random
inhibitors division
byorthe of
helpdata set
toofmaximum will bevalue
regression done ofthrough
2D quantitative ofseveral
A QSAR study on thiophenyl derivatives as SGLT2 inhibitors cycles asthan
potential
in order
equal
antidiabetic
to get the agents
best QSARwas model. pIC50with
performed This study
They also demonstrate efcacy in vivo when structure-
training
thirty-three
administered will
activity
(2) relationship
the minimum (2D
value QSAR).
of pICThe ofenergy-minimized
test set should be
compounds. Comparison of the obtained results indicated the set, 50
superiority of the genetic algorithm over the simulated
orally inannealing
rats or mice. They induce a glucosuric help inhigher
geometry rational
was drug
used
than designing
or for
equaltheto of theseof
calculation
minimum derivatives
the various
value as2D
of pIC50 of
One of the
response, andmainstepwisefeatures of diabetesvariable
forward-backward is the elevation
method for SGLT2
offeature
descriptors
training
selection.such
set.as
The topological,
best 2D QSAR model shape and geometrical,
showed satisfactory
blood
the result of the blockade of renal glucose reabsorption, inhibitors
statistical and for the eradication of T2DM. The unicolumn
sugar with
and its deleterious
parameters for the dataconsequences in a �variety
set (�2 = 0.8499, statistics
of and
2 = 0.8267, pred �2 of
physicochemical the training
= 0.7729) parameters anddescriptors
with four test
suchsets are reported
as individual
describing (H- in
the
tissues nature lead
consequently of to reduction of the blood glucose Table 2.
Acceptor
[1]. Thus,
level substituent
control of groups and the glucose
the plasma environmentlevelofisthe of substitution
The
count,site.
that maximum Evaluation
induces
H-Donor and of
hyperglycemia
count, the
minimum model
XlogP, implied
value
[6].
retentionin that
Medical electron-rich
training
index and test
complications
(Chi),
substitution set
utmost
and improvement of insulin sensitivity [20]. Quantitative element associated with T2DM include cardiovascular disease,
position improves the inhibitory activity. The good predictive 3D-QSAR models by k-nearest neighbor (kNN) method
importance
structure-activity in therelationship
for molecular treatment of(QSAR) this disease.
studiesIncan recent
be were
stroke,
count, compared in a wayestate
estate numbers, that contribution, and
years, to
utilized feld predict
analysis eye irritation
(MFA) potential as an
have cross-validated nephropathy,
alignment-
coefcient �2 value of 0.7663 and retinopathy,
predicted �renal
2 value failure, andThe results
of 0.7386.
the ideahave
alternative thatshowed
affecting glucose absorption in the amputations of
independent descriptors were used as predictor
intestine
in
1. silico method,
that
Introduction just as
thiophenyl it has
groups arebeen used successfully
necessary for activity and tohalogen,
the extremities
variables),bulky,asand[7].
lessInbulky
recent years,
groups much attention
in thiophenyl nucleus has
and/or the
predict enhanced
severalglucose thereabsorption
other toxicological in endpoints
the kidneyfor might
some be they
been were found to be appropriate for the development
a
time biological
[21]. Hence,activity. These studies
in continuation are efforts
to our promising for the development
[22–67] of novel SGLT2 inhibitor,
given to sodium-dependent
of glucose which may have potent
cotransporters
possible
in
2.1. Data antidiabetic
way
Set.to Thecontrol
biological
the sugar
data level
set washaschosen
evolved. from a models.
(SGLTs), A considerable number of the 265
activity.
Diabetes QSAR studies for angiotensin II AT1 receptor, physicochemical
developing
series mediators of reabsorption of glucose in the human body.
comprises
antitubercular
of thirty-three a group
thiophenyl
agents, of metabolic
antimalarial
derivatives
disorders
activity,
as SGLT2characterised
inhibitors parameters,
Sodium-dependent 300 alignment
glucose cotransporter
type parameters, 2 (SGLT2)
and 99is a
by
antimicrobial
as high-capacity, low-afnity transporter expressed
atoms
2. Materials
chronic
activity,
potential and Methods
hyperglycaemia
antibacterial
antidiabetic activity,
agents
with reported
disorders
COX inhibitors,
by
in Lee
the et and al.so[68]. types
selectively
count descriptors calculations were done using
metabolism
forth.
The in the S1 domain of the proximal tubule in the kidney
the
of carbohydrate,
biological
In this study,
activitywe fat,
values
have
and [IC protein
taken
50 (nM)]
thiophenyl
thatreported
resultderivatives
inindefects in V-life,
and is MDSresponsible
3.5. Thefor preprocessing
90% of renal of glucose
the independent
reuptake.
secretion and action of insulin [2]. Dysfunction and
for
nanomolar Sodium dependent
variables (i.e., 2D descriptors)
glucose cotransporter
was done by 1 (SGLT1),
removingon
failurewere converted
performing
units 2D and 3Dtoquantitative
their molar structural-activity
units pIC50 and the other hand, is a low-capacity, high-afnity
of various organs, especially the eyes, kidneys, nerves, invariable
rela-
subse- transporter(constant column) which resulted in a total of
and
tionship
quently used
analysis as theanddependent
calculations variable
in order fortotheunderstand
QSAR distributed
216 molecular in the
descriptors
kidney, gut, to be andusedotherfortissues,
QSAR analysis.
heart,stereoelectronic
their
analysis. and the blood vessels properties.are the
Genetic
usualalgorithm
complications The responsible
various alignment-independent descriptors were
of converted to pIC50 for the QSAR analysis along with also
(GA),
The for the remaining 10% of glucose reuptake [8]. Na+
diabetes [3,
simulated
the annealing
4]. Diabetes (SA),isand mainly
stepwise
divided forward-
into four -glucose
calculated. In this study to calculate AI descriptors, we
main
backward
structure of the compounds in the series are listed in cotransporter (SGLT) is a membrane protein that plays
have
types 1including
variable
Table selection insulin-dependent
methods have been diabetes
employedmellitus for an important
used the following
role inattributes:
the reabsorption
2 (double of bonded
glucose atom),
in the 3
(type 1), with asterisk). The test compounds were
selec-
(marked kidneys.
(triple bonded
Sodium-dependent
atom), C, N, O,glucose S, H, F,cotransporters
Cl, Br, and I, and
non-insulin-dependent
tion
selected
of relevant descriptors. diabetesThe mellitus
obtained(type results 2), (SGLTs),
the distance range of 0–7. The QSAR models were built
gestational
provide
manually such that the structural diversity and wide mediators of reabsorption of glucose in the kidney, have
with
diabetes,
further
range insight
and other
into somespecifc benefcial
types [5].information
Diabetes in mellitus therecently
consideration
emerged of asthenovelapplicability
drug targets of the
for the
descriptor
type
structural
of activity in the data set were included. In this paper, a for treatment
the of
2 (T2DM)
modifcations
series of thiophenyl
accounts
to design for
compounds
almost
new potential
90%withof substitutions
diabetes
SGLT2 inhibitors.
cases,at X activity.
diabetesVarious
[9]. SGLT types
is known
of physicochemical
to have three descriptors
isoforms
with R
Moreover,
and new compounds with high predictive (SGLT1,
have
the property
activities
position of thiophenyl
of insulin moiety
resistance are and
subjected
beta-cell to SGLT2,
been calculated
and SGLT3) which
[10–12].
are shown
SGLT1inisthe expressed
data sheet
dysfunction
were
examining
designed. the primarily
(Table 3).
International Journal of Medicinal 3
Chemistry

Table 1: Structure and biological activity of thiophenyl derivatives hSGLT2

inhibitors.
X

S R

O
HO

HO
OH

OH

S. number X R IC50 pIC50

OMe

1 Cl 86.5 7.0629

OEt

2 Cl 34.6 7.4609

OH

3∗ Cl 140 6.8538

4 Cl 65.0 7.1870

5 Cl 54.6 7.2628

6∗ Cl 111 6.9546

7 Cl 94 7.0268

O
O

8 Cl 115 6.9393
4 International Journal of Medicinal
Chemistry

Table 1: Continued.
X

S R

O
HO

HO
OH

OH

S. number X R IC50 pIC50

O
O

9 Cl 70.7 7.1505

SMe

10∗ Cl 12.8 7.8927

SEt

11 Cl 48.4 7.3151

12 Cl 11.9 7.9244

13 Cl 57.2 7.2426

O O

14∗ Cl 162 6.7904

15 Cl 60.2 7.2204

16 Cl 4.47 8.3496
International Journal of Medicinal 5
Chemistry

Table 1: Continued.
X

S R

O
HO

HO
OH

OH

S. number X R IC50 pIC50

17 Cl 10.3 7.9871

18 Cl F 91.3 7.0395
S

OMe

19∗ H 11.5 7.9393

OEt

20 H 8.73 8.0589

OH

21 H 50.2 7.2992

22∗ H 27.5 7.5606

23 H F 21.1 7.6757
S

OEt

24 H 71.4 7.1463
6 International Journal of Medicinal
Chemistry

Table 1: Continued.
X

S R

O
HO

HO
OH

OH

S. number X R IC50 pIC50

OMe

25∗ H 68.9 7.1617

Cl

26 H 451 6.3458

t-Bu

27 H 88.3 7.0540

28 H 69.6 7.1573

29∗ H 24.8 7.6055

30 H 59.8 7.2232

31 Br 12.4 7.9065

32 OMe 49.5 7.3053

International Journal of Medicinal 7
Chemistry
Table 1: Continued.
X

S R

O
HO

HO
OH

OH

S. number X R IC50 pIC50

33∗ Me 29.3 7.5331

∗Structures with (∗ ) are included in test set

validation.

Table 2: Unicolumn statistics of training and test set for 2D QSAR

studies.
Data set Average Max Min Std. dev. Sum
2D QSAR
Training 1.6127 2.6542 0.6503 0.4426 41.9299
Test 1.7765 2.2095 1.1072 0.3889 12.4355
3D QSAR
Training 1.5401 2.2095 0.6503 0.4163 36.9636
Test 1.9335 2.6542 1.3945 0.3452 17.4018
2.4. Calculating 3D Descriptors. Energy minimized and
geometry optimized structures of molecules were
aligned by and standard error for predicted �2 (pred �2 se). The
the template-based method [72] where a template internal
structure cross-validated predictability (�2 ) was evaluated by the
is defned and used as a basis for alignment of a set of equa-
molecules, and a reference molecule is chosen on which tion given below:2 2
∑ (�� − �� )
the other molecules of the data set get aligned � = 1 − 2
, (1)
considering ∑ (� � − �mean )

the chosen template. In the present study, we aligned

where �� , �� are the actual and predicted activity
the
of the �th
database by ftting all of the compounds on most active
molecule in the training set, respectively, and �mean is
compound 16 (Figure 1(a)) as an alignment template
the
using a
average activity of all molecules in the training set. For
common substructure with the V-life MDS routine
external validation, activity of each molecule in the test
database
set
alignment. The superimposition of all molecules based
was predicted using the model generated 2 from the
on ∑ (�� − �� )
training Pred �
2
= 1 − ,
minimizing root mean square deviation (RMSD) is shown 2 (2)
set. The pred �2 value is calculated
∑ (�� − �meanas follows:
)
in
Figure 1(b).
where �� , �� are the actual and predicted activity
The steric, electrostatic, and hydrophobic felds were
of the �th
calculated at each lattice intersection of a regularly
molecule in the test set, respectively, and �mean is the
spaced
average
˚
activity of all molecules in the training set. Both
grid of 2.0 A. Methyl probe of charge +1 with 10.0
summations
kcal/mole
2.5. External Validation for 2D QSAR Models. The QSAR are over all molecules in the test set. Thus, the pred �2
electrostatic
models were and 30.0 kcal/mole
assessed steric of
by the number and hydrophobic
cross- value is
cutoff
validated �2 indicative of theofpredictive
2.6. Evaluation power ofofthe
the Quantitative current
Models. model
Among
wasleave-one-out
by used for felds generation.
method [73] (�This resulted in
2 ), cross-validated
based
several
calculation
standard of on the external
search teststepwise
algorithms, set. (SW) forward variable
4500 (�
error feld descriptors
2 se), predicted (1500 for each test
�2 for external steric,
set (pred �2 ),
electrostatic, and selection
hydrophobic which theoretically form a continuum) for
all
the compounds in separate columns (Table 3).
8 International Journal of Medicinal
Chemistry
Table 3: Selected descriptors values for QSAR
models.
SsCH3 count T C Cl 1 S 1044 E 184 E 1139 H 1153
13.35725 1 −0.00234 −0.00103 0.070427 0.068996
14.89175 1 −0.00252 −0.03374 0.039663 0.110293
13.05024 1 −0.00332 −0.0034 0.045418 0.156499
12.66982 1 −0.00271 −0.07924 0.015991 0.205079
13.19176 1 −0.00718 −0.05341 0.01499 0.156726
13.4038 1 −0.00276 −0.02181 0.126288 0.213233
13.05024 1 −0.00272 −0.11568 0.049628 0.170716
14.11489 1 −0.00229 −0.04659 0.03041 0.190254
13.19176 1 −0.0037 −0.05161 0.007637 0.195659
13.85076 1 −0.00269 −0.09992 0.148325 0.170358
13.95209 1 −0.002 −0.05398 0.046206 0.178527
13.49164 1 −0.00307 −0.04381 0.022364 0.196138
13.8452 1 −0.00213 −0.03588 0.01794 0.135694
13.11122 1 −0.00219 −0.05675 0.025435 0.117383
11.57052 1 −0.0077 −0.03111 0.040108 0.13964
13.05024 1 −0.00148 −0.06435 0.079842 0.242054
13.05024 1 −0.00131 −0.03439 0.040536 0.219392
12.37104 1 −0.00205 −0.00201 0.013029 0.200942
13.85076 0 −0.00288 −0.02182 0.060327 0.072212
13.49164 0 −0.00392 −0.07865 0.052315 0.116761
11.65703 0 −0.00194 −0.0367 0.068249 0.151754
12.45755 1 −0.00638 −0.03708 0.025638 0.202708
13.35725 0 −0.00325 −0.0005 0.016382 0.181212
12.3107 0 −0.00272 −0.06519 0.139852 0.220012
13.11122 0 −0.00186 −0.04364 0.073036 0.168351
12.30098 0 −0.00406 −0.01576 0.034676 0.18917
13.1015 1 −0.00226 −0.03521 0.000739 0.211024
11.56143 1 −0.00212 −0.07403 0.159932 0.171132
12.11965 1 −0.00331 −0.02547 0.047064 0.188637
12.3107 1 −0.00178 −0.03979 0.004465 0.186037
12.92017 1 −0.0024 −0.03588 0.030881 0.137912
13.11122 1 −0.01145 −0.05675 0.032062 0.126376
13.35725 1 −0.00109 −0.00103 0.070427 0.137455
14.89175 1 −0.0011 −0.03374 0.039663 0.232216
13.05024 1 −0.00082 −0.0034 0.045418 0.257901
12.66982 1 −0.00144 −0.07924 0.015991 0.22212
13.19176 1 −0.00178 −0.05341 0.01499 0.144703
13.4038 1 −0.00071 −0.02181 0.126288 0.132103
13.05024 1 −0.00234 −0.11568 0.049628 0.068996
14.11489 1 −0.00252 −0.04659 0.03041 0.110293
13.19176 1 −0.00332 −0.05161 0.007637 0.156499

method, genetic algorithms (GA), and simulated

annealing
(SA) based feature selection procedures are most
popular for
building QSAR models and can explain the situation of �2 . According to Golbraikh and Tropsha, models are
more considered acceptable, if they satisfy all of the following
effectively. The models were also subjected to the test conditions:
for
criteria of external validation as suggested by Golbraikh (1) �2 > 0.5,
and (2) �2 > 0.6.
Tropsha [71]. To know predictive potential of the
models,
squared correlation coefcient values between the When the observed values of the test set compounds
observed (�-axis) are plotted against the predicted values of the
and predicted values of the test set compounds with
intercept
2
(� ) were calculated. Interchange of the axes gives the
value
International Journal of Medicinal 9
Chemistry
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0 SsCH3 T C C1 LUMO SaaSE
−0.1 count 1 energy index
−0.2
−0.3
−0.4

(a) (b) (c)

8.5
Predict
8
ed y = 0.951x + 0.371
activiti
7.5
es
7

6.5

6
6 8

Observed activities

(d) (e)

8.5
Predict
ed 8 y = 0.874x − 0.082
activiti
7.5
es
7
6.5
6
6 8
Observed activities

(f) (g)
8.5
Predict
ed 8
y = 0.912x + 0.144
activiti
7.5
es
7
6.5
6
6 8

Observed activities

(h) (i)
8.5
Predict
ed 8
y = 0.801x + 1.483
activiti
7.5
es
7
6.5
6
6 8
Observed activities

Training set
Test set

(j)

Figure 1: (a) Thiophenyl ring (template structure). (b) Alignment of thiophenyl derivatives. (c) Contribution charts of the
descriptors for
the 2D QSAR model-1. (d) Plot of observed versus predicted activity by 2D QSAR model-1. (e) Contribution plot for steric and
electrostatic
interactions GA-PLS model. (f) Plot of observed versus predicted activity by 3D QSAR GA-PLS model. (g) Contribution plot for
steric and
electrostatic interactions SA-PLS model. (h) Plot of observed versus predicted activity by 3D QSAR SA-PLS model. (i)
Contribution plot for
steric, hydrophobic, and electrostatic interactions SW-PLS model. (j) Plot of observed versus predicted activity by 3D QSAR SW-
PLS model.
 The steric and electrostatic interaction energies are
The signifcant model with �2 = 0.8499 was com-
10
considered, as International
puted at the lattice points Journal
of the gridof Medicinal
using a methyl
model-1 showed an internal predictive power (�2 = probe Chemistry
0.8267)
compounds (�-axis) setting intercept to zero, the slope of thecharge +1. The
correlation bestbetween
matrix GA-kNN the MFAphysicochemical
3D QSAR model
and
of the a predictivity for the external test set (pred � 2= that
param-
0.7729)
ftted line gives the value of �. has
etersa and
�2 ofthe
0.7663 and pred
biological �2 of for
activity 0.7386 was are
model-1 considered.
of about 77%. The �-test = 45.897 shows the statistical The points
presented generated in GA-kNN MFA 3D QSAR model
3. Results and
signifcance Discussion
of 99.99% of the model which means that are S 1044
in Table (−0.0317,
5. The −0.0317),
contribution chartEof184 (−0.2885,
selected descriptors
the −0.2885),
is and
QSAR studyofwas
probability performed
failure on thiophenyl
of the model C-arylIn
is 1 in 10000. S 931 (−0.0306, −0.0306). Figure 1(e) shows
represented in Figure 1(c). Also, the graph for the
observed
glucoside
addition, contribution
versus
derivatives
the for theirtest
randomization SGLT2 inhibitors
shows confdenceas potential
of ∼99.9% plot of theactivity
three models
predicted for thefor the is
series electrostatic and steric
plotted in Figure 1(d)
antidia-
that the felds,
which
betic agents.
generated Comparison
model of the and
is not random obtained
henceresults
it is chosen respectively, and indicates relative regions of the local
shows good correlation.
indicated
as felds
Molecular felds are the steric, electrostatic, and
superiority
the QSAR model. of Genetic
the genetic algorithm over
algorithm-PLS model theindicates around
hydro- the aligned molecules leading to activity
simulated
the pIC50 =
positive 0.6451 (±0.1584)
contribution of SsCHSsCH 3 count
3 count, + 0.4287 variation in
and SaaSE- phobic interaction energies which are used to develop a
annealing (SA) and
index (±0.0851) T Cstepwise forward-backward
Cl 1 − 0.2574 (±0.0036) LUMO variable the model. For electrostatic feld and steric felds, the
model pIC
for 50
3D =QSAR.
S 1044In(−0.0317,
this study,−0.0317)
3D QSAR−models
E 184
method
showed for feature
energy
that increase selection:
in the values of these descriptors lattice were generated
(−0.2885,by kNN-MFA in conjunction with genetic
is + 0.2789 (±0.0487) SaaSE-index; points generated inSthe
algorithms (GA), −
−0.2885) 931model
simulated are E−0.0306);
annealing
(−0.0306, 184 (−0.2885,
(SA), and stepwise
benefcial for =
�training 26,
the �test =
SGLT2 7, degreeFrom
inhibitors. of freedom
the above = 21, �2 −0.2885)� = 26, � = 7, �
(SW) forward-backward selection methods:
training test 2 = 0.7663, �2 se = 0.3389,
model,= and S 1044
�-test(−0.0317,
= 38.4683, −0.0317)
and predand �2 =S 0.7386.
931 (−0.0306,
0.8499,
it is clear �2 =descriptor
that the 0.8267, �-test =145.8975,
T C Cl contributes�2 se = −0.0306).
0.3098,
positively These points suggested the signifcance of electrostatic
�2 se =inhibitors
to the SGLT2 0.3464, pred �2 = 0.7729,
activity, and pred �2 to
which corresponds se = prop-
count 0.6158. erties as indicated in the ranges in parentheses for
of number of carbon atoms separated from any chlorine maximum
atom by 1-bond distance. Thus, the presence of chloro SGLT2 inhibitory activity. The negative value for E 184
substituents (like in compounds 1–18) would increase means
the that electron-withdrawing substituents in this region are
activity. Descriptor SsCH3 count indicates that increase favorable and would increase SGLT2 inhibitory activity,
in as
methyl group of R position may lead to an increase in shown by the presence of chlorine group in the active
the com-
activity. Its positive value suggests that increasing the pounds. Therefore, less steric and more steric
number substituents
of such carbons will lead to better SGLT2 inhibitors. This were preferred at the position of generated data points
suggests that substituents such as methyl would S 1044
increase and S 931, respectively, for enhancing the biological
the activity. The above results are in close agreement activity
with of thiophenyl pharmacophore. Two data points
the experimental observations, where compounds 10, generated
15, 16, at the position of R around thiophenyl nucleus were
17, and 28–33 at the R position produce SGLT2 steric
inhibitors. points S 1044 and S 931 which indicates that less steric
Molecules with negative coefcient LUMO energy can or
accept less bulky substituents are favorable on this site. On the
electrons more easily than those having high LUMO other
energy. hand, less electronegative groups such as hydroxyl and
The SaaSE-index (∼24%) shows the sulphur atom nitro.
connected The electrostatic blue ball model around R positions of
with two aromatic bonds in the molecule and is the
inversely thiophenyl suggested the electron-withdrawing groups
proportional to the activity. This further suggests that on
the this position benefted potency; this may be the reason
increase in electronegative atom environment adjacent why
to compounds with double bonds at R positions had higher
indicated sulphur atom would result in increase in the potencies than other compounds. In addition, a red
activ- contour
ity. In addition, in agreement with QSAR model, near the position suggested the electron-withdrawing
presence of sub-
more bulky and hydrophilic substituents like methoxy at stituent would increase the activity. Therefore, the –OH
ring at R
R led to an increase in SGLT2 potency. On the other position resulted in signifcant increased activity.
hand, Electron-
more lipophilic halogens like chloro in this X position withdrawing nature of the electronegative chloro atom
retain does
the SGLT2 inhibition activities. The residuals (observed- contribute to the SGLT2 inhibitory activity of the
predicted activity) were found to be minimal and are molecule.
pre- The graph for observed versus predicted activity for the
sented in Table 4. The statistical results of best model series is plotted in Figure 1(f) which shows good
and correlation.
International Journal of Medicinal 11
Chemistry
Table 4: Comparative observed and predicted activities (LOO) of thiophenyl SGLT2
inhibitors.
2D model-1 3D model-GA-PLS 3D model-SA-PLS 3D model-SW-PLS
Com pIC50
Pred.Res. Pred.Res. Pred.Res. Pred.Res.
7.1845−0.1216 7.1847−0.1218 7.1718−0.1089 7.05530.0076
1 7.0629 7.4978−0.0369 7.36790.093 7.12830.3326 7.22890.232
2 7.4609 6.8737−0.0199 6.75650.0973 6.9154−0.0616 6.72950.1243
3 6.8538 7.2319−0.0449 7.07890.1081 7.11220.0748 7.10470.0823
4 7.187 7.25640.0064 7.4543−0.1915 7.24280.02 7.3525−0.0897
5 7.2628 6.94480.0098 6.91210.0425 6.93930.0153 6.9875−0.0329
6 6.9546 7.1464−0.1196 7.0723−0.0455 7.1174−0.0906 7.1653−0.1385
7 7.0268 6.9815−0.0422 6.9644−0.0251 6.91630.023 6.91280.0265
8 6.9393 7.1854−0.0349 7.07460.0759 7.1649−0.0144 7.1761−0.0256
9 7.1505 7.82890.0638 7.86940.0233 7.82650.0662 7.86560.0271
10 7.8927 7.3283−0.0132 7.3729−0.0578 7.3247−0.0096 7.18260.1325
11 7.3151 7.91350.0109 7.9918−0.0674 7.9418−0.0174 7.9356−0.0112
12 7.9244 7.2857−0.0431 7.2498−0.0072 7.3821−0.1395 7.21480.0278
13 7.2426 6.78450.0059 6.7954−0.005 6.7991−0.0087 6.78940.001
14 6.7904 7.21380.0066 7.2525−0.0321 7.2862−0.0658 7.19660.0238
15 7.2204 8.3791−0.0295 8.3875−0.0379 8.34240.0072 8.34460.005
16 8.3496 7.98450.0026 7.9949−0.0078 7.93260.0545 7.96740.0197
17 7.9871 7.1278−0.0883 7.1392−0.0997 7.0734−0.0339 7.0673−0.0278
18 7.0395 7.92430.015 7.9418−0.0025 7.93880.0005 7.91320.0261
19 7.9393 7.92750.1314 8.0692−0.0103 8.0872−0.0283 8.1471−0.0882
20 8.0589 7.29610.0031 7.14910.1501 7.14490.1543 7.23070.0685
21 7.2992 7.5884−0.0278 7.42810.1325 7.51610.0445 7.53190.0287
22 7.5606 7.6978−0.0221 7.59460.0811 7.59880.0769 7.59750.0782
23 7.6757 7.13260.0137 7.13840.0079 7.2864−0.1401 7.11250.0338
24 7.1463 7.09680.0649 7.1948−0.0331 7.12890.0328 7.1799−0.0182
25 7.1617 6.32910.0167 6.32590.0199 6.16430.1815 7.4332−0.0874
26 6.3458 7.1853−0.1313 6.93420.1198 6.97680.0772 6.90870.1453
27 7.054 7.2218−0.0645 7.02020.1371 7.2738−0.1165 7.2684−0.1111
28 7.1573 7.6283−0.0228 7.6163−0.0108 7.60230.0032 7.6132−0.0077
29 7.6055 7.2346−0.0114 7.2458−0.0226 7.2282−0.005 7.2547−0.0315
30 7.2232 7.85790.0486 7.88810.0184 7.81940.0871 7.86580.0407
31 7.9065 7.29820.0071 7.04240.2629 7.19210.1132 7.8928−0.5875
32 7.3053 7.47280.0603 7.34660.1865 7.33080.2023 7.27330.2598
33 7.5331

Table 5: Correlation matrix between descriptors present in the best QSAR

model-1.
Parameter pIC50 SsCH3 count T C Cl 1 LUMO energy SaaSE-index
pIC50 1.0000
SsCH3 count 0.5431
T C Cl 1 0.2165 1.0000
LUMO energy 0.4964 0.4312 1.0000
SaaSE-index 0.3731 0.5731 0.6823
0.4633 1.0000
0.6591 0.8504 1.0000

was determined by Pred �2 , which is 0.6938. The points

The residuals (observed-predicted activity) were found gen-
to be erated in SA kNN-MFA 3D QSAR model are E 1139
minimal and are
�training presented
= 26, �test = 7,in�2Table 4:
= 0.7254, �2 se = 0.3795, (0.1545,
�-test = 31.8965, and pred �2 = 0.6938. 0.1545) and S 646 (−0.3001, −0.3001), that is,
electrostatic
and steric interaction felds at lattice points 1139 and S
3D data points generated, which contribute to SA
646,
kNN-
respectively. Positive values for E 1139 show that
MFA QSAR model, are shown in Figure 1(g). The external
electron-
predictability of the above 3D QSAR model using the
donating groups on the thiophenyl ring increase
test set
biological
activity of compounds. On the other hand, thiophenyl
ring
might be substituted with either electron-withdrawing or
 [21] S. Kar and K. Roy, “Predictive toxicology using QSAR: a
perspective,” Journal of the Indian Chemical Society, vol. 87,
no.
12, pp. 1455–1515, 2010.
12
International Journal of Medicinal International Journal of Medicinal 13
[22] M. C. Sharma, D. V. Kohli, S. C. Chaturvedi, and S.
Chemistry Chemistry
Sharma, “Molecular modelling studies of some substitued 2-
butylbenzimidazoles angiotensin II receptor antagonists as
electron-donating
HbA1� levels in db/db groups micewithout loss of
and prolongs theactivity.
survival3D of would lead to better
antihypertensive understanding
agents,” Digest Journal of the QSAR models.
of Nanomaterials
QSAR stroke- The 2D and 3D QSAR suggested the presence of
and
proneshowed
studies rats,” Pharmacological
requirement of Research,
steric groupvol. 63,atno.R 4, pp. negative Biostructures, vol. 4, no. 4, pp. 843–856, 2009.
284–
position. steric
[23] M. potential
C. Sharma,at D.RV.position
Kohli, N. in thiophenyl
K. Sahu, nucleus,
S. Sharma, and S. that
[8] Y . Kanai,
293, 2011. W. S. Lee, G. You, D. Brown, and M. A. Hediger,
“Thegraph for observed versus predicted activity for the is, C. Chaturvedi, “2D-QSAR studies of some 1, 3, 4-thidiazole-
The
series
human kidney low afnity Na+/glucose cotransporter SGLT2. R position 2yl azetidine 2-one as antimicrobial
in thiophenyl nucleus should activity,” Digest
acquire Journal
less
is plotted in Figure 1(h). The residuals (observed-
Delineation of the major renal re-absorptive mechanism for of
steric
D- �training = 26, �test = 7, �2 = 0.7359, �2 se = Nanomaterials and Biostructures, vol. 4, no. 2, pp. 339–347,
predicted or less bulky substituents are favorable as well as
glucose,” The Journal of Clinical Investigation, vol. 93, no. 1, 2009.
activity) 0.3562,
were found�- to be minimal and are presented in [24] according
[9]
pp. O. Marsenic, “Glucoseand control by�the kidney: an emerging M. C. Sharma, D. V. Kohli, S. Sharma, and S. C. Chaturvedi,
Table test = 30.8743,
4: pred 2 = 0.6743. to models. The constructed 3Dstructure
“Two dimensional-quantitative QSAR models activity and
target
397–404,
The 3D model 1994. using the SW-kNN-MFA analysis
in diabetes,” American Journal of Kidney Diseases, vol. 53, structure-
relationships
method activity relationship (SAR)
-2, 3 diarylthiophenes analyses
as selective of theinhibitors,”
cox-1/-2 compounds
no. 5,
�2 was
[10] pp.
found
W.875–883,
S. Lee, Y to be 0.7359
. 2009.
Kanai, whichand
R. G. Wells, suggests that the
M. A. Hediger, “The Digest
used
model
high Journal
in the of Nanomaterials
study suggested that andan Biostructures,
electronegative vol. 4, and
no. 3, pp.
could
afnitybe useful
Na+ /glucosefor predicting
cotransporter. SGLT2 inhibitor for
Re-evaluation of such
function 459–469,
bulky 2009.
and distribution of expression,” The Journal of Biological
thio- [25] A. Dhakad, M. C. Sharma, S. C. Chaturvedi, and S. Sharma,
Chem- substituent at R position and one less bulky substituent
“3D-QSAR studies, biological evaluation studies on some
phenyl
[11]
derivatives.
B. Mackenzie,
These points suggested
M.16,
Panayotova-Heiermann,
the
D. D. Loo, J. E. at X, R position of thiophenyl analogs are required to
istry, vol. 269, no. pp. 12032–12039, 1994. substituted 3-Chloro-1-[5-(5-chloro-2-phenyl-benzimidazole-
signifcance design novel SGLT2
Lever, 1-ylmethyl)-[1, 3, 4]inhibitors. We emphasize in this
thiadiazole-2-yl]-azetidin-2-one as
and requirement
and E. M. Wright,of“SAAT1 electrostatic
is a low andafnitysteric
Na+ and
/glucose study
poten-
hydrophobic
cotrans- thattialthe E 1139 electron-donating
antimicrobial activity,” Digest Journal groups on the
of Nanomaterials
properties
porter and asnotmentioned
an amino acid in the ranges for
transporter. A structure- and
thiophenyl
reinterpretation,”
activity Biostructures, vol. 4, no.activity
2, pp. 275–284, 2009.
[12] E. M.
The Wright
Journal of and E. Turk,
Biological “The sodium/glucose
Chemistry, vol. 269, pp.cotransport
22488– ring increase biological of compounds. Electron-
relationship and biological activity of thiophenyl [26] M. C. Sharma, D. V. Kohali, N. K. Sahu, S. Sharma, and
family SLC5,” Pfl¨ gers Archiv, vol. 447, no. 5, pp. 510–518, withdrawing
22491, groups are highly favorable. Furthermore,
S. C. Chaturvedi, “QSAR, synthesis and biological activity
analogues. thestudies of some thiazolidinones derivatives,” Digest Journal
2004.u1994.
It shows
[13] a positive
K. Katsuno, contribution
Y. Fujimori, Y. Takemura towards
et al.,the activity, a kNN-MFA
“Sergliflozin,
novel
which
Conflict maps
of offered enough information to
of Interests
selective inhibitor of low-afnity sodium glucose understand
Nanomaterials and Biostructures, vol. 4, no. 1, pp. 223–232,
indicates that H 1153 hydrophobic substitution at R the structure-activity
The
2009. authors declare no relationship
conflict of and identifed
cotransporter
position structural
(SGLT2), validates the critical role of SGLT2 in renal glucose [27] M. C. Sharma, S. Sharma, P. Sharma, and A. Kumar, “Study
interests.
favors activity.and
reabsorption Figure 1(i) showed
modulates plasmathe hydrophobic
glucose level,” Journal features of physicochemical
influencing properties-inducible
the inhibitory activity. nitricThe
oxide
contour
of Acknowledgments
synthase
correlation
thePharmacology
[14] presence
B. Mackenzie, of and
aD.yellow
D. Loo,contour
Experimental covering the
Therapeutics,
M. Panayotova-Heiermann, R 320,
vol. andno. relationship of substituted quinazolinamines analogs:
E. of the results obtained from 3D QSAR study successfully
positions
1, pp.
M. Wright, of “Biophysical characteristics of the pig kidney Na +pharma- The authors wish to express gratitude to V-life Science
explored
cophore the primitive and
identifcation structure-activity
QSAR studies,” relationship.
Arabian Journal of
/the323–330,
ring indicating
2007. that hydrophobic substituents may Technologies Pvt. Ltd. for providing the trial version
TheChemistry, 2013.
glucose
be well cotransporter SGLT2 reveal a common mechanism software
for fndings
[28] M. C. can
Sharmabe quite
and D.useful
V. Kohli, to“An
aidapproach
the designing
to design of
tolerated in that region. The R position of the thiophenyl for the study. The authors thank the referees for their
SGLT1 and SGLT2,” The Journal of Biological Chemistry, vol. novel antihypertensive agents by 2D QSAR studies on series of
critical
was sub-
antidiabetic agents with high predicted potent activity.
[15] W. N. Washburn, “Development of the renal glucose
271,
surrounded comments and for suggesting improvements.
no. 51, pp.by
reabsorp- a yellow contour
32678–32683, 1996. which also indicated stituted benzimidazoles derivatives as angiotensin II
thattion inhibitors: a new mechanism for the pharmacotherapy
References
receptor
hydrophobic groups at this position may increase
of [1]antagonists,”
A. Ceriello, “Proactive
Arabian Journal
study: (r)evolution
of Chemistry, in 2011.
the therapy of
diabetes mellitus type 2,” Journal of Medicinal Chemistry, [29]diabetes?”
activity. M. C. Sharma Diabetic
and D. Medicine,
V. Kohli, vol.
“QSAR 22, analysis
no. 11, pp.and1463–
3D
[16]
vol. G.bulky
Less52, Toggenburger,
substituents M. Kessler, and G. Semenza,
are tolerated “Phlorizin
at thiophenyl ringas 1464,QSAR 2005.
kNNMFA approach on a series of substituted
Na+ ,d-glucose cotransporter. quinolines
a
no.
meaningprobe
7, pp. of1785–1794,
thatthe small-intestinal
increasing 2009.size of the groups substituted [2] K. G. Alberti and P. Z. Zimmet, “Defnition, diagnosis and
A classifcation
derivatives asofangiotensin
diabetes mellitus
II receptor
and antagonists,”
its complications. Arabian
Part
in model,” Biochimica et Biophysica Acta, vol. 688, no. 2, pp. 1:
Journal
diagnosis
of Chemistry,
and classifcation
2011. of diabetes mellitus
these regions reduces SGLT2 inhibitory activity, since S [30]
557– provisional
M. C. Sharma and D. V. Kohli, “QSAR studies on substituted
190
QSAR
[17] 571, study
A. Oku,
1982. K.was
Ueta, performed
K. Arakawaon et thiophenyl C-aryl
al., “Antihyperglycemic report
benzimidazoles
of a WHO as consultation,”
angiotensin Diabetic
II receptor Medicine,
antagonists:vol. 15,
kNN-
effect
have negative values. Positive values for the
glucoside pp.MFA approach,” Arabian Journal of Chemistry, 2011.
of T-1095 via
hydrophobic
derivatives for inhibition
1153their SGLT2 of renal NA+ -glucose
inhibitors cotransporters [31]
as potential 539–553,
M. C. Sharma
1998. and D. V. Kohli, “Two dimensional and k-
in
4. Conclusions
lattice
antidia- points around ring indicate that SGLT2 inhibitory Nearest [3] V. Beletate, R. P. El Dib, and A. N. Atallah, “Zinc
streptozotocin-induced diabetic rats,” Biological and supplementa-
neighbor molecular feld analysis approach on substituted
activity
betic
[18]
Pharma-Y. agents.
could Genetic
Fujimori, beKatsuno,
K. increased
algorithmsby substituting
K. Ojima (GA),
et simulated
more
al., “Sergliflozin tri- tion for the prevention of type 2 diabetes mellitus,”
hydrophobic
annealing
etabonate,
ceutical Bulletin, vol. 23, no. 12, pp. 1434–1437, 2000. Cochrane
azolone derivatives: an insight into the structural
groups
(SA), andin stepwise
a selective these
SGLT2 regions.
(SW) forward-backward
inhibitor, Similarly,
improvesthe positive
glycemic selection
values
control in of requirement
Database of Systematic Reviews, vol. 1, Article ID
streptozotocin-induced
elec-
meth- diabetic rats and Zucker fatty rats,” CD005525,
for the angiotensin II receptor antagonist,” Journal of Saudi
European
trostatic
ods Journal
havedescriptors
been of Pharmacology,
employed suggested thevol.
for selection 609,
requirementno. 1–3,
of relevant ofpp. 2007.
Chemical Society, 2011.
148–
electropos-
descriptors. [32]
[4] T.
M.Kuzuya,
C. SharmaS. Nakagawa,
and D. V. Kohli,
J. Satoh
“QSARet al.,
analysis
“Report ofofimidazo
the Com-
[19]154,
Y. Fujimori,
2009. K. Katsuno, I. Nakashima, Y. Ishikawa-Takemura,
itive
Comparison
or sterically
of thebulky
obtained
groups results
at the indicated
position the
of mittee
[4,5-b] onpyridine
the classifcation
substituted and a- phenoxyphenylacetic
diagnostic criteria of acids as
H. Fujikura, and M. Isaji, “Remogliflozin etabonate, in a novel
generated
superiority diabetes
angiotensin II AT1 receptor antagonists,” Journal of Saudi
category of selective low-afnity sodium glucose
data
of thepoint
genetic
cotransporter E 1090 algorithm
aroundover thiophenyl
the stepwise method forfor mellitus,”
pharmacophore Chemical Society,
Diabetes2011.Research and Clinical Practice, vol. 55,
[33]
no. M.
1, C. Sharma and D. V. Kohli, “Predicting substituted 2-
maximum
feature activity. The
(SGLT2) inhibitors, electrostatic
exhibits antidiabetic E efcacy
1090 data point
in rodent
pp.
butylbenzimidazoles
65–85, 2002. derivatives as angiotensin II receptor
models,”2D
gener-
selection. Journal
QSAR of Pharmacology
further revealed and that
Experimental
a specifc [5]antagonists:
T. Kuzuya and 3D-QSAR
A. Matsuda,and pharmacophore
“Classifcation ofmodeling,”
diabetes on the
Therapeu-
ated which indicates sterically bulky groups such as
group Journal
basis of etiologies versus degree of insulin defciency,”
[20] tics,
benzene,
or type vol.
L. Rossetti,327,D.no.
of descriptor 1, pp.
Smith, isG. 268–276,
not I. Shulman,
sufcient 2008.D.toPapachristou,
capture theand true Diabetes
of Saudi Chemical Society, 2011.
R. A. DeFronzo, “Correction of hyperglycemia with phlorizin [34]Care,
methyl,
factors responsible
ethyl, and isopropyl
for the activity
was required
in the set at R ofposition.
inhibitor M. C.vol.
Sharma
20, no.and2,D.pp.V.219–220,
Kohli, “3D1997.QSAR studies of some
normalizes tissues sensitivity to insulin in diabetic rats,”
The
compounds. This study also revealed that SsCH 3 count, sub- [6] D. Porte Jr., “Clinical importance of insulin secretion and its
Journal interaction
stituted imidazolinones
with insulin resistance
derivativesinangiotensin
the treatment II receptor
of type 2
graph
along for observed
of Clinical Investigation,versus vol.predicted
79, no. 5, pp. activity for the1987.
1510–1515, diabetes
antagonists,”
mellitus
World
andApplied
its complications,”
Sciences Journal, vol. 12, no.
series
with LUMO energy and SaaSE-index, forms a powerful
11,
Diabetes/Metabolism
pp.
is plotted in Figure 1(j). The residuals (observed-
tool Research
2129–2134, and2011.
Reviews, vol. 17, pp. 181–188, 2001.
predicted
to improve a QSAR model. This study used T C Cl 1 to [35] [7] W.
M. Zhang,
C. SharmaA. Welihinda,
and D. V. Kohli,
J. Mechanic
“Exploration
et al., “EGT1442,
of quantitativea
activity) were
investigate whether
found a tosimilarity
be minimal based
and set aregeneration
presented in potent structure activity relationship studies on a series of
Table 4.
method substituted
and selective SGLT2 inhibitor, attenuates blood glucose and
 [50] M. C. Sharma and D. V. Kohli, “A comprehensive structure-
activity analysis 2,3,5-trisubstituted 4,5-dihydro-4-oxo-3H-
imidazo [4,5-c] pyridine derivatives as angiotensin II
receptor
antagonists: using 2D- and 3D-QSAR approach,” Medicinal
14 International
Chemistry Research, Journal
vol. 22, no. 2, of
pp.Medicinal
588–605, 2013.
Chemistry
[51] M. C. Sharma and D. V. Kohli, “Comprehensive two and
three-dimensional QSAR studies of 3-substituted 6-butyl-
quinazolinones as angiotensin II AT1 receptor antagonists,” 1,2dihydropyridin-2-ones derivatives as angiotensin II
World Applied Sciences Journal, vol. 12, no. 11, pp. 2111– receptor
2119, 2011. antagonists,” Medicinal Chemistry Research, vol. 22, no. 3,
[36] M. C. Sharma and D. V. Kohli, “3D QSAR kNNMFA approach
pp.
studies on series of substituted piperidin-2-one biphenyl
1107–1123, 2013.
tetrazoles as angiotensin II receptor antagonists,” American-
[52] M. C. Sharma, “Structural requirements of some 2-(1-
Eurasian Journal of Toxicological Sciences, vol. 3, pp. 75–84,
propyl-
2011.
[37] M. C. Sharma and D. V. Kohli, “3D QSAR approach on piperidin-4-yl)-1H-benzimidazole-4-carboxamide derivatives
substituted isoxazolidines derivatives as angiotensin II as poly (ADP-ribose) polymerase (PARP) for the treatment of
receptor cancer: QSAR approach,” Interdisciplinary Sciences:
antagonist,” American-Eurasian Journal of Toxicological Sci- Computa-
ences, vol. 3, no. 2, pp. 85–91, 2011. tional Life Sciences, 2014.
[38] M. C. Sharma and D. V. Kohli, “2D- and 3D- QSAR studies [53] M. C. Sharma, S. Sharma, N. K. Sahu, and D. V. Kohli,
of substituted 4H-pyrido [1, 2-a] pyrimidin-4-ones “QSAR studies of some substituted imidazolinones
angiotensin angiotensin
II receptor antagonists,” European Journal of Toxicological II receptor antagonists using Partial Least Squares
Sci- Regression
[39]ences,
M. C. vol.
Sharma,
3, pp.“Structural insight for (6-oxo-3-pyridazinyl)-
92–100, 2011. (PLSR) method based feature selection,” Journal of Saudi
benzimidazoles derivatives as angiotensin II receptor antag- Chem-
onists: QSAR, pharmacophore identifcation and kNNMFA ical Society, vol. 17, no. 2, pp. 219–225, 2013.
approach,” Journal of Saudi Chemical Society, 2012. [54] M. C. Sharma, “Discovery of potent antihypertensive lig-
ands substituted imidazolyl biphenyl sulfonylureas analogs
[40] M. C. Sharma and D. V. Kohli, “A comprehensive structure- as
activity analysis of 5-Carboxyl Imidazolyl biphenyl Sulfony- angiotensin II AT1 receptor antagonists by molecular
lureas derivatives angiotensin AT1 receptor antagonists: 2D- modelling
and 3D-QSAR approach,” Arabian Journal of Chemistry, studies,” Interdisciplinary Sciences: Computational Life
2012.
[41] M. C. Sharma and D. V. Kohli, “Comprehensive structure- Sciences,
activity relationship analysis of isoxazolinyl and 2014.
isoxazolidinyl [55] M. C. Sharma, D. V. Kohli, and S. Sharma, “Molecular
substituted quinazolinone derivatives as angiotensin II modeling
receptor studies of substituted 2,4,5-trisubstituted triazolinones aryl
[42]antagonists,”
M. C. SharmaJournal
and D.ofV.Saudi
Kohli,Chemical
“QSAR analysis of2012.
Society, 2-alkyl- and
4-(biphenylmethoxy) quinolines as angiotensin II receptor nonaryl derivatives as angiotensin II AT1 receptor
antagonists,” Oxidation Communications, vol. 35, no. 4, pp. antagonists,”
928– Journal of Chemistry, vol. 2013, Article ID 427181, 14 pages,
944, 2012. 2013.
[43] M. C. Sharma and D. V. Kohli, “Insight into the structural [56] M. C. Sharma, “Discovery of new lead pyrimidines
requirement of aryltriazolinone derivatives as angiotensin II derivatives
AT1 receptor: 2D and 3D-QSAR k-nearest neighbor molecular as potential cannabinoid CB1 receptor antagonistic through
feld analysis approach,” Medicinal Chemistry Research, vol. molecular modeling and pharmacophore approach,”
21, Medicinal
no. 10, pp. 2837–2853, 2012. Chemistry Research, vol. 23, no. 5, pp. 2287–2311, 2014.
[44] M. C. Sharma and D. V. Kohli, “Predicting 2,3-Dihydro- [57] M. C. Sharma, “Comparative pharmacophore modeling and
quinazolinonesderivatives as angiotensin II receptor antago- QSAR studies for structural requirements of some
nists: 2D QSAR approach,” Oxidation Communications, vol. substituted
35, 2-aminopyridines derivatives as inhibitors nitric oxide syn-
no. 3, pp. 722–734, 2012. thases,” Interdisciplinary Sciences: Computational Life
[45] M. C. Sharma, S. Sharma, and D. V. Kohlp, “Qsar approach Sciences,
insight the structural requirement of substituted quinazoli- 2014.
nones derivatives as angiotensin ii receptor antagonists,” [58] M. C. Sharma, “A structure-activity relationship study of
Oxi- imidazole-5-carboxylic acids derivatives as angiotensin II
dation Communications, vol. 35, no. 3, pp. 694–707, 2012. receptor antagonists combining 2D and 3D QSAR methods,”
[46] M. C. Sharma, S. Sharma, and D. V. Kohli, “Qsar studies
Interdisciplinary Sciences: Computational Life Sciences,
of 2-alkylbenzimidazole derivatives as angiotensin ii receptor
2014.
antagonists,” Oxidation Communications, vol. 35, no. 3, pp.
[59] M. C. Sharma, S. Sharma, P. Sharma, A. Kumar, and K.
708–
S. Bhadoriya, “Structural insights for substituted acyl sul-
721, 2012.
[47] M. C. Sharma, S. Sharma, and K. S. Bhadoriya, “QSAR anal- fonamides and acyl sulfamides derivatives of imidazole as
yses and pharmacophore studies of tetrazole and angiotensin II receptor antagonists using molecular
sulfonamide modeling
analogs of imidazo[4,5-b]pyridine using simulated annealing approach,” Journal of the Taiwan Institute of Chemical
based feature selection,” Journal of Saudi Chemical Society, Engineers,
[48]
2012. M. C. Sharma and D. V. Kohli, “Rationalization of 2-alkyl- vol. 45, no. 1, pp. 12–23, 2014.
benzimidazoles bearing a N-phenyl pyrrole moiety as novel [60] M. C. Sharma, “Molecular modelling studies for the
Angiotensin II AT1 receptor antagonists—a QSAR approach,” discovery
Oxidation Communications, vol. 36, no. 1, pp. 190–204, of new substituted pyridines derivatives with angiotensin II
2013. AT1
[49] M. C. Sharma and D. V. Kohli, “Quantitative structure- receptor antagonists,” Interdisciplinary Sciences,
activity Computational
relationship analysis of series of substituted piperidin-2-one Life Sciences, vol. 6, no. 3, pp. 197–207, 2014.
biphenyl tetrazoles analogues as novel angiotensin ii [61] M. C. Sharma, S. Sharma, N. K. Sahu, and D. V. Kohli,
receptor “3D QSAR kNN-MFA studies on 6-substituted benzimidazoles
antagonists,” Oxidation Communications, vol. 36, no. 1, pp. derivatives as nonpeptide angiotensin II receptor
176– antagonists: a
189, 2013. rational approach to antihypertensive agents,” Journal of
Saudi
Chemical Society, vol. 17, no. 2, pp. 167–176, 2013.
[62] M. C. Sharma, “Prospective QSAR-based prediction mod-
International Journal of Medicinal 15
Chemistry

[63] M. C. Sharma, S. Sharma, P. Sharma, and A. Kumar, “Phar-

macophore and QSAR modeling of some structurally diverse
azaaurones derivatives as anti-malarial activity,” Medicinal
Chemistry Research, vol. 23, no. 1, pp. 181–198, 2014.
[64] M. C. Sharma, “Rationalization of physicochemical
characters
and pharmacophore approach of 3-(4-biphenylmethyl) 4, 5-
dihydro-4-oxo-3H-imidazo [4, 5-c] pyridines analogs toward
angiotensin II AT receptor antagonists,” Interdisciplinary Sci-
ences: Computational Life Sciences. In press.
[65] M. C. Sharma, “Molecular modeling studies of substituted
3,4-
dihydroxychalcone derivatives as 5-lipoxygenase and
cyclooxy-
genase inhibitors,” Medicinal Chemistry Research, vol. 23,
no. 4,
pp. 1797–1818, 2014.
[66] M. C. Sharma, “A structure-activity relationship study of
naph-
thoquinone derivatives as antitubercular agents using
molecu-
lar modelling techniques,” Interdisciplinary Sciences:
Computa-
tional Life Sciences. In press.
[67] M. C. Sharma, “Identifcation of 3-nitro-2, 4, 6-trihydroxy-
benzamide derivatives as photosynthetic electron transport
inhibitors by QSAR and pharmacophore studies,” Interdisci-
plinary Sciences: Computational Life Sciences, 2014.
[68] S. H. Lee, K. S. Song, J. Y. Kim et al., “Novel thiophenyl C-
aryl
glucoside SGLT2 inhibitors as potential antidiabetic agents,”
Bioorganic and Medicinal Chemistry, vol. 19, no. 19, pp.
5813–
5832, 2011.
[69] VLife MDS 3.5, Molecular Design Suite, Vlife Sciences Tech-
nologies Pvt. Ltd., Pune, India, 2008,
http://www.vlifesciences
.com/.
[70] T. A. Halgren, “Merck molecular force feld. III. Molecular
geometries and vibrational frequencies for MMFF94,” Journal
of Computational Chemistry, vol. 17, no. 5-6, pp. 553–586,
1996.
[71] A. Golbraikh and A. Tropsha, “Predictive QSAR modeling
based on diversity sampling of experimental datasets for
the
training and test set selection,” Journal of Computer-Aided
Molecular Design, vol. 16, no. 5-6, pp. 357–369, 2002.
[72] S. Ajmani, K. Jadhav, and S. A. Kulkarni, “Three-dimensional
QSAR using the k-nearest neighbor method and its
interpreta-
tion,” Journal of Chemical Information and Modeling, vol. 46,
no.
1, pp. 24–31, 2006.
[73] R. D. Cramer III, D. E. Patterson, and J. D. Bunce,
“Comparative
molecular feld analysis (CoMFA). 1. Effect of shape on
binding
of steroids to carrier proteins,” Journal of the American
Chemical
Society, vol. 110, no. 18, pp. 5959–5967, 1988.
 Photoener
International Journal of
Organic Chemistry
International Journal of International Journal of Advances in
Medicinal Chemistry International Analytical Chemistry Physical Chemistry
gy
Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

International Journal of

Carbohydrate Journal of
Chemistry Quantum Chemistry
Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Submit your manuscripts

at
http://www.hindawi.com
Journal of
The Scientifc Analytical Methods
World Journal in Chemistry
Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of International Journal of International Journal of Journal of Bioinorganic Chemistry

Spectroscopy Inorganic Chemistry Electrochemistry Applied Chemistry and Applications
Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of Chromatography Journal of Journal of International Journal of

Theoretical Chemistry Research International Catalysts Chemistry Spectroscopy
Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014