Novel diamidino-substituted derivatives of phenyl benzothiazolyl and dibenzothiazolyl furans and thiophenes: Synthesis, antiproliferative and DNA binding properties

Article abstract of DOI:10.1021/jm901441b

A series of new diamidino-, diisopropylamidino-, and diimidazolinyl- substituted derivatives of phenyl benzothiazolyl and dibenzothiazolyl furans and thiophenes were successfully prepared and evaluated for their antiproliferative activity on tumor cell lines in vitro, DNA binding propensity, and sequence selectivity as well as cellular distribution. A strong antiproliferative effect of the tested compounds was observed on all tested cell lines in a concentration-dependent response pattern. In general, imidazolinyl-substituted derivatives and/or the thiophene core were in correlation with increased antiproliferative activity. Two compounds (2b and 3b) were chosen for biological studies due to their differential antiproliferative properties. The DNA binding properties of this new series of compounds were assessed and evidenced their efficient minor groove binding properties with preferential interaction at AT-rich sites. Both compounds also present nuclear subcellular localization, suggesting that their cellular mode of action implies localization in the DNA compartment and direct inhibition of DNA replication and induction of apoptosis.

Full text of DOI:10.1021/jm901441b

2418 J. Med. Chem. 2010, 53, 2418–2432
DOI: 10.1021/jm901441b
Novel Diamidino-Substituted Derivatives of Phenyl Benzothiazolyl and Dibenzothiazolyl Furans and
Thiophenes: Synthesis, Antiproliferative and DNA Binding Properties
§
Livio Racane, Vesna Tralic-Kulenovic, Sandra Kraljevic Pavelic, Ivana Ratkaj, Paul Peixoto,
z
ꢀ ‡
ꢀ
ꢀ ‡
ꢀ
ꢀ §
,z
Raja Nhili,z Sabine Depauw,z Marie-Paule Hildebrand,z,^ Marie-Helene David-Cordonnier,*
ꢀ ꢁ
,†
ꢂ
Kresimir Pavelic, and Grace Karminski-Zamola*
ꢀ §
†
ꢀ
Department of Organic Chemistry, Faculty of Chemical Engineering and Technology, University of Zagreb, Marulicev trg 20, P.O. Box 177,
HR-10000 Zagreb, Croatia, Department of Applied Chemistry, Faculty of Textile Technology, University of Zagreb, baruna Filipovica 28a 10000
‡
ꢀ
§
z
ꢀ
ꢂ
ꢀ
Zagreb, Croatia, Department of Biotechnology, University of Rijeka, Trg brace Mazuranica 10, 51000 Rijeka, Croatia, INSERM U837,
Jean-Pierre Aubert Research Centre (JPARC), Team “Molecular and Cellular Targeting for Cancer Treatment”, IRCL, Place de Verdun,
F-59045 Lille cedex, France, and ^Institut pour la Recherche sur le Cancer de Lille (IRCL), Place de Verdun, F-59045 Lille cedex, France
Received September 29, 2009
A series of new diamidino-, diisopropylamidino-, and diimidazolinyl-substituted derivatives of phenyl
benzothiazolyl and dibenzothiazolyl furans and thiophenes were successfully prepared and evaluated
for their antiproliferative activity on tumor cell lines in vitro, DNA binding propensity, and sequence
selectivity as well as cellular distribution. A strong antiproliferative effect of the tested compounds was
observed on all tested cell lines in a concentration-dependent response pattern. In general, imidazolinyl-
substituted derivatives and/or the thiophene core were in correlation with increased antiproliferative
activity. Two compounds (2b and 3b) were chosen for biological studies due to their differential
antiproliferative properties. The DNA binding properties of this new series of compounds were assessed
and evidenced their efficient minor groove binding properties with preferential interaction at AT-rich
sites. Both compounds also present nuclear subcellular localization, suggesting that their cellular mode
of action implies localization in the DNA compartment and direct inhibition of DNA replication and
induction of apoptosis.
Introduction
cancer cell lines MCF-7a and HeLa. In the alkaline comet assay,
some of these compounds showed dose-dependent DNA dama-
ging activity.16 Further, the newly prepared 2-acetyl-3-(6-meth-
oxybenzothiazo)-2-yl-aminoacrylonitrile showed significant
antiproliferative activity and strongly induced programmed cell
death in leukemia cells.17 The most recent article on the synthesis
of corresponding amino phenyl-substituted benzothiazoles
described the identification of promising scaffolds that are able
to inhibit different kinases with IC50 values in the nanomolar
range.18 Besides the benzothiazole core and its numerous poten-
tial biological implications, the central rings and attached sub-
stituents play a major role in the binding affinity and selectivity.
Indeed, for the particular case of DNA binding of benzothiazole
derivatives, engrafting amidine extremities to those structures, as
positively charged substituents, orients the function of the
molecule toward the binding to an electronegatively charged
biological molecule such as DNA in a manner similar to that of
berenil, pentamidine, stilbamidine, furamidine, and other di-
amidine heterocycles.19,20 Such diamidine, or other imidazoline,
charge effect was particularly well-studied for the furamidine
aromatic diamidine series developed by Boykin’s lab. Particu-
larly, the presence of terminal charges was shown to facilitate the
binding of the molecule to the minor groove of the DNA helix
but also the cellular distribution of the compounds toward the
nucleus.21,22 Positively charged extremities are not the sole
structural elements for strong DNA binding, and the three-
dimensional organization of the molecule is crucial for sequence-
selective DNA binding. It should ideally have a crescent shape,
In the last two decades, several heterocyclic compounds
from the series of benzothiazoles were synthesized and their
biological and pharmacological activity has consequently
been investigated. These compounds were extensively studied
for their antiallergic,1 anti-inflammatory,1,2 antitumor,3-7
and analgesic8,9 activities. Considering their mechanism of ac-
tion, it was shown that benzothiazole derivatives act as tyrosine
kinase10-13 and topoisomerases I and II inhibitors.14,15 Therefore,
various benzothiazole compounds are considerably interesting
due to their diverse pharmaceutical uses. A novel series of
optically active thiourea and their 2-aminobenzothiazole deriva-
tives were recently reported to exert a strong in vitro cytotoxicity
against mouse Ehrlich Ascites carcinoma (EAC) and two human
*To whom correspondence should be addressed. Phone: þ33 3 20 16 92
23. Fax: þ33 3 20 16 92 29. E-mail: marie-helene.david@inserm.fr. Phone:
þþ38514597215. Fax: þþ39514597224. E-mail: gzamola@fkit.hr.
aAbbreviations: bp, base pair; BPE, buffer phosphate EDTA; CD,
circular dichroism; CT-DNA, calf thymus DNA; cpd, compound;
dATP, deoxyadenosine triphosphate; dGTP, deoxyguanosine tripho-
sphate; DMEM, Dulbecco’s modified Eagle medium; DMSO, dimethyl
sulfoxide; EDTA, ethylene diamine tetraacetic acid; FBS, fetal bovine
serum; HeLa, cervical carcinoma; HEp-2, epidermoid carcinoma, lar-
ynx; HepG2, hepatocellular carcinoma; HP, hairpin; HT-29, human
colon carcinoma cells; ICD, induced circular dichroism; JK, normal
diploid human fibroblasts; MCF-7, breast epithelial adenocarcinoma,
metastatic; MiaPaCa-2, pancreatic carcinoma; MTT, 3-(4,5-di-
methylthiazole-2-yl)-2,5-diphenyl tetrazolium bromide, SW620, color-
ectal adenocarcinoma, metastatic; TBE, tris borate EDTA; Tm, melting
temperature.
r
pubs.acs.org/jmc
Published on Web 02/19/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2419
Scheme 1
Scheme 2
which allows the molecule to optimally fit the curve of the DNA
minor groove to optimize both van der Waals and electrostatic
contacts23 even if some linear derivatives could also fit the
minor groove through an original dimerization process, as
evidenced for dication 4,40-(pyrimidine-2,5-diyl)dibenzimid-
amide (DB1242),23 or mediated through a water molecule as
for the semilinear dicationic compounds 2-(40-carbaminidoylbi-
phenyl-4-yl)-1H-benzo[d]imidazole-6-carboximidamide (DB921)24
and 2-[4-(4-carbaminidoylphenoxy)phenyl]-1H-benzo[d]imid-
azole-6-carboximidamide (RT-29).25
Our previously obtained results showed that antiprolifera-
tive activity of cyano-, amidino-,26 and amino27-substituted
2-phenylbenzothiazole derivatives strongly depends on the
position of the substituent on the 2-phenylbenzothiazole
skeleton, as well as on the type of amidino substituent
attached. We found that, in a series of unsubstituted,
N-isopropyl-substituted, as well as 2-imidazolinyl mono- and
bisamidino derivatives of 2-phenylbenzothiazole, N-isopropyl-
substituted amidine possesses less pronounced antiproliferative
activity on tested tumor cells.
In relation with the above considerations, we designed
and efficiently synthesized new diamidino-, diisopropylami-
dino-, and diimidazolinyl-substituted derivatives of phenyl
benzothiazolyl and dibenzothiazolyl furans and thiophenes
and evaluated their antiproliferative activity on tumor
cell lines in vitro, DNA binding propensity, and sequence
selectivity as well as cellular distribution. In addition, two
compounds were chosen according to their differential effect
for further biological studies, including the cell cycle analysis
and apoptosis induction in order to reveal a more detailed
picture on the possible antiproliferative mechanisms and/or
targets.
Results and Discussion
Chemistry. The synthesis of the bisamidino-substituted
benzothiazolyl derivatives was performed by two different
approaches according to Schemes 1 and 2. Unsymmetrical
bisamidino monobenzothiazolyl derivatives 2a-3b were
prepared from corresponding bisnitriles 1a and 1b by Pinner
reaction previously described.28 Following this approach for
the synthesis of symmetrical bisamidino dibenzothiazolyl
derivatives, the corresponding bisnitriles have to be pre-
pared. By condensation reaction of 4-amino-3-sulfanylbenzo-
nitrile29 with dichloride of 2,5-furandicarboxylic acid, only
bisnitrile 7 was prepared. Unfortunately, this approach
failed when bisnitrile 7 was employed in the first step of the
Pinner reaction due to its very low solubility even by using
solvents such as 2-methoxyethanol and 2-(2-ethoxyethoxy)-
ethanol and prolonged reaction time up to 3 weeks. The
compound 7 did not convert into an imidoyl ether hydro-
chloride, and only a starting nitrile was recovered.
Recently, a second approach for the synthesis of amidino-
substituted benzothiazolyl derivatives was developed.30 We
found a simple and efficient method for the synthesis of
amidino 8 and 2-imidazolinyl-substituted 2-aminothio-
phenol 9 as the key precursor for this new synthetic approach
outlined in Scheme 2. Condensation reaction of commer-
cially available 2,5-furan and 2,5-thiophene dicarboxylic
acid and amidino-substituted 2-aminothiophenole 8 was
performed in polyphosphoric acid (Method A). The bisami-
dino dibenzothiazolyl compounds 10a-11b were isolated as
hydrochloride salts in a low to moderate yield of about
30-60%. To improve the yield, we tried to carry out the
reaction of amidino-substituted 2-aminothiophenole 8 with

2420 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Racaneꢀ et al.
Table 1. Inhibition Effects of Compounds 2a-11b on the Growth of Tumor Cells In Vitroa
IC50b (μM)
cell lines
substance no.
HeLa
MCF-7
HEp-2
HepG2
SW620
MiaPaCa-2
JK
2a
2b
13.91
4.11
0.13
0.05
6.52
0.15
4.89
0.12
0.24
0.04
0.03
0.17
0.19
0.05
0.37
0.29
51.39
30.73
0.37
72.43
48.19
0.40
50.56
38.13
0.39
0.50
0.04
61.05
8.72
0.02
3a
3b
<0.01
0.17
0.01
0.15
>100
0.47
0.09
7.97
0.19
>100
0.65
0.01
0.09
10a
10b
11a
11b
0.29
30.53
0.21
<0.01
0.06
0.06
<0.01
0.07
<0.01
>100
0.17
>100
0.27
a The results are presented as IC50 values in μM. The cell growth rate was evaluated by performing the MTT assay: experimentally determined
absorbance values were transformed into a cell percentage growth (PG) using the formulas proposed by National Cancer Institute and described
previously in Gazivoda et al.39 b IC50; 50% inhibitory concentration, or compound concentration required to inhibit tumor cell proliferation by 50%.
The IC50 values were calculated from dose-response curves using linear regression analysis by fitting the mean test concentrations that give PG values
above and below the reference value. If, however, all of the tested concentrations produce PGs exceeding the respective reference level of effect (e.g., PG
value of 50) for a given cell line, the highest tested concentration is assigned as the default value (>100). Each test point was performed in quadruplicate
in three individual experiments.
corresponding diacyl chlorides 5a and 5b in acetic acid
(Method B). Conversion of furan and thiophene dicarbo-
xylic acids into diacyl chlorides was achieved with thionyl
chloride quantitatively, and the corresponding dichlorides
were used in the condensation reaction. The overall yield for
Method B was much better than in Method A, and for
isolated diamidino dihydrochloride salts 10a-11b, yield
was 72-88%. The structures of the new compounds were
confirmed by spectroscopic methods: IR, 1H NMR, and 13
NMR spectra, as well as MS and elemental analysis.
C
Figure 1. Cyano-disubstituted, isopropylamidino-disubstituted-,
imdazolinyl-disubstituted 2-(4-phenyl)-5-(6-benzothiazolyl)furan
andthiophene (1a-3b), cyano-disubstituted-, amidino-disubstituted,
and imidazolinyl-disubstiuted 2,5-(6-benzothiazolyl)furans and thio-
phenes (7-11b).
Antiproliferative Activity. The antiproliferative effect of
compounds 2a, 2b, 3a, 3b, 10a, 10b, 11a, and 11b was
evaluated in vitro on a panel of seven human cell lines derived
from different cancer types including HeLa (cervical
carcinoma), SW620 (colorectal adenocarcinoma, metastatic),
MiaPaCa-2 (pancreatic carcinoma), MCF-7 (breast epithelial
adenocarcinoma, metastatic), HEP-2 (epidermoid carcinoma,
larynx), HepG2 (hepatocellular carcinoma), or JK (normal
diploid human fibroblasts). Our previous results obtained by
employing different benzothiazole derivatives have shown a
strong antiproliferative activity for this class of compounds on
several tumor cell lines.7,26,27 The results obtained in this study
were therefore partially expected and are summarized and
presented in Table 1. The majority of compounds demon-
strated a strong antiproliferative effect on all tested cell lines
and exerted a concentration-dependent response pattern.
Especially, this was observed on HeLa, MCF-7, and MiaPa-
Ca-2 cell lines, where all compounds strongly inhibited the cell
growth at submicromolar concentrations and higher (Table 1).
In particular, the strongest antiproliferative effect was ob-
served for three imidazolinyl-substituted derivatives 3a, 3b,
and 11b and for one isopropylamidino-substituted derivative
with a thiophene core, 10b. However, all of these compounds
showed a nonselective cytotoxic activity. Interestingly, the
isopropylamidino-substituted derivative 2a was the least cyto-
toxic for normal human fibroblasts among the tested com-
pound panel, while derivatives with a furan core, 10a and 11a,
showed the weakest antiproliferative effect specifically on the
growth of SW620 and HEp-2 cell lines. The isopropylamidino-
substituted derivative 2b with a thiophene core exerted a strong
differential antiproliferative effect both on the MCF-7 and
MiaPaCa-2 cells, while showing a lower cytotoxicity on nor-
mal human fibroblasts in comparison to other highly active
compounds (Figure 2). All together, it has been concluded that
imidazolinyl-substituted derivatives and/or the thiophene core
are in correlation with increased antiproliferative activity of
bisamidino-substituted benzothiazolyl derivatives. A similar
observation was described by Racane et al., where the activity
of imidazolinyl derivatives was better than that monitored for
the isopropylamidino derivatives.26 Further on, all tested
compounds might act as DNA groove binders due to their
angular structure (confirmed by further experiments and pre-
sented in the section DNA Binding Properties), making the cell
DNA a dominant target for their antiproliferative activity.
Additional mechanistic studies, including the cell cycle analysis
and apoptosis induction assay, were thus performed in order to
reveal a more detailed picture on the possible antiproliferative
mechanisms and/or targets. Two derivatives were chosen for
that purpose: (i) 2b due to the observed differential effect on
the growth of MCF-7 and MiaPaCa-2 cell lines and a lower
cytotoxicity exerted on normal human fibroblasts in compar-
ison to other tested compounds, and (ii) 3b, as one imidazo-
linyl-substituted derivative with a thiophene core that showed
less cytotoxic effect on MCF-7, MiaPaCa-2, and normal hu-
man fibroblasts.
Effects of 2b and 3b on the Cell Cycle of MiaPaCa-2 and
MCF-7 Cell Lines. The results of the flow cytometric analysis
revealed strong perturbations in the cell cycle of MiaPaCa-2
and MCF-7 induced upon treatments with different concen-
trations of 2b and 3b (Table 2). The observed perturbations
varied for each substance and in dependence to the concen-
tration used, substantiating the impact of structural differ-
ences and concentration on the activity.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2421
Figure 2. Concentration-response profiles obtained on HeLa, MCF-7, MiaPaCa-2, and JK cells treated with the tested compounds at
different concentrations. The percentages of growth (PG) were calculated.
In MCF-7 cells (Table 2), the lower concentration of
compound 2b (10 μM) induced an increase in the S-phase
population and a substantial decrease in the G2/M cell
population accompanied by a moderate rise in the sub-G1
cell population. Oppositely, the higher tested concentration
of compound 2b (50 μM) induced a drastic accumulation of
cells in the G2/M phase accompanied by a concomitant
decrease of cells in the G1 phase and an accentuated increase
of cell number in the sub-G1 phase. A same pattern was
observed upon treatment of MCF-7 cells with compound 2b
at both tested concentrations after 48 h, which was accom-
panied by a drastic decrease of cell number in the S phase for
treatment with higher 2b concentration. It seems thus, that
the main antiproliferative mechanism of 2b depends on the
concentration used and involves a cell cycle arrest in the S
phase induced upon treatment with lower concentration and
the G2/M delay upon treatment with higher concentration.
In MiaPaCa-2 cells (Table 2), compound 3b (at concen-
tration 1 μM) induced a marked increase in the G2/M phase
after the 24 h treatment accompanied by a concomitant
decrease in the G1 and S cell population. Similarly, an
increase in the G2/M phase was induced after the 48 h
treatment with the same concentration, even if an increase
in the S phase was observed, as well. A marked increase in the
sub-G1 was observed for both concentrations upon 24 and
48 h treatments. Additionally, MiaPaCa-2 cells treated with
a higher concentration of 3b (5 μM) after 48 h accumulated in
the G1 phase, which was accompanied with a decrease in the
G2/M population. The antiproliferative mechanism of 3b on
MiaPaCa-2 cells thus depends on the concentration used and
involves a cell cycle arrest in the G2/M phase induced upon
treatment with lower concentration, the G1 delay upon
treatment with higher concentration, and a substantial rise
of cells in the sub-G1 phase in all treatments.
In MCF-7 cells (Table 2), compound 3b upon the 24 h
treatment provoked a marked decrease in G1 population,
which was more accentuated in the treatment with higher 3b
concentration. This G1 decrease was accompanied by a rise
in sub-G1, S, and G2/M cell populations. Similarly, a
decrease in the G1 population was observed for both tested
concentrations upon the 48 h treatment, which was followed
by a moderate rise of cells in the sub-G1 phase and a
substantial rise of cells in the S phase. The antiproliferative
mechanism of 3b on MCF-7 cells is thus somewhat different
in comparison to the activity observed on MiaPaCa-2 cells. It
involves a moderate rise of cells in the sub-G1 phase in all
treatments, a rise in cell number in the S-phase in all
treatments, and a G2/M arrest upon the 48 h treatment with
the higher concentration. Similar results were obtained with
the class of cyano- and amidinobenzothiazole derivatives
described in our previous paper, where both G1 and G2/M
delays accompanied by a reduction of cell number in the
S-phase were described.31 However, the increase of cells in
the S-phase observed in our experiments might be explained
by previously described mechanisms of action for this class
of compounds that include the hampering of DNA replica-
tion processes in the cytoplasm and nucleus (e.g., through
inhibition of topoisomerase I and II and/or by direct binding
to DNA).20,31,32 Briefly, the damaged cells block the cell
cycle prior to mitosis, and if the damage is too high, they
ultimately die, which might be the case in our experiments, as
well. Indeed, we observed that both compounds induced an

2422 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Racaneꢀ et al.
Table 2. Flow Cytometric Analysis of the MCF-7 and MiaPaCa-2 Cell Cycle upon the 24 and 48 h Treatments with Compound 2b at Concentrations 10
and 50 μM and Compound 3b at Concentrations of 1 and 5 μMa
compound 2b
cell line/treatment
MCF-7
cell percentage (% ( standard deviation)c
sub-G1
G1
S
G2/M
control 24 h
10 μM 24 h
50 μM 24 h
control 48 h
10 μM 48 h
50 μM 48 h
2.4 ( 0.5
6.7 ( 0.1b
18.5 ( 5.3b
2.7 ( 0.1
46.3 ( 2
31.8 ( 2.2
41.5 ( 1.9b
44.5 ( 6.4
41.4 ( 2.3
83.5 ( 2.1b
12.3 ( 0.5b
27.6 ( 0.9
9.5 ( 1b
47.4 ( 1.4
1.1 ( 0.7b
47 ( 2.4
63 ( 5.2b
15.7 ( 0.8
6.6 ( 2.4b
32.3 ( 1.3b
2.1 ( 0.1b
8.6 ( 0.1b
9.9 ( 4.6b
54.5 ( 2.1
compound 3b
MiaPaCa-2
control 24 h
1 μM 24 h
5 μM 24 h
control 48 h
1 μM 48 h
5 μM 48 h
15.2 ( 1.8
33.8 ( 1.4b
21.8 ( 0.12b
14.0 ( 0.5
55.1 ( 0.2b
40.7 ( 0.5b
31.7 ( 1.6
18.3 ( 0.6b
37.9 ( 8.4
44.6 ( 1.1
11.4 ( 0.1b
55.1 ( 2.1b
50.1 ( 0.6
36.8 ( 2.5b
51.6 ( 2.3
29.6 ( 1.5
44.7 ( 0.4b
35 ( 1.4
18.3 ( 1.1
44.9 ( 2.1b
10.5 ( 6.2
27.4 ( 1.8
43.9 ( 0.5b
10.4 ( 1.3b
MCF-7
control 24 h
1 μM 24 h
5 μM 24 h
control 48 h
1 μM 48 h
5 μM 48 h
2.4 ( 0.5
10.1 ( 5.1
8.6 ( 0.9b
2.6 ( 0.1
7.7 ( 0.1b
6.3 ( 0.1b
44.8 ( 1.3
26.6 ( 1.5b
8 ( 3.3b
46.6 ( 1.8
32.6 ( 1.2b
9.9 ( 2.3b
31.1 ( 2.8
40.1 ( 1.6
45.8 ( 0.9b
38.8 ( 1.3
45.8 ( 1.3b
69.7 ( 1.8b
23.9 ( 4.2
34.2 ( 1.1
46.2 ( 4.2b
15.3 ( 1.3
13.9 ( 1.3b
20.5 ( 0.6b
a Cells were stained with propidium iodide and analyzed with Becton Dickinson FACScalibur flow cytometer (10 000 counts were measured for each
condition). The results are presented as cell percentages (%). b Statistically significant: statistical analysis was performed in Microsoft Excel by using the
ANOVA at p < 0.05. c The percentage of the cells in each cell cycle phase was based on the obtained DNA histograms and determined using the WinMDI
2.9 and Cylchred software.
Table 3. Results of the Annexin V Assay Performed on MCF-7 and
MiaPaCa-2 Cells upon 24 and 48 h Treatments with Compounds 2b
and 3ba
increase of sub-G1 MCF-7 and MiaPaCa-2 cell populations,
which is indicative of apoptosis. The most drastic rise in sub-
G1 was, however, seen in MiaPaCa-2 cells treated with 3b.
The annexin V assay was hence performed to confirm
apoptosis, and morphological changes of cells were moni-
tored by microscopy.
compound 2b
compound 3b
MCF-7
(%)
MCF-7
(%)
MiaPaCa-2
(%)
Annexin V Assay. Due to high antiproliferative activity
observed for tested compounds, the results of annexin V
assay were expected and showed a marked increase in the
number of apoptotic cells upon treatments with all tested
concentrations of 2b and 3b (Table 3). However, higher
concentrations were more effective in apoptosis induction
in both cell lines upon the 48 h treatment. Therefore, the
number of apoptotic MCF-7 and even more for MiaPaCa-2
cells, rose substantially upon the 24 h treatment with com-
pound 3b. A strong effect on apoptosis induction was
observed for compound 2b on the MCF-7 cells, as well,
where the number of apoptotic cells reached 79% upon the
24 h treatment and 86% upon the 48 h treatment. Higher
concentrations of compound 3b induced a similar rise of
apoptotic cells after 48 h in MCF-7 and MiaPaCa-2 cells,
where it reached 74 and 81%, respectively.
The observed percentages are higher than percentages of
cells in the sub-G1 phase documented in the cell cycle
analysis probably because annexin V assay does not discri-
minate between diverse types of cell death mechanisms and it
rather generally indicates dying cells. The cell death suppres-
sion is extremely relevant for tumor development and pro-
gression. The mechanisms of cell death and cell survival are
complex and involve much more than apoptosis and/or
necrosis with cross-talk and other programmed processes
such as autophagy, postmitotic death, or entosis.33,34 Indeed,
control 24 h
10 μM
50 μM
control 48 h
10 μM
50 μM
15
76
94
13
52
99
control 24 h
1 μM
5 μM
control 48 h
1 μM
5 μM
9
40
54
8
18
57
68
15
42
96
26
82
a The results are presented as percentages of cells positive for annexin
and/or propidium iodide (apoptotic cells) per total counted cell number.
99% of MCF-7 cells treated with higher concentration of 2b
were stained both green and red (Figure 3), which might be
indicative for necrosis and apoptosis, while the cell cycle
analysis for this treatment showed a rather moderate in-
crease in the sub-G1 cell population. Moreover, MCF-7 cells
treated with compound 2b increased several times in volume
(Figure 4), which is linked to the S and G2/M arrests revealed
by the cell cycle analysis. Thus, MCF-7 cells probably died
through a distinct cell death mechanism that involves the
block of cells from S-G2 transition upon treatment with
lower 2b concentration and impaired cell division through
the G2/M phase arrest upon treatment with higher concen-
tration of 2b. It has already been shown that antitumor
compounds might cause such a block of cells from the S-G2
transition by decreasing the rate of DNA synthesis through
diverse mechanisms, including inhibition of topoiso-
merase I35 or DNA polymerase.36 However, in this study,

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2423
Figure 4. MCF-7 (A,B) and MiaPaCa-2 (C) cells treated with com-
pounds 2b (A) and 3b (B,C) after 24 and 48 h treatments. MCF-7 cells
treated with compound 2b increased several times in volume, while a
change of cell morphology that directly points to apoptosis (shrinkage
of cells, chromatin condensation) was observed for MCF-7 and
MiaPaCa-2 cells treated with compound 3b (control: untreated cells).
Figure 3. Representative fluorescence microscopy pictures of
apoptotic MCF-7 (A,B) and MiaPaCa-2 (C) cells treated with
compounds 2b (A) and 3b (B,C) after the 24 and 48 h treatments.
The apoptotic cells are labeled with green fluorescein and/or red
propidium iodide (control: untreated cells).
Table 4. Variation of DNA Melting Temperature (ΔTm) of CT-DNA
or Poly(dAdT)2 Induced by the Various Compounds (Cpd)
ΔTma (ꢀC)
Cpd
CT-DNA
(dAdT)2
we have shown that neither one tested compound, including
compounds 2b and 3b, exerted intercalative properties and/
or topoisomerase I induced relaxation. Nevertheless, it has
been shown that both compounds act as groove binders and
thereby directly inhibit DNA replication and induce apop-
tosis (described later on in the text). Moreover, we showed
that all compounds easily penetrate through the cell mem-
brane and accumulate in the nucleus where they influence the
DNA synthesis observed as sustained S or G2 arrest in the
treatment of cells with 2b and 3b. The compound 3b, how-
ever, accumulates as well in the perinucleus, probably caus-
ing additional perinuclear DNA destabilization (e.g.,
through interaction with other partners). This phenomenon
might explain the observed difference in the mechanism of
action that involves a stronger apoptotic response of cells
treated with 3b and accumulation in the sub-G1 phase.
Indeed, a change of cell morphology that points to apoptosis
(shrinkage of cells, chromatin condensation) was observed
for both MCF-7 and MiaPaCa-2 cells treated with com-
pound 3b (Figures 3 and 4).
2a
2b
12.7
14.2
18.0
13.9
9.5
27.1
26.5
32.4
24.3
23.2
5.1
3a
3b
10a
10b
11a
11b
7.6
17.3
2.3
25.3
1.4
a Variation of the Tm values for 20 μM of CT-DNA or poly(dAdT)2
incubated with 2 μM of various compounds (drug/base pair ratio = 0.1)
versus incubation alone. The reaction is conducted at room temperature
in 1 mL of BPE buffer (6 mM Na2HPO4, 2 mM NaH2PO4, 1 mM EDTA,
pH 7.1).
DNA Binding Properties. The DNA binding propensity
for the new series of benzothiazole derivatives was first
assessed using UV spectrometry. DNA melting tempera-
ture (Table 4) and modification of the absorption spectra
of each compound in the presence or absence of CT-DNA
(Supplementary Figure 1A) was first studied. A Benesi-
Hildebrand derived plot of [CT-DNA]/(εFree - εObs) over

2424 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Racaneꢀ et al.
Figure 5. Fluorescence quenching upon DNA binding. Panel A: Fluorescence measurements are expressed as arbitrary units (a.u.) over
wavelength (nm). A fixed concentration of compounds (1 μM for 2b, 3b, and 10a or 5 μM for the others) was incubated alone (upper lanes) or
with increasing concentrations of CT-DNA from 1 to 30 μM (up to down arrows). The corresponding Stern-Volmer plots are presented as
embedded panels expressing the decrease of fluorescence of the unbound compound (F0/F) of CT-DNA concentration expressed in μM. Panel
B: Stern-Volmer constants (KSV). KSV values for the various compounds are calculated from the slope and intercept values deduced from
Stern-Volmer plots presented in panel A and are expressed in M.
[CT-DNA] was constructed where εFree and εObs are the
extinction coefficients for free compound and that obtained
from UV-vis spectrometry measurement at the various CT-
DNA concentrations ([CT-DNA]) (Supplementary Figure
1B). The ratio of intercept to slope gave the equilibrium
constants for the formation of various drug/DNA com-
plexes, which are presented in the bottom part of Supple-
mentary Figure 1B). The binding constant evidenced that
compounds 2a, 2b, 3a, 3b, and 11a are the strongest
DNA binders. The variation of melting temperature studies
using CT-DNA identified compounds 2a, 2b, 3a, 3b, 10a, and
11a as strong DNA binders with ΔTm of more than or

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2425
Figure 6. Circular dichroism measurements performed in the presence of CT-DNA. Graded concentrations of various benzothiazole
derivatives from 1 to 25 μM (2a to 3b) or from 1 to 50 μM (10a to 11b) were added to samples containing CT-DNA alone (dashed bold
lanes). Each lane corresponds to an average of three measurements.
around 10 ꢀC at drug/CT-DNA ratios (r) of 0.1. The same
studies performed using double-strand poly(dAdT)2 (r = 0.1)
evidenced the same series as very strong binders of AT-rich
sequences.
pounds. Indeed, however, each compound failed to present
any intrinsic CD (data not shown), a large positive induced
CD (ICD) appears at around 400 nm, within the absorption
band of the compound, as evidenced from Supplementary
Figure 1. This positive ICD is very high, particularly for
compounds 2a, 2b, 3a, 3b, 10a, and 11a, and may suggest a
binding of the compound in one groove of the helix. This
positive ICD is associated with a negative ICD at band
around 300 nm for some compounds (2a, 3a, 10a, and
11a), suggesting some differences in the DNA interaction
between the furan (a) and the thiophene (b) series. Com-
pounds 10b and 11b induce particular bisignate excitonic
ICDs characterized by equal negative and positive values on
either side of their respective absorption maxima. This only
appears at the highest drug/DNA ratio and could possibly be
due to exciton coupling interactions between closely bound
chromophores as part of dye-dye interaction. We then
focused on the ICD by increasing compounds 2a, 2b, 3a, 3b, 10a,
and 11a on either poly(dAdT)2 (Supplementary Figure 2A) or
All compounds present intrinsic fluorescence properties
that could be used to evaluate their respective DNA binding
propensities. Their DNA binding activity is associated with a
decrease of the fluorescence intensities of benzothiazole
derivatives with various intensities, as evidenced in Figure 5A.
The quenching constants KSV (Figure 5B) were deduced
from Stern-Volmer plots (Figure 5A, embedded panels).
The binding mode (intercalation, groove binding) was
then evaluated using circular dichroism (CD) measurements
(Figure 6) and topoisomerase I induced relaxation assays
(data not shown). The CD spectrum of CT-DNA (200 μM,
dashed bold lane) presents the negative peak at ∼245 nm and
the positive peak at ∼275 nm typical for a right-handed helix.
This classical CT-DNA-induced CD spectrum dramatically
changed upon addition of increasing concentration of com-

2426 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Racaneꢀ et al.
Figure 7. DNase I footprinting assays. Panel A: Radio-labeled 265 bp DNA fragment was incubated alone (lanes “DNA”) or graded
concentrations of the various indicated benzothiazole derivatives from 0.1 to 1 μM (as indicated on the top of each lanes). The G-track lanes
labeled “G” were used as markers for guanines to locate the footprint areas and to establish the scale indicated from 40 to 130 bp. Panel B:
Densitometric analysis. Differential cleavage plots derived from the gels are presented in (A). Gray boxes localize the sites protected on the gels.
poly(dGdC)2 (Supplementary Figure 2B). Using poly-
(dAdT)2, the addition of increasing concentrations of each
tested compound evidenced very strong positive ICD in the
band of compound absorbance, as well as a strong drop onthe
positive CD from the DNA (270-280 nm). Such strong
positive ICD in the absorption band of the compounds argues
for a binding in one groove of the DNA helix. Using the
highest concentrations of compounds (more than 20 μM and
particularly with compounds 2b and 3b), a negative peak
at ∼365 nm appears that suggests a possible second mode of
binding to the DNA (Figure 6). The possibility of a second
binding mode was also suggested by the UV/visible absorp-
tion spectrometry experiments (Supplementary Figure 1) that
showed an evolution of the CD spectra for drug concentra-
tions >20 μM with spectra losing the first isosbestic point
obtained at lower drug concentrations (409 and 412 nm for 2b
and 3b, respectively) for a second isosbestic point observed at
higher wavelength using the highest drug/DNA ratios (418
and 432 nm for 2b and 3b, respectively). By contrast, oppo-
sitely signed CD bands in the visible spectral region using
poly(dGdC)2 may correspond to strong excitonic effects
appearing only at the highest drug/DNA ratios using com-
pounds 2a, 2b, and 3b (more than 20 μM of CT-DNA).
Altogether, the CD spectra give arguments for groove binding
of compounds 2a, 2b, 3a, 3b, 10a, and 11a from CT-DNA
experiments, more likely on AT-rich sequences.
Topoisomerase I induced relaxation assays were also
performed as described in Peixoto et al.37 to determine

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2427
possible DNA intercalation properties of the tested com-
pounds. None of the compounds presented any typical DNA
intercalation profile from this experiment (data not shown),
comfirming the absence of intercalation between adjacent
base pairs of CT-DNA shown CD experiments (Figure 6).
DNA Sequence Selectivity. Sequence-specific DNA binding
properties were precisely assessed using DNase I footprinting
assays (Figure 7 and Supplementary Figures 3-5). Revela-
tion of the gels showed some protected regions (footprints) in
the presence of increasing concentrations of compounds 2a,
2b, 3a, 10a, and 11a and to a much lesser extent compound 3b
but no footprints in the presence of compounds 10b and 11b
(Figure 7A). Densitometric analyses of the gels evidenced
that the protected portions of the gel span AT-rich sequences
(Figure 7B). Particularly, 50-78ATTA75, 50-95ATTA92, and
50-104ATTT101 sites are well-recognized by the furan com-
pounds 2a and 3a. Compound 11a presents a more particular
profile with recognition of sites 50-102ATTA98 and 50-59AACA56
additionally to the other 50-78ATTA75 and 50-95ATTA92
sites. The thiophene compounds present lesser (2b and 3b)
to no more (10b, 11b) binding selectivity than their furan
counterparts. Similarly, DNase I footprinting experiments
performed using the 257 bp (Supplementary Figures 3 and 4)
and 117 bp (Supplementary Figure 5) DNA fragments
confirm the preference for AT-rich site recognition with
some sequence selectivity depending on the A or T succes-
sions (Supplementary Figure 5) but also evidenced binding
for compounds 2b and 3a on the 50- AACT site, alighting the
complexity of sequence selectivity of those compounds
(Supplementary Figures 3 and 4).
Binding to those DNA sequences was analyzed using
melting temperature studies by employing hairpin (HP)
oligonucleotides containing each of the sequences as single
binding sites. Table 5 shows compounds 2a, 2b, 3a, and
11a as strong binders for AT-rich sequences: 50-AATT,
50-TTAA, and 50ATTA sites, with a preference for 50-AATT.
However, some binding to GC-rich sites could also occur as
evidenced using the HP-GCGC sequence using compounds
2a to 10a. Mixed 50-ACCA sequences were also tested but
showed weaker DNA stabilization potencies using the tested
compounds.
Table 5. Variation of DNA Melting Temperature (ΔTm) of Hairpin
(HP) Oligonucleotides Induced by Various Compounds (Cpd)a
ΔTmb (ꢀC)
Cpd HP-ATTA HP-AATT HP-TTAA HP-ACCA HP-GCGC
2a
2b
4.1
5.7
6.8
0.8
1.1
3.3
4.4
1.1
11.4
10.1
15.5
3.5
7.7
6.7
8.6
2.1
1.8
0.6
4.3
0
2.8
2.6
3.6
1.5
1.1
0
9.8
4.1
6.6
3.5
4.2
0.6
1.6
0
3a
3b
10a
10b
11a
11b
2.6
1.8
9.3
0.4
2.1
0
a Different HP oligonucleotides containing in the stem the binding
sites as specified. b Variation of the Tm values for 2 μM of 22-mer hairpin
oligonucleotides forming a 9 bp stem that varies for the indicated 4-mer
sequence incubated with 2 μM of the various compounds (drug/oligo-
nucleotide ratio = 1) versus incubation alone. The reaction is conducted
at room temperature in 1 mL of BPE buffer (6 mM Na2HPO4, 2 mM
NaH2PO4, 1 mM EDTA, pH 7.1).
Cellular Penetration and Distribution. In order to deter-
mine whether different compounds might efficiently pass
Figure 8. Cell penetration and subcellular localization using fluorescence microscopy. HT-29 human colon carcinoma cells were treated for
16 h with 5 μM of various benzothiazole derivatives specified on each panel side prior to further treatment with Mito Fluor Red 588 used to
localize the cytoplasm compartment (middle panels of each series). Single fluorescence analysis of the compounds was established in blue for all
compounds with exceptions for compound 11b, which fluoresces only in green, and compound 3b, which fluoresces both in blue and in green
(left panels of each series). Superposition analyses are presented on the right panels of each series. Enlarged pictures of significant cells from
each 3b panel (as specified using white boxes) are shown in embedded panels at the bottom. The size marker is given in the first panel.

2428 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Racaneꢀ et al.
through the cellular and nuclear membranes, we used their
fluorescence properties to explore their potential subcellular
localization. HT-29 cells were treated with 5 μM of each drug
as a fixed concentration of compounds that is sufficient to
visualize the compounds using fluorescence microscopy.
Mito Fluor Red 588 was used concomitantly as a mitochon-
drion marker to visualize the cytoplasm. Figure 8 shows the
good cellular penetration of this new series and shows the
nuclear localization of each tested compound as single
fluorescence analyses (left panels of each column) or as
superposition with the marker for cytoplasm (right panels
of each column). Of particular interest, compound 3b fluor-
esces both in green and in blue and reveals a particular
localization: when analyzed in green, compound 3b locates
in the nucleus, as determined for other compounds, but when
analyzed in blue, it locates at the perinuclear level, as
evidenced in the surrounding panels. This particular locali-
zation difference might reflect interactions of 3b with two
different types of partners.
in correlation with the second isosbestic point evidenced at
those concentrations using UV/visible spectrometry
(Supplementary Figure 1). Both 10b and 11b weak binders
bore a thiophene central core instead of the furan within the
respective strong binders 10a and 11a. The change of the oxygen
atom for a sulfur induces an enlargement of the angle of the
two amidino or imidazolinyl benzothiazole “arms” of the
molecules, which could impair a good fitting of the thiophene-
containing molecules on the DNA helix and thus reduce
their DNA binding activities, as evidenced using Tm measure-
ments (Table 5) and DNase I footprinting activities (Figure 7
and Supplementary Figures 3-5). Similarly, the same change of
the furan core to a thiophene moiety in the imidazolinyl phenyl
benzothiazole series (compare 3a to 3b) induces changes in
sequence specificity: the furan compound 3a strongly interacted
with AT-sites such as the 50-ATTA sequence, whereas the
thiophene compound 3b failed and showed preference for GC
sites as evidenced from Tm measurements on hairpin oligonu-
cleotides. A similar modification does not affect DNA binding
affinity and selectivity in the isopropyl amidine series (2a versus
2b). This might be due to higher degree of flexibility of those
extremities from comparison with amidinyl and imidazolinyl
ends to fit with the DNA groove, more likely the minor one.
Identification of the binding mode of the different compounds
argues for an orientation in the helix axis with the compounds
setting in the groove of the DNA helix, as evidenced from
CD spectra (Figure 6 and Supplementary Figure 2). Such
groove binding occurs on AT-rich sequences, as evidenced
from footprinting experiments (Figure 7 and Supplementary
Figures 3-5) and CD spectra using poly(dAdT)2 from
comparison with poly(dGdC)2 (Supplementary Figure 2).
At the highest drug/DNA ratio, interactions of 2b and 3b with
poly(dAdT)2 result in a positive excitonic effect, suggesting
stacking of the molecule in a right-handed orientation parallel
to that of DNA helix. Not all AT-containing sequences are
equally recognized by those compounds, as evidenced using
DNase I footprinting experiments (Figure 7 and Supplementary
Figures 3-5) and melting temperature using hairpin oligonu-
cleotides containing specifically designed sequences (Table 5).
However, in the presence of pure GC oligonucleotides
(poly(dGdC)2), the interaction seems different with negative
excitonic cotton effects using compound 2a whereas (i) com-
pounds 2b and 3b present positive exciton chirality, suggesting a
right-handed orientation of the molecule similar to that of the
DNA helix; (ii) compounds 10a and 11a present only negative
ICD, suggesting chromophore stacking or intercalation process;
and (iii) compound 3a shows of a mixture of right-handed (in
agreement with a fitting in the groove) and chromophore
stacking or intercalative orientations.
Conclusion
In this paper, the synthesis of new amidino- and imidazo-
linyl-substituted 2,5-dibenzothiazolyl furans and thiophenes
(10a, 10b, 11a, and 11b) from 2,5-dicarbonylchlorides of furan
or thiophene dicarboxylic acids with zwitterionic amidino- or
imidazolinyl-substituted aminothiophenoles 8 and 9 has been
described. A strong antiproliferative effect of the tested
compounds (2a, 2b, 3a, 3b, 10a, 10b, 11a, 11b) was observed
on all tested cell lines in a concentration-dependent response
pattern. Especially, this was observed on HeLa, MCF-7, and
MiaPaCa-2 cell lines where all compounds strongly inhibited
the cell growth at submicromolar concentrations and higher.
In particular, the strongest antiproliferative effect, in addition
to a nonselective cytotoxic activity, was observed for three
imidazolinyl-substituted derivatives 3a, 3b, and 11b and for
the amidino-substituted thiophene derivative 10b. The amidino-
substituted thiophene derivative 2b exerted a strong differential
antiproliferative effect both on the MCF-7 and MiaPaCa-2
cells, while showing a lower cytotoxicity on normal human
fibroblasts in comparison to other highly active compounds.
From DNA binding studies, compounds 2a, 2b, 3a, 3b, 10a, and
11a were identified as strong DNA binders whereas compounds
10b and 11b were weaker DNA binding agents (Table 4 and
Figures 6 and 7). Using compounds 2a to 3b, an isosbestic point
is seen for drug concentrations from 0.1 μM up to 4-20 μM
depending on the compound, suggesting that a single mode of
binding occurs at those concentrations (Supplementary
Figure 1A). Using higher DNA concentrations (more than 30
or 40 μM), a second isosbestic point occurs at higher wave-
length, suggesting a second mode of binding at those drug/DNA
ratios. This is in agreement with observations from DNase I
footprinting experiments showing clear sequence selectivity on
AT-rich sequences (Figure 7 and Supplementary Figures 3-5)
at the lowest concentrations but nonsequence specific binding at
highest drug concentration (data not shown). Using DNA
concentrations below 20 μM, the CD spectra show a large
positive ICD in the presence of both CT-DNA and poly-
(dAdT)2 but no ICD using poly(dGdC)2, suggesting a binding
within the DNA groove of AT-rich sequences but not that
of GC-rich sites. At concentrations greater than 20 μM of
the active compounds 2b and 3b, an additional negative ICD
is observed using both CT-DNA and poly(dAdT)2, whereas
a bisignate excitonic ICD is generated using poly(dGdC)2,
arguing for a second binding mode (sequence-independent)
Compounds 2b and 3b were chosen for biological studies due
to their differential antiproliferative properties. Both com-
pounds accumulate in the nucleus and act as groove binders,
thereby directly inhibiting DNA replication and inducing the S
and G2 arrests or the G1 arrest if employing 3b at higher
concentration. Apoptosis induction was observed for both
compounds. The biological results presented herein are in line
with our previous studies showing that this class of compounds
has strong antitumor properties, which might be enhanced in
dependence on the position of the substituent on the 2-phenyl-
benzothiazole skeleton, as well as on the type of amidino
substituent attached. Newly designed and efficiently synthe-
sized series of diamidino-, diisopropylamidino-, and diimida-
zolinyl-substituted derivatives of phenyl benzothiazolyl and
dibenzothiazolyl furans and thiophenes described in this study

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2429
have shown to be a highly promising class of antitumor
compounds. Nevertheless, it was generally observed that the
imidazolinyl-substituted derivatives and/or the thiophene core
were in correlation with increased antiproliferative activity.
2,20-(Thiophene-2,5-diyl)bis(1,3-benzothiazole-6-carboximid-
amide) dihydrochloride (10b): Compound 10b was prepared
using general Methods A and B described above. Crystallization
from a water/acetone mixture afforded 0.343 g (61.2%, Method
A) and 0.495 g (88.2%, Method B) of yellow crystals: mp >300
ꢀC; 1H NMR (300 MHz, DMSO-d6) (δ ppm) 9.55 (s, 4H,
disappeared on addition of D2O, 4Hamidine), 9.34 (s, 4H, dis-
appeared on addition of D2O, 4Hamidine), 8.73 (d, 2H, J = 1.6
Hz, 2Hbenzothiazole), 8.25 (d, 2H, J = 8.6 Hz, 2Hbenzothiazole), 8.12
(s, 2H, 2Hthiophene), 7.98 (dd, 2H, J = 1.8 Hz, J = 8.6 Hz,
2Hbenzothiazole).
2,20-(Furan-2,5-diyl)bis[6-(4,5-dihydro-1H-imidazol-2-yl)-1,3-
benzothiazole] dihydrochloride (11a): Compound 11a was pre-
pared using general Methods A and B described above. Crystal-
lization from a water/acetone mixture afforded 0.181 g (30.3%,
Method A) and 0.451 g (75.6%, Method B) of yellow crystals:
mp >300 ꢀC; 1H NMR (300 MHz, DMSO-d6) (δ ppm) 10.70 (br
s, 4H, disappeared on addition of D2O, 4Hamidine), 8.79 (s, 2H,
2Hbenzothiazole), 8.33 (d, 2H, J = 8.6 Hz, 2Hbenzothiazole), 7.99 (d,
2H, J = 8.6 Hz, 2Hbenzothiazole), 7.80 (s, 2H, 2Hfuran), 4.02 (s, 8H,
CH2).
2,20-(Thiophene-2,5-diyl)bis[6-(4,5-dihydro-1H-imidazol-2-yl)-
1,3-benzothiazole] dihydrochloride (11b): Compound 11b was
prepared using general Methods A and B described above.
Crystallization from a water/acetone mixture afforded 0.218 g
(35.6%, Method A) and 0.484 g (78.8%, Method B) of yellow
crystals: mp >300 ꢀC; 1H NMR (300 MHz, DMSO-d6) (δ ppm)
10.75 (br s, 4H, disappeared on addition of D2O, 4Hamidine), 8.86
(s, 2H, 2Hbenzothiazole), 8.29 (d, 2H, J = 8.6 Hz, 2Hbenzothiazole),
8.16 (s, 2H, 2Hthiophene), 8.11 (d, 2H, J = 8.6 Hz, 2Hbenzothiazole),
4.05 (s, 8H, CH2).
Antitumor Activity Assays. The cell lines HeLa (cervical
carcinoma), SW620 (colorectal adenocarcinoma, metastatic),
MiaPaCa-2 (pancreatic carcinoma), MCF-7 (breast epithelial
adenocarcinoma, metastatic), HEp-2 (epidermoid carcinoma,
larynx), HepG2 (hepatocellular carcinoma), and JK (normal
diploid human fibroblasts) were cultured as monolayers and
maintained in Dulbecco’s modified Eagle medium (DMEM)
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.
For the growth inhibition activity assay, the panel cell lines
were inoculated onto a series of standard 96-well microtiter
plates on day 0, at 3000-5000 cells per well according to the
doubling times of specific cell line. Test agents were then added
in five 10-fold dilutions (1 ꢀ 10-8 to 1 ꢀ 10-4 M) and incubated
for further 72 h. Working dilutions were freshly prepared on the
day of testing in the growth medium. The solvent (DMSO) was
also tested for eventual inhibitory activity by adjusting its con-
centration to be the same as in the working concentrations
(DMSO concentration never exceeded 0.1%). After 72 h of
incubation, the cell growth rate was evaluated by performing
the MTT assay: experimentally determined absorbance values
were transformed into a cell percentage growth (PG) using the
formulas proposed by National Cancer Institute and described
previously in Gazivoda et al.38 This method directly relies on
control cells behaving normally at the day of assay because it
compares the growth of treated cells with the growth of untreated
cells in control wells on the same plate;the results are therefore a
percentile difference from the calculated expected value.
The IC50 and LC50 values for each compound were calculated
from dose-response curves using linear regression analysis by
fitting the mean test concentrations that give PG values above
and below the reference value. If, however, all of the tested
concentrations produce PGs exceeding the respective reference
level of effect (e.g., PG value of 50) for a given cell line, the
highest tested concentration is assigned as the default value (in
the screening data report that default value is preceded by a “>”
sign). Each test point was performed in quadruplicate in three
individual experiments. The results were statistically analyzed
Experimental Section
Chemistry. Melting points were determinate on a Koffler hot
stage microscope. IR spectra were recorded on a Nicolet magna
760, with KBr disks or Bruker Vertex 70 FTIR spectrophoto-
meter with an ATR sampling accessory. 1H and 13C NMR
spectra were recorded on Bruker Avance DPX 300 or Bruker
AV-600 spectrometers using TMS as an internal standard and
the deuterated solvents indicated were used. Mass spectra were
recorded with an Agilent 1100 Series LC/MSD Trap SL spectro-
meter using electrospray ionization (ESI). Elemental analyses
for carbon, hydrogen, and nitrogen were performed at the
microanalytical laboratories of the Rudjer Boskovic Institute.
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. Synthesis of unsym-
metrical bisnitriles 1a and 1b and bisamidines 2a, 2b, 3a, and
3b was achieved as described in Scheme 1 according to the
literature.28,29 Dichlorocarbonyl compounds 5a and 5b were
prepared from dicarboxylic acid 4a and 4b and thionyl chloride
prior to use. Compound 7 was prepared from 4-amino-3-
sulfanylbenzonitrile 629 and furan-2,5-dicarbonyl dichloride
5a and reported in Supporting Information. The 5-amidinium-
2-aminobenzothiolate 8 and 5-(imidazolinium-2-yl)-2-amino-
benzothiolate hydrate 9 described as reagent in Scheme 2 were
prepared according to the literature.30
General Methods for Preparation of Compounds 10a, 10b, 11a,
and 11b. Method A: A mixture of 5-amidinium-2-aminobenzo-
thiolate 8 (0.335 g, 2.0 mmol) or 5-(imidazolinium-2-yl)-2-
aminobenzothiolate hydrate 9 (0.423 g, 2.0 mmol) and furan-
2,5-dicarboxylic acid 4a (0.156 g, 1.0 mmol) or thiophene-2,5-
dicarboxylic acid 4b (0.172 g, 1 mmol) in polyphosphoric acid
(8-10 g) was heated with stirring for 2 h at 180 ꢀC. The reaction
mixture was cooled and poured in water. The resulting precipi-
tate was filtered off, washed with NaHCO3 solution (10%), and
dried. The products were crystallized from 10:1 mixture of
glacial acetic acid/concentrated hydrochloric acid mixture
(charcoal), filtered off, washed with acetone, and dried in
vacuum. The pure products were obtained by crystallization
from appropriate solvents.
Method B: A mixture of 5-amidinium-2-aminobenzothiolate
8 (0.335 g, 2.0 mmol) or 5-(imidazolinyl-2-yl)-2-aminobenzo-
thiolate hydrate 9 (0.423 g, 2.0 mmol) and furan-2,5-dicarbonyl
dichloride 5a (0.192 g, 1.0 mmol) or thiophene-2,5-dicarbonyl
dichloride 5b (0.208 g, 1.0 mmol) in glacial acetic acid (25 mL)
was refluxed for 4 h under nitrogen. The reaction mixture was
cooled and the crystals were filtered off, washed with diethy-
lether, and dried. The crystals were suspended in 2 M hydro-
chloric acid (25 mL), heated to boil, cooled and the resulting
hydrochloride salts filtered off, washed with acetone, and dried
in vacuum. The pure products were obtained by crystallization
from appropriate solvents.
2,20-(Furan-2,5-diyl)bis(1,3-benzothiazole-6-carboximidamide) di-
hydrochloride (10a): Compound 10a was prepared using general
Methods A and B described above. Crystallization from a
water/acetone mixture afforded 0.296 g (54.3%, Method A)
and 0.393 g (72.1%, Method B) of yellow crystals: mp >300 ꢀC;
1H NMR (300 MHz, DMSO-d6) (δ ppm) 9.48 (s, 4H, disap-
peared on addition of D2O, 4Hamidine), 9.17 (s, 4H, disappeared
on addition of D2O, 4Hamidine), 8.66 (d, 2H, J = 1.4 Hz,
2Hbenzothiazole), 8.24 (d, 2H, J = 8.5 Hz, 2Hbenzothiazole), 7.90
(dd, 2H, J = 1.6 Hz, J = 8.6 Hz, 2Hbenzothiazole), 7.73 (s, 2H,
2Hfuran).

2430 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Racaneꢀ et al.
(ANOVA, Tukey posthoc test at p < 0.05). Finally, the effects
of the tested substances were evaluated by plotting the mean
percentage growth for each cell type in comparison to control on
dose-response graphs.
The variations of melting temperature (ΔTm) of CT-DNA or
poly(dAdT)2 induced by the presence of the various compounds
were calculated from the melting temperature measurement of
20 μM CT-DNA or poly(dAdT)2 incubated alone (control Tm)
or with 2 μM of the various compounds (drug/base pair ratio of
0.1) in 1 mL of BPE buffer. The ΔTm induced by the various
compounds on specific sites was measured at a drug/number of
site ratio of 0.1 using 2 μM of 22-mer hairpin oligonucleotides
(the 9 bp stem generated from autohybridization is underlined)
designed to specifically contain one site of interest (in bold):
ATTA (50-CGCATTACGTCTCCGTAATGCG-30); AATT
(50-CGCAATTCGTCTCCGAATTGCG-30); TTAA (50-CGC-
TTAACGTCTCCGTTAAGCG-30); GCGC (50-AGTGCGC-
TGTCTCCAGCGCACT-30) and ACCA (50-CGCACCACGT-
CTCCGTGGTGCG-30) in the absence or presence of 2 μM of the
specified compounds. The absorbency of DNA at 260 nm was
measured in quartz cells using the Uvikon 943 spectrophotometer
thermostatted with a Neslab RTE111 cryostat. Absorption value was
measured for each sample every minute over a range of 20 to 100 ꢀC
with an increment of 1 ꢀCpermin. TheTm values were obtained from
the midpoint of the hyperchromic transition, and the variation of
melting temperature was calculated from the equation ΔTm values =
Cell Cycle Analysis. A total of 5 ꢀ 105 cells were seeded per
Petri dish (10 cm in diameter, Sarstedt, Germany). After 24 h,
MCF-7 cells were treated with the compound 2b at concentra-
tions of 10 and 50 μM, and MiaPaCa-2 and MCF-7 cells were
treated with compound 3b at concentrations of 1 and 5 μM. After
24 and 48 h, the attached cells were trypsinized, combined with
floating cells, washed with PBS, and fixed with 70% ethanol.
Immediately before the analysis, the cells were washed again with
PBS and stained with 1 μg/mL of propidium iodide (PI) with the
addition of 0.2 μg/mL of RNase A. The stained cells were then
analyzed with Becton Dickinson FACScalibur flow cytometer
(10000 counts were measured). Each test point was performed in
duplicate. The percentage of the cells in each cell cycle phase was
based on the obtainedDNA histograms anddeterminedusingthe
WinMDI 2.9 and Cylchred software. Statistical analysis was
performed in Microsoft Excel by using the ANOVA at p < 0.05.
Detection of Apoptosis, Annexin V Assay. Detection and
quantification of apoptotic cells at a single cell level was
performed using Annexin V-Fluos staining kit (Roche), accord-
ing to the manufacturer’s recommendations. The MCF-7 cells
were seeded in chamber slides (4 ꢀ 104 cells/well, Lab-Tk II
Chamber slide, Nunc, SAD) and treated with compounds 2b
(concentrations 10 and 50 μM) and compound 3b (1 and 5 μM).
The MiaPaCa-2 cells were also seeded in 6-well plates (4 ꢀ 104
cells/well, Lab-Tk II Chamber slide, Nunc, SAD) and treated
only with compound 3b (1 and 5 μM). After 24 and 48 h,
the growth medium was removed from wells and 100 μL of
incubation buffer, containing Annexin V-Flous labeling reagent
and propidium iodide (PI) solution, was added per well and
incubated for 15 min. The cells were then washed with PBS with
the addition of 2% FCS, pelleted, and resuspended in two
staining solutions prepared in the Hepes buffer containing either
AnnexinV-fluoresceinlabeling reagent, propidium iodide(PI), or
their combination. Camptothecin (20 μM, Sigma) was used as the
control apoptosis-inducing agent. The cells were then analyzed
under the fluorescent microscope. The results are represented as
percentages of cells positive to annexin and/or propidium iodide
per total cell number. At least 100 cells were counted in each well.
UV/Visible Absorption Spectroscopy and DNA Melting Temp-
erature Studies. CT-DNA, double-stranded poly(dAdT)2, and
poly(dGdC)2 oligonucleotides were purchased from Sigma.
CT-DNA was prepared as previously described (David-
Cordonnier et al.). Compounds 2a, 3a, 3b, 10b, 11a, and 11b
were dissolved at 10 mM in DMSO, whereas compounds 2b and
10a were prepared as 5 and 4 mM solutions, respectively. Each
compound was further extemporarily diluted in the appropriate
aqueous reaction buffers.
Tm[DrugþDNA] - Tm[DNA alone]
.
Fluorescence Spectroscopy. Fluorescence spectral measure-
ment were recorded using 1 or 5 μM of the fluorescent drugs (as
specified in the figure legend) incubated in 1 mL of BPE buffer in
the presence or absence of increasing concentrations of CT-
DNA (0.1 to 30 μM of base pairs or below if precipitation of the
CT-DNA was observed in the sample) in a quartz cuvette of
10 mm path length. The excitation wavelengths were 380
(compounds 2a and 2b), 386 (compounds 3a, 3b and 10a), 318
(compound 11a), 388 (compound 10b), and 392 nm (compound
11b). The quenching constant KSV was deduced from
Stern-Volmer plots expressing the ratio of fluorescence of the
compound alone (F0) over the fluorescence of the compound in
the presence of CT-DNA (F) as a function of CT-DNA con-
centration. In this configuration, F0/F = 1 þ KSV[CT-DNA],
where the slope KSV is considered as an equilibrium constant for
the static quenching process.
Circular Dichroism. A fixed concentration of CT-DNA
(200 μM of base pairs) was incubated with or without
(control) increasing concentrations of the different drugs
(from 1 to 25 μM for compounds 2a, 2b, and 3a or from 1 to
50 μM for the other compounds) in 1 mL of sodium cacodylate
(1 mM, pH 7.0). The analyses using poly(dAdT)2 or poly-
(dGdC)2 (200 μM of base pairs) were performed using incre-
mented drug concentration from 2 to 80 μM. The CD spectra
were collected from 530 to 230 nm with a resolution of 0.1 nm, in
a quartz cell of 10 mm path length, using a J-810 Jasco spectro-
polarimeter at 20 ꢀC controlled by a PTC-424S/L peltier type
cell changer (Jasco).
UV/visible spectral absorption measurements were con-
ducted using a fixed concentration (20 μM) of the various tested
drugs incubated in 1 mL of BPE buffer (6 mM Na2HPO4, 2 mM
NaH2PO4, 1 mM EDTA, pH 7.1) in the presence or absence of
graded concentrations of CT-DNA (from 0.1 to 200 μM of base
pairs) in a quartz cuvette of 10 mm path length. The UV/visible
spectra were recorded from 300 to 450 nm using a Uvikon 943
spectrophotometer and are referenced against a cuvette contain-
ing the same graded concentrations of CT-DNA as that in the
cuvette sample to directly substrate intrinsic CT-DNA absorp-
tion. The binding constant, K, was determined from the spectro-
scopic titration data using the following equation [CT-
DNase I Footprinting. DNase I footprinting experiments were
performed essentially as described in Peixoto et al.39 Briefly, the
265 and 117 bp DNA fragments were obtained from double
digestion of the pBS plasmid (Stratagene, La Jolla, CA) at
EcoRI and PvuII restriction sites for 1 h with the corresponding
enzymes in their respective buffers. Similarly, the 257 bp DNA
fragment was generated from BamHI and PciI digestion of
pGL3cat-basic vector (Promega). The generated DNA fragments
were 30-end labeled using 5 µL of R-[32P]-dATP (265 and 117 bp
fragments) or R-[32P]-dGTP (3000 Ci/mmol each, GE Health-
care, Buckinghamshire, England) and 10 units of Klenow
enzyme (BioLabs) for 30 min at 37 ꢀC. The resulting radio-
labeled DNA fragments were then separated from the plasmid
remnant on a 6% polyacrylamide gel under native conditions in
TBE buffer (89 mM Tris base, 89 mM boric acid, 2.5 mM Na2
EDTA, pH 8.3). The portion of the gel containing the DNA
fragments was cut off from the gel, crushed, and dialyzed
overnight against 400 μL of elution buffer (10 mM Tris-HCl
pH 8.0, 1 mM EDTA, 100 mM NaCl). The DNA was recovered
DNA]/(εFree - εObs) = [CT-DNA]/(εBound - εFree) þ 1/K(εBound
-
ε
Free), where εObs εFree and εBound correspond to
,
,
A /[compound], the extinction coefficient for the free com-
pound, and the extinction coefficient for the compound fully
bound to CT-DNA, respectively. By plotting [CT-DNA]/(εFree
- εObs) versus [CT-DNA] (Supplementary Figure 1B), K is given
by the ratio of the slope to the intercept.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6 2431
from polyacrylamide gel by filtration through a Millipore 0.22
μm membrane and subsequently precipitated with cold ethanol
followed by centrifugation. Appropriate concentrations of the
various ligands (as indicated in the figures legend) were incu-
bated with the radio-labeled DNA fragments for a 15 min
incubation at 37 ꢀC to ensure equilibrium prior to digestion of
the DNA by the addition of DNase I (final concentration 0.001
unit/mL) in 20 mM NaCl, 2 mM MgCl2, 2 mM MnCl2, pH 7.3.
After 3 min of digestion, the reaction was stopped by freeze-
drying. Samples were lyophilized and subsequently dissolved in
4 μL of denaturing loading buffer (80% formamide solution
containing tracking dyes). The DNA samples were heated at
90 ꢀC for 4 min and rapidly chilled on ice for 4 min prior to
electrophoresis. The generated DNA cleavage products were
resolved on a denaturing polyacrylamide gel (0.35 mm thick, 8%
polyacrylamide containing 8 M urea). After a 90 min electro-
phoresis at 65 W in TBE buffer, gels were soaked in 10% acetic
acid for 10 min, transferred to Whatman 3 MM paper, and dried
under vacuum at 80 ꢀC. A Molecular Dynamics STORM 860
was used to collect data from storage screens exposed to dried
gels overnight at room temperature. The identity of the bases
was established from comparison of the relative position of the
bands to the guanines sequencing standard (G-track) classically
obtained using dimethylsulfate and piperidine treatment of the
same DNA fragment. Quantifications of the footprint were
performed using ImageQuant 3.3 software.
Fluorescence Microscopy. HT-29 human colon carcinoma
cells (ATCC) were maintained as monolayers in 150 cm2 culture
flasks using culture medium consisting of DMEM-glutaMAX
medium supplemented with 10% fetal bovine serum, penicillin
(100 IU/mL), and streptomycin (100 μg/mL); 5 ꢀ 104 HT-29
cells were cultured in Lab-Tek II Chamber Slide (Nunc Int.)
incubated with 5 μM of the tested compounds for 16 h at 37 ꢀC in
culture medium. Cells were then rinsed three times with ice-cold
PBS prior to be incubated in the dark with 250 nM of the
fluorescent cytoplasmic probe Mito Fluor Red 588 (Molecular
Probes) for 30 min at 37 ꢀC. After a further three washings with
PBS, a drop of Vectashield antifade solution (AbCys S.A.,
France) was then added and the treated portion of the slide
was visualized immediately by fluorescence microscopy using an
ApoTome microscope (Zeiss) with an X63 oil-immersion objec-
tive. Images were captured using the software AxioVision with
an excitation wavelength of 358 nm for an emission at 461 nm in
blue for all compounds but 11b, for which the excitation
wavelength was 495 nm for an emission at 519 nm (green).
Compound 3b was analyzed at both blue and green fluore-
scence. The used excitation and emission wavelengths for Mito
Fluor Red 588 were 588 and 622 nm, respectively.
well as the UV/visible absorbance spectrometry data, the
DNase I footprinting assay on the 257 bp DNA fragment and
the densitometric analyses of DNase I footprinting assays on the
257 and 117 bp DNA fragments. This material is available free
of charge via the Internet at http://pubs.acs.org.
References
(1) Ban, M.; Taguchi, H.; Katsushima, T.; Takahashi, M.; Shinoda, K.;
Watanabe, A.; Tominaga, T. Novel antiallergic and antiinflammatory
agents. Part I: Synthesis and pharmacology of glycolic amide deriva-
tives. Bioorg. Med. Chem. 1998, 6, 1069–1076.
(2) Papadopoulou, C.; Geronikaki, A.; Hadjipavlou-Litina, D.
Synthesis and biological evaluation of new thiazolyl/benzothiazo-
lyl-amides, derivatives of 4-phenyl-piperazine. Farmaco 2005, 60,
969–973.
(3) Chung, Y.; Shin, Y.-K.; Zhan, C.-G.; Lee, S.; Cho, H. Synthesis
and evaluation of antitumor activity of 2- and 6-[(1,3-benzothiazol-
2-yl)aminomethyl]-5,8-dimethoxy-1,4-naphthoquinone derivatives.
Arch. Pharm. Res. 2004, 27, 893–900.
(4) McFadyen, M. C. E.; Melvin, W. T.; Murray, G. I. Cytochrome
P450 enzymes: novel options for cancer therapeutics. Mol. Cancer
Ther. 2004, 3, 363–371.
(5) Yoshida, M.; Hayakawa, I.; Hayashi, N.; Agatsuma, T.; Oda, Y.;
Tanzawa, F.; Iwasaki, S.; Koyama, K.; Furukawa, H.; Kurakata,
S. Synthesis and biological evaluation of benzothiazole derivatives
as potent antitumor agents. Bioorg. Med. Chem. Lett. 2005, 15,
3328–3332.
ꢀ
ꢂ
ꢀ
ꢂ ꢀ
(6) Starcevic, K.; Caleta, I.; Cincic, D.; Kaitner, B.; Kralj, M.; Ester,
K.; Karminski-Zamola, G. Synthesis, crystal structure determina-
tion and antiproliferative evaluation of novel benzazoyl benza-
mides. Heterocycles 2006, 68, 2285–2299.
ꢀ
ꢂ
ꢀ
ꢀ
ꢀ
(7) Caleta, I.; Kralj, M.; Bertosa, B.; Tomic, S.; Pavlovic, G.; Pavelic,
K.; Karminski-Zamola, G. Novel cyano and amidinobenzothiazole
derivatives: synthesis, antitumor evaluationand X-ray quantitative
structure-activity relationship (QSAR) analysis. J. Med. Chem. 2009,
52, 1744–1756.
(8) Baell, J. B.; Forsyth, S. A.; Gable, R. W.; Norton, R. S.; Mulder,
R. J. Design and synthesis of type-III mimetics of ω-conotoxin
GVIA. J. Comput.-Aided Mol. Des. 2002, 15, 1119–1136.
(9) Westway, S. M.; Thompson, M.; Rami, H. K.; Stemp, G.; Trouw,
L. S.; Mitchell, D. J.; Seal, J. T.; Medhurst, S. J.; Lappin, S. C.;
Biggs, J.; Wright, J.; Arpino, S.; Jerman, J. C.; Cryan, J. E.;
Holland, V.; Winborn, K. Y.; Coleman, T.; Stevens, A. J.; Davis,
J. B.; Gunthorpe, M. J. Design and synthesis of 6-phenylnicotina-
mide derivatives as antagonists of TRPV1. Bioorg. Med. Chem.
Lett. 2008, 18, 5609–5613.
(10) Das, J.; Lin, J.; Moquin, R. V.; Shen, Z.; Spergel, S. H.; Wityak, J.;
Doweyko, A. M.; DeFex, H. F.; Fang, Q.; Pang, S.; Pitt, S.; Ren
Shen, D.; Schieven, G. L.; Barrish, J. C. Molecular design, synth-
esis, and structure-activity relationships leading to the potent and
selective P56lck inhibitor BMS-243117. Bioorg. Med. Chem. Lett.
2003, 13, 2145–2149.
(11) Das, J.; Moquin, R. V.; Lin, J.; Liu, C.; Doweyko, A. M.; DeFex,
H. F.; Fang, Q.; Pang, S.; Pitt, S.; Ren Shen, D.; Schieven, G. L.;
Barrish, J. C.; Wityak, J. Discovery of 2-amino-heteroaryl-ben-
zothiazole-6-anilides as potent p56lck inhibitors. Bioorg. Med.
Chem. Lett. 2003, 13, 2587–2590.
(12) Gaillard, P.; Jeanclaude-Etter, I.; Ardissone, V.; Arkinstall, S.;
Cambet, Y.; Camps, M.; Chabert, C.; Church, D.; Cirillo, R.;
Gretener, D.; Halazy, S.; Nichols, A.; Szyndralewiez, C.; Vitte,
P.-A.; Gotteland, J.-P. Design and synthesis of the first generation
of novel potent, selective, and in vivo active (benzothiazol-2-
yl)acetonitrile inhibitors of the c-Jun N-terminal kinase. J. Med.
Chem. 2005, 48, 4596–4607.
Acknowledgment. We greatly appreciate the financial help
of the Croatian Ministry of Science Education and Sports
(Projects 125-0982464-1356, 098-0982464-2393, 117-0000000-
ꢀ
3283). M.-H.D.-C. thanks the Fonds europeen de develop-
pement regional (FEDER, European Community) and the
ꢀ
ꢀ
ꢀ
Region Nord-Pas de Calais for a grant and a fellowship to
S.D., the Ligue Nationale contre le Cancer (Comite du Nord)
ꢀ
(13) Liu, C.; Lin, J.; Pitt, S.; Zhang, R. F.; Sack, J. S.; Kiefer, S. E.; Kish,
K.; Doweyko, A. M.; Zhang, H.; Marathe, P. H.; Trzaskos, J.;
Mckinnon, M.; Dodd, J. H.; Barrish, J. C.; Schieven, G. L.;
Leftheris, K. Benzothiazole based inhibitors of p38a MAP kinase.
Bioorg. Med. Chem. Lett. 2008, 18, 1874–1879.
and the Institut pour la Recherche sur le Cancer de Lille
(IRCL) for grants. R.N. is supported by the CHRU de Lille
ꢀ
and the Conseil Regional Nord-Pas de Calais for her Ph.D.
(14) Pinar, A.; Yurdakul, P.; Yildiz, I.; Temiz-Arpaci, O.; Acan, N. L.;
Aki-Sener, E.; Yalcin, I. Some fused heterocyclic compounds as
eukaryotic topoisomerase II inhibitors. Biochem. Biophys. Res.
Commun. 2004, 317, 670–674.
(15) Abdel-Aziz, M.; Matsuda, K.; Otsuka, M.; Uyeda, M.; Okawara,
T.; Suzuki, K. Inhibitory activities against topoisomerase I & II by
polyhydroxybenzoyl amide derivarives and their structure-
activity relationship. Bioorg. Med. Chem. Lett. 2004, 14, 1669–1672.
(16) Manjula, S. N.; Malleshappa, N.; Noolvi, K.; Vipan Parihar, S. A.;
Reddy, M.; Ramani, V.; Gadad, A. K.; Singh, G. N.; Gopalan
Kutty, C.; Rao, M. Synthesis and antitumor activity of optically
fellowship. We are grateful to the IRCL for a postdoctoral
fellowship to P.P. and technical expertise (M.-P.H.). The
ꢀ
ꢀ
ꢀ
Institut de Medecine Predictive et de Recherche Therapeu-
tique (IMPRT)-IFR114 is acknowledged for access to the
Molecular Dynamics STORM 860 equipment.
Supporting Information Available: Experimental and spectro-
scopic data for cyano derivative 7, elemental analysis and
spectroscopic data for new compounds (10a, 10b, 11a, 11b) as

2432 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 6
Racaneꢀ et al.
ꢁ
ꢀ
ꢀ
ꢂ
ꢀ
active thiourea and their 2-aminobenzothiazole derivatives: a
novel class of anticancer agents. Eur. J. Med. Chem. 2009, 44,
2923–2929.
(28) Racane, L.; Tralic-Kulenovic, V.; Fiser-Jakic, L.; Boykin, W. D.;
Karminski-Zamola, G. Synthesis of bis-substituted amidinoben-
zothiazoles as potential anti-HIV agents. Heterocycles 2001, 55,
2085–2098.
(17) Repicky, A.; Jantova, S.; Cipak, L. Apoptosis induced by 2-acetyl-
3-(6-methoxybenzothiazo)-2-yl-amino-acrylonitrile in human leu-
kemia cells involves ROS-mitochondrial mediated death signaling
and activation of p38 MAPK. Cancer Lett. 2009, 277, 55–63.
ꢀ
ꢀ
ꢁ
ꢂ
ꢀ
(29) Tralic-Kulenovic, V; Karminski-Zamola, G.; Racane, L.; Fiser-Jakic,
L. Synthesis of symmetric and unsymetric bis cyanosubstituted hetero-
cycles. Heterocycl. Commun. 1998, 4, 423–427.
€
ꢁ
ꢀ
ꢀ
ꢀ
ꢀ
(18) Tasler, S.; Muller, O.; Wieber, T.; Herz, T.; Krauss, R.; Totzke, F.;
(30) Racane, L.; Tralic-Kulenovic, V.; Mihalic, Z.; Pavlovic, G.;
Karminski-Zamola, G. Synthesis of new amidino-substituted
2-aminothiophenole: mild basic ring opening of benzothiazole.
Tetrahedron 2008, 64, 11594–11602.
0
benzothiazoles as kinase inhibitors: hit identification and scaffold
€
Kubbutat, M. H. G.; Schachtele, C. N-Substituted 2 -(aminoaryl)-
hopping. Bioorg. Med. Chem. Lett. 2009, 19, 1349–1356.
(19) Chaires, J. B.; Ren, J.; Hamelberg, D.; Kumar, A.; Pandya, V.;
Boykin, D. W.; Wilson, W. D. Structural selectivity of aromatic
diamidines. J. Med. Chem. 2004, 47, 5729–5742.
(31) Pinar, A.; Yurdakul, P.; Yildiz, I; Temi-Arpaci, O.; Acan, N. L.;
Aki-Sener, E; Yalcin, I. Some fused heterocyclic compounds as
eukaryotic topoisomerase II inhibitors. Biochem. Biophys. Res.
Commun. 2004, 317, 670–674.
(20) Cai, X.; Gray, P. J, Jr.; Von Hoff, D. D. DNA minor groove
binders: back in the groove. Cancer Treat. Rev. 2009, 35, 437–450.
(21) Lansiaux, A.; Dassonneville, L.; Facompre, M.; Kumar, A.;
(32) Abdel-Aziz, M.; Matsuda, K.; Otsuka, M.; Uyeda, M.; Okawara,
T.; Suzuki, K. Inhibitory activities against topoisomerase I & II
by ployhydroxybenzoyl amide derivatives and their structure-
activity relationship. Bioorg. Med. Chem. Lett. 2004, 14, 1669–
1672.
ꢀ
Stephens, C. E.; Bajic, M.; Tanious, F.; Wilson, W. D.; Boykin,
D. W.; Bailly, C. Distribution of furamidine analogues in tumor
cells: influence of the number of positive charges. J. Med. Chem.
2002, 45, 1994–2002.
(33) Okada, H.; Mak, T. W. Pathways of apoptotic and non-apoptotic
death in tumour cells. Nat. Rev. Cancer 2004, 4, 592–603.
(34) Dıaz, L. F.; Chiong, M.; Quest, A. F. G.; Lavandero, S.; Stutzin, A.
´
Mechanisms of cell death: molecular insights and therapeutic
(22) Lansiaux, A.; Tanious, F; Mishal, Z.; Dassonneville, L.; Kumar,
A.; Stephens, C. E.; Hu, Q.; Wilson, W. D.; Boykin, D. W.; Bailly,
C. Distribution of furamidine analogues in tumor cells: targeting of
the nucleus or mitochondria depending on the amidine substitu-
tion. Cancer Res. 2002, 62, 7219–7229.
(23) Munde, M.; Ismail, M. A.; Arafa, R.; Peixoto, P.; Collar, C. J.; Liu,
Y.; Hu, L.; David-Cordonnier, M.-H.; Lansiaux, A.; Bailly, C.;
Boykin, D. W.; Wilson, W. D. Design of DNA minor groove
binding diamidines that recognize GC base pair sequences: a novel
dimeric-hinge motif. J. Am. Chem. Soc. 2007, 129, 13732–13743.
(24) Miao, Y.; Lee, M. P.; Parkinson, G. N.; Batista-Parra, A.; Ismail,
M. A.; Neidle, S.; Boykin, D. W.; Wilson, W. D. Out-of-shape
DNA minor groove binders: induced fit interactions of heterocyclic
dications with the DNA minor groove. Biochemistry 2005, 44,
14701–14708.
perspectives. Cell Death Differ. 2005, 12, 1449–1456.
(35) Zhang, J.; Cheng, C.; He, C.-L.; Zhou, Y.-J.; Cao, Y. The expres-
sion of Bcl-XL, Bcl- XS and p27Kip1 in topotecan-induced apop-
tosis in hepatoblastoma HepG2 cell line. Cancer Invest. 2008, 26,
456–463.
(36) Joe, A. K.; Liu, H.; Suzui, M.; Vural, M. E.; Xiao, D.; Weinstein, I.
B. Resveratrol induces growth inhibition, S-phase arrest, apopto-
sis, and changes in biomarker expression in several human cancer
cell lines. Clin. Cancer Res. 2002, 8, 893–903.
(37) Peixoto, P.; Bailly, C.; David-Cordonnier, M.-H. In Methods in
Molecular Biology Series, Drug-DNA Interactions; Fox, K. R., Ed.;
Humana Press: Totowa, NJ, 2010; Vol. 613, pp 235-256.
ꢀ ꢀ ꢂ
(38) Gazivoda, T.; Raic-Malic, S.; Kristafor, V.; Makuc, D.; Plavec, J.;
(25) Tanious, F. A.; Laine, W.; Peixoto, P.; Bailly, C.; Goodwin, K. D.;
Lewis, M. A.; Long, E. C.; Georgiadis, M. M.; Tidwell, R. R.;
Wilson, W. D. Unusually strong binding to the DNA minor groove
by a highly twisted benzimidazole diphenylether: induced fit and
bound water. Biochemistry 2007, 46, 6944–6956.
ꢀ
ꢀ
ꢀ
Bratulic, S.; Kraljevic Pavelic, S.; Pavelic, K.; Naesens, L.;
Andrei, G.; Snoeck, R.; Balzarini, J.; Mintas, M. The novel
pyrimidine and purine derivatives of l-ascorbic acid: synthesis,
one- and two-dimensional 1H and 13C NMR study, cytostatic
and antiviral evaluation. Bioorg. Med. Chem. 2008, 16, 5624–5634.
(39) 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 transcrip-
tion factors Pit-1 and Brn-3 by selective binding of a phenyl-furan-
benzimidazole dication. Nucleic Acids Res. 2008, 36, 3341–3353.
ꢁ
ꢀ
ꢀ
(26) Racane, L.; Tralic-Kulenovic, V.; Kitson, R. P.; Karminski-Zamola,
G. Synthesis and antiproliferative activity of cyano and amidino sub-
stituted 2-phenylbenzothiazoles. Monatsh. Chem. 2006, 137, 1571–1577.
ꢁ
ꢀ
ꢀ
ꢀ
(27) Racane, L.; Stojkovic, R.; Tralic-Kulenovic, V.; Karminski-Zamola,
G. Synthesis and antitumor evaluation of novel derivatives of 6-amino-
2-phenylbenzothiazoles. Molecules 2006, 11, 325–333.