Synthesis and biological validation of novel pyrazole derivatives with anticancer activity guided by 3D-QSAR analysis

Article abstract of DOI:10.1016/j.bmc.2012.01.032

Using literature data on anticancer activity of pyrazole derivatives, 3D-QSAR models were developed and 3D-QSAR analysis was performed. The 3D-QSAR analysis enabled identification of molecular properties that have the highest impact on antitumor activity against lung cancer cells. The results of 3D-QSAR analysis were taken into account while new compounds were designed. Obtained 3D-QSAR models were used for prediction of activity of new compounds. In this way, design of new compounds was guided by 3D-QSAR analysis which was performed on literature data. Ten new pyrazole derivatives were synthesised and their antitumor activities against A549 and NCIH23 lung cancer cells were validated. In order to obtain full profile of anticancer activity, cells viability (MTS) assays were combined with cell proliferation (BrdU) assays which measure actively dividing cells in treated sample. Experimental measurements showed good agreement between predicted and measured activities for majority of compounds. Also, anticancer activities of new pyrazole derivatives pointed to the chemical groups that can be useful in designing antitumor molecules. Substitution of hydrazine linker with rigid, 1,2,4-oxadiazole moiety resulted in compound 10, which has low (if any) cytotoxic activity and high potential cytostatic activity. Therefore, compound 10 presents a good starting point for design of new, more potent and safer anticancer therapeutics.

Full text of DOI:10.1016/j.bmc.2012.01.032

Bioorganic & Medicinal Chemistry 20 (2012) 2101–2110
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and biological validation of novel pyrazole derivatives
with anticancer activity guided by 3D-QSAR analysis
a,
b,
a
b,
´
´
ˇ
´
´
Ines Vujasinovic , Andrea Paravic-Radicevic , Kata Mlinaric-Majerski , Karmen Brajša
,
Branimir Bertoša c,
⇑
a
-
ˇ
Department of Organic Chemistry and Biochemistry, Ruder Boškovi´c Institute, Bijenicka cesta 54, 10000 Zagreb, Croatia
b GlaxoSmithKline Research Centre Zagreb, Prilaz b. Filipovic´a 29, Zagreb, Croatia
c
-
ˇ
´
Department of Physical Chemistry, Ruder Boškovic Institute, Bijenicka cesta 54, 10000 Zagreb, Croatia
a r t i c l e i n f o
a b s t r a c t
Article history:
Using literature data on anticancer activity of pyrazole derivatives, 3D-QSAR models were developed and
3D-QSAR analysis was performed. The 3D-QSAR analysis enabled identification of molecular properties
that have the highest impact on antitumor activity against lung cancer cells. The results of 3D-QSAR anal-
ysis were taken into account while new compounds were designed. Obtained 3D-QSAR models were used
for prediction of activity of new compounds. In this way, design of new compounds was guided by
3D-QSAR analysis which was performed on literature data. Ten new pyrazole derivatives were synthes-
ised and their antitumor activities against A549 and NCIH23 lung cancer cells were validated. In order to
obtain full profile of anticancer activity, cells viability (MTS) assays were combined with cell proliferation
(BrdU) assays which measure actively dividing cells in treated sample. Experimental measurements
showed good agreement between predicted and measured activities for majority of compounds. Also,
anticancer activities of new pyrazole derivatives pointed to the chemical groups that can be useful in
designing antitumor molecules. Substitution of hydrazine linker with rigid, 1,2,4-oxadiazole moiety
resulted in compound 10, which has low (if any) cytotoxic activity and high potential cytostatic activity.
Therefore, compound 10 presents a good starting point for design of new, more potent and safer antican-
cer therapeutics.
Received 8 December 2011
Revised 18 January 2012
Accepted 19 January 2012
Available online 28 January 2012
Keywords:
QSAR
VolSurf+
Pyrazole
Amidooxime
Antitumor activity
A549 and NCIH23 lung cancer cells
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
The pyrazole ring is a prominent structural motif found in numer-
ous pharmaceutically active compounds. Due to the easy preparation
Lung cancer remains major global health problem,1 accounting
for more than a million annual deaths worldwide. It is the second
most commonly occurring form of cancer in most Western coun-
tries and it is the leading cancer-related cause of death. In contrast
to the mortality rate in men, which began declining more than
20 years ago, women’s lung cancer mortality rates have been rising
for over the last decades and are just recently beginning to stabi-
lize.2 Since current treatment modalities are inadequate, the discov-
ery and development of new, more active, more selective and less
toxic compounds for the treatment of malignancy are one of the
most important goals in medicinal chemistry. In fact, cancer chemo-
therapy has entered a new era of molecularly targeted therapeutics,
which is highly selective and not associated with the serious toxic-
ities of conventional cytotoxic drugs.3
and rich biological activity, pyrazole framework plays an essential role
in biologically active compounds and therefore represents an interest-
ing template for combinatorial4,5 as well as medicinal chemistry.6–8
Numerous representatives of this heterocycle exhibit anti-viral/
antitumor,9,10 antibacterial,11–13 antiinflamatory,6 analgesic,14 fungi-
static,15 and anti-hyperglycemic activity.16,17 Also, some arylpyrazoles
were reported to have non-nucleoside HIV-1 reverse transcriptase
inhibitory activity.18 Extensive studies have been devoted to arylpy-
razole derivatives such as Celecoxib, a well-known cyclooxygenase-2
inhibitor.19–21
Large amount of experimental data on anticancer activity of
pyrazole derivatives can be useful in designing new compounds
with anticancer activity. Computational chemistry tools can help
in extracting the most useful information from large amount of
experimental data and enable their usage in designing new com-
pounds. Quantitative Structure–Activity Relationship (QSAR) stud-
ies have often been applied to find correlations between biological
activities and molecular descriptors for different classes of com-
pounds.22–24 Such correlations can enable identification of the
molecular properties which are the most important for biological
⇑
Corresponding author. Tel.: +385 1 456 10 25; fax: +385 1 468 02 45.
E-mail address: bbertosa@irb.hr (B. Bertoša).
Present address: Galapagos Research Centre, Prilaz b. Filipovic´a 29, Zagreb,
Croatia.
0968-0896/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2012.01.032

´
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
2102
activity. In the same time, obtained models can be used for predict-
ing activity of the compound whose activity has not been
measured.
the highest impact on compound’s activity. Model 2 actually con-
stitutes external validation of model 1, dataset compounds were
divided into the training set (used to build model) and test set
(used for external prediction, compounds: S5, S14, S21, S25, S29,
S39, S44, S45, S46, S49 (Table S1 in Supplementary data)) using
random number generator (www.random.org). Model 1 was also
used for the prediction of the activity of new, yet unsynthetised
compounds.
In the presented work, we have used VolSurf+ program to perform
3D-QSAR analysis on the series of pyrazole compounds25–30 with mea-
sured antitumor activity towards A549 lung cancer cells. Up to our
knowledge, this is the first attempt to use this dataset25–28 for gener-
ating 3D-QSAR models in order to identify the most important molec-
ular properties for the antitumor activity of pyrazole rings. VolSurf+
software has already been used for predicting antitumor activity of no-
vel heterocyclic compounds and experiments proved its reliability.31
The main advantages of the VolSurf+ based procedure are: (i) it does
not depend on the alignment of the molecules and (ii) the molecular
descriptors used in building QSAR model have clear physical or chem-
ical meaning.32,33 Results of 3D-QSAR analysis helped us to design new
compounds, to predict their biological activities and to make a priori-
tization in synthesis. In this way, 3D-QSAR analysis served us as guide-
line in planning experiments.
Ten new pyrazole derivatives were synthesised and their anti-
tumor activity against two lung cancer cell lines (A549 and
NCIH23) were measured. Beside the A549 lung cancer cells (which
were used for building 3D-QSAR models), activity of new com-
pounds was also tested against NCIH23, human cell lines adenocar-
cinoma, that are commonly used in determination of antitumor
activity of new chemical entities. Antitumor activity was charac-
terized using cells viability (MTS) assays and cell proliferation
(BrdU) assays. Cell viability indicates healthy cells in cell popula-
tion and such assays cannot distinguish if the treated cells are in
active phase of proliferation or not. Therefore, to have full profile
of anticancer activity, cells viability assays should be combined
with cell proliferation assays which measure actively dividing cells
in treated sample. Using such approach, compound with low (if
any) cytotoxic activity and high potential cytostatic activity was
found.
2.1.2. Identification of the most important descriptors
Principal component analysis (PCA) was performed on the
molecular descriptors of dataset compounds used for building
model 1. The first two principal components (PCs) explained 84%
of X-matrix variation. PCA loadings plot shows contribution of each
descriptor to the first two PCs (Fig. 2). Since PCs are constructed in
a way that the first few components describe majority of the X-ma-
trix variance, their loadings describe descriptors with the highest
contribution to the overall variance in the X-space PCs (which im-
plies that their variation between the compounds is the most
pronounced).
Following descriptors were identified as the ones with the high-
est variation between the compounds: W1–W3 (hydrophilic vol-
umes), WN1 (weak H-bond interaction in which compound
interacts as H-bond acceptor), WO1 (weak H-bond interaction in
which compound interacts as H-bond donor), V (volume), D1
(hydrophobic regions), HSA (the sum of hydrophobic surface
areas), S (surface) and MW (molecular mass).
The influence of each descriptor to the QSAR model can be esti-
mated from its PLS coefficient, while the real impact of a descriptor
on the biological activity is given as the product of the descriptor’s
value and its PLS coefficient. In order to investigate impact of each
descriptor on the biological activity, product between average va-
lue of the descriptor (calculated for the dataset used to built the
model) and its associated PLS coefficient was calculated for models
1 (Fig. 3) and 2 (Fig. S1 in the Supplementary data). In this way, fol-
lowing descriptors were found to have positive impact on antitu-
mor activity against A549 cells: hydrophilic volumes which are
descriptors related to polarizability and dispersion forces (W2,
W3 W4), molecular volume (V), molecular mass (MW), hydropho-
bic interactions at energy level of ꢀ0.2 kcal/mol (D1), the ability of
the compound to make weak H-bond interaction with carbonyl
group (WO1) and percentage of unionized species at pH 7 and 8
(%FU7, %FU8). The latter finding is in accordance with recently pub-
lished investigation that identified presence of a protonable func-
tion as one of the major structural conditions for antitumor
activity of 8-hydroxyquinoline substituted benzylamines.34 Aver-
age values of some of positively correlated descriptors of the data-
set compounds with the pIC50 values higher than 5.00
(compounds: S41, S43, S45, S47–S51) are: V—999.5 Å3, MW—
455.0 g/mol, W2—896.4 kcal/mol, D1—396 kcal/mol, WO1—
164.7 kcal/mol, %FU7—96.5%, %FU8—73.2%. In case of the least ac-
tive dataset compounds used for building model 1 (compounds
with pIC50 lower than 4.20: S1, S15, S17, S18, S21, S23, S35), the
average values of the same descriptors are: V—848.7 Å3, MW—
376.4 g/mol, W2—898.5 kcal/mol, D1—255.9 kcal/mol, WO1—
247.9 kcal/mol, %FU7—99.9 %, %FU8—99.7%. Interestingly, descrip-
tor related to hydrophilic volume at the lowest energy value
(ꢀ0.2 kcal/mol) (W1) had negative impact on activity. Other
descriptors with the highest negative correlation with antitumor
activity against A549 cells are: ability of the compound to make
weak H-bond interaction in which it interacts as H-bond acceptor
(WN1), molecular surface (S) and percentage of unionized species
at pH 10 (%FU10). The findings that %FU7 and %FU8 descriptors
have positive, while %FU10 has negative correlation with activity
indicate a weak acidic function as favorable for the antitumor
activity of investigated compounds. Average values of some of neg-
2. Results and discussion
2.1. 3D-QSAR analysis
3D-QSAR analysis was performed on the 51 compounds whose
antitumor activity against A549 lung cancer cells was described in
the literature.25–28 Two main goals of the performed 3D-QSAR
analysis were: (1) identification of molecular properties that have
the highest impact on antitumor activity against A549 cancer cells,
(2) prediction of antitumor activity of new, yet unsynthesized,
compounds. Using 3D-QSAR analysis, new compounds were de-
signed and their antitumor activities were predicted. In this way,
3D-QSAR analysis served us as a guideline in choosing new candi-
dates for synthesis and biological measurements. In the same time,
the results of biological measurements of new compounds pre-
sented in this paper enabled validation of the 3D-QSAR approach.
2.1.1. 3D-QSAR models
Several different 3D-QSAR models were built using the VolSurf+
program,32,33 detailed description of the procedure is given in Sec-
tion 4. Models in which raw data were used and models in which
completely inactive compounds were omitted were found to be
more reliable than all other generated models (Table 1). The latter
can be explained by the fact that in some cases lack of compound’s
activity is caused by insolubility, not by inadequate structural
properties of the compound, and by the fact that the exact activity
of inactive compounds was not given in the literature (Table S1 in
Supplementary data).25–28 Models 1 and 2 (Table 1, Fig. 1), both ob-
tained using raw data, were shown to be the most trusted ones and
were used for identification of the molecular descriptors that have

´
2103
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
Table 1
Statistical properties of 3D-QSAR models
e
Model
nOa
nLVb
R2
SDECc
Q2d
SDEP
SDEP-extf
0.21
nOEg
9
Scalingh
1i
2i
3
40
31
51
51
40
38
38
31
10
6
6
5
4
6
5
4
6
2
0.93
0.95
0.84
0.87
0.95
0.88
0.89
0.96
0.91
0.10
0.09
0.26
0.23
0.09
0.21
0.20
0.08
0.23
0.86 (0.81)j
0.85 (0.81)j
0.75 (0.77)j
0.78 (0.75)j
0.83 (0.75)j
0.77 (0.74)j
0.76 (0.68)j
0.79 (0.73)j
0.66 (0.64)j
0.15 (0.17)j
0.16 (0.18)j
0.32 (0.30)j
0.30 (0.32)j
0.16 (0.20)j
0.30 (0.31)j
0.30 (0.35)j
0.18 (0.21)j
0.44 (0.45)j
ꢀ
ꢀ
ꢀ
+
4
5i
6i
7i
8i
9k
+
0.37
0.29
0.14
0.45
13
13
9
ꢀ
+
+
+
41
a
b
c
d
e
f
Number of objects used to built the model (training set).
Number of latent variables.
SDEC is standard deviation of error of calculation.
Q2 is the cross-validated predictive performance.
SDEP is the standard deviation in cross-validated prediction.
SDEP for the external prediction.
Number of objects used for external prediction (test set).
Models in which autoscaling pretreatment procedure was applied (+) and models in which raw data was used (ꢀ).
Models in which inactive compounds (S24–S33, S39) were not included in training set nor in test set.
Values in brackets were obtained using random groups cross validation procedure.
g
h
i
j
k
Model was obtained using just ten compounds which were chosen as the representatives of the dataset according to the PCA results.
Figure 1. Predicted versus experimental antitumor activity, black dots present
training set compounds, yellow dots present test set compounds.
atively correlated descriptors of the dataset compounds with
the pIC50 values higher than 5.00 (compounds S41, S43, S45,
S47–S51) are: WN1—527.9 kcal/mol, W1—1587.5 kcal/mol, S—
673.4 Å2, %FU10—2.6%. The average values of the same descriptors
in the case of the least active dataset compounds used for building
model 1 (S1, S15, S17, S18, S21, S23, S35), are: WN1—605.6 kcal/
mol, W1—1432.1 kcal/mol, S—580.7 Å2, %FU10—78.1%.
Figure 2. PCA loadings plot for the first two principal components. PCA was
performed on the dataset used to build model 1.
important descriptors with positive correlation to activity were
related to polarizability and dispersion forces (W2, W3, W4). We
assumed that introducing atoms with high electronegativity, such
as oxygen, fluorine and nitrogen, into B ring of general formula
(Fig. 4), would enhance these properties and improve biological
activity. For that purpose compounds 1–6 were prepared. This
modification also affected other descriptors with positive correla-
tion to antitumor activity, such as molecular volume (V) and
molecular mass (MW). Hydrophobic interactions (D1) were in-
creased by introduction of additional aromatic ring. Further, the
ability of a compound to make weak H-bond interaction with car-
bonyl group (WO1) and percentage of unionized species at pH 7
and 8 (%FU7, %FU8) were also found as descriptors with positive
correlation to antitumor activity. Both factors indicate that pres-
ence of basic group is important. Thus, we replaced hydrazine
group as basic linker with more basic amidooxime group. For that
purpose compounds 7–10 were prepared. In addition, amidooxime
2.2. Design of new compounds
Taking the advantage of previously reported structures,26–31 all
proposed compounds were designed with the same general core
(1-arylmethyl-3-aryl-1H-pyrazole), fixed
A ring and different
modifications at position of B ring and linker (Fig. 4). Activity of
each new, proposed compound was predicted using 3D-QSAR
models and the most promising compounds were subjected to syn-
thesis and biological tests (compounds 1–10, Table 2).
Ring A contains 2-chloropyridine or tert-butylbenzyl in N-1 po-
sition since it was found that such compounds inhibited viability of
A549 cell.27 Modifications of B ring and linker were made in an
attempt to emphasize properties that were found by 3D-QSAR
analysis as positively correlated with antitumor activity. The most

´
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
2104
Figure 3. Products of the descriptor’s average value (calculated for the dataset used to built the model) and the associated PLS coefficient of model 1. Descriptors with the
highest impact on the activity are labeled; list and description of all 128 Volsurf+ descriptors is given in the VolSurf+ manual.32,33
A ring
Cl
A
N
N
B ring
N
F
OH
OH
OH
B
F
CN
Linker
Cl
O
NH2
O
N
O
HN
N
N
N
O
Figure 4. The common structural formula of the title compounds with colored portions indicated the modification sites.
compounds are able to complex with various metal ions35,36 which
is the property related with antitumor activity and chelation with
ions in cell.27 Finally, according to literature,37 amidooxime group
is less mutagenic and toxic than hydrazine group.
hydrolysis under microwave conditions at 110 °C for 30 min, then
acylated with phosphorus chloride and finally converted to amides
15a-b by introducing gaseous ammonia at room temperature. In
the next step, amides 15a-b, were converted into nitriles 16a-b
using POCl3. Reaction of the nitriles 16a-b with hydroxylamine
hydrochloride in the presence of triethylamine under microwave
heating, afforded amidooxime derivatives 7 and 8 in a good yields.
Finally, condensation of amidooxime 8 with 2,3-dihydroxybenzoic
acid at room temperature afforded desired O-acylamidooxime 9.
However, at reflux temperature, with the same reagents, 1,2,4-oxa-
diazole derivative 10 was obtained as main product.
2.3. Chemistry
The synthetic pathways used for preparation of the tested com-
pounds 1–10 (Table 2) are shown in Scheme 1 and Scheme 2. The
starting material, ethyl 5-(4-chlorophenyl)-2H-pyrazole-3-carbox-
ylate (11) was prepared from commercially available 4-substituted
acetophenone and diethyl oxalate, with hydrazine in the presence
of acetic acid at room temperature, according to procedures re-
ported in literature.25,26 In the next step, it was used for prepara-
tion of key intermediates 12a-b, by alkylation with 2-chloro-5-
chloromethylpyridine for 12a and 1-tert-butyl-4-(chloro-
methyl)benzene for 12b. We have applied rapid and efficient
microwave irradiation for the synthesis of desired compounds
12a-b. Compared to the conventional heating,26 the reaction time
was shorten to 30 min instead of 6–14 h. The reaction of interme-
diates 12a-b with hydrazine hydrate gave carbohydrazide deriva-
tives 13a-b that were further treated with diverse aldehydes to
afford final compounds 1–6 in good yields >80% (Scheme 1).
To prepare compounds 7–10 (Scheme 2), starting intermediates
12a-b, were converted into corresponding acids 14a-b by alkali
All newly synthesized compounds were characterized by spec-
tral and LC–MS analyses which were in full agreement with the
proposed structures.
2.4. Biological evaluation
All prepared compounds 1–10 were tested on viability of A549
cell lines to compare predicted and experimental value (Table 2).
In addition, all compounds were also tested on additional lung
cancer cell line NCIH23 (adenocarcinoma, non-small cell lung
cancer). These two cell lines, A549 and NCIH23, are adenocarci-
nomic human cell lines most commonly used in determination
of antitumor screening activity of new chemical entities. While
A549 cells are human cancer alveolar basal epithelial cells,

´
2105
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
Table 2
New compounds, designed according to the results of 3D-QSAR analysis, and their predicted and measured antitumor activities against A549 lung cancer cells
R1
N
N
R2
Compound
Model 1 pIC50
(l
M) (pred)a Model 2 pIC50
(l
M) (pred)a MTS assay pIC50 M) (exp)b
(l
Cl
R1
R2
Cl
Cl
O
HO
HO
HO
1
2
4.82
4.90
4.82
4.91
5.33
5.28
N
N
N
H
N
N
N
O
OH
OH
N
H
O
N
H
3
4
5
4.92
4.78
4.77
4.92
4.79
4.79
5.11
4.80
<4.00
Cl
O
F
F
N
N
H
N
N
F
O
N
H
F
Cl
Cl
O
6
7
N
N
4.77
4.50
4.81
4.47
<4.00
4.30
N
H
N
CN
NH
NH
HO
NH
8
9
4.57
5.01
4.56
5.01
4.75
4.08
NH
HO
NH2
HO
OH
N
O
O
N
O
N
10
4.92
4.93
<4.00
HO
OH
a
pIC50 values of antitumor activity against A549 lung cancer cells predicted by model 1 and model 2.
pIC50 values of antitumor activity against A549 lung cancer cells measured experimentally.
b
NCIH23 are lung cancer cells with K-ras and p53 gene mutations
and high expression of c-myc gene, which are the most frequent
events in process of carcinogenesis.38
Influence of tested compounds on viability of A549 and
NCIH23 cells was investigated using Promega MTS assay which
measures metabolic activity of the cells and is widely used as a
reliable way for measuring number of viable cells.39 In this test,
metabolic activity in cell population is measured via incubation
with tetrasolium salt that is cleaved into a colored formazan
product by metabolic active cells. Effect of new compounds on
proliferation of A549 and NCIH23 cancer cells was investigated
by BrdU assay which measures incorporation of labeled DNA pre-
cursor 5-bromo-20-deoxyuridine (BrdU) into genomic DNA during
S phase of cell cycle. The results for each of tested compounds are
reported as growth percentages compared with the untreated
control cells after 48 h of drug exposure. IC50 values were calcu-
lated using the program GraFit 5.0.12, and average values from
three independent experiments are listed in Table 3.
Compound 1 was chosen as internal standard to test the com-
patibility of biological results found in literature27 with our results.

´
2106
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
R
X
R
X
H
O
N
O
N
e
a, b
c
d
N
N
O
N
O
O
N
Cl
Cl
O
N
NH2
H
Cl
Cl
11
12a X=N, R=Cl
12b X=C, R=t-Bu
13a X=N, R=Cl
13b X=C, R=t-Bu
R
X
Ar =
HO
HO
HO
F
N
N
O
Ar
N
N
H
F
CN
Cl
a
b
c
d
1 X=N, R=Cl, Ar=a
2 X=N, R=Cl, Ar=b
3 X=C, R=t-Bu, Ar=b
4 X=N, R=Cl, Ar=c
5 X=C, R=t-Bu, Ar=c
6 X=N, R=Cl, Ar=d
Scheme 1. Reagents and conditions: (a) diethyl oxalate, NaOEt (21% in EtOH), rt, 10 min; (b) hydrazin hydrate, AcOH, rt, 30 min; (c) 2-chloro-5-(chloromethyl)pyridine or 1-
(chloromethyl)-4-(1,1-dimethylethyl)benzene, K2CO3, CH3CN, microwave, 120 °C, 30 min; (d) hydrazine hydrate, MeOH, 65 °C, 24 h; (e) aldehyde, EtOH, 30 °C, 5 h.
R
R
R
R
X
X
X
X
a
d
b, c
N
N
O
N
N
N
O
N
O
N
N
CN
OH
NH2
O
Cl
Cl
Cl
Cl
12a X=N, R=Cl
16a X=N, R=Cl
15a X=N, R=Cl
14a X=N, R=Cl
12b X=C, R=t-Bu
16b X=C, R=t-Bu
15b X=C, R=t-Bu
14b X=C, R=t-Bu
e
R
X
f
g
HO
OH
H
N
N
H
N
N
N
O
O
N
N
N
N
OH
O
OH
NH
N
NH
OH
Cl
Cl
Cl
7 X=N, R=Cl
8 X=C, R=t-Bu
9
10
Scheme 2. Reagents and conditions: (a) NaOH, EtOH, microwave, 110 °C, 30 min; (b) POCl3, 120 °C, 2 h; (c) NH3 gas, THF, rt, 20 min; (d) POCl3, pyridine, rt, 2 h; (e)
hydroxyamine hydrochloride, Et3N, EtOH, microwave, 80 °C, 30 min; (f) DCC, DMAP, CH3CN, carboxylic acid, rt, 24 h; (g) DCC, DMAP, CH3CN, carboxylic acid, 50 °C, 24 h.

´
2107
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
Excellent correlation between published data and our measure-
ment of its activity was obtained (Table 3). Due to later, compound
1 was found to be a good system for evaluation of newly synthe-
sized compounds.
between predicted and measured activity was obtained for com-
pounds 7 and 8 (Table 2). Their activities are particularly interest-
ing because these derivatives do not contain B ring, a structural
feature present in all other structures, and hydrazine linker was re-
placed with amidooxime linker. Therefore, the good inhibitory
activity of 7 and 8 is mostly determined by compound basicity
and possibility to make weak H-bond interactions with carbonyl
group, properties provided by amidooxime substitution. Acylation
of compound 8 led to compound 9 which have been predicted as
strongly active compound. Surprisingly, it showed lower activity
on viability test against A549 cells then reference compound 1
and parent compound 8. Possible explanation for lower activity
of 9 in MTS assay than it was predicted could be its spontaneous
conversion into compound 10. During preparation of compound
9, it was noticed that higher temperature or prolonged time of stir-
ring (>24 h) afforded compound 10 as main product. Since com-
pound 10 shows no activity on inhibition of A549 cell lines
(Table 3), it is very likely that lower activity of 9 is consequence
of its partial conversion into 10. However, compound 9 demon-
strated high proliferation inhibition on A549 cell line and showed
notable activity on both cells assays in case of NCIH23 cell line.
These results showed importance of different in vitro assays for
more accurate evaluation of biological activity.
Results obtained by MTS and BrdU assays (Table 3) showed
similar effects of the new compounds on growth of both, A549
and NCIH23, cell lines. Exception is compound 5 which is inactive
on viability and proliferating assay on A549, inactive in viability
assay on NCIH23, but active on proliferative BrdU assay and
therefore need further profiling. Also, compound 6 was not active
in both assays (due to low solubility). For all compounds, inhibi-
tion of growth measured with MTS assay has higher IC50 values
than inhibition measured with BrdU assay. Compounds 1, 2, 3,
4, 7 and 9 are active on both cell lines and inhibit growth as well
as proliferation of cell population. Compound 8 showed the same
range of growth inhibition in metabolic and proliferating assays.
Compound 10 showed no activity in MTS assay on both cell lines,
but showed strong activity in proliferation BrdU assay for both
cell lines. This results points to the absence of cytotoxic activity
and potential cytostatic activity of compound 10.
2.5. Anticancer activity of new pyrazole derivative
Comparison of predicted and experimentally measured antitu-
mor activities of new compounds using MTS assay is presented
in Table 2. Very good agreement between measured and predicted
activities was observed for compound 1. Introduction of additional
hydroxyl group into compound 1, led to compounds 2 and 3 which
were experimentally determined as the most potent ones, even
Finally, compound 10, as the only one with rigid linker in the
structure, proved to be the most interesting compound from ther-
apeutic point of view. Compound 10 showed low activities in MTS
assays, but it showed high activities in BrdU assays (Table 3). Sig-
nificant activities in BrdU assays reveal its potential cytostatic
activity that, in combination with absence or very low cytotoxic
activity, makes this compound a good starting point for further de-
sign of new, more potent and safer anticancer therapeutics.
more active than it was predicted (pIC50-pred = 4.90, pIC50
-
exp = 5.28 and pIC50-pred = 4.92, pIC50-exp = 5.11, respectively).
Also, compounds 2 and 3, are colored yellow and that can poten-
tially interfere with readouts of MTS assay. The same problem is
not connected to the BrdU proliferative assay (Table 3). Complete
agreement between predicted and experimentally determined
3. Conclusion
The presented studies demonstrate how 3D-QSAR analysis can
serve as a guideline for synthesis of new compounds. Apparently,
such approach proved to be useful in planning experiments and
it made experimental search for new compounds with antitumor
activity more efficient and faster. Predicted activities were in
accordance with experimentally determined ones for the majority
of compounds. In some cases, discrepancy between predicted and
measured activities was caused by low solubility of the compounds
in the media used for biological tests.
Antitumor activities of new pyrazole derivatives showed that
hydroxy and hydrazine groups (which might be responsible for
metabolic instability and toxic effect) can be substituted with fluo-
rine and amidooxime, respectively, without affecting the activity.
Substitution of hydrazine linker with rigid, 1,2,4-oxadiazoline moi-
ety resulted with the most interesting compound from the thera-
peutic point of view according to the additional biological
evaluation. Although compound 10 was inactive on inhibition of
A549 cells, it was found to have potential cytostatic activity due
to observed significant activities in BrdU assays. The later, together
with its low (if any) cytotoxic activity, makes compound 10 a
promising lead compound for future design of new anticancer
therapeutics.
activity was obtained for compound 4 (pIC50-pred = 4.78, pIC50
-
exp = 4.80), which was obtained from compound 2 by substitution
of hydroxyl groups with fluorine atoms. Its activity demonstrates
that it is possible to substitute hydroxyl group liable to glucoroni-
cation with more drug-like fluoro group40 and to maintain activity.
Unfortunately, tert-butylbenzyl analogue 5, proved to be inactive
on viability and proliferating assay on A549 (Table 3). Its inactivity
is probably consequence of its strong tendency to precipitate in
culture media, which implies that compound 5 might be active
due to structural features, but because of insolubility in aqueous
media it is impossible to experimentally determine its activity.
Since compound 5 showed some activity on inhibition of prolifer-
ation of NCIH23 (Table 3), its further profiling is needed. The com-
pound 6 was not active in both assays (Table 3). Its inactivity could
be attributed to very low solubility in DMSO media. Fair correlation
Table 3
Growth inhibitory and cell proliferation (IC50) of A549 and NCIH23 cells for the
compounds 1–10 at 48 h
Compound
A549
BrdU assay
IC50 M)
NCIH23
BrdU assay
IC50 M)
MTS assay
MTS assay
IC50
(l
M)
(l
IC50
(lM)
(l
1
2
3
4
5
6
7
8
9
10
4.70
5.20
7.80
16.01
>100
>100
50.50
17.70
83.22
>100
0.30
0.70
0.40
9.05
>100
>100
38.29
17.00
26.75
1.54
3.04
3.35
3.91
13.50
>100
>100
44.39
16.49
21.88
>100
0.27
0.57
0.3
4. Experimental section
4.1. D-QSAR analysis
7.86
21.88
>100
27.94
12.28
15.63
0.36
4.1.1. Dataset
Dataset of 51 compounds25–28 with measured antitumor activ-
ity against A549 lung cancer cells, (Table S1 in Supplementary
data), was used for building and validation of 3D-QSAR models.

´
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
2108
All the compounds from the dataset have common structural fea-
tures, 1-arylmethyl-3-aryl-1H-pyrazole as the main core, while
modifications were made at position N1, C3 and C5 where differ-
ently substituted aromatic rings were introduced. Negative loga-
rithmic value (pIC50) of measured antitumor activities was used
in deriving 3D-QSAR models. For the inactive compounds whose
IC50 values are not explicitly given in the literature, but just stated
as inactive or estimated as ‘>5 ꢁ 10ꢀ4 mol/dm3’, pIC50 was set to
3.30.
4.2. Chemistry
All solvents and reagents were used as supplied, unless noted
otherwise. Yields given are isolated yields. NMR spectra were re-
corded at 25 °C in DMSO-d6 with TMS as the internal standard on
Bruker Avance DRX500 spectrometer equipped with 5 mm diame-
ter inverse detection probe with z-gradient accessory, as well as
Bruker Avance DPX300 spectrometer using dual 1H probe. Melting
points were obtained using a Kofler apparatus and are uncorrected.
Mass spectra were recorded on Varian MAT 311 instrument (FAB),
and Platform LCZ or LCQ Deca instruments (ESI). The analyses were
performed using the following condition: An Acquity UPLC BEH
4.1.2. Calculation of molecular descriptors
For each compound, 3D structure was generated using Vol-
Surf+ program32,33 and Molecular Interaction Fields (MIFs) were
calculated by the program GRID.41 Grid spacing was set to 0.5 Å
and the following probes were used: H2O (the water molecule),
O (sp2 carbonyl oxygen atom), N1 (neutral NH group (e.g.,
amide)) and DRY (the hydrophobic probe). From the MIFs, Vol-
Surf+ derived series of 128 descriptors (independent of molecular
alignment) that refer to molecular size and shape, to hydrophilic
and hydrophobic regions and to the balance between them, to the
‘charge state’, to lipophilicity, and molecular diffusion, logP, logD,
to the presence/distribution of pharmacophoric groups as well as
to some other relevant ADME properties. VolSurf+ descriptors are
based on 3-D Molecular Interaction Fields, but are compressed to
alignment independent descriptors. The definition of all 128 Vol-
Surf+ descriptors is given in the VolSurf+ manual.32,33
C18 analytical column (50 mm ꢁ 2.1 mm, 1.7
lm packing diame-
ter) on Quattro micro™ (Micromass) mass spectrometer, using
external calibration) with the following two solvents; solvent A,
0.1% v/v solution of formic acid in water; solvent B, 0.1% v/v solu-
tion of formic acid in acetonitrile; start 97% A, linear gradient to
100% B in 1.5 min, then 0.4 min at 100% B, then linear gradient to
97% A in 0.1 min. Total run time 2 min. All tested compounds have
a purity of at least 90% according to NMR and UPLC analyses.
4.2.1. General procedure for the synthesis of 1–6
The mixture of carbohydrazide 13a-b (0.1 mmol) and corre-
sponding aldehyde (0.1 mmol) in EtOH (1 ml) was stirred at
30 °C for 5 h. The reaction mixture was concentrated under re-
duced pressure and purified by flash chromatography (0–100%
EtOAc/cyclohexane) affording 1–6 as pure compounds.
4.2.1.1. 3-(4-Chlorophenyl)-1-[(6-chloro-3-pyridinyl)methyl]-
N0-[(1E)-(2-hydroxyphenyl)methylidene]-1H-pyrazole-5-carbo-
4.1.3. Building and validation of 3D-QSAR models
The relationship between the structure based molecular
descriptors and biological activity of the dataset compounds
was determined using the partial least square (PLS) analysis.
The number of significant latent variables (nLV) and quality of
the models were determined using the leave-one-out (LOO) and
random group’s cross-validation procedures. In the later case, 5
random groups and 30 iterations were used. Standard deviation
of error of calculation (SDEC) and standard deviation of error of
prediction (SDEP) were calculated for each model. In order to
make more realistic validation of the predictive power of the
models, external validation was also performed. For that purpose,
dataset compounds were divided into training set (used to built
the model) and test set (used for external prediction). Test set
compounds were randomly chosen from the overall dataset using
random number generator (www.random.org). The size of the
test set was approximately 1/4 of the overall dataset. The SDEP
for external validation was calculated using homemade program.
In order to identify the descriptors with the highest (positive or
negative) impact on biological activity of the compounds, prod-
ucts between PLS coefficient and average value of the descriptor
were observed. Similar approach for identifying the most impor-
tant descriptors has been used in investigation of P-glycoprotein
interacting drugs.42
hydrazide (1).
The title compound 1 was obtained as a white
solid, from carbohydrazide 13a and 2-hydroxybenzaldehyde fol-
lowing the general procedure described above. Yield: 92%; mp
231–233 °C;27 1H NMR (500 MHz, DMSO) d: 5.82 (s, 2H, CH2),
6.85–6.99 (m, 2H, 5-ArH), 7.32 (t, 1H, 5-ArH), 7.46–7.57 (m, 4H,
1-ArH, 3-ArH), 7.59–7.61 (m, 1H, 5-ArH), 7.75 (d, 1H, J = 7.35, 1-
ArH), 7.82 (d, 2H, J = 8.35, 3-ArH), 8.39 (s, 1H, 1-ArH), 8.64 (s, 1H,
@CH), 10.93 (s, 1H, OH), 12.20 (s, 1H, NH); 13C NMR (125 MHz,
DMSO) d: 51.2, 105.5, 116.5, 118.8, 119.5, 124.4, 126.9, 128.9,
129.0, 130.9, 131.8, 132.7, 132.8, 135.4, 139.1, 148.3, 148.5,
149.1, 149.5, 155.2, 157.4; MS (ESI): m/z (94%) 466.0 [M+H]+.
4.2.1.2. 3-(4-Chlorophenyl)-1-[(6-chloro-3-pyridinyl)methyl]-N0
-[(1E)-(2,3-dihydroxy-phenyl)methylidene]-1H-pyrazole-5-car-
bohydrazide (2).
The title compound 2 was obtained as a
yellow solid, from carbohydrazide 13a and 2,3-dihydroxybenzal-
dehyde following the general procedure described above. Yield:
95%; mp 150–152 °C; 1H NMR (500 MHz, DMSO) d: 5.85 (s, 2H,
CH2), 6.73 (t, 1H, 5-ArH), 6.86 (d, 1H, J = 7.60, 5-ArH), 7.02 (d,
1H, J = 7.60, 5-ArH), 7.48–7.54 (m, 4H, 3-ArH, 1-ArH, 4-H), 7.73–
7.78 (m, 1H, 1-ArH), 7.82 (d, 2H, J = 8.36, 3-ArH), 8.38 (s, 1H, 1-
ArH), 8.61 (s, 1H, @CH), 9.35 (s, 1H, OH), 10.62 (s, 1H, OH), 12.17
(s, 1H, NH); 13C NMR (125 MHz, DMSO) d: 51.2, 105.5, 117.5,
119.0, 119.3, 124.4, 126.9, 128.9, 129.0, 130.9, 132.7, 132.8,
135.4, 139.1, 145.7, 146.1, 148.3, 148.5, 149.1, 149.5, 155.2; MS
(ESI): m/z (100%) 482.0 [M+H]+.
In deriving QSAR models, both pretreatment procedures that
are available in VolSurf+ program (raw and autoscaling) were
tested. Using the autoscaling pretreatment every variable is mean
centered and scaled to give it unit variance. With raw pretreat-
ment procedure, raw data matrix is used. Also, models that in-
clude inactive compounds, as well as the models without
inactive compounds were derived (Table 1).
4.2.1.3.
methylidene]-1-{[4-(1,1-dimethylethyl)phenyl]methyl}-1H-
pyrazole-5-carbohydrazide (3). The title compound 3 was
3-(4-Chlorophenyl)-N0-[(1E)-(2,3-dihydroxyphenyl)
Principal component analysis (PCA) was performed on the
molecular descriptors of dataset compounds used for building mod-
els 1 and 2. PCA loadings were used for identifying the descriptors
with the highest contribution to the first two principal components.
Usefulness of PCA in QSAR analysis has already been proved in
several cases.31,43
obtained as a yellow solid, from carbohydrazide 13b and 2,3-dihy-
droxy-benzaldehyde following the general procedure described
above. Yield: 92%; 1H NMR (500 MHz, DMSO) d: 1.23 (s, 9H,
3Me), 5.77 (s, 2H, CH2), 6.73 (t, 1H, 5-ArH), 6.83–6.88 (m, 1H, 5-
ArH), 6.98–7.04 (m, 1H, 5-ArH), 7.20 (d, 2H, J = 7.90, 1-ArH), 7.34
(d, 2H, J = 7.90, 1-ArH), 7.49 (t, 1H, 4-H), 7.52 (d, 2H, J = 8.35, 3-

´
2109
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
ArH), 7.82 (d, 2H, J = 8.35, 3-ArH), 8.60 (s, 1H, @CH), 9.31 (s, 1H,
OH), 10.56 (s, 1H, OH), 12.15 (s, 1H, NH); 13C NMR (125 MHz,
DMSO) d: 31.1, 34.3, 53.7, 105.5, 117.5, 119.0, 119.3, 125.2,
126.9, 127.2, 128.9, 129.0, 130.9, 132.7, 134.1, 135.4, 145.7,
146.1, 148.3, 148.5, 150.3, 155.2; MS (ESI): m/z (98%) 501.3 [M+H]+.
white solid, from 16b (50 mg, 0.14 mmol), following the method
described for 7.
1H NMR (300 MHz, DMSO) d: 1.23 (s, 9H, 3Me), 5.63 (s, 2H, CH2),
5.95 (s, 2H, NH, OH), 7.06–7.15 (m, 3H, 4-H, 1-ArH), 7.26–7.33 (m,
2H, 1-ArH), 7.45 (d, 2H, J = 8.52, 3-ArH), 7.76 (d, 2H, J = 8.52, 3-
ArH), 9.88 (s, 1H, NH); 13C NMR (75 MHz, DMSO) d: 31.1, 34.2,
54.3, 104.1, 125.4 126.8, 127.1, 128.9, 131.4, 132.3, 134.1, 137.3,
144.2, 148.5, 150.1; MS (ESI): m/z (90%) 383.4 [M+H]+.
4.2.1.4. 3-(4-Chlorophenyl)-1-[(6-chloro-3-pyridinyl)methyl]-
N0-[(1E)-(2,4-difluoro-phenyl)methylidene]-1H-pyrazole-5-car-
bohydrazide (4).
The title compound 4 was obtained as a
white solid, from carbohydrazide 13a and 2,4-difluorobenzalde-
hyde following the general procedure described above. Yield: 80%;
mp 218–221 °C; 1H NMR (300 MHz, DMSO) d: 5.81 (s, 2H, CH2),
7.22 (t, 1H, 5-ArH), 7.39 (t, 1H, 5-ArH), 7.45–7.57 (m, 4H, 3-ArH,
1-ArH, 4-H), 7.69–7.77 (m, 1H, 1-ArH), 7.82 (d, 2H, J = 8.28, 3-
ArH), 7.90–8.04 (m, 1H, 5-ArH), 8.37 (s, 1H, 1-ArH), 8.59 (s, 1H,
@CH), 12.25 (s, 1H, NH); 13C NMR (75 MHz, DMSO) d: 51.2, 104.6,
105.5, 112.7, 112.9, 118.4, 124.4, 126.9, 128.0, 129.0, 130.9, 132.6,
132.8, 135.5, 139.1, 140.7, 148.5, 149.1, 149.6, 155.5, 162.6; MS
(ESI): m/z (97%) 486.0 [M+H]+.
4.2.2.3.
phenyl]methyl}-1H-pyrazol-5-yl)-1,2,4-oxadiazol-5-yl]-1,2-ben-
zenediol (9). To a solution of amidooxime 8(15.3 mg, 0.04 mmol),
3-[3-(3-(4-Chlorophenyl)-1-{[4-(1,1-dimethylethyl)
N,N0-dicyclohexylcarbodiimide DCC (8.5 mg, 0.041 mmol) and dimethyl-
aminopyridine DMAP (0.4 mg, 0.0032 mmol) in dry acetonitrile (1.5 ml)
was added carboxylic acid (6.3 mg, 0.041 mmol) under vigorous stirring.
The reaction mixture was stirred at rt for 24 h. After reaction was concen-
trated the residue was purified by flash chromatography (0–100% EtOAc/
cyclohexane) to afford pure 9 (12 mg, 58%) as a white solid.
1H NMR (300 MHz, DMSO) d: 1.22 (s, 9H, 3Me), 5.55 (d, 2H,
2xNH), 5.85 (s, 2H, CH2), 6.77 (t, 1H, 5-ArH), 7.04 (d, 1H, J = 7.50,
5-ArH), 7.20 (s, 1H, 4-H), 7.25–7.38 (m, 4H, 1-ArH), 7.48 (d, 2H,
J = 8.50, 3-ArH), 7.62 (d, 1H, J = 7.50, 5-ArH), 7.81 (d, 2H, J = 8.50,
3-ArH), 9.48 (s, 1H, OH), 10.46 (s, 1H, OH); 13C NMR (75 MHz,
DMSO) d: 31.1, 34.2, 54.3, 104.1, 117.4, 119.1, 119.3, 125.4 126.8,
127.1, 128.9, 131.4, 132.3, 134.1, 137.3, 145.7, 146.1, 146.4,
148.5, 150.1, 155.6, 174.2; MS (ESI): m/z (93%) 519.4 [M+H]+
4.2.1.5. 3-(4-Chlorophenyl)-N0-[(1E)-(2,4-difluorophenyl)meth-
ylidene]-1-{[4-(1,1-dimethylethyl)phenyl]methyl}-1H-pyra-
zole-5-carbohydrazide (5).
The title compound 5 was
obtained as a white solid, from carbohydrazide 13b and 2,4-diflu-
orobenzaldehyde following the general procedure described above.
Yield: 78%; mp 74–76 °C; 1H NMR (300 MHz, DMSO) d: 1.21 (s, 9H,
3Me), 5.74 (s, 2H, CH2), 7.11–7.25 (m, 3H, 1-ArH, 5-ArH),7.26–7.42
(m, 3H, 1-ArH, 5-ArH), 7.43–7.57 (m, 3H, 3-ArH, 4-H), 7.81 (d, 2H,
J = 8.35, 3-ArH), 7.90–8.03 (m, 1H, 5-ArH), 8.58 (s, 1H, @CH), 12.06
(s, 1H, NH); 13C NMR (75 MHz, DMSO) d: 31.2, 34.3, 54.2, 104.6,
105.5, 112.7, 112.9, 118.4, 125.4, 126.9, 127.1, 128.0, 129.0,
130.9, 132.6, 134.0, 135.5, 140.7, 148.5, 150.2, 155.5, 162.6; MS
(ESI): m/z (100%) 507.5 [M+H]+.
4.2.2.4.
3-[3-(3-(4-Chlorophenyl)-1-{[4-(1,1-dimethylethyl)
phenyl]methyl}-1H-pyrazol-5-yl)-1,2,4-oxadiazol-5-yl]-1,2-ben-
zenediol (10).
0.04 mmol),
To a solution of amidooxime 8 (15.3 mg,
N,N0-dicyclohexylcarbodiimide
DCC (8.5 mg,
0.041 mmol) and dimethylaminopyridine DMAP (0.4 mg,
0.0032 mmol) in dry acetonitrile (1.5 ml) was added carboxylic
acid (6.3 mg, 0.041 mmol) under vigorous stirring. The reaction
mixture was stirred at 120 °C for 4 h. After reaction was concen-
trated the residue was purified by flash chromatography (0–100%
EtOAc/cyclohexane) to afford pure 10 (15 mg, 74%) as a white solid.
1H NMR (500 MHz, DMSO) d: 1.22 (s, 9H, 3Me), 5.85 (s, 2H, CH2),
7.14 (d, 1H, J = 8.30, 5-ArH), 7.22 (d, 2H, J = 8.30, 1-ArH), 7.33 (d,
2H, J = 8.30, 1-ArH), 7.50 (d, 2H, J = 8.50, 3-ArH), 7.55 (t, 1H, 5-
ArH), 7.62 (1, 1H, 4-H), 7.96 (d, 2H, J = 8.50, 3-ArH), 8.01 (d, 1H,
J = 7.50, 5-ArH), 9.40 (s, 1H, OH), 10.74 (s, 1H, OH); 13C NMR
(125 MHz, DMSO) d: 31.1, 34.2, 54.3, 106.7, 117.5, 119.0, 119.3,
125.4, 127.0, 127.1, 128.8, 130.4, 132.7, 134.0, 135.1, 145.7,
146.1, 149.5, 150.1, 157.4, 160.1, 175.0; MS (ESI): m/z (98%)
501.3 [M+H]+.
4.2.1.6. 3-(4-Chlorophenyl)-1-[(6-chloro-3-pyridinyl)methyl]-
N0-[(1E)-(4-cyanophenyl)methylidene]-1H-pyrazole-5-carbo-
hydrazide (6).
The title compound 6 was obtained as a white
solid, from carbohydrazide 13a and 4-formylbenzonitrile following
the general procedure described above. Yield: 84%; mp 238–
240 °C; 1H NMR (300 MHz, DMSO) d: 5.80 (s, 2H, CH2), 7.44–7.56
(m, 4H, 5-ArH), 7.70–7.77 (m, 1H, 4-H), 7.78–7.82 (m, 2H, 1-
ArH), 7.92 (s, 4H, 3-ArH), 8.36 (s, 1H, 1-ArH), 8.45 (s, 1H, @CH),
12.24 (s, 1H, NH); 13C NMR (75 MHz, DMSO) d: 51.2, 59.8, 105.5,
112.2, 118.6, 124.4, 126.9, 127.8, 129.0, 130.9, 132.7, 132.8,
135.4, 138.2, 138.4, 139.1, 146.7, 148.5, 149.1, 149.5; MS (ESI):
m/z (97%) 475.0 [M+H]+.
4.2.2. Synthesis of compounds 7–10
4.3. Biology
4.2.2.1. 3-(4-Chlorophenyl)-1-[(6-chloro-3-pyridinyl)methyl]-
N-hydroxy-1H-pyrazole-5-carboximidamide (7).
A mixture
4.3.1. Cell culture
of 16a (54 mg, 0.16 mmol), hydroxyamine hydrochloride (34 mg,
0.49 mmol) and Et3N (0.22 ml, 1.6 mmol) in ethanol (4 ml) was
heated under microwave irradiation at 80 °C for 30 min. After reac-
tion was concentrated the residue was purified by flash chroma-
tography (0–100% EtOAc/cyclohexane) to afford pure 7 (47 mg,
80%) as a white solid; mp 170–172 °C; 1H NMR (300 MHz, DMSO)
d: 5.69 (s, 2H, CH2), 6.01 (s, 2H, NH, OH), 7.16 (s, 1H, 4-H), 7.42–
7.52 (m, 3H, 1-ArH, 3-ArH), 7.65–7.67 (m, 1H, 1-ArH), 7.76 (d,
2H, J = 8.50, 3-ArH), 8.30–8.34 (m, 1H, 1-ArH), 9.95 (s, 1H, NH);
13C NMR (75 MHz, DMSO) d: 51.0, 103.9, 124.2, 126.7, 128.8,
131.5, 132.3, 132.9, 137.5, 138.9, 144.3, 148.6, 149.0, 149.3; MS
(ESI): m/z (100%) 362.3 [M+H]+.
The human cancer cell lines A549 and NCIH23 were purchased
from ATCC and grown in ATCC recommended cell media supple-
mented with 10% heat-inactivated fetal calf serum (BIOWEST’Cat.
No.: S181) and 1% antibiotic-antimycotic (Gibco, Cat. No. 15240-
062) in an atmosphere of 5% CO2 at 37 °C. A549 cells were grown
in F12K Nutrient Mixture Kaighn‘s modification (Cat. No.
21127022), and NCIH23 in RPMI Medium 1640 (GIBCO Cat. No.
A10491–01). All proliferation assays were performed in RPMI
1640 medium.
4.3.2. Compound preparation
All tested compounds were dissolved in DMSO to obtain 10 mM
stock solutions. Final concentrations of tested compounds were in
4.2.2.2.
methyl}-N-hydroxy-1H-pyrazole-5-carboximidamide
(8). The title compound 8 (41 mg, 74%) was obtained as a
3-(4-Chlorophenyl)-1-{[4-(1,1-dimethylethyl)phenyl]
range of 100–0.2
lM and with constant concentration of 1% DMSO
at all compound dilutions.

´
2110
I. Vujasinovic et al. / Bioorg. Med. Chem. 20 (2012) 2101–2110
7. Regan, J.; Breitfelder, S.; Cirillo, P.; Gilmore, T.; Graham, A. G.; Hickey, E.; Klaus,
4.3.3. MTS assay
B.; Madwed, J.; Moriak, M.; Moss, N.; Pargellis, C.; Pav, S.; Proto, A.; Swinamer,
A.; Tang, L.; Torcellini, C. J. Med. Chem. 2002, 45, 2994.
Cancer cells were seeded at concentration of 5 ꢁ 103 in 96 well-
flat bottom plates in medium F12 K (A549) or RPMI1640 (NCIH23)
culture medium supplemented with 10% FBS and incubated in an
atmosphere of 5% CO2 at 37 °C for 3 h. The cells were then treated
8. Pevarello, P.; Brasca, M. G.; Amici, R.; Orsini, P.; Traquandi, G.; Corti, L.; Piutti,
C.; Sansonna, P.; Villa, M.; Pierce, B. S.; Pulici, M.; Giordano, P.; Martina, K.;
Fritzen, E. L.; Nugent, R. A.; Casale, E.; Cameron, A.; Ciomei, M.; Roletto, F.;
Isacchi, A.; Fogliatto, G.; Pesenti, E.; Pastori, W.; Marsiglio, A.; Leach, K. L.; Clare,
P. M.; Fiorentini, F.; Varasi, M.; Vulpetti, A.; Warpehoski, M. A. J. Med. Chem.
2004, 47, 3367.
9. Manfredini, S.; Bazzanini, R.; Baraldi, P. G.; Guarneri, M.; Simoni, D.; Marongiu,
M. E.; Pani, A.; Tramontano, E.; La Colla, P. J. Med. Chem. 1992, 35, 917.
10. Park, H.-A.; Lee, K.; Park, S.-J.; Ahn, B.; Lee, J.-C.; Cho, H. Y.; Lee, K.-I. Bioorg. Med.
Chem. Lett. 2005, 15, 3307.
11. Tanitame, A.; Oyamada, Y.; Ofuji, K.; Fujimoto, M.; Iwai, N.; Hiyama, Y.; Suzuki,
K.; Ito, H.; Wachi, M.; Yamagishi, J. J. Med. Chem. 2004, 47, 3693.
12. Genin, M. J.; Allwine, D. A.; Anderson, D. J.; Barbachyn, M. R.; Emmert, D. E.;
Garmon, S. A.; Graber, D. R.; Grega, K. C.; Hester, J. B.; Hutchinson, D. K.; Morris,
J.; Reischer, R. J.; Ford, C. W.; Zurenko, G. E.; Hamel, J. C.; Schaadt, R. D.; Stapert,
D.; Yagi, B. H. J. Med. Chem. 2000, 43, 953.
with tested compounds at final concentrations of 100–0.2 lM for
48 h. After incubation cell proliferation and viability was measured
by the (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphe-
nyl)-2-(4-sulfophenyl)00H-tetrazolium, inner salt; MTS based assay
(Promega). Twenty microliters of MTS solution was added to the
cells and incubated at 37 °C in CO2 incubator. The conversion of
MTS into aqueous soluble formazan is accomplished by dehidro-
genase enzymes found in metabolically active cells. The quantity
of formazan product measured by the amount of 490 nm absor-
bance is directly proportional to the number of living cells in
culture.
13. Bekhit, A. A.; Fahmy, H. T. Y.; Rostom, S. A. F.; Baraka, A. M. Eur. J. Med. Chem.
2003, 38, 27.
14. Menozzi, G.; Mosti, L.; Fossa, P.; Mattioli, F.; Ghia, M. J. Heterocycl. Chem. 1997,
34, 963.
15. Sridhar, R.; Perumal, P. T.; Etti, S.; Shanmugam, G.; Ponnuswamy, M. N.;
Prabavathy, V. R.; Mathivanan, N. Bioorg. Med. Chem. Lett. 2004, 14, 6035.
16. Kees, K. L.; Fitzgerald, J. J.; Steiner, K. E.; Mattes, J. F.; Mihan, B.; Tosi, T.;
Mondoro, D.; McCaleb, M. L. J. Med. Chem. 1996, 39, 3920.
17. Bebernitz, G. R.; Argentieri, G.; Battle, B.; Brennan, C.; Balkan, B.; Burkey, B. F.;
Eckhardt, M.; Gao, J.; Kapa, P.; Strohschein, R. J.; Schuster, H. F.; Wilson, M.; Xu,
D. D. J. Med. Chem. 2001, 44, 2601.
18. Genin, M. J.; Biles, C.; Keiser, B. J.; Poppe, S. M.; Swaney, S. M.; Tarpley, W. G.;
Yagi, Y.; Romero, D. L. J. Med. Chem. 2000, 43, 1034.
4.3.4. BrdU assay (Amersham cell proliferation Biotrak ELISA
system, RPN250)
Cancer cells were seeded at concentration of 5 ꢁ 103 in 96 well-
flat bottom plates in medium RPMI1640 (GIBCO Cat. No. A10491-
01) supplemented with 10% FBS and incubated in an atmosphere
of 5% CO2 at 37 °C for 3 h. The cells were then treated with tested
compounds at final concentrations of 100–0.2
lM for 48 h. BrdU
was added to the cells in final concentration of 10
lM and reincu-
19. Hashimoto, H.; Imamura, K.; Haruta, J.-I.; Wakitani, K. J. Med. Chem. 2002, 45,
1511.
bated for 3 h. During the labeling period the pyrimidine analogue
BrdU is incorporated in place of thymidine into DNA of proliferat-
ing cells. After removing the culture medium the cells are fixed and
the DNA denatured for 30 min at rt. Solution for fixation was re-
20. Habeeb, A. G.; Rao, P. N. P.; Knaus, E. E. J. Med. Chem. 2001, 44, 3039.
21. Sakya, S. M.; DeMello, K. M. L.; Minich, M. L.; Rast, B.; Shavnya, A.; Rafka, R. J.;
Koss, D. A.; Cheng, H.; Li, J.; Jaynes, B. H.; Ziegler, C. B.; Mann, D. W.; Petras, C.
F.; Seibel, S. B.; Silvia, A. M.; George, D. M.; Lund, L. A.; St. Denis, S.; Hickman,
A.; Michelle, L.; Haven, M. L.; Lynch, M. P. Bioorg. Med. Chem. Lett. 2006, 16, 288.
moved and cells were treated with 200
for 30 min at rt following treatment with 100
l
l of blocking reagents
l peroxidase labeled
ˇ
22. Caleta, I.; Kralj, M.; Marjanovic´, M.; Bertoša, B.; Tomic´, S.; Pavlovic´, G.; Pavelc´,
l
K.; Karminski-Zamola, G. J. Med. Chem. 2009, 52, 1744.
BrdU solution for 90 min. The immune complexes are detected by
the subsequent substrate reaction and the resultant color was read
at 450 nm using Envision plate reader. The absorbance values cor-
relate directly to the amount of DNA synthesis and thereby to the
number of proliferating cells. Reduction of DNA synthesis was
calculated relative to control (untreated cells), the IC50 values were
calculated using GraFit 5.0.12 software.
23. Bertoša, B.; Kojic´-Prodic´, B.; Ramek, M.; Piperaki, S.; Tsantili-Kakoulidou, A.;
´
Wade, R.; Tomic, S. J. Chem. Inf. Model. 2003, 43, 1532.
24. Bertoša, B.; Aleksic´, M.; Karminiski-Zamola, G.; Tomic´, S. Int. J. Pharm. 2010,
394, 106.
25. Wei, F.; Zhao, B.-X.; Huang, B.; Zhang, L.; Sun, C.-H.; Dong, W.-L.; Shin, D.-S.;
Miao, J.-Y. Bioorg. Med. Chem. Lett. 2006, 16, 6342.
26. Xia, Y.; Dong, Z.-W.; Zhao, B.-X.; Ge, X.; Meng, N.; Shin, D.-S.; Miao, J.-Y. Bioorg.
Med. Chem. 2007, 15, 6893.
27. Xia, Y.; Fan, C.-D.; Zhao, B.-X.; Zhao, J.; Shin, D.-S.; Miao, J.-Y. Eur. J. Med. Chem.
2008, 43, 2347.
28. Fan, C.-D.; Zhao, B.-X.; Wei, F.; Zhang, G.-H.; Dong, W.-L.; Miao, J.-Y. Bioorg.
Med. Chem. Lett. 2008, 18, 3860.
Acknowledgment
29. Ding, X.-L.; Zhang, H.-Y.; Oi, L.; Zhao, B.-X.; Lian, S.; Lv, H.-S.; Miao, J.-Y. Bioorg.
Med. Chem. Lett. 2009, 19, 5325.
30. Zheng, L.-W.; Wu, L.-L.; Zhao, B.-X.; Dong, W.-L.; Miao, J.-Y. Bioorg. Med. Chem.
Lett. 1957, 2009, 17.
31. Fortuna, C. G.; Barresi, V.; Berellini, G.; Musumarra, G. Bioorg. Med. Chem. 2008,
16, 4150.
We gratefully acknowledge support by the Ministry of Science,
Education and Sport of the Republic of Croatia (Projects 098-
1191344-2860 and 098-0982933-2911).
32. Cruciani, G.; Crivori, P.; Carrupt, P.-A.; Testa, B. J. Mol. Struct.: THEOCHEM. 2000,
503, 17.
Supplementary data
33. Cruciani, G.; Pastor, M.; Guba, W. Eur. J. Pharm. Sci. 2000, 11, S29.
34. Madonna, S.; Beclin, C.; Laras, Y.; Moret, V.; Marcowycz, A.; Lamoral-Theys, D.;
Dubois, J.; Barthelemy-Requin, M.; Lenglet, G.; Depauw, S.; Cresteil, T.; Aubert,
G.; Monnier, V.; Kiss, R.; David-Cordonnier, M.-H.; Kraus, J.-L. Eur. J. Med. Chem.
2010, 45, 623.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2012.01.032.
References and notes
35. Eloy, F.; Lenaers, R. Chem. Rev. 1962, 62, 155.
36. Pearse, G. A., Jr.; Pflaum, R. T. J. Am. Chem. Soc. 1959, 81, 6505.
37. Malca-Mor, L.; Stark, A.-A. Appl. Environ. Microbiol. 1982, 44, 801.
38. Brooks, D. G.; James, R. M.; Patek, C. E.; Williamson, J.; Arends, M. J. Oncogene
2001, 20, 2144.
39. Cory, A. H.; Owen, T. C.; Barltrop, J. A.; Cory, J. G. Cancer Comm. 1991, 3, 207.
40. Müller, K.; Faeh, C.; Diedrich, F. Science 1881, 2007, 317.
41. Goodford, P. J. J. Med. Chem. 1985, 28, 849.
1. An Ipsos MORI report for the Global Lung Cancer Coalition, June 2010.
2. Jemal, A.; Siegel, R.; Xu, J.; Ward, E. CA Cancer J. Clin. 2010, 60, 277.
3. Seymour, L. Cancer Treat. Rev. 1999, 25, 301.
4. Tietze, L. F.; Steinmetz, A.; Balkenhohl, F. Bioorg. Med. Chem. Lett. 1997, 7, 1303.
5. Grosche, P.; Holtzel, A.; Walk, T. B.; Trautwein, A. W.; Jung, G. Synthesis 1961,
1999.
6. Penning, T. D.; Talley, J. J.; Bertenshaw, S. R.; Carter, J. S.; Collins, P. W.; Docter,
S.; Graneto, M. J.; Lee, L. F.; Malecha, J. W.; Miyashiro, J. M.; Rogers, R. S.; Rogier,
D. J.; Yu, S. S.; Anderson, G. D.; Burton, E. G.; Cogburn, J. N.; Gregory, S. A.;
Koboldt, C. M.; Perkins, W. E.; Seibert, K.; Veenhuizen, A. W.; Zhang, Y. Y.;
Isakson, P. C. J. Med. Chem. 1997, 40, 1347.
42. Zhuang, X.-M.; Xiao, J.-H.; Li, J.-T.; Zhang, Z.-Q.; Ruan, J.-X. Int. J. Pharm. 2006,
309, 109.
´
43. Tomic, S.; Bertoša, B.; Wang, T.; Wade, R. C. Proteins: Struct. Funct. Bioinf. 2007,
67(2), 435.