Anilides and quinolones with nitrogen-bearing substituents from benzothiophene and thienothiophene series: Synthesis, photochemical synthesis, cytostatic evaluation, 3D-derived QSAR analysis and DNA-binding properties

European Journal of Medicinal Chemistry 71 (2014) 267e281

Contents lists available at ScienceDirect

European Journal of Medicinal Chemistry

journal homepage: h ttp://www.elsevier.com/locate/ejmech

Original article

Anilides and quinolones with nitrogen-bearing substituents from

benzothiophene and thienothiophene series: Synthesis,

photochemical synthesis, cytostatic evaluation, 3D-derived QSAR

analysis and DNA-binding properties

b

,

c

,

d

e

ꢀ ^a

ꢁ

**

Maja Aleksic , Branimir Bertosa

, Raja Nhili , Sabine Depauw , Irena Martin-Kleiner ,

d

e

ꢀ ^b

Marie-Hélène David-Cordonnier , Sanja Tomic , Marijeta Kralj ,

Grace Karminski-Zamola ^a

,

*

a

ꢀ

Department of Organic Chemistry, Faculty of Chemical Engineering and Technology, University of Zagreb, Marulicev trg 20, P.O. Box 177, HR-10000 Zagreb,

Croatia

b

ꢁ

ꢀ

ꢁ

Division of Physical Chemistry, “RuCer Boskovic” Institute, Bijenicka Cesta 54, P.O. Box 1016, HR-10000 Zagreb, Croatia

^cDivision of Physical Chemistry, Faculty of Natural Sciences, University of Zagreb, Horvatovac 102A, HR-10000 Zagreb, Croatia

^dINSERM U837-JPARC (Jean-Pierre Aubert Research Center), Team “Molecular and Cellular Targeting for Cancer Treatment”, Université Lille Nord de France,

IFR-114, Institut pour la Recherche sur le Cancer de Lille, Place de Verdun, F-59045 Lille Cedex, France

e

ꢁ

ꢀ

ꢁ

Division of Molecular Medicine, “RuCer Boskovic” Institute, Bijenicka Cesta 54, P.O. Box 1016, HR-10000 Zagreb, Croatia

a r t i c l e i n f o

a b s t r a c t

Article history:

Received 4 June 2013

Received in revised form

4 November 2013

Accepted 7 November 2013

Available online 15 November 2013

A series of new anilides (2aec, 4e7, 17aec, 18) and quinolones (3aeb, 8aeb, 9aeb, 10e15, 19) with

nitrogen-bearing substituents from benzo[b]thiophene and thieno[2,3-c]thiophene series are prepared.

Benzo[b]thieno[2,3-c]- and thieno[3⁰,2⁰:4,5]thieno[2,3-c]quinolones (3aeb, 8aeb) are synthesized by the

reaction of photochemical dehydrohalogenation from corresponding anilides. Anilides and quinolones

were tested for the antiproliferative activity. Fused quinolones bearing protonated aminium group,

quaternary ammonium group, N-methylated and protonated aminium group, amino and protonated

amino group (8a, 9b, 10e12) showed very prominent anticancer activity, whereby the hydrochloride salt

of N⁰,N⁰-dimethylaminopropyl-substituted quinolone (14) was the most active one, having the IC₅₀

concentration at submicromolar range in accordance with previous QSAR predictions. On the other hand,

ﬂexible anilides were among the less active. Chemometric analysis of investigated compounds was

performed. 3D-derived QSAR analysis identiﬁed solubility, metabolitic stability and the possibility of the

compound to be ionized at pH 4e8 as molecular properties that are positively correlated with anticancer

activity of investigated compounds, while molecular ﬂexibility, polarizability and sum of hydrophobic

surface areas were found to be negatively correlated. Anilides 2aeb, 4e7 and quinolones 3aeb, 8aeb, 9b

and 10e14 were evaluated for DNA binding propensities and topoisomerases I/II inhibition as part of

their mechanism of action. Among the anilides, only compound 7 presented some DNA binding pro-

pensity whereas the quinolones 8b, 9b and 10e14 intercalate in the DNA base pairs, compounds 8b, 9b

and 14 being the most efﬁcient ones. The strongest DNA intercalators, compounds 8b, 9b and 14, were

clearly distinguished from the other compounds according to their molecular descriptors by the PCA and

PLS analysis.

Keywords:

Amides

Quinolones

Antitumor evaluation

QSAR analysis

DNA-binding properties

1. Introduction

Together with the biological aspects of cancer treatment,

* Corresponding author. Tel.: þ385 14597215; fax: þ385 14597250.

** Corresponding author. Division of Physical Chemistry, Faculty of Natural Sci-

ences, University of Zagreb, Horvatovac 102A, HR-10000 Zagreb, Croatia. Tel.: þ385

14606132; fax: þ385 14606131.

chemotherapy using small molecules or bioactive natural products

is still the useful manner of cancer treatment, whereby the major

cellular targets are DNA and tubulin, along with various protein

kinases [1e3]. Therefore, the search for novel small organic mole-

cules with either better activity and/or selectivity is still of great

ꢁ

E-mail addresses: bbertosa@irb.hr (B. Bertosa), gzamola@fkit.hr (G. Karminski-

Zamola).

http://dx.doi.org/10.1016/j.ejmech.2013.11.010

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

268

R

importance for developing novel anticancer drugs. Amides and

quinolones exhibit several pharmacological activities and have

therefore attracted considerable attention from medicinal and

synthetic organic chemists [4e7]. Recently some other compounds

derived from benzo[b]thiophene series showed as potent tubulin

polymerization inhibitors [8]. Some similar compounds from ben-

zothiophene series namely 3-(aryl)benzothieno[2,3-c]pyran-1-

ones (tricyclic lactones) were prepared by a tandem one-pot

Sonogashira coupling and intramolecular cyclization. Tricyclic lac-

tones obtained were evaluated for their capacity to inhibit the

in vitro growth of three human tumor cell lines, MCF-7 (breast

adenocarcinoma), NCI-H460 (non-small cell lung cancer) and SF-

268 (CNS cancer). Most of the compounds showed a high growth

inhibitory effect on all the tested cell lines, with GI₅₀values in the

Cl

O

N

O

R

S

NH

R

S

R₁= H, Br, CH₃, NO₂, COOCH₂CH₃,

R

₂= CONH(CH₂)₃NH⁺(CH₃)₂Cl^-,

R₃= H, CH₃

Fig. 1. Earlier prepared benzo[b]thienyl-carboxanilides and benzo[b]thieno[2,3-c]

quinolones.

mM range. A structureeactivity relationship was established for the

Sonogashira products and for the tricyclic lactones, namely related

to the presence and position of substituents (OMe and/or F) in the

benzothiophene moiety or in the phenyl ring [9]. New organo-

metalic compounds from benzo[b]thiophene series with antitumor

properties were prepared [10], as well as a series of N-(2-(5-ﬂuoro-

2-(4-ﬂuorophenylthio)benzo[b]thiophen-3-yl)ethyl) acylamides,

which were evaluated for their binding afﬁnity and intrinsic ac-

tivity at melatonin receptors [11]. Also, benzothiophenes offer in-

terest to pharma industry as scaffolds for synthesis of raloxifene

and relevant structural analogues of selective estrogen receptor

modulators (SERM) [12].

part of their mechanism of action assessment. We have shown that

benzo[b]thienyl-carboxanilide with 2-imidazolinyl-substituent

binds to DNA as a groove binder, while its cyclic analog binds as an

intercalator.

Using this knowledge from our previous investigations [18], we

synthesized new compounds with a new nitrogen-bearing sub-

stituent. Thus, in this work we introduced amino, protonated

amino, acetamido, N,N-dimethylamino groups, as well as aminium

or quaternary ammonium salts in the anilide part of the substituted

carboxanilides and quinolones of the benzo[b]thiophene and

thieno[2,3-b]thiophene series. Quinolon nitrogen was in some

compounds substituted with methyl or N,N-dimethylaminopropyl

groups. These changes were made in an attempt to affect properties

found by previous QSAR analysis as important for activity, such as,

integy moments and solubility. To elucidate the impact of the

variation of thiophene vs. benzene ring, we also prepared nitro,

acetamido and N,N-dimethylamino substituted thieno[2,3-b]

thienyl-carboxanilides and thieno[3⁰,2⁰:4,5]thieno[2,3-c]quinolone.

Volsurf [19] based QSAR analysis was applied in order to help us

to identify which molecular properties of the studied compounds

have the largest impact on antitumor activity. Similar approaches

for building 3D-derived QSAR models have already been proven as

useful in investigation of different classes of the potentially bio-

logically active compounds [20e24] and can be valuable in

designing new compounds with enhanced anticancer activity [24].

Moreover, this paper presents an example of usefulness of the

approach in which QSAR analysis serves as a guideline in designing

new compounds with increased anticancer activity: the most active

compound presented in this paper was designed according to the

previous QSAR analysis [25].

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[3⁰,2⁰:4,5]thieno- and benzo[b]thieno[2,3-c]

quinolones, containing a N,N-dimethylaminopropyl substituent in

the amine part of the molecule and on the quinolone nitrogen

[13,14]. Moreover, we also reported the synthesis and anti-

proliferative activity of various cyano- and/or amidino-substituted

benzo[b]thieno[2,3-c]quinolones and their acyclic precursors [15],

since it is well known that amidines are structural parts of

numerous compounds of 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 pre-

cursors, which did not intercalate at all. In order to examine impact

of thiophene vs. benzene ring, we have also prepared amidino-

substituted thieno[3⁰,2⁰:4,5]thieno- and thieno[2⁰,3⁰:4,5]thieno

[2,3-c]quinolones, as well as their acyclic precursors, whereby both

cyclic and acyclic compounds showed pronounced antitumor ac-

tivity [16]. 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, respec-

tively, resulted with increased antitumor activity. Some of the

naphthyridone compounds were active in the submicromolar

range, with special selectivity to MiaPaCa-2, HeLa and SW620 cells.

Compounds with high DNA binding capacity were identiﬁed as the

most promising inhibitors of tumor growth and of topoisomerase

activity whereby the 2-imidazolinyl-substituted derivatives

showed the most prominent activity [17].

2. Results and discussion

2.1. Chemistry

Prepared compounds were synthesized according to the pro-

cedures shown in Schemes 1 and 2 by the conventional methods of

organic synthesis for the preparation of similar heterocyclic com-

pounds [13e18].

Further on, a class of substituted benzo[b]thienyl- and thieno

[2,3-b]thienyl-carboxanilides and benzo[b]thieno[2,3-c]- and

thieno[3⁰,2⁰:4,5]thieno[2,3-c]quinolones prepared (Fig. 1) in the

multistep synthesis and photochemical synthesis was tested for the

antiproliferative activity on various tumor cell lines [18]. Using the

experimentally obtained antitumor measurements, 3D-derived

QSAR models were obtained and molecular properties that have

the highest impact on antitumor activity of quinolones were

identiﬁed. Some of the prepared compounds were evaluated for

DNA binding propensities and topoisomerases I and II inhibition as

Starting from the 3-chlorobenzo[b]thiophene-2-carbonyl chlo-

ride 1 and 3-chlorothieno[2,3-b] thiophene-2-carbonyl chloride 16

via condensation reaction with substituted anilines corresponding

carboxanilides 2aec and 17aec were prepared. Carboxanilides 2ae

b were photochemically cyclized into corresponding quinolones

3aeb. Compound 2b was also hydrogenated in the presence of

palladium into compound 4. Carboxanilide 2c was methylated with

methyl iodide to obtain quaternary ammonium salt 7. Quinolone 3a

was hydrolized with NaOH into compound 11. N-alkylation of 3a

with N,N-dimethyaminopropylchloride hydrochloride in the

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

269

H₃COCHN

Cl

O

N

NH₃⁺Cl^-

S

NH

CH₃

N

O

S

5

x HCl

H₃C

14

HCl(g)/EtOH

Cl

O₂N

O

H₃COCHN

S

4

NH

NH₂

NH

O

N

CH₃

S

H₂/Pd

N

O

S

3b

H₃C

13

hν

Cl

MeOH/toluene

O

(CH₃)₂N(CH₂)₃ClxHCl

NaH/DMF

S

NH

NO₂

2b

H₃COC(H₃C)N

CH₃I/NaH

DMF

H₃COCHN

Cl

hν

MeOH/toluene

R

H₂N

O

COCl

CH₃

N

NH

Et₃N

toluene

S

NH

NHCOCH₃

1

2a

O

S

R = NHCOCH₃,

NO₂,

15

3a

Cl

NaOH/

MeOH

O

N(CH₃)₂

^-Cl⁺H₃N

H₂N

S

NH

N(CH₃)₂

2c

HCl(g)/EtOH

CH₃I

aceton/diethylether

NH

O

NH

O

HCl(g)/EtOH

S

Cl

11

12

O

CH₃

NH⁺Cl^-

S

NH

N⁺(CH₃)₃I^-

S

NH

CH₃

6

7

R

CH₃

NH⁺Cl^-

hν

MeOH/toluene

hν

H₃C

EtOH/H₂O

R

CH₃

N

HCl_(g)/EtOH

CH₃

N

O

S

CH₃I/NaH

DMF

NH

O

S

9a R = N(CH₃)₂

O

10

S

R

8a R = N⁺H(CH₃)₂Cl^-

8b R = N⁺(CH₃)₃I^-

CH₃

N

O

S

9b R = N⁺(CH₃)₃I^-

Scheme 1. Synthesis of substituted 3-cholorobenzo[b]thiophene-2-carboxamides 2aec, 4e7 and benzo[b]thieno[2,3-c]quinolones 8aeb, 9aeb, 10e15.

H₃COCHN

Cl

H₂N

R

O

hν

COCl

NH

O

S

NH

R

MeOH/

toluene

R = NHCOCH3, NO2,

N(CH3)2/Et3N

16

S

17a R=NHCOCH₃

17b R=NO₂

17c R=N(CH₃)₂

19

HCl(g)/EtOH

Cl

O

CH₃

NH⁺Cl^-

CH₃

S

NH

18

Scheme 2. Synthesis of 3-cholorothieno[2,3-b]thiophene-2-carboxamides 17aec, 18 and thieno[3⁰,2⁰:4,5]thieno[2,3-c]quinolone 19.

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

270

presence of NaH gave the expected quinolone 13. Carboxanilides 6,

7 and 17a were photochemically cyclized into corresponding

quinolones 8a, 8b and 19. Quinolones 3a, 8aeb were methylated in

the presence of NaH in DMF on the lactame nitrogen into N-methyl

compounds 15, 9aeb. During the methylation reaction of com-

pound 8a the N,N-dimethylamino hydrochloride substituent was

transformed into the N,N-dimethylamino substituent in the 9a so

that the 9a, without identiﬁcation, was protonated with HCl_ginto

aminium salt 10. Also carboxanilides 2c, 4, 17a and quinolones 11

and 13 were protonated into hydrochloride salts.

Structures of the compounds were conﬁrmed by the ¹H and ¹³

C

NMR analysis. The formation of the amide bond was supported by

the appearance of a one-proton singlet at 9.88 ppme11.15 ppm in

the ¹H NMR spectra of carboxanilides 2aec, 4e7, 17aec, 18. The

disappearance of characteristic one-proton singlet of NH quinolone

hydrogen at 12.21 ppme12.59 ppm conﬁrmed that a mono-

alkylation reaction had taken place for quinolones 3a, 8a, and 8b.

2.2. Antiproliferative activity

Nineteen novel compounds were tested for their potential

antiproliferative effect on the three tumor cell lines. In general they

exert low to modest effect. However, the majority of the tested

compounds were slightly soluble and, besides, the precipitates

produced from compound 6, 2b and 17b interfered with the MTT

assay, leading to false negative results; therefore these results

should be interpreted with caution. In general, quinolones were

more active compared to the carboxanilides, but no difference was

observed in the activity of their hydrochloride salts. Among car-

boxanilides compound 7 had the most pronounced effect. The most

active compounds, with the IC₅₀values in the micromolar range,

were quinolones 8a, 9b,10,11, 12, whereby the hydrochloride salt of

N,N-dimethylamino-propyl-substituted quionolone 14 was the

most active one, having the IC₅₀concentration at submicromolar

range (0.2e0.6 mM) (Fig. 2). Compound 14 was designed according

to our previous QSAR analysis as compound with potentially high

anticancer activity [25]. Experiments presented in this work

conﬁrmed previous predictions; compound 14 is the most active of

all compounds presented in this paper.

2.3. Analysis of the 3D structureeactivity relationship

Detailed chemometric analysis was performed in order to

elucidate the structural variables important for the anticancer ac-

tivity of the analyzed compounds and, possibly, to gain deeper

understanding of the molecular mechanism through which anti-

cancer activity is achieved.

Fig. 3. PCA scoring plot using: a) autoscaling data, b) raw data.

2.3.1. Principal component analysis and 3D-derived QSAR models

Principal component analysis (PCA) was performed on the

overall dataset of 19 compounds. The ﬁrst two principal compo-

nents explained 38% of variance in the case of autoscaled data and

72% in the case when the raw data was used. Distribution of the

compounds in the PC space (PC scores) was examined in order to

check similarity between the molecules and their possible

grouping (Fig. 3). PCA is also useful tool for analyzing variation

between the descriptors of the dataset. Since PCs are constructed in

a way that the ﬁrst few components describe majority of the vari-

ance among descriptors (X-matrix), their loadings describe de-

scriptors with the highest contribution to the overall variance in the

X-space (Supplement material Fig. S3).

Using antitumor activity against HCT 116 (colon carcinoma)

cells, 3D-derived QSAR models 1A and 1B were derived (Fig. S1a in

Fig. 2. The concentrationeresponse curve of the most active compound 14.

PG ¼ percentage of growth.

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

271

Supplement material and Fig. 4a, respectively). In case of model 1A,

autoscaling procedure was applied, while model 1B was derived

using raw data. 3D-derived QSAR models 2A and 2B were built

using antitumor activity against MCF-7 (breast carcinoma) cells

(Fig. S1b in Supplement material and Fig. 4b, respectively). Model

2A was built using autoscaling procedure, while model 2B was built

using raw data. Statistical properties of derived models are sum-

marized in Table 2.

omitted. Compound 8b tends to precipitate and it is hard to

determine its anticancer activity. In cases of HCT 116 and H460 cells

it was only possible to estimate its IC₅₀as larger than 100 mM, while

in case of MCF cell line it was measured with an error of 92%

(Table 1). Compound 14 was designed according to our previous

QSAR analysis [25] and it was show to be the most active of all

compounds presented in this paper. Differences in its chemical

structure comparing to the rest of the presented compounds were

recognized by statistical analysis. Principal component analysis

(PCA) performed on raw data revealed its isolated position in the

space of physico-chemical descriptors (Fig. 3b) and PLS analysis

revealed it as an outlier in the derived models. Beside compounds

14 and 8b, PCA performed on the autoscaled descriptors also clearly

distinguished compounds 7 and 9b from the rest of the compounds

in the space of physico-chemical descriptors (Fig. 3a). In order to

further improve quality of the models, compounds 7 and 9b were

also omitted from the training set. Predictive ability of such models

was largely improved (values in brackets in Table 2 and Fig. S2 in

Supplement material) comparing to the models built using 17

compounds.

Models 1A, 1B, 2A and 2B were derived using 17 compounds

since compounds 8b and 14 were identiﬁed as outliers and were

a)

8

7

9b

2.3.2. Molecular descriptors with the highest impact on the

antitumor activity

6

PCA loadings plot (Supplement material Fig. S3) revealed

following descriptors as the ones with the highest variation among

the studied compounds: W1eW3 (hydrophilic regions), V (vol-

ume), WN1 (H-bond acceptor descriptor, calculated with amide

group as the probe), S (surface), HSA (the sum of hydrophobic

surface areas), WO1 (H-bond donor descriptor, calculated with

carbonyl oxygen as the probe) and %FU4e%FU8 (the percentage of

unionized species calculated at pH 4, 5, 6, 7 and 8).

Partial Least Square (PLS) analysis provides information on

quantitative inﬂuence of each descriptor to the derived QSAR

models. In case of models derived using autoscaling procedure,

inﬂuence of each descriptor to the compound’s activity can be

estimated directly from its PLS coefﬁcient. In case of models derived

using raw data matrix, the real impact of a descriptor on the bio-

logical activity is given as the product of the descriptor’s value and

its PLS coefﬁcient. From autoscaling models 1A and 2A (Fig. 5)

descriptors related to the solubility properties (LgS10, L4LgS) and

hydrophobic regions (CD8eCD6, D8, D7) were identiﬁed as the

descriptors that have the highest positive inﬂuence on compound’s

anticancer activity. The same observation was experimentally

conﬁrmed since the major obstacle for testing anticancer activity of

presented compounds was their low solubility. The same models

identiﬁed descriptors related to the molecular ﬂexibility (FLEX,

FLEX RB, G) as descriptors that decrease anticancer activity of

investigated compounds. The later implies that decrease in com-

pound’s ﬂexibility should increase its anticancer activity.

Models obtained using raw data, 1B and 2B identiﬁed solubility

(L0lgS) and metabolic stability (MetStab) as descriptors that in-

crease anticancer activity of investigated compounds (Fig. 6).

Polarizability and dispersion forces (W1eW3), sum of hydrophobic

surfaces (HSA), molecular mass (MW) and surface (S) were deter-

mined as descriptors with negative impact on anticancer activity by

models 1B and 2B. Interestingly, while possibility of compounds to

accept H-bond (WN1) and percentage of unionized species at pH 7

(%FU7) are according to model 1B descriptors that improve anti-

cancer activity of the studied compounds, model 2B identiﬁed

them as descriptors that are negatively correlated with antitumor

activity. This might indicate that optimal antitumor activities

toward HCT 116 and MCF-7 cell lines are achieved through different

molecular mechanisms.

10

12

3a

5

11

8a

3b

5

15

4

3

2

17b

6

7

2a

18

2b

17a

2

3

4

5

6

7

8

pIC₅₀- experimental

b)

8

7

6

5

4

3

2

11

12

9b

4

3b

2b

8a

3a

6

5

10

7

18

17b

2a

15

17a

2

3

4

5

6

7

8

pIC₅₀- experimental

In our previous studies, accomplished for different datasets of

amides and quinolones [18,25], V (volume), S (surface) and HSA

Fig. 4. Predicted vs. experimental antitumor activity (expressed as pIC₅₀) in case of: a)

model 1B, b) model 2B.

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

272

Table 1

IC₅₀values (in

affect antitumor activity for of all compounds: those presented here

m

M) of compounds 2ae18.

as well as amidino-substituted amides and quinolones studied

earlier. These ﬁnding was also experimentally observed since low

solubility of investigated compounds often presents the largest

obstacle in obtaining active compounds. Thus, increase in com-

pound’s solubility should increase its anticancer activity.

a

Compounds

IC₅₀

(

mM)

Cell lines

HCT116

Scheme 1

MCF-7

H460

2a

2b

3a

3b

4

5

6

7

8a

8b

9b

10

11

12

14

15

17a

17b

18

>100

ꢁ100^b

30 ꢂ 9

22 ꢂ 4

ꢁ100

>100^b

76 ꢂ 23

49 ꢂ 31

ꢁ100

55 ꢂ 38^b

3. Molecular mechanism of action

>100

18 ꢂ 2

>100

As part of their molecular mechanism of action, some com-

pounds were evaluated for their DNA binding properties. In a

general manner, binding to DNA correlates with a stabilization of

the hybridization of the two strands of the helix that could be

visualized as part of the increase of the temperature that is required

to separate the two strands. DNA melting temperature studies are

commonly performed with this purpose as a quick and informative

tool for DNA binding. Calf thymus DNA (CT-DNA) was incubated

alone or with a ﬁxed concentration of each compound and the

absorbance was measured using a reference cuvette containing the

same buffer alone or with the same compound at the same con-

centration to only get an insight on the hyperchromicity associated

with DNA, but not with the compound itself. From this series, only

the quinolones 8b and 14 present a strong DNA binding propensity,

ꢁ100

24 ꢂ 24

>100^b

17 ꢂ 2

7 ꢂ 1

>100^b

17 ꢂ 2

7 ꢂ 1

>100^b

20 ꢂ 7

37 ꢂ 35

>100

50 ꢂ 46

10 ꢂ 3

16 ꢂ 1

2 ꢂ 1

5 ꢂ 0.4

18 ꢂ 3

13 ꢂ 1

9 ꢂ 3

5 ꢂ 1

4 ꢂ 0.2

3 ꢂ 0.2

0.2 ꢂ 0.002

14 ꢂ 0.7

ꢁ100

2 ꢂ 0.1

0.6 ꢂ 0.3

ꢁ100

6 ꢂ 2

0.3 ꢂ 0.06

40 ꢂ 24

>100

>100^b

>100

ꢁ100^b

>100

>100^b

>100

a

IC₅₀: the concentration that causes 50% growth inhibition.

The precipitates interfered with the MTT test.

b

as evidenced by increase of the temperature of 10 ^ꢀC (

DT_m) for a

drug/DNA ratio R of only 0.25, and up to 16.9 ^ꢀC and 19 ^ꢀC,

respectively, at R ¼ 1 (Table 3). By contrast, the quinolone iodide 9b

(sum of hydrophobic surfaces) were identiﬁed as descriptors that

are positively correlated with antitumor activity at MCF-7 cell line,

while according to this study (models 1B and 2B) they have

negative impact on antitumor activity. Average values of these

descriptors for the dataset compounds are signiﬁcantly lower in

presents much weaker DT_mvalue. Similarly, the quinolones 10 and

11 as well as the carboxianilide 7 only have low DNA binding

properties. Therefore, among the carboxianilides, only compound 7

presents some DNA binding propensity. Interestingly, this com-

pound was also the most active carboxanilide on cells as evidenced

in MTT assays (Table 1). Finally, compounds 2aeb, 3a, 4e6, 8a and

12 failed to protect the DNA helix from heat-induced destabiliza-

tion suggesting no DNA binding propensity (data not shown). In

terms of structure activity relationships, the addition of a methyl

group on the quinolone iodine 8b to give 9b decreases the efﬁcacy

3

ꢂ

case of the compounds presented in this paper (V ¼ 657.5 A ,

2

ꢂ

2

ꢂ

S ¼ 446.7 A , HSA ¼ 381.8 A ) comparing to the average values of

3

ꢂ

2

ꢂ

the compounds used in previous studies (V ¼ 759.3 A , S ¼ 517.9 A ,

2

ꢂ

HSA ¼ 438.0 A ). Apparently, increase of the size of the presented

compounds should increase their anticancer activities toward MCF-

7 cell line. Descriptors with the highest negative impact on anti-

tumor activity for the presented compounds (the same cell line,

MCF-7) are common with the descriptors found for the previous

two groups of compounds: W1eW2 (hydrophilic volumes

describing polarizability and dispersion forces), WO1 (H-bond

donor descriptor, calculated with carbonyl group as the probe),

WN1 (H-bond acceptor descriptor, calculated with amide group as

the probe). Further decrease of these descriptors should lead to

further increase of the compound’s anticancer activity against MCF-

7 cells. According to models derived using autoscaling procedure,

descriptors related to the solubility (L2LgS, L3LgS, L4LgS) positively

of DNA binding with the

and the fully relaxed plasmid DNA that appeared at 0.5

8b, but 20 M using 9b. However, this conclusion does not apply to

D

T_mvalue that drops from 16.9 to 5.9 ^ꢀC

M using

m

the protonated N,N-dimethylamino substituted quinolone 8a

which does not intercalate nor bind DNA whereas the protonated

N,N-dimethylamino substituted N-methyl-quinolone 10 present

some (even if weak) DNA binding properties and weakly

intercalates.

UV/visible spectra analyses of the absorbancy was measured

using the various tested compounds (20

mM) in the absence or

presence of increasing concentrations from 1 to 100

(top to bottom).

mM of CT-DNA

Comparison of the spectra in the absence (dashed lanes) and

presence of 100 M of CT-DNA (dotted lanes) evidenced strong

Table 2

Statistical properties of 3D-derived QSAR models.

m

hypochromic effect ranging from 23% to 58% upon binding to the

DNA (Fig. 7). The hypochromic effect is associated with a bath-

ochromic effect (red shifts) using the most efﬁcent DNA binding

compounds 8b, 9b and 14 as exempliﬁed by the increase of

maximum wavelength (Dl) of up to 6 nm at the upper CT-DNA

concentration (drug/DNA ration of 0.2).

Induced red shift associated with hypochromic effect suggests

interaction of the aromatic system of compounds 8b, 9b and 14

with the base pairs. Therefore the compounds were evaluated for

DNA intercalation properties using the properties of topoisomerase

I to differentially relax supercoiled DNA in the presence of DNA

intercalators [26]. All compounds were evaluated at the same

Model nO^a

nLV^bR²

SDEC^c

Q^2d

SDEP^e

1A

1B

2A

2B

17 (15)^f1 (5)^f0.77 (0.98)^f0.37 (0.10)^f0.58 (0.87)^f0.51 (0.29)^f

17 (15)^f7 (6)^f0.95 (0.98)^f0.17 (0.12)^f0.62 (0.84)^f0.48 (0.32)^f

17 (15)^f8 (6)^f0.99 (0.99)^f0.07 (0.07)^f0.69 (0.73)^f0.41 (0.39)^f

17 (15)^f3 (2)^f0.79 (0.76)^f0.33 (0.37)^f0.58 (0.58)^f0.47 (0.49)^f

a

Number of objects used to built the model.

Number of latent variables.

b

c

SDEC is standard deviation of error of calculation.

d

Q²is the cross-validated predictive performance and is given by

P

Q²¼ 1 ꢃ ð _{n ¼ 1}ðy_expðiÞꢃ y_predðiÞÞ²= _{n ¼ 1}ðy_expðiÞꢃ hy_expiÞ²Þ; where y_pred(i)corre-

sponds to the predicted and y_exp(i)to the experimentally determined inhibition,

pIC₅₀for the compound i, respectively.

n

e

SDEP is the standard deviation in cross-validated prediction and is given by

qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

P

n

SDEP ¼

ð

_{i ¼ 1}ðy_expðiÞꢃ y_predðiÞÞ²=nÞ.

concentration range from 1 to 50

was then adjusted for the strongest DNA intercalating compounds

to lower concentrations from 0.2 to 5 M (Fig. 8, bottom panel).

mM (Fig. 8) and the concentration

f

Calculations for the model in which two additional outliers (7, 9b) have been

expelled.

m

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

273

a)

0,04

0,03

0,02

0,01

0

CD6-CD8

D8

MetStab

LgS10

DRDRAC

D7

WN6

IW1-IW3

DRDRDR

DRDODO

-0,01

-0,02

-0,03

-0,04

FLEX RB

FLEX

G

Volsurf descriptors

b)

0,2

L4lgS

DD4

0,15

0,1

R

DRDODO

DRDRAC

IW1-IW2

0,05

0

-0,05

-0,1

-0,15

IW3

FLEX

G

DD1

FLEX RB

Volsurf descriptors

Fig. 5. PLS coefﬁcients of 3D-derived QSAR model: a) 1A, b) 2A. Descriptors with the highest impact on the activity are labeled; list and description of all 128 VolSurf þ descriptors is

given in the VolSurf þ manual [27].

From this experiment, three type of compounds could be classiﬁed

depending on their DNA base pair intercalation efﬁciency. As a ﬁrst

group, the carboxianilide 4, as well as the quinolones 3b and 8a do

not intercalate at all into the DNA helix. In a second group, the

quinolones 3a, 10e12 as well as the carboxanilide 7 evidence weak

DNA intercalating properties as visualized by fully relaxed plasmid

(data not shown). Compounds 7, 10, 11 and 14 induce a change in

the intensity of the positive (275 nm) and/or the negative (245 nm)

peaks of the CD spectra for normal B-DNA conformation, suggesting

modiﬁcations in the base stacking and the ellipticity of the DNA

helix.

Based on the absence of isosbestic points in UV/Visible titration

studies using those compounds (Fig. 7), it was attempted that

compounds 9b, 8b and 14 do not present a single binding mode.

The DNA relaxation assay is a functional test. It clearly evidenced

those drugs as potent and functional DNA intercalators. By contrast,

circular dichroism experiment, as a biophysical analysis, gives

structural information on the orientation of the compound rela-

tively to the DNA helix (groove binding, intercalation between

adjacent base pairs). Thus, the absence of ICD suggests that the

intercalation evidenced using the topoisomerase I-induced DNA

relaxation experiment may be associated with another mode of

binding such as interaction with the phosphate backbone using the

positively charges arms (quaternary ammonium substituent for 8b

and 9b or protonated N,N-dimethylamino-propyl for 14), inter-

fering with correct intercalation of the planar chromophore be-

tween adjacent base pairs that therefore could not result in a

negative ICD.

DNAs obtained using 20e50

mM of compounds. The last group

contains the quinolones 9b, 8b and 14 as efﬁcient DNA inter-

calators. This is in agreement with both the increase of the DNA

melting temperature (Table 3) and the red shifts of the peaks of

absorbance obtained in UV/Visible spectroscopy measurements

(Fig. 7). Particularly, compounds 8b and 14 are highly efﬁcient DNA

intercalators with fully relaxed circular DNA (Rel form, Fig. 8 bottom

panel) generated in the presence of only 0.5e1 mM of drug. Those

two compounds presented the highest DNA stabilization properties

with more than 10 ^ꢀC required for DNA denaturation from the

weaker evaluated drug/DNA ratio (R ¼ 0.25) to the highest one

(R ¼ 1) (Table 3).

Compounds 9b, 8b and 14 were therefore evaluated for poten-

tial modiﬁcations of the circular dichroism (CD) spectra for CT-DNA,

but none of those compounds clearly evidenced typical negative

induced CD (ICD) attempted for a DNA intercalating compound

ꢀ

274

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

a)

3

2

WN1

%FU7

1

L0lgS

0

-1

-2

-3

-4

-5

-6

-7

-8

D1

S

W1

W2

Volsurf descriptors

b)

0.5

MetStab

%FU7

0

-0.5

-1

D1

WN1

MW

HSA

S

V

W2

-1.5

-2

-2.5

-3

-3.5

-4

W1

Volsurf descriptors

Fig. 6. Products of the descriptors average value (calculated for the dataset used to built the model) and the associated PLS coefﬁcient of 3D-derived QSAR model: a) 1B and b) 2B.

Descriptors with the highest impact on the activity are labeled; list and description of all 128 VolSurf þ descriptors is given in the VolSurf þ manual [27].

In order to further evaluate the mode of binding to DNA of this

new series of compounds, we then used DNaseI footprinting assays

to assess if their potential sequence selectivity. Compounds 8b and

14 did not reveal any sequence-selective binding but non-selective

Table 3

binding inducing a global inhibition of DNaseI cleavage along the

DNA melting temperature for CT-DNA incubated with the indicated compounds at

the various drug/DNA ratio (R).

whole DNA sequence using 2e5

mM of compound (Supplement

material Fig. S4). Compound 9b gave similar global inhibition,

starting at much higher concentration (data not shown) in agree-

ment with a weaker DNA binding propensity as evidenced using

thermal melting temperature studies (Table 3).

Since DNA intercalation could result in an inhibition of top-

oisomerases DNA cleavage activities, we ﬁnally looked at top-

oisomerases inhibition. For topoisomerase I, topoisomerase I-

Drug/DNA ratio

D

T_m(^ꢀC)

7

8b

9b

10

11

14

R ¼ 0.25

R ¼ 0.5

R ¼ 1

10

12.8

16.9

10

12.7

19

1

3

2.9

5.9

0

3

2

P^a

a

P ¼ precipitation.

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

275

Fig. 7. UV/Visible spectra analyses of the binding to DNA of selected compounds. Compounds selected from thermal melting temperature studies (20

mM) were incubated in BPE

buffer with or without increasing concentrations of CT-DNA from 0.1 to 100 mM. Spectra were collected against a cuvette containing the same concentration of CT-DNA in the same

buffer to substract the absorbance of CT-DNA in the course of the acquisition process. Dashed lanes correspond to samples containing the drug without CT-DNA; plain lanes to the

addition of CT-DNA (0, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 2, 4, 6, 8, 10, 20, 40, 60, 80

mM, top to bottom); dotted lanes to the addition of 100

m

M of CT-DNA; H is the percentage of hypochromicity;

Dl is the wavelength shift at the indicated arrows where Dl

¼

l_peak

ꢃ

l_peak

.

(drug alone)

(100mM

CT-DNA)

induced DNA relaxation assays failed to evidence any accumulation

of nicked DNA that should co-migrate with the nicked form (Fig. 8).

None of the compounds evidenced an accumulation of the upper

band which intensity is similar to that obtained in the lanes con-

taining the DNA alone (“DNA”) or incubated with topoisomerase I

in the absence of any compounds (lanes “0” and “Topo I”), even at

the highest drug concentrations, suggesting that none of those

compounds are potent topoisomerase I poisoning drugs.

We then focussed on topoisomerase II as a target. Inhibition of

topoisomerase II is revealed by the generation of a linear band

when the compound is a poison of topoisomerase II, as obtained

using the reference drug etoposide used as a positive control

(Fig. 9). This latter effect is not visualized using those compounds

suggesting that none of them are topoisomerase II inhibitors. Only

compound 7 presents a small increase in the linear band using

binders. In the QSAR analysis presented in this paper, these de-

scriptors were also found to be positively correlated with com-

pound’s activity (Fig. 5) by models 1A and 2A. Further, integy

moments of compounds 8b (IW1 ¼ 0.09, IW2 ¼ 0.42, IW3 ¼ 1.21)

and 14 (IW1 ¼ 0.12, IW2 ¼ 0.48, IW3 ¼ 1.27), which are the only

compounds from this series that present a strong DNA binding

propensity, are signiﬁcantly higher than the average integy mo-

ments of the rest of the compounds presented in this paper

(IW1 ¼ 0.07, IW2 ¼ 0.21, IW3 ¼ 0.71). The same is partly true for 3D

pharmacophoric descriptors related to H-bonding: DRDRDO (8b:

0.00, 14: 21.7, average: 14.9), DRDRDR (8b: 27.4, 14: 22.2, average:

18.2), DRACDO (8b: 0.00, 14: 12.2, average: 16.4), DRDODO (8b:

0.00, 14: 0.00, average: 7.34), DRDRAC (8b:26.7, 14: 26.9, average:

20.2). Thus, previous and presented chemometric analyses are in

agreement and are supported by experimental results since com-

pounds that were experimentally found as DNA binders are

discriminated from the rest of the compounds by the same de-

scriptors in both analyses [18].

50

mM of compound.

The DNA binding studies clearly evidence the quinolones 8b and

14 as strong DNA intercalators, the N-methyl-quinolone 9b as a

medium intercalator whereas the quinolones 10, 11 and 12 as well

the carboxanilide 7 are much weaker DNA intercalators.

4. Conclusions

PCA analysis (Fig. 3) clearly distinguished compounds 14, 8b and

9b from the rest of the investigated compounds in the space of

molecular descriptors. In order to identify molecular properties

responsible for the differences in DNA binding properties of

investigated compounds, we compared their descriptors with the

descriptors found for the rest of investigated compounds (Table 4,

Fig. S5 in the supplement). The largest differences are between the

descriptors related to percentage of unionized species at pH 4e8 (%

FU4e8). The lower values of %FU4e8 descriptors enhance com-

pound’s DNA binding properties, which is understandable since

compound’s preference to exist in ionized (positively charged) form

should enhance its interaction with DNA.

Novel compounds from the benzothiophene and thienothio-

phene series with nitrogen-bearing substituents were synthesized

and tested on three tumor cell lines. The majority of the compounds

was slightly soluble and exerted low to modest antiproliferative

effect. All carboxanillides, except 7 were less active compared to the

quinolones prepared from them. The most active quinolones were

8a, 9b, 10e12 and 14. Compound 14 was the most active compound

from this series, having the IC₅₀concentrations z0.3 mM. Since this

compound was designed by previous QSAR analysis and predicted

as active compound it presents an excellent example of veriﬁcation

of chemometric analysis by laboratory methods.

Previous 3D-QSAR analysis [18] pointed to the integy moments,

which indicate concentration of hydrophilic regions on one part of

the molecule, and 3D pharmacophoric descriptors related to the

hydrogen bond formation potency as the descriptors that are

positively correlated with the biological activity. These descriptors,

combined with the large planar platform, are desirable for DNA

Detailed chemometric analysis was performed on the dataset

consisting of the investigated compounds. The 3D-derived QSAR

analysis and PCA enabled identiﬁcation of the descriptors important

for anticancer activity of investigated compounds. Solubility, metab-

olitic stability and the possibility of the compound to be ionized at pH

4e8 were found as molecular properties whose increase should lead

ꢀ

276

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

Topo I

14

4

11

12

Rel

topo

Sc

Topo I

8b

7

10

9b

Rel

topo

Sc

Topo I

12

3a

8a

3b

Rel

topo

Sc

Topo I

8b

14

Rel

topo

Sc

Fig. 8. Topoisomerase I-induced DNA relaxation. Supercoiled pUC19 plasmid was incubated or not (“0” lanes) with increasing concentrations of the indicated compounds at the

concentration speciﬁed on the top of each lanes ( M) prior to be subjected to relaxation using topoisomerase I (“Topo I”). DNA samples were separated on an agarose gel that was

m

post-stained using ethidium bromide to visualize the different circular DNA bands (Sc, supercoiled; topo, topoisomers; Rel, relaxed DNA). “DNA”, supercoiled plasmid untreated with

topoisomerase I).

to increase of anticancer activity of the investigated compounds.

Molecular ﬂexibility, polarizability and sum of hydrophobic surfaces

were recognized as the descriptors whose decrease should lead to

increase of anticancer activity of the investigated compounds. This

ﬁndings can be used as guidelines in designing new compounds with

increased anticancer activity, similarly as previous QSAR analysis was

used for designing of compound 14.

The DNA binding propensities of studied compounds (anilides

2aeb, 4e7 and quinolones 3aeb, 8a, 10e14) were investigated

using DNA melting temperature studies, UV/Visible spectropho-

tometry, CD spectrometry and inhibition of topoisomerase I and II

tests. Compounds 8b, 9b and 14 were identiﬁed as the compounds

with strong DNA binding propensity. The same compounds were

distinguished from the rest of the compounds by chemometric

tools (PCA and PLS analysis). The highest difference in their mo-

lecular properties comparing to the rest of the compounds were

found in their possibility to exist in ionized form at pH 4e8,

concentration of hydrophilic regions on one part of the molecule

and their hydrogen bond formation potency. These descriptors,

combined with the large planar platform, are desirable for DNA

binders and should be considered in future attempts to design

DNA binders.

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

277

Fig. 9. Topoisomerase II-induced poisoning effect. Supercoiled pUC19 plasmid was incubated or not (“DNA” lanes) with the indicated compounds at the concentration speciﬁed on

the top of each lanes prior to be subjected to DNA cleavage by topoisomerase II (“Topo II”). Etoposide (“Etop” lanes) was used as a positive control for topoisomerase II poisoning.

DNA samples were separated on an agarose gel containing ethidium bromide to visualize the different DNA forms (Nck, nicked DNA; Lin, linear DNA; Sc, supercoiled plasmid; Rel,

relaxed DNA).

5. Experimental

J ¼ 9.1 Hz, H_arom.); ¹³C NMR (150 MHz, DMSO-d₆) (

d ppm): 158.0,

142.2, 142.0, 141.4, 134.8, 134, 7, 133.9, 127.9, 126.5, 125.0, 124.4,

123.6, 119.2, 117.5, 116.6; elemental analysis calcd. (%) for

5.1. Chemistry

C₁₅H₈N₂O₃S: C 60.80, H 2.72, N 9.45; found C 61.08, H 2.52, N 9.60.

Melting points were recorded SMP11 Bibby and Büchi 535

apparatus and are uncorrected. IR spectra were recorded on FTIR-

ATR spectrophotometer. ¹H and ¹³C NMR spectra were recorded

on Varian Gemini 300 or Varian Gemini 600 spectrophotometers at

300, 600, 150 and 75 MHz, respectively. All NMR spectra were

measured in DMSO-d₆solutions using TMS as an internal standard.

Elemental analyses for carbon, hydrogen and nitrogen were per-

formed on a PerkineElmer 2400 elemental analyzer and a Perkine

Elmer, Series II, CHNS analyzer 2400. All compounds were routinely

checked by TLC with Merck silica gel 60F-254 glass plates. In pre-

parative 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 cut-off ﬁlter of wavelengths below 280 nm.

5.1.1.3. 2-Acetamido-6-oxo-5,6-dihydrothieno[3⁰,2⁰:4,5]thienyl[2,3-c]

quinolin 19. A solution of 17a (0.10 g, 0.29 mmol) in methanol

(60 ml) and toluene (60 ml) was irradiated for 10 h. The solution

was concentrated and the crude product was puriﬁed by column

chromatography with dichloromethaneemethanol mixture to

obtain 0.03 g (28%) of ivory solid; m.p. >300 ^ꢀC; IR

n

/cm^ꢃ1: 3358,

3112, 1647, 1593; ¹H NMR (300 MHz, DMSO-d₆) (

d

ppm): 12.00 (s,

1H, H_quinolone), 10.17 (s, 1H, H_amide), 8.86 (d, 1H, J ¼ 2.1 Hz, H_arom.),

8.04 (d, 1H, J ¼ 5.3 Hz, H_thiophene), 7.96 (d, 1H, J ¼ 5.4 Hz, H_thiophene),

7.75 (dd, 1H, J₁¼ 8.9 Hz, J₂¼ 2.1 Hz, H_arom.), 7.45 (d, 1H, J ¼ 8.9 Hz,

H

_arom.), 1.25 (s, 3H, CH₃); ¹³C NMR (75 MHz, DMSO-d₆) (

d ppm):

165.4 (s), 159.3 (s), 143.9 (s), 138.6 (s), 136.2 (s), 133.9 (s), 132.7 (s),

131.5 (d), 130.7 (d), 128.3 (d), 121.2 (d), 119.0 (s), 117.6 (d), 116.3 (s),

24.5 (q); elemental analysis calcd. (%) for C₁₅H₁₀N₂O₂S₂: C 57.31, H

3.21, N 8.91; found C 57.43, H 3.11, N 9.15.

5.1.1. General method for the synthesis of substituted 6-oxo-5,6-

dihydro [1]benzothieno[2,3-c]quinolines (3aeb) and 6-oxo-5,6-

dihydrothieno[3⁰,2⁰:4,5]thienyl[2,3-c]quinoline (19)

5.1.2. Preparation of N-(4⁰-aminophenyl)-3-chlorobenzo[b]

thiophene-2-carboxamide 4

A

solution of N-substituted 3-chlorobenzo[b]thiophene-2-

carboxamide derivatives (2aeb) in methanol/toluene was irradi-

ated at room temperature with a 400-W high pressure mercury

lamp for 2.5e10 h. The solution was concentrated and the obtained

solid was ﬁltered off.

A solution of 2b (1.11 g, 3.33 mmol) in THF (70 ml) and 10% PdeC

(0.56 g) was hydrogenated until the required quantity of H₂was

taken up. The solution was ﬁltered through Celite to remove cata-

lyst and the solvent was removed under reduced pressure. The

crude product was puriﬁed by recrystallization from ethanol to

5.1.1.1. 2-Acetamido-6-oxo-5,6-dihydro [1]benzothieno[2,3-c]quino-

line 3a. A solution of 2a (0.10 g, 0.29 mmol) in methanol (15 ml)

and toluene (20 ml) was irradiated for 2.5 h, 0.05 g (56%) of white

obtain 0.79 g (78%) of light yellow solid; m.p. 163e164 ^ꢀC; IR

cm^ꢃ1: 3309, 3190, 1628, 1596, 1533; ¹H NMR (300 MHz, DMSO-d₆)

ppm): 10.10 (s, 1H, H_amide), 8.15e8.12 (m, 1H, H_arom.), 7.94e7.91

n/

(d

solid was obtained; m.p. >300 ^ꢀC; IR

n

/cm^ꢃ1: 3263, 3111,1651, 1605,

ppm): 12.21 (s, 1H, H_quino-

(m, 1H, H_arom.), 7.62-7.59 (m, 2H, H_arom.), 7.36 (d, 2H, J ¼ 8.7 Hz,

1566; ¹H NMR (300 MHz, DMSO-d₆) (

d

H

_arom.), 6.57 (d, 2H, J ¼ 8.7 Hz, H_arom.), 5.05 (s, 2H, NH₂); ¹³C NMR

_lone), 10.23 (s, 1H, H_amide), 9.17 (d, 1H, J ¼ 1.7 Hz, H_arom.), 8.76 (d, 1H,

J ¼ 7.4 Hz, H_arom.), 8.29 (dd, 1H, J₁¼7.1 Hz, J₂¼ 1.7 Hz, H_arom.), 7.79e

7.67 (m, 3H, H_arom.), 7.49 (d, 1H, J ¼ 8.8 Hz, H_arom.), 2.14 (s, 3H, CH₃);

(75 MHz, DMSO-d₆) (d ppm): 158.0, 145.8, 136.5, 135.8, 132.9, 127.3,

127.2, 126.0, 123.4, 122.4, 121.9, 118.7, 113.7 (3C); elemental analysis

calcd. (%) for C₁₅H₁₁ClN₂OS: C 59.50, H 3.66, N 9.25; found C 59.62,

H 3.53, N 9.45.

¹³C NMR (75 MHz, DMSO-d₆) (

d ppm): 169.0, 158.0, 141.9, 136.0,

135.8, 135.0, 133.8, 133.1, 127.9, 126.3, 125.3, 124.8, 120.8, 117.7, 117.3,

112.9, 24.6; elemental analysis calcd. (%) for C₁₇H₁₂N₂O₂S: C 66.22,

H 3.92, N 9.08; found C 66.07, H 4.17, N 8.98.

5.1.3. Preparation of N-[4⁰-(N⁰,N⁰,N⁰-trimethylamino)phenyl]-3-

chlorobenzo[b]thiophene-2-carboxamide iodide 7

The suspension of 2c (1.00 g, 3.02 mmol) in a mixture of acetone

(120 ml) and diethyl ether (50 ml) was heated with methyl iodide

(0.59 ml, 9.48 mmol) for 18 h. The solution was concentrated and

the crude product was ﬁltered off and washed with diethyl ether to

5.1.1.2. 2-Nitro-6-oxo-5,6-dihydro

3b. A solution of 2b (0.10 g, 0.30 mmol) in methanol (15 ml) and

[1]benzothieno[2,3-c]quinoline

toluene (25 ml) was irradiated for 3 h, 0.02 g (24%) of brown solid

was obtained; m.p. >300 ^ꢀC; IR

n

/cm^ꢃ1: 2969, 2829, 1643, 1587,

obtain 0.98 g (69%) of white solid; m.p. 227e230 ^ꢀC; IR

n

/cm^ꢃ1

:

1532; ¹H NMR (600 MHz, DMSO-d₆) (

d

ppm): 12.78 (s, 1H, H_quino-

3396, 3003, 1651, 1610, 1537; ¹H NMR (300 MHz, DMSO-d₆) (

d

_lone), 9.44 (s, 1H, H_arom.), 8.77 (d, 1H, J ¼ 7.6 Hz, H_arom.), 8.43 (dd, 1H,

J₁¼ 9.1 Hz, J₂¼ 2.3 Hz, H_arom.), 8.32 (d, 1H, J ¼ 8.0 Hz, H_arom.), 7.80 (t,

1H, J ¼ 7.7 Hz, H_arom.), 7.73 (t, 1H, J ¼ 7.6 Hz, H_arom.), 7.70 (d, 1H,

ppm): 10.89 (s, 1H, H_amide), 8.21e8.18 (m, 1H, H_arom.), 8.02e7.91 (m,

5H, H_arom.), 7.68e7.64 (m, 2H, H_arom.), 3.62 (s, 9H, CH₃); ¹³C NMR

(75 MHz, DMSO-d₆) (d ppm): 159.5, 142.8, 139.3, 136.8, 135.6, 131.4,

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

278

5.1.6.1. 2-(N⁰,N⁰-dimethylamino)-5-methyl-6-oxo-5,6-dihydro

[1]

Table 4

The average values of the descriptors that differ the most between the DNA inter-

calators (compounds 14, 8b and 9b) and the rest of the compounds.

benzothieno[2,3-c]quinoline 9a. From 8a (0.27 g, 0.62 mmol) in

DMF (10 ml), sodium hydride (0.13 g, 5.34 mmol) and methyl iodide

(0.09 g, 0.61 mmol), after recrystallization from ethanol 0.06 g

Descriptor^a

Average value

for compounds

14, 8b and 9b

Average value

for the rest of

compounds presented

in this paper

Average value

for the previously

investigated

(20%) of light brown solid was obtained; m.p. >300 ^ꢀC; IR

n

/cm^ꢃ1

:

3029, 2885, 2516, 1618, 1564; ¹H NMR (300 MHz, DMSO-d₆) (

d

compounds

ppm): 8.72 (d, 1H, J ¼ 8.4 Hz, H_arom.), 8.24 (d, 1H, J ¼ 8.3 Hz, H_arom.),

3

ꢂ

V/A

749.9

498.2

1209.8

669.0

328.6

100.7

654.2

445.5

1158.7

715.0

383.6

139.0

316.8

379.0

72.2

80.9

88.9

92.8

91.3

759.3

517.9

1308.5

788.4

425.8

197.0

364.7

438.0

19.6

19.8

20.4

20.6

20.0

7.84 (s, 1H, H_arom.), 7.73e7.66 (m, 2H, H_arom.), 7.62 (d, 1H, J ¼ 9.4 Hz,

2

ꢂ

S/A

3

H

_arom.), 7.25 (d, 1H, J ¼ 9.2 Hz, H_arom.), 3.77 (s, 3H, CH₃), 3.09 (s, 6H,

ꢂ

W1/A

CH₃), ¹³C NMR (75 MHz, DMSO-d₆) (

d ppm): 158.2, 141.7, 141.3,

3

ꢂ

W2/A

3

ꢂ

W3/A

138.3, 135.8, 133.9, 132.8, 127.5, 126.1 (2C), 125.7, 124.2 (2C), 118.3,

117.7, 44.6 (2C), 30.3; elemental analysis calcd. (%) for C₁₈H₁₆N₂OS:

C 70.10, H 5.23, N 9.08; found C 69.96, H 5.47, N 9.32.

3

ꢂ

WO1/A

MW

342.4

445.2

2

ꢂ

HSA/A

%FU4/%

%FU5/%

%FU6/%

%FU7/%

%FU8/%

0.00017

0.00170

0.01699

0.16911

1.61728

5.1.6.2. 5-Methyl-2-(N⁰,N⁰,N⁰-trimethylamino)-6-oxo-5,6-dihydro [1]

benzothieno[2,3-c]quinoline iodide 9b. From 8b (0.27 g, 0.62 mmol)

in DMF (10 ml), sodium hydride (0.13 g, 5.34 mmol) and methyl

iodide (0.09 g, 0.61 mmol), after recrystallization from ethanol

a

V, volume; S, surface; W1eW3, hydrophilic volumes describing polarizability

and dispersion forces; WO1, H-bond donor descriptor calculated with carbonyl

oxygen as the probe; MW, molecular mass; HSA, sum of hydrophobic surfaces; %

FU4e%FU8, percentage of unionized species at pH 4e8.

0.06 g (20%) of white solid was obtained; m.p. 278e282 ^ꢀC; IR

n

/

cm^ꢃ1: 3439, 2925, 2827, 1581; ¹H NMR (300 MHz, DMSO-d₆) (

d

ppm): 8.95e8.92 (m, 2H, H_arom.), 8.37e8.31 (m, 2H, H_arom.), 8.01 (d,

1H, J ¼ 9.6 Hz, H_arom.), 7.80e7.73 (m, 2H, H_arom.), 3.88 (s, 3H, CH₃),

3.83 (s, 9H, CH₃); ¹³C NMR (75 MHz, DMSO-d₆) (

d ppm): 157.4,141.6,

127.8, 126.2, 123.6, 122.6, 121.3 (2C), 120.7 (2C), 120.2, 56.5 (3C);

elemental analysis calcd. (%) for C₁₈H₁₈ClN₂OS: C 45.73, H 3.84, N

5.93; found C 46.01, H 3.60, N 6.07.

141.5,138.5,135.6,134.6,133.9,127.9,126.3,126.2,124.2,121.3,118.1,

117.9, 115.6, 56.7 (3C), 30.3, elemental analysis calcd. (%) for

C₁₉H₁₉IN₂OS: C 50.67, H 4.25, N 6.22; found C 50.43, H 4.53, N 6.12.

5.1.6.3. 2-(N⁰-methylacetamido)-5-methyl-6-okso-5,6-dihidro

[1]

5.1.4. Preparation of 2-(N⁰,N⁰-dimethylamino)-6-oxo-5,6-dihydro

[1]benzothieno[2,3-c] quinoline hydrochloride 8a

benzothieno[2,3-c]quinoline 15. From 3a (0.45 g, 1.46 mmol) in DMF

(25 ml), sodium hydride (0.29 g, 12.09 mmol) and methyl iodide

(0.21 g, 1.45 mmol) the resulting solution was concentrated and the

crude product was puriﬁed by column chromatography with

dichloromethaneemethanol mixture to obtain 0.02 g (4%) of ivory

A solution of 6 (0.91 g, 2.48 mmol) in methanol (160 ml) and

toluene (160 ml) was irradiated for 3 h. The solution was concen-

trated and the crude product was puriﬁed by column chromatog-

raphy with dichloromethaneemethanol mixture to obtain 0.28 g

solid; m.p. >300 ^ꢀC; IR

(300 MHz, DMSO-d₆) (

n

d

/cm^ꢃ1: 3372, 3083, 1632, 1554; ¹H NMR

ppm): 9.01 (bs, 1H, H_arom.), 8.73 (s, 1H,

(35%) of brown solid; m.p. >300 ^ꢀC; IR

n

/cm^ꢃ1: 3392, 3224, 1652;

ppm): 12.35 (s, 1H, H_quinolone), 8.95

¹H NMR (300 MHz, DMSO-d₆) (

d

H

_arom.), 8.28e8.25 (m, 1H, H_arom.), 7.84 (d, 1H, J ¼ 8.9 Hz, H_arom.),

7.70e7.69 (m, 3H, H_arom.), 3.84 (s, 3H, CH₃), 3.39 (bs, 3H, OCH₃), 1.85

(s, 3H, CH₃); ¹³C NMR (75 MHz, DMSO-d₆) (

ppm): 168.6, 157.8,

(m, 1H, H_arom.), 8.77 (s, 1H, H_arom.), 8.31e8.28 (m, 1H, H_arom.), 7.73e

7.62 (m, 4H, H_arom.), 3.72 (bs, 1H, NH^þ), 3.22 (s, 6H, CH₃); ¹³C NMR

d

(75 MHz, DMSO-d₆) (d ppm): 158.0, 142.2, 141.8, 136.0, 135.8, 135.7,

142.0, 139.8 (2C), 135.6 (2C), 134.7, 128.4, 128.0, 126.5 (2C), 124.4,

122.6, 119.4, 117.7, 30.5 (2C), 22.9; elemental analysis calcd. (%) for

133.7, 128.1, 126.5 (2C), 126.0, 124.7, 119.4, 118.5, 118.1, 44.4 (2C);

elemental analysis calcd. (%) for C₁₇H₁₅ClN₂OS: C 61.72, H 4.57, N

8.47; found C 62.06, H 4.26, N 8.70.

C₁₉H₁₆N₂O₂S: C 67.84, H 4.79, N 8.33; found C 68.14, H 4.64, N 8.18.

5.1.7. Preparation of 2-Amino-6-okso-5,6-dihidro [1]benzothieno

[2,3-c]quinoline 11

5.1.5. Preparation of 2-(N⁰,N⁰,N⁰-trimethylamino)-6-oxo-5,6-

dihydro [1]benzothieno[2,3-c]quinoline iodide 8b

To a solution of 3a (0.80 g, 2.60 mmol) in methanol (55 ml) was

added saturated solution of sodium hydroxid (1.52 g, 38.00 mmol).

After reﬂuxing for 24 h the solution was concentrated and the crude

product was puriﬁed by column chromatography with dichloro-

methaneemethanol mixture to obtain 0.04 g (6%) of light yellow

A solution of 7 (0.20 g, 0.42 mmol) in ethanol (80 ml) and water

(80 ml) was irradiated for 2 h, 0.15 g (81%) of white solid was ob-

tained; m.p. >300 ^ꢀC; IR

(300 MHz, DMSO-d₆) (

(m, 1H, H_arom.), 8.89 (d, 1H, J ¼ 2.5 Hz, H_arom.), 8.35e8.32 (m, 1H,

n

/cm^ꢃ1: 3473, 3125, 3006, 1643; ¹H NMR

d

ppm): 12.59 (s, 1H, H_quinolone), 8.97e8.94

solid; m.p. > 300 ^ꢀC; IR

n

/cm^ꢃ1: 3325, 2974, 2842, 1649, 1596, 1509;

ppm): 11.95 (s, 1H, H_quinolone),

H

_arom.), 8.24 (dd, 1H, J₁¼ 9.3 Hz, J₂¼ 2.5 Hz, H_arom.), 7.79e7.76 (m,

¹H NMR (300 MHz, DMSO-d₆) (

d

2H, H_arom.),7.73 (d, 1H, J ¼ 9.3 Hz, H_arom.), 3.79 (s, 9H, CH₃); ¹³C NMR

8.87e8.84 (m, 1H, H_arom.), 8.27e8.24 (m, 1H, H_arom.), 7.98 (d, 1H,

J ¼ 2.0 Hz, H_arom.), 7.72e7.65 (m, 2H, H_arom.), 7.29 (d, 1H, J ¼ 8.7 Hz,

(75 MHz, DMSO-d₆) (d ppm): 158.3, 141.9, 141.7, 138.3, 135.5, 135.1,

134.3, 128.3, 126.7, 126.5, 124.7, 121.6, 118.5, 117.4, 115.6, 57.3 (3C);

elemental analysis calcd. (%) for C₁₈H₁₇IN₂OS: C 49.55, H 3.93, N

6.42; found C 49.30, H 4.19, N 6.15.

H

_arom.), 6.92 (dd, 1H, J₁¼ 8.7 Hz, J₂¼ 2.1 Hz, H_arom.), 5.25 (s, 2H,

NH₂); ¹³C NMR (75 MHz, DMSO-d₆) (

d ppm): 157.8 (s), 141.3 (s),

139.2 (s), 138.0 (s), 134.2 (s), 134.0 (s), 133.1 (s), 127.9 (d), 126.2 (d),

125.0 (d), 124.6 (d), 122.2(d), 118.7 (d), 117.3 (d), 115.3 (s); elemental

analysis calcd. (%) for C₁₅H₁₀N₂OS: C 67.65, H 3.78, N 10.52; found C

67.95, H 3.92, N 10.37.

5.1.6. General method for the synthesis of 5-methyl-6-oxo-5,6-

dihydro [1]benzothieno[2,3-c]quinolines (9aeb, 15)

To a cold solution of corresponding quinolin-6-ones in anhy-

drous DMF, sodium hydride as 60e65% oil dispersion was added in

two portions. After stirring for 15 min, methyl iodide was added to

the solution and the reaction mixture was stirred for 2 h at 0e5 ^ꢀC

and 24 h at room temperature. The products were ﬁltered off and

recrystallizated from ethanol.

5.1.8. 2-Acetamido-5-[3-(N⁰,N⁰-dimethylamino)proply]-6-oxo-5,6-

dihydro [1]benzothieno[2,3-c]quinoline 13

To a cold solution of 3a (0.30 g, 0.97 mmol) in anhydrous DMF

(18 ml), sodium hydride (0.19 g, 8.09 mmol) as 60e65% oil

dispersion was added in two portions. After stirring for 15 min a

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

279

solution of 3-(dimethylamino)propyl chloride hydrochloride

(0.40 g, 2.53 mmol) in DMF (18 ml) was added to the solution and

the reaction mixture was stirred for 2 h at 0e5 ^ꢀC and 24 h at room

temperature. The solid was separated by ﬁltration and the solvent

was removed under reduced pressure. The crude product was

recrystallized from ethanol to obtain 0.85 g (22%) of light yellow

119.0, 116.7, 113.5, 54.5, 42.5 (2C), 39.2, 24.6, 23.2; elemental anal-

ysis calcd. (%) for C₂₂H₂₄ClN₃O₂S: C 61.46, H 5.63, N 9.77; found C

61.54, H 5.51, N 10.04.

5.2. Antiproliferative evaluation assay

solid; m.p. 244e246 ^ꢀC; IR

1583; ¹H NMR (600 MHz, DMSO-d₆) (

9.25 (s, 1H, H_arom.), 8.80e8.78 (m, 1H, H_arom.), 8.29e8.27 (m, 1H,

n

/cm^ꢃ1: 3325, 2949, 2768, 1688, 1619,

ppm): 10.32 (s, 1H, H_amide),

The experiments were carried out on three human cell lines,

which are derived from three cancer types. The following cell lines

were used: HCT 116 (colon carcinoma), H 460 (lung carcinoma) and

MCF-7 (breast carcinoma). The cells were cultured as monolayers

and maintained in Dulbecco’s modiﬁed Eagle medium (DMEM)

d

H

_arom.), 7.89 (d, 1H, J ¼ 9.2 Hz, H_arom.), 7.79 (d, 1H, J ¼ 9.4 Hz, H_arom.),

7.74e7.70 (m, 2H, H_arom.), 4.42 (t, 2H, J ¼ 7.4 Hz, CH₂), 2.38 (t, 2H,

J ¼ 6.7 Hz, CH₂), 2.18 (s, 6H, CH₃), 2.16 (s, 3H, CH₃), 1.89e1.80 (m, 2H,

supplemented with 10% fetal bovine serum (FBS), 2 mM

L-gluta-

CH₂); ¹³C NMR (150 MHz, DMSO-d₆) (

d

ppm): 168.6, 156.9, 141.6,

mine, 100 U/ml penicillin and 100 g/ml streptomycin in a hu-

m

135.2, 134.6, 134.4, 133.1, 132.2, 127.4, 125.9, 125.0, 124.3, 120.3,

118.5, 116.2, 113.1, 56.3, 45.1, 40.6, 25.5, 24.1; elemental analysis

calcd. (%) for C₂₂H₂₃N₃O₂S: C 67.15, H 5.98, N 10.68; found C 67.35, H

5.63, N 10.78.

midiﬁed atmosphere with 5% CO₂at 37 ^ꢀC. The cell lines were

inoculated onto a series of standard 96-well microtiter plates on

day 0, at 3 ꢄ 10⁴cells/mL (HCT 116, H 460) to 5 ꢄ10⁴cells/mL (MCF-

7), depending on the doubling times of a speciﬁc cell line. After

24 h, cells were treated with indicated compounds and concen-

trations where highest DMSO concentration at <1%. Following a

72 h incubation, an MTT assay (Sigma) was performed per standard

protocol, as described previously [15e18]. The absorbance (A) of the

microtiter plate was measured on a microplate reader at 570 nm

where absorbance is directly proportional to the number of living,

metabolically active cells. Each treatment was performed in tech-

nical quadruplicates, with at least 2 biological replicates. The re-

sults are expressed as IC₅₀, which is the concentration necessary for

50% of inhibition. The IC₅₀values for each compound are calculated

from concentrationeresponse curves using linear regression anal-

ysis by ﬁtting the test concentrations that give percentage of

growth (PG) values above and below the reference value (i.e. 50%).

5.1.9. General method for the synthesis of and 6-oxo-5,6-dihydro

[1]benzothieno[2,3-c]quinoline hydrochlorides (10,12,14)

A stirred suspension of compounds 9a, 11 and 13 in absolute

ethanol was saturated with HCl(g). After 20 h of stirring resulting

product was ﬁltered off and washed with diethyl ether to obtain 6-

oxo-5,6-dihydro [1]benzothieno[2,3-c]quinoline hydrochlorides.

5.1.9.1. 2-(N⁰,N⁰-dimethylamino)-5-methyl-6-oxo-5,6-dihydro

[1]

benzothieno[2,3-c]quinoline hydrochloride 10. Compound 10 was

prepared using above described method from 9a (0.04 g,

0.14 mmol) in absolute ethanol (10 ml) to obtain 0.03 g (67%) of

brown solid; m.p. >300 ^ꢀC; IR

n

/cm^ꢃ1: 3400, 2548, 2445, 1633,

ppm): 9.04e8.92 (m, 2H,

1587; ¹H NMR (300 MHz, DMSO-d₆) (

d

H

_arom.), 8.19 (d, 1H, J ¼ 8.6 Hz, H_arom.), 8.00 (bs, 1H, H_arom.), 7.80 (d,

5.3. Computational analysis

1H, J ¼ 9.2 Hz, H_arom.), 7.76e7.59 (m, 2H, H_arom.), 3.88 (bs, 1H, NH^þ),

3.74 (s, 3H, CH₃), 3.22 (s, 6H, CH₃); ¹³C NMR (75 MHz, DMSO-d₆) (

d

Measured anticancer activity of the presented compounds

against MCF-7 (breast carcinoma), HCT116 (colon carcinoma) and

H460 (lung carcinoma) were used to build 3D-derived QSAR

models. Negative logarithmic value of the concentration that cau-

ses 50% growth inhibition of the cell lines (pIC₅₀) was used as

measure of compound’s anticancer activity in the 3D-derived QSAR

models. For the poorly active compounds whose IC₅₀values were

not explicitly measured, but just estimated as ꢁ100, pIC₅₀was set

to 3.301. Volsurf þ program [19] was applied for generation of

molecular descriptors for each compound (for detailed description

of all 128 VolSurf þ descriptors see the VolSurf þ manual [27]). Grid

ppm): 157.5, 141.8 (2C), 138.8, 135.2, 134.3, 133.3, 128.0, 126.4 (2C),

126.2, 124.5 (2C), 118.6, 118.3, 45.4 (2C), 30.4; elemental analysis

calcd. (%) for C₁₈H₁₇ClN₂OS: C 62.69, H 4.97, N 8.12; found C 62.89, H

4.82, N 8.24.

5.1.9.2. 2-Amino-6-okso-5,6-dihidro [1]benzothieno[2,3-c]quinoline

hydrochloride 12. Compound 12 was prepared using above

described method from 11 (0.08 g, 0.30 mmol) in absolute ethanol

(15 ml) to obtain 0.05 g (53%) of light brown solid; m.p. >300 ^ꢀC; IR

n

/cm^ꢃ1: 3331, 2964, 2858, 1652; ¹H NMR (600 MHz, DMSO-d₆) (

d

ppm): 12.41 (s, 1H, H_quinolone), 10.17 (bs, 3H, NH3^þ) 8.90e8.88 (m,

2H, H_arom.), 8.32e8.29 (m, 1H, H_arom.), 7.73e7.71 (M, 2H, H_arom.),

7.63 (d, 1H, J ¼ 8.7 Hz, H_arom.), 7.52 (dd, 1H, J₁¼ 8.7 Hz, J₂¼ 2.0 Hz,

spacing was set to 0.5 A and the following probes were used: H2O

ꢂ

(the water molecule), O (sp²carbonyl oxygen atom), N1 (neutral NH

group (e.g. amide)) and DRY (the hydrophobic probe).

H

_arom.); ¹³C NMR (75 MHz, DMSO-d₆) (

d

ppm): 158.6 (s), 140.8 (s),

For each cell line, 3D-derived QSAR models were derived using

either autoscaled variables (models labeled as A) or raw data

(models labeled as B). Partial Least Square (PLS) analysis was

applied to ﬁnd the relationship between the 3D structure-based

molecular descriptors and measured antitumor activity. In order

to identify the descriptors with the highest (positive or negative)

impact on biological activity of the compounds, PLS coefﬁcients of

the 3D-derived QSAR models were analyzed using the procedure

described in previous work [18,24,25]. In case of the models derived

using autoscaling procedure, the inﬂuence of each descriptor to the

QSAR model was estimated from its PLS coefﬁcient. In case of

models derived using raw data (1B and 2B), product between

average value of the descriptor (calculated for the dataset used to

build the model) and its associated PLS coefﬁcient was calculated

for each descriptor.

139.3 (s), 137.6 (s), 134.7 (s), 133.9 (s), 131.2 (s), 128.3 (d), 126.9 (d),

125.3 (d), 124.0 (d), 121.8 (d), 119.2 (d), 117.1 (d), 116.0 (s); elemental

analysis calcd. (%) for C₁₅H₁₁ClN₂OS: C 59.50, H 3.66, N 9.25; found C

59.15, H 3.81, N 9.45.

5.1.9.3. 2-Acetamido-5-[3-(N⁰,N⁰-dimethylamino)propyl]-6-oxo-5,6-

dihydro

[1]benzothieno[2,3-c]quinoline

hydrochloride

14.

Compound 14 was prepared using above described method from 13

(0.06 g, 0.15 mmol) in absolute ethanol (10 ml) to obtain 0.06 g

(94%) of white solid; m.p. 222e224 ^ꢀC; IR /cm^ꢃ1: 3373, 3066, 2660,

n

1628, 1587; ¹H NMR (300 MHz, DMSO-d₆) (

H

d ppm): 10.56 (bs, 2H,

_amide, NH^þ), 9.12 (s, 1H, H_arom.), 8.77e8.75 (m, 1H, H_arom.), 8.21e

8.18 (m, 1H, H_arom.), 7.92 (d, 1H, J ¼ 8.5 Hz, H_arom.), 7.76 (d, 1H,

J ¼ 9.2 Hz, H_arom.), 7.66e7.63 (m, 2H, H_arom.), 4.44 (bs, 2H, CH₂),

3.25e3.18 (m, 2H, CH₂), 2.73 (s, 3H, CH₃), 2.48 (bs, 2H, CH₂), 2.14 (s,

Principal component analysis (PCA) was performed in order to

determine distribution of the compounds in the space of the mo-

lecular descriptors (PCA scores plot) and to ﬁnd contribution of

6H, CH₃); ¹³C NMR (75 MHz, DMSO-d₆) (

d

ppm): 169.2, 157.6, 142.1,

135.5, 135.4, 135.2, 133.2, 132.3, 128.0, 126.4, 125.7, 124.6, 120.7,

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

280

each descriptor to the ﬁrst two principal components (PCA loadings

plot). Usefulness of PCA in 3D-derived QSAR analysis has already

been proved in several cases [18,25,28,29].

bilateral Hubert Curien partnership between Croatian and French

institutions (Cogito program) as the Egide project N^ꢀ24765PH. M.-

H.D.-C. thanks the Association pour la Recherche sur le Cancer

(ARC) Ligue Nationale Contre le Cancer (Comité du Pas-de-Calais,

Septentrion), the Fonds Européen de Développement Régional

(FEDER, European Community) and the Région Nord/Pas-de-Calais

for grants (M-H.D.-C.), the Institut pour la Recherche sur le Cancer

de Lille (IRCL) for Ph.D and post-doctoral fellowships (R.N.) and

technicien expertise (S.D.).

5.4. DNA binding properties and topoisomerase assays

5.4.1. Spectroscopic analyses of DNA binding

CT-DNA (Sigma Aldrich, France) was dissolved in water and

dialyzed overnight prior to use. The analyzed compounds were

prepared as 10 mM stock solutions in DMSO, aliquoted and stored

at ꢃ20 ^ꢀC and then freshly diluted in the appropriate aqueous

buffer.

Appendix A. Supplementary material

For UV/visible spectroscopy, tested compounds (20 mM) were

diluted 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 con-

Supplementary material related to this article can be found at

http://dx.doi.org/10.1016/j.ejmech.2013.11.010.

centrations of CT-DNA (from 10 to 100

m

M with 10

mM steps and

References

then from 100 to 200 M with steps of 20

m

mM of base pairs). The UV/

Visible spectra were recorded from 230 nm to 530 nm in a quartz

cuvette of 10 mm pathlength using an Uvikon XL spectrophotom-

eter and referenced against a cuvette containing DNA at identical

concentration.

For melting temperature studies, the absorbency of DNA was

measured at 260 nm in quartz cells using an Uvikon 943 spectro-

photometer thermostated with a Neslab RTE111 cryostat every min

over a range of 20e100 ^ꢀC with an increment of 1 ^ꢀC per min The

Tm values were obtained from the midpoint of the hyperchromic

transition obtained from ﬁrst-derivative plots. The variation of

[1] P. Diana, A. Martorana, P. Barraja, A. Montalbano, G. Dattolo, G. Cirrincione,

F. Dall ’Acqua, A. Salvador, D. Vedaldi, G. Basso, G. Viola, Isoindolo[2,1-a ]qui-

noxaline derivatives, novel potent antitumor agents with dual inhibition of

tubulin polymerization and topoisomerase I, J. Med. Chem. 51 (2008) 2387e

2399.

[2] L. Wilson, M.A. Jordan, Microtubule dynamics: taking aim at a moving target,

Chem. Biol. 2 (1995) 569e573.

[3] I.V. Magedov, M. Manpadi, M.A. Ogasawara, A.S. Dhawan, S. Rogelj, S. Van

Slambrouck, W.F. Steelant, N.M. Evdokimov, P.Y. Uglinskii, E.M. Elias, E.J. Knee,

P. Tongwa, M.Y. Antipin, A. Kornienko, Structural simpli ﬁ cation of bioactive

natural products with multicomponent synthesis. 2. Antiproliferative and

antitubulin activities of pyrano[3,2-c ] pyridones and pyrano[3,2-c ]quinolones,

J. Med. Chem. 51 (2008) 2561e2570.

[4] A.A. Sagardoy, M.J. Gil, R. Villar, M.J. Vinas, A. Arrazola, I. Encio, V. Martinez-

Merino, Benzo[b ]thiophene-6-carboxamide 1,1-dioxides: inhibitors of human

cancer cell growth at nanomolar concentrations, Bioorg. Med. Chem. 18

(2010) 5701e5707.

[5] B. Letafat, S. Emami, N. Mohammadhosseini, M.A. Faramarzi, N. Samadi,

A. Sha ﬁ ee, A. Foroumadi, Synthesis and antibacterial activity of new N-[2-

(thiophen-3-yl)ethyl]piperazinyl Quinolones, Chem. Pharm. Bull. 55 (2007)

894e898.

[6] S.W. Wang, S.L. Pan, Y.C. Huang, J.H. Guh, P.C. Chiang, D.Y. Huang, S.C. Kuo,

K.H. Lee, C.M. Teng, CHM-1, a novel synthetic quinolone with potent and

selective antimitotic antitumor activity against human hepatocellular carci-

noma in vitro and in vivo, Mol. Cancer Ther. 7 (2008) 350e360.

[7] H. Hoang, D.V. LaBarbera, K.A. Mohammed, C.M. Ireland, E.B. Skibo, Synthesis

and biological evaluation of imidazoquinoxalinones, imidazole analogues of

pyrrolo- iminoquinone marine natural products, J. Med. Chem. 50 (2007)

4561e4571.

[8] R. Romagnoli, P.G. Baraldi, M.D. Carrion, O. Cruz-Lopez, M. Tolomeo,

S. Grimaudo, A. Di Cristina, M.R. Pipitone, J. Balzarini, A. Brancale, E. Hamel,

Substituted 2-(3⁰,4⁰,5⁰- trimethoxybenzoyl)-benzo[b]thiophene derivatives as

potent tubulin polymerization inhibitors, Bioorg. Med. Chem. 18 (2010)

5114e5122.

[9] M.J.R.P. Queiroz, R.C. Calhelha, L.A. Vale-Silva, E. Pinto, M.S.J. Nascimento,

Synthesis of novel 3-(aryl)benzothieno[2,3-c ]pyran-1-ones from Sonogashira

products and intramolecular cyclization: antitumoral activity evaluation, Eur.

J. Med. Chem. 44 (2009) 1893e1899.

melting temperature (

DT_m) were measured by substracting the

melting temperature measurement of 20

mM of CT-DNA or poly(-

dAdT)₂incubated alone (control T_m) to that obtained with DNA

incubated with increasing concentrations of the various tested

compounds (R ¼ drug/base pair ratio from 0.25 to 1) in 1 mL BPE

buffer (

D

T_mvalues ¼ T_{m[Drug þ DNA]}ꢃ T_{m[DNA alone]}).

For circular dichroism, the various drugs (50

bated with or without (control) a ﬁxed or increasing concentrations

of CT-DNA (from 0.1 to 100 M) in BPE. The CD spectra were

mM) were incu-

m

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) as

described previously [30].

5.4.2. DNase I footprinting

DNase I footprinting experiments were conducted essentially as

previously described [31]. After an exposition of the gels to storage

screen for the appropriated delay at room temperature, the results

were collected using a Molecular Dynamics STORM 860.

[10] A.P. Ferreirra, J.L.F. da Silva, M.P. Duarte, M.F.M. da Piedade, M.P. Robalo,

S.G. Harjivan, C. Marzano, V. Gandin, M.M. Marques, Synthesis and charac-

terization of new organometallic Benzo[b ]thiophene derivatives with poten-

tial antitumor properties, Organometalics 28 (2009) 5412e5423.

[11] C. Mésangeau, M. Fraise, P. Delagrange, D.H. Caignard, J.A. Boutin, P. Berthelot,

S. Yous, Preparation and pharmacological evaluation of a novel series of 2-

(phenylthio)benzo[b ]thiophenes as selective MT₂receptor ligands, Eur. J.

Med. Chem. 46 (2011) 1835e1840.

5.4.3. Topoisomerase-mediated DNA relaxation and topoisomerase

II cleavage assay

Topoisomerase I-mediated DNA relaxation experiment was

performed as previously described [26] using supercoiled pUC19

plasmid DNA and human topoisomerase I (Topogen, USA). The re-

actions were stopped upon addition of SDS and proteinase K prior

to gel electrophoresis on a 1% agarose gel for 2 h at 120 V in TBE

buffer. Gels were then stained in a bath containing ethidium bro-

mide before being washed and photographed under UV light.

Topoisomerase II DNA cleavage assays were performed essentially

as described with etoposide as a positive control for Topo II

poisoning effect.

[12] S. Dadiboyena, Recent advances in the synthesis of raloxifene: a selective

estrogen receptor modulator, Eur. J. Med. Chem. 51 (2012) 17e34.

ꢀ

[13] J. Dogan Koruznjak, N. Slade, B. Zamola, K. Pavelic, G. Karminski-Zamola,

Synthesis, photochemical synthesis and antitumor evaluation of novel de-

rivatives of thieno[3⁰,2⁰: 4, 5] thieno[2,3-c ]quinolones, Chem. Pharm. Bull. 50

(2002) 656e660.

ꢁ

ꢀ

[14] J. Dogan Koruznjak, M. Grdisa, N. Slade, B. Zamola, K. Pavelic, G. Karminski-

Zamola, Novel derivatives of benzo[b ]thieno[2,3-c ]quinolones: synthesis,

photochemical synthesis and antitumor evaluation, J. Med. Chem. 45 (2003)

4516e4524.

Acknowledgments

ꢁ

ꢀ

[15] I. Jarak, M. Kralj, L. Suman, G. Pavlovic, J. Dogan, I. Piantanida, M. Zinic,

ꢀ

K. Pavelic, G. Karminski-Zamola, Novel cyano- and N -isopropylamidino-

substituted derivatives of Benzo[b ]thiophene- 2-carboxanilides and benzo[b]

thieno[2,3-c ]quinolones: synthesis, photochemical synthesis, crystal structure

determination and antitumor evaluation. Part 2, J. Med. Chem. 48 (2005)

2346e2360.

We greatly appreciate the ﬁnancial support of the Croatian

Ministry of Science Education and Sports (projects 125-0982464-

1356, 098-098-1192344-2860, 098-098-2464-2514) and the

ꢀ

M. Aleksic et al. / European Journal of Medicinal Chemistry 71 (2014) 267e281

281

ꢁ

ꢀ

[16] I. Jarak, M. Kralj, I. Piantanida, L. Suman, M. Zinic, K. Pavelic, G. Karminski-Zamola,

Novel cyano- and amidino-substituted derivatives of thieno[2, 3-b ]- and thieno

[3, 2- b]thiophene-2-carboxanilides and Thieno[3⁰,2⁰:4,5]thieno- and Thieno

[2⁰,3⁰:4, 5]thieno[2,3- c]quinolones: synthesis, photochemical synthesis, DNA

binding and antitumor evaluation, Bioorg. Med. Chem. 14 (2006) 2859e2868.

molecules based on similarity of their interaction ﬁ elds and the log P and log D

values: application to auxins, J. Chem. Inf. Model 43 (2003) 1532e1541.

ꢀ

ꢁ

ꢀ

ꢁ

[24] I. Vujasinovic, A. Paravic-Radicevic, K. Mlinaric-Majerski, K. Brajsa, B. Bertosa,

Synthesis and biological validation of novel pyrazole derivatives with anti-

cancer activity guided by 3D-QSAR analysis, Bioorg. Med. Chem. 20 (2012)

2101e2110.

ꢀ

[17] K. Ester, M. Hranjec, I. Piantanida, I. Caleta, I. Jarak, K. Pavelic, M. Kralj,

G. Karminski-Zamola, Novel derivatives of Pyridylbenzo[b]thiophene-2-

carboxamides and benzo[b]thieno[2,3-c]naphthyridin-2-ones: minor struc-

tural variations Provoke major differences of antitumor action mechanisms,

J. Med. Chem. 52 (2009) 2482e2492.

ꢁ

ꢀ

[25] B. Bertosa, M. Aleksic, G. Karminiski-Zamola, S. Tomi&cacute, QSAR analysis of

antitumor active amides and quinolones from thiophene series, Int. J. Pharm.

394 (2010) 106e114.

[26] P. Peixoto, C. Bailly, M.-H. David-Cordonnier, Topoisomerase I-mediated DNA

relaxation as a tool to study intercalation of small molecules into supercoiled

DNA, Methods Mol. Biol. 613 (2010) 235e256.

ꢀ

ꢁ

[18] M. Aleksic, B. Bertosa, R. Nhili, L. Uzelac, I. Jarak, S. Depauw, M.H. David

ꢀ

Cordonnier, M. Kralj, S. Tomic, G. Karminski-Zamola, Novel substituted ben-

zothiophene and thienothiophene carboxanilides and quinolones: synthesis,

photochemical synthesis, DNA binding properties, antitumor evaluation and

3D-derived QSAR analysis, J. Med. Chem. 55 (2012) 5044e5060.

[19] G. Cruciani, P. Crivori, P.-A. Carrupt, B. Testa, Molecular ﬁ elds in quantitative

structure-permeation relationships: the VolSurf approach, J. Mol. Struct.

Theochem. 503 (2000) 17e30.

[20] C.G. Fortuna, V. Barresi, G. Berellini, G. Musumarra, Design and synthesis of

trans 2-(furan-2-yl)vinyl heteroaromatic iodides with antitumour activity,

Bioorg. Med. Chem. 16 (2008) 4150e4159.

[27] G. Cruciani, M. Pastor, W. Guba, VolSurf: a new tool for pharmacokinetic

optimization of lead compounds, Eur. J. Pharm. Sci. 11 (2000) S29eS39.

ꢀ

ꢁ

ꢀ

[28] S. Tomic, B. Bertosa, B. Kojic-Prodic, I. Kolosvary, Stereoselectivity of Bur-

kholderia cepacia lipase towards secondary alcohols: molecular modelling and

3D QSAR approach, Tetrahedron: Asymmetry 15 (2004) 1163e1172.

ꢀ

ꢁ

[29] S. Tomic, B. Bertosa, T. Wang, R.C. Wade, COMBINE analysis of the speci ﬁ city of

binding of Ras proteins to their effectors, Proteins Struct. Funct. Bioinform. 67

(2) (2007) 435e447.

[30] T. Lemster, U. Pindur, S. Depauw, G. Lenglet, C. Dassi, M.-H. David-Cor-

donnier, 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. 44 (2009)

3235e3252.

[31] P. Peixoto, Y. Liu, S. Depauw, M.-P. Hildebrand, D.W. Boykin, C. Bailly,

W.D. Wilson, M.-H. David-Cordonnier, 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. 36 (2008) 3341e

3353.

[21] X.-M. Zhuang, J.-H. Xiao, J.-T. Li, Z.-Q. Zhang, J.-X. Ruan, A simpli ﬁ ed model to

predict P -glycoprotein interacting drugs from 3D molecular interaction ﬁ eld,

Int. J. Pharm. 309 (2006) 109e114.

ꢁ

ꢀ

ꢁ

ꢀ

[22] I. Caleta, M. Kralj, M. Marjanovic, B. Bertosa, S. Tomic, G. Pavlovic, K. Pavelc,

G. Karminski-Zamola, Novel cyano- and amidinobenzothiazole derivatives:

synthesis, antitumor evaluation, and X-ray and quantitative structur-

e ꢃ activity relationship (QSAR) analysis, J. Med. Chem. 52 (2009) 1744e1756.

ꢁ

ꢀ

[23] B. Bertosa, B. Kojic-Prodic, M. Ramek, S. Piperaki, A. Tsantili-Kakoulidou,

R. Wade, S. Tomi&cacute, A new approach to predict the biological activity of

Article Doi

Anilides and quinolones with nitrogen-bearing substituents from benzothiophene and thienothiophene series: Synthesis, photochemical synthesis, cytostatic evaluation, 3D-derived QSAR analysis and DNA-binding properties

DOI: 10.1016/j.ejmech.2013.11.010

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.ejmech.2013.11.010

Full text of DOI:10.1016/j.ejmech.2013.11.010

Products guided by the article

R&D Labs maybe for 162090-86-0

Relevant to this article

Hot Product