A convenient synthesis and molecular modeling study of novel purine and pyrimidine derivatives as CDK2/cyclin A3 inhibitors

Article abstract of DOI:10.1016/j.bmc.2010.08.033

A series of novel purine and pyrimidine derivatives were prepared and biologically evaluated for their in vitro anti-CDK2/cyclin A3 and antitumor activities in Ehrlich ascites carcinoma (EAC) cell based assay. The novel purine derivatives 13a,b demonstrated potent inhibitor activities with IC₅₀ values of 14 ± 9 and 13 ± 9 μM, respectively. Additionally, compound 15a showed the highest potency (IC₅₀ = 10 ± 6 μM) in EAC cell based assay. Molecular modeling study, including fitting to a 3D-pharmacophore model and their docking into cyclin dependant kinase2 (CDK2) active site showed high fit values and docking scores.

Full text of DOI:10.1016/j.bmc.2010.08.033

Bioorganic & Medicinal Chemistry 18 (2010) 7639–7650
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
A convenient synthesis and molecular modeling study of novel purine
and pyrimidine derivatives as CDK2/cyclin A3 inhibitors
Abdel-Sattar S. Hamad Elgazwy a, , Nasser S. M. Ismail b, Heba S. A. Elzahabi c
⇑
a Department of Chemistry, Faculty of Science, University of Ain Shams, Abbassia 11566, Cairo, Egypt
b Department of Pharmaceutical Chemistry, Faculty of Pharmacy University of Ain Shams, Abbassia 11566, Cairo, Egypt
c Department of Pharmaceutical Chemistry, Faculty of Pharmacy (Girls Branch), Al azahar University, Nasser City, Cairo, Egypt
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel purine and pyrimidine derivatives were prepared and biologically evaluated for their
in vitro anti-CDK2/cyclin A3 and antitumor activities in Ehrlich ascites carcinoma (EAC) cell based assay.
The novel purine derivatives 13a,b demonstrated potent inhibitor activities with IC50 values of 14 9 and
Received 27 June 2010
Revised 10 August 2010
Accepted 16 August 2010
Available online 26 August 2010
13
9
lM, respectively. Additionally, compound 15a showed the highest potency (IC50 = 10
6 lM) in
EAC cell based assay. Molecular modeling study, including fitting to a 3D-pharmacophore model and
their docking into cyclin dependant kinase2 (CDK2) active site showed high fit values and docking scores.
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Diaminomaleonitrile (DAMN)
Purine
Pyrimidine
CDK2
Molecular modeling
Antitumor
Docking
Bioinformatics
1. Introduction
In particular, this article introduces the main synthetic princi-
ples for the generation of the purine ring system and 2,6,9-trisub-
Diaminomaleonitrile (DAMN) and its N-substituted derivatives
are very useful compounds in heterocyclic synthesis.1–6 During
the course of our investigations on the use of DAMN in heterocyclic
synthesis, we have designed new approaches to 4-cyano-1,3-dihy-
dro-2-oxo-2H-imidazole-5-(N1-tosyl)carboxamide 11 (Scheme 1)
as a reactive precursor of thiopurine.7 In some of these cases,
new DAMN derivatives and N-({[(Z)-2-amino-1,2-dicyanovinyl]-
amino}carbonyl)-4-methylbenzenesulfonamide 10, were used as
key intermediates and we give herein a report on these compounds
in more detail. Recently, there has been described various natural
manifestations of purine systems, that is, methylated, higher-alkyl-
ated, and glycosylated forms.8 These comprise the purine alkaloids,
cytokinines, as well as the purine nucleoside antibiotics. In part,
the compounds described were isolated from natural sources al-
ready long ago. However, some have been reported only during
the last few years. In brief, the biological activities of most of the
purine derivatives are briefly described9 as potential anticancer,
anti-HIV-1, antimicrobial agents and in some cases, syntheses are
formulated.
stituted purines as inhibitors of cyclin/CDK complexes.10 Also,
these compounds are potent inhibitors of human cellular prolifer-
ation. They are useful in treating a disorder mediated by elevated
levels of cell proliferation in a mammal compared to a healthy
one by administering an effective dose. In the treatment of prolif-
erative diseases the interruption of the cell cycle is one approach.
The phases of the cell cycle are driven by cyclin-dependent ki-
nases.11–15 Upon complexation with its activating proteins, cyclin
E or cyclin A, cycline-dependent kinase2 (CDK2) modulates the
activity of many cellular substrates via phosphorylation on Ser
and/or Thr residues.16 In complex with cyclin E, cycline-dependent
kinase2 (CDK2) plays a paramount role during the G1/S transition
of the cell cycle while in complex with cyclin A, it facilitates the
progression of the S phase of the cell cycle. Recent evidence also
suggests that CDK2 may have a crucial role in the G2 phase of
the cell cycle.17,18 The importance of cycline-dependent kinase2
(CDK2) for cell cycle progression has led to an active pursuit of
small molecule inhibitors of this enzyme as a possible treatment
against cancer and other hyper-proliferative disorders.19 Our cur-
rent investigation was based on; first, using a structure-guided
strategy based on cycline-dependent kinase2 (CDK2)20–26 was as
appropriate means to generate CDK2 inhibitors that might prove
useful for the therapy of proliferative disorders.
⇑
Corresponding author. Tel.: +202 26173044; fax: +202 24831836.
E-mail addresses: aselgazwy@yahoo.com, elgazwy@sci.asu.edu.eg, elgazwy@
usa.com (Abdel-Sattar S. Hamad Elgazwy).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.08.033

7640
Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
O
N=C=O
O
S
N
N
N
H2N
C
C
H
N
C
O
N
N
HN
O
NH
H2N
H2N
C
C
Et3N (0.5 eq) or
DBU (0.5 mol%)
N
H
Me
NH
S O
O
NH
S O
O
CH3CN
stirring in acetonitrile
º
N2/0 C/
for 24 h.
20 mn
9
10
H3C
11
H3C
O
º
DBU (2 eq) at 0 C
stirring for 5 h
H3C
CH2Y
12a-b
DBU or Et3N
(3 eqv.)/RT/1 days
12a-b
and then 24 h. at RT
a) Y = H
b) Y = -COCH3
CH3
CH3
O
O
O
O
S
N
S
CH3
CH3
CH3
N
N
CH3
N
O
O
NH
O
NH
N
H
N
H
O
O
NH2
13a
NH2
13b
Scheme 1.
Several cores have been reported as potent CDK inhibitors
including purines, pyrimidine and quinazolines (Fig. 1).27–30 It
has previously been found that the presence of hetero-bicyclic ring
systems was essential to gain access to the phosphate binding
region of the ATP-kinases pocket.
particular interest for the purine ring system, the CDK2 complex
with ligand 5 (PDB code: 1H0V) is available from the PDB.30 This
example demonstrates the successful use of the fully active cyclin
complex in a prospective drug design program.
Second, the molecular modeling work on the cycline-dependent
kinase2 (CDK2) in complex with ligand (5) suggested a binding
mode (Fig. 2), where the bicyclic system was located in the ATP
binding site, making polar interactions with the kinase binding
motif. A large number of crystal structures were available for
human CDK2 in complex with small ligands which bind deeply
within the ATP site, and which interact with the kinase motif, of
2. Results and discussion
2.1. Chemistry
The design and synthesis of monosubstituted and disubstituted
pyrimidine, urea, purine and annelated imidazole derivatives with
anticancer activity are described.31 N-Alkylation with various side
O
CO2H
N
N
HN
Cl
N
H
HN
N
NH
N
N
N
N
N
N
R1
HO
N
N
H
N
HO
N
H
O S O
NH2
R2
Olomoucine (2), R1 = H, R2 = Me
Purvalanol B (1)
(4) NU6102
Roscovitine (3), R1 = ethyl, R2 = CH(CH3)2
O
HN
O
O
O
O
N
N
N
N
N
N
O
N
H2N
N
N
H
NH2
H2N
N
N
NH2
H2N
N
H2N
N
H
NU6027 (6)
(5)
NU6034 (7)
NU2058 (8)
Figure 1. Chemical structures of potent cycline-dependent kinase2 (CDK2) inhibitors.

Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
7641
Figure 2. 2D interaction diagram of ligand (5) with CDK2 enzyme represented using MOE program with the essential amino acid residues at the binding site are tagged in
circles.
chains in the imidazole and pyrimidine rings led to a series of mon-
osubstituted purine products. For preparation of disubstituted pur-
ine derivatives, appropriately substituted tosyl isocyanides were
used. The biological activity of all the compounds was evaluated
against a number of cancer cell lines. The sequence of reactions fol-
lowed in the synthesis of the target compounds is illustrated in
Scheme 1.
This paper described a facile one step synthesis of urea deriva-
tive 10, which was cyclized using a catalytic amount of DBU in ace-
tonitrile to give 11 in fairly high yield (93%). Compounds 10 and 11
were versatile precursors for fused nitrogen heterocycles, espe-
cially purine8 or thiopurine ring.7 The formation of 11 can be envis-
aged through the mechanism described by Ernst Schaumann
et al.32 and us,7 The mechanism illustrated that compound 11 is
afforded via rearrangement upon base catalyzed cyclization, as
outlined in Scheme 2.
In our research group, we are interested in studying the reactiv-
ity of the isolated urea derivatives 10. For example, treating 10
with aldehydes or ketones in the presence of a base such as DBU
(1,8-diazabicyclo[5.4.0]undec-7-ene) or Et3N (triethylamine)
either in catalytic or equivalent amount. A previous detailed
study33 of the reaction of DAMN with isocyanates together with
either an aldehyde or ketone in the presence of triethylamine re-
ported that the products of these reactions were pyrimidino[5,4-
d]pyrimidines. Similarly, we have studied the reaction of 10 with
ketone 12 or aldehyde 14 derivatives in the presence of a base as
outlined in Schemes 3 and 4, respectively. This reaction takes place
either via pathway A or B.
In pathway A, double bond isomerization is proposed under
base catalysis during the condensation of 10 with ketone. Interme-
diate E could not be isolated and this was attributed to its rapid
cyclization to pyrimidine F or its tautomer G and subsequent cycli-
zation followed by oxidation in air to give the isolated product,
assigned structure 13a,b. The same product is also obtained from
intermediate F after prolonged hydrolysis of the cyano group using
hydrogen peroxide in acetic acid.
In pathway B, we are unable to isolate the proposed intermedi-
ate A (Scheme 3). In the presence of base, it was postulated33 that
isomerization occurs around the C@C bond to give an intermediate
C, which cyclized rapidly to 13a,b; intermediate C could never be
isolated.
It can be seen from Scheme 1 that the reaction of DAMN with
tosyl isocyanate proceeded was expected to give the urea deriva-
tives 10 in excellent yields. Compound 10 has been previously
described by Ohtsuka33 and our spectroscopic data are in complete
agreement with his data and the characteristic nitrile absorption at
2254 and 2210 cmÀ1 and the carbonyl absorption at 1663, 1648
and 1611 cmÀ1 in the IR spectrum. On reaction of 10 with alde-
hydes or ketones in the presence of triethylamine or DBU at room
temperature the products of these reactions were isolated and
characterized by physical tools such as IR, 1H NMR, 13C NMR spec-
tral analysis and allowed us to establish the structures of the
products.
The reaction of urea derivatives of DAMN with aldehydes in the
presence of DBU or Et3N, afforded white–pale yellow solids precip-
itated in good yields. The reaction is inconsistent with the reaction
described by Ohtsuka33 and assigned in pyrimidine structure
16a–d (in case of aldehyde precursors). We are able to isolate an
intermediate 15a–d which thermally stable in hot acetonitrile
and cyclized rapidly to pyrimidine 16a–d, as shown in Scheme 4.
This article reports on the full assignment of NMR spectra for all
products deduced from the correlations established in the 2-D
(COSY, HSQC, HMBC) experiments, especially for the imine hydro-
gen and the spectroscopy results obtained on these compounds
products were satisfactory. The 1H NMR spectrum showed the
presence of two broad singlets at d 6.64 and 9.88 ppm due to the
H2N
CN
H2N
CN
NH
H
N
H2N
CN
NH
CN
NH
O
N
DBU (5%)
HN
O
C
N
C
H
N
N
H
º
at 0 C/2 hrs.
O
NH
Tos
N
NH
Tos
NH
Tos
O
Tos
11
10
Scheme 2.

7642
Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
Tos
NH
R1
R2
Tos
NH
O
O
Pathway A
Pathway B
HN
CN
NC
N
HN
H2N
CN
CN
O
Et3N (0.5 eq) or
DBU (0.5 mol%)
NH
HN
R1
C N
OH
HN
O
R1
R2
N
O
12
R2
Tos
10
A
R1
R2
E
Et3N (3 eq)
12
a) R1= R2 = CH3
b) R1= CH3, R2 = CH2COCH3
Tos
NH
H2O
O
N
N
C
HN
CN
H
N
H
O
R2
R1
N
C NH
N
Tos
N
H
O
R1
B
R2
F
hydrolysis
NH
Tos
NH
O
HO
C
HN
C N
NH2
H
N
N
O
R2
R1
N
N
N
H
O
Tos
R1
R2
G
O
Tos
C
O
NH2
NH
N
Tos
N
R1
R2
H
N
HN
H
NH
N
R2
O
O
H2N
N
N
H
N
N
Tos
R1
N
D
O
R1
R2
O
NH2
13a,b
Scheme 3.
D
A
C
F
N
N
H2N
C
N
N
C
CHO
B
Et3N (3 eq) stirring
in EtOH for 1 days,
-H2O
F
A
B
HN
O
C
HN
CN
NH
+
NH
D
O
S
O
O
C
S
O
O
14a-d
H3C
a) A = B = D = F = H, C = OMe
b) A = C = F = OMe, B = D = H
c) A = C = F = H, B = D = OMe
d) C = D = F = H, A = B = OMe
10
H3C
15a-d
CH3
B
B
F
A
C
A
C
D
O
O
S
NH
N
D
NH
N
C
F
Tos
O
N
C
HN
N
O
N
N
N
16a-d
Et3NH
Scheme 4.
amine protons and a singlet at d 7.28 for the HC proton of the imid-
azole ring and the 13C NMR spectrum was fully consistent with the
assigned structure. The infrared spectrum confirmed the presence
of the NH and C@N stretching vibrations within the region of
3420–3160, and 1650–1640 cmÀ1, respectively. The infrared spec-
trum also showed a sharp absorption band at 2200–2220 cmÀ1 for

Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
7643
the C„N stretching vibration. The 8-oxo-6-carboxamido-1,2-dihy-
dropurine 13a,b was prepared by stirring a suspension of the corre-
sponding urea 10 with a slight excess of acetone or acetylacetone in
ethanol at room temperature (Scheme 1). The reactions were mon-
itored by TLC (9:1 CHCl3/EtOH) and reaction times varied between
20 min and 24 h, depending upon the solvent used for the reaction
and the rate of precipitation, these 1,2-dihydropurines can be iso-
lated as solids in color from orange to yellow. Compounds 13a,b
were re-crystallized from mixture of ethanol/methanol (1:1) and
gave pale yellow to off white crystals, respectively. These were fully
characterized by TLC, IR and 1H NMR, 13C NMR and mass spectros-
copy. The infrared spectrum confirmed the presence of the NH and
C@N stretching vibrations within the region of 3400–3100, and
1660–1650 cmÀ1, respectively. The C@O of the amide group ap-
peared at 1695–1710 cmÀ1 as a strong band. The high resolution
mass spectrum gave a molecular ion peak at 363, 377 (M+1)+. Which
fits with the expected molecular weight of 362, 376 for the 1,2-dihy-
dropurine 13a,b. In the 1H NMR spectra of the isolated compounds
13a and 13b, the amide protons were observed in the region of d
7.69–7.85 ppm and in several cases the assignment were confirmed
by D2O exchange. The H-2 proton appeared as a broad singlet at d
4.95–4.97 ppm and the aromatic protons showed the expected pat-
terns in the range of d 6.85–7.30 ppm. The proton of the midazole
ring appeared as a sharp singlets in the range of d 7.42–7.50 ppm.
The 13C NMR spectrum of these 1,2-dihydropurine had the expected
number of bands, with the C-2 carbon at d 76.2–76.5, C-4 carbon at d
16a–d, diversely affect the activity. The recorded data showed that
both compounds 15a,b and their corresponding pyrimidinyl ana-
logues 16a,b, respectively, displayed the same IC50 (Table 3). In
these derivatives the OCH3 substituent of phenyl methylene group
oriented in either para or ortho/para direction.
On the other hand, when OCH3 group takes the ortho/meta
directing group, in this case, the corresponding pyrimidinyl deriv-
atives 16c,d exerted higher inhibitory activity than their corre-
sponding 15c,d, respectively, compared to the ligand 5.
2.2.2. In vitro antitumor activity
Additionally, the target compounds were examined against the
viability of EAC cell line, compared to ligand 5 (Table 3). The re-
corded GI50 values revealed that urea derivative 10 significantly
showed higher activity than ligand 5. Introduction of p-OCH3 phe-
nyl methylene moiety to compound 10 (GI50 = 12
15a potentiated the antitumor activity (GI50 = 10
l
M), affording
lM). It was ob-
served that both diaminomaleonitrile urea derivatives 10 and
15a possess acyclic skeleton. Also introduction of cyclic skeleton,
for example, purine ring system as in 13a,b or pyrimidinyl moiety
as in 16b showed considerable antitumor activity (GI50 = 15–
16
13b as well as pyrimidine derivative 16b showed the same antitu-
mor activity (GI50 = 16 M). Concerning phenyl methylene deriva-
lM). Another interesting observation is that purine derivative
l
tives 15a–d, it was found that 15a (the most active compound) has
only one OCH3 group in para position. Increasing the number of
OCH3 groups and varying their positions, for example, 15b–d led
to a dramatic reduction in the activity. While the p-OCH3 and
s-tri-OCH3 substituted phenyl methylene moiety when combined
with the cyclic pyrimidine nucleus, for example, 16a,b showed
higher activity than ligand 5.
136.4–137.2, C-5 carbon at
d 122.2–122.4, C-6 carbon at d
159.8–160.3, C-8 carbon at d 148.4–148.7, C@O carbon at d 166.4–
166.8 ppm.
2.2. Biological activity study
The newly synthesized compounds 10, 11, 13a,b, 15a–d, 16a–d
were evaluated in vitro as CDK2/cyclin A3 inhibitors as well as the
in vitro study against the viability of EAC cell line.
2.3. A molecular modeling study
Pharmacophore model was generated using the Discovery
Studio 2.5 software (Accelrys Inc., San Diego, CA, USA). Molecular
2.2.1. In vitro CDK2/cyclin A3 inhibition activity
docking into the prepared enzyme was performed using the MOE
.
The obtained data (Table 3, Fig. 6) revealed that, we have
synthesized some new potent CDK2/cyclin A3 inhibitors, for exam-
ple, compounds 10, 13a,b and 16c comparing to ligand 5,30 in
particular when the bicyclic purine ring system is held constant,
for example, 13a,b (lowest docking and IC50 value). Acyclic system
of diaminomaleonitrile urea, for example, compound 10, as well as
cyclic pyrimidine derivative, for example, 16c display lower inhib-
itory activity than purine derivatives 13a,b, recording that com-
pound 10 produced higher activity than pyrimidine derivative
16c. Compound 13a adopted a preferential binding mode than its
acetyl analogue 13b as well as ligand 5. Both of the latter com-
pounds illustrated a three hydrogen bond while 13a adopted four
hydrogen bonds (Fig. 5). The purine ring system significantly
potentiated the activity of 13a,b via hydrogen bonding interaction
so exerting promising binding mode model, low binding energy to-
gether with low IC50. Also, 13a preferentially formed two hydrogen
bonds through its sulfoxide group while that of 13b analogue
formed only one. Also, replacement of (1) hydrogen of methyl
group in 13a by acetyl group producing 13b, a little bit diminished
in vitro cycline-dependent kinase2 (CDK2) inhibitory activity.
These could be reasonable explanations for the highest activity ex-
erted by 13a. Here, diaminomaleonitrile urea derivative 10 is con-
sidered as a pivotal compound that is used as a precursor for the
synthesis of the other test compounds 11, 15a–d and 16a–d. So
2.3.1. Generation of CDK2 inhibitor hypothesis
The pharmacophore modeling method, as a key tool of com-
puter aided drug design, has been widely used in lead discovery
and optimization.34–36 Hip-Hop algorithm, which identifies com-
mon chemical features from a set of ligands without the use of
affinity data, was used to develop the pharmacophore model for
CDK2 inhibitor. The feature-based pharmacophore model was
mapped from a set of highly selective ligands (Fig. 1). The set of
conformational models of each structure of the training set was
performed using the prepare ligand protocol and was used to gen-
erate the common feature hypotheses.
2.3.2. Validation of the generated pharmacophore
Ten hypotheses were generated37 and the one ranked number 4
was chosen as the valid ideal hypothesis (Fig. 3) based on the
following: (a) The hypothesis showed full mapping of all its
features without any steric clashes together with high fit values
with the training set (compounds 1–8, e.g., see Fig. 4a), (b) Retro-
spectively, the simulated fit values of test set compounds (10, 11,
13a,b, 15a–d, 16a–d) with hypothesis 4 were more consistent with
the experimental results than the others38 (c) The database search
study for examining the affinity of such hypothesis with the molec-
ular structures of MiniMaybridge databases revealed that only 19
hits have been retrieved from the databases (2000 compounds).
Such a low number of the recognized database molecules by gen-
erated hypothesis may give an additional advantage and selectivity
to our hypothesis. Such an ideal hypothesis encompassed four
that, incorporating the two amino groups in 10 (IC50 = 32
imidazolidinone ring in 11 led to dramatic decrease in CDK2 inhib-
itory activity (IC50 = 45 M). Also, masking the amino group with
different aromatic aldehydes in 15a–d, reduced the activity
(IC50 = 38–43 M). Next, introduction of pyrimidinyl structure in
lM) into
l
l

7644
Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
Figure 3. (A and B) Constraint distances and angles of CDK2 inhibitors hypothesis. The chemical features colored light blue, green and violet represent HY, HBA and HBD,
respectively.
features namely; hydrogen bonding acceptor (HBA, green color),
two hydrogen bonding donor (HBD1 and HBD2, violet color) and
hydrophobic features (HY, blue color).
acids in the predicted binding pocket.30 The performance of the
docking method on CDK inhibitors was evaluated by re-docking
crystal ligand with 0.00974 RMSD value.44 The docking process
was carried out for the test set compounds (10, 11, 13a,b, 15a–d,
16a–d) using the compound energy as a scoring function.
In the flexible-ligand-rigid enzyme docking, the enzyme was
represented by potential energy maps, namely, electrostatic,
hydrogen bond, hydrophobic, and van der Waals. Interactive dock-
ing was carried out for all the conformers of each compound of the
test set to the selected active site of CDK2.
Previously, Vadivelan et al.39 reported a ligand-based Pharma-
cophore model for CDK2 inhibitors that contains the same kind
and number of features; two hydrogen bonding acceptor (HBA),
one hydrogen bonding donor (HBD) and one hydrophobic feature
(HY). Yet, the reported co-crystal structure of CDK2/ATP as well
as the crystal structure of CDK2 with several inhibitors40–43 dis-
proved the latter model. These crystal structures revealed that
the mode of binding in the CDK2/ATP binding pocket along the
residues Glu81–Leu83 is represented as a donor–acceptor–donor
motif. This binding mode is consistent with our generated pharma-
cophore which contains HBD1, HBA, and HBD2 in a similar manner.
In this article we reported the constraint distances and the an-
gles between the essential features existed in the generated
hypothesis, as shown in Figure 3 and Table 1.
The predicted binding energies of the compounds are recorded
in the Table 2. The docking of ligand 5 and highly active molecule
13a into the active site of CDK2 was performed (Fig. 5).
The docking results suggested that; first, the p-tosyl group will
increase the hydrophobic binding interaction with the deep hydro-
phobic pocket created by His 84, Phe 80, Glu81 and Ala 84. Second,
the hydrogen bonding interactions have been found, between the
crucial features of compounds with the high docking scores and
N–H group of Lys33, N–H of Lys89 and Asp86. Moreover, replace-
ment of the purine ring system by imidazolidinone or opening
the heterocyclic ring will decrease the docking value.
Alignment study of docked compound 13a and ligand 5 with
the binding pocket of CDK2 protein (Fig. 5D) revealed that (i) pur-
ine ring system of compound 13a was perfectly aligned with pur-
ine nucleus of ligand 5, (ii) carboxamide side chain superimposed
with O-pyrrolidin-2-one methyl substituent of ligand 5, (iii) addi-
tionally, both the ligand 5 and compound 13a make the same
hydrogen bonding interaction with Asp86 and Lys33.
The structures of the test set 10, 11, 13a,b, 15a–d, 16a–d were
built and prepared using the Discovery Studio software and their
conformational models were generated in the energy range of
20 kcal/mol above the estimated global energy minima. The map-
ping was carried out using Best Fit algorithm, during the Com-
pare/Fit process. Different mapping for all the conformers of each
compound to the hypothesis were visualized and the fit values of
the best-fitting conformers were recorded (Fig. 4, Table 2).
2.3.3. Molecular docking studies of CDK2 inhibitors and binding
conformation
All dock runs were conducted using MOE software. The 3D
structure of the enzyme was used to detail intermolecular interac-
tions between the ligand and the target protein. An automated
docking study was carried out using the crystal structure of inhib-
itor 5/CDK2 complex; having resolution of 1.95 Å. The prepared
protein was used in the determination of the important amino
Each docked compound was assigned a score according to its fit
in the LBP (Table 2).
2.3.4. Conclusion of molecular modeling
The above molecular docking study provides useful information
for understanding the structural features of CDK2 inhibitors. The

Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
7645
Figure 4. (a–c) Mapping of CDK2 inhibitors pharmacophore with lead compound 8 (NU2058) and compounds 13a and 13b, respectively.
Table 1
Constraint distances and angles between the features of generated CDK2 inhibitors hypothesis
Dimensions
Features of CDK2 inhibitors hypothesis
HBD1–HBD2, 4.787; HBD1–HBA, 4.731; HBD2–HBA, 4.650; HBD1–HY, 8.488; HBA–HY, 4.081
HBA–HBD1–HBD2, 30.698; HBD1–HBD2–HBA, 60.660; PI–HY2–HY3, 90.74; HBA–HBD1 vector,
58.814.; HBD2–HBA vector, 53.370
Constraint distances (Å) between features
Constraint angles (Å) between features
binding mode of the newly constructed CDK2 inhibitors pharmaco-
phore model was compatible with that obtained from the reported
crystal structures.40–43 The most potent purine derivatives 13a,b
in the in vitro antitumor activity. Among the test compounds a
contrast biological relations are observed between a cyclic urea
derivative 10 and cyclic purine derivatives 13a,b. Presence of lipo-
philic function (LF) such as para-tosyl group in compounds (10,
13a,b) in addition 2-methyl and 2-(2-oxopropyl) moieties of cyclic
purine derivatives 13a,b; showed the highest cycline-dependent
kinase2 (CDK2) inhibitory activity together with appreciated anti-
tumor activity against EAC cells for 13a,b and vice versa for com-
pound 10. Electron donating group (EDG) on para position of
benzylidene moiety for Compounds 15a–d; increase antitumor
agent against EAC (15a). While the insertion of such group in other
position provided lower potency (15b–d). 2,3-Dimethoxy substitu-
tion on benzylidene moiety of pyrimidine derivatives (16d) in-
duced steric clash and revealed lower potency as shown in (Fig. 7).
(IC50 values of 14 9 and 13
9 lM, respectively) showed the
highest docking scores and fit values. This was extended to the suc-
cessful designing of highly active analogs of purine derivatives
against CDK2.
3. Structural activity relationship (SAR)
The structural activity relationship (SAR) of newly synthesized
compounds 10, 11, 13a,b, 15a–d and 16a–d, explored the impor-
tance of the planer bicyclic purine system in cycline-dependent
kinase2 CDK2/cyclin A3 inhibition activity and to certain extent

7646
Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
Table 2
15
Best fit value and docking scour for each compound in the test set
(12, 13, 15a,b, 17a–d, 18a–d) mapped with active site of CDK2
14
13
12
11
10
13a
16d
Compd No.
Docking value (kcal/mol)
Fit value
16a
16c
Ligand 5
10
À12.01
À12.97
À13.51
À13.18
À11.28
À10.66
À10.86
À11.97
À12.91
À11.44
À12.87
À13.19
À10.67
3.89
2.93
3.79
3.87
2.76
1.99
1.87
2.10
2.45
2.34
1.92
2.81
2.66
13b
15d
10
13a
13b
15a
15b
15c
15d
16a
16b
16c
16d
11
ligand 5
16b
15a
15c
11
15b
In vitro IC50 (µM)
Figure 6. The in vitro IC50 of the test set compounds versus their docking scores
against cycline-dependent kinase2 CDK2/cyclin A3 docking scores of ligand 5 and
the newly synthesized compounds mapped with active site of cycline-dependent
kinase2 (CDK2) are diagramed against their corresponding IC50/lM in vitro CDK2
4. Experimental section
4.1. Chemistry
inhibitory activity. Except for pyrimidine derivatives 16a–d, the in vitro CDK2
inhibitory activity of the designed compounds was found to approximately match
with the docking values comparing to ligand 5. As an example of pyrimidine
derivatives, it was compound 16d that did not exhibit significant in vitro activity
against CDK2 (Table 3) in spite of having a better docking score value than the
reference ligand 5. This could be explained by a clash region predicted to exist
between 16d and some amino acid residues of CDK2 (Fig. 7). This clash region was
constructed through the existence of some atoms of 16d outside the binding pocket.
DAMN was obtained from Aldrich and used without further
purification. The aromatic aldehydes were obtained from Merck.
The reactions were carried out by mixing the reagents directly
from the bottle with a solvent. Infrared spectra were recorded on
a Perkin–Elmer Paragon FT-IR instrument. NMR spectra were
recorded in DMSO-d6 at room temperature on a Bruker AMX 400
Avance operating at 400 (1H) and 100 MHz (13C), respectively.
Melting points were taken with an Electro-thermal Apparatus
and are uncorrected.
sphere, in a round-bottom flask equipped with a serum cap, tolu-
ene sulfonyl-isocyanate (1 mL, 5.88 mmol) was added drop-wise
with a syringe through the serum cap, and 10 min later a precipi-
tate start to appear. The mixture was stirred at room temperature
for 24 h, when the cream solid was filtered and washed with ace-
tonitrile and diethyl ether, the product was identified as the title
compound (1.60 g) of white powder. The product characterization
was; Yield: 94%; mp >300 °C (from ethanol, lit.3,11: >300 °C); IR:
(Nujol mull) 2255 (m, CN), 2209 (s, CN), 1645 (s, C@O), 1611
4.1.1. N-{[(Z)-2-Amino-1,2-dicyanovinyl]amino}carbonyl]-4-
methylbenzenesulfonamide (10)
A solution of DAMN (0.60 g, 5.55 mmol) in dry acetonitrile
(10 mL) was kept stirring in an ice bath, under nitrogen atmo-
Figure 5. (A–C) The proposed binding mode of compounds 13a (A), 13b (C) and 10 (B) inside the active site of CDK2 resulting from docking, respectively. The most important
amino acids are shown together with their respective numbers. Compound 13a form four hydrogen bonds with tow Lys89 and two with Lys33 and Asp86. (D) Alignment of
docked compound 13a (green color) and ligand 5 with the binding pocket.

Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
7647
Table 3
4.1.4. 2-Methyl-9-[(4-methylphenyl)sulfonyl]-8-oxo-2-(2-oxo-
propyl)-2,7,8,9-tetrahydro-1H-purine-6-carboxamide (13b)
DBU (274 L, 0.274 g, 2.20 mmol) was added to a suspension of
diaminomaleonitrile urea (0.70 g, 2.28 mmol) in acetyl acetone
(205.6 L, 2 mmol) at 0 °C, and the reaction mixture was stirred at
CDK2 inhibition and the cytotoxic activity of the newly synthe-
sized compounds on EAC cells
l
Compounds
Cell growth inhibition
CDK2/cyclin A3
(IC50/lM)
(GI50
/
lM)
l
10
11
12
28
15
16
10
26
25
2 7
17
16
19
21
18
8
6
5
5
6
6
9
6
2
1
6
6
5
32
45
14
13
39
5
5
9
9
8
room temperature in a flask equipped with a screem cap under
nitrogen. A solid precipitate started to develop after 3 h, and the
reaction mixture was stirred for another 24 h. The solid was filtered
and washed with acetone to give the title compound (0.28 g, 46%
yield) as pinkish color. TLC (silica gel; EtOH/CH2Cl2 1:9) Rf = 0.72.
Characterization; mp 330–232 °C dec; IR (Nujol mull) 3450 (NH),
3290 (NH2), 1735 (s, C@O); 1670 (s, C@O); 1640 (s, C@O); 1H
NMR: (DMSO-d6, 300 MHz) d 1.64 (s, 3H, –C–CH3), 2.21 (s, 2H,
–CH2–), 2,42 (s, 3H, CH3), 6.98 (s, 2H, NH2), 7.36 (d, 2H, J = 7.8 Hz,
CHTos), 7.86 (d, 2H, J = 7.8 Hz, CHTos), 9.70 (s, 1H, NH), 11.32
(s, 1H, NH); 13C NMR: d 21.3, 24.5, 27.5, 67.8, 69.0, 114.6, 126.9,
127.2, 129.9, 140.7, 143.5, 148.67, 151.1, 160.3, 206.3. Anal. Calcd
for C17H19N5O5S (mol. wt. 405.43): C, 50.36; H, 4.72; N, 17.27; S,
7.91. Found: C, 50.49; H, 4.92; N, 17.85; S, 8.13.
13a
13b
15a
15b
15c
15d
16a
16b
16c
38 13
42 13
43 13
39
38
7
7
35 10
36 10
35 4%
16d
Ligand 530
IC50 value; corresponds to the compound concentration causing
50% mortality in net cells.
IC50 >100 l
g mL is considered inactive.30 the biological value of
ligand 5 as reported in lit.
4.1.5. General procedure for the synthesis of 1-(tosyl-3-[2-(-(2,3,
4,5,6-pentasubsituttedbenzylidene-amino)-1,2-dicycano-vinyl]-
urea (15a–d) and N-[5-(2,3,4,5,6-pentasubstitutedphenylmeth-
ylene-amino)-6-cyano-2-oxo-2,3-dihydro-pyrimidin-4-yl]-4-
methyl-benzenesulfonamide (16a–d)
(s, C@C); 1H NMR: (DMSO-d6, 300 MHz) d 2,4 (s, 3H, CH3), 7.38 (s,
2H, NH2), 7.41 (d, 2H, J = 7.9 Hz, Ar-H), 7.77 (s, 1H, –NHCO), 7.79 (d,
2H, J = 7.9 Hz, Ar-H), 11.18 (s, H, CONHTos); 13C NMR: d 106.1,
113.7, 117.0, 129.7, 149.8, 127.5, 129.5, 137.0, 144.0, 21.0. Anal.
Calcd for C12H11N5O3S (mol. wt. 305.3): C, 47.21; H, 3.63; N,
22.94; S, 10.50. Found: C, 47.12; H, 3.61; N, 22.89; S, 10.56.
Under vigorous stirring
a mixture of diaminomaleonitrile
DAMN-urea (2 mmol) and the respective corresponding aldehydes
14a–d (4 mmol) were suspended in ethanol (5 mL) and triethyl-
amine (4 mmol) was slowly added at 0 °C for 10 min and stirring
the reaction mixture was continued for 2–3 h at room temperature
until the color changed from yellow to white then to red after 5 h.
which were allowed to stand overnight until a precipitate ap-
peared. The solid was filtered off, washed with ethanol and dried
under reduced pressure and recrystallized from methanol to pro-
duce the first portion 15a–d whereupon the reaction mixtures
solidified in good yields. The A homogeneous solution of mother li-
quor was continued stirring for 24 h until the second portion of
crude products obtained which was filtered off and washed with
ethanol and recrystallized from methanol. The products, unless
otherwise noted, were yellow solids of 16a–d.
4.1.2. 5-Cyano-N-[(4-methylphenyl)sulfonyl]-2-oxo-2,3-dihydro-
1H-imidazole-4-carboximidamide (11)
To a suspension of compound 10 (0.35 g, 1.14 mmol) in acetoni-
trile (2 mL) under nitrogen atmosphere at 0 °C, DBU (43 mL,
0.043 g, 0.28 mmol) was added and then stirred at rt for 2 h. The
solid product was collected by filtration and washed with acetoni-
trile to give compound 11 as a cream white solid. Recrystallization
from ethanol gave colorless powder. Yield: 91% (0.32 g); mp 263–
265 °C (from ethanol, lit.7 262–263 °C); IR: kmax/cmÀ1: 3285, 3070,
2236,1724, 1663, 1658; 1H NMR: (DMSO-d6, 300 MHz) d 2,6 (s, 3H,
CH3), 7.36 (d, 2H, J = 8.7 Hz, CHTos), 7.78 (d, 2H, J = 8.7 Hz, CHTos),
8.41 (s, 1H, NH), 11.3 (s, 1H, NH), 11.7 (s, 1H, NH), 11.51 (s, H,
CONHTos); 13C NMR: d 96.9, 110.8 (CN), 126.6 (CN), 150.9, 157.6
(C@N), 126.2, 129.3, 139.1, 142.7, 21.0. Anal. Calcd for C12H11N5O3S
(mol. wt. 305.3): C, 47.21; H, 3.63; N, 22.94; S, 10.50. Found: C,
47.23; H, 3.72; N, 22.82; S, 10.59. The spectral data were consistent
with literature values.7
4.1.6. N-{[((Z)-1,2-Dicyano-2-{[(1E)-(4-methoxyphenyl)methy-
lene]amino}vinyl) amino]carbonyl}-4-methylbenzenesulfona-
mide (15a)
Reddish solid, yield: (0.60 g, 70%); mp >300 °C dec (from meth-
anol); IR (Nujol mull) 3353 (s, NH), 3259 (s, NH), 2955, 2854, 2205
(CN), 1919.3 (CN), 1655 (s, C@O); 1600 (s, C@O); 1527 (s, C@O),
1461; 1H NMR: (DMSO-d6, 300 MHz) d 2.35 (s, 3H, CH3), 3.85 (s,
3H, OCH3), 7.13 (d, 2H, J = 8.7 Hz, CHAr), 7.25 (s, 1H, –N@CH),
7.36 (d, 2H, J = 8.4 Hz, CHTos), 7.71 (d, 2H, J = 8.1 Hz, CHTos), 7.87
(d, 2H, J = 8.7 Hz, CHAr), 9.10 (s, 1H, NH), 9.86 (s, 1H, NH); 13C
NMR (DMSO-d6, 750 MHz): d 21.0, 55.6, 98.0, 110.8, 114.5, 126.2,
126.5, 126.6, 128.0, 129.3, 129.5, 132.3, 139.1, 142.7, 150.9,
151.6, 163.1. Anal. Calcd for C20H17N5O4S: C, 56.73; H, 4.05; N,
16.54, S, 7.57. Found: C, 56.87; H, 4.19; N, 16.43; S, 7.90.
4.1.3. 2,2-Dimethyl-9-[(4-methylphenyl)sulfonyl]-8-oxo-2,7,8,9-
tetrahydro-1H-purine-6-carboxamide (13a)
DBU (274 lL, 0.274 g, 2.20 mmol) was added to a suspension of
diaminomaleonitrile urea (0.35 g, 1.14 mmol) in acetone (4 mL) at
0 °C, and the reaction mixture was stirred at room temperature in a
flask equipped with a screem caps under nitrogen. A solid precip-
itate started to develop after 3 h, and the reaction mixture was stir-
red for another 24 h. The solid was filtered and washed with
acetone and diethyl ether to give the title compound (0.15 g, 45%
yield) as a cream yellow color. TLC (silica gel; ethanol/CH2Cl2;
1:9) Rf = 0.72. Characterization; mp 235–237 °C dec; IR (Nujol
mull) 3446 (NH), 3289.8 (NH2), 2970, 2955, 1679 (s, C@O);
1636.4 (s, C@O), 1567.6; 1H NMR: (DMSO-d6, 300 MHz) d 1.61
(s, 6H, -C(CH3)2), 2,42 (s, 3H, CH3), 6.98 (s, 2H, NH2), 7.40 (d, 2H,
J = 7.8 Hz, CHTos), 7.92 (d, 2H, J = 8.1 Hz, CHTos), 9.71 (s, 1H, NH),
11.35 (s, 1H, NH); 13C NMR: d 21.3, 27.5, 69.0, 114.6, 126.9,
127.2, 129.9, 140.7, 143.5, 148.67, 151.1, 160.3. Anal. Calcd for
4.1.7. N-{[((Z)-1,2-Dicyano-2-{[(1E)-(2,4,6-trimethoxyphenyl)-
methylene]amino}vinyl) amino]carbonyl}-4-methylbenzene-
sulfonamide (15b)
Yield: (0.75 g, 77%); mp >300 °C dec (from methanol); IR (Nujol
mull) 3530 (s, NH), 3279.8 (br s, NH), 2922, 2844, 2235.9 (CN),
1720 (s, C@O); 1663.9 (s, C@O); 1620.2, 1556.8 (s, C@O); 1H
NMR: (DMSO-d6, 300 MHz) d 2.34 (s, 3H, CH3), 3.85 (s, 3H,
OCH3), 3.81 (s, 6H, 2OCH3), 7.12 (d, 2H, J = 8.7 Hz, CHAr), 7.26 (s,
1H, –N@CH), 7.36 (d, 2H, J = 8.4 Hz, CHTos), 7.79 (d, 2H,
J = 8.4 Hz, CHTos), 7.87 (d, 2H, J = 8.7 Hz, CHAr), 8.37 (s, NH), 8.40
C15H17N5O4S (mol. wt. 363.39): C, 49.58; H, 4.72; N, 19.27; S,
8.82. Found: C, 49.52; H, 5.15; N, 19.30; S, 8.39.

7648
Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
Figure 7. (A and B) Docking of compound 16d to the active site of CDK2 shows its hydrogen bond interactions and its clash with binding site (C) Mapping of CDK2 inhibitors
pharmacophore with 16d, with non optimum fitting interaction.
(s, 2H, CHAr), 11.50 (br s, NH); 13C NMR (DMSO-d6, 750 MHz): d
21.2, 55.5, 55.6, 105, 112.9, 114.8, 126.9, 128.5, 129.9, 130.3,
132.3, 139.1, 141.4, 151.1, 153.2, 165.5, 167.1. Anal. Calcd for
4.1.9. N-{[((Z)-1,2-Dicyano-2-{[(1E)-(2,3-dimethoxyphenyl)-
methylene]amino}vinyl) amino]carbonyl}-4-methylbenzene-
sulfonamide (15d)
C
22H21N5O6S (mol. wt. 483.5): C, 54.65; H, 4.38; N, 14.48, S, 6.63.
Yield: (0.78 g, 86%); mp >300 °C dec (from methanol); IR (Nujol
mull) 3285 (br s, NH), 2233 (CN), 1723 (s, C@O); 1665, 1559; 1H
NMR: (DMSO-d6, 300 MHz) d 2.35 (s, 3H, CH3), 3.85 (s, 6H,
2OCH3), 7.13 (d, 1H, J = 8.7 Hz, CHAr), 7.25 (s, 1H, –N@CH), 7.35
(d, 2H, J = 8.4 Hz, CHTos), 7.73 (d, 1H, J = 8.7 Hz, CHAr), 7.80
(d, 2H, J = 8.4 Hz, CHTos), 7.88 (dd, 1H, J = 2.8, 8.7 Hz, CHAr),
78.37 (s, NH), 8.40 (s, 2H, CHAr), 11.50 (br s, NH); 13C NMR
(DMSO-d6, 750 MHz): d 21.2, 55.5, 106, 113.1, 114.5, 127.1,
128.3, 129.7, 130.6, 132.5, 139.5, 141.6, 151.5, 153.9, 164.5,
166.1. Anal. Calcd for C21H19N5O5S (mol. wt. 453.47): C, 55.62; H,
4.22; N, 15.44, S, 7.07. Found: C, 55.38; H, 4.26; N, 15.24; S, 6.98.
Found: C, 54.47; H, 4.23; N, 14.33; S, 6.55.
4.1.8. N-{[((Z)-1,2-Dicyano-2-{[(1E)-(3,5-dimethoxyphenyl)-
methylene]amino}vinyl) amino]carbonyl}-4-methylbenzene-
sulfonamide (15c)
Pink solid, yield: (0.86 g, 94%); mp >300 °C dec (from metha-
nol); IR (Nujol mull) 3280 (br s, NH), 2236 (CN), 1720 (s, C@O);
1664, 1557; 1H NMR: (DMSO-d6, 300 MHz) d 2.34 (s, 3H, CH3),
3.84 (s, 6H, 2OCH3), 7.15 (d, 2H, J = 8.7 Hz, CHAr), 7.27 (s, 1H, –
N@CH), 7.35 (d, 2H, J = 8.4 Hz, CHTos), 7.73 (s, 1H, CHAr), 7.80
(d, 2H, J = 8.4 Hz, CHTos), 7.88 (d, 2H, J = 8.7 Hz, CHAr), 78.37
(s, NH), 8.40 (s, 2H, CHAr), 11.50 (br s, NH); 13C NMR (DMSO-d6,
750 MHz): d 21.2, 55.5, 106, 113.1, 114.5, 127.1, 128.3, 129.7,
130.6, 132.5, 139.5, 141.6, 151.5, 153.9, 164.5, 166.1. Anal. Calcd
for C21H19N5O5S (mol. wt. 453.47): C, 55.62; H, 4.22; N, 15.44, S,
7.07. Found: C, 55.68; H, 4.32; N, 15.42; S, 7.05.
4.1.10. 6-Amino-5-{[(1E)-(4-methoxyphenyl)methylene]-
amino}-1-[(4-methylphenyl)sulfonyl]-2-oxo-1,2-dihydro-
pyrimidine-4-carbonitrile (16a)
Reddish solid re-crystallization from acetonitrile, yield: (0.55 g,
65%); mp >300 °C dec (from methanol); IR (Nujol mull) 3355–3260

Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
7649
(br s, NH2), 2230 (CN), 1655 (s, C@O); 1610, 1550, 1461; 1H NMR:
(DMSO-d6, 300 MHz) d 2.35 (s, 3H, CH3), 3.84 (s, 3H, OCH3), 7.12
(d, 2H, J = 8.7 Hz, CHAr), 7.26 (s, 1H, –N@CH), 7.36 (d, 2H,
J = 8.4 Hz, CHTos), 7.72 (d, 2H, J = 8.1 Hz, CHTos), 7.88 (d, 2H,
J = 8.7 Hz, CHAr), 9.10–9.86 (br s, 2H, NH2); 13C NMR (DMSO-d6,
750 MHz): d 21.5, 55.8, 97.9, 110.5, 114.2, 126.3, 126.6, 126.7,
128.1, 129.5, 129.8, 132.2, 139.4, 142.1, 151.3, 151.9, 163.7. Anal.
Calcd for C20H17N5O4S: C, 56.73; H, 4.05; N, 16.54, S, 7.57. Found:
C, 56.87; H, 4.19; N, 16.43; S, 7.90.
lM (Table 3), which is the concentration required inhibiting the
enzyme activity by 50% under the assay conditions used.
4.2.1.1. Experimental procedure. Ninety-six-well high-binding
microtiter ELISA plates were coated overnight at 4 °C with mouse
anti-GST in PBS. The plates were decanted and blocked with ELISA
buffer (0.05% bovine g-globulins [BGG] in 50 mM Tris-buffered sal-
ine [pH7.5] with 0.05% Tween 20 and 0.1 mM sodium ortho vana-
date). The cyclin-dependent kinase reaction was performed in
96-well non-binding microtiter plates. ATP, CDK2/cyclinA3 (HIS
tagged and phosphor T160), and GST-Rb were diluted in kinase
reaction buffer (25 mM Tris-buffered saline [pH7.5] with 10 mM
MgCl2, 5 mM b-glycerophosphate, 2 mM DTT, 0.1 mM sodium
ortho vanadate, and 0.005% Nonidet P-40). Final concentrations of
ATP and CDK2 were 200 mM and 300 ppm, respectively. Com-
pounds were diluted in 10% DMSO in kinase reaction buffer. ATP
was added to each well of the kinase reaction plate, followed in se-
quence by GST-Rb and test compound serial dilutions. The cycline-
dependent kinase reaction was started with the addition of CDK2/
cyclin A3 to the kinase reaction plate and stopped with kinase stop
buffer (50 mM EDTA in ELISA buffer). The ELISA plates were
washed with ELISA wash buffer (50 mM Tris-buffered saline [pH
7.5] with 0.05% Tween 20 and 1 mM MnCl2). The reactions were di-
luted 1:10 into kinase stop buffer in the ELISA plate. Phosphory-
lated Rb was detected with rabbit anti-phospho-Rb (Ser795)
followed by (HRP)-linked IgG anti-rabbit antibody. The plate was
developed with tetra-methyl benzylidene as substrate and the
absorbance was measured at 450 nm.
4.1.11. 6-Amino-5-{[(1E)-(2,4,6-trimethoxyphenyl)methylene]-
amino}-1-[(4-methylphenyl) sulfonyl]-2-oxo-1,2-dihydropyri-
midine-4-carbonitrile (16b)
Yield: (0.65 g, 67%); mp >300 °C dec (from methanol); IR (Nujol
mull) 3335–3280 (br s, NH2), 2235 (CN), 1664 (s, C@O); 1620.2,
1557; 1H NMR: (DMSO-d6, 300 MHz) d 2.35 (s, 3H, CH3), 3.85
(s, 3H, OCH3), 3.82 (s, 6H, 2OCH3), 7.15 (d, 2H, J = 8.7 Hz, CHAr),
7.25 (s, 1H, –N@CH), 7.36 (d, 2H, J = 8.4 Hz, CHTos), 7.80 (d, 2H,
J = 8.4 Hz, CHTos), 7.89 (d, 2H, J = 8.7 Hz, CHAr), 8.37 (s, NH2),
8.40 (s, 2H, CHAr); 13C NMR (DMSO-d6, 750 MHz): d 21.0, 55.3,
55.5, 104, 112.1, 114.4, 126.5, 128.6, 129.4, 130.6, 132.8, 139.6,
141.5, 151.8, 153.9, 163.7, 166.3. Anal. Calcd for C22H21N5O6S
(mol. wt. 483.5): C, 54.65; H, 4.38; N, 14.48, S, 6.63. Found: C,
54.55; H, 4.16; N, 14.23; S, 6.46.
4.1.12. 6-Amino-5-{[(1E)-(3,5-dimethoxyphenyl)methylene]-
amino}-1-[(4-methylphenyl) sulfonyl]-2-oxo-1,2-dihydropyri-
midine-4-carbonitrile (16c)
Yield: (0.66 g, 72%); mp >300 °C dec (from methanol); IR (Nujol
mull) 3280 (br s, NH2), 2236 (CN), 1665 (s, C@O), 1557; 1H NMR:
(DMSO-d6, 300 MHz) d 2.35 (s, 3H, CH3), 3.83 (s, 6H, 2OCH3), 7.15
(d, 2H, J = 8.7 Hz, CHAr), 7.26 (s, 1H, –N@CH), 7.36 (d, 2H,
J = 8.4 Hz, CHTos), 7.77 (s, 1H, CHAr), 7.81 (d, 2H, J = 8.4 Hz, CHTos),
4.2.2. In vitro antitumor assay
All chemicals and reagents are supplied from Sigma–Aldrich
(SIGMA–ALDRICH Chemie GmbH, Steinheim, Germany). Animal
house and biochemical equipments have been made available by
the University of Karachi, Pakistan. Female Swiss albino mice
weighing 25–30 g were used in this study (The Holding Company
for Biological Products and Vaccines Karachi, Pakistan. Mice were
housed at a constant temperature (24 2 °C) with alternating
12 h-light and dark cycles and fed standard laboratory food and
water. Tests were made in consideration of the internationally va-
lid guidelines. The Medical Centre for Research, Karachi, Pakistan is
concerned with biological and animal studies which have an ap-
proval of an institution responsible for animal ethics.
7.87 (d, 2H, J = 8.7 Hz, CHAr), 78.37 (s, NH2), 8.41 (s, 2H, CHAr); 13
C
NMR (DMSO-d6, 750 MHz): d 21.2, 55.5, 105, 112.1, 114.2, 126.8,
128.1, 129.4, 130.2, 132.8, 139.35, 141.2, 151.7, 153.5, 163.9,
164.6. Anal. Calcd for C21H19N5O5S (mol. wt. 453.47): C, 55.62; H,
4.22; N, 15.44, S, 7.07. Found: C, 55.68; H, 4.32; N, 15.42; S, 7.05.
4.1.13. 6-Amino-5-{[(1E)-(2,3-dimethoxyphenyl)methylene]-
amino}-1-[(4-methylphenyl) sulfonyl]-2-oxo-1,2-dihydropyri-
midine-4-carbonitrile (16d)
Yield: (0.58 g, 63%); mp >300 °C dec (from methanol); IR (Nujol
mull) 3285 (br s, NH2), 2230 (CN), 1665 (s, C@O), 1559; 1H NMR:
(DMSO-d6, 300 MHz) d 2.34 (s, 3H, CH3), 3.84 (s, 3H, OCH3), 3.81
(s, 3H, OCH3), 7.12 (d, 1H, J = 8.7 Hz, CHAr), 7.26 (s, 1H, –N@CH),
7.36 (d, 2H, J = 8.4 Hz, CHTos), 7.74 (d, 1H, J = 8.7 Hz, CHAr), 7.81
(d, 2H, J = 8.4 Hz, CHTos), 7.87 (dd, 1H, J = 2.8, 8.7 Hz, CHAr),
78.37 (s, NH2); 13C NMR (DMSO-d6, 750 MHz): d 21.1, 55.3,
105.6, 112.1, 114.2, 126.1, 128.5, 129.3, 130.9, 132.3, 139.6,
141.4, 151.3, 153.7, 163.9, 165.3. Anal. Calcd for C21H19N5O5S
(mol. wt. 453.47): C, 55.62; H, 4.22; N, 15.44, S, 7.07. Found: C,
55.68; H, 4.23; N, 15.40; S, 7.04.
4.2.2.1. Cell growth inhibition assay. The in-vitro growth inhibi-
tory activity of the test compounds against EAC cell line was evalu-
ated in NCI. The evaluation depends on using the standard 48 h
exposure assay. Determination of GI50/lM of the test compounds
was performed by preparation of serial dilutions ranged from 10À9
to 10À4 M47 in a mixture of dimethyl-sulfoxide (DMSO)/saline. Rela-
tionships between the profile of cell growth inhibition produced by
the test compounds and that of the established CDK inhibitor ligand
5, was investigated using the COMPARE algorithm study.48 Ehrlich
ascites carcinoma (EAC) cells were obtained by needle aspiration
of ascetic fluid from the pre-inoculated mice under aseptic condi-
tions. Tumor cells suspension (2.5 Â 106 per mL) was prepared in
saline. The parent line was kindly supplied by the Dow Medical
College (DMC), Diagnostics Lab., Karachi University, Pakistan. The
tumor cells were maintained by weekly intra-peritoneal transplan-
tation of cells.
4.2. Biological assays
4.2.1. CDK2/cyclin A3 inhibition assay
Inhibition of CDK2/cyclin A3 was assayed as previously de-
scribed45 using enzyme prepared from starfish oocytes (Marthaste-
rias glacials) and an assay buffer comprised of 50 mM Tris–HCl 7.5,
containing 5 mM MgCl2. Cycline-dependent kinase2 CDK2/cyclin
A3 was prepared as previously described46 (cyclin A3 is a C-termi-
nal cyclin A fragment encoding residues 171–432). The final ATP
4.3. Molecular modeling
4.3.1. Generation of CDK2 inhibitor hypothesis
The training set was selected as described above (Fig. 1). The
generation of the pharmacophore model for CDK2 inhibitors was
accomplished using Discovery Studio 2.5 software (Accelrys Inc.,
concentration in (CDK2) assay was 12.5
lM. The inhibitory activity
for the test compound was explored as the measurement of IC50
/

7650
Abdel-Sattar S. Hamad Elgazwy et al. / Bioorg. Med. Chem. 18 (2010) 7639–7650
11. Pines, J. Adv. Cancer Res. 1995, 66, 181.
12. Sielecki, T. M.; Boylan, J. F.; Benfield, P. A.; Trainor, G. L. J. Med. Chem. 2000, 43,
1.
San Diego, CA, USA). Molecules were built and conformational
models for each compound were generated automatically using
the Prepare Ligands protocol. The prepared training set molecules
were used for generation 3D pharmacophore using common fea-
tures hypothesis generation protocol (Hip-Hop module). In this
study, hydrogen bonding acceptor, two hydrogen bonding donor
and hydrophobic region, were used as the chemical features, which
were reported to be crucial for the CDK2 inhibitors activity. By this
process we specified the crucial features required for binding with
the CDK2 enzyme.
13. Arris, C. E.; Boyle, F. T.; Calvert, A. H.; Nicola, J. C.; Endicott, J. A.; Garman, A. F.;
Gibson, A. E.; Golding, B. T.; Grant, S.; Griffin, R. J.; Jewsbury, P.; Johnson, L. N.;
Lawrie, A. M.; Newell, D. R.; Noble, M. E. N.; Schultz, R.; Wu, Y. J. Med. Chem.
2000, 43, 2797.
14. Richardson, C. M.; Williamson, D. S.; Parratt, M. J.; Borgognoni, J.; Cansfield, A.
D.; Dokurno, P.; Francis, J. L.; Howes, R.; Moor, J. D.; Murray, J. B.; Robertson, A.;
Surgenor, A. E.; Torrance, C. J. Bioorg. Med. Chem. Lett. 2006, 16, 1353.
15. Harper, J. W.; Adams, P. D. Chem. Rev. 2001, 101, 2511.
16. Luk, K. C.; Simcox, M. E.; Schutt, A.; Rowan, K.; Thompson, T.; Chen, Y.;
Kammlott, U.; DePinto, W.; Dunten, P.; Dermatakis, A. Bioorg. Med. Chem. Lett.
2004, 14, 913.
17. Castanedo, G.; Clark, K.; Wang, S.; Tsui, V.; Wong, M.; Nicholas, J.;
Wickramasinghe, D.; Marsters, J. C.; Sutherlina, D. Bioorg. Med. Chem. Lett.
2006, 16, 1716.
18. Tang, J.; Shewchuk, L. M.; Sato, H.; Hasegawa, M.; Washio, Y.; Nishigaki, N.
Bioorg. Med. Chem. Lett. 2003, 13, 2985.
19. Senderowicz, A. M.; Headlee, D.; Stinson, S. F.; Lush, R. M.; Kalil, N.; Villalba, L.;
Hill, K.; Steinberg, S. M.; Figg, W. D.; Tompkins, A.; Arbuck, S. G.; Sausville, E. A.
J. Clin. Oncol. 1998, 16, 2986.
20. Massagué, J. Nature 2004, 432, 298.
21. Grana, X.; Reddy, E. P. Oncogene 1995, 11, 211.
22. Morgan, D. O. Nature 1995, 374, 131.
23. Dynlacht, B. Transcrip. Factors 2001, 85.
24. Pevarello, P.; Villa, M. Expert Opin. Therapeutic Patents 2005, 15, 675–703.
25. Shapiro, G. I. J. Clin. Oncol. 2006, 24, 1770.
26. Malumbres, M.; Barbacid, M. Nat. Rev. Cancer 2009, 9, 153.
27. Zaharevitz, D.; Kunick, C.; Schultz, C.; Meijer, L.; Leost, M.; Gussio, R.;
Senderowicz, A.; Lahusen, T.; Sausville, E. Am. Assoc. Cancer Res. 1999, 10.
Philadelphia, PA.
4.3.2. Molecular docking studies of CDK2 inhibitors
Docking Study was performed using the MOE software. Down-
loading the crystal structure of CDK2 enzyme complexes with
inhibitor 5 was carried out from protein data bank website (PDB)
entry 1H0V.30 Regularization and optimization for protein and li-
gand were performed. Determination of the essential amino acids
in binding site was carried out and compared with that present
in literature. The performance of the docking method was evalu-
ated by re-docking crystal ligand into the assigned active site of
CDK2 to determine RMSD value. Interactive docking using was car-
ried out for all the conformers of each compound of the test set
(10, 11, 13a,b, 15a–d, 16a–d) to the selected active site. Each
docked compound was assigned a score according to its fit in the
ligand binding pocket (LBP) and its binding mod.
28. Shewchuk, L.; Hassell, A.; Wisely, B.; Rocque, W.; Holmes, W.; Veal, J.; Kuyper,
L. F. J. Med. Chem 2000, 43, 133.
29. Davies, T. J.; Bentley, J.; Arris, C. E.; Boyle, F. T.; Curtin, N. J.; Endicott, J. A.;
Gibson, A. E.; Golding, B. T.; Griffin, R. J.; Hard castle, I. R.; Jews bury, P.;
Johnson, L. N.; Mesguiche, V.; Newell, D. R.; Noble, M. E. M.; Tucker, J. A.; Wang,
L.; Whitfield, H. J. Nat. Struct. Biol. 2002, 9, 745.
30. Gibson, A. E.; Arris, C. E.; Bentley, J.; Boyle, F. T.; Curtin, N. J.; Davies, T. G.;
Endicott, J. A.; Golding, B. T.; Grant, S.; Griffin, R. J.; Jews bury, P.; Johnson, L. N.;
Mesguiche, V.; Newell, D. R.; Noble, M. E. M.; Tucker, J. A.; Whitfield, H. J. J. Med.
Chem. 2002, 45, 3381.
31. Dolle, R. E.; Le Bourdonnec, B.; Goodman, A. J.; Morales, G. A.; Salvino, J. M.;
Zhang, W. J. Comb. Chem. 2007, 9, 855.
32. Schaumann, E.; Grabley, S. Chem. Ber. 1980, 113, 934.
33. Ohtsuka, Y. J. Org. Chem. 1976, 41, 629.
34. Wermuth, C. G. Pharmacophore and Pharmacophore Searches; Wiley-VCH, 2006.
pp 3–13.
35. Ekins, S.; Mestres, J.; Testa, B. J. Pharmacol. 2007, 152, 9.
36. Ekins, S.; Mestres, J.; Testa, B. J. Pharmacol. 2007, 152, 21.
37. Raimundo, J. M.; Sudo, R. T.; Pontes, L. B.; Antunes, F.; Trachez, M. M.; Zapata-
Sudo, G. Eur. J. Pharmacol. 2006, 530, 117.
5. Conclusion
In conclusion, the manuscript describes the investigation of a
series of novel purine, urea, imidazole and pyrimidine derivatives,
which were prepared in good yield by using diaminomaleonitrile
and tosylisocyanate in acetonitrile. Molecular modeling studies,
including fitting to a 3D-pharmacophore model their docking into
cyclin-dependent kinase2 (CDK2) active site were performed to
understand the structural features of CDK2 inhibitors. Biological
evaluation for both in vitro CDK2/cyclin A3 inhibition activity
and antitumor activity in Ehrlich ascites carcinoma (EAC) cell
based assay were also carried out. The most potent novel purine
derivatives were found to show the highest docking scores and
fit values, which appeared coherent with the obtained biological
data.
38. Ismail, N. S. M.; Salah El Dine, R.; Hattori, M.; Takahashi, K.; Ihara, M. Bioorg.
Med. Chem. 2008, 16, 7877.
39. Vadivelan, S.; Sinha, B. N.; Irudayam, S. J.; Jagarlapudi, S. A. R. P. J. Chem. Inf.
Model. 2007, 47, 1526.
40. Schulze-Gahmen, U.; De Bondt, H.; Kim, S.-H. J. Med. Chem. 1996, 39, 4540.
41. Hoessel, R.; Leclerc, S.; Endicott, J. A.; Nobel, M. E. M.; Lawrie, A.; Tunnah, P.;
Leost, M.; Damiens, E.; Marie, D.; Marko, D.; Niederberger, E.; Tang, W.;
Eisenbrand, G.; Meijer, L. Nat. Cell Biol. 1999, 1, 60.
42. Bramson, N. H.; Corona, J.; Davis, S. T.; Dickerson, S. H.; Edelstein, M.; Frye, S.
V.; Gampe, R. T.; Harris, P. A.; Hassel, A.; Holmes, W. D.; Hunter, R. N.; Lackey,
K. E.; Lovejoy, B.; Luzzio, M. J.; Montana, V.; Rocque, W. J.; Rusnak, D.;
Shewchuck, L.; Veal, J. M.; Walker, D. H.; Kuyper, L. F. J. Med. Chem. 2001, 44,
4339.
Acknowledgements
The authors gratefully acknowledge Professor Soomro, Dow
Medical College (DMC), Diagnostics Lab., Karachi University for
the evaluation the antitumor activity for the newly synthesized
compounds.
References and notes
43. Huang, P.; Ramphal, J.; Rice, A.; Tang, F.; Liang, C.; Wei, J.; McMahon, G.; Tang,
C. Proc. Am. Assoc. Cancer Res. 2001, 42. Abst. 3622.
1. Al-Azmi, A.; Elassar, A. Z. A.; Booth, B. L. Tetrahedron 2003, 59, 2749.
2. Alves, M. J.; Booth, B. L.; Proença, M. F. J. Chem. Soc., Perkin Trans. 2 1990, 1705.
3. Booth, B. L.; Dias, A. M.; Proença, M. F.; Zaki, M. E. A. J. Org. Chem. 2001, 66,
8436.
44. Abd El-Salam, O. I.; Abou El Ella, D. A.; Ismail, N. S. M.; Abdullah, M. Pharmazie
2009, 64, 147.
45. Vesely, J.; Havlicek, L.; Strnad, M.; Blow, J. J.; Donella-Deana, A.; Pinna, L.;
Letham, D. S.; Kato, J.-Y.; Detivaud, L.; Leclerc, S.; Meijer, L. Eur. J. Biochem.
1994, 224, 771.
4. Ohtsuka, Y. J. Org. Chem. 1978, 43, 3231.
5. Ohtsuka, Y. J. Org. Chem. 1979, 44, 827.
6. Erian, A. W. Chem. Rev. 1993, 93, 1991.
46. Brown, N. R.; Noble, M. E. M.; Lawrie, A. M.; Morris, M. C.; Tunnah, P.; Divita, G.;
Johnson, L. N.; Endicott, J. A. J. Biol. Chem. 1999, 274, 8746.
47. Monks, A.; Scudiero, D.; Skehan, P.; Shoemaker, R.; Paull, K.; Vistica, D.; Hose,
C.; Langley, J.; Cronise, P.; Vaigro-Wolff, A.; Gray-Goodrich, M.; Campbell, H.;
Mayo, J.; Boyd, M. J. Natl. Cancer Inst. 1991, 83, 757.
7. Hamad, A. S.; Derbala, H. A. Y. J. Heterocycl. Chem. 2001, 38, 939.
8. Rosemeyer, H. Chem. Biodiv. 2004, 1, 361; Bartoccini, F.; Cabri, W.; Celona, D.;
Minetti, P.; Piersanti, G.; Tarzia, G. J. Org. Chem. 2010, 75, 5398.
9. Rida, S. M.; Ashour, F. A.; El-Hawash, S. A. M.; El-Semary, M. M.; Badr, M. H.
Arch. Pharm. 2007, 340, 185.
48. Paull, K. D.; Shoemaker, R. H.; Hodes, L.; Monks, A.; Scudiero, D. A.; Rubinstein,
L.; Plowman, J.; Boyd, M. R. J. Natl. Cancer Inst. 1989, 81, 1088.
10. Trova, M. P.; Schenectady, N.Y., US Patent 6969720–B2, 2005.