Synthesis and antitumor evaluation of novel diarylsulfonylurea derivatives: Molecular modeling applications

Article abstract of DOI:10.1016/j.ejmech.2009.11.014

Some new ethyl 2-[3-(4-unsubstituted or 4-substituted phenylsufonyl)ureido]-5,6,7,8-tetrahydro-4H-cyclohepta[b]thiophene-3-carboxylates 3a-c, 2-[3-(phenylsulfonyl)ureido]-5,6,7,8-tetrahydro-4H-cyclohepta[b]thiophene-3-carbohydrazides 4a-f, 3-phenylsulfony

Full text of DOI:10.1016/j.ejmech.2009.11.014

European Journal of Medicinal Chemistry 45 (2010) 689–697
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis and antitumor evaluation of novel diarylsulfonylurea derivatives:
Molecular modeling applications
,
b
c
Magda A. El-Sherbeny a , Alaa A.-M. Abdel-Aziz , Musa A. Ahmed
*
a Department of Medicinal Chemistry, Faculty of Pharmacy, University of Mansoura, Mansoura 35516, Egypt
b Department of Pharmaceutical Chemistry, College of Pharmacy, King Saud University, Riyadh 11451, Saudi Arabia
c Department of Medicinal Chemistry, Garyounis University ,Benghazi, Libya
a r t i c l e i n f o
a b s t r a c t
Article history:
Some new ethyl 2-[3-(4-unsubstituted or 4-substituted phenylsufonyl)ureido]-5,6,7,8-tetrahydro-4H-
cyclohepta[b]thiophene-3-carboxylates 3a–c, 2-[3-(phenylsulfonyl)ureido]-5,6,7,8-tetrahydro-4H-cyclo-
hepta[b]thiophene-3-carbohydrazides 4a–f, 3-phenylsulfonyl-6,7,8,9-tetrahydro-5H-cyclohepta[10,20:4,5]
Received 29 December 2008
Received in revised form
27 October 2009
Accepted 5 November 2009
Available online 12 November 2009
thieno[2,3-d]pyrimidine-2,4(1H,3H)dione
5 and 3-(phenylsulfonyl)-1-[3-(alkylaminocarbonyl or
substituted piperazinylcarbonyl)-5,6,7,8-tetrahydro-4H-cyclohepta[b]thiophen-2-yl]ureas 6a–d have
been synthesized and tested for their antitumor activity. Among these compounds, 4a, 4c and 4d
exhibited a broad spectrum antitumor activity with full panel (MG-MID) median growth inhibition (GI50
)
Keywords:
Synthesis
of 3.5, 4.9 and 4.0 M respectively. In addition, compounds 4c, 4d, 6c and 6d proved to be of moderate
m
selectivity toward colon cancer cell lines with ratios of 3.1, 1.2, 3.6 and 3.0 respectively. Molecular
modeling and pharmacophore prediction methods are used to study the antitumor activity of the most
active compounds compared with the least active species by means of the molecular mechanic method.
Ó 2009 Elsevier Masson SAS. All rights reserved.
Arylsulfonylurea
Antitumor
Molecular modeling
1. Introduction
occurrence of the unexpected methemoglobinemia and anemia in
phase II, which are caused by the aniline class metabolites [17,18].
In the course of identifying various chemical substances which
may serve as leads for designing novel antitumor agents, we are
particularly interested in the present work with diaryl-
sulfonylureas. Diarylsulfonylureas have been identified as a new
class of cancer chemotherapeutic agents with significant thera-
peutic efficacy against solid tumors [1–13]. Among these
compounds, remarkable effectiveness of LY186641 (A, Sulofenur)
and LY295501 (B) has been demonstrated against advanced colon
adenocarcinoma xenografts intrinsically resistant to virtually all
standard chemotherapeutic agents and had very significant activity
against several pediatric tumors grown as xenografts (Scheme 1)
[5,8,9,14–18]. Even though the mechanism of actions of these
compounds has not been elucidated, it is estimated that diary-
lsulfonylureas act on mitochondria and are different from other
clinically used antineoplastic agents in view of potency and toxicity
[15,16]. Diarylsufonylureas show neither side effects exhibited by
other anticancer agents, nor the cross-resistance in multi-drug
resistant cell lines. Despite some of these merits, the development
of diarylsulfonylurea has been seriously hampered due to the
In view of these facts and as a continuation of the previous efforts
aimed at the development of new antitumor agents, a new series of
ethyl 2-[3-(4-unsubstituted or 4-substituted phenylsufonyl)
ureido]- 5,6,7,8-tetrahydro-4H-cyclohepta[b]thiophene-3-carbox-
ylates 3a–c, 2-[3-(phenylsulfonyl)ureido]-5,6,7,8-tetrahydro-4H-
cyclohepta[b]thiophene-3-carbohydrazides 4a–f, 3-phenylsul
fonyl-6,7,8,9-tetrahydro-5H-cyclohepta[10,20:4,5]thieno[2,3-d]pyri
midine-2,4(1H,3H)dione 5 and 3-(phenylsulfonyl)-1-[3-(alkylami-
nocarbonyl or substituted piperazinylcarbonyl)-5,6,7,8-tetrahydro-
4H-cyclohepta[b]thiophen-2-yl]ureas 6a–d were designed to
explore the effect of the isosteric substitution of the aryl moieties
upon the antitumor activity (Schemes 1 and 2). Flexible alignment
and pharmacophore prediction methodology was used to identify
the structural features required for antitumor properties.
2. Results and discussion
2.1. Chemistry
Scheme 2 outlines the synthetic pathway used to obtain
compounds 3a–c, 4a–f, 5 and 6a–d. The starting material ethyl
2-amino-5,6,7,8-tetrahydro-4H-cyclohepta[b]thiophene-3-carbox-
ylate 1 was prepared according to Gewald’s procedure [19] via the
* Corresponding author. Tel.: þ20 50 2259700; fax: þ20 50 2247496.
E-mail address: magda_m20@yahoo.com (M.A. El-Sherbeny).
0223-5234/$ – see front matter Ó 2009 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2009.11.014

690
M.A. El-Sherbeny et al. / European Journal of Medicinal Chemistry 45 (2010) 689–697
(Scheme 2). Treatment of 3a with either hydrazine hydrate or the
appropriate phenyl hydrazine in ethanol afforded the correspond-
ing carbohydrazide derivatives 4a–f in a moderate yield (49%–66%).
On the other hand, 3-phenylsulfonyl-6,7,8,9-tetrahydro-5H-cyclo-
hepta[10,20:4,5]thieno[2,3-d]pyrimidine-2,4(1H,3H)-dione
5 was
obtained in this work with 50% yield, via the intramolecular cycli-
zation of 3a in sodium ethoxide. Moreover the interaction of 3a
with appropriate amine in 2-propanol furnished compounds 6a–
d (51%–67% yield).
2.2. Biological activity
2.2.1. In vitro antitumor evaluation
All the synthesized compounds were subjected tothe NCI in vitro
disease oriented human cells screening panel assay [21–23]. About
60 cell lines of nine tumor subpanels were incubated with five
concentrations (0.01–100 mM) for each compound and were used to
create log concentration versus % growth inhibition curves. Three
response parameters (GI50, TGI and LC50) were calculated for each
cell line (Tables 1 and 2). The GI50 value corresponds to compound’s
concentration causing 50% decrease in net cell growth, the TGI value
is the compound’s concentration resulting in total growth inhibition
and the LC50 value is the compound’s concentration causing a net
50% loss of initial cells at the end of the incubation period (48 h).
Subpanel and full panel mean graph midpoint values (MG-MID) for
certain agents are the average of individual real and default GI50, TGI
or LC50 values of all cell lines in the subpanel or the full panel,
respectively [23]. The NCI antitumor drug discovery screen has been
designed to distinguish between broad spectrum antitumor
compounds and tumor or subpanel-selective agents [22].
Scheme 1. Reported (A and B) and proposed (4, 5 and 6) diarylsulfonylureas.
reaction of cycloheptanone, ethyl cyanoacetate and sulfur in the
presence of diethylamine. Interaction of 1 with KOCN in dilute
acetic acid following the reported procedure [20] yielded the cor-
responding ureido derivative 2 in 70% yield. Upon heating 2 with
the appropriate benzene sulfonylchloride in pyridine, ethyl 2-[3-(4-
unsubstituted or 4-substituted phenylsufonyl)ureido]-5,6,7,8-tet-
rahydro-4H-cyclohepta[b]thiophene-3-carboxylates 3a–c were
obtained in a relatively good yield (55%–68%). Moreover, compound
3a was also prepared in a better yield (70%) through the reaction of
the amino ester 1 with benzene sulfonylisocyanate in toluene
The preliminary anticancer screening data of NCI using a 3-cell
lines panel, revealed that, only compounds 4a–d, 5 and 6a–d which
have the ability to reduce the growth of any one of the cell lines to
32% or less are passed on for the evaluation in full panel of 60 cell
lines (Tables 1 and 2). In present study, the active analogs showed
Scheme 2. Synthesis of the proposed diarylsulfonylureas.

M.A. El-Sherbeny et al. / European Journal of Medicinal Chemistry 45 (2010) 689–697
691
Table 1
Growth inhibitory concentration (GI50) of some selected cell lines (
2.2.2. Structural activity relationship and influence of lipophilic and
steric factors
m
M)a.
2.2.2.1. Structural activity relationship. The activity of the tested
compounds could be correlated to structure variation and modifi-
cations. The obtained screening results showed that, among the
tested compounds, compounds 4a–d are the most active members
with GI50 (MG-MID) values of 3.5, 17.0, 4.9 and 4.0 mM respectively.
Regarding these compounds, 2-[3-(phenylsulfonyl)ureido]-
5,6,7,8tetrahydro-4H-cyclohept[b]thiophen-3-carbohydrazide 4a is
the most active one of this series, GI50 value 3.5 mM, upon
replacement of the hydrazino moiety of 4a with phenylhydrazino
group as in 4b this result in a remarkable decrease in the magnitude
of antitumor activity, GI50 (MG-MID) value 17.0 mM. On the other
Comp.
No
Non small
cell lung
cancer
Colon
cancer
HCC-2998
Melanoma
UACC-257
Ovarian
cancer
OVCAR-
3
Breast
cancer
MCF-7
EKVX
b
b
4a
4b
4c
4d
5
6a
6c
6d
–
–
–
–
0.90
6.3
1.6
2.8
–
4.2
b
b
b
–
–
–
–
–
–
–
15.2
14.8
–
–
b
b
b
b
b
b
2.9
–
–
–
–
8.3
b
b
b
3.8
6.2
19.5
–
–
–
–
b
b
b
b
–
b
b
b
b
–
–
15.3
16.9
b
b
b
11.9
a
Data obtained from NCI’s in vitro disease oriented human tumor cell screen.
GI50 > 100 mmol/l.
hand, introduction of electronegative moiety (either Br or Cl) at
position 4 of the phenyl ring of 4b, this afforded 4c,d with
a remarkable increase in the antitumor activity, GI50 (MG-MID)
b
a distinctive potential pattern of selectivity as well as broad spec-
trum antitumoractivity. With regard to selectivityagainst individual
cell lines, compound 5 and 6a showed effectiveness against non-
small cell lung cancer EKVX with GI50 values of 3.8, and 6.2 mM
respectively (Table 1). Colon cancer HCC-2998 proved to be sensitive
values, 4.9 and 4.0 mM respectively. Moreover, also compounds 4c,d
showed moderate selectivity toward colon cancer cell lines with
ratios of 3.1 and 3.3 respectively. Upon cyclization of compound 3a
to the corresponding cycloheptathienopyrimidindione, compound
5 was obtained with a dramatic decrease in the magnitude of
toward compounds 4a–d with GI50 concentration range of 0.9–
antitumor activity GI50 (MG-MID) value, 26.0 mM. In case of
6.3
the same cell line with GI50 values of 19.5, 15.3 and 16.9
respectively. Regarding melanoma UACC-257, higher sensitivity was
observed with only compound 4a with GI50 value of 2.8 M.
Compounds 4b,c and 6d are effective against ovarian cancer OVCAR-
3 with GI50 values of 15.2, 14.8 and 11.9 M respectively. Breast
cancer MCF-7 was sensitive toward compounds 4a and 4d with GI50
values of 4.2 and 8.3 M respectively (Table 1). With regard to broad
spectrum antitumor activity, most of the active compounds showed
GI50 values (MG-MID) < 100 M against leukemia, non-small cell
mM. Moreover compounds 5, 6c,d proved to be effective against
compounds 6a–d, the presence of substituted piperazino moiety at
position -3 favored the antitumor activity rather than the presence
of propyl or butylamino groups at the same position, as shown by
mM
m
6c,d > 6a,b with GI50 (MG-MID) values of 33.2, 34.6 > 60.8, 77.5 mM
respectively. In addition, the same compounds 6c,d showed
moderate selectivity toward colon cancer cell lines with a ratios of
3.6 and 1.2 respectively.
m
m
2.2.2.2. Lipinski’s rule of five and the effect of lipophilic and steric
parameters [24–27]. We calculated the compliance of compounds
to the Lipinski’s rule of five [24]. Briefly, this simple rule is based on
the observation that most biological active drugs have a molecular
weight (MW) of 500 or less, a logP no higher than 5, five or fewer
hydrogen bond donor sites and ten or fewer hydrogen bond
acceptor sites (N and O atoms). The results disclosed in Table 3
show that all synthesized compounds comply with these rules.
Theoretically, these compounds should present good passive oral
absorption and differences in their bioactivity cannot be attributed
to this property.
The introduction of substituted amide or substituted hydrazide
fragments attached to thiophene core and the variation of the
substituents on these fragment have allowed evaluating the influ-
ence of lipophilicity and steric parameters at the pharmacophoric
part of the molecules. Table 3 gathers antitumor activity as well as
values of ClogP (lipophilic factors) and molar refractometry (steric
factors), determined by using HyperChem program [28] for each
m
lung, colon, melanoma, ovarian, renal, prostate and breast cancer
subpanel cell lines (Table 2). Compounds 4a,c,d showed the most
effective growth inhibition GI50 (MG-MID) values of 3.5, 4.9 and
4.0
with values of 17.0 and 26
least effective members of this series with GI50 (MG-MID) values of
60.8, 77.5, 33.2 and 34.6 M respectively (Table 2).
The ratio obtained by dividing the compound’s full panel MG-
MID ( M) by its individual subpanel MG-MID concentration ( M) is
mM respectively, while compounds 4b, 5 showed GI50 (MG-MID)
mM respectively, compounds 6a–d are the
m
m
m
considered as a measure for the compound selectivity [23]. Ratios
between 1.2 and 6 refer to moderate selectivity, while ratios greater
than 6 indicate selectivity toward the corresponding cell line-
subpanel. Most of the tested compounds proved to be non selective
with broad spectrum antitumor activity against the nine tumor
subpanels used, except compounds 4c,d and 6c,d that showed
moderate selectivity at the GI50 level toward colon cancer cell lines
with a ratios of 3.1, 1.2, 3.6 and 3.0 respectively (Table 2).
Table 2
Median growth inhibitory concentration (GI50 mM) of in vitro subpanel tumor cell lines.
Comp.
Subpanel tumor cell linesa
I
II
III
IV
V
VI
VII
VII
IX
4.5
18.5
5.3
MG-MIDb
c
4a
4b
4c
4d
5
6a
6b
6c
6d
30.3
18.6
–
30.6
42.2
42.2
–
10.5
16.0
11.5
8.6
3.5
7.0
–
4.5
1.5
–
15.8
–
–
14.5
34.2
90.2
92.6
80.2
60.3
3.5
17.0
4.9
4.0
26
60.8
77.5
33.2
34.6
30.9
25.5
12.0
81.3
72.5
85.3
56.2
55.8
32.3
31.3
4.2
26.2
65.3
76.7
61.2
52.3
15.2
10.2
7.5
34.2
76.5
–
26.3
11.5
–
50.2
80.3
72.3
42.3
65.2
1.6 (3.1)d
1.2 (1.23)
20.5
5.9
9.3
22.5
70.3
85.2
49.2
58.2
54.2
30.3
22.3
36.2
85.3
65.2
9.2 (3.6)
11.5 (3.0)
–
–
31.3
36.2
a
I, leukemia; II, non-small cell lung cancer; III, colon cancer; IV, CNS cancer; V, melanoma; VI, ovarian cancer; VII, renal cancer; VIII, prostate cancer; IX, breast cancer.
b
c
GI50 full panel mean-graph midpoint (
m
mol/L).
M.
GI50 selectivity ratios are shown in parentheses.
Compound showed GI50 values > 100
m
d

692
M.A. El-Sherbeny et al. / European Journal of Medicinal Chemistry 45 (2010) 689–697
Table 3
difference in activity between 4b and 4c or 4d. It is clear that lip-
ophilicity and molar refractometry of the molecules is not a crucial
factor for the activity and differences in their bioactivity cannot be
attributed to these parameters (Table 3).
Molar refractometry and calculated Lipinski’s rule of five for compounds under
biological investigation.
MRb
Parameter
Nviolatiosg
a
Comp No
GI50
LogPc
MWd
nONe
nOHNHf
4a
4b
4c
4d
5
6a
6b
6c
6d
3.5
17.0
4.9
4.0
26
60.8
77.5
33.2
34.6
105.7
127.9
135.6
132.5
94.6
114.5
118.9
126.2
146.7
2.5
2.7
4.9
4.7
1.7
1.7
2.2
1.0
3.2
408.5
484.6
563.5
519.1
376.5
435.6
449.6
476.6
538.7
8
8
8
8
6
7
7
8
8
5
4
4
4
1
3
3
2
2
0
0
1
1
0
0
0
0
1
2.3. Molecular modeling study
2.3.1. Conformational analysis
As an attempt to gain a better insight into the molecular struc-
tures of the most active compounds 4a,c,d; the moderately active 4b
and 6c; and the least active compound 6b, conformational analysis
has been performed by use of the MMþ forcefield [29] (calculations
in vacuo, bond dipole option for electrostatics, Polak-Ribiere algo-
rithm, and RMS gradient of 0.01 kcal/Åmol) as implemented in
HyperChem 5.1 [28]. The most stable conformer was fully optimized
with AM1 semi-empirical molecular orbital calculation [30]. The
global minimum was confirmed as true minimum not saddle point
by the absence of negative eigen value of the Hessian through
frequency calculation [26]. The calculation results showed that the
lowest energy-minimized structures of the compounds under
investigation exhibited a common arrangement of the arylsulfonyl
moiety and a critical distance of such moiety from the thiophene
core. Such group arranged in a perpendicular manner to the thio-
phene core and stabilized by hydrogen bonding with NH group of
urea fragment. Moreover, the aryl moietyattached tothiophene core
was out of the plane of thiophene ring (Fig. 1).
a
Data taken from Table 2.
Molar refractometry (Å3).
Calculated lipophilicity.
Molecular weight.
Number of hydrogen bond acceptor.
Number of hydrogen bond donor.
b
c
d
e
f
g
Number of violation from Lipinski’s rule of five.
compound. Within the series of compounds 4a–d we have
observed sharp increase of the antitumor activity when molar
refractometry increases from 127.9 cm3/mol (4b) to 135.6 cm3/mol
(4c). Such phenomenon was observed within other series such as
compounds 6b and 6d with molar refractometry values 118.9 cm3/
mol and 147.7 cm3/mol respectively. On the other hand the molar
refractometry cannot explain the difference in biological activity of
compounds 4a and 4b. Although lipophilicity does not exert
a significant effect on activity for compounds 4a–d, 5 and 6a–
d variations of lipophilicity produced by aromatic substituents (H,
Br and Cl) notably alter antitumor activity as shown by the
2.3.2. Flexible alignment [31–34]
To probe similarity between the three-dimensional structures of
active compounds 4a–d, flexible alignment was employed (Fig. 2).
Fig. 1. Lowest energy conformers of the most active compounds 4a, 4c, 4d and the least active compound 6b as representative examples showing the intramolecular hydrogen
bonds.

M.A. El-Sherbeny et al. / European Journal of Medicinal Chemistry 45 (2010) 689–697
693
Fig. 2. Flexible alignments of the most active compounds (Left panel): 4a (in green), 4b (in pink), 4c (in blue) and 4d (in red). Right panel showed flexible alignment of the most
active compounds 4a (in green), 4d (in red) and the least active compound 6b (in white).
Our initial approach was to employ MOE/flexible alignment [35] to
automatically generate superpositions of the compounds under
investigation with minimal user bias. Using MOE/MMFF94, 200
conformers of each compound were generated and minimized with
a distance-dependant dielectric model. A low energy set of 100 was
selected for further analysis. Conformations of compound 4d were
generated using distance geometry and optimized with MMFF94
[36]. Nine low energy, maximally dissimilar structures were
selected for comparison to the other compounds 4a–c. After
assigning MMFF94 charges to all molecules, flexible alignment was
employed to scan and rank overlays of compounds 4a–d based on
steric, electrostatic field, hydrophobic areas overlap, hydrogen
bond donors and acceptors overlap. From the top scoring super-
position, several sets most consistent with the structure activity
relationships of the four antitumor compounds were selected and
subjected to more refined searching using MOE/flexible alignment
module [35]. Since the molecules are highly flexible, the limited set
of conformers used in the analysis was not capable of achieving
complete atom to atom superposition. A further refinement was
generated by constraining the functional group mapping suggested
from the initial alignment. The calculated energy difference
between the native conformers derived from the flexible alignment
and the final conformers in Fig. 2 is minimal. In fact, the refined
structure of compound 4d was 0.7 kcal/mol lower in energy than
Fig. 3. Electrostatic maps (left panels) and hydrophobic maps (right panels) of the lowest energy conformers for the most active compounds 4a (upper panel) and the least active
compound 6b (lower panel); maps are color coded: red for a hydrogen bond and a hydrophilic region, cyan for a medium polar region, and green for a hydrophobic region.

694
M.A. El-Sherbeny et al. / European Journal of Medicinal Chemistry 45 (2010) 689–697
the starting structure, and compounds 4a–c was 1.0, 0.8 and
0.5 kcal/mol lower, respectively. A common feature of the MOE-
generated alignments is the superposition of the arylsulfonyl
fragments of the compounds with a slight deviation (0.2–0.9 Å).
Moreover, the CONHNH fragment of the compound 4d maps to the
CONHNH fragment of the other compounds 4a–c (0.25–0.75 Å).
The thiophene ring and urea fragment aligns fairly well for the four
compounds with almost complete superposition of the thiophene
rings of compounds 4a–d (0.4 Å).
Analysis of inactive molecule is an important step to understand
the essential features for a given activity. It is clear that compound
6b was flexibly aligned in a different manner when compared to the
active compounds explaining why such compound was the least
active as antitumor agents and showing the importance of both
CONHNH moiety and the aromatic ring attached to this CONHNH
group for activity (Fig. 2).
2.3.3. Electrostatic and hydrophobic mappings
By the same way, antitumor compounds 4c,d and the least
active compound 6b were subjected to flexible alignments. As
shown in Fig. 2 (right panel), all active compounds have the same
feature of alignment as we described above showing five points of
similarities; thiophene moiety, arylsulfonyl group, urea fragment,
CONHNH group directly attached to thiophene moiety as hydrogen
bond donor, and aromatic ring attached to such CONHNH moiety.
In an attempt to understand the lowest activity of compound 6b
and the highest antitumor activity of 4a, electrostatic and hydro-
phobic mappings has been carried out for the lowest energy
conformers, to examine the similarity and dissimilarity in elec-
tronic, electrostatic binding characteristics of the surface of the
molecules and conformational properties (Fig. 3; upper left panel).
Comparison of the electrostatic mappings of 4a and 6b show that
Fig. 4. The most active compounds 4a, 4c and 4d, mapped to the pharmacophore model for antitumor activity (Left lower panel). Right lower panel showed the least active
compound 6b, mapped to the pharmacophore model for antitumor activity. Upper panel showed the best predicted pharmacophore features and geometries which are required for
antitumor activity. Pharmacophore features are color coded: orange for hydrophobic aromatics, red for a hydrogen bond acceptor, and blue a hydrogen bond donor feature.

M.A. El-Sherbeny et al. / European Journal of Medicinal Chemistry 45 (2010) 689–697
695
compound 4a possess an increased negative charge regions located
on the sulfonyl urea and hydrazide groups representing hydrogen
bond regions (in red) 4a. In contrast, the lowest activity of
compound 6b maybe attribute to the difference in electrostatic
mapping in which the red hydrogen bond area was located mainly
on the sulfonyl urea fragment (Fig. 3; lower left panel).
group as well as nonbonded distances among two hydrophobic
aromatics and centroid of the thiophene align on the predicted
pharmacophore maps were found (Fig. 4). This pharmacophoric
assumption was in consistence with reported results for antitumor
arylsylfonylurea [4,12,13].
Similarly, hydrophobic mappings of the most active compound
4a showed that the hydrophobic region in green was distributed on
both side of aromatic ring of arylsulfonyl and arylhydrazide moie-
ties while the hydrophilic region in red was located mainly on
sulfonylurea and hydrazide fragments (Fig. 3; upper right panel).
On the other hand the hydrophobic distributions of the lowest
active compound 6b occupy an aliphatic butyl fragment indicating
of hydrophobic dissimilarity of such compound (Fig. 3; lower right
panel).
It is clear that the related charge distribution, electrostatic and
hydrophobic mappings suggesting a similar activity of the mole-
cules. As a result, we find that compounds 4a,c,d all can adopt
conformations that have identical distances and orientations and
that show sufficient agreement of the overall structure and of the
electrostatic potential pattern.
3. Conclusion
The present work led to the development of novel antitumor
molecules and three of the novel agents 4a,c,d, have shown higher
antitumor activities than all the other tested compounds. Flexible
alignment was conducted on the basis of experimental data from
which six featured pharmacophore model was developed. The
model was mapped onto the antitumor molecules and compared
with least active compounds. This model justifies the importance of
the main pharmacophore groups (particularly, the arylsulfonyl urea
fragment and the arylhydrazide moiety) as well as of their relative
distance. In fact, while the substitution pattern on the phenyl ring
considerably affect the activity, spatial considerations are particu-
larly important for the basic hydrazine group, since their spatial
locations in regard to the plane of thiophene core are critical for the
biological activity. Therefore, our pharmacophore model could be
very useful for the virtual screening in the development of new
diaryl sulfonyl urea as antitumor agents.
2.3.4. Pharmacophore prediction [31,34]
For pharmacophore generation and prediction, compound 4d
(Fig. 4) served as the reference to which all conformations of each
analogue were aligned. All structures were built de novo using 2D/
3-D editor sketcher in MOE [35]. Conformational models were
calculated using a 15 Kcal energy cut off (minimization conver-
gence criteria during conformational analysis: energy con-
vergence ¼ 0.01 kcal/mol, gradient convergence ¼ 0.01 kcal/mol).
The number of conformers generated for each substrate was
limited to a maximum of 100. All molecules with their associated
conformations were regrouped including the biological data.
Hypothesis generation was performed using low energy
conformers of the molecules. After assignment of possible phar-
macophore elements for each analogue the calculation and analysis
were carried out using the MOE program [35] and a superposition
of the molecules, including the assigned elements, was attempted
(Fig. 4). Several runs of calculation were repeated to generate
pharmacophore maps. For each run a distinct number of specified
pharmacophore elements were adapted. All adapted models
showed that the acceptor atoms of the sulfonyl and the donor
atoms of urea and hydrazino nitrogen fragments as hydrophilic
element and the aromatic rings as hydrophobic element were well
superimposed within the set distance tolerance. This confirms the
important role of the hydrophilic and hydrophobic moieties for
antitumor activity.
Models for 4d and its analogues 4a and 4c (Fig. 4; lower left
panel) possess pharmacophore elements in hydrophobic aromatic
fragments, the hydrazino basic nitrogen which is out of the thio-
phene plane as well as sulfonylurea in the same plane of thiophene
core. In contrast, model for 6b (Fig. 4; lower right panel) showed
common elements only in hydrophobic aromatic moiety and
sulfonylurea fragment. Since the arylhydrazino part of compound
4d plays an important role in activity, models 6b was excluded from
further considerations. Therefore, model 4d was considered as the
representative pharmacophore map of the antitumor activity.
According to the pharmacophore generated by MOE [35] the
minimal structural requirements for antitumor activity (Fig. 4)
consist of two aromatic rings which are distance 9.0 Å from each
other, a basic nitrogen atom (hydrazino moiety), a hydrophobic
region directly attached to the basic centre, a H-bonding acceptor
group (sulfonyl) attached to H-donor urea fragment. For all the
active molecules, reasonable nonbonded distances among the
hydrazine basic nitrogen, sulfonyl group and the H-donor urea
4. Experimental
4.1. Chemistry
Melting points (ꢀC, uncorrected), fisher-johns apparatus, 1H NMR
spectra (TMS as internal standard, chemical shifts in ppm) on a varian
EM 360 (90 MHZ) instrument. Elemental analysis (C, H, N) were
performed at the microanalytical center, Cairo university, all the
compoundswerewithin0.4%of thetheoreticalvalues.Ethyl2-amino-
5,6,7,8-tetrahydro-4H-cyclohepta [b]thiophen -3-carboxylate 1 have
been synthesized according to the procedure described in [37].
4.1.1. Ethyl 2-ureido-5,6,7,8-tetrahydro-4H-cyclohepta[b]
thiophene-3-carboxylate (2)
To a stirred solution of 1 (2.40 g, 0.01 mol) in acetic acid (10 mL)
and water (50 ml), KOCN (1.62 g, 0.02 mol) in water (20 mL) was
added dropwise over a period of 15 min. The reaction mixture was
stirred at room temperature for 6 h, allowed to stand at room
temperature for 10 h, and then diluted with an equal volume of
water. The separated solid was filtered, washed with water, dried
and crystallized from chloroform to yield 70% white solid, mp 130–
133 ꢀC (Ethanol); 1H NMR (DMSO-d6)
d: 1.12 (t, 3H, J ¼ 7.0 Hz, CH3),
1.31–2.10 (m, 6H,–CH2–C5, C6, C7), 2.42–2.90 (m, 4H,–CH2–C4, C8),
4.16–4.32 (q, 2H, J ¼ 7.0 Hz, –CH2CH3), 5.46 (brs, 2H, NH2), 10.09
(brs, 1H, NH). Microanalysis (C13H18N2O3S); anal.calcd.: C, 55.30; H,
6.43; N, 9.92. Found: C, 55.55; H, 6.54; N, 9.66.
4.1.2. Ethyl 2-[3-(4-unsubstituted or 4-substituted phenylsulfonyl)
ureido]-5,6,7,8-tetrahydro-4H-cyclohepta[b]thiophene-3-
carboxylates (3a–c)
A solution of 2 (2.82 g, 0.01 mol) and the appropriate sulfonyl
chloride (0.02 mol) in pyridine (10 mL) was heated under reflux for
3 h. Solvent was then evaporated in vacuo and the remaining
residue was then triturated with ice-water (50 mL) and filtered. The
obtained residue was dried and crystallized. 3a: 68% yield as white
solid, mp 143–145 ꢀC (Ethanol); 1H NMR (DMSO-d6)
d: 1.14 (t, 3H,
J ¼ 7.1 Hz, –CH2CH3), 1.49–1.74 (m, 6H, –CH2–C5, C6, C7), 2.60–2.93
(m, 4H, –CH2–C4, C8) 4.28 (q, 2H, J ¼ 7.0 Hz, –CH2CH3), 7.64–7.96
(m, 5H, Ar-H), 10.55 (brs, 1H, NH), 12.29 (brs, 1H, NH). Microanalysis
(C19H22N2O5S2); anal.calcd.: C, 54.01; H, 5.25; N, 6.63. Found: C,

696
M.A. El-Sherbeny et al. / European Journal of Medicinal Chemistry 45 (2010) 689–697
54.32; H, 5.52; N, 6.83. 3b: 54% yield as white solid, mp 165–167 ꢀC
(CHCl3–Ethanol); 1H NMR (DMSO-d6)
–CH2CH3), 1.55–1.82 (m, 6H, –CH2–C5, C6, C7), 2.83–2.83 (m, 4H,
–CH2–C4, C8), 4.38 (q, 2H, J ¼ 7.0 Hz, –CH2CH3), 7.23–7.85 (m, 5H,
Ar-H, NH), 8.93 (brs, 1H, NH). Microanalysis (C19H21BrN2O5S2);
anal.calcd.: C, 45.51; H, 4.22; N, 5.59. Found: C, 45.42; H, 4.52; N,
5.46. 3c: 55% yield as white solid, mp 157–159 ꢀC (CHCl3-Ethanol);
4.1.6. 3-Phenylsulfonyl-6,7,8,9-tetrhydro-5H-
d: 1.11 (t, 3H, J ¼ 7.0 Hz,
cyclohepta[10,20:4,5]thieno[2,3-d]pyrimidine-2,4(1H,3H)dione (5)
Compound 3a (4.23 g, 0.01 mol) was added portion wise to
a stirred mixture of sodium metal (0.5 g) in absolute ethanol
(50 mL). The reaction mixture was heated under reflux for 10 h,
evaporated in vacuo, 20 mL of water was then added and the
mixture was neutralized to pH 6. The separated solid was filtered,
dried and crystallized from ethanol to afford 50% of 5 as white solid,
1HNMR (DMSO-d6)
d
: 1.15 (t, 3H, J ¼ 7.1 Hz, –CH2CH3), 1.47–1.88 (m,
6H, –CH2–C5, C6, C7), 2.28 (s, 3H, CH3), 2.51–2.95 (m, 4H, –CH2–C4,
C8), 4.29 (q, 2H, J ¼ 7.1 Hz, –CH2CH3), 7.13–7.56 (m, 4H, Ar-H), 9.54
(brs, 1H, NH), 10.03 (brs, 1H, NH). Microanalysis (C20H24N2O5S2);
anal.calcd.: C, 49.94; H, 4.63; N, 6.13. Found: C, 50.33; H, 4.41; N,
6.23.
mp 176–177 ꢀC; 1H NMR (DMSO-d6)
d: 1.59–1.82 (m, 6H –CH2–
C6,C7, C8), 2.71–3.15 (m, 4H, –CH2–C5, C9), 7.34–7.95 (m, 5H, Ar-H),
8,57 (brs, 1H, NH). Microanalysis (C17H18N4O3S2); anal.calcd.: C,
54.24; H, 4.28; N, 7.44. Found: C, 54.53; H, 4.42; N, 7.25.
4.1.7. 3-(Phenylsulfonyl)-1-[3-(alkylaminocarbonyl or substituted
piperazinylcarbonyl)-5,6,7.8- tetrahydro-4H-cyclohepta[b]
thiophen-2-yl]ureas (6a–d)
4.1.3. Ethyl 2-[3-(phenylsufonyl)ureido]-5,6,7,8-tetrahydro-4H-
cyclohepta-[b]thiophene-3-carboxylate (3a)
Benzenesulfonyl isocyanate (1.83 g, 1.34 ml, 0.01 mol) was
added dropwise to a stirred solution of 1 (2.40 g, 0.01 mol) and
triethylamine (1.0 mL, 0.01 mol) in dry toluene (50 mL). The reac-
tion mixture was heated under reflux for 5 h, then evaporated in
vacuo and the obtained residue was triturated with ice-water,
filtered, dried and crystallized from ethanol to afford 70% of 3a.
A mixture of 3a (4.23 g, 0.01 mol) and the appropriate amine
(0.01 5 mol) in 2-propanol (50 mL) was heated under reflux for
8 h. On cooling, the separated solid was collected by filtration,
dried and crystallized. 6a: 67% yield as white solid, mp 218–
220 ꢀC (Ethanol); 1H NMR (DMSO-d6)
d: 0.98–1.04 (t, 3H,
J ¼ 7.0 Hz, –CH2CH3), 1.25–1.32 (m, 2H, –CH2CH3), 1.52–2.01 (m, 6
H, –CH2–C5, C6, C7), 2.74–2.92 (m, 4H,–CH2–C4, C8), 4.23–4.25
(m, 2H, –HNCH2), 7.40–7.76 (m, 5H, Ar-H), 9.05 (brs, 1H, NH),
10.13 (brs, 1H, NH), 12.15 (brs, 1H, NH). Microanalysis
(C20H25N3O4S2); anal.calcd.: C, 55.15; H, 5.79; N, 9.65. Found: C,
55.32; H, 6.00; N, 9.83. 6b: 63% yield as white solid, mp 168–
4.1.4. 2-[3-(Phenylsulfonyl)ureido]-5,6,7,8-tetrahydro-4H-
cyclohepta[b]thiophene-3-carbohydrazide (4a)
Hydrazine hydrate 98% (2.5 mL, 0.05 mol) was added drop wise
to 3a (4.23 g, 0.01 mol) in ethanol (50 mL). The reaction mixture
was then heated under reflux for 4 h, evaporated in vacuo and the
obtained residue was triturated with ice-water (50 ml), filtered,
dried and crystallized. 4a: 58% yield as white solid, mp 201–203 ꢀC
169 ꢀC (Ethanol); 1H NMR (DMSO-d6)
d
: 0.99–1.12 (t, 3H J ¼ 7.1 Hz,
–CH2CH3), 1.27–1.33 (m, 2H,–CH2CH3), 1.38–1.89 (m, 8H,–CH2–C5,
C6, C7 and CH2CH2CH3), 2.62–2.86 (m, 4H,–CH2–C4, C8), 3.97 (q,
2H, –HNCH2), 6.93 (brs, 1H, NH), 7.35–7.95 (m, 6H, Ar-H, NH),
10.32 (brs, 1H, NH), 11.72 (brs, 1H, NH). Microanalysis
(C21H27N3O4S2); anal.calcd.: C, 56.10; H, 6.05; N, 9.35. Found: C,
56.46; H, 6.32; N, 9.23. 6c: 51% yield as white solid, mp 145–
(EtOAc); 1H NMR (DMSO-d6)
d:1.31–2.10 (m, 6H, –CH2–C5, C6, C7),
2.38–2.95 (m, 4H, –CH2–C4, C8), 4.22 (brs, 2H, NH2), 7.23–7.97 (m,
6H, Ar-H, NH), 10.32 (brs, 1H, NH), 11.98 (brs, 1H, NH). Microanalysis
(C17H20N4O4S2); anal.calcd.: C, 49.98; H, 4.93; N, 13.72. Found: C,
49.66; H, 4.83; N, 13.53.
146 ꢀC (CHCl3); 1H NMR (DMSO-d6)
d: 1.52–2.21 (m, 6H, –CH2–C5,
C6, C7), 2.32 (s, 3H, N–CH3), 2.43–2.95 (m, 12H, –CH2–C4, C8, and
piperazine-H), 7.38–7.93 (m, 5H, Ar-H), 10.53 (brs, 1H, NH),
11.83(brs, 1H, NH). Microanalysis (C22H28N4O4S2); anal.calcd.: C,
55.44; H, 5.92; N, 11.76. Found: C, 55.65; H, 6.12; N, 11.55. 6d: 59%
yield as white solid, mp 202–205 ꢀC (CHCl3); 1H NMR (DMSO-d6)
4.1.5. 2-[3-(phenylsulfonyl)ureido]-5,6,7,8-tetrahydro-4H-
cyclohepta-[b]thiophene-3-carbohydrazides (4b–f)
Compound 3a (4.23 g, 0.01 mol) was added portion wise to
a stirred solution of the appropriate phenyl hydrazine (0.015 mol)
and triethylamine (5.06 mL, 0.05 mol) in ethanol (50 mL). The
reaction mixture was heated under reflux for 6 h, on cooling, the
separated solid was filtered, washed with ice-water, dried and
crystallized. 4b: 49% yield as white solid, mp 192–193 ꢀC (Pet.
d
: 1.49–1.98 (m, 6H –CH2–C5, C6, C7), 2.44–2.93 (m, 12H, –CH2–
C4, C8 and piperazine-H), 7.43–7.82 (m,10H, Ar-H), 9.53 (brs, 1H,
NH), 12.21 (brs, 1H, NH). Microanalysis (C27H30N4O4S2); anal.-
calcd.: C, 60.20; H, 5.61; N, 10.40. Found: C, 59.85; H, 5.34; N,
10.22.
ether-benzene); 1H NMR (DMSO-d6)
d: 1.42–1.83 (m, 6H, –CH2–C5,
C6, C7), 2.61–2.94 (m, 4H, –CH2–C4, C8), 7.23–8.02 (m, 11H, Ar-H,
NH), 9.31 (brs, 1H, NH), 10.23 (brs, 1H, NH), 12.23 (brs, 1H.NH).
Microanalysis (C23H24N4O4S2); anal.calcd.: C, 57.01; H, 4.99; N,
11.56. Found: C, 57.32; H, 4.56; N, 11.34. 4c: 52% yield as white solid,
4.2. Biological evaluation
Under a sterile condition, cell lines were grown in RPMI 1640
media (Gibco, NY, USA) supplemented with 10% fetal bovine serum
(biocell, CA, USA), 5 ꢁ105 cell/mL was used to test the growth
inhibition activity of the synthesized compounds. The concentra-
mp 183–185 ꢀC (Ethanol); 1H NMR (DMSO-d6)
d: 1.53–2.11 (m, 6H,
–CH2–C5, C6, C7), 2.43–2.94 (m, 4H, –CH2–C4, C8), 7.20–7.99 (m,
9H, Ar-H), 8.23–9.41 (brm, 2H, NH), 10.20 (brs, 1H, NH), 12.12 (brs,
1H, NH). Microanalysis (C23H23BrN4O4S2); anal.calcd.: C, 49.02; H,
4.11; N, 9.94. Found: C, 49.23; H, 4.36; N, 9.76. 4d: 56% yield as
tions of the compounds ranging from 0.01 to 100 mM were prepared
in phosphate buffer saline. Each compound was initially solubilized
in dimethyl sulfoxide (DMSO); however, each final dilution con-
tained less than 1% DMSO. Solutions of different concentrations
(0.2 mL) were pipetted into separate well of a microtiter tray in
duplicate. Cell culture (1.8 mL) containing a cell population of
6 ꢁ 104 cells/mL was pipetted into each well. Controls, containing
only phosphate buffer saline and DMSO at identical dilutions, were
also prepared in the same manner. These cultures were incubated
in a humidified incubator at 37 ꢀC. The incubator was supplied with
5% CO2 atmosphere. After 48 h, cells in each well were diluted 10
Times with saline and counted by using a coulter counter. The
counts were corrected for the dilution [21–23].
white solid, mp 168–169 ꢀC (Ethanol); 1H NMR (DMSO-d6)
d: 1.43–
1.94 (m, 6H, –CH2–C5, C6, C7), 2.39–2.86 (m, 4H, –CH2–C4, C8),
7.14–7.95 (m, 9H, Ar-H), 8.43–9.45 (brm, 2H, NH), 10.64 (brs, 1H,
NH),12.32 (brs, 1H, NH). Microanalysis (C23H23ClN4O4S2); anal.-
calcd.: C, 53.22; H, 4.47; N, 10.79. Found: C, 53.52; H, 4.32; N, 10.53.
4e: 66% yield as white solid, mp 206–209 ꢀC (EtOAc); 1H NMR
(DMSO-d6) d: 1.23–1.79 (m, 6H, –CH2–C5, C6, C7), 2.41–2.83 (m, 4H,
–CH2–C4, C8), 7.32–7.56 (m, 9H, Ar-H), 8.23 (brs, 1H, NH), 9.70–
10.44 (brm, 2H, NH), 12.35 (brs, 1H, NH). Microanalysis
(C23H23N5O6S2); anal.calcd.: C, 52.16; H, 4.38; N, 13.22. Found: C,
52.43; H, 4.25; N, 13.13.

M.A. El-Sherbeny et al. / European Journal of Medicinal Chemistry 45 (2010) 689–697
697
[4] I.-W. Kim, C.-K. Lee, H.S. Kim, S.-H. Jung, Arch. Pharm. Res. 26 (2003) 9–14.
4.3. Molecular modeling methods
[5] P.J. Houghton, J.A. Houghton, Investig. New Drugs 14 (1996) 271–280.
[6] E.Y. Moon, S.K. Seong, S.H. Jung, M. Lee, D.K. Lee, D.K. Rhee, S. Pyo, S.J. Yoon,
Cancer Lett. 140 (1999) 177–187.
4.3.1. Conformational search
[7] S.-H. Jung, J.-S. Song, H.-S. Lee, S.-U. Choi, C.-O. Lee, Bioorg. Med. Chem. Lett. 6
(1996) 2553–2558.
[8] F. Mohamadi, M.M. Spees, G.B. Grindey, J. Med. Chem. 35 (1992) 3012–3016.
[9] F. Mohamadi, M.M. Spees, G.B. Grindey, Bioorg. Med. Chem. Lett. 2 (1992)
987–992.
[10] S. Kim, J.H. Park, S.-Y. Koo, J.I. Kim, M.-H. Kim, J.E. Kim, K. Jo, H.G. Choi, S.B. Lee,
S.-H. Jung, Bioorg. Med. Chem. Lett. 14 (2004) 6075–6078.
[11] A. Scozzafava, T. Owa, A. Mastrolorenzo, C.T. Supuran, Curr. Med. Chem. 10
(2003) 925–953.
[12] H.-Y.P. Choo, J.-S. Lim, Y. Kam, S.Y. Kim, J. Lee, Eur. J. Med. Chem. 36 (2001)
829–836.
[13] H.-Y.P. Choo, S. Choi, S.-H. Jung, H.Y. Koh, A.N. Pae, Bioorg. Med. Chem. 11
(2003) 4585–4589.
[14] P.J. Houghton, J.A. Houghton, L. Myers, P. Cheshire, J.J. Howbert, G.B. Grindey,
Cancer Chemother. Pharmacol. 25 (1989) 84–88.
[15] P.J. Houghton, F.C. Bailey, G.S. Germain, G.B. Grindey, J.J. Howbert,
J.A. Houghton, Biochem. Pharmacol. 39 (1990) 1187–1192.
[16] P.J. Houghton, F.C. Bailey, J.A. Houghton, K.G. Murti, J.J. Howbert, G.B. Grindey,
Cancer Res. 50 (1990) 664–668.
Initial structures for the molecules 4a–d and 6b were con-
structed using the HyperChem program version 5.1. The MMþ
(calculations in vacuo, bond dipole option for electrostatics, Pola-
keRibiere algorithm, and RMS gradient of 0.01 kcal/mol) confor-
mational searching in torsional space was performed using the
multiconformer method [38]. Energy minima for compounds 4a–
d and 6b were determined by a semi-empirical method AM1 (as
implemented in HyperChem 5.1). The conformations thus obtained
were confirmed as minima by vibrational analysis. Atom-centred
charges for each molecule were computed from the AM1 wave
functions (HyperChem 5.1) by the procedure of Orozco and Luque
[39], which provides derived charges that resemble those obtain-
able from ab initio 6-31G* calculations.
4.3.2. Flexible alignment and pharmacophore prediction
[17] J.J. Howbert, Drugs Future 16 (1991) 517–520.
Flexible alignment and pharmacophore prediction of
compounds 4a–d and 6b was carried out with the software
‘Molecular Operating Environment’ (MOE of Chemical Computing
Group Inc., on a Core 2 duo 1.83 GHz workstation) [35]. The
molecules were built using the Builder module of MOE. Their
geometry was optimized by using the MMFF94 forcefield followed
by a flexible alignment using systematic conformational search.
Lowest energy aligned conformation(s) were identified through the
analysis module of MOE and the distances between the pharma-
cophoric elements were measured.
[18] W.J. Ehlhardt, J.J. Howbert, Drug Metab. Dispos. 19 (1991) 366–369.
[19] K. Gewald, Chem. Ber. 98 (1965) 3571–3574.
[20] F. Kurzer, in: H. Gilman, A. Blatt (Eds.), Organic Synthesis, second ed. John
wiley and sons, Inc., London, 1956, p. 49.
[21] M. Grever, S. Sehepartz, B. Chabners, Semin. Oncol. 19 (1992) 622–627.
[22] A. Monks, D. Schudiero, P. Skehan, J. Natl. Cancer Inst. 83 (1991) 757–760.
[23] M. Boyd, K. Paull, Drug Dev. Res. 43 (1995) 91–108.
[24] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Adv. Drug Deliv. Rev. 46
(2001) 3–26.
[25] A.A.-M. Abdel-Aziz, H.I. El-Subbagh, T. Kunieda, Bioorg. Med. Chem. 13 (2005)
4929–4935.
[26] A.A.-M. Abdel-Aziz, Eur. J. Med. Chem. 42 (2007) 614–626.
[27] S.T. Al-Rashood, I.A. Aboldahab, M.N. Nagi, L.A. Abouzeid, A.A.M. Abdel-Aziz,
S.G. Abdel-Hamide, K.M. Youssef, A.M. Al-Obaid, H.I. El-Subbagh, Bioorg. Med.
Chem. 14 (2006) 8608–8621.
Acknowledgments
[28] HyperChem: Molecular Modeling System Release 5.1. Hypercube Inc., Florida,
USA, 1997.
The authors would like to express their gratitude and thanks to
the National Cancer Institute (NCI), Bethesda Maryland, USA for
doing the antitumor testing of the new compounds.
Our sincere acknowledgments to Chemical Computing Group
Inc, 1010 Sherbrooke Street West, Suite 910, Montreal, H3A 2R7,
Canada., for its valuable agreement to use the package of MOE
2008.10 software.
[29] S. Profeta, N.L. Allinger, J. Am. Chem. Soc. 107 (1985) 1907–1918.
[30] M.J.S. Dewar, E.G. Zoebisch, E.F. Healy, J.J.P. Stewart, J. Am. Chem. Soc. 107
(1985) 3902–3909.
[31] A.M. Al-Obaid, S.G. Abdel-Hamide, H.A. El-Kashef, A.A.-M. Abdel-Aziz,
A.S. El-Azab, H.A. Al-Khamees, H.I. El-Subbagh, Eur. J. Med. Chem. 44 (2009)
2379–2391.
[32] P. Labute, C. Williams, M. Feher, E. Sourial, J.M. Schmidt, J. Med. Chem. 44
(2001) 1483–1490.
[33] S.K. Kearsley, G.M. Smith, Tetrahedron Comput. Methodol.
615–633.
3 (1990)
[34] W. Gerhard, S. Thomas, B. Fabian, L. Thierry, Drug Discov. Today 13 (2008)
23–29.
References
[35] MOE, 2008.10 of Chemical Computing Group. Inc.
[36] T.A. Halgren, J. Comput. Chem. 17 (1996) 490–519.
[37] M. Perrisin, C. Luu Dac, G. Narcisse, F. Bakri-Logeals, F. Hu Guet, Eur. J. Med.
Chem. Chimica. Therapeutica 15 (1980) 413–419.
[38] M. Lipton, W.C. Still, J. Comput. Chem. 9 (1988) 343–355.
[39] M. Orozco, F.J. Luque, J. Comput. Chem. 11 (1990) 909–923.
[1] A. Scozzafava, A. Mastrolorenzo, C.T. Supuran, Bioorg. Med. Chem. Lett. 10
(2000) 1887–1891.
[2] P. Nussbaumer, D. Geyl, A. Horvath, P. Lehr, B. Wolff, A. Billich, Bioorg. Med.
Chem. Lett. 13 (2003) 3673–3677.
[3] P. Lehr, A. Billich, B. Wolff, P. Nussbaumer, Bioorg. Med. Chem. Lett. 15 (2005)
1235–1238.