Novel substituted benzothiophene and thienothiophene carboxanilides and quinolones: Synthesis, photochemical synthesis, DNA-binding properties, antitumor evaluation and 3D-derived QSAR analysis

Article abstract of DOI:10.1021/jm300505h

A series of new N,N-dimethylaminopropyl- and 2-imidazolinyl-substituted derivatives of benzo[b]thienyl- and thieno[2,3-b]thienylcarboxanilides and benzo[b]thieno[2,3-c]- and thieno[3′,2′:4,5]thieno[2,3-c]quinolones were prepared. Quinolones were prepared by the reaction of photochemical dehydrohalogenation of corresponding anilides. Carboxanilides and quinolones were tested for the antiproliferative activity. 2-Imidazolinyl-substituted derivatives showed very prominent activity. By use of the experimentally obtained antitumor measurements, 3D-derived QSAR analysis was performed for the set of compounds. Higly predicitive 3D-derived QSAR models were obtained, and molecular properties that have the highest impact on antitumor activity were identified. Carboxanilides 6a-c and quinolones 9a-c and 11a were evaluated for DNA binding propensities and topoisomerases I and II inhibition as part of their mechanism of action assessment. The evaluated differences in the mode of action nicely correlate with the results of the 3D-QSAR analysis. Taken together, the results indicate which modifications of the compounds from the series should further improve their anticancer properties.

Full text of DOI:10.1021/jm300505h

Article
pubs.acs.org/jmc
Novel Substituted Benzothiophene and Thienothiophene
Carboxanilides and Quinolones: Synthesis, Photochemical Synthesis,
DNA-Binding Properties, Antitumor Evaluation and 3D-Derived QSAR
Analysis
Maja Aleksic,† Branimir Bertosa,‡ Raja Nhili,⊥ Lidija Uzelac,§ Ivana Jarak,†,# Sabine Depauw,⊥
́
̌
Marie-Helene David-Cordonnier,⊥ Marijeta Kralj,§ Sanja Tomic,*,‡ and Grace Karminski-Zamola*,†
́
̀
́
†Department of Organic Chemistry, Faculty of Chemical Engineering and Technology, University of Zagreb, Marulice
Box 177, HR-10000 Zagreb, Croatia
‡Division of Physical Chemistry and §Division of Molecular Medicine, “Ruđer Bosk
́
v trg 20, P.O.
̌
ovic”
́
Institute, Bijenick
̌
a cesta 54, P.O. Box 1016,
HR-10000 Zagreb, Croatia
⊥INSERM U837-JPARC (Jean-Pierre Aubert Research Center), Team “Molecular and Cellular Targeting for Cancer Treatment”,
Universite
France
́
Lille Nord de France, IFR-114, Institut pour la Recherche sur le Cancer de Lille, Place de Verdun, F-59045 Lille Cedex,
S
* Supporting Information
ABSTRACT: A series of new N,N-dimethylaminopropyl- and 2-imidazolinyl-substituted derivatives of benzo[b]thienyl- and
thieno[2,3-b]thienylcarboxanilides and benzo[b]thieno[2,3-c]- and thieno[3′,2′:4,5]thieno[2,3-c]quinolones were prepared.
Quinolones were prepared by the reaction of photochemical dehydrohalogenation of corresponding anilides. Carboxanilides and
quinolones were tested for the antiproliferative activity. 2-Imidazolinyl-substituted derivatives showed very prominent activity. By
use of the experimentally obtained antitumor measurements, 3D-derived QSAR analysis was performed for the set of
compounds. Higly predicitive 3D-derived QSAR models were obtained, and molecular properties that have the highest impact on
antitumor activity were identified. Carboxanilides 6a−c and quinolones 9a−c and 11a were evaluated for DNA binding
propensities and topoisomerases I and II inhibition as part of their mechanism of action assessment. The evaluated differences in
the mode of action nicely correlate with the results of the 3D-QSAR analysis. Taken together, the results indicate which
modifications of the compounds from the series should further improve their anticancer properties.
biological interest as various medical and biochemical agents.
Amidino-substituted benzo[b]thieno[2,3-c]quinolones, strong
ds-DNA/RNA intercalators, showed in general stronger and
more selective antitumor activity than acyclic precursors, which
did not intercalate at all. Compounds with one amidino group
as substituent have shown the best antiproliferative effect on
the cell lines tested. This could be explained by the permanent
positive charge placed on the amidino group, which may start a
number of additional interactions. Unexpectedly, the introduc-
tion of two amidino groups diminished the biological activity,
possibly because of steric effects. To elucidate the impact of
variation of thiophene vs benzene ring, we have also prepared
INTRODUCTION
■
Substituted heterocyclic quinolones exhibit several pharmaco-
logical activities and have therefore attracted considerable
attention from medicinal and synthetic organic chemists.1−4 As
a part of our continuing search for potential anticancer agents
related to heterocyclic quinolones, we have previously reported
synthesis and strong inhibitory activities on several human
tumor cell lines of thieno[3′,2′:4,5]thieno- and benzo[b]thieno-
[2,3-c]quinolones, containing a N,N-dimethylaminopropyl
substituent in the amide part of the molecule and on the
quinolone nitrogen (Figure 1).5,6 Moreover, we also reported
the synthesis and antiproliferative activity of various cyano-
and/or amidino-substituted benzo[b]thieno[2,3-c]quionolones
and their acyclic precursors,7 since it is well-known that
amidines are structural parts of numerous compounds of
Received: December 13, 2011
Published: May 24, 2012
© 2012 American Chemical Society
5044
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Figure 1. Structures of earlier prepared carboxamides, quinolones, and naphthyridones.5,6
Figure 2. Substituted 3-chlorobenzo[b]thiophene-2-carboxamides and benzo[b]thieno[2,3-c]quinolones.
amidino-substituted thieno[3′,2′:4,5]thieno- and thieno-
[2′,3′:4,5]thieno[2,3-c]quinolones as well as their acyclic
precursors, whereby both cyclic and acyclic compounds showed
pronounced antitumor activity.8 Replacement of the benzene
ring with pyridine, in the amino part of the amides and
naphthyridones, and the acyclic amidino groups with cyclic
amidino 2-imidazolinyl group resulted in an increase of the
antitumor activity. Some of the naphthyridone compounds
were active in the submicromolar range, with special selectivity
to MiaPaCa-2a, HeLa, and SW620 cells. The most promising
compounds inhibited tumor growth and topoisomerase activity,
which was related to their high DNA binding capacity, whereby
the 2-imidazolinyl-substituted derivatives showed the most
prominent activity.9 All of the above-mentioned considerations
prompted us to prepare novel N,N-dimethylaminopropyl- and
2-imidazolinyl-substituted derivatives of benzo[b]thieno-,
thieno[3′,2′:4,5]thieno-, and thieno[2′,3′:4,5]thieno[2,3-c]-
quinolones (Figures 2 and 3).
Figure 3. Substituted 3-chlorothieno[2,3-b]- and 3-chlorothieno[3,2-
b]thiophene-2-carboxamides and thieno[3′,2′:4,5]thieno- and thieno-
[2′,3′:4,5]thieno[2,3-c]quinolones.
performed the 3D-derived QSAR analysis using the Volsurf
based procedure. The main advantage of this approach is usage
of descriptors that have clear chemical or physical meaning and
are independent of the molecular superposition. Besides
structural properties, VolSurf+ descriptors describe the
ADME (absorption, distribution, metabolism, and excretion)
properties of the molecule as well. Similar approaches for
In order to identify molecular properties that have the largest
impact on antitumor activity of the investigated compounds, we
5045
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Scheme 1. Synthesis of 3-Chlorobenzo[b]thiophene-2-carboxamides and Benzo[b]thieno[2,3-c]quinolones
building 3D-derived QSAR models have already been proven to
be successful and useful in investigation of different classes of
the potentially biologically active compounds.10−12
compounds 5a−e. Since the protonation of compound 4e
was conducted in ethanol, a transesterification reaction gave
rise to the corresponding ethylated ester 5e, which was
confirmed by the disappearance of the methyl at the position of
1
RESULTS AND DISCUSSION
3.93 ppm in the H NMR spectra of compound 4e and the
■
appearance of a quadruplet at 4.35 ppm and a triplet 1.33 ppm,
Chemistry. All compounds (2a−19) shown in Figures 2
and 3 were prepared according to the Schemes 1 and 2, starting
from substituted 3-chlorobenzo[b]thiophenecarbonyl chlorides
(1a−f)6 and 3-chlorothieno[2,3-b]thiophene- or 3-chloro[3,2-
b]thiophene-2-carbonyl chlorides.8 Chlorides 1a−e gave in the
reaction with p-aminobenzoic acid 6-substituted N-(4-carbox-
yphenyl)-3-chlorobenzo[b]thiophene-2-carboxamides 2a−e,
which in the reaction with SOCl2 gave corresponding acid
chlorides 6-substituted N-(4-chlorocarbonylphenyl)-3-
chlorobenzo[b]thiophene-2-carboxamides 3a−e. 6-Substituted
2-{N-[4-N-(3-dimethylamino)propyl]carbamoyl}-3-
chlorobenzo[b]thiophene anilides 4a−e were prepared from
3a−e in the reaction with N,N-dimethylaminopropylamine.
The formation of the new amide bond was supported by the
appearance of a new triplet peak at approximately 8.5 ppm in
1
correspondingly, in the H NMR spectra of the compound 5e.
Carboxanilides 5a−e were photochemically transformed into 9-
substituted 2-[N-3-(dimethylamino)propyl]carbamoyl-6-oxo-
5,6-dihydrobenzothieno[2,3-c]quinoline hydrochlorides 7a−e
in methanol. Acid chlorides 1a and 1f reacted with p-
aminobenzonitrile to give the corresponding cyano-substituted
carboxamides 2f−h.7 Cyano-substituted carboxamides 2f−h
were, in the Pinner reaction, transformed into corresponding 2-
imidazolinyl-substituted carboxamides 6a−c. The cyano-sub-
stituted carboxamides 2f−h were photochemically converted in
MeOH/toluene medium into cyclic benzo[b]thieno[2,3-c]-
quinolones 8a−c in yields of 61−92.5%, while 2-imdazolinyl-
substituted carboxamide hydrochloride salts 6a−c were photo-
chemically converted into the corresponding benzo[b]thieno-
[2,3-c]quinolone hydrochloride salts 9a−c.
1
the H NMR spectra of compounds 4a−e. The hydrochloride
salts of N,N-dimethylaminopropyl-substituted carboxamides
5a−e were prepared by protonation 4a−e with the saturated
ethanolic or methanolic solutions of gaseous HCl. Protonation
of tertiary amino nitrogen resulted in the appearance of a broad
singlet at around 10 ppm in the 1H NMR spectra of
The photochemical dehydrohalogenation reaction of com-
pounds 6a and 6c was conducted in water to obtain 9a and 9c
with yields of 32.4−96.4%. The Pinner reaction of cyano-
substituted quinolone 8a gave 9a with 77% yields, but in the
5046
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
reaction 2-(imidazolin-2-yl)-5-methyl-6-oxo-5,6-dihydro[1]-
benzothieno[2,3-c]quinolin-6-one hydrochloride 11a and 5-
methyl-9-(imidazolin-2-yl)-6-oxo-5,6-dihydro[1]benzothieno-
[2,3-c]quinolin-6-one dihydrochloride 11b with yields of 43.9%
and 55.5%.
Scheme 2. Synthesis of 3-Chlorothieno[2,3-b]- and 3-
Chlorothieno[3,2-b]thiophene-2-carboxamides and
Thieno[3′,2′:4,5]thieno- and Thieno[2′,3′:4,5]thieno[2,3-
c]quinolones
N-(4′-Cyanophenyl)-3-chlorothieno[3,2-b]thiophene carbox-
anilide 12 and N-(4′-cyanophenyl)-3-chlorothieno[2,3-b]-
thiophene carboxanilide 13 (Scheme 2) were prepared from
the earlier described 3-chlorothieno[3,2-b]thiophene-2-carbon-
yl chloride or 3-chlorothieno[2,3-b]thiophene-2-carbonyl
chloride8 and 4-aminobenzonitrile. In the Pinner reaction,
they were converted into the corresponding 2-imdazolinyl-
substituted carboxanilides, N-[4′-(2-imidazolinyl)phenyl]-3-
chlorothieno[3,2-b]thiophene carboxanilide 14 and N-[4′-(2-
imidazolinyl)phenyl]-3-chlorothieno[2,3-b]thiophene carboxa-
nilide 15, as hydrochloride salts. Carboxanilide 12 was
photochemically cyclized in methanol/toluene medium into
2-cyano-6-oxo-5,6-dihydrothieno[3′,2′:4,5]thieno[2,3-c]-
quinoline 16 with a yield of 85% which, after Pinner reaction,
gave 2-(2-imidazolinyl)-6-oxo-5,6-dihydrothieno[3′,2′:4,5]-
thieno[2,3-c]quinoline hydrochloride salt 17 in a yield of
56.9%. The same quinolone 17 was obtained from carbox-
anilide 14 in a photochemical dehydrohalogenation reaction in
MeOH with a yield of 39.1%. Cyano-substituted quinolone 16
was N-methylated into quinolone 18, which after the Pinner
reaction gave 2-(2-imidazolinyl)-5-methy-6-oxo-5,6-
dihydrothieno[3′,2′:4,5]thieno[2,3-c]quinoline 19 in a yield of
65.1%. All 2-imidazolinyl-substituted carboxanilides, as well as
1
quinolones, show characteristic peaks in H NMR spectra for
cases of 8b and 8c into 9b and 9c, the reaction was not
successful.
Cyano-substituted quinolones 8a and 8b were N-methylated
with methyl iodide in the presence of NaH to obtain N-
methylated quinolones 10a and 10b, which gave in the Pinner
NH and NH protonated 2-imidazolinyl group at about 10−10.9
ppm.
Antiproliferative Activity. The tested compounds showed
diverse but mostly strong antiproliferative effect on the
Table 1. In Vitro Inhibition of Tumor Cell Growth by Compounds 5a−19
a
IC50 (μM)
compd
MOLT-4
NT
HCT116
SW620
NT
MCF-7
H460
5a
7
3
7
2
3
2
5
7
3
2
2
2
2
1
0.1
0.2
0.3
0.2
0.1
0.2
5b
5c
2
2
2
2
0.2
0.1
0.4
0.5
2
3
2
3
0.02
1
2
2
1
2
0.01
0.1
0.1
0.4
0.1
4
5d
5e
6a
0.02
0.03
0.2
0.1
2.1 1.4
0.9
1.9 0.1
1.8 0.1
2.8 0.5
1.7 0.02
1.2 0.1
1.3 0.2
1.4 0.1
1.6 0.01
0.1
1.5 0.1
6b
6c
2
2
1.4 0.1
2.6 0.4
7a
1
1
1
2
1
0.1
0.2
0.4
0.2
0.7
1
0.2
1
0.1
2
2
2
3
3
0.1
0.1
0.1
1
2
0.4
0.1
0.1
4
7b
7c
0.4 0.03
0.4 0.01
0.4 0.01
2
1
2
2
0.01
0.02
0.7
2
7d
7e
9a
2
1
0.01
0.5
23
3
0.2
1
0.8 0.4
0.8 0.3
0.9 0.3
1.3 0.1
≥10
0.7 0.2
1.2 0.2
≥10
1.4 0.1
1.5 0.1
≥10
1.6 0.1
1.2 0.1
≥10
9b
9c
3
0.4
10a
10b
11a
14
15
18
19
8.7 0.8
≥10
0.4 0.03
1.8 0.2
3.3 0.8
4.8 0.3
>10
>10
>10
>10
>10
>10
>10
>10
0.4 0.1
1.9 0.6
3.2 0.2
>10
0.2 0.03
0.9 0.2
2.1 0.4
>10
1.3 0.2
1.8 0.2
2.3 0.1
>10
0.6 0.02
0.4 0.1
1.3 0.1
>10
2
0.04
1
0.1
0.3 0.1
2
0.2
1
0.3
a
IC50: the concentration that causes 50% growth inhibition SD.
5047
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Figure 4. Concentration−response profiles of compounds 5e, 7d, 11a, and 14. PG = percentage of growth.
Figure 5. PCA scores plots derived using autoscaling pretreatment.
presented panel of cell lines (Table 1 and Figure 4). All
compounds except nitro-substituted carboxanilide 5d, unsub-
stituted 7a, and carbethoxy-substituted 7e quinolones along
with 2-imidazolinyl-substituted carboxianilides 6a−c and 14
and quinolones 9a and 9b, 11a, and 19 precipitated when
diluted with the cell culture medium at the maximal tested
5048
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
concentration. Because of the lower solubility in aqueous
medium of compounds 9c, 10a, 10b, and 15, the highest tested
concentration was 10−5 M.
(see the methodological part/computational analysis). The
models obtained for the antitumor activity of the compounds
on the MCF-7 cell line (breast carcinoma) were named 1 (A, B,
C, and D). Those for the antitumor activity on HCT116 (colon
carcinoma) were named 2 (A and B). Models for the antitumor
activity on SW620 (colon carcinoma) were named 3 (A, B, C,
and D). Models obtained for the antitumor activity on H460
cells (lung carcinoma) were named 4 (A and B), and models
derived for MOLT-4 cell line (T-lymphoblast leukemia) were
named 5. Statistical properties of the models derived using the
extended data set are summarized in Table 2. The properties of
the models derived using only compounds from the present
study are given in Supporting Information Table S3.
Modest antiproliferative effect was achieved with compounds
5a, 9c, 10a, and 10b. All other compounds showed very strong
and mostly nonselective cytotoxic effect to all tested tumor cell
lines. Compounds 5e, carbethoxy-substituted N,N-dimethyl-
aminopropylcarbamoylbenzo[b]thiophene anilide, and 7d,
nitro-substituted N,N-dimethylaminopropylcarbamoylbenzo-
[b]thieno[2,3-c]quinolin-6-one showed somewhat differential
effect on the tumor cell lines (Figure 4A,B). Importantly, the
strongest antiproliferative effect (IC50 in the submicromolar
range) showed the 2-imidazolinyl-substituted derivatives 9a,
11a, and 14 (Figure 4C,D).
No significant difference in IC50 values was observed between
carboxanilide (acyclic) and quinolone (cyclic) derivatives.
Similarly, our previously published results also showed similar
IC50 values after the treatment with acyclic (pyridylbenzo[b]-
thiophene-2-carboxamides) and cyclic (benzo[b]thieno[2,3-
c]naphthyridin-2-ones) compounds.9 Still, a minor structural
modification (the position of herocyclic nitrogen with respect
to amidine substitutent) altered the target specificity and
consequently the mode of action.9 Yet again, the 2-
imidazolinyl-substituted derivatives showed the most prom-
inent activity; i.e., the 2-imidazolinyl substituent acts as the best
pharmacophoric group in a diverse set of compounds.9,13,14
Analysis of the 3D Structure−Activity Relationship. In
order to elucidate the structural variables important for the
anticancer activity of the analyzed compounds, we performed a
detailed chemometric analysis.
Table 2. Statistical Properties of the Four Selected 3D-
Derived QSAR Models Derived for the Extended Data Set
a
d
e
f
g
h
model
nO
nLV
R2
SDEC
Q2
SDEP
b
1C
30
30
30
30
5
6
6
7
0.93
0.87
0.94
0.91
0.15
0.21
0.17
0.21
0.68
0.60
0.67
0.71
0.32
0.36
0.39
0.36
c
1D
b
3C
c
3D
a
1 and 3 are models for MCF-7 and SW620 cell lines, respectively.
b
c
3D-derived QSAR models obtained for the autoscaled data. 3D-
d
derived QSAR models obtained for the raw data. Number of objects
e
f
used to built the model. Number of latent variables. SDEC is
g
standard deviation of error of calculation. Q2 is the cross-validated
predictive performance and is given by Q2 = 1 − (∑nn=1(yexp(i)
−
ypred(i))2)/∑nn=1(yexp(i) − ⟨yexp⟩)2, where ypred(i) corresponds to the
predicted and yexp(i) to the experimentally determined inhibition pIC50
h
for the compound i, respectively. SDEP is the standard deviation in
cross-validated prediction and is given by SDEP = [(∑nn=1(yexp(i)
−
Principal Component Analysis (PCA). Principal component
analysis (PCA) was performed on the overall data set of 22
compounds using autoscaling pretreatment. The first two
principal components explained 61.3% of variance (the first
three PCs explained 71.6%).
ypred(i))2)/n]1/2
.
Models derived for the extended (30 molecules) data set
span a large range of biological activities, and although more
robust than the above-discussed models, they have good
internal and external predictive performances (Table 2).
Correlations between measured predicted activities by these
models are displayed in Figures 6 and 7.
Compounds 5a and 9c turned out to be outliers, and they
were omitted from all models. Activity measurements for the
compound 5a are/were very problematic (see Table 1), and the
9c structural variables significantly differ from those of the
other poorly active compounds, as shown by the PC analysis
and that might be ascribed to poor solubility of 9c.
External prediction of the 3D-derived QSAR models was
tested using compound 19 as well as compounds 5a and 9c
(Table 3). The external predictive performances of the models
derived using the large data set are better (for compound 19)
or comparable with the external prediction achieved by the
models derived considering only compounds from the present
study.
Molecular Descriptors with the Highest Impact on the
Antitumor Activity. In the 3D-derived QSAR models built with
the autoscaled structural variables, the influence of the
descriptors to the model can be estimated from their PLS
coefficients. Plots with the PLS coefficient for the models 1A,
2A, 3A, 4A, 1C, and 3C (Supporting Information Figures S3
and S4) revealed descriptors related to hydrophilic properties
of the compounds, their solubility, and the possibility to make
H-bonds as the most important ones. The following descriptors
were found to have the highest positive impact on the activity:
integy moments (IW1−IW3) which indicate concentration of
Distribution of the compounds in the PC space (PC scores)
was examined in order to try to find possible grouping of the
molecules. From the PCA scores plots (Figure 5) it can be seen
that all poorly to moderately active compounds (10a, 10b, 18)
are grouped together. In the space of the first two PCs (Figure
5), they are gathered in the upper-right quadrant. The
exception is 9c, a quinolone with two 2-imidazolinyl
substituents, with comparable activity but at the PC plots far
from them (Figure 5). However, it is separated from the other
compounds as well. The closest compound is carboxanilide 6c
which, similar to 9c, has two 2-imidazolinyl substituents.
Apparently, the structural descriptors of 9c differ from the other
quinolones, and they are mostly determined by the 2-
imidazolinyl substituents.
3D-Derived QSAR Models. The 3D-based QSAR models
were derived for the compounds analyzed in this work but also
for the larger set of compounds including 10 compounds
selected from the previous publications7,9 (Supporting
Information Table S2). For this purpose compounds were
selected that are structurally similar to the presented ones (i.e.,
do not significantly expand the space described by the first two
principal components shown in Figure 5; see Supporting
Information Figure S1) and that, at the same time, cover a wide
range of activity. However, previously the antitumor activity
tests were performed on different sets of cell lines, with only
MCF-7 and SW620 lines common to all studies.
The 3D-derived QSAR models were distinguished according
to the variable pretreatment, number of objects, and cell lines
5049
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Figure 6. Predicted vs experimental antitumor activity expressed as
pIC50, which is the negative logarithmic value of concentration that
causes 50% growth inhibition of the MCF-7 tumor cell lines: (A)
model 1C; (B) model 1D. Some of the compounds are not labeled on
the graph for clarity.
Figure 7. Predicted vs experimental antitumor activity expressed as
pIC50, which is the negative logarithmic value of concentration that
causes 50% growth inhibition of the SW620 tumor cell lines: (A)
model 3C; (B) model 3D. The compounds from the previous study
are labeled as “S”.
hydrophilic regions on one part of the molecule, 3D
pharmacophoric descriptors related to H-bonding (DRDRDO,
DRDRDR, DRACDO, DRDODO, DRDRAC), and solubility
properties (L2LgS). Descriptors that were found to have
negative correlation with antitumor activity according to the
autoscaling 3D-derived QSAR models were lipophilicity (log P,
especially in the case of the QSAR models derived for the 30-
object data set; see S4), percentage of un-ionized species at pH
range from 4 to 7 (%FU4−%FU7), the ratio of the polar
surface area (PSA) to the surface (S) (PSAR), the ratio of the
polar surface area (PSA) to the hydrophobic surface area
(HSA) (PHSAR), and the capacity factors (CW1−CW3)
which represent the ratio of the hydrophilic volume to the total
molecular surface.
Molecular Mechanism of Action. Compounds 6a−c, 9a−
c, and 11a were selected for further biochemical analysis to
investigate their activity regarding DNA binding and topo-
isomerases targeting of a small series of those compounds as
part of their mechanism of action.
DNA Binding Properties. We first focused on the DNA
binding ability of this series of molecules using spectrophoto-
metric measurements. For this purpose, the DNA binding
propensity was investigated globally through their capacity to
increase the melting temperature of double strand DNA during
the process of heat denaturation. The induced variations (ΔTm)
were determined for different compound/bp ratios, and the
results are presented in Table 4.
5050
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
single mode of binding in this concentration range. For all
other compounds, no isosbestic points were observed when
increasing the DNA concentration, suggesting that they do not
present a unique mode of binding to the DNA.
Table 3. Antitumor Activity of Compounds 19 (Top Rows),
5a (Three Bottom Rows), and 9c against Four Cell Lines
Predicted by 3D-Derived QSAR Models
a
IC50 (μM)
To get an insight into the mode of binding (groove binding
versus intercalation) of these compounds to the DNA, we
tested them for both circular dichroism and topoisomerase I
induced relaxation of supercoiled plasmid DNA (Figure 9).
Circular dichroism spectra were collected in BPE buffer
supplemented or not with increasing concentration of NaCl.
Incubation of 6c in the CT-DNA solution resulted in a large,
positively induced CD (ICD) peak with its maximum at ∼325
nm (Figure 9C, lanes light blue, pink, and purple) suggesting
the groove binding of the biimidazolinyl-substituted carbox-
anilide 6c. This strong positive ICD is also present upon
incubation in buffer containing 10 mM (Figure 9C, orange line)
or 100 mM NaCl (Figure 9D), suggesting strong binding in the
groove, even at high salt concentrations. Compounds 6a and 6b
induced a weak bisignate ICD around 340 and 355−390 nm,
respectively, suggesting intermolecular stacking of these
molecules in the presence of CT-DNA at the highest
concentration. Once again, this modified CD is also observed
upon addition of 10 (orange lines) and 100 mM (red lines)
NaCl.
The 3D-derived QSAR analysis revealed that integy mo-
ments, which indicate concentration of hydrophilic regions on
one part of the molecule, and 3D pharmacophoric descriptors
related to the hydrogen bond formation potency are positively
correlated with the biological activity. These descriptors,
combined with the large planar platform, are desirable for
DNA binders, and in agreement with the measured DNA
binding properties, they are accentuated in compounds 11a and
9c. Activity of 9c was significantly overestimated by the 3D-
derived QSAR models where it was considered as an outlier.
The experiments on DNA binding properties showed that this
compound is indeed a strong DNA binder with minor groove
and intercalative profiles. However, this DNA binding
propensity does not lead to a particularly strong antitumor
activity relative to the other compounds. A possible reason
might be low solubility.
HCT116
SW620
MCF-7
H460
0.3
experimental
1
0.1
0.3 0.1
2
0.2
1
predicted by models
1A, 2A, 3A, 4A
b
1.15
1.51
1.02
0.91
0.38
0.51
NT
2.45
2.69
1.86
2.88
7
1.26
3.39
c
1B, 2B, 3B, 4B
b
1C, 3C
c
1D, 3D
experimental
predicted by models
1A, 2A, 3A, 4A
7
3
7
3
0.1
b
1.82
2.88
1.18
2.10
0.72
1.82
≥10
1.03
1.41
1.78
3.16
≥10
1.21
2.14
c
1B, 2B, 3B, 4B
b
1C, 3C
c
1D, 3D
experimental
≥10
≥10
predicted by models
b
1A, 2A, 3A, 4A)
0.55
1.96
0.35
1.68
0.78
2.04
0.28
1.18
1.45
3.16
1.72
1.28
c
1B, 2B, 3B, 4B
b
1C, 3C
c
1D, 3D
a
b
The concentration that causes 50% growth inhibition. 3D QSAR
models obtained for the autoscaled data. 3D QSAR models obtained
for the raw data.
c
From this series, the carboxanilide 6c and quinolone 9c
derivatives bearing two 2-imidazolinyl substituents were
identified as the most potent compound to stabilize the DNA
helix with ΔTm larger than 20 °C at R = 0.25. Their
corresponding monoimidazolinyl substituted analogues (6a,b
and 9a,b) as well as the N-methylated derivative 11a were less
potent in the DNA helix stabilization. Interestingly, the
quinolones (9a,b and 11a) were more potent than the
carboxanilides (6a,b) with ΔTm between 10 and 20 °C at R
= 0.25 for the former and below 10 °C for the later derivatives.
Similarly, UV/visible spectrometry identified the lowest
hypochromic effects using the monoimidazolinyl substituted
carboxanilides 6a,b (Figure 8). An isosbestic point was
obtained using compound 6c at CT-DNA up to 100 μM
(exemplified panel, as indicated by the arrows), suggesting a
The DNA intercalation potency of this series was evaluated
using topoisomerase I induced DNA relaxation assays (Figure
10). Incubating a supercoiled plasmid DNA with an
Table 4. Variations of the DNA Melting Temperature upon Binding of the Compounds (ΔTm = Tm[Drug+DNA] − Tm[DNAalone]
)
a
Determined at Different Drug/bp Ratios (R)
a
nd: not determined. na: could not be analyzed. Light gray: strong binding when 10 °C < ΔTm < 20°C. Dark gray: very strong binding with ΔTm >
20°C.
5051
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Figure 8. UV/vis spectra of drug−DNA interaction. An amount of 20 μM each compound was incubated alone (top lines) or with step by step
increasing concentrations of CT-DNA from 10 to 200 μM (6a and 6c) or from 0.1 to 40 μM (6b, 9a−c, and 11a). Each spectrum is recorded from
230 to 430 nm relative to a cuvette containing the same amount of CT-DNA, with the upper right panel being an example of the previous panel
(arrows) focusing on the isosbetic point indicated with the presence of increasing concentrations of 6c (up to 100 μM).
Figure 9. Circular dichroism spectra. CT-DNA (200 μM, blue lines) was incubated with 50 μM 6a,b, 9a−c, or 11a in BPE buffer (purple lines)
supplemented with 10 mM (orange lines) or 100 mM (red lines) NaCl. For compound 6c, CT-DNA was incubated in BPE supplemented with 100
mM NaCl (D) or 10 mM NaCl (orange line in C) or in BPE (all other lanes in C) in the presence of 10 μM (light blue lines), 20 μM (pink lines), or
50 μM (purple, orange or red lines for BPE, BPE + 10 mM NaCl, or BPE + 100 mM NaCl buffers, respectively). The absence of intrinsic CD of the
molecules is visualized as green lines.
intercalating drug led first to a decrease of the number of
supercoils in topoisomers, up to a fully relaxed (circular) DNA
followed by switching back to topoisomers with a progressive
increase of the number of supercoils in the positive orientation,
opposite to the negative supercoils naturally occurring in
plasmids produced in bacteria.15 The experiment showed that
the condensed structures 9a−c and 11a are potent DNA
intercalating drugs. The intermediate stage corresponding to
the fully relaxed form (“Rel”, Figure 10) is reached using weak
concentrations of these compounds (0.5−2 μM), close to that
obtained using ethidium bromide (0.25 μM) in similar
experiments.15 In contrast, molecules with the carboxanilide
core 6a−c characterized by higher flexibility (and a smaller
coplanar platform) showed lower DNA intercalating properties,
as evidenced from fully relaxed DNA that is reached using ∼10
μM (6a, 6c) or ∼50 μM (6b).
The potential sequence specificity of the compounds was first
tested using DNase I footprinting. Sequence-selective binding
was observed for binding of the compounds to the DNA groove
(minor), regarding its circular dichroism spectra, 6c. Densito-
metric analysis revealed weak sequence selectivity of 6c for the
AT-rich sites (the gray boxed regions in the graph at the
bottom of Figure 11). The other tested compounds have not
shown any sequence selectivity but global inhibition of DNase I
cleavage, typical for DNA intercalators. Therefore, DNase I
footprinting is in agreement with previous results.
To comfirm the potent binding of 6c to AT-rich sequences,
we performed both melting temperature and spectroscopic
(UV/visible and circular dichroism) studies on poly(dAdT)2
5052
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
W5), and their potency to take part in hydrogen bond
formation as proton donors is 2 times higher than the potency
of 11a (descriptor WO1). On the other hand, values of %FU
descriptors, which were found to negatively correlate with the
biological activity by QSAR analysis, point to the possibility of
11a to become neutral at high pH values. Neutrality could not
be considered as a desirable characteristic for DNA binders but
can facilitate binding to the hydrophobic pocket of a protein,
suggesting that this compound could exert its cytotoxic effect
through another mechanism of action than 6c and 9c.
Cell Cycle Perturbations. We selected compounds 6a, 6c,
9a, and 11a as the most active and interesting representatives of
carboxanilides and quinolones that, as discussed earlier, showed
interesting and different DNA binding features. All compounds
were tested for cell cycle perturbation on the HCT 116 cell line
for 24 and 48 h using flow cytometry. Concentration that was
used was ∼IC50 (1 μM) or slightly higher than IC50 (5 μM) of
all compounds (Figure 15). All of the compounds induced
time- and concentration-dependent effects on the cell cycle.
Interestingly, only 6a at the higher concentration induced
statistically significant accumulation of cells in the G1 phase of
the cell cycle, pointing to cellular target(s) other than DNA for
this compound, which is in agreement with its low DNA-
binding activity but not with the Topo II inhibitory activity. All
other compounds induced accumulation of cells in the G2/M
phases, accompanied by the reduction of cells in the S phase,
correlating with the DNA damage mechanism. Namely, the
accumulation of cells in G2/M phases (a G2/M phase arrest)
points to a damage in the cell, which occurs during the S phase
(DNA replication) or before mitosis (aberrant mitotic spindle
formation). Thus, it is usually triggered by agents that induce
DNA damage or mitotic spindle damage. As expected, 6c
displayed a rather moderate effect on the cell cycle
perturbations, with an increase in the G2/M, after the
treatment with the higher concentration and after a prolonged
period of treatment. The quinolones 9a and 11a induced a
strong G2/M arrest, along with a drastic S phase reduction,
whereas compound 9a, which also induced apoptosis (indicated
by the subG1 percentage), is more effective. This was expected
from their high antiproliferative activity (submicromolar IC50)
and strong intercalation into double stranded DNA. However,
since 11a was shown to have different relative physicochemical
characteristics from the other compounds used to build the
QSAR model and since it seems that its high cytotoxicity is
achieved by different mode of action, it could also be possible
that its activity may interfere with the cellular tubulin level
which, leading to mitotic spindle damage, can also induce G2/
M arrest.
Figure 10. Topoisomerase I induced DNA relaxation. Various
concentrations of compounds (as indicated in micromolar at the top
of each lane) were incubated with supercoiled (Sc) pUC19 plasmid
prior to treatment with topoisomerase I enzyme (Topo I) to reveal
relaxed plasmid (Rel) and various topoisomers (topo). DNA,
untreated supercoiled plasmid DNA.
(Figure 12). The high ΔTm values obtained at low drug/DNA
ratios are evidence of the strong 6c binding to poly(dAdT)2.
CD spectra with a maximum peak shifting from ∼330 nm
(maximum at 20 μM, green line) to ∼365 nm (max at 100 μM,
red line) followed by rapid decrease from 100 to 200 μM
indicated binding of 6c into the groove of poly(dAdT)2 DNA.
Topoisomerase I or II Inhibition Properties. Functional
evaluation of topoisomerase I inhibition by compounds 6a−c,
9a−c, and 11a was first evaluated in terms of poisoning activity
on denaturing polyacrylamide gels using CPT as a reference
drug. It was found (Supporting Information Figure S5A) that
none of the tested compounds poisoned topoisomerase I (no
cleaved band was visualized). Further, we found that their
potency to suppress the topoisomerase I activity is related with
their binding to the DNA (Supporting Information Figure
S5B). The most potent compounds were 6c and 9b with full
inhibition obtained at 10 μM. Such suppressor activity is in
correlation with their strong DNA binding propensities, as
indicated from DNA melting temperature studies (Table 4).
Finally we evaluated compound activity as topoisomerase II
inhibitors. Etoposide was used as a reference drug. Its ability to
block the ligation step is classically revealed by the appearance
of a new DNA band corresponding to linearized plasmid DNA
(“Lin”). We found (Figure 13) that compounds 6a−c and 9a−
c, but not 11a, efficiently inhibited topoisomerase II enzyme.
Quantification relative to the same concentration (50 μM) of
etoposide (lanes “Etop”) showed that 6a was as potent as
etoposide on Topo II.
CONCLUSION
■
A series of carboxanilides (5a−e, 6a−c, and 14) and quinolones
(7a−e, 9a−c, 10a, 10b, 11a, 11b, 17, and 19) were prepared
and tested for the antiproliferative activity on various tumor cell
lines. In general, all compounds except cyano-substituted
quinolones showed very prominent activity, whereby 2-
imidazolinyl-substituted ones were among the most active.
Further, some of them, 6a−c, 9a−c, and 11a, were evaluated
for DNA binding propensities and topoisomerases I and II
inhibition in order to learn about their mechanism of action.
The 3D-derived QSAR analysis was performed for the data
set consisting of compounds from present and previous studies.
The highly predictive 3D-derived QSAR models were derived,
and molecular properties that have the highest impact on
We found compounds 6c, 9c, and 11a as the most
outstanding ones, the first two as the most potent DNA binder
and the third as the only one that did not inhibit topoisomerase
II activity. The results obtained for DNA binding propensities
and topoisomerases I and II inhibition prompted us to compare
descriptors for 6c, 9c, and 11a (Figure 14). Apparently, 6c and
9c are more hydrophilic than the third one (descriptors W1−
5053
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Figure 11. DNase I footprinting assays. Denaturing polyacrylamide gel in the presence of the various tested compounds (top) and the densitometric
analysis (bottom) of the sequence-selective binding of 6c on the 3′-end radiolabeled 265 bp DNA fragment. Gray boxes exemplify compound
localization on the DNA helix relative to its base pair composition. DNA, DNase-I untreated 265 bp fragment used as quality control. The G-track
was used to localize guanines in the DNA fragment and to deduce the migration position of each base within the known sequence.
antitumor activity of carboxanilides and quinolones were
identified. The results of 3D-derived QSAR analysis nicely
agree with the results obtained for DNA binding propensities
and topoisomerases I and II inhibition measurements. The
most potent DNA binders, 6c and 9c, are grouped together in
the PCA score plot (Figure 5). In the same plot compound
11a, for which no effect on topoisomerase II inhibition was
found, is close to the origin and isolated from the other
compounds. However, while 6c binds to DNA as a groove
binder, 9c binds to DNA as an intercalator. By five 3D-derived
QSAR models (1A, 1B, 1C, 4A, 4B) 6c was predicted to be the
most active compound in the series. Among the compounds in
the training set, the activities of compounds 6c and 7a were, on
average, the most overestimated. Compound 11a, for which no
effect on topoisomerase II inhibition was found, is on the PCA
score plot close to the origin and isolated from the other
compounds, indicating that its mode of action might differ from
those of the other active compounds. Its activity was much
better predicted with the models derived for the 30-objects data
set than with the models considering only compounds from the
present study, suggesting robustness of the previous. The cell
cycle perturbation experiments showed that 6c displayed rather
a moderate effect on the cell cycle perturbations while the
quinolones 9a and 11a induced a strong G2/M arrest along
with a drastic S phase reduction whereas compound 9a which
also induced apoptosis was more effective. Since 11a was
shown to have different relative physicochemical characteristics
from the other compounds used to build the QSAR model, it
seems that its high cytotoxicity is achieved by a different mode
of action. So it could be possible that its activity may interfere
with the cellular tubulin level which, leading to mitotic spindle
damage, can also induce G2/M arrest. From the obtained
analysis it is clear that there is no unique mode of action of the
active compounds analyzed in the present study to achieve their
activity. Taken together, the results indicate the most potent
directions of development of new anticancer compounds (from
the same series), namely, increasing concentration of hydro-
philic regions and their clustering at a certain region of the
molecular surface, improving H-bonding efficiency, and
increasing solubility. Specifically, the presence of 2-imidazolinyl
at the fourth position from the amido group (peptide bond)
boosts biological activity of thieno compounds.
EXPERIMENTAL SECTION
■
Chemistry. Melting points were determined on a Koffler hot stage
microscope and are uncorrected. IR spectra were recorded on FTIR-
5054
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Figure 12. Sequence-selective binding of 6c to AT-rich DNA. (A) Variation of the absorbency at 260 nm during the heating process of samples
containing poly(dAdT)2 alone (black) or incubated with increasing drug/DNA ratios of 6c. The ΔTm values derived from the graph are presented in
the table. (B) UV/visible spectrophotometry of 6c (20 μM) alone (black line) or incubated with increasing concentrations of poly(dAdT)2 from 10
to 200 μM. Major changes are observed below 100 μM (red line) as exemplified by green arrows. Minor spectral modifications appearing between
100 and 200 μM are localized using green arrows. (C) CD spectra of poly(dAdT)2 alone (black lines) or incubated with increasing concentrations of
6c from 1 to 20 μM (green lines), from 20 to 100 μM (red lines), and from 100 to 200 μM (blue lines).
and a Perkin-Elmer series II CHNS analyzer 2400. Where analyses are
indicated only as symbols of elements, analytical results obtained are
within 0.4% of the theoretical value. All compounds were routinely
checked by TLC with Merck silica gel 60F-254 glass plates. In
preparative photochemical experiments the irradiation was performed
at room temperature with a water-cooled immersion well with an
“Origin Hanau” 400 W high pressure mercury arc lamp using Pyrex
glass as a cutoff filter of wavelengths below 280 nm. All compounds
were routinely checked by TLC with Merck silica gel 60F-254 glass
plates.
General Method for the Synthesis of 2-Imidazolinyl
Substituted 3-Chlorobenzo[b]thiophene-2-carboxamide Hy-
drochlorides (6a−c). A stirred suspension of the corresponding
carboxamide in absolute ethanol or 2-methoxyethanol was cooled in
an ice−salt bath and was saturated with HCl gas. The reaction mixture
was maintained at room temperature until the nitrile band disappeared
(monitored by IR analysis at 2200 cm−1) and than diluted with diethyl
ether. The crude imidate was filtered off and was immediately
suspended in absolute ethanol. Ethylenediamine was added, and the
Figure 13. Topoisomerase II poisoning effect. The various tested
mixture was stirred at reflux for 24 h. The crude product was then
compounds (50 μM) were incubated with 350 ng of pUC19
filtered off, washed with diethyl ether to give a powder which was
supercoiled plasmid DNA prior to addition of Topo II as described.16
suspended in absolute ethanol and saturated with HCl (g). The
After migration, the gel was quantified for the percentage of
reaction mixture was stirred at room temperature for 24 h. The
topoisomerase II poisoning effect normalized to the reference drug
products were filtered off.
etoposide used at the same concentration.
N-[4′-(N′-2-Imidazolinyl)phenyl]-3-chlorobenzo[b]-
thiophene-2-carboxamide Hydrochloride (6a). Compound 6a is
prepared using above-described method. A solution of compound 2f7
(0.60 g, 1.92 mmol) in absolute ethanol (25 mL) was saturated with
ATR, Nicolet Magna 760, Perkin-Elmer 297, and Perkin-Elmer
Spectrum 1 spectrophotometers, with KBr disks. H and 13C NMR
1
HCl gas and stirred until the nitrile band disappeared. The crude
imidate was filtered off and was immediately suspended in absolute
ethanol (25 mL). Ethylenediamine (0.45 mL, 6.75 mmol) was added.
The mixture was stirred at reflux for 24 h, and 0.50 g (61%) of white
solid was obtained, mp >270 °C. IR ν/cm−1: 3078, 2968, 1647, 1605,
1529. H NMR (DMSO-d6) (δ ppm): 11.14 (s, 1H, NHamide), 10.84
(s, 2H, NHamidine), 8.20−8.18 (m, 1H, Harom), 8.16 (d, 2H, J = 8.9 Hz,
Harom), 7.99 (d, 2H, J = 8.9 Hz, Harom), 7,97−7,96 (m, 1H, Harom),
spectra were recorded on Varian Gemini 300, Bruker Avance DPX
300, and Bruker Avance DRX 500 spectrometers equipped with a 5
mm broadband probe or on a 14.1 T Varian 600 system (Varian, Palo
Alto, CA) equipped with a 3 mm broadband probe using TMS as an
internal standard in DMSO-d6. Mass spectra were recorded with an
Agilent 1100 series LC/MSD Trap SL spectrometer using electrospray
ionization (ESI). Elemental analyses for carbon, hydrogen, and
nitrogen were performed on a Perkin-Elmer 2400 elemental analyzer
1
5055
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Figure 14. Volsurf+ descriptors for compounds 6c (blue), 9c (red), and 11a (green). Only descriptors with absolute values above 15 are shown, and
only the most pronounced ones are labeled. Labels of the descriptors with values notably differing among 6c (blue), 9c (red), and 11a (green) are
given above the axis. List and description of all 128 Volsurf+ descriptors are given in the VolSurf+ manual.17,18
7.66−7.64 (m, 2H, Harom), 4.00 (s, 4H, CH2imidazolyl). 13C NMR
(DMSO-d6) (δ ppm): 164.0 (s), 159.7 (s), 143.6 (s), 136.8 (s), 135.6
(s), 131.3 (s), 129.9 (d, 2C), 127.8 (d), 126.2 (d), 123.6 (d), 122.6
(d), 120.5 (s), 119.7 (d, 2C), 117.3 (s), 44.2 (t, 2C).
mL) was saturated with HCl gas and stirred until the nitrile band
disappeared. The crude imidate was filtered off and was immediately
suspended in absolute ethanol (15 mL). Ethylenediamine (0.24 mL,
3.66 mmol) was added, and the mixture was stirred at reflux for 24 h.
The crude product was then filtered off, washed with diethyl ether to
give a powder which was suspended in absolute ethanol and saturated
with HCl (g). The reaction mixture was stirred at room temperature
for 24 h, and 0.30 g (78%) of yellow powder was obtained, mp >285
°C.
6-(N″-2-Imidazolinyl)-N-phenyl-3-chlorobenzo[b]thiophene-
2-carboxamide Hydrochloride (6b). A mixture of compound 2g7
(0.50 g, 1.60 mmol) in dry ethanol (25 mL) was saturated with HCl
gas and stirred until the nitrile band disappeared. The crude imidate
was filtered off and was immediately suspended in absolute ethanol
(25 mL). Ethylenediamine (0.35 mL, 5.25 mmol) was added. The
mixture was stirred at reflux for 24 h, and 0.24 g (35%) of white solid
was obtained, mp >270 °C. IR ν/cm−1: 3050, 2937, 2870, 1644, 1613,
1520. 1H NMR (DMSO-d6) (δ ppm): 10.83 (s, 2H, NHamidine), 10.75
(s, 1H, NHamide), 8.83 (s, 1H, Harom), 8.22 (d, 1H, J = 8.5 Hz, Harom),
8.12 (d, 1H, J = 8.5 Hz, Harom), 7,73 (d, 2H, J = 7.8 Hz, Harom), 7.41 (t,
2H, J = 7.4 Hz, Harom), 7.19 (t, 1H, J = 7.4 Hz, Harom), 4.07 (s, 4H,
CH2imidazolyl). 13C NMR (DMSO-d6) (δ ppm): 164.7 (s), 158.3 (s),
139.4 (s), 138.0 (s), 136.8 (s), 136.4 (s), 128.9 (d, 3C), 125.4 (d),
124.9 (d), 124.7 (d), 123.4 (d), 120.9 (s), 120.3 (s), 119.6 (s), 44.6 (t,
2C).
Method B. A solution of acyclic imidazolinyl-substituted 3-
chlorobenzo[b]thiophene-2-carboxamide 6a (0.10 g, 0.23 mmol) in
water (35 mL) was irradiated for 2 h at room temperature. The
reaction mixture was concentrated under the reduced pressure and
filtered off, and 0.08 g (96%) of yellow solid was obtained, mp >285
1
°C. IR ν/cm−1: 3061, 2957, 1664, 1597. H NMR (DMSO-d6) (δ
ppm): 12.72 (s, 1H, NHquinolone), 10.93 (s, 2H, NHamidine.), 9.13 (t, 1H,
J1 = 4.9 Hz, J2 = 3.7 Hz, Harom), 9.02 (s, 1H, Harom), 8.28 (t, 1H, J1 =
3.9 Hz, J2 = 5.1 Hz, Harom), 8.14 (d, 1H, J = 8.6 Hz, Harom), 7.70 (t, 3H,
J1 = 5.7 Hz, J2 = 2.7 Hz, Harom), 4.07 (s, 4H, CH2imidazolyl). 13C NMR
(DMSO-d6) (δ ppm): 164.3 (s), 158.1 (s), 141.9 (s), 141.4 (s), 135.3
(s), 135.0 (s), 133.8 (s), 128.5 (d), 128.0 (d), 126.9 (d), 126.2 (d),
124.3 (d), 124.2 (d), 117.5 (d), 117.3 (s), 116.0 (s), 44.7 (t, 2C).
General Method for the Synthesis of 2- and 9-Substituted 6-
Oxo-5,6-dihydro[1]benzothieno[2,3-c]quinolin-6-ones (9b, c).
A solution of 3-chlorobenzo[b]thiophene-2-carboxamide derivatives
(6b,c) in methanol or water was irradiated at room temperature with a
400 W high-pressure mercury lamp for 1.5 and 2 h. The solution was
concentrated, and the obtained solid was filtered off.
9-(2-Imidazolinyl)-6-oxo-5,6-dihydro[1]benzothieno[2,3-c]-
quinolin-6-one Hydrochloride (9b). A solution of 6b (0.10 g, 0.26
mmol) in methanol (80 mL) was irradiated for 2 h, and 0.08 g (88%)
of yellow solid was obtained, mp >285 °C. IR ν/cm−1: 3132, 2988,
1644, 1618. 1H NMR (DMSO-d6) (δ ppm): 12.46 (s, 1H, NHquinolone),
10.89 (s, 2H, NHamidine.), 9.20 (d, 1H, J = 8.8 Hz, Harom), 8.92 (s, 1H,
Harom), 8.84 (d, 1H, J = 8.0 Hz, Harom), 8.21−8.17 (m, 1H, Harom),
7.67−7.59 (m, 2H, Harom), 7.48−7.43 (m, 1H, Harom), 4.09 (s, 4H,
CH2imidazolyl). 13C NMR (DMSO-d6) (δ ppm): 164.8 (s), 157.9 (s),
141.4 (s), 139.8 (s), 138.2 (s), 135.9 (s), 129.8 (d), 127.0 (d), 125.5
(d, 2C), 124.2 (d), 123.4 (d), 121.0 (s), 117.4 (s), 117.3 (d), 45.0 (t,
2C).
6-(N″-2-Imidazolinyl)-N-[4′-(N′-2-imidazolinyl)phenyl]-3-
chlorobenzo[b]thiophene-2-carboxamide Dihydrocloride (6c).
A mixture of compound 2h7 (0.35 g, 1.04 mmol) in dry 2-
methoxyethanol (40 mL) was saturated with HCl gas and stirred
until the nitrile band disappeared. The crude imidate was filtered off
and was immediately suspended in absolute ethanol (30 mL).
Ethylenediamine (0.42 mL, 6.22 mmol) was added. The mixture
was stirred at reflux for 24 h, and 0.24 g (47%) of white solid was
obtained, mp >285 °C. IR ν/cm−1: 3050, 2911, 2751, 2680, 1659,
1
1608, 1510. H NMR (DMSO-d6) (δ ppm): 11.33 (s, 1H, NHamide),
10.97 (s, 2H, NHamidine), 10.69 (s, 2H, NHamidine), 8.92 (s, 1H, Harom),
8.24 (d, 1H, J = 8.5 Hz, Harom), 8.17 (dd, 1H, J1 = 8.6 Hz, J2 = 1.3 Hz,
Harom), 8.09 (d, 2H, J = 8.9 Hz, Harom), 7.99 (d, 2H, J = 8.9 Hz, Harom),
4.06 (s, 4H, CH2imidazolyl), 4.00 (s, 4H, CH2imidazolyl). 13C NMR
(DMSO-d6) (δ ppm): 165.0 (s), 164.5 (s), 159.6 (s), 143.9 (s), 139.7
(s), 137.0 (s), 136.3 (s), 130.4 (d, 2C), 126.0 (d), 125.5 (d), 124.0
(d), 121.6 (s), 120.4 (d, 2C), 118.0 (s), 45.0 (t, 2C), 44.7 (t, 2C).
2-(2-Imidazolinyl)-6-oxo-5,6-dihydro[1]benzothieno[2,3-c]-
quinolin-6-one Hydrochloride (9a). Method A. For the synthesis
of 9a was applied the method described for the compounds 6a−c. A
stirred suspension of 8a (0.30 g, 1.09 mmol) in absolute ethanol (20
5056
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
Figure 15. Effects of compound 6a, 6c, 9a, and 11a at 1 and 5 μM on the cell cycle distribution of HCT116 cells after the 24 and 48 h treatments.
The histograms represent the percentages of cells in the respective cell cycle phase (G1, S, and G2/M), along with the percentage of cells in the
subG1 (dead cells) obtained by flow cytometry.
g, 0.20 mmol) in water (75 mL) was irradiated for 1.5 h, and 0.03 g
2,9-(2-Diimidazolinyl)-6-oxo-5,6-dihydro[1]benzothieno-
[2,3-c]quinolin-6-one Hydrochloride (9c). A solution of 6c (0.10
(32%) of light brown solid was obtained, mp >285 °C. IR ν/cm−1:
5057
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
1
3127, 2973, 1654, 1598. H NMR (DMSO-d6) (δ ppm): 12.77 (bs,
and the experimentally determined antitumor activities against MCF-7,
SW620, HCT116, H460, and MOLT-4 cell lines. From the 23
synthesized compounds whose antitumor activities are presented in
this paper, 20 of them were used for building the 3D-derived QSAR
models. Besides, the 3D QSAR models were built considering only 17
(very) active compounds as well as a large set of 30 compounds, 20
from this publication and 10 from previous publications. Compounds
5a and 9c were shown to be outliers in all models and therefore were
excluded from the data set. Compound 19 was synthesized and
analyzed afterward and was used for the external validation of the
derived models. Negative logarithmic values of concentrations that
cause 50% growth inhibition of the cell lines (pIC50) were used as a
measure of biological activity in the 3D-derived QSAR models. For the
poorly active compounds whose IC50 values were not explicitly
measured but just estimated as >10 or ≥10, pIC50 was set to 4.301.
For each compound, the 3D structure was generated using the
VolSurf+ program17 and molecular interaction fields (MIFs) were
calculated by the program GRID.19 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, VolSurf+ derived a
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, to molecular diffusion, to log P, to log D, to the presence/
distribution of pharmacophoric groups, and to some other relevant
ADME properties. The definition of all 128 VolSurf+ descriptors is
given in the VolSurf+ manual. The 3D-derived QSAR models were
derived for the raw data (models labeled as B, D, and F) as well as for
the autoscaled ones (models labeled as A, C, and E). By the
autoscaling pretreatment, every variable is the mean centered and
scaled to give unit variance. Models A, B, C, and D were derived for
compounds from the present study, the first two for the 20-object data
set and the last two for the 17-objects data set. Models E and F were
derived for 30-object data set comprising 20 compounds from the
present study and 10 compounds from previous studies.7,9 Models for
different cel lines are distinguished by numbers 1−5 (MCF-7,
HCT116, SW620, H460, MOLT-4). The 3D-derived QSAR models,
i.e., the relationship between the 3D structure-based molecular
descriptors and measured antitumor activity, were 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) cross-validation procedures. Standard deviation
of error of calculation (SDEC) and standard deviation of error of
prediction (SDEP) were calculated for each model.
In order to identify the descriptors with the highest (positive or
negative) impact on biological activity of the compounds, PLS
coefficients were analyzed for the 3D-derived QSAR models built
using autoscaling (models 1A, 2A, 3A, 4A) pretreatment.
Principal component analysis (PCA) was performed in order to
determine the distribution of the compounds in the space of the
molecular descriptors and with an attempt to find an explanation for
the outliers 5a and 9c. Usefulness of PCA in 3D-derived QSAR
analysis has already been proved in several cases.10,20,21
Spectroscopic Evaluation of DNA Binding. UV/Visible
Spectrometry and DNA Melting Temperature Studies. CT-DNA
and double-stranded poly(dAdT)2 oligonucleotide (Sigma Aldrich,
France) were dissolved in water and dyalized overnight prior to use.
The analyzed compounds were prepared as 10 mM stock solutions in
DMSO, aliquoted, and stored at −20 °C and then freshly diluted in
the appropriate aqueous buffer.
The various compounds were prepared at 20 μM in 1 mL of BPE
buffer (6 mM Na2HPO4, 2 mM NaH2PO4, 1 mM EDTA, pH 7.1) in
the presence or absence of increasing concentrations of CT-DNA or
poly(dAdT)2 (from 10 to 100 μM with 10 μM steps and then from
100 to 200 μM with steps of 20 μM of base pairs). The UV/visible
spectra were recorded from 230 to 430 nm in a quartz cuvette of 10
mm path length using an Uvikon XL spectrophotometer and
referenced against a cuvette containing DNA at identical concen-
tration. The Tm values were obtained from the midpoint of the
1H, NHquinolone), 10.90 (bs, 4H, NHamidine), 9.49 (d, 1H, J = 8.6 Hz,
Harom), 9.17 (s, 1H, Harom), 8.94 (s, 1H, Harom), 8.30 (d, 1H, J = 8.3
Hz, Harom), 8.18 (d, 1H, J = 8.6 Hz, Harom), 7.79 (d, 1H, J = 8.6 Hz,
Harom), 4.10 (d, 8H, J = 4.0 Hz, CH2imidazolyl). 13C NMR (DMSO-d6)
(δ ppm): 164.9 (s), 164.8 (s), 158.1 (s), 142.2 (s), 141.5 (s), 135.2
(s), 129.3 (d), 127.8 (d), 125.6 (d), 125.0 (d), 121.9 (d), 121.5 (s),
118.1 (d), 117.6 (s), 117.4 (s), 116.8 (s), 45.1 (t, 2C), 44.9 (t, 2C).
2-(2-Imidazolinyl)-5-methyl-6-oxo-5,6-dihydro[1]-
benzothieno[2,3-c]quinolin-6-one Hydrocloride (11a). For the
synthesis of 11a was applied the method described for the compounds
6a−c. A mixture of compound 10a (0.15 g, 0.52 mmol) in dry ethanol
(20 mL) was saturated with HCl gas and stirred until the nitrile band
disappeared. The crude imidate was filtered off and was immediately
suspended in absolute ethanol (20 mL). Ethylenediamine (0.11 mL,
1.63 mmol) was added. The mixture was stirred at reflux for 24 h, and
0.11 g (56%) of white solid was obtained, mp >285 °C. IR ν/cm−1:
1
3210, 3101, 2973, 1649, 1618. H NMR (DMSO-d6) (δ ppm): 11.21
(bs, 2H, NHamidine), 9.27−9.23 (m, 1H, Harom), 9.18 (s, 1H, Harom),
8.41−8.29 (m, 2H, Harom), 8.04−8.00 (m, 1H, Harom), 7.76−7.72 (m,
2H, Harom), 4.09 (s, 3H, CH3), 3.87 (s, 4H, CH2imidazolyl). 13C NMR
(DMSO-d6) (δ ppm): 164.6 (s), 158.0 (s), 142.6 (s), 141.9 (s), 135.2
(s), 134.5 (s), 129.3 (d), 128.4 (d), 127.3 (d), 126.6 (d), 124.9 (d),
124.5 (d), 118.7 (s), 117.8 (d), 116.7 (s), 44.9 (t, 2C), 30.8 (q).
Antitumor Evaluation Assay. The experiments were carried out
on five human cell lines, which are derived from four cancer types. The
following cell lines were used: SW620 and HCT116 (colon
carcinoma), H460 (lung carcinoma), MCF-7 (breast carcinoma),
and MOLT-4 (T-lymphoblast leukemia). MCF-7, SW620, HCT116,
and H460 cells were cultured as monolayers and maintained in
Dulbecco’s modified Eagle's medium (DMEM), while MOLT-4 cells
were cultured in suspension in RPMI medium, both supplemented
with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 U/mL
penicillin, and 100 μg/mL streptomycin in a humidified atmosphere
with 5% CO2 at 37 °C.
The growth inhibition activity was assessed as described
previously.7−9,11,13,14 The panel cell lines were inoculated onto a
series of standard 96-well microtiter plates on day 0 at 1 × 104 to 3 ×
104 cells/mL, depending on the doubling times of a specific cell line.
Test agents were then added in 10-fold dilutions (10−8−10−4 M) and
incubated for a further 72 h. Working dilutions were freshly prepared
on the day of testing. After 72 h of incubation the cell growth rate was
evaluated by performing the MTT assay, which detects dehydrogenise
activity in viable cells. The absorbance (OD, optical density) was
measured on a microplate reader at 570 nm. The absorbance is directly
proportional to the number of living, metabolically active cells. The
percentage of growth (PG) of the cell lines was calculated according to
one of the following two expressions:
If (mean ODtest − mean ODtzero) ≥ 0, then PG = 100 × (mean
ODtest − mean ODtzero)/(mean ODctrl − meanODtzero). If (mean
ODtest − mean ODtzero) < 0, then PG = 100 × (mean ODtest − mean
ODtzero)/ODtzero, where the mean ODtzero is the average of optical
density measurements before exposure of cells to the test compound,
the mean ODtest is the average of optical density measurements after
the desired period of time, and the mean ODctrl is the average of
optical density measurements after the desired period of time with no
exposure of cells to the test compound.
Each test was performed in quadruplicate in at least two individual
experiments. The results are expressed as IC50, which is the
concentration necessary for 50% of inhibition. The IC50 values for
each compound are calculated from concentration−response curves
using linear regression analysis by fitting the test concentrations that
give PG values above and below the reference value (i.e., 50%). If,
however, for a given cell line all of the tested concentrations produce
PGs exceeding the respective reference level of effect (e.g., PG value of
50), then the highest tested concentration is assigned as the default
value, which is preceded by a “>” sign. Each result is a mean value of
three separate experiments.
Computational Analysis. The 3D-derived QSAR models were
derived using the Volsurf+ program17 generated molecular descriptors
5058
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
hyperchromic transition obtained from first-derivative plots. The
variation of melting temperature (ΔTm) was measured by subtracting
the melting temperature measurement of 20 μM CT-DNA or
poly(dAdT)2 incubated alone (control Tm) from that obtained with
DNA incubated with increasing concentrations of the various tested
compounds (drug/base pair ratio of 0.01−0.5) in 1 mL of BPE buffer
(ΔTm = Tm[Drug+DNA] − Tm[DNA alone]). The absorbency of DNA was
measured at 260 nm in quartz cells using a Uvikon 943
spectrophotometer thermostated with a Neslab RTE111 cryostat
every minute over a range of 20−100 °C with an increment of 1 °C
per min, and Tm values were deduced from the midpoint of the
hyperchromic transition.
Circular Dichroism. To evaluate the orientation of the drug relative
to the helicity of the DNA, the various drugs (50 μM) were incubated
with or without (control) a fixed or increasing concentration of CT-
DNA or poly(dAdT)2 (from 1 to 200 μM) in BPE or BPE
supplemented with 10 or 100 mM NaCl as specified in the figure
captions. The CD spectra were collected in a quartz cell of 10 mm
path length from 480 to 230 nm using a J-810 Jasco
spectropolarimeter at a controlled temperature of 20 °C fixed by a
PTC-424S/L Peltier type cell changer (Jasco) essentially as described
previously.22
Biochemical Approaches for DNA Binding. Radiolabeled DNA
Fragments. The 117 and 265 bp EcoRI and PvuII DNA fragments
were obtained from double digestion of the pBluscript plasmid
(Stratagene, La Jolla, CA) using EcoRI (two sites) and PvuII (1 site)
restriction enzymes for 1 h in their respective buffers. The generated
DNA fragment was 3′-end labeled at the generated 3′ available termini
of the EcoRI cleaved sites upon addition of α-[32P]dATP (3000 Ci/
mmol, PerkinElmer, France) using Klenow enzyme for 30 min at 37
°C. After separation of the two fragments of interest from the plasmid
remnant on a 6% native polyacrylamide gel in TBE buffer (89 mM
Tris base, 89 mM boric acid, 2.5 mM Na2 EDTA, pH 8.3), the
portions of the gel containing the 117 and 265 bp radiolabeled DNA
fragments were localized using a Molecular Dynamics STORM 860
and cut off from the gel, crushed, and dialyzed overnight. The eluted
radioactive materials were filtered through a 0.22 μm membrane
(Millipore). DNA fragments were finally ethanol-precipitated, dried,
and dissolved in an appropriate amount of Milli-Q water.
DNase I Footprinting. DNase I footprinting experiments were
performed essentially as previously described.23 Increasing concen-
trations of the various compounds were incubated with the 265-bp
radiolabeled DNA fragment for 15 min at 37 °C to ensure equilibrium
prior to DNA digestion upon addition of DNase I (final concentration,
0.002 unit/mL) in reaction buffer (20 mM NaCl, 2 mM MgCl2, 2 mM
MnCl2, pH 7.3). After mild digestion (4 min), the reaction was
stopped by freeze-drying, lyophilization, and subsequent dissolution in
4 μL of denaturing loading buffer (80% formamide solution containing
tracking dyes) prior to the mixture being heated at 90 °C for 4 min
and chilled on ice for another 4 min. DNA cleavage products were
resolved under polyacrylamide electrophoresis in denaturing con-
ditions (0.35 mm thick, 8% polyacrylamide containing 8 M urea) for
90 min at 65 W in TBE buffer. Gels were finally soaked in 10% acetic
acid, transferred to Whatman 3 MM paper, and dried under vacuum at
80 °C. After an exposition of the gels to storage screen for the
appropriate delay at room temperature, the results were collected
using a Molecular Dynamics STORM 860. Each base was localized
from comparison with the bands from guanines sequencing standard
(G-track) classically obtained using dimethyl sulfate (DMS) and
piperidine treatment of the same DNA fragment.
for 2 h at 120 V in TBE buffer. Gels were then stained in a bath
containing ethidium bromide before they were washed and photo-
graphed under UV light.
Topoisomerase I DNA cleavage assays were performed in the same
condition exept that the samples were loaded on a 1% agarose gel
prepared with ethidium bromide and that camptothecin (CPT, 20
μM) was used as a control for topoisomerase I poisoning effect.
In order to localize potential cleavage sites induced by topoisomer-
ase I poisoning or suppressor activities, the reaction was conducted
using the 3′-end-labeled 117 base pairs DNA fragment as already
published24 in the absence or presence of 20 μM CPT, respectively.
Topoisomerase II DNA cleavage assays were performed essentially
as described,16 with etoposide as a positive control for the Topo II
poisoning effect.
Cell Cycle Analysis. Tumor cells (2 × 105) were seeded per well
into a six-well plate. After 24 h the tested compounds were added at
various concentrations (as shown in the Results and Discussion). After
the desired length of time, the attached cells were trypsinized,
combined with floating cells, washed with phosphate buffer saline
(PBS), fixed with 70% ethanol, and stored at −20 °C. Immediately
before the analysis, the cells were washed with PBS and stained with
50 μg/mL propidium iodide (PI) with the addition of 0.2 μg/μL
RNase A. The stained cells were then analyzed with Becton Dickinson
FACScalibur (Becton Dickenson) flow cytometer (20 000 counts were
measured). The percentage of the cells in each cell cycle phase was
determined using the ModFit LT software (Verity Software House)
based on the DNA histograms. The tests were performed in duplicate
and repeated at least twice.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental details, spectroscopic data, and elemental analysis
results for compounds 2a−e, 3a−e, 4a−e, 5a−e, 7a−e, 8a−c,
10a, 10b, 11b, 14−19; elemental analysis results for 6a−c, 9a−
c, and 11a; antitumor activity of the compounds from the
previous publications (S1−10); statistical properties of the 3D-
derived QSAR models obtained using only the compounds
analyzed in the present study; PCA scores plot derived for 30-
data set; figures with predicted vs experimental antitumor
activity for models 1A, 1B, 3A, 3B, 4A, and 4B; PLS
coefficients for models 1A, 2A, 3A, 4A, 1C, and 3C; products
of the descriptor average values and the associated PLS
coefficients for models 1B, 2B, 3B, and 4B; and the
topoisomerase I poisoning of the various compound using
the 3′-end radiolabeled 117bp DNA fragment. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*For S.T.: phone, ++38514571251; fax, ++39514680245; e-
mail, sanja.tomic@irb.hr. For G.K.-Z.: phone, ++38514597215;
fax, ++39514597224; e-mail, gzamola@fkit.hr.
Present Address
#Center for Neuroscience and Cell Biology, Department of
Zoology, University of Coimbra, 3004-517 Coimbra, Portugal.
Notes
The authors declare no competing financial interest.
Topoisomerase I Mediated DNA Relaxation and Topo-
isomerases Cleavage Assays. Topoisomerase I mediated DNA
relaxation experiments were performed as previously described.15
Graded concentrations of the tested compounds were incubated in the
presence of supercoiled pUC19 plasmid DNA prior to the addition of
4 units of human topoisomerase I (Topogen, U.S.) at 37 °C for 45
min in relaxation buffer. The reactions were stopped upon addition of
SDS (0.25%) and proteinase K (250 μg/mL) for a 30 min incubation
at 50 °C. After addition of electrophoresis dye mixture, the DNA
forms were separated on a 1% agarose gel without ethidium bromide
ACKNOWLEDGMENTS
■
We greatly appreciate the financial support of the Croatian
Ministry of Science Education and Sports (Projects 125-
0982464-1356, 098-1191344-2860, 098-0982464-2514) and the
bilateral Hubert Curien partnership between Croatian and
French institutions (Cogito program) as the Egide Project No.
24765PH. The authors thank the Association Laurette Fugain,
5059
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060

Journal of Medicinal Chemistry
Article
(13) Hranjec, M.; Kralj, M.; Piantanida, I.; Sedic, M.; Suman, L.;
Pavelic, K.; Karminski-Zamola, G. Novel Cyano- And Amidino-
́
the Ligue Nationale Contre le Cancer (Comite
́
du Pas-de
eloppement
ional (FEDER, European Community), and the Region
Nord-Pas de Calais for grants (M-H.D.-C.), the Universite de
Lille-2 together with the Region Nord-Pas de Calais for a Ph.D.
Calais, Septentrion), the Fonds Europee
́ ́
n de Dev
Substituted Derivatives of Styryl-2-benzimidazoles and Benzimidazo-
[1,2-a]quinolines. Synthesis, Photochemical Synthesis, DNA Binding
and Antitumor Evaluation, Part 3. J. Med. Chem. 2007, 50, 5696−5711.
(14) Hranjec, M.; Piantanida, I.; Kralj, M.; Suman, L.; Pavelic, K.;
Karminski-Zamola, G. Novel Amidino-Substituted Thienyl- and
Furylvinyl-benzimidazole Derivatives and Their Photochemical
Conversion into Corresponding Diaza-cyclopenta[c]fluorenes. Syn-
thesis, Interactions with DNA and RNA and Antitumor Evaluation.
Part 4. J. Med. Chem. 2008, 51, 4899−4910.
Reg
́
́
́
́
fellowship to R.N., and the Institut pour la Recherche sur le
Cancer de Lille (IRCL) for technical expertise (S.D.), as well as
the IMPRT-IFR114 for access to the Strom 860 equipment.
REFERENCES
■
(15) Peixoto, P.; Bailly, C.; David-Cordonnier, M.-H. Topoisomerase
I-Mediated DNA Relaxation as a Tool To Study Intercalation of Small
Molecules into Supercoiled DNA. Methods Mol. Biol. 2010, 613, 235−
256.
(1) Letafat, B.; Emami, S.; Mohammadhosseini, N.; Faramarzi, M. A.;
Samadi, N.; Shafiee, A.; Foroumadi, A. Synthesis and Antibacterial
Activity of New N-[2-(thiophen-3-yl)ethyl]piperazinyl Quinolones.
Chem. Pharm. Bull. 2007, 55, 894−898.
́
(16) Serbetci̧ , T.; Depauw, S.; Prado, S.; Poree, F.-X.; Hildebrand,
(2) Wang, S. W.; Pan, S. L.; Huang, Y. C.; Guh, J. H.; Chiang, P. C.;
Huang, D. Y.; Kuo, S. C.; Lee, K. H.; Teng, C. M. CHM-1, a Novel
Synthetic Quinolone with Potent and Selective Antimitotic Antitumor
Activity against Human Hepatocellular Carcinoma in Vitro and in
Vivo. Mol. Cancer Ther. 2008, 7, 350−360.
(3) Hoang, H.; LaBarbera, D. V.; Mohammed, K. A.; Ireland, C. M.;
Skibo, E. B. Synthesis and Biological Evaluation of Imidazoquinox-
alinones, Imidazole Analogues of Pyrrolo-iminoquinone Marine
Natural Products. J. Med. Chem. 2007, 50, 4561−4571.
M.-P.; David-Cordonnier, M.-H; Michel, S; Tillequin, F. Benzo[c]-
[1,7] and [1,8]phenanthrolines Substituted with an Aminoalkyl Chain:
Synthesis and Evaluation of the Cytotoxicity and the Topoisomerase 1
Inhibition. Eur. J. Med. Chem. 2010, 45, 2547−2558.
(17) Cruciani, G.; Crivori, P.; Carrupt, P.-A.; Testa, B. Molecular
Fields in Quantitative Structure−Permeation Relationships: The
VolSurf Approach. J. Mol. Struct.: THEOCHEM 2000, 503, 17−30.
(18) Cruciani, G.; Pastor, M.; Guba, W. VolSurf: A New Tool for
Pharmacokinetic Optimization of Lead Compounds. Eur. J. Pharm. Sci.
2000, 11, S29−S39.
(4) Chen, M. H.; Fitzgerald, P.; Singh, S. B.; O’Neill, E. A.; Schwartz,
C. D.; Thompson, C. M.; O’Keefe, S. J.; Zaller, D. M.; Doherty, J. B.
Synthesis and Biological Activity of Quinolinone and Dihydroquino-
linone p38 MAP Kinase Inhibitors. Bioorg. Med. Chem. Lett. 2008, 18,
2222−2226.
(19) Goodford, P. J. A Computational Procedure for Determining
Energetically Favorable Binding Sites on Biologically Important
Macromolecules. J. Med. Chem. 1985, 28, 849−857.
́ ̌ ́ ́
(20) Tomic, S.; Bertosa, B.; Kojic-Prodic, B.; Kolosvary, I.
́
(5) Dogan Koruznjak, J.; Slade, N.; Zamola, B.; Pavelic, K.;
Stereoselectivity of Burkholderia cepacia Lipase towards Secondary
Alcohols: Molecular Modelling and 3D QSAR Approach. Tetrahedron:
Asymmetry 2004, 15, 1163−1172.
Karminski-Zamola, G. Synthesis, Photochemical Synthesis and
Antitumor Evaluation of Novel Derivatives of Thieno[3′,2′:4,5]thieno-
[2,3-c]quinolones. Chem. Pharm. Bull. 2002, 50, 656−660.
́ ̌
(21) Tomic, S.; Bertosa, B.; Wang, T.; Wade, R. C. COMBINE
̌ ́
(6) Dogan Koruznjak, J.; Grdisa, M.; Slade, N.; Zamola, B.; Pavelic,
Analysis of the Specificity of Binding of Ras Proteins to Their
Effectors. Proteins: Struct., Funct., Bioinf. 2007, 67 (2), 435−447.
(22) Lemster, T.; Pindur, U.; Depauw, S.; Lenglet, G.; Dassi, C.;
David-Cordonnier, M.-H. Photochemical Electrocyclisation of 3-
Vinylindoles to Pyrido[2,3-a]-, Pyrido[4,3-a]- and Thieno[2,3-a]-
carbazoles: Design, Synthesis, DNA-Binding and Antitumor Cell
Cytotoxicity. Eur. J. Med. Chem. 2009, 44, 3235−3252.
K.; Karminski-Zamola, G. Novel Derivatives of Benzo[b]thieno[2,3-
c]quinolones: Synthesis, Photochemical Synthesis and Antitumor
Evaluation. J. Med. Chem. 2003, 45, 4516−4524.
̌
(7) Jarak, I.; Kralj, M.; Suman, L.; Pavlovic,
́
G.; Dogan, J.; Piantanida,
̌
I.; Zinic,
́
M.; Pavelic,
́
K.; Karminski-Zamola, G. Novel Cyano- and N-
Isopropylamidino-Substituted Derivatives of Benzo[b]thiophene- 2-
carboxanilides and Benzo[b]thieno[2,3-c]quinolones: Synthesis, Pho-
tochemical Synthesis, Crystal Structure Determination and Antitumor
Evaluation. Part 2. J. Med. Chem. 2005, 48, 2346−2360.
(23) Peixoto, P.; Liu, Y.; Depauw, S.; Hildebrand, M.-P.; Boykin, D.
W.; Bailly, C.; Wilson, W. D.; David-Cordonnier, M.-H. Direct
Inhibition of the DNA Binding Activity of POU Transcription Factors
Pit-1 and Brn-3 by Selective Binding of a Phenyl-furan-benzimidazole
Dication. Nucleic Acids Res. 2008, 36, 3341−3353.
̌
̌
́ ́
(8) Jarak, I.; Kralj, M.; Piantanida, I.; Suman, L.; Zinic, M.; Pavelic,
K.; Karminski-Zamola, G. Novel Cyano- and Amidino-Substituted
Derivatives of Thieno[2, 3-b]- and Thieno[3, 2-b]thiophene-2-
carboxanilides and Thieno[3′,2′:4,5]thieno- and Thieno[2′,3′:4, 5]-
thieno[2,3-c]quinolones: Synthesis, Photochemical Synthesis, DNA
Binding and Antitumor Evaluation. Bioorg. Med. Chem. 2006, 14,
2859−2868.
(24) Genes
̀
, C.; Lenglet, G.; Depauw, S.; Nhili, R.; Prado, S.; David-
, F.-H. Synthesis
Cordonnier, M.-H.; Michel, S.; Tillequin, F.; Poree
́
and Biological Evaluation of N-Substituted Benzo[c]phenanthrolines
and Benzo[c]phenanthrolinones as Antiproliferative Agents. Eur. J.
Med. Chem. 2011, 46, 2117−2131.
́
(9) Ester, K.; Hranjec, M.; Piantanida, I.; Caleta, I.; Jarak, I.; Pavelic,
K.; Kralj, M.; Karminski-Zamola, G. Novel Derivatives of
Pyridylbenzo[b]thiophene-2-carboxamides and Benzo[b]thieno[2,3-
c]naphthyridin-2-ones: Minor Structural Variations Provoke Major
Differences of Antitumor Action Mechanisms. J. Med. Chem. 2009, 52,
2482−2492.
̌ ́ ́
(10) Bertosa, B.; Aleksic, M.; Karminiski-Zamola, G.; Tomic, S.
QSAR Analysis of Antitumor Active Amides and Quinolones from
Thiophene Series. Int. J. Pharm. 2010, 394, 106−114.
̌
́ ̌ ́
(11) Caleta, I.; Kralj, M.; Marjanovic, M.; Bertosa, B.; Tomic, S.;
́ ́
Pavlovic, G.; Pavelc, K.; Karminski-Zamola, G. Novel Cyano- and
Amidinobenzothiazole Derivatives: Synthesis, Antitumor Evaluation,
and X-ray and Quantitative Structure−Activity Relationship (QSAR)
Analysis. J. Med. Chem. 2009, 52, 1744−1756.
(12) Bertosa
Kakoulidou, A.; Wade, R.; Tomic,
Biological Activity of Molecules Based on Similarity of Their
Interaction Fields and the logP and logD Values: Application to
Auxins. J. Chem. Inf. Model. 2003, 43, 1532−1541.
̌
, B.; Kojic-
́
Prodic,
́
B.; Ramek, M.; Piperaki, S.; Tsantili-
S. A New Approach To Predict the
́
5060
dx.doi.org/10.1021/jm300505h | J. Med. Chem. 2012, 55, 5044−5060