A study of the properties, reactivity and anticancer activity of novel N-methylated-3-thiazolyl or 3-thienyl carbazoles and their Pd(II) and Pt(II) complexes

Article abstract of DOI:10.1016/j.jinorgbio.2018.03.008

The synthesis and characterization of two hybrid N-methylated carbazole derivatives containing a thiazolyl or a thienyl ring is reported. The thiazolyl derivative has been also characterised by X-ray diffraction analysis. The study of its reactivity in front of [MCl₂(dmso)₂] (M = Pd or Pt) or Na₂[PdCl₄] in methanol has allowed us to isolate and characterize its complexes. However, for the thienyl analogue, the formation of any Pd(II) or Pt(II) complex was not detected, indicating that it is less prone to bind to the M(II) ions than its thiazolyl analogue. Density Functional Theory (DFT) and Time-Dependent Density Functional Theory (TD-DFT) calculations have also been carried out in order to rationalize the influence of the nature of the thiazolyl or thienyl group on the electronic delocalization. Molecular mechanics calculations show that the free rotation of the thiazolyl in relation to the carbazole requires a greater energy income than for its thienyl analogue. Studies of the cytotoxic activity of the new compounds on colon (HCT116) and breast (MDA-MB231 and MCF7) cancer cell lines show that the thiazolyl carbazole ligand and its Pt(II) complex are the most active agents of the series and in the MCF7 line their potency is higher than that of cisplatin. In the non-tumoral human skin fibroblast BJ cell line, all the compounds were less toxic than cisplatin. Their potential ability to modify the electrophoretic mobility of pBluescript SK⁺ plasmid DNA and to act as inhibitors of Topoisomerases I and IIα or cathepsin B has also been investigated.

Full text of DOI:10.1016/j.jinorgbio.2018.03.008

Accepted Manuscript
A study of the properties, reactivity and antitumor activity of
novel N-methylated-3-thiazolyl or 3-thienyl carbazoles and their
Pd(II) and Pt(II) complexes
M. Reig, R. Bosque, M. Font-Bardía, C. Calvis, R. Messeguer, L.
Baldomà, J. Badía, D. Velasco, C. López
PII:
S0162-0134(17)30799-7
doi:10.1016/j.jinorgbio.2018.03.008
JIB 10459
DOI:
Reference:
To appear in:
Journal of Inorganic Biochemistry
Received date:
Revised date:
Accepted date:
14 November 2017
14 February 2018
12 March 2018
Please cite this article as: M. Reig, R. Bosque, M. Font-Bardía, C. Calvis, R. Messeguer,
L. Baldomà, J. Badía, D. Velasco, C. López , A study of the properties, reactivity and
antitumor activity of novel N-methylated-3-thiazolyl or 3-thienyl carbazoles and their
Pd(II) and Pt(II) complexes. The address for the corresponding author was captured
as affiliation for all authors. Please check if appropriate. Jib(2017), doi:10.1016/
j.jinorgbio.2018.03.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.

ACCEPTED MANUSCRIPT
A study of the properties, reactivity and antitumor activity of novel N-
methylated-3-thiazolyl or 3-thienyl carbazoles and their Pd(II) and Pt(II)
complexes
a
b
c
d
d
e
e
M. Reig, R. Bosque, M. Font-Bardía, C. Calvis, R. Messeguer, L. Baldomà, J. Badía, D.
a
b
Velasco and C. López*
a
2
Grup de Materials Orgànics, Institut de Nanociència i Nanotecnologia (IN UB), Secció de Química
Orgànica, Departament de Química Inorgànica i Orgànica, Facultat de Química, Martí i Franquès 1-11, E-
08028 Barcelona, Spain
b
Secció de Química Inorgànica, Departament de Química Inorgànica i Orgànica, Facultat de Química,
Martí i Franquès 1-11, E-08028 Barcelona, Spain
c
Unitat de Difracció de Raigs-X, Centres Científics i Tecnològics (CCiT), Universitat de Barcelona, Solé i
Sabaris 1–3, E-08028 Barcelona, Spain
d
Biomed Division LEITAT Technological Center, Parc Científic de Barcelona, Edifici Hèlix, Baldiri i
Reixach 13-21, E-08028 Barcelona, Spain
e
Departament de Bioquímica i Biologia Molecular, Facultat de Farmàcia, Av. Joan XXIII s/n, E-08028
Barcelona, Spain
Keywords
Carbazole – Pd(II)-based complexes – Pt(II)-based complexes – Cancer / Antitumor activity
Abstract
The synthesis and characterization of two hybrid N-methylated carbazole derivatives
containing a thiazolyl or a thienyl ring on position 3 is reported. The thiazolyl derivative has been
also characterised by X-Ray diffraction. The study of its reactivity in front of [MCl (dmso) ] (M =
2
2
Pd or Pt) or Na [PdCl ] in methanol has allowed us to isolate and characterize its complexes.
2
4
However, when the reactions were performed using the thienyl analogue, no evidences of the
formation of any Pd(II) or Pt(II) complex were detected. Thus indicating that the latter is less
prone to bind to the M(II) ions than its thiazolyl analogue. DFT and TD-DFT calculations have
also been carried out in order to: a) rationalize the influence of the nature of the thiazolyl or
1

ACCEPTED MANUSCRIPT
thienyl group on the electronic delocalization and b) assign the bands detected in their electronic
spectra. Molecular mechanics calculations show that the free rotation of the thiazolyl in relation
to the carbazole requires a greater energy income than for its thienyl analogue. Comparative
studies of the cytotoxic activity of the new ligands and the metal complexes on colon (HCT116)
and two breast (MDA-MB231 and MCF7) cancer cell lines are also reported. In order to get
additional information on mode of action of the new products, comparative studies on their
+
potential ability to: a) modify the electrophoretic mobility of pBluescript SK plasmid DNA and b)
act as inhibitors of Topoisomerases I and II or cathepsin B have also been carried out.
1. Introduction
Cancer is among the leading causes of morbidity and death that unfortunately affects
millions of persons worldwide (i.e. more than one million each year in USA [1]). The American
Cancer Society estimates an incidence of ca. 1.7 million new cases for 2017 and more than 0.6
million deaths [2] mainly produced by colorectal, breast, lung and ovarian cancers [2-4]). Every
cancer type needs a specific treatment protocol that usually involves chemotherapy (CT) [5],
radiotherapy and/or surgery. The development of new antitumor drugs with improved activities
and lower side effects than those used nowadays in CT is still one of the main challenges of
current research in medicinal chemistry. Among the variety of strategies used nowadays in drug
discovery [6], those with greater expectations are based on: a) natural products and/ or b) new
synthetic products with several bioactive arrays (i.e. by incorporation of an additional bioactive
unit in the backbones of commercially available pharmaceuticals or known drugs and commonly
known as “molecular hybridization approach”) [6].
On the other hand, it is well-known that heterocycles and their derivatives are one of the
most important types of organic compounds due to their outstanding physical and photo-optical
properties, their rich reactivity, their utility as ligands in Coordination and Organometallic
Chemistry and their multiple applications in a variety of fields, including medicinal chemistry [7-
9]. Heterocyclic cores are present in huge range of natural and marketed antimicrobial, anti-
2

ACCEPTED MANUSCRIPT
inflammatory, antiviral, anticancer, antihypertensive, antimalarial, anti-HIV, antidepressant,
antihelmintic drugs, among others. Their relevance in new drugs design and development is
undeniable [7]. For instance, among all the anticancer drugs approved by FDA during the late
five years ca. 75% have heterocyclic arrays with N and / or S atoms or polycyclic aromatic
compounds with heterocyclic fragments [8-10].
Carbazole (Fig. 1), thiazole and thiophene are probably three of the most important
scaffolds in drug design and discovery [11-13]. These units are present in diverse bioactive
synthetic -and even in naturally occurring products. For instance, Ellipticine and Glybomine-C
(
Fig. 1) isolated from plants, are potent cytotoxic agents in several cancer cell lines [14] and the
discovery that N-alkylation of Ellipticine enhanced inhibition growth activity has stimulated the
interest on N-substituted carbazoles [11,14]. The number of potent cytotoxic (substituted and /or
anellated) carbazoles reported in the last 2 years has grown exponentially and according to a
recent review published by Caruso et al. [11d]: “Carbazoles are promising scenarios for breast
cancer treatments”.
Fig. 1. Carbazole and two naturally occurring carbazole derivatives (Ellipticine and Glybomine-B and C)
with potent cytotoxic activities in front of several cancer cell lines.
In addition to carbazoles, and in a lesser extent, thiazole and thiophene derivatives are
gaining increasing interest as “central cores” or as “pendant” groups in drug engineering and
specially as promising platforms or “templates” to build up new and more efficient antitumor
agents that could overcome or at least reduce the main problems (i.e. drug resistance, toxicity,
or other undesirable side effects) associated to drugs used currently. Cis-[PtCl (NH ) ] (cisplatin)
2
3 2
and Doxorubicin [4,15,16] are two CT agents used in cancer treatments that may generate
3

ACCEPTED MANUSCRIPT
severe-side effects (i.e. nephrotoxicity, neurotoxicity, the increase of blood pressure, severe
nausea, vomiting, or diarrhoea produced by cisplatin or severe heart problems (cardiomyopathy)
associated to doxorubicin [15,16]).
Despite of the interest in electronics and materials science arisen by hybrid carbazole /
thienyl derivatives [17-20] and the potential synergic effect of the presence of two bioactive
arrays [i.e. the N-substituted carbazole and a thiazole (or a thienyl) unit] in the same molecule
that could be relevant in drug design [21] and in photodynamic therapy (PTD) [22], studies on
their biological activities are scarce. Moreover, it is well-known that heterocycles are valuable
ligands in Coordination and Organometallic Chemistry [8] and their binding to a transition metal
m+
ion (M ) commonly affects their properties and their biological and / or catalytic activities. For
instance, Pd(II) and Pt(II) complexes with heterocyclic ligands (i.e. pyrazoles, indoles) showing
greater cytotoxic activity than the free ligands have been reported [23]. However, parallel studies
on hybrid carbazole / thiazole or thiophene derivatives have not been investigated yet.
Therefore, there is a lack of information on their coordination ability and the effect produced by
the binding of these ions on their properties and especially on their cytotoxic activity
.
In this paper, we present two new carbazole derivatives: 9-methyl-3-(2-thiazolyl)-9H-
carbazole (1a) and 9-methyl-3-(2-thienyl)-9H-carbazole (1b) shown in Scheme 1, a study of their
reactivity in front of Pd(II) and Pt(II), their spectroscopic properties and the antitumor activity of
the free ligands and the new Pd(II) and Pt(II) complexes.
2
2
2
. Experimental
.1. Chemistry
.1.1. Materials and methods
[
MCl (dmso) ] (M = Pd or Pt) and 3-iodo-9H-carbazole were prepared as described
2 2
previously [24-26], and the remaining reagents were obtained from commercial sources and
used as received. The success of the synthesis of the Pt(II) compound 3a is strongly dependent
on the quality of the methanol, the presence of water produces the formation of metallic
4

ACCEPTED MANUSCRIPT
platinum, other undesirable minor by-products and a significant decrease in the yield. Thus, the
use of high quality MeOH (HPLC grade) is required. The remaining solvents used were dried
and distilled before use [27]. During the preparation of the complexes (2a and 3a), the reaction
flask was protected from the light with aluminium foil. Elemental analysis were carried out at the
Centres Científics i Tecnològics (CCiT, Univ. Barcelona) with an Eager 1108 microanalyzer.
+
Mass spectra (ESI ) were performed at the Servei d’Espectrometria de Masses (Univ. de
Barcelona) using a LC/MSD-TOF Agilent Technologies instrument. UV-vis. spectra of CH Cl
2
2
solutions of the free ligands (1a and 1b) and complexes 2a and 3a were recorded at 298 K with
a Varian Cary UV-Vis-NIR 500E spectrometer and their emission spectra were obtained on a
PTI fluorimeter equipped with a 220B lamp power supply, a 815 photomultiplier detection system
and a Felix 32 software at 298 K in CH Cl solutions. 1,4-Bis(5-phenyl-2-oxazolyl)benzene
2
2
(
POPOP) dissolved in cyclohexane was used as a standard for the fluorescence quantum yield
1
13
1
determination (exc = 300 nm, ФPOPOP =0.93). H and C{ H}-NMR spectra were recorded at 298
K in acetone-d for the precursor (3-iodo-9-methyl-9H-carbazole) or in CDCl (in the remaining
6
3
1
13
1
cases) with a Varian Mercury 400 MHz and a Bruker 400 MHz Avance III [for H and C{ H}]
195
1
and a Bruker 250 MHz and a Bruker 400 Avance III HD (for Pt{ H}) spectrometers. Chemical
1
13
shifts are given in δ values (ppm) using the solvent peaks as internal references ( H and C)
195
1
and H PtCl in D O ( Pt{ H}) and coupling constants (J) are given in Hz. Abbreviations used for
2
6
2
the multiplicities are as follows: s (singlet), d (doublet), dd (doublet of doublets), t (triplet), q
1
(
quadruplet) and m (multiplet). The atom numbering system used in the assignment of H and
13
1
C{ H}-NMR data is shown in Chart 1.
Chart 1. Atom labelling scheme for ligands 1a and 1b.
5

ACCEPTED MANUSCRIPT
2.1.2. Synthesis of ligands 1a and 1b
-3
2
.1.2.1. Synthesis of the precursor 3-iodo-9-methyl-9H-carbazole. NaH (240 mg, 6.00  10 mol,
-
60% dispersion in mineral oil) was added to a solution of 3-iodo-9H-carbazole (1.60 g, 5.46  10
3
mol) in anhydrous DMF (10 mL) under nitrogen atmosphere. The solution was stirred at room
-3
temperature for 30 minutes. Then, iodomethane (374 μL, 6.00  10 mol) was added and the
mixture was stirred at room temperature for 30 minutes and then treated with water. The
aqueous layer was extracted with CH Cl and the organic layer was dried over Na SO , filtered
2
2
2
4
off and the solvent was distilled off under reduced pressure. The crude was purified by flash
column chromatography using a mixture of hexane and ethyl acetate (20 : 1 v/v) as the eluent to
1
give 3-iodo-9-methyl-9H-carbazole (1.45 g, 86%). H NMR (400 MHz, acetone-d )  (ppm): 8.49
6
4
4
3
5
3
4
2
(
d, J = 1.7, 1H, H ), 8.17 (d, J = 7.6, 1H, H ), 7.74 (dd, J = 8.6, J = 1.7, 1H, H ),
H-H
H-H
H-H
H-H
3
8
7
3
1
7
.55 (d, J = 8.2, 1H, H ), 7.53-7.48 (m, 1H, H ), 7.42 (d, J = 8.6, 1H, H ), 7.26-7.22 (m,
H-H
H-H
6
+
1
H, H ), 3.91 (s, NMe, 3H). CI-MS (m/z): calc. for C H IN (M + H) 308.0, found: 308.0.
13
11
2.1.2.2. Synthesis of 9-methyl-3-(2-thiazolyl)-9H-carbazole (1a). A mixture of 9-methyl-3-iodo-
-3
-3
9
H-carbazole (1.24 g, 4.04  10 mol), 2-(tributylstannyl)thiazole (1.81 g, 4.84  10 mol) and
-3
Pd(PPh ) (231 mg, 0.20  10 mol) in anhydrous DMF (10 mL) was heated to 100 °C under a
3
4
nitrogen atmosphere for 20 h. Then, the reaction mixture was cooled down to room temperature,
treated with water and the product was extracted with dichloromethane. The organic layer was
dried over Na SO , filtered off and the solvent was distilled off under reduced pressure. The
2
4
crude was purified by flash column chromatography using a mixture of hexane and CH Cl (5 : 1
2
2
1
v/v) as the eluent to give compound 1a (420 mg, 39%). H NMR (400 MHz, CDCl )  (ppm): 8.73
3
4
4
3
5
3
4
2
(
d, J = 1.7, 1H, H ), 8.16 (d, J = 8.0, 1H, H ), 8.10 (dd, J = 8.6 , J = 1.7 Hz, 1H, H ),
H-H H-H H-H H-H
3
4’
7
3
1
3
7
.87 (d, J = 3.3, 1H, H ), 7.54-7.50 (m, 1H, H ), 7.44 (d, J = 8.6, 1H, H ), 7.43 (d, J
=
H-H
H-H
H-H
8
6
5’
13
8
.2, 1H, H ), 7.31-7.27 (m, 2H, H and H ), 3.89 (s, 3H, NMe). C NMR (100 MHz, CDCl ) 
3
2
’
4’
9a
8a
7
3
2
(
(
ppm): 170.0 (C ), 143.4 (C ), 142.2 (C ), 141.7 (C ), 126.5 (C ), 125.1 (C ), 124.9 (C ), 123.3
4a
5a
5
6
4
5’
1
8
C ), 123.0 (C ), 120.8 (C ), 119.7 (C ), 119.0 (C ), 117.8 (C ), 108.9 (2C, C and C ), 29.4 (N-
6

ACCEPTED MANUSCRIPT
+
CH ). HRMS (ESI-MS) (m/z): calc. for C H N S (M + H) : 265.0794, found: 265.0796.
3
16 13
2
Elemental Anal. (%). Calc. for C H N S (MW = 264.34). C, 72.70; H, 4.58; N, 10.60 and S,
16
12
2
12.13; found: C, 72.65; H, 4.65; N, 10.53; and S, 11.86.
2.1.2.3. Synthesis of 9-methyl-3-(2-thienyl)-9H-carbazole (1b). A mixture of 9-methyl-3-iodo-9H-
-3
-3
carbazole (771 mg, 2.51  10 mol), 2-(tributylstannyl)thiophene (1.12 g, 3.00  10 mol) and
-3
Pd(PPh ) (139 mg, 0.12  10 mol) in anhydrous DMF (10 mL) was heated to 100 °C under a
3
4
nitrogen atmosphere for 24 h. Then, the reaction mixture was cooled down to room temperature,
treated with water and the product was extracted with dichloromethane. The organic layer was
dried over Na SO , filtered off and the solvent was distilled off under reduced pressure. The
2
4
crude was purified by flash column chromatography using a mixture of hexane and
1
dichloromethane (9 : 1 v/v) as the eluent to give compound 1b (343 mg, 52%). H NMR (400
4
4
3
5
3
MHz, CDCl )  (ppm): 8.32 (d, J = 1.8, 1H, H ), 8.14 (d, J = 7.7, 1H, H ), 7.75 (dd, J
=
=
3
H-H
H-H
H-H
4
2
7
3
8
3
8
.5 Hz, J = 1.8, 1H, H ), 7.52-7.48 (m, 1H, H ), 7.41 (d, J = 8.1, 1H, H ), 7.40 (d, J
H-H H-H
H-H
1
3
4
3’
5’
6
8
.5 Hz, 1H, H ), 7.35 (dd, J = 3.6, J = 1.0, 1H, H ), 7.28-7.24 (m, 2H, H and H ), 7.11
H-H
H-H
3
3
4’
13
(
dd, J = 5.1, J
= 3.6, 1H, H ), 3.87 (s, 3H, NMe). C NMR (100 MHz, CDCl )  (ppm):
H-H
H-H
3
2’
8a
9a
4’
7
3
2
5’
1
46.0 (C ), 141.6 (C ), 140.7 (C ), 128.1 (C ), 126.2 (C ), 125.9 (C ), 124.5 (C ), 123.8 (C ),
4a
5a
3’
5
6
4
1
8
1
23.3 (C ), 122.9 (C ), 122.2 (C ), 120.6 (C ), 119.3 (C ), 118.0 (C ), 108.9 (C or C ), 108.8
1
8
+
(
C or C ), 29.4 (NMe). HRMS (ESI-MS) (m/z): calc. for C H NS (M + H) 264.0841, found:
17 14
2
64.0843; elemental Anal (%). Calc. for C H NS (MW = 263.34). C, 77.53; H, 4.98; N, 5.32 and
17 13
S, 12.70; found: C, 72.45; H, 5.04; N, 5.25; and S, 12.61.
2.1.3. Preparation of the complexes 2a and 3a
2.1.3.1. Synthesis of compound 2a. This compound was obtained using two alternative
procedures that differ in the nature of the starting Pd(II) complex used as reagent:
PdCl (dmso) ] or Na [PdCl ] {methods a) and b), respectively}. Method b) allows the isolation of
[
2
2
2
4
compound 2a with a higher yield (and at lower temperatures) than using method a).
7

ACCEPTED MANUSCRIPT
-4
Method a) [PdCl (dmso) ] (63 mg, 1.9  10 mol) was treated with 30 mL of methanol,
2
2
-4
refluxed until complete dissolution and filtered out. Then, 50 mg (1.9  10 mol) of carbazole 1a,
were added to the hot filtrate and the mixture was refluxed for 1 h. After this period, the resulting
solution was allowed to cool down to room temperature and the solid formed was collected by
filtration, air-dried and later on dried in vacuum for 2 days. (Yield: 38 mg, 57 %).
-5
Method b) A solution containing Na [PdCl ] (28 mg, 9.5  10 mol) and 20 mL of
2
4
-4
methanol was added to another one formed by ligand 1a (50 mg, 1.9  10 mol) and 5 mL of
methanol. The resulting mixture was stirred for 24 h at 298 K. After this period, the solid formed
1
was collected by filtration and dried as in Method a) (Yield: 55 mg, 82 %). H NMR (400 MHz,
4
4
3
4
2
CDCl )  (ppm): 9.22 (d, J = 1.8, 2H, 2H ), 8.47 (dd, J = 8.5, J = 1.8, 2H, 2H ), 8.19 (d,
3
H-H
H-H
H-H
3
5
3
4’
1
7
8
5’
JH-H = 7.7, 2H, 2H ), 8.08 (d, J = 3.7, 2H, 2H ), 7.58-7.44 (m, 8H, 2H , 2H , 2H and 2H ),
H-H
3
6
7
.32 (t, J = 7.7, 2H, 2H ), 3.93 (s, 6H, 2NMe). Elemental Anal. (%). Calc. for C H Cl N PdS
H-H 32 24 2 4 2
(
MW = 706.01). C, 54.44; H, 3.43; N, 7.94 and S, 9.08; found: C, 54.50; H, 3.50; N, 8.03 and S,
8.85.
-4
2
.1.3.2. Synthesis of compound 3a. cis-[PtCl (dmso) ] (80 mg, 1.9  10 mol) was suspended in
2
2
30 mL of methanol, until complete dissolution. Then, the hot solution was filtered out and the
-4
filtrate was poured into a methanol solution (5 mL) of ligand 1a (50 mg, 1.9  10 mol). The
reaction flask was protected from light with aluminium foil and the mixture was refluxed for 1 h
and filtered. Then, the filtrate was afterwards concentrated to dryness on a rotary evaporator and
the residue was dried in vacuum for 24 h. After this period, the solid was dissolved in the
minimum amount of CH Cl (ca. 15 mL) and passed through a short SiO column (5.0 cm  1.5
2
2
2
cm). Elution with CH Cl released a pale yellowish band that was collected and concentrated to
2
2
195
1
dryness on a rotary evaporator giving 3a (yield: 69 mg, 60 %). Pt{ H}-NMR data (54 MHz,
1
4
4
CDCl )  (ppm): -2983 (s). H NMR-data (400 MHz, CDCl )  (ppm): 9.22 (d, J = 1.8, 1H, H ),
3
3
H-H
3
4
2
3
5
3
8
.47 (dd, J = 8.5, J = 1.8, 1H, H ), 8.20 (d, J = 7.7, 1H, H ), 8.08 (d, J = 3.7, 1H,
H-H
H-H
H-H
H-H
4’
1
7
3
8
3
5’
H ), 7.58-7.54 (m, 2H, H and H ), 7.47 (d, J
= 8.2, 1H, H ), 7.46 (d, J
= 3.7, 1H, H ),
H-H
H-H
8

ACCEPTED MANUSCRIPT
6
7
.34-7.30 (m, 1H, H ), 3.94 (s, 3H, NMe), 3.37 (s, 6H, dmso). ESI-MS (m/z): calc. for
+
C H Cl N OPtS (M + H) 608.0, found: 608.0. Elemental Anal. (%). Calc. for:
18
19
2
2
2
C H Cl N OPtS (MW = 608.46): C, 35.53; H, 2.98; N, 4.60 and S, 10.54; found C, 35.59; H,
18
18
2
2
2
3.05; N, 4.63 and S, 10.60.
2
.1.3.3. Synthesis of the two isomers of [PtCl (1a)(dmso)] {trans-(3a) and cis-(4a)}. NaAcO (16
2
-4
mg, 1.9  10 mol) was dissolved in 5 mL of methanol at 298 K and then added dropwise to a
-4
-4
mixture formed by carbazole 1a (50 mg, 1.9  10 mol), cis-[PtCl (dmso) ] (80 mg, 1.9  10
2
2
mol) and 25 mL of toluene. The flask was protected from the light with aluminium foil and
refluxed for 3 days. After this period the deep brown solution was filtered through a Celite pad,
and the filtrate was concentrated on a rotary evaporator. The dark residue was dried in vacuum
for 24 h, dissolved in CH Cl (ca. 30 mL) and finally passed through a short (5.0 cm  1.5 cm)
2
2
SiO column. Elution with CH Cl released a wide pale yellow band that was collected in portions
2
2
2
(
ca. 25 mL / each). The first collected fractions ca. 120 mL gave after concentration 11 mg of 3a;
while the remaining subsequent fractions eluted (ca. 200 mL) gave, after concentration a solid
195
1
(
26 mg) containing isomers 3a and 4a in a ca. equimolar ratio. %). Pt{ H}-NMR data (54 MHz,
CDCl , see also Fig. S1, A)  (ppm): -2980 (s) (trans- isomer, 3a) and -2932 (s) (cis- isomer, 4a);
3
1
4
4
H NMR data (400 MHz, CDCl )  (ppm): (see also Fig. S1, B): 9.22 (d, J = 1.8, 1H, H of 3a),
3
H-H
3
4
2
3
5
3
8
.47 (dd, J = 8.5 , J = 1.8 , 1H, H of 3a), 8.20 (d, J = 7.7, 1H, H of 3a), 8.08 (d, J
=
H-H
H-H
H-H
H-H
4’
1
7
3
8
3
.7 , 1H, H of 3a), 7.58-7.54 (m, 2H, H and H of 3a), 7.47 (d, J = 8.2, 1H, H of 3a), 7.46
H-H
3
5’
6
(
d, J = 3.7 , 1H, H ), 7.34-7.30 (m, 1H, H of 3a), 3.98 (s, 3H, NMe of 4a); 3.94 (s, 3H, NMe
H-H
of 3a), 3.37 (s, 6H, Me(dmso) of 3a); 3.25 [s, 3H, Me(dmso) of 4a]; and 2.29 [s, 3H, Me(dmso) of
+
4
a]. ESI-MS (m/z): calc. for C H Cl N OPtS (M + H) = 608.0; found: 608.0. Elemental Anal.
18
19
2
2
2
(
%). Calc. for: C H Cl N OPtS (MW = 608.46): C, 35.53; H, 2.98; N, 4.60 and S, 10.54; found
18 18 2 2 2
C, 35.59; H, 3.15; N, 4.54 and S, 10.37.
2.2. Crystallography
9

ACCEPTED MANUSCRIPT
A colourless prism-like specimen of C H N S (1a) (sizes in Table 1) was used for the X-
16
12
2
ray crystallographic analysis. The X-ray intensity data were measured on a D8 Venture system
equipped with a multilayer monochromator and a Mo microfocus (λ = 0.71073 Å). The frames
were integrated with the Bruker SAINT software package using a narrow-frame algorithm. The
integration of the data using an orthorhombic unit cell yielded a total of 7259 reflections to a
maximum θ angle of 27.50° (0.77 Å resolution), of which 2855 were independent (average
redundancy 2.543, completeness = 99.9%, R = 3.42%, Rsig = 4.25%) and 2482 (86.94%) were
int
2
greater than 2σ(F ). The final cell constants given in Table 1 are based upon the refinement of
the XYZ-centroids of reflections above 20 σ(I). The calculated minimum and maximum
transmission coefficients (based on crystal size) are 0.6711 and 0.7456. The structure was
solved using the Bruker SHELXTL Software Package, and refined using SHELXL [28], using the
space group P2 2 2 , with Z = 4 for the formula unit, C H N S. The final anisotropic full-matrix
1
1
1
16 12
2
2
least-squares refinement on F with 173 variables converged at R = 3.58%, for the observed
1
data and wR = 8.01% for all data. The goodness-of-fit was 1.074. The largest peak in the final
2
3
3
difference electron density synthesis was 0.241 e-/Å and the largest hole was -0.234 e-/Å with
-
3
an RMS deviation of 0.053 e /Å . Further details concerning the resolution and refinement of the
crystal structure are presented in Table 1. CCDC-1560144 contains the crystallographic data of
this paper. These data can be obtained from the Cambridge Crystallographic Data Centre via:
www.ccdc.cam.ac.uk/data.request.cif
10

ACCEPTED MANUSCRIPT
Table 1. Crystal data and details of the refinement for compound 1a.
Empirical formula
Formula weight
Temperature / K
16 12 2
C H N S
264.34
100(2)

/ Å
0.71073
Crystal sizes / mm mm  mm
0.292  0.168  0.048
Orthorhombic
Crystal system
Space group
a / Å
1 1 1
P2 2 2
6.0702(2)
b / Å
12.8577(5)
c / Å
16.0721(6)
α = β = γ / deg.
90
3
V / Å
1254.41(8)
Z
4
-
3
Density (calculated)/ Mgm
1.400
0.243
-
1

/ mm
F(000)
range for data collection / deg.
552

from 2.028 to 27.498
Index ranges
-7 h  7, -16  k  16, -20  l  20
99.9 %
Completeness to  = 25.242°
Absorption correction
Semi-empirical from equivalents
0.7456 and 0.6711
Max. and min. transmission
Refinement method
2
Full-matrix least-squares on F
Data / restraints / parameters
Goodness-of-fit on F2
2855 / 0 / 173
1.074
1
Final R indices [I >2(I)]
R indices (all data)
R = 0.0358, wR
2
= 0.0743
R
1
= 0.0471, wR
0.04(4)
0.241 and -0.234
2
= 0.0801
Absolute structure parameter
Largest diff. peak and hole/ e.Å
-
3
2.3. Computational details
The conformational map has been searched at the molecular mechanics level using the
augmented MMFF94 method [29] as implemented in Spartan [30]. The dihedral angle S1-C2-
C3-C4 () has been sampled every 5 deg. and the remaining geometric parameters have been
11

ACCEPTED MANUSCRIPT
fully optimized. DFT [31] calculations have been performed using the B3LYP functional [32,33]
implemented in the Gaussian 03 software [34] and the 6-31G* basis set [35], including
polarization functions for the non-hydrogen atoms.
2
.4. Biological Studies
.4.1. Cell culture
2
Colon adenocarcinoma (HCT116) cells (from the American Type Culture Collection) and
breast cancer (MDA-MB231 and MCF7) cells (from European Collection of Cell Cultures,
ECACC) were used for all the experiments. Cells were grown as a monolayer culture in
minimum essential medium (DMEM with L-glutamine, without glucose and without sodium
pyruvate) in the presence of 10 % heat-inactivated fetal calf serum, 10 mM of D-glucose and
0.1% streptomycin/penicillin in standard culture conditions.
2.4.2. Cell viability assays
For these studies, compounds were dissolved in 100 % DMSO at 50 mM as stock
solution; then, serial dilutions have been done in DMSO (1:1) (in this way DMSO concentration
in cell media was always the same); finally, 1:500 dilutions of the serial dilutions of compounds
on cell media were done. The assay was carried out as described by Givens et al. [36]. In brief,
MDA-MB231 and MCF7 cells were plated at 5000 cells/well or 10000 cells/well respectively, in
100 µL media in tissue culture 96 well plates (Cultek). After 24 h, medium was replaced by 100
µL/well of serial dilution of drugs. Each point concentration was run in triplicate. Reagent blanks,
containing media plus colorimetric reagent without cells were run on each plate. Blank values
were subtracted from test values and were routinely 5-10 % of uninhibited control values. Plates
were incubated for 72 h. Hexosamidase activity was measured according to the following
protocol: the media containing the cells was removed and cells were washed once with PBS 60
µL of substrate solution (p-nitrophenol-N-acetyl-β-D-glucosamide 7.5 mM [Sigma N-9376],
sodium citrate 0.1 M, pH = 5.0, 0.25 % Triton X-100) was added to each well and incubated at
37 °C for 1-2 hours; after this incubation time, a bright yellow colour appeared; then, plates could
12

ACCEPTED MANUSCRIPT
be developed by adding 90 µL of developer solution (Glycine 50 mM, pH = 10.4; EDTA 5 mM),
and absorbance was recorded at 410 nm.
2.4.3. DNA migration studies
A stock solution (10 mM) of each compound was prepared in high purity DMSO. Then,
serial dilutions were made in MilliQwater (1:1). Plasmid pBluescript SKþ (Stratagene) was
obtained using a QIAGEN plasmid midi kit as described by the manufacturer. Interaction of
+
drugs with pBluescript SK plasmid DNA was analysed by agarose gel electrophoresis following
a modification of the method described by Abdullah et al. [37]. PlasmidDNA aliquots (40 mg/mL)
were incubated in TE buffer (10 mM TrisHCl, 1 mM EDTA, pH 7.5) with different concentrations
of compounds 1a, 1b, 2a and 3a ranging from 0 M to 200 M at 37 C for 24 h. Final DMSO
concentration in the reactions was always lower than 1%. For comparison, cisplatin and ethidium
bromide (EB) were used as reference controls. Aliquots of 20 mL of the incubated solutions of
compounds containing 0.8 mg of DNA were subjected to 1% agarose gel electrophoresis in TAE
buffer (40 mM Trisacetate, 2 mM EDTA, pH 8.0). The gel was stained in TAE buffer containing
ethidium bromide (ET, 0.5 mg/mL) and visualized and photographed under UV light.
2.4.4. DNA Topoisomerase I and Topoisomerase IIα inhibition assays
Topoisomerase I-based experiments were performed as described previously [38].
Supercoiled pBluescript DNA, obtained as described above, was treated with Topoisomerase I
in the absence or presence of increasing concentrations of compounds 1a, 1b and 2a. Assay
mixtures contained supercoiled pBluescript DNA (0.8 μg), calf thymus Topoisomerase I (3 units)
and compounds 1a, 1b or 2a (0–200 μM) in 20 μL of relaxation buffer Tris-HCl buffer (pH 7.5)
containing 175 mM KCl, 5 mM MgCl and 0.1 mM EDTA. Ethidium bromide (EB, 10 μM) was
2
used as a control of intercalating agents and etoposide (E, 100 μM) as a control of the non-
intercalating agent. Reactions were incubated for 30 min at 37 °C and stopped by the addition of
2
μL of agarose gel loading buffer. Samples were then subjected to electrophoresis and DNA
bands stained with ethidium bromide as described above.
13

ACCEPTED MANUSCRIPT
To distinguish whether compounds act as Topoisomerase inhibitors or DNA intercalators
the conversion of relaxed DNA to a supercoiled state caused by the compounds was analysed in
the presence of Topoisomerase I. Relaxed DNA was obtained by incubation of supercoiled DNA
with Topoisomerase I as described above. Assay mixtures (20 μL) contained: relaxed DNA,
Topoisomerase I (3 units) and compound (50 μM or 100 μM). Reactions were incubated 20 min
at 37 °C and stopped as described above. Ethidium bromide (10 μM) was used as a control of
intercalative drug.
The DNA Topoisomerase IIα inhibitory activity of the compounds tested in this study was
measured as follows. Supercoiled pBluescript DNA was incubated with Topoisomerase IIα
(
Affymetrix) in the absence or presence of increasing concentrations of compounds under
analysis. Assay mixtures contained supercoiled pBluescript DNA (0.3 μg), Topoisomerase IIα (4
units) and the tested compounds (0 - 100 μM) in 20 μL of 1x Topo II reaction buffer (PN73592).
Etoposide was used as a control of Topo IIα inhibitor. Reactions were incubated for 45 min at 37
°
C and stopped by the addition of 2 μL of agarose gel loading buffer. Samples were then
subjected to electrophoresis and DNA bands stained with ethidium bromide as described before.
2.4.5. Cathepsin B inhibition assay
The colorimetric cathepsin B assay was performed as described by Casini et al. [39] with
few modifications. Briefly, the reaction mixture contained 100 mM sodium phosphate (pH 6.0), 1
mM EDTA and 200 μM sodium N-carbobenzoxy-L-lysine p-nitrophenyl ester as the substrate.
To have the enzyme catalytically active before each experiment the cysteine in the active
site was reduced by treatment with dithiothreitol (DTT). For this purpose, 5 mM DTT was added
to the cathepsin B sample, before dilution, and incubated 1 h at 30 °C. To test the inhibitory
effect of the compounds on cathepsin B, activity measurements were performed in triplicate
using fixed concentrations of enzyme (500 nM) and substrate (200 μM). The compounds were
used at concentrations ranging from 5 to 100 μM. Previous to the addition of substrate,
cathepsin B was incubated with the different compounds at 25 °C for 2 h. The cysteine
proteinase inhibitor E-64 was used as a positive control of cathepsin B inhibition. Complete
14

ACCEPTED MANUSCRIPT
inhibition was achieved at 10 μM concentration of E-64. Activity was measured over 90 s at 326
nm on a UV-spectrophotometer.
3
3
3
. Results and discussion
.1. Synthesis and characterization
.1.1. Synthesis of the ligands
The new carbazole derivatives: 9-methyl-3-(2-thiazolyl)-9H-carbazole (1a) and 9-methyl-
-(2-thienyl)-9H-carbazole (1b) were prepared from commercially available carbazole in a three-
3
step-sequence of reactions (Scheme 1), that involved the iodination of the 9H-carbazole [26]
followed by the alkylation to produce the iodo-9-methyl-9H-carbazole [40], that later on reacted
with either 2-(tributylstannyl)thiazole (for 1a) [41] or 2-(tributylstannyl)thiophene (for 1b) via Stille
coupling reaction [42] to produce the final products. All compounds were entirely characterized
1
13
1
by H NMR and C{ H} NMR spectroscopies, mass spectrometry and elemental analyses.
15

ACCEPTED MANUSCRIPT
Scheme 1. Synthesis of carbazoles 1a and 1b. Reagents and conditions: i) KI, KIO , acetic acid, reflux. ii)
3
NaH, DMF, room temperature followed by treatment with iodomethane, in DMF at room temperature. iii) 2-
(
Tributylstannyl)thiazole, [Pd(PPh ) ], DMF, 100 °C. iv) 2-(Tributylstannyl)thiophene, [Pd(PPh ) ], DMF,
3 4 3 4
100 °C.
The crystal structure of compound 1a (Fig. 2) confirmed the presence of the thiazolyl
group on position 3. In compound 1a, the nitrogen atom of the thiazolyl unit (N1) is on the same
side as the Me group. As a consequence of this arrangement of groups, the N1 atom is proximal
to the hydrogen atom H12 of the carbazole unit while the S1 atom is relatively close to the H4
atom. The distances N1···H12 (2.580 Å) and S1···H4 (2.751 Å) are smaller than the sum of the
van der Waals radii of the atoms involved (N, 1.55 Å; H, 0.95 Å and S, 1.85 Å) [43,44]. Thus
suggesting the existence of non-conventional C-H···N and C-H···S intramolecular hydrogen
bonds [45], similar to those found in most 2-phenylthiazole derivatives [43,46].
Fig. 2. Molecular structure and atom labelling scheme for the new hybrid carbazole-thiazole ligand (1a).
The thiazolyl group is planar and slightly twisted (ca. 13.9°) in relation to the main plane
of the carbazole array. In the crystal, the relative orientation of the molecules (Fig. 3, A) allows
··· interactions between the heterocyclic array of a unit at (x,y,z) and the substituted phenyl
ring of another one at (1+x, y, z) (the distance between the centroids of these rings is 3.90 Å). In
addition, one of the hydrogen atoms of the methyl group is at only 2.77 Å from the centroid of the
phenyl ring of a parallel unit, indicating the existence of intermolecular C-H··· contacts. As a
consequence of this, the assembly of the molecules results in pillars (Fig. 3, A). These structural
units are connected by additional C-H··· short contacts (Fig. 3, B) involving the H3 atom of the
heterocyclic units in one of the pillars and the centroids of the thiazolyl groups of another one.
16

ACCEPTED MANUSCRIPT
Fig. 3. Schematic view of: A) the assembly of a molecule of 1a, sited at (x, y, z) and another unit at (-1+ x,
y, z) by - stacking between the thiazolyl and the phenyl ring of the carbazole (in purple) and C-H···
short contacts (green dotted lines) involving one of the methyl protons (H10) and the propagation of these
interactions along the crystal to give pillars; B) simplified view of connectivity of the pillars through C-H···
contacts.
3.1.2. Coordination capability of the new hybrid carbazoles 1a and 1b
In view of their potential biological activities (i.e. antitumor, antibacterial), we decided to
evaluate the coordination abilities of the new carbazoles in front of the Pd(II) and Pt(II) ions. In a
first stage, we selected ligand 1a and studied its reactivity with [MCl (dmso) ] (M = Pd or Pt) or
2
2
Na [PdCl ] under different experimental conditions [Table 2 (entries I-VI) and Scheme 2]. When
2
4
[
PdCl (dmso) ] was treated with ligand 1a [in molar ratios Pd(II): 1a = 1 or 2] in refluxing
2 2
methanol for 1 h a pale yellowish precipitate (hereinafter referred to as 2a) was formed with a
yield of 57% in the case of using a molar ratio of 1:1 (Table 2, entry I). Elemental analyses of 2a
and NMR characterization agreed with those expected for [PdCl (1a) ]. Compound 2a is a stable
2
2
solid at room temperature and exhibits low solubility in CHCl or CH Cl . Compound 2a can be
3
2
2
obtained with a higher yield of 82% and at room temperature using Na [PdCl ], instead of the
2
4
[
PdCl (dmso) ], a two-fold excess of carbazole 1a and methanol as solvent (Table 2, entry II).
2 2
17

ACCEPTED MANUSCRIPT
Table 2. Summary of experimental conditions [reagents, molar ratios (1a:Pd(II) or 1a:Pt(II)), solvents,
temperature (T), reaction time (t, in h)] used in the study of the reactivity of carbazole 1a with
[
PdCl (dmso) ], Na [PdCl ] or cis-[PtCl (dmso) ].
2
2
2
4
2
2
Entry Reagents
molar ratios
(1:1) or (2:1)
(1:1) or (2:1)
(1:1)
Solvent
MeOH
T
t
Final products
I
1a : [PdCl
2
(dmso)
2
]
reflux
1
2a
2a
3a
3a
II
1a : Na [PdCl
2
4
]
MeOH
298 K 24
III
IV
V
VI
1a : cis-[PtCl
1a : cis-[PtCl
1a : cis-[PtCl
1a : cis-[PtCl
2
2
2
2
(dmso)
(dmso)
(dmso)
(dmso)
2
2
2
2
]
MeOH
reflux
reflux
reflux
reflux
1
]
(1:1)
MeOH
24
24
72
a
a
b
] : NaOAc
] : NaOAc
(1:1:1)
Toluene /MeOH
Toluene /MeOH
3a and 4a
c
(1:1:1)
3a and 4a
a
b
c
A 5:1 mixture. Only traces of compound 4a were detected. Integration of the signals observed in the
1
H-NMR spectrum of the raw material indicated that the molar ratio 3a:4a was 1.7.
Scheme 2. Synthesis of the complexes. Reagents and conditions: i) [PdCl (dmso) ] in refluxing methanol
2
2
(
1 h) or Na [PdCl ] in methanol at 298 K, 24 h [molar ratios Pd(II):1a = 1:1 and 1:2, respectively]; ii) cis-
2 4
[
PtCl (dmso) ] in refluxing methanol (1 h) and iii) equimolar amounts of cis-[PtCl (dmso) ] and NaOAc in a
2 2 2 2
toluene: MeOH (5:1) mixture under reflux {see text and Table 2, (entries V and VI)}.
In order to compare the effect of the binding of the M(II) ion to the carbazole 1a on the
antitumor activity, we also studied the reactivity of 1a in front of Pt(II). Treatment of equimolar
amounts of 1a and cis-[PtCl (dmso) ] in methanol (HPLC grade) under reflux for 1 h, followed by
2
2
the work-up of a SiO column chromatography gave a yellowish solid (3a) (Table 2, entry III and
2
Scheme 2). Its elemental analyses were consistent with those expected for [PtCl (1a)(dmso)].
2
18

ACCEPTED MANUSCRIPT
195
1
Moreover, the position of the singlet observed in the Pt{ H}-NMR spectrum of 3a ( = -2983
ppm) agrees with those of related Pt(II) complexes with a “PtCl (Nheterocycle)(Sdmso)” core [23b-
2
1
d,47]. Its H-NMR spectrum (Fig. S2) showed two singlets of relative intensities 1:2 in the high
field region. The less intense one is assigned to the methylic protons of the ligand at  = 4.0
ppm; while the other corresponds to the six protons of the dmso ligand. This finding is
characteristic of trans- isomers of [PtCl (N-donor ligand)(dmso)] [23b-d,47], thus indicating that
2
compound 3a is trans-[PtCl (1a)(dmso)]. It should be noted that when the reaction was
2
performed using longer reaction times no evidences of the formation of any other Pt(II)
1
compound were detected by H-NMR
.
Since it is well-known that the presence of a base such as NaOAc and mixtures of
toluene / methanol (5:1) as solvent may induce the formation of the cis- isomers of compounds
[
PtCl (N-donor ligand)(dmso)] or even cycloplatinated complexes [23b-e,47,48], we also
2
investigated whether for ligand 1a the addition of NaOAc could affect the nature of the final Pt(II)
product. When equimolar amounts of 1a, cis-[PtCl (dmso) ] and NaOAc were refluxed in a
2
2
1
mixture of toluene : methanol (5:1) for 72 h (Table 2, entry VI and Scheme 2), the H-NMR
spectrum of the raw material in CDCl at 298 K (Fig. S4) revealed the coexistence of 3a and a
3
minor product (hereinafter referred to as 4a). The work-up of a column chromatography allowed
195
1
us to isolate complex 3a and a solid containing a mixture of 3a and 4a. The Pt{ H} NMR
spectrum of the solid dissolved in CDCl at 298 K (Fig. S1, A) showed two singlets (one at  = -
3
2980 ppm (due to 3a) and the other at  = -2932 ppm assigned to compound 4a). Their chemical
shifts suggest that the environment of the Pt(II) atoms in 3a and 4a should be very similar.
Moreover, the separation between the two singlets (ca. 41 ppm), falls in the typical range
1
reported for trans- and cis- isomers of [Pt(N-donor)Cl (dmso)] compounds. Besides that, its H-
2
NMR spectrum (Fig. S1, B) revealed that for 4a the resonances due to the protons of the dmso
-
ligand appeared as two singlets, in good agreement with a cis- disposition of the Cl ligands. On
these basis we assumed that 4a is the cis- isomer of [PtCl (1a)(dmso)]. Unfortunately, attempts
2
19

ACCEPTED MANUSCRIPT
to isolate 4a in its pure form, by fractional crystallization or subsequent column chromatography
failed.
1
Comparison of H-NMR spectra of the new complexes (2a, 3a and 4a) with that of the
2
4
parent ligand 1a reveals that the resonances due to the H and H protons of the carbazole array
were highly affected by the binding of the nitrogen to the Pt(II) ion. It should be noted that: a) the
formation of the Pt-N(thiazole) bond requires the cleavage of the intramolecular C11-H12···N
hydrogen bond, b) in cis- and trans- isomers of [PtCl (N-heterocycle)(dmso)] complexes, the
2
heterocycle is orthogonal to the main coordination plane of the ligand [23d-e,49,50], and c)
-
frequently ancillary ligands (Cl or dmso) are involved in additional C-H ···X [X = Cl or O(dmso)]
contacts with the neutral N-donor ligand [49,50]. All these findings could explain the variations
observed in the chemical shifts of the protons adjacent to position 3 in complexes 3a and 3b.
In order to compare the potential coordination ability of the two new carbazoles, the
reactivity of the thienyl derivative 1b with [MCl (dmso) ] and Na [PdCl ] was studied under
2
2
2
4
identical conditions as for 1a (described above and shown in Scheme 2) and using identical
conditions as those shown in Table 2 (entries I - IV). However, none of these studies allowed us
1
neither the isolation or even the detection by H-NMR of any Pd(II) or Pt(II) complex, thus
suggesting that thiazole-substituted carbazole 1a has a greater coordination ability than the
thienyl analogue 1b.
3.2. Electronic spectra and optical properties
Absorption spectra of CH Cl solutions of 1a and 1b at 298 K (Table 3 and Fig. S5, A)
2
2
showed two intense bands in the range 250–450 nm that are characteristic of carbazoles. The
corresponding UV-vis spectra of the complexes 2a and 3a (Fig. S5, B and Table 3) exhibited two
intense absorption bands in the range 300   < 350 nm. One of them shifted to lower energies
in relation to the free ligand being for the Pt(II) complex (3a) the magnitude of the shift bigger
than for the Pd(II) complex 2a (Table 3). These findings suggest that this absorption band is due
to a metal perturbed intra-ligand electronic transition (MPILET). The second absorption band, at
20

ACCEPTED MANUSCRIPT
higher energies, is practically coincident with that of the free ligand. The spectra of compounds
2a and 3a (Fig. S5, B) also exhibited an additional and poorly resolved absorption band at lower
wavelengths [280   < 290 nm].
Table 3. Absorption and emission properties of the free carbazoles (1a and 1b) and the Pd(II) and Pt(II)
complexes (2a and 3a, respectively) in CH Cl [Wavelengths  (in nm), logarithms of the extinction
2
2
i
coefficients (log  ), emission wavelengths [ (in nm) after excitation at exc = 300 nm] and quantum
i
em
yields ().
Absorption
Emission
spectroscopic data
spectroscopic data
Compd.
 (log )
em

1
1
2
3
a
b
a
a
328 (4.3), 301 (4.5), 279 sh ( 4.2)
317 ( 4.2), 299 (4.5)
370 sh, 384
376, 394
370 sh, 385
370 sh, 386
0.22
0.12
0.02
< 0.01
334 (3.6), 300 (3.8), 284 ( 3.8)
345 (4.2), 301 (4.4), 287 (4.4)
The emission spectra of 1a, 1b, 2a and 3a were recorded in CH Cl solution at 298 K.
2
2
Upon excitation at exc = 300 nm, the free ligands 1a and 1b exhibited emission bands in the
range 370-395 nm (Fig. S6 and Table 3). The thienyl-based derivative 1b showed a
bathochromic shift of the wavelength of maximum emission of 10 nm in comparison to the
thiazolyl-based ligand 1a, according to the strongest electron-donating character of the thienyl
unit. It should be noted that the quantum yield of 1a (Table 3) is significantly higher than that of
1b. Complexes 2a and 3a exhibited also emission bands consistent with that of ligand 1a, but
their fluorescence quantum yields [51] decreased considerably in relation to the free ligand (1a)
.
3.3. Computational studies
In order to elucidate the effect produced by the thiazolyl or thienyl groups of compounds
1a and 1b on the electronic delocalization, computational calculations based on the DFT
methodology were undertaken [31]. Calculations were carried out using the B3LYP hybrid
functional [32,33] and the 6-31G* basis set [35] implemented in the Gaussian03 program [34]. In
a first stage, geometries of compounds 1a and 1b were optimized without imposing any
21

ACCEPTED MANUSCRIPT
restriction. Final atomic coordinates for the optimized geometries are included as supplementary
information (Tables S1-S2). Bond lengths and angles of the optimised geometry of 1a were
consistent with those obtained from the X-ray studies (the differences did not clearly exceed 3)
and those of 1b are in the range reported for related carbazoles with mono or polythienyl units
on position 3 [20,50].
Molecular orbital (MO) calculations of the optimized geometries revealed that highest
occupied molecular orbital (HOMO) (Fig. 4) of 1a and 1b are very similar except for a tiny
difference in the contribution of the atomic orbitals of the sulphur atom. The LUMO (Fig. 4) of 1a
is mainly centred on the thiazolyl unit and two of the rings of the carbazole; while in 1b, the
contribution of the thienyl decreases in relation to that of the thiazole in 1a. Moreover, the
HOMO-LUMO gap (E) of 1b (4.32 eV) is higher than for 1a (4.14 eV). These findings suggest
that the replacement of the thiazolyl ring of 1a by the thienyl in 1b reduces the electronic
delocalization mentioned above.
Fig. 4. Highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) for
the new carbazole derivatives (1a and 1b).
In the optimized geometry of 1b the intramolecular separation between the S1 atom and
the hydrogen atom of phenyl ring (2.848 Å) is larger than in 1a [S1···H: 2.748 Å (optimized
22

ACCEPTED MANUSCRIPT
geometry) or 2.751 Å (from the crystal structure)] and the angle formed by the heterocycle and
the carbazole is 30.3 bigger than in 1a. Since it is well-known that deviations from planarity
affects the electronic delocalization, the properties of the compounds and their potential utility,
we also calculated the energy of the molecules for different orientations of the attached
heterocycle versus the carbazole unit using molecular mechanics. These arrangements were
generated by modifying the torsion angle defined by the set of atoms S1-C2’-C3-C4 (hereinafter
referred to as ) from 0 to 360. The results shown in Fig. 5, reveal that for 1a the minimum
energy corresponds to  values in the ranges (0    16 and 344   360), that is to say
close to co-planarity, similar to that found in the crystal structure  = 13.5 and with the S1 atom
and NMe group located in opposite sides.
Fig. 5. Plot of the energy of the molecule of 1a (blue) or 1b (red) versus the value of the torsion angle S1-
C2’-C3-C4 ().
The energy barrier to achieve an orthogonal arrangement of the thiazole ( = 90 or
270) is rather high (9.1 kJ/mol). The conformer with the N1 and N2 atoms on opposite sides, [
values between 164 and 196] is slightly less stable than for  = 0  16  (Fig. 5). The
differences between the energies of both conformers determined from molecular mechanics
calculations and DFT are 0.4 and 0.5 Kcal/mol, respectively.
In contrast with the results obtained for 1a, in 1b the most favourable orientation of the
thienyl unit is far away from co-planarity and corresponds to φ values in the ranges 128° - 133°
23

ACCEPTED MANUSCRIPT
and 232° - 237°. The energy barriers to achieve coplanar arrangements (Fig. 5) are smaller than
that obtained for 1a (9.1 kJ/mol); consequently, from an energetic point of view, the free rotation
of the thienyl unit, that requires a smaller energy income, is more likely to occur than that of the
thiazolyl ring of 1a. In addition, time dependent DFT (TD-DFT) calculations were performed to
achieve the assignment of the bans observed in the UV-vis spectra (Table S4 and Fig. S7).
3.4. Biological studies
3.4.1. Antiproliferative assay
We have evaluated the cytotoxic activity of ligands 1a and 1b and the new complexes 2a
and 3a in front of the colon cell line HCT116 and two breast cancer cell lines [the triple negative
ER, PR and no HER2 over expression) MDA-MB231 and the MCF7]. The effects of the new
(
products on the growth of the three cell lines and that of cisplatin, used as positive control, were
assessed after 72 h and the results are presented in Table 4 and Fig. 6.
The comparison of the in vitro cytotoxic activities of the free carbazoles 1a and 1b in the
HCT116 cell line revealed that the replacement of the thienyl ring (in 1b) by the thiazolyl unit of
1a produced a significant enhancement of the cytotoxic potency. This trend is practically
identical to those observed in the two breast (MDA-MB231 and the MCF7) cancer cell lines and
could be attributed to several factors. One of these could be the lipophilicity that, as shown in
Table 4, is expected to be slightly different for the two systems. It should be noted that in the
MCF7 cell line ligand 1a is (ca. 9.5 times) more potent than cisplatin, the non-alkylated 9H-
carbazole (IC > 40 M) and even than doxorubicin (IC = 7.3 M).
50
50
a
Table 4. Cytotoxic activities (IC50 values , M) on HCT-116 (colon), MDA-MB-231 and MCF7 (breast) cell
lines, for the palladium(II) and platinum(II) complexes derived from 1a (2a and 3a, respectively) and for
b
c
cisplatin. For comparison purposes, Clog P values and lipophilic efficiencies (LipE ) of compounds 1a,
1b, 2a and 3a on the MCF7 cell line are also included (nd = not determined).
24

ACCEPTED MANUSCRIPT
a
IC50 values in the cancer cell lines
b
c
Compounds
Carbazoles
ClogP
LipE
HCT116
MDA-MB231
MCF7
1
1
a
b
31  2
9.4  2.6
51  nd
2.0  0.5
47  nd
4.42
5.73
1.27
50  nd
-1.39
Complexes
2
3
a
a
> 100
> 100
24  2
9.42
4.33
-------
-4.80
1.29
34  nd
40  4.4
37  nd
6.5  2.4
2.4  2.2
19  4.5
Cisplatin
-------
a
b
Data are shown as the mean values of two experiments performed in triplicate. ClogP is the calculated
logarithmic value of the n-octanol/water partition coefficient and was calculated using the
c
ChemBioDrawUltra computer program. LipE indexes were calculated as LipE = - log(IC )- ClogP [52]
50
and using the IC50 values obtained in the MCF7 cell line.
In order to compare the effect produced by the binding of the Pd(II) or Pt(II) we also
examined the effect produced by the complexes 2a and 3a on identical cell lines. As shown in
Table 4 and Fig. 6, the Pd(II) complex 2a did not show any relevant antiproliferative activity (IC50
>
100 µM) in the HCT116 and MDA-MB213 cell lines. In the MCF7 it was more active, but its
potency was slightly smaller (IC = 24  2.2 M) than that of the reference cisplatin (IC = 19 
50
50
4.5 M). In contrast, the Pt(II) complex (3a) exhibited a higher inhibitory growth effect, being 9
times more potent than the reference drug in the MCF7 breast cancer cell line.
Fig. 6. Comparative plot of the IC50 values (in µM) of the new carbazoles (1a and 1b), the Pd(II) and Pt(II)
complexes derived from 1a and cisplatin in front of the colon cancer cell line (HCT116) and the two breast
cancer cell lines (MDA-MB231 and MCF7).
25

ACCEPTED MANUSCRIPT
It is well-known that the preparation of new products with improved cytotoxic potency is
not the unique requirement in medicinal chemistry and drug design, other factors such as the
lipophilicity that contributes to the ADMET (absorption, distribution, metabolism excretion and
toxicity) properties of drugs also plays a crucial role. Nowadays, the lipophilic efficiency (LipE)
index [52] that includes lipophilicity and potency is becoming more and more popular in drug
design and optimization, because it allows to normalize the observed potency with changes in
the lipophilicity.
In view of this, we calculated the clog values for the new compounds and their LipE index
in the MCF7 cell line. The results (Table 4) reveal that for the compounds characterized in this
work the LipE index increases as follow 2a << 1b < 1a  3a. The Pd(II) complex 2a that shows
low solubility it is simultaneously the most lipophilic and the less potent compound of the series.
In contrast with these results, for the carbazole-thiazole ligand 1a and its trans-[PtCl (1a)(dmso)]
2
complex 3a there is an effective combination of their cytotoxic potency in MCF7 and lipophilicity,
and on these basis they are promising scaffolds in the search of optimized drugs. Chemical
modifications of the core of ligand 1a, its binding to the Pt(II) atom or even changes on the
ancillary ligands bound to it in 3a may allow to tune the lipophilicity and to improve the lipophilic
efficiency.
3.4.2. Additional studies to elucidate the mechanism of action
In the majority of the described cases of cytotoxic carbazoles, they act as DNA-
intercalators or as Topoisomerase I/II (or telomerases) inhibitors [53], although other
mechanisms of action involving different targets {i.e. estrogen receptors (ER) or cyclin
dependent kinases (CDK), among others} have also been postulated [10,11,54]. To examine
whether the presence of the thiazole (in 1a) or the thienyl unit (in 1b) in the free ligand and the
binding of 1a to the Pd(II) or Pt(II) ions in complexes 2a and 3a, could have an important role in
the mechanism of action, additional experiments were performed.
To elucidate whether compounds 1a, 1b, 2a and 3a act as DNA intercalators or as
Topoisomerase I or II inhibitors, three different sets of experiments were undertaken. In a first
26

ACCEPTED MANUSCRIPT
stage it was examined if the new compounds could induce changes in the electrophoretic
+
mobility of the supercoiled closed form (ccc) of pBluescript SK plasmid DNA. For DNA migration
studies, the plasmid was incubated with compounds 1a, 1b, 2a and 3a at increasing
concentrations ranging from 0 µM to 200 µM. For comparison purposes, incubation of DNA with
cisplatin (CDDP) or ethidium bromide (EB) was also performed. As expected, cisplatin greatly
altered the electrophoretic mobility of pBluescript DNA at all concentrations tested. As depicted
in Fig. 7, the free ligands (1a and 1b) and the Pd(II) compound (2a) were not effective. Only the
Pt(II) complex (3a) produced a significant effect on the electrophoretic mobility of native
pBluescript DNA at concentrations greater than 100 µM.
OC
+
Fig. 7. Interaction of pBluescript SK plasmid DNA (40 µg/ml) with increasing concentrations of
compounds 1a, 1b, 2a and 3a, ethidium bromide (EB) and cisplatin. Lane 1: DNA only; Lane 2: 1 M;
Lane 3: 2.5 M; Lane 4: 5 M and Lane 5: 10 M; Lane 6: 25 M; Lane 7: 50 M; Lane 8: 100 M; Lane
9: 200 M. ccc represents the supercoiled closed circular form and oc the open circular form.
Secondly, a Topoisomerase based gel assay was performed to evaluate the ability of
compounds 1a, 1b and 2a to intercalate into DNA or to act as DNA Topoisomerase I inhibitors.
For that, supercoiled pBluescript plasmid DNA was incubated with Topoisomerase I in the
presence of increasing concentrations of the compounds under study. The results are presented
in Fig. 8, where ethidium bromide (EB) was used as an intercalator control. The analysed
compounds did not prevent unwinding of DNA indicating that they are neither intercalators nor
Topoisomerase I inhibitors.
27

ACCEPTED MANUSCRIPT
Fig. 8. Analysis of the new ligands (1a and 1b) and compound 2a as potential DNA intercalators or
Topoisomerase I inhibitors. Conversion of supercoiled pBluescript SK+ DNA (40 µg/mL) to relaxed DNA
by the action of Topoisomerase I (3 units) in the absence or in the presence of increasing amounts of
compounds. E = 100 M etoposide; EB = 10 M Ethidium Bromide; SC = supercoiled DNA as control; R =
relaxed DNA by the action of Topoisomerase I as control; Lane 1: 25 M; Lane 2: 100 M; Lane 3: 200
M; ccc= closed circular form and oc = open circular form.
As mentioned above, another important target for antitumor agents is the Topoisomerase
II, which is associated with solving the topological constraints of DNA by transiently cleaving
both strands of the double helix [53]. In humans there are two Topoisomerase II isoenzymes, IIα
and IIβ. Here we study the capability of compounds 1a, 1b and 2a as catalytic inhibitors of
Topoisomerase IIα. The inhibitory activity was evaluated by measuring the extent of enzyme
mediated relaxed DNA after treatment with 200 µM of 1a, 1b or 2a compounds. Only the Pd(II)
complex 2a showed at this concentration inhibitory activity (Fig. 9).
E
P
T 1a 1b 2a
R
ccc
Fig 9. Topoisomerase-II  inhibitory activity of compounds 1a, 1b and 2a. Reactions contained
supercoiled plasmid DNA, Topoisomerase IIα (4 units) and 200 µM of the indicated compound. Control
reactions were performed in the presence of: E, etoposide at 100 µM; P: SC Plasmid DNA only; T:
reaction performed with plasmid DNA and Topoisomerase IIα (4 units)
.
Other mechanism of action implies Cathepsin B, which is a cysteine metaloprotease that
could be involved in metastasis, angiogenesis and tumour progression. Examples of Pd(II) and
28

ACCEPTED MANUSCRIPT
Pt(II) complexes as inhibitors of Cathepsin B have been reported [55]. However, none of the new
compounds presented in this work (1a, 1b and 2a) inhibited the enzyme activity at 100 µM
concentration.
Overall the biological studies undertaken with the new compounds 1a, 1b, 2a and 3a
provide conclusive evidences. The DNA migration studies revealed that only Pt(II) complex (3a)
modifies the electrophoretic mobility of the plasmid in a similar way as cisplatin but at higher
concentrations. Experimental results also revealed that neither compounds 1a, with higher
cytotoxic activity than cisplatin in the tested cancer lines HCT116 and MCF7, nor compounds 1b
or 2a operate as intercalators or as inhibitors of cathepsin B or Topoisomerase I. However, the
Pd(II) complex (2a) inhibits the activity of Topoisomerase II (at 200 µM concentration).
As a summary, among the new compounds, 1a and its Pt(II) complex are the most active
in the assayed HCT116, MDA-MB231 and MCF7 cancer cell lines. Moreover, compound 1a,
clearly more potent than 3a, has an additional interest due to its low toxiticy on the normal and
non-tumoral human skin fibroblast BJ cell line (IC > 100 µM for 1a versus IC = 23 ꢀ 2 µM for
50
50
cisplatin under identical experimental conditions). This finding enhances the interest and
relevance of carbazole 1a for further and additional biological studies.
4. Conclusions
Here we have presented two new N-methylated and 3-substituted carbazoles with a
thiazolyl (1a) or a thienyl (1b) unit and comparative studies of their properties and reactivity in
front of Na [PdCl ] or [MCl (dmso) ] (M = Pd or Pt) and biological activities. The obtained results
2
4
2
2
proved that compound 1a is clearly more reactive than 1b and has a greater coordination
capability towards the Pd(II) and Pt(II) ions, leading to trans-[PdCl (1a) ] (2a) and the geometrical
2
2
isomers {trans- (3a) or cis- (4a)} of [PtCl (1a)(dmso)]. Computational studies have allowed us to
2
explain the effect produced by the thiazolyl and thienyl units on: a) their relative orientation in
relation to the backbone of the carbazole, b) the electronic delocalization, c) the distribution of
charge and d) the origin of the bands observed in their absorption spectra and their assignment.
29