Virtual screening and synthesis of new chemical scaffolds as VEGFR-2 kinase inhibitors

Article abstract of DOI:10.1055/s-0032-1323759

Background: VEGFR-2 tyrosine kinase inhibitors are currently receiving high interest in drug discovery process as anticancer agents. We have used virtual screening techniques in order to discover new scaffolds that can be used for developing new VEGFR-2 k

Full text of DOI:10.1055/s-0032-1323759

554 Original Article
Virtual Screening and Synthesis of New Chemical
Scaffolds as VEGFR-2 Kinase Inhibitors
Authors
M. S. Elsayed1, M. E. El-Araby2, R. T. Serya1, K. A. M. Abouzid1
Affiliations
1 Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Ainshams University, Abbasia, Cairo, Egypt
2 Department of Organic Chemistry, Faculty of Pharmacy Helwan University, Helwan, Egypt
Key words
Abstract
and adopted to screen the selected subset. Initial
▶
SEA prediction
pharmacophore
docking
●
hits mapped to the pharmacophore were filtered
using docking and scoring. Selected compounds
were synthesized and biologically tested.
Results: Compound 9 showed very good cyto-
toxicity profile against the NCI 60 cancer cell
lines, while compound 8 showed reasonable
inhibition of VEGFR-2 tyrosine kinase.
▼
▶
●
Background: VEGFR-2 tyrosine kinase inhibi-
tors are currently receiving high interest in drug
discovery process as anticancer agents. We have
used virtual screening techniques in order to dis-
cover new scaffolds that can be used for develop-
ing new VEGFR-2 kinase inhibitors.
▶
●
▶
●
VEGFR-2
Method: Similarity ensemble approach was
used to reduce the chemical space of ZINC data-
base to select a subset of compounds. A validated
structure-based pharmacophore was developed
Conclusion: Stepwise virtual screening of data-
bases such as ZINC may result in new scaffolds
for developing VEGFR-2 kinase inhibitors.
molecule inhibitors has been proved as an effec-
tive strategy for the treatment of cancer. Inhibi-
tion of this enzyme was associated with the
reduction of the tumor mass and tumor cells pro-
gression as a result of blood supply shortage [6].
FDA has already approved several small molecule
inhibitors of VEGFR-2 for the treatment of differ-
ent types of cancers [7]. Current inhibitors of
VEGFR-2 tyrosine kinase are adopting different
chemical scaffolds. Quinazoline, quinoline, pyri-
dine, pthalazine, benzimidazole, and indolinone
–based scaffolds were used in the synthesis of
VEGFR-2 inhibitors [8].
In this study we have used several virtual screen-
ing computational techniques in order to dis-
cover new scaffolds for developing VEGFR-2
inhibitors. A subset of compounds from ZINC
database that was predicted by SEA prediction to
have VEGFR-2 inhibitory activity was obtained.
SEA (similarity ensemble approach) [9] predic-
tion is a service provided by ZINC database indi-
cates the expected biological activity of some
chemical structure of this massive library. The
subset was filtered using a validated structure-
based pharmacophore built based on the ligand
protein interactions of a subnanomolar VEGFR-2
inhibitor having napthamide scaffold [10]. The
compounds mapped to the pharmacophore were
further reduced using molecular docking and
Introduction
▼
Selective chemotherapeutic agents are the main
target of anticancer drug discovery process these
days. Several strategies were adopted in order to
discover new selective anticancer agent of rea-
sonable safety profile [1]. One of the major
strategies currently receiving high interest, is tar-
geting the neovascularization process induced by
the tumor tissues. The rapid growth rate of can-
cer cells and their massive consumption of nutri-
ents and oxygen lead to their starvation after
short time of progression [2]. The consequences
of this phenomenon will lead the center of the
tumor mass to death. As a defense mechanism
cancer cells produce what's called vascular
endothelial growth factors. These growth factors
have the ability to induce neovascularization
process by interacting with some receptors on
the endothelial cells of the nearby blood vessels
pushing them to divide and form new capillaries
towards the tumor mass [3,4]. VEGFR-2 (vascular
endothelial growth factor receptor 2) plays a
major role in this process. VEGFR-2 is a tyrosine
kinase receptor involved in many signal trans-
duction pathways in the endothelial cells of vas-
cular system controlling their division and
progression [5]. Targeting the intracellular tyro-
sine kinase domain of this receptor using small
received 01.06.2012
accepted 13.08.2012
Bibliography
DOI http://dx.doi.org/
10.1055/s-0032-1323759
Published online:
September 21, 2012
Arzneimittelforschung 2012;
62: 554–560
© Georg Thieme Verlag KG
Stuttgart · New York
ISSN 0004-4172
Correspondence
Khaled A. M. Abouzid
Department of Pharmaceutical
Chemistry
Faculty of Pharmacy
Ainshams University Abbasia
Cairo
Egypt
Tel.:+2012/221 65624
Fax: +202/250 80728
Khaled.abouzid@pharm.asu.
edu.eg
Elsayed MS et al. Virtual screening and Synthesis… Arzneimittelforschung 2012; 62: 554–560

Original Article 555
will be mentioned. 6 features were added to the query root of
Fig. 1 An overview of
the screening model
used in this study.
the pharmacophore CATALYST [13]. Ligand pharmacophore
mapping algorithm integrated in Accelry’s discovery studio soft-
ware was used for screening of the dataset. The compounds
were mapped using rigid mapping method with zero maximum
omitted features and the fit values were recorded.
Docking and scaffold selection
Docking was the next stage after the pharmacophore filtration.
GOLD 4.1[14] was used for the docking purpose. The kinase struc-
ture was prepared using protein preparation protocol of Accelry’s
discovery studio; the protein structure was protonated, the water
molecules were deleted and the structure was minimized using
500 steps of conjugate gradient minimization method. The
mapped ligands were docked using 50000 iterations and they
were scored using ChemScore scoring function in the kinase
mode of GOLD. The docked ligands were arranged according to
their scores and the top 10% in the docking score was visually
inspected.
In order to select a suitable scaffold for synthesis and biological
evaluation, reported molecules as VEGFR-2 kinase inhibitors
were neglected, compounds inspired by scaffolds commonly
used in the design of kinase inhibitors were also neglected and
finally synthetic feasibility was the last factor to select the new
scaffold.
scored using ChemScore scoring function. Visual inspection of
the resulted hits directed us to novel benzo[d]isothiazole
1,1-dioxide scaffold. Some derivatives of this scaffold and bio-
isosterically related isoindole scaffold were synthesized and bio-
logically evaluated against VEGFR-2 tyrosine kinase beside some
other kinases. 3 compounds were selected by NCI screening pro-
Results and Discussion
▼
gram [11] and their cytotoxic activities were evaluated against
Screening model construction and validation
▶
60 different cancer cell lines. Fig. 1 illustrates the screening
●
Virtual screening has been used to discover many hits for various
biological targets. Virtual screening approach in drug discovery
depends on a computational technique that can reduce the mas-
sive chemical space of structures in the direction of a certain bio-
logical target. Ligand-based approaches of virtual screening use
the data obtained from the chemical structures of ligands acting
on the biological target to obtain a novel compound of similar
activity, Meanwhile structure-based approaches relies on the
information obtained from the 3D structure of the binding site of
the biological target [15]. Hybrid techniques also have been devel-
oped in order to use both sources of information. In this study we
have used a hybrid between ligand-based and structure-based
approaches of virtual screening. SEA prediction is a ligand-based
strategy that was used as a very fast prediction tool followed by a
structure-based pharmacophore and molecular docking.
SEA prediction or similarity ensemble approach was developed
by Michael et al. [9]. as a comparison tool that depends on the
similarity between the chemical natures of ligands. This tech-
nique was developed based on 65000 ligands annotated for drug
targets and it was originally made to explore the polypharma-
cology exerted by some chemical compounds in the biological
system. It was tailored as a service that can give a prediction of
the biological activity of any chemical compound. This tech-
nique was adopted to reduce the chemical space of the ZINC
database to those structures predicted to have VEGFR-2 inhibi-
tory activity. Approximately 150060 compounds were predicted
by SEA to be inhibitors of VEGFR-2 tyrosine kinase.
process used in this study.
Materials and Methods for the Screening Model
▼
Ligands preparation and conformations generation
SEA predicted subset of compounds to have VEGFR-2 inhibitory
activity was obtained from ZINC database website. The ligands
structures were prepared according to the ligand preparation
protocol of Accelry’s Discovery Studio 2.5.5 [12]. According to
this protocol, structures of ligands were corrected to get rid of
errors and bad valences. No tautomers or isomers were gener-
ated for each ligand; structures were protonated at pH 7.4 and
saved. For each ligand 255 conformations were generated with
energy threshold of 20kcal/mol using FAST conformation gen-
eration method.
Pharmacophore generation and ligands mapping
A structure-pharmacophore was manually built using the phar-
macophore building tools found in Accelry’s discovery studio
software CATALYST modules. The reported interactions between
naphthamide based potent VEGFR-2 inhibitor (PDB ID: 3B8Q)
and the protein were used in the pharmacophore building proc-
ess. The protein structure was downloaded from PDB website
and it was prepared according to protein preparation protocol in
Accelry’s discovery studio 2.5. Missing residues and atoms were
corrected and then water molecules were deleted and hydrogen
atoms were added according to role based criteria. The protein
structure was minimized using 500 steps of conjugate gradient
minimization algorithm to remove any clashes in the structure,
then it was used for pharmacophore building and docking as
In order to reduce the number of this subset, a rapid validated
technique of screening was used. Structure-based pharmacoph-
ore is a technique developed by Wolber and Langer [16] as a
screening tool to represents an alternative for computationally
expensive molecular docking. The technique depends on the
Elsayed MS et al. Virtual screening and Synthesis… Arzneimittelforschung 2012; 62: 554–560

556 Original Article
Fig. 2 a Crystal structure of a potent VEGFR-2 kinase inhibitor together with the constructed pharmacophore; b The pharmacophore used in the screening
process consists of 6 features namely 2 hydrogen bond acceptors HBA, 2 hydrophobic aromatic features HA, one ring aromatic RA and one hydrogen bond
donor HBD feature; c mapping of compound 5a (ZINC25418828) with the generated pharmacophore.
from ZINC database for common biological targets to be used in
validation purposes of docking but it can be also used in phar-
macophore validation especially structure-based pharmacoph-
ores. Enrichment factor (EF) is a form of pharmacophore
validation and it represents the ability of a pharmacophore to
accurately recognize active hits [23]. The enrichment factor (EF)
of the pharmacophore was calculated using the following equa-
tion:
TP
TP + FP n
N
EF =
Where (TP) is the number of true positives, (FP) is the number of
false positives, (N) is the number of the total compounds and (n)
is the number of the total positive compounds.
This pharmacophore showed an enrichment factor of 28.3 at 2%.
Furthermore receiver operating characteristic curve (ROC) met-
ric was calculated for this pharmacophore. This validation tech-
nique allows determination of the ability of the pharmacophore
to retrieve active compounds i.e., its sensitivity and in the same
time the ability of the pharmacophore to exclude inactives i.e.
specificity [24].
Fig. 3 The receiver operating characteristic curve (ROC) generated for
the pharmacophore (AUC=0.922).
extraction of the ligand protein interactions in the form of phar-
macophoric features. Structure-based pharmacophore can be
built either automatically using some algorithms that extract
the interactions between the ligand and the protein [17] or
manually based on the reported SAR and interactions. The lig-
and-protein interactions of a potent VEGFR-2 inhibitor based on
N
AUC = Se(x)[(1᎐ Sp)(x) ᎐ (1᎐ Sp)(x ᎐ 1)]
∑
x=2
naphthamide scaffold co-crystalized with VEGFR-2 tyrosine
▶
kinase domain (PDB ID: 3B8Q) [10] ( Fig. 2a) were used to
●
manually build a structure-based pharmacophore.. As shown
Se(x) is the percent of the true positives vs. the total positives at
rank position x, (1−Sp)(x) is the percent of the false positives vs.
the total negatives at rank position x.
The ROC curve is a function of (1−Specificity) vs. the Sensitivity,
and the area under the ROC curve (AUC) is the important way of
measuring the performance of the test. The closer the area under
the curve to the unity, the model is more likely to retrieve actives
▶
in Fig. 2b it was constructed from 6 different pharmacophoric
●
features based on the reported important interactions with the
kinase.
A pharmacophore cannot be used for screening purpose without
testing its ability to retrieve active compounds and to exclude
inactive. In order to quantitatively and statistically validate this
pharmacophore model, it was challenged against 2 sets of active
compounds and decoys. The inhibitor used in the pharmacoph-
ore construction is from type 2 kinase inhibitors i.e., it is target-
ing the inactive conformation of the kinase and hence it was
validated using a diverse active dataset of type 2 inhibitors [18].
A diverse dataset of 41 reported VEGFR-2 tyrosine kinase type 2
inhibitors were retrieved from the literature and deployed as the
active set [19–21]. As the decoy set, 2906 of VEGFR-2 kinase
decoys were used, this decoy dataset was obtained from the
directory of useful decoys (DUD) database [22]. DUD was created
and to exclude inactives. This pharmacophore showed area
▶
under the curve of 0.92 and its ROC curve is shown in Fig. 3.
●
Only 1675 compounds mapped the pharmacophore from the
subset predicted to have VEGFR-2 inhibitory activity by SEA.
Shape or exclusion volume constrains were not added to the
pharmacophore in order to keep its scope as large as possible. An
additional screening technique should be used in order to filter
these compounds that can’t fit the binding site of the kinase.
Molecular docking is based on 2 main complementarities
between ligand and the protein; namely the shape and the phys-
Elsayed MS et al. Virtual screening and Synthesis… Arzneimittelforschung 2012; 62: 554–560

Original Article 557
icochemical complementarities [23]. Mentioning that this phar-
macophore was built inside the binding site of the kinase; the
coordinates of the pharmacophore are the same of the binding
site. Molecular docking was used just to exclude the non-fitting
compounds. Only 256 compounds could fit inside the kinase
binding site with reasonable docking score.
After careful Visual inspection of the resulted hits and after
exclusion of the reported compounds from the hits, a novel hit
ZINC ID: ZINC25418828 (compound 5a) based on benzo[d]iso-
rine was chlorinated using phosphorous pentachloride by fusion
to give 3-chloropseudosaccharine 4. Compounds 5a–c were
obtained by reacting amine derivatives (3a-c)with compound 4
in refluxing dioxane and purification of the resulted precipitate
[26].
Carbonyl group are historically recognized as a biosiostere for
the sulfoxide group [27]. Based on the previous fact 2 derivatives
adopting the isoindolinone scaffold instead of benzo[d]isothia-
zole 1,1-dioxide scaffold were synthesized in a similar fashion to
test the effect of this biosisteric replacement. It has been also
known that biphenyl urea is a very common fragment used in
type 2 VEGFR-2 kinase inhibitors [28]. Fragment 7 [1-(4-
aminophenyl)-3-(2,4-dichlorophenyl)urea] is a biphenyl urea
derivative that was used in this scheme to test the effect of this
thiazole 1,1-dioxide scaffold was chosen for synthesis and bio-
▶
logical evaluation. Fig. 2c shows mapping of this compound
●
▶
●
with the pharmacophore model.
Fig. 4 shows the binding
mode of this compound inside the active site of VEGFR-2 kinase.
Chemistry
active fragment on the biological activity of this series. Accord-
▶
In order to synthesize the hit compound 5a (ZINC25418828) and
2 other derivatives (5b and 5c) the steps illustrated in Fig. 5
ing to Fig. 6, phthalimide was chlorinated using phosphorous
●
▶
●
pentachloride by fusion at 150°C to give the chloroderivative 6
which was reacted with 2 different fragments (3a and 7) in
refluxing dioxane to give compounds 8 and 9. These compounds
were mapped against the pharmacophore and they were docked
inside the VEGFR-2 kinase binding site and showed excellent
results as compounds 5a–c.
were adopted. First p-Nitrobenzoyl chloride was reacted with
different aromatic amines in DCM in presence of Triethylamine
in a typical amide formation to give compounds 2a–c [25]. The
amide derivatives were reduced using Pd/charcoal catalytic
hydrogenation to produce the amino derivatives 3a–c. Saccha-
Fig. 4 Binding mode
of compound 5a in
VEGFR-2 tyrosine
Biological evaluation
The structures of the synthesized compounds were submitted to
the NCI (National cancer institute Bethesda, Maryland, USA)
anticancer screening program [11]. 3 compounds representing
this work were selected for initial 10μM one dose percent inhi-
kinase binding site.
bition assay on 60-tumor cell line panel. Results of this initial
▶
screening are presented in Table 1.
●
B o t h 5b and 5c exhibited very low cytotoxic effect against the vari-
ous types of cell lines. Compound 9 showed very good activity
against different types of cell lines for example the growth of differ-
Fig. 5 Synthetic pathway for compounds 5a-c;
Reagents and conditions: a Ar-NH2, TEA in DCM,
rt overnight; b Pd/H2 EthAc/2 bars 1h; c Dioxane,
reflux 3h; d PCl5, reflux 180°C 1h.
Elsayed MS et al. Virtual screening and Synthesis… Arzneimittelforschung 2012; 62: 554–560

558 Original Article
Table 1 percent growth inhibition exerted by compounds 5b, c and 9
against the NCI 60 cancer cell lines.
ent types of leukemia cell lines was inhibited with values between
33–75% giving an indication that this compound may be a promis-
ing lead for developing anti-leukemic drug. The compound also
showed very good activity against HCT-116 colon cancer cell line
whose growth was inhibited by 95% although VEGFR-2 kinase is not
over expressed in this kind of cell lines, [6] indicating that this com-
pound may have other mechanism of action. Regarding melanoma
this compound inhibited the growth of LOX IMVI cell line with 74%.
The enzymatic activities of the synthesized compounds were
Compound/NSC no
Cell Name
5b
762535
5c
762536
9
762537
Leukemia
CCRF-CEM
HL-60(TB)
MOLT-4
RPMI-8226
SR
7
−11
−13
−2
9
−6
−3
14
4
61
34
75
6
10
34
tested against VEGFR-2 tyrosine kinase using radiolabeled ATP
Non-Small Cell Lung Cancer
A549/ATCC
EKVX
▶
6
−6
1
NT
0
−13
−3
−4
9
3
−2
−4
NT
5
−5
−2
−3
7
30
24
15
7
20
53
50
45
30
assay technique to show their inhibitory activity.
Table 2
●
shows the percent inhibition exerted by our compounds against
some tyrosine kinases including VEGFR-2 tyrosine kinase. Com-
pound 8 showed promising inhibition against VEGFR-2 tyrosine
kinase whose activity was inhibited by 30%. The promising
activity of compound 8 which is a classical bioisostere of com-
pound 5a encouraged us to improve this compound. The isoin-
dolinone was proved to be a better scaffold than benzo[d]
isothiazole1,1-dioxide in the enzymatic activity. In order to
improve compound 8, a biphenyl urea fragment compound 7
[1-(4-aminophenyl)-3-(2,4-dichlorophenyl)urea ] was added to
the isoindolinone scaffold instead of the amide fragments 3a.
Biphenyl urea is a common fragment used in type 2 VEGFR-2
kinase inhibitors. Unfortunately compound 9 didn’t show any
improvement in activity against VEGFR-2 kinase but showed
very promising activity against cell lines indicating that, this
compound may have other mechanism of action. Compound 8
may be suitable for lead optimization process as it can be con-
sidered as a lead compound. The enzymatic activity of the rest of
the synthesized compounds ranged from weak to moderate
inhibition. The enzyme inhibition activity of the synthesized
compounds was not as expected from the screening model.
HOP-62
HOP-92
NCI-H226
NCI-H23
NCI-H322M
NCI-H460
NCI-H522
Colon Cancer
COLO 205
HCC-2998
HCT-116
HCT-15
HT29
KM12
SW-620
CNS Cancer
SF-268
−13
−13
0
5
3
−21
−3
−2
0
0
11
−2
42
33
95
27
49
39
43
2
−12
4
8
−5
7
9
−4
−2
−6
5
1
9
18
27
2
8
15
77
SF-295
SF-539
SNB-19
SNB-75
U251
−2
Melanoma
LOX IMVI
MALME-3M
M14
MDA-MB-435
SK-MEL-2
SK-MEL-28
SK-MEL-5
UACC-257
UACC-62
Ovarian Cancer
IGROV1
O V C A R - 3
O V C A R - 5
OVCAR-8
NCI/ADR-RES
SK-OV-3
Renal Cancer
786-0
A498
A C H N
CAKI-1
RXF 393
SN12C
TK-10
UO-31
−5
−10
−5
−3
−4
−8
4
−4
2
0
0
0
0
0
0
4
75
19
14
19
−6
−6
5
Conclusion
▼
A virtual screening model was developed based on some tech-
niques of database searching. A combination of ligand-based
approaches represented by SEA prediction, structure-based
approaches represented by Docking and hybrid technique as
structure-based pharmacophore were used to construct the
model. The screening process resulted in a new building scaffold
for developing VEGFR-2 kinase inhibitors based on benzo[d]iso-
thiazole-1,1-dioxide and this scaffold was bioisostered to isoin-
dole scaffold. Although the enzymatic activity of the synthesized
compounds were not satisfactory, compound 9 showed very
4
−1
3
8
0
−8
−16
0
1
6
−14
−9
−4
4
−3
7
18
13
37
11
52
26
4
2
−1
−3
−17
2
−2
8
−1
−11
0
5
−8
5
21
19
31
18
11
22
24
20
3
3
Prostate Cancer
PC-3
DU-145
5
−6
0
−9
40
43
Breast Cancer
MCF7
MDA-MB-231/ATCC
HS 578T
BT-549
T-47D
12
−14
−5
9
3
3
10
−2
−1
11
−2
−7
0
41
48
15
19
23
45
30
Fig. 6 Synthetic pathway for compounds 8 and 9; Reagents and condi-
tions: a PCl5, 150°C, 0.5h; b 7 or 3a in Dioxane, 3h reflux.
MDA-MB-468
MEAN % Growth inhibition
−1
Elsayed MS et al. Virtual screening and Synthesis… Arzneimittelforschung 2012; 62: 554–560

Original Article 559
6.99 (d, J=8.1Hz, 1H).; 13C NMR (126MHz, DMSO-d6) δ 164.62,
Table 2 Enzyme inhibition activity of the synthesized compounds.*
157.00, 140.61, 135.34, 135.06, 133.87, 133.40, 130.52, 129.62,
128.96, 128.69, 128.44, 128.18, 127.97, 127.42, 126.94, 124.07,
121.52; HR-MS (ESI): m/z Calculated for (C20H14ClN3O3S) [M-H]-
410.0366, found 410.0347.
VEGFR-2
Aurora A
B-RAf
FLT-3
5a
13%
1 7 %
1 5 %
30%
19%
73%
3%
N T
4%
N T
N T
NT
7%
NT
20%
N T
5b
5c
5 %
9%
4%
65%
N T
8
18%
NT
4-[(1,1-dioxidobenzo[d]isothiazol-3-yl)amino]-N-(4-methoxyphe-
nyl)benzamide (5c): The product was separated as white powder
(0.29g, 70%); m.p=265–266°C; 1H NMR (500MHz, DMSO-d6) δ
10.20 (s, 1H, NH), 10.09 (s, 1H, NH), 8.76 (d, J=7.3Hz, 1H, ArH),
8.19–8.07 (m, 3H, ArH), 8.01–7.85 (m, 2H, ArH), 7.80–7.59 (m, 2H,
ArH), 7.24 (d, J=8.3Hz, 2H, ArH), 6.94 (m, 2H, ArH), 3.75 (s, 3H,
OCH3); 13C NMR (126MHz, DMSO-d6) δ 164.35, 156.98, 155.69,
140.59,140.27, 133.82, 132.35, 131.60, 129.14, 128.49, 128.27,
124.19, 122.13, 121.45, 119.75, 113.70, 55.15.; HR-MS (ESI): m/z
Calculated for (C21H17N3O4S) [M-H]- 406.0862, found 406.0848.
9
staurosporin
NT
(*: data represented as the mean of 2 separate values of % inhibition exerted by the
tested compounds at 10μM conc, (NT) indicates No treatment)
good cytotoxicity profile against the NCI 60 cancer cell lines
which can validate the model in active molecules retrieval.
Experimental
3-chloro-1H-isoindol-1-one (6): (5g, 30mmol) of phthalimide
was mixed in a glass mortar with 10g (48mmol) of phosphorus
pentachloride, the mixture was transferred to 250mL flask and
heated in an oil bath at 150°C for about 0.5h. The phosphorus
oxychloride formed was distilled off and the residue was poured
on cruched ice 100g and the solid was extracted with chloro-
form (30mL). The organic layer was dried over anhydrous
sodium sulphate and evaporated under vacuum to give 6 as
orange residue which was used as it is in the next reaction.
▼
Chemistry
All chemicals used were purchased from Aldrich (USA) and used
without further purification. Melting points were determined in
one end open capillary tubes using Stuart Scientific apparatus
and were uncorrected. 1H spectra were run at 500MHz and 13
C
spectra were run at 126MHz in dimethylsulfoxide (DMSO-d6).
Chemical shifts are quoted in δ and were related to that of the
TMS. High resolution mass spectrometry (HR-MS) was per-
formed on an LTQ FTMS (ThermoFisher Scientific, Bremen, Ger-
many) equipped with an electrospray source (ESI-FT) with
1ppm accuracy. Analytical thin layer chromatography (TLC) was
performed on silica gel 60 F254 packed on Aluminium sheets,
purchased from Merck (Merck, Darmstadt, Germany), the devel-
oping solvents were DCM/MeOH (9:1), with visualization under
U.V. light (254nm). Compounds 2a–c, 3a–c [25,29,30], 4 [31]
and 7 [32,33] were synthesized according to the literature pro-
cedures and showed similar spectral data.
1-(4-aminophenyl)-3-(2,4-dichlorophenyl)urea (7): The prod-
uct was separated as brownish yellow powder (0.8g, 60.1%);
m.p >300°C; 1H NMR (500MHz, DMSO-d6) δ 8.97 (s, 1H), 8.28–
8.13 (m, 2H), 7.59 (d, J=2.5Hz, 1H), 7.36 (dd, J=9.0, 2.5Hz, 1H),
7.16–6.99 (m, 2H), 6.58–6.45 (m, 2H), 4.83 (s, 2H).; 13C NMR
(126MHz, DMSO-d6) δ 152.17, 144.39, 135.65, 128.41, 127.97,
127.53, 125.40, 122.00, 121.62, 120.70, 120.43, 114.13; HR-MS
(ESI): m/z Calculated for (C13H11Cl2N3O) [M-H]- 294.0201, found
294.0204.
General procedure for synthesis of compounds 5a–c,
8 and 9
4-[(1-oxo-1H-isoindol-3-yl)amino]-N-(m-tolyl)benzamide
(8): This compound was prepared according to the same pro-
cedure used for the synthesis of compounds (5a–c). The product
was separated as yellow crystals (0.16g, 65%); m.p=195–196°C;
1H NMR (500MHz, DMSO-d6) δ 10.32 (s, 1H,NH), 9.74 (s, 1H,
NH), 8.20–8.06 (m, 4H, ArH), 8.07–7.88 (m, 3H, ArH), 7.70–7.47
(m, 2H, ArH), 7.34–7.09 (m, 3H, ArH), 2.29 (s, 3H, CH3).; 13C NMR
(126MHz, DMSO-d6) δ 169.23, 137.73, 137.62, 134.82, 134.34,
134.21, 132.57, 131.51, 129.29, 128.68, 126.94, 123.56, 123.42,
122.91, 120.90, 120.66, 117.55, 113.26, 21.21; HR-MS (ESI): m/z
Calculated for (C22H17N3O2) [M-H]- 354.1243, found 354.1239.
Compound 4,6 (1 mmol) was dissolved in dioxane 5mL and a
solution of (amines 3a-c,7) 1 mmol in 5mL dioxane was added
drop wise while stirring to the previous solution. The mixture
was heated under reflux for 3h and cooled. The resulted solid was
filtered washed with dioxane twice and washed with diethyl
ether and dried to give compounds (5a–c, 8, 9) in a pure form.[26]
4-[(1,1-dioxidobenzo[d]isothiazol-3-yl)amino]-N-(m-tolyl)ben-
zamide (5a): The product was separated as white powder
(0.29g, 74%); m.p=225° C; 1H NMR (500MHz, DMSO-d6) δ
11.41 (s, 1H, NH), 10.25 (s, 1H, NH), 10.15 (s, 1H, ArH), 8.79 (d,
J=7.1Hz, 1H, ArH), 8.12 (m, 3H, ArH), 7.98 (d, J=8.0Hz, 1H, ArH),
7.92 (m, 1H, ArH), 7.73–7.49 (m, 2H, ArH), 7.36–7.12 (m, 2H,
ArH), 6.92 (t, J=8.8Hz, 1H, ArH), 2.31 (s, 1H, CH3); 13C NMR
(126MHz, DMSO-d6) δ 164.71, 157.00, 140.57, 140.40, 139.08,
137.69, 133.82, 133.36, 131.52, 130.59, 129.24, 128.59, 128.21,
124.43, 121.45, 120.91, 120.07, 117.56, 21.20.; HR-MS (ESI): m/z
Calculated for (C21H17N3O3S) [M-H]- 390.0913, found 390.0899.
1-(2,4-dichlorophenyl)-3-{4-[(1-oxo-1H-isoindol-3-yl)amino]
phenyl}urea (9): This compound was prepared according to
the same procedure used for the synthesis of compounds (5a–c)
The product was separated as reddish powder (0.15g, 56%);
m.p >300°C; 1H NMR (500MHz, DMSO-d6) δ 10.62 (s, 1H), 9.72
(s, 1H), 8.50 (s, 1H), 8.22 (d, J=9.0Hz, 1H), 8.00 (d, J=7.5Hz, 1H),
7.96–7.89 (m, 1H), 7.88–7.82 (m, 1H), 7.76–7.71 (m, 1H), 7.69 (d,
J=8.8Hz, 2H), 7.62 (t, J=3.1Hz, 1H), 7.51–7.47 (m, 2H), 7.39 (dd,
J=8.9, 2.5Hz, 1H).; 13C NMR (126MHz, DMSO-d6) δ 152.02,
139.36, 135.57, 135.29, 134.63, 134.30, 134.03, 133.27, 133.00,
131.10, 128.54, 127.56, 125.93, 122.91, 122.58, 122.07, 120.73,
118.49, 117.52, 110.22; HR-MS (ESI): m/z Calculated for
(C21H14Cl2N4O2) [M-H]- 423.0416, found 423.0420.
N-(2-chlorophenyl)-4-[(1,1-dioxidobenzo[d]isothiazol-3-yl)
amino]benzamide (5b): The product was separated as white
1
powder (0.28g, 68%); m.p=236° C; H NMR (500MHz, DMSO-
d6) δ 11.28 (s, 1H), 10.12 (s, 1H), 8.69 (d, J=7.4Hz, 1H), 8.12 (m,
2H), 8.01–7.79 (m, 4H), 7.71–7.47 (m, 2H), 7.46–7.19 (m, 2H),
Elsayed MS et al. Virtual screening and Synthesis… Arzneimittelforschung 2012; 62: 554–560

560 Original Article
7 Goodman VL, Rock EP, Dagher R et al. Approval Summary: Sunitinib
for the Treatment of Imatinib Refractory or Intolerant Gastrointestinal
Stromal Tumors and Advanced Renal Cell Carcinoma. Clin Can Res
2007; 13: 1367–1373
8 Boyer SJ. Small molecule inhibitors of KDR (VEGFR-2) kinase: an over-
view of structure activity relationships. Curr Top Med Chem 2002;
2: 973–1000
Biological evaluation
In-vitro cytotoxicity assay: The cytotoxicity of the synthesized
compounds was assayed using the standardized assay procedure
of the national cancer institute NCI (National cancer institute
Bethesda, Maryland, USA) which was described elsewhere.[11]
9 Keiser MJ, Roth BL, Armbruster BN et al. Relating protein pharmacology
by ligand chemistry. Nat Biotech 2007; 25: 197–206
10 Harmange J-C, Weiss MM, Germain J et al. Naphthamides as Novel and
Potent Vascular Endothelial Growth Factor Receptor Tyrosine Kinase
Inhibitors: Design, Synthesis, and Evaluation. J Med Chem 2008; 51:
1649–1667
11 Shoemaker RH. The NCI60 human tumour cell line anticancer drug
screen. Nat Rev Cancer 2006; 6: 813–823
1 2 Accelry’s Discovery Studio 2.5.5. S a n D i e g o : A c c e l r y s S o f t w a r e I n c .,
2010
13 Hecker EA, Duraiswami C, Andrea TA et al. Use of Catalyst Pharmacoph-
ore Models for Screening of Large Combinatorial Libraries. J Chem
Info and Comp Scie 2002; 42: 11204–11211
14 Scott RB. Cancer chemotherapy – the first twenty-five years. Br Med
J 1970; 4: 259–265
15 Krüger DM, Evers A. Comparison of Structure- and Ligand-Based Vir-
tual Screening Protocols Considering Hit List Complementarity and
Enrichment Factors. ChemMedChem 2010; 5: 148–158
16 Wolber G, Langer T. LigandScout: 3-d pharmacophores derived from
protein-bound Ligands and their use as virtual screening filters.
J Chem Info and Mod 2005; 45: 160–169
17 Wolber G, Langer T. LigandScout: Interactive automated pharmacoph-
ore model generation from ligand-target complexes. Abstracts Papers
of the American Chemical Society 2005; 229: U611–U611
18 Liu Y, Gray NS. Rational design of inhibitors that bind to inactive kinase
conformations. Nat Chem Biol 2006; 2: 358–364
In-vitro enzyme inhibition assay: In-vitro kinase inhibition
was determined for the synthesized compounds by KINEXUS
Corporation, Vancouver, British Columbia, Canada. Kinexus uses
a radioactive assay format for profiling evaluation of protein
kinase targets and all assays are performed in a designated
radioactive working area. Protein kinase assays were performed
at 30°C for 20–40min (depending on the target) in a final vol-
ume of 25 according to the following assay reaction receipt: 5μL
of diluted active protein kinase target (~10–50nM final protein
concentration in the assay), 5μl of peptide substrate, 5μl kinase
assay buffer, 5μl of compound (various concentrations) and 5μl
of 33P-ATP (250μM stock solution, 0.8μl Ci). The assay was initia-
ted by the addition of 33P-ATP and the reaction mixture incu-
bated at 30℃ for 20–40min, depending on the protein kinase
target. After the incubation period, the assay was terminated by
spotting 10μl of the reaction mixture onto multiscreen phos-
phocellulose P81 plate. The Multiscreen phosphocellulose P81
plate was washed 3 times for approximately 15min each in a 1%
phosphoric acid solution. The radioactivity on the P81 plate was
counted in the presence of scintillation fluid in a Trilux scintilla-
tion counter. Blank control was set up for each protein kinase
target which included all the assay components except the addi-
tion of appropriate substrate (replace with equal volume of
kinase assay buffer). The corrected activity for kinase target was
determined by removing the blank control value.
19 Choquette D, Teffera Y, Polverino A et al. Discovery of novel 1,2,3,4-tet-
rahydroisoquinolines and 3,4-dihydroisoquinoline-1(2H)-ones as
potent and selective inhibitors of KDR: Synthesis, SAR, and pharma-
cokinetic properties. Bioorg & Med Chem Lett 2008; 18: 4054–4058
20 Potashman MH, Bready J, Coxon A et al. Design, Synthesis, and Evalu-
ation of Orally Active Benzimidazoles and Benzoxazoles as Vascular
Endothelial Growth Factor-2 Receptor Tyrosine Kinase Inhibitors.
J Med Chem 2007; 50: 4351–4373
21 Frey RR, Curtin ML, Albert DH et al. 7-Aminopyrazolo[1,5-a]pyrimi-
dines as Potent Multitargeted Receptor Tyrosine Kinase Inhibitors.
J Med Chem 2008; 51: 3777–3787
22 Huang N, Shoichet BK, Irwin JJ. Benchmarking Sets for Molecular Dock-
ing. J Med Chem 2006; 49: 6789–6801
Acknowledgments
▼
The authors would like to thank Ahmed H. El-Khatib and Michael
W. Linscheid Humboldt-Universitaet zu Berlin, Department of
Chemistry, Berlin, Germany for carrying out the spectral analy-
sis of the synthesized compounds. Also the authors would like to
thank the National Cancer Institute, USA for carrying out the
cytotoxic evaluation of the compounds.
23 Hawkins PCD, Skillman AG, Nicholls A. Comparison of Shape-Matching
and Docking as Virtual Screening Tools. J Med Chem 2006; 50: 74–82
24 Triballeau N, Acher F, Brabet I et al. Virtual Screening Workflow
Development Guided by the “Receiver Operating Characteristic” Curve
Approach. Application to High-Throughput Docking on Metabotropic
Glutamate Receptor Subtype 4. J Med Chem 2005; 48: 2534–2547
25 Kakuta H, Zheng X, Oda H et al. Cyclooxygenase-1-selective inhibi-
tors are attractive candidates for analgesics that do not cause gastric
damage. design and in vitro/in vivo evaluation of a benzamide-type
cyclooxygenase-1 selective inhibitor. J Med Chem 2008; 51: 2400–2411
26 Rode HB, Sprang T, Besch A et al. Pseudosaccharin amine deriva-
tives: synthesis and elastase inhibitory activity. Pharmazie 2005;
60: 723–731
27 Romagnoli R, Baraldi PG, Remusat V et al. Synthesis and biological
evaluation of 2-amino-3-(3′, 4′, 5′-trimethoxy-phenylsulfonyl)-5-aryl
thiophenes as a new class of antitubulin agents. Med Chem 2007;
3: 507–512
28 Dumas J, Smith RA, Lowinger TB. Recent developments in the discovery
of protein kinase inhibitors from the urea class. Curr Opin Drug Discov
Devel 2004; 7: 600–616
29 Hu W-P, Chen Y-K, Yu H-S et al. Synthesis, and biological evaluation
of 2-(4-aminophenyl)benzothiazole derivatives as photosensitizing
agents. Bioorg and Med Chem 2010; 18: 6197–6207
30 Clark CR, Sansom RT, Lin C-M et al. Anticonvulsant Activity of Some
4-Aminobenzanilides. J Med Chem 1985; 28: 1259–1262
31 Bulman Page PC, Bethell D, Stocks PA et al. Sulfur Oxidation Mediated
by Imine Derivatives. Synlett 1997; 12: 1355–1358
32 Linn JA, Timothy L, Stevens KL et al. Lawrence Pyrimidine compounds
useful as kinase inhibitors. Patent WO/2008/024634 2008
33 Naegeli C, Tyabji A, Conrad L. Über den Umsatz aromatischer Isocyan-
säure-ester mit aromatischen Aminen. Helvetica Chimica Acta 1938;
21: 1127–1143
Conflict of Interest
▼
The authors report no conflicts of interest. The authors alone are
responsible for the content and writing of the paper.
References
1 Saijo N, Tamura T, Nishio K. Strategy for the development of novel
anticancer drugs. Canc Chem and Pharm 2003; 52: 97–101
2 Siemann DW. Tumor Vasculature: a Target for Anticancer Therapies.
John Wiley & Sons, Ltd, 2006; 1–8
3 Grunewald M, Avraham I, Dor Y et al. VEGF-Induced Adult Neovascu-
larization: Recruitment, Retention, and Role of Accessory Cells. Cell
2006; 124: 175–189
4 Holmes DI, Zachary I. The vascular endothelial growth factor (VEGF)
family: angiogenic factors in health and disease. Genome Biol 2005;
6: 209
5 Silva SR, Bowen KA, Rychahou PG et al. VEGFR-2 expression in carci-
noid cancer cells and its role in tumor growth and metastasis. Intl J
Canc 2011; 128: 1045–1056
6 Dev IK, Dornsife RE, Hopper TM et al. Antitumour efficacy of VEGFR2
tyrosine kinase inhibitor correlates with expression of VEGF and its
receptor VEGFR2 in tumour models. Br J Cancer 2004; 91: 1391–1398
Elsayed MS et al. Virtual screening and Synthesis… Arzneimittelforschung 2012; 62: 554–560