α?N?heterocyclic thiosemicarbazone Fe(III) complex: Characterization of its antitumor activity and identification of anticancer mechanism

Article abstract of DOI:10.1016/j.ejmech.2016.07.041

We synthesized an α?N?heterocyclic thiosemicarbazone ligand (L) and its Fe complex (C1) and assessed their chemical and biological properties in order to understand their marked activity. Electrochemical studies and ascorbate oxidation studies demonstrate

Full text of DOI:10.1016/j.ejmech.2016.07.041

European Journal of Medicinal Chemistry 123 (2016) 354e364
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
a
ꢀNꢀheterocyclic thiosemicarbazone Fe(III) complex:
Characterization of its antitumor activity and identification of
anticancer mechanism
Yi Gou, Jun Wang, Shifang Chen, Zhan Zhang, Yao Zhang, Wei Zhang, Feng Yang*
School of Pharmacy, Nantong University, Nantong, Jiangsu, China
a r t i c l e i n f o
a b s t r a c t
Article history:
We synthesized an
a
ꢀNꢀheterocyclic thiosemicarbazone ligand (L) and its Fe complex (C1) and assessed
Received 17 August 2015
Received in revised form
12 July 2016
Accepted 19 July 2016
Available online 26 July 2016
their chemical and biological properties in order to understand their marked activity. Electrochemical
studies and ascorbate oxidation studies demonstrated that C1 shows considerable redox activity, and
FeIII/II redox potentials was within the range accessible to cellular oxidants and reductants. Absorption
spectral, emission spectral and viscosity analysis reveal that L and C1 interacted with DNA through
intercalation and C1 exhibited a higher DNA binding ability. Agarose gel electrophoresis experiments
indicated that C1 exhibited the highest pBR322 DNA cleaving ability. In vitro, C1 showed significantly
more anticancer activity than the ligand alone. Moreover, C1 induces production of reactive oxygen
species (ROS) and DNA damage, resulting in activation of the p53 pathway, cell cycle arrest at the S phase,
and mitochondria-mediated apoptosis by regulating the expression of Bclꢀ2 family proteins.
© 2016 Elsevier Masson SAS. All rights reserved.
Keywords:
Iron(III) complex
a
ꢀNꢀheterocyclic thiosemicarbazone
DNA cleavage
Anticancer activity
Anticancer mechanism
1. Introduction
transferrin, at a rapid rate [12e14]; and (2) cancer cells express
higher levels of the Fe-containing enzyme such as the ribonucleo-
Metal coordination complexes offer biological and chemical
diversity that is distinct from that of organic drugs [1e3]. For de-
cades, cisplatin and its derivatives are used in more than 50% of the
treatment regimes for patients suffering from cancer [4,5]. How-
ever, the toxicities coupled with acquired and intrinsic drug resis-
tance, have hampered theirs broad clinical application [6e8]. These
unresolved disadvantages stimulate research on most desirably
with reduced side effects, more efficacious, a broader spectrum of
indication and higher selectivity metalꢀbased anticancer
compounds.
Iron (Fe) is an essential trace element in biological systems and
catalyzes key reactions involved in metabolism, oxygen transport,
respiration, and DNA synthesis [9]. Many studies have shown that
rapidly growing tumor cells, such as leukemia and neuroblastoma,
have particularly high demands for Fe [10,11]. It is believed that the
antitumor activity of Fe is twoꢀfold: (1) neoplastic cells generally
proliferate at a faster rate than their normal counterparts and ex-
press higher levels of the transferrin receptor 1, which is necessary
for Fe cellular uptake of from serum Fe-transport protein,
tide reductase [15], being a catalyst of the rate-limiting step of DNA
synthesis [16]. These characteristics suggest that iron represents a
potential therapeutic target for cancer treatment.
On the other hand, thiosemicarbazones are an emerging class of
ligands that show pronounced and selective antitumor activity and
can overcome resistance to standard chemotherapy [17e23].
Especially
a
ꢀNꢀheterocyclic thiosemicarbazones with a nitrogen
containing heterocycle in a position to the thiosemicarbazide side
chain exert significant antitumor activity in vitro and in vivo [24,25].
For instance, Triapine (3-aminopyridine-2-carboxaldehyde thio-
semicarbazone) has been tested in a variety of tumor cell lines in
the preclinical setting and is currently undergoing different phase I
and II clinical trials [26]. A recent study of a phase II trial of triapine
(105 mg/m2/day) demonstrated that the overall response rate was
49%, with a complete remission rate of 24% [27]. Interestingly, the
biological activities of thiosemicarbazone can be modified by the
linkage to important metal ions such as Fe, Ni, Cu and Zn [28e30].
Thus, development of thiosemicarbazones compounds that chelate
Fe has become a therapeutic strategy showing significant promise.
However, previous studies have illustrated that complexation of
some thiosemicarbazones ligands such as Bp44 mT, Dp44 mT,
NBp4mT, HDp4pT and NBpT ligands with Fe can result in
* Corresponding author. 9 Seyuan Road, Nantong, Jiangsu, 226019, China.
E-mail address: fyang@mailbox.gxnu.edu.cn (F. Yang).
http://dx.doi.org/10.1016/j.ejmech.2016.07.041
0223-5234/© 2016 Elsevier Masson SAS. All rights reserved.

Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
355
significantly decreased antiproliferative activity when compared to
the free ligands [20,31,32]. In contrast, similar thiosemicarbazones
ligands such as BpT and HDp4mT showing less activity than their Fe
complexes are often noticed [31,32]. Therefore, the mechanisms
involved in the antitumor activities of thiosemicarbazones Fe
complexes are far from clear.
the SHELXTL version 5.1 [33]. All of the nonꢀhydrogen atoms were
refined anisotropically. All hydrogen atoms were placed in
geometrically ideal positions and constrained to ride on their
parent atoms. The crystallographic data for complexes L and C1 are
summarized in Table 1, and selected angles and bond lengths are
given in Table 2. Crystallographic data for the structural analyses
have been deposited at the Cambridge Crystallographic Data
Centre, reference numbers 1418471 for L and 1418470 for C1. The
crystallographic data can be obtained free of charge from the
Cambridge Crystallographic Data Centre via http://www.ccdc.cam.
ac.uk/data_request/cif.
In an effort to discover novel thiosemicarbazones derivatives
with potent anticancer activity and to elucidate their mechanism of
action, thiosemicarbazone bearing condensed
a
ꢀNꢀheterocyclic
moiety ligand (L) and its Fe complex ([Fe(L)2]NO3) were synthe-
sized. We not only comprehensively characterized their chemical
and biological properties but also determined Fe complex mecha-
nism of action for HepG2 cells in vitro. Thus, our results are
important for further understanding the pronounced biological
activity of these complexes.
2.1.3. Density functional theory (DFT) calculations
All of the calculations were done with the GAMESS suite of
codes [34] and the atomic coordinates were obtained from the
Xꢀray structures (Table S1). Full geometry optimization of all
structures was carried out at the density functional theory (DFT)
level of theory using B3LYP method [35,36]. Symmetry operations
were not applied for all structures. Geometries and the reported
Gibbs energies were optimized employing the 6e31 þ G(d,p) basis
set for all atoms. Solvation effects were taken into account using the
integral equation formalism (IEFꢀPCM) [37] model.
2. Experimental section
All solvents and chemicals used were of high purity and avail-
able from commercial sources. Water used in the reactions was
distilled prior to use. Elemental analyses (C, N and H) were per-
formed on a PerkinꢀElmer 2400 analyser. Infrared (IR) spectra of all
samples were recorded using KBr pellets on a Nexus 870 FTꢀIR
spectrophotometer in the frequency of 400e4000 cmꢀ1. UVevi-
sible spectra were measured in DMSO-buffer solution on a Cary 1E
UVeVisible spectrophotometer in the 200e800 nm range.
2.2. Cyclic voltammetry
Cyclic voltammetry was carried out with a BAS100B/W poten-
tiostat. The electrochemical measurements were performed in a
conventional two compartment three electrode cell with a glassy
carbon as a working electrode, an aqueous AgeAgCl as a reference
electrode (Eо ¼ 196 mV vs NHE), and platinum (Pt) wire as a counter
electrode. To ensure the solubility of this complex, the complex C1
was at 1 mM in MeCN/H2O (70:30 v/v). The supporting electrolyte
was Et4NClO4 (0.1 M), and the solutions were purged with nitrogen
prior to measurement.
2.1. Synthesis and structures of L ligand and Fe(III) complex
2.1.1. Synthesis of L ligand and Fe(III) complex
Synthesis of L. A solution of N-phenylhydrazinecarbothioamide
(1.67 g, 10 mmol) in MeOH (15 mL) was added to a methanolic
solution (15 mL) of quinoline-2-carbaldehyde (1.57 g, 10 mmol) and
the mixture refluxed for 1 h. The mixture was kept in air for a week,
forming yellow block crystals. Yield: 71%. Anal. Calcd for
C
17H14N4OS (322.38): C, 63.33; H, 4.38 and N,17.38. Found: C, 62.95;
2.3. Ascorbate oxidation assay
H, 4.67 and N, 17.27. IR (KBr, cmꢀ1): 3651w, 3126s, 2977 m, 1594vs,
1535vs, 1443 m, 1398 m, 1344s, 1311 m, 1186s, 1099 m, 1026s,
932 m, 895 m, 827s, 788w, 690s, 621 m, 460 m. 1Н NMR ([D6]
Ascorbate oxidation assay were investigated as described in an
earlier report [38]. Briefly, ascorbic acid (100
immediately prior to an experiment and incubated in the presence of
Fe(NO3)3 (10
M), the L ligand (ligand:Fe(III) ratios ¼ 0.1,1, or 3) and a
M). The absorbance at 265 nm
mM) was prepared
DMSO):
d
¼ 12.20 (s, 1H), 10.39 (s, 1H), 8.62 (d, J ¼ 8.7 Hz, 1H),
8.43e8.33 (m, 2H), 8.02 (dd, J ¼ 17.0, 8.2 Hz, 2H), 7.78 (t, J ¼ 7.6 Hz,
m
1H), 7.67e7.53 (m, 3H), 7.41 (t, J ¼ 7.8 Hz, 2H) and 7.25 (t, J ¼ 7.3 Hz,
50ꢀfold molar excess of citrate (500
m
1H). 13C NMR ([D6]DMSO):
d
¼ 177.06, 154.25, 147.85, 143.56,
was recorded after 10 and 40 min at room temperature, and the
decrease of intensity between these time points was calculated [39].
139.44, 136.69, 130.41, 129.29, 128.61, 128.39, 128.38, 127.69, 126.78,
126.13 and 118.91 ppm.
Synthesis of [Fe(L)2]NO3 (C1). The L ligand (2 mmol) was dis-
solved in MeOH (15 mL). Then 0.482 g of Fe(NO3)3 was added to the
basic ligand solution, and the mixture was gently refluxed for
30 min at room temperature to give a celadon solution and then
filtered. The filtrate was kept in air for a week, forming blank block
crystals. The crystals were isolated, washed three times with
distilled water and dried in a vacuum desiccator containing anhy-
drous CaCl2. Yield: 63%. Anal. Calcd for C34H26FeN9O3S2 (728.61): C,
56.04; H, 3.49 and N, 17.30. Found: C, 55.31; H, 3.53 and N, 17.17. IR
(KBr, cmꢀ1): 3647w, 2974 m, 1596vs, 1540vs, 1502s, 1471s, 1424s,
1317s, 1185s, 1107s, 993 m, 936 m, 878 m, 825 m, 748s, 689 m,
619 m, 532w, 497 m.
Table 1
Crystal data for complexes HL and C1.
Complex
HL
C1
Empirical formula
Molecular weight
Crystal system
Space group
a (Å)
C17H14N4OS
322.38
monoclinic
P21/c
10.519(5)
7.173(3)
22.858(12)
90.00
98.85(5)
90.00
C34H26FeN9O3S2
728.61
orthorhombic
Pbca
16.3360(4)
18.0196(4)
21.4266(5)
90.00
b (Å)
c (Å)
a
b
g
(o)
(o)
(o)
90.00
90.00
T (K)
296.15
1704.0(14)
4
1.257
672
293.15
6307.3(2)
8
1.535
3000
V (Å3)
Z
2.1.2. Crystal structures determination of Fe(III) complex
Single crystal Xꢀray crystallographic data were collected on a
rcalc. (g$cmꢀ3
F(000)
)
Bruker SMART Apex II CCD diffractometer at room temperature
(Mo-Ka) (mmꢀ1
)
0.199
0.664
using graphite-monochromated MoꢀK
a
(
l
¼ 0.71073 Å) radiation.
m
Data/restraint/parameters
2991/0/217
1.002
0.0957, 0.2164
6455/0/442
1.038
0.0513, 0.0970
Empirical adsorption corrections were applied to all data using
SADABS program. The structures were solved by direct methods
and refined with by fullꢀmatrix leastꢀsquares methods on F2 using
Goodness-of-fit on F2
Final R1, wR2 [I > 2
s(I)]

356
Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
Table 2
of CT-DNA alone and
the complex.
h is the viscosity of CT-DNA in the presence of
Selected bond lengths [Å] and angles [ꢂ] in complexes HL and C1.
HL
S1eC7
N2eC7
N2eN3
N4eC9
N4eC17
C1
Fe1e S2
Fe1eN5
Fe1eN2
Fe1eN1
Fe1eN6
Fe1eS1
S2eFe1eS1
N5eFe1eS2
N6eFe1eS2
N6eFe1eN5
N6eFe1eS1
1.672(7)
1.351(8)
1.379(7)
1.312(7)
1.352(8)
N2eC7eS1
N1eC7eS1
C7eN2eN3
N1eC eN2
C8eN3eN2
118.0(5)
126.8(6)
120.6(6)
115.2(6)
116.8(6)
2.5. DNA cleavage experiment
The cleavage of DNA was studied by agarose gel electrophoresis.
Reactions using supercoiled pBR322 plasmid DNA in 1 ꢁ TBE
(TrisꢀborateꢀEDTA) buffer (pH 7.26) was treated with the com-
2.2109(9)
2.056(2)
1.918(3)
2.062(2)
1.912(3)
2.2170(9)
97.19(4)
163.89(8)
83.97(8)
80.51(10)
90.11(8)
N5eFe1eN1
N5eFe1eS1
N2eFe1eS2
N2eFe1eN5
N2eFe1eN1
N2eFe1eS1
N1eFe1eS2
N1eFe1eS1
N6eFe1eN2
N6eFe1eN1
91.19(10)
87.18(7)
88.33(8)
107.60(10)
80.99(10)
83.61(8)
plexes L and C1 (5e15
total volume of 20
m
M) by dilution with the 1 ꢁ TBE buffer to a
m
L. The mixture was then incubated for 3 h at
37 ꢂC, and then the loading buffer was added. The resulting solu-
tions were electrophoresed for 50 min at 100 V on 1% agarose gels
containing 5
m
L goldview in 1 ꢁ TBE buffer. Then, the gel bands
88.96(7)
were visualized and photographed on a BIO-RAD Laboratories
Segrate gel imaging system.
163.24(8)
169.39(11)
106.08(10)
2.6. Cell lines
The human breast cancer cell lines MCF-7, drug-resistant MCF-
7/ADR cells, human cervical cancer cell lines HeLa, human liver
hepatocellular carcinoma cell lines HepG2 and human liver cell
lines HL-7702 (purchased from the American Type Culture Collec-
tion and the German Collection of Microorganisms and Cell Cul-
tures) were maintained in DMEM or RPMI-1640 supplemented
with 10% fetal bovine serum (FBS), 50 U/mL penicillin, 50 mg/mL
streptomycin at 37 ꢂC and 5% CO2.
2.4. DNA binding experiments
2.4.1. UV spectroscopy
Electronic absorption spectra were determined in the range of
240e650 nm by gradually increasing the amounts of calf-thymus
DNA (CT-DNA) to the complexes (20
mM) in 5 mM Tris-
HClꢀ50 mM NaCl buffer. Then, the resulting solutions were allowed
to equilibrate at room temperature for 5 min. In order to affirm
quantitatively the affinity of the complexes L and C1 bound to DNA,
the binding constant (Kb) was obtained by using the following
equation [40].
2.7. 3ꢀ(4,5ꢀDimethylthiazolꢀ2ꢀyl)ꢀ2,5ꢀdiphenyltetrazolium
bromide (MTT) assays
All compounds were dissolved in PBS with 0.5% DMSO and then
tested. The 100
m
L of cell suspensions at a density of 5 ꢁ 104 cells/
[DNA]/(εaꢀεf) ¼ [DNA]/(εbꢀεf) þ 1/Kb(εbꢀεf)
(1)
mL was seeded in triplicate in 96ꢀwell plates and incubated for
24 h at 37 ꢂC in 5% carbon dioxide atmosphere. Then the medium
was removed and replaced with the respective medium with 10%
FBS containing the compounds at the appropriate concentrations
where [DNA] is the concentration of DNA in base pairs and the
apparent molar extinction coefficients εa, εb and εf correspond to
Aobs/[complex], the extinction coefficient of free (unbound) and the
fully bound complexes, respectively.
and incubated at 37 ꢂC under conditions of 5% CO2 for 48 h. 10
mL
MTT dye (5 mg/mL) was added to each well. After incubation for
another 4 h, the absorbance was read by enzyme labeling instru-
ment with 570/630 nm double wavelength measurement. The
cytotoxicity was evaluated based on the percentage of cell survival
compared with the negative control. The final IC50 values were
calculated by the Bliss method (n ¼ 5). All the tests were repeated
in at least three independent experiments.
2.4.2. Fluorescence spectroscopy
The competitive binding study was carried out by maintaining
the ethidium bromide (EB, 2.5 mM) and CT-DNA (4 mM) in Tris-HCl
buffer (pH 7.2), and increasing the concentrations of the complexes
L and C1. The fluorescence spectra were recorded between 550 and
800 nm at room temperature with the excitation wavelength set at
530 nm. The complexes L and C1 did not emit fluorescence at the
excitation and emission wavelengths. For every addition, the
sample was shaken and allowed to stand for 5 min, and then the
fluorescence emission spectra were recorded. Further, the apparent
binding constant (Kapp) has been calculated from equation (2) [41].
2.8. Anticancer mechanism of Fe(III) complex
2.8.1. Apoptosis by flow cytometry
The apoptotic events induced by C1 were determined by
annexin V staining and PI staining according to the manufacturer's
protocol for the Annexin VꢀFITC Apoptosis Detection Kit (Abcam).
Briefly, HepG2 cells were seeded in 6-well plates at 2.0 ꢁ 105 cells/
well in 2 mL of complete DMEM and cultured for 24 h. The cells
were incubated at 5% CO2 and 37 ꢂC with the C1 at indicated con-
centrations. The HepG2 cells without the treatment were used as a
KEtBr [EB] ¼ Kapp [complex]
(2)
where [EB] ¼ 2.5
m
M and KEtBr is 1 ꢁ 107
M
ꢀ1; [complex] is the
concentration of the complex causing 50% reduction in the emis-
sion intensity of EB.
control. After 24 h, cells were harvested and resuspended in 200
mL
1 ꢁ annexin Vꢀbinding buffer. Next, 5
m
L each of annexin V and PI
2.4.3. Viscosity experiments
were added to each sample and incubated on ice for 15 min. The
samples were analyzed by flow cytometry (FACScan, Bection
Dickinson, San Jose, CA). The rate of cell apoptosis was analyzed.
The relative change in viscosity was measured using an Ubbe-
lodhe viscometer, immersed in a thermostated water-bath main-
tained at 30 0.5 ꢂC. The concentration of DNA was 200
mM and
increasing concentrations of complexes were used. Each complex
2.8.2. Acridine orange/ethidium bromide (AO/EB) double staining
solution was measured three times and an average flow time was
HepG2 cells were treated with C1 (0.2 mM and 0.4 mM) at a fixed
incubation time of 24 h, respectively. After incubation, cells were
washed twice with PBS and fixed with 4% paraformaldehyde. Then,
1/3
calculated. Data are presented as (
h
/ho)
versus the ratio of the
concentration of the complexes to CT-DNA, where ho is the viscosity

Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
357
the cells were stained with AO/EB staining solution for 5 min. After
2.8.6. Western blot analysis
washing twice with PBS, the cells of morphological observation
were obtained under a reflected fluorescence microscope (Nikon
MF30 LED, Japan).
HepG2 cells were seeded into 3.5 cm dishes for 24 h, and then
exposed to C1 at indicated concentrations for 24 h at 37 ꢂC. Cells
were harvested and washed with iceꢀcold PBS three, then were
lysed in radioimmunoprecipitation assay (RIPA) buffer. The protein
concentration of the supernatant was determined by BCA (bicin-
choninic acid) assay. Equal amounts of cellular total proteins were
separated on 10% SDSꢀpolyacrylamide gel electrophoresis and then
transferred onto poly(vinylidene difluoride) (PVDF) membranes
(Millipore, MA, USA). The PVDF membranes were blocked with 5%
(w/v) nonꢀfat milk in TBST buffer (20 mM Tris, pH 8.0,150 mM NaCl,
and 0.05% Tween 20) for 1 h. Then, the membranes were incubated
with the primary antibodies (Cell Signalling Technology and Santa
Cruz) overnight at 4 ꢂC. After a subsequent washing step, the
membrane is incubated with the appropriate secondary antibodies
conjugated with horseradish peroxidase (Cell Signalling Technol-
ogy) for 1 h at room temperature and washed for three times with
TBST. The immunoreactivity was detected using Amersham ECL Plus
(Amersham) western blotting detection reagents.
2.8.3. Cell cycle distribution analysis
Cell cycle distribution was analyzed by PI staining and flow
cytometry. HepG2 cells were exposed to C1 at the indicated con-
centrations. After 24 h incubation, the cells were collected, washed
for twice with ice-cold PBS, fixed with 70% ethanol at 4 ꢂC overnight
and treated with Rnase A for 35 min at 37 ꢂC, followed by PI staining
for 15 min in the dark. Percentage of cells in different cell cycle
phases was measured by flow cytometry using a 488 nm laser
(FACScan, Bection Dickinson, San Jose, CA).
2.8.4. Intracellular reactive oxygen species (ROS) measurements
Intracellular ROS generation was determined using
20,70ꢀdichlorodihydroꢀfluorescein diacetate (H2DCFꢀDA) (Beyo-
time Institute of Biotechnology, Haimen, China). Briefly, 2 mL
HepG2 cells (2 ꢁ 105 cells per well) were seeded in 6-well plates
and cultured for 24 h. Then cells were incubated with C1 at indi-
cated concentrations for 24 h at 37 ꢂC. Cells were collected for flow
cytometric assessment. The fluorescence intensity was monitored
with excitation wavelength at 488 nm and emission wavelength at
525 nm.
2.9. Statistical analysis
All biological experiments were repeated 3 to 5 times. Student's
t-test was applied to evaluate the significance of differences
measured. The representative results were expressed as mean
standard deviations (SD) and considered to be significant when
p < 0.05.
2.8.5. The change of mitochondrial membrane potential assay
Mitochondrial membrane potential was measured by a fluo-
rescent dye JCꢀ1 (Beyotime, Haimen, China). HepG2 cells were
treated with different concentrations C1 in 6ꢀwell plates and PBS
was used as a control. Cells were harvested and washed twice with
PBS after 24 h of incubation. Subsequently, cells were stained with
3. Results and discussion
3.1. L ligand and Fe(III) complex characterization
The structure of L was solved in the space group P21/c and
monoclinic molecule with one independent molecule in a unit cell.
Fig. 1A shows the molecular structure of L, which is intersectant
with the dihedral angle between the phenyl ring and pyridine ring
of 49.4ꢂ. The monomeric units of L are arranged in a dimer fashion
1 mL of JCꢀ1 stock solution (10
mg/mL). Assays were initiated by
incubating HepG2 cells with JCꢀ1 for 30 min at 37 ꢂC in the dark
and the fluorescence of separated cells was detected with a flow
cytometer (FACScan, Bection Dickinson, San Jose, CA).
Fig. 1. (A) Ortep drawings of L with 50% probability displacement ellipsoids. Hydrogen atoms and H2O are omitted for clarity. (B) The NeH/S interactions in L (Symmetry code:
i ¼ 1 ꢀ x, 2 ꢀ y, ꢀ z). (C) The local coordination environment of C1. Displacement ellipsoids are drawn at 50% probability. Hydrogen atoms and NOꢀ3 are omitted for clarity. (D) A
polymeric chain formed by Hꢀbonding interactions in C1.

358
Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
Fig. 2. (A) Cyclic voltammograms of the C1 prepared in this investigation. Sweep rate, 100 mV/s; solvent MeCN:H2O (7:3) with 0.1 M Et4NClO4. (B) Effect of the Fe(III) complex of the
L ligand on ascorbate oxidation. Results are mean SD (three experiments, (**) p < 0.01).
Fig. 3. (A) Absorption spectra of L and complex C1 (20
mM) upon the titration of CT-DNA (0e70
mM). (B) Fluorescence quenching curves of ethidium bromide bound to DNA: ligand L
and the complex C1. [DNA] ¼ 5
mM, [EB] ¼ 2.5
m
M, and [complex] ¼ 0e40 M. (C) Relative viscosity increments of CT-DNA (200
m
mM) solution bound with L and C1 with increasing
the [complex]/[DNA] ratio.

Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
359
(Fig. 1B) by the weaker N2eH2/S1i (N2/S1i ¼ 2.396 Å and the
N2eH2/S1i angle is 164.0ꢂ, symmetry code: (i) 1 ꢀ x, 2 ꢀ y, ꢀ z)
hydrogen bonds.
positive control [38]. In the study, a range of ironꢀbinding equiv-
alents ratios (the L ligand is tridentate chelators and form 1:2 iron/
ligand complex, while EDTA is hexadentate and forms 1:1 iron/
ligand complex) were examined, namely 0.1, 1, and 3. As shown in
Fig. 2B, the iron EDTA complex demonstrated the greatest activity
at increasing ascorbate oxidation. For C1 complex, a marked in-
crease in ascorbate oxidation was found upon increasing the iron-
ꢀbinding equivalents ratios from 0.1 to 1 and 3. This result is also in
agreement with the cyclic voltammetry data presented, which
demonstrated that the C1 complex, similar other thio-
semicarbazones Fe complexes [38,47], shows considerable redox
activity.
Compound C1 crystallizes in the orthorhombic system with
space group Pbca. As shown in Fig. 1C, the iron atom was pseudo-
ꢀoctahedrally coordinated by two N2S tridentate L ligand. All FeꢀX
bond distances are very similar to the related thiosemicarbazone
Schiff baseꢀiron complexes [42e44]. Here, the FeeN(pyridine)
distances [Fe1ꢀN1 ¼ 2.062 Å and Fe1ꢀN5 ¼ 2.056 Å ] are relatively
longer than the FeeN(azo) distances [Fe1eN2 ¼ 1.918 Å, and
Fe1ꢀN6 ¼ 1.913 Å], indicating the different strength of the bond
formed by each of the coordinated nitrogen atoms. The difference
in FeeN bond lengths may arise from constraints involved in
chelateꢀring formation or from the differing positions which
N(pyridine) and N(azo) occupy in the coordination polyhedron, as
well as from differential Fe/N back donation. The four
fiveꢀmembered chelate rings are planar, and the mean planes
through the first and second L ligand atoms are almost perpen-
dicular to each other. In the solid state, C1 was linked into a
3.4. DNA binding properties
DNA is the primary intracellular target of antitumor drugs, thus
the interaction between DNA and metallodrug is of paramount
importance in understanding the mechanism. In order to have an
insight on the binding propensity and binding mode, the mode and
propensity for binding of free ligand and complex C1 to calf-thymus
DNA (CT-DNA) were studied with multiple techniques such as ab-
sorption and emission spectral studies. Upon addition of DNA to the
solution of L, the absorption bands at 340 nm exhibited a hypo-
chromism of about 6.1% without any shift in the wavelength of
absorption (Fig. 3A). However, complex C1 exhibited a hypo-
chromism of about 11.6% with a hypsochromic shift of 1 nm at
446 nm (Fig. 3A), which indicated that the complex was bound to
CT-DNA via an intercalative mode. In order to affirm quantitatively
the affinity of the complexes bound to DNA, the intrinsic binding
constants (Kb) of the compounds with DNA was obtained by using
eqn. (1). The Kb values were found to be 0.83 ꢁ 104 Mꢀ1 for L and
1.16 ꢁ 104 Mꢀ1 for C1. The observed values of Kb revealed that the
ligand and the C1 complex bind to DNA via intercalative mode [48].
To further investigate the interaction mode between the Fe
complex and DNA, we carried out EB fluorescence displacement
experiments. As shown in Fig. 3A, C1 has UV absorption at excita-
tion wavelength (530 nm) and thus some inner-filter effect is
present, which results in a decrease in the observed fluorescence
intensity. This effect can be approximately corrected with the
oneꢀdimensional
polymeric
chain
by
N8eH8A$$$O3i
(N8/O3i ¼ 2.917 Å and the N8eH8A$$$O3i angle is 172.2ꢂ, sym-
metry code: (i) 0.5 ꢀ x, ꢀ0.5 þ y, z) and weaker N4eH4/O2ii
(N4/O2ii ¼ 2.955 Å and the N4eH4/O2ii angle is 164.6ꢂ, sym-
metry code: (ii) ꢀ0.5 þ x, y, 1.5 ꢀ z) hydrogen bonds (Fig. 1D).
L ligand can be represented by two tautomeric forms, the thione
form and thiol form (the transfer of the NeNHeC proton to the
sulfur atom). In the case, the CeS bonds (C7eS1 and C25eS2) of
average 1.745 Å in the compound C1 (thiol form) were longer than
this in the L ligand (C7eS1 ¼ 1.672 Å, thione form), indicating that
the thiocarbonyl moiety in the compound C1 adopted a tautomeric
form and acted as a monoꢀnegative ligand, which support thiolate
formation in the ligands on complexation. To shed light on the
energies involved in the process of conformational change of L
ligand in methanol, quantum chemical calculations were per-
formed. The calculations show that withdrawal of H from the
NeNHeC fragment of L and its location on the S atom provokes an
energy increase of 13.73 kcal/mol. This is in agreement with pre-
viously obtained result on similar systems [45]. The result implies
that in solution the L ligand exists almost exclusively in the thione
form and that forming the HSꢀC tautomer prior to complexation is
necessary.
3.2. Cyclic voltammetry
The redox activity has been linked with cytotoxicity of com-
plexes of the thiosemicarbazones, particularly their Fe complexes
[31,46]. By use of cyclic voltammetry, the FeII/III redox potentials
were determined in a mixture of H2O and MeCN (3:7 v/v) to ensure
sufficient solubility. From inspection of Fig. 2A, the electrochem-
istry is complicated, and the voltametric waves are only qua-
siꢀreversible. In relation to biological activity, the measured FeII/III
redox potentials of [Fe(L)2]þ, like that of other similar thio-
semicarbazones Fe complexes such as [Fe(DpT)2]þ, which lie within
the range accessible to both cellular oxidants and reductants, and
both the ferric and the ferrous forms are chemically stable [30].
Thus, C1 could redox cycle to generate intracellular ROS.
3.3. Ascorbate oxidation studies
The cyclic voltammetry experiments reported above suggest
that the C1 complex can undergo facile interconversion between
the ferrous and ferric states. Hence, it was important to confirm the
ability of the C1 complex to mediate the oxidation of a physiological
substrate. Thus, the oxidation of ascorbate catalyzed by the C1
complex was assessed, and EDTA was also included to act as
Fig. 4. Agarose gel electrophoresis patterns for the cleavage of pBR322 DNA by com-
plex Fe(NO3)3, L and C1 (from top to bottom) at pH 7.0 at 37 ꢂC. Lane 1: DNA alone;
Lanes 2e4: DNA with complexes at the concentrations of 10, 15, 20 mM, respectively.

360
Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
l
l
convenient formula Fcorr ¼ Fobs ꢁ 10A( ex)/2 ꢁ 10A( em)/2 (where Fcorr
the stronger binding ability toward DNA than free ligand L, which is
consistent with the electronic absorption spectral results. More-
over, the values of Kapp were less than the binding constant of the
is the correct fluorescence, Fobs the measured fluorescence in-
tensity, and A(lex) and A(lem) the values of absorbance at the
emission and excitation wavelength, respectively) [49,50]. The
emission spectra of the EBeDNA system in the absence and pres-
ence of L and C1 (corrected) are shown in Fig. 3B. The addition of
the complexes results in a decrease of the fluorescence intensity of
the emission band of the DNA-EB system indicate that the com-
plexes would efficiently compete with EB for intercalative binding
sites on DNA by replacing EB. According to eqn. (2), Kapp values are
evaluated as 1.31 ꢁ 105 Mꢀ1 and 3.67 ꢁ 105 Mꢀ1 for L and complex
C1, respectively. The higher values of Kapp for complex C1 indicated
classical intercalators and metallointercalators (107 ꢀ1), espe-
M
cially for L, which suggested that the interaction between the
planar aromatic L and DNA was a moderate intercalative mode.
To further investigate the binding nature of the complexes with
DNA, viscosity measurements on the solutions of DNA incubated
with the L and complex C1 have been carried out. Viscosity mea-
surement is regarded as the most critical method to detect the
binding mode between DNA and small agents, which can monitor
the DNA length changes, classical intercalation leads to an
Table 3
ICa50
(mM) of HL and C1 for the selected cells for 48 h.
Compound
Cell growth inhibition, IC50 SD (mM)
HepG2
HeLa
HL-7702
MCF-7
MCF-7/ADR
RFb
HL
>40
>40
>40
>40
>40
C1
0.23 0.02
22.47 1.72
0.46 0.02
27.57 1.81
1.62 0.12
10.39 1.23
1.1 0.07
20.7 1.91
2.78 0.17
1.38 0.05
21.39 1.68
>40
1.2
1.1
>14
Cisplatin
Doxorubicin
aIC50 values are presented as the mean SD from three separated experiments. RFb ¼ IC50 (MCF-7)/IC50 (MCF-7/ADR).
Fig. 5. (A) Immunoblotting analysis of proteins (
(B) Percentage expression levels of
cytometry after HepG2 cells were treated with vehicle and C1 at indicated concentrations for 24 h and stained with H2DCFDA. (D) Quantification of the flow cytometric results in (C)
showing the percentage of cells with increased intracellular DCF oxidation compared to control cells. Results are the mean SD (n ¼ 3): (*) p < 0.05, (**) p < 0.01, (***) p < 0.001.
g
H2AX, phosꢀCHK1, phosꢀCHK2 and phosꢀp53) related to the DNA damage pathway. ꢀActin was assessed as a loading control.
b
g
H2AX, phosꢀCHK1, phosꢀCHK2 and phosꢀp53. The percentage values are those relative to the control. (C) Analysis of ROS levels by flow

Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
361
Fig. 6. (A) Cell cycle contributions resulting from treatment with C1 (0.2 and 0.4
of CDK2 and cyclin A.
Results are the mean SD (n ¼ 3): (**) p < 0.01.
mM) for 24 h. (B) Representative Western blots of the effects of C1 on the protein expression levels
ꢀActin was assessed as a loading control. (C) Percentage expression levels of CDK2 and cyclin A. The percentage values are those relative to the control.
b
increased DNA viscosity and groove binding or electrostatic mode
results in not apparent alteration in DNA viscosity [51,52]. The
relative viscosities of CT-DNA in the presence of the complexes are
shown in Fig. 3C. Upon continuous addition of the complexes, the
relative viscosity of DNA increases efficiently. The result suggests
complex could bind to DNA by intercalation, which rules out the
groove binding interactions or electrostatic between DNA and
complex [53].
complexation on the antiproliferative efficacy of L ligand with
Fe(III), we performed MTT assays using various cancer cells and one
normal liver cell HL-7702. In comparison with the free ligand, the
C1 demonstrated significantly increased antiproliferative activity
after a 48 h incubation, while no pronounced antiproliferative ef-
fect was observed with Fe(NO3)3 alone (Table 3). The result may be
that relative to the ligand, the Fe complex is more able to cross the
cell membrane because of its predicted greater lipophilicity [31,56].
In addition, in MCF-7/ADR cells, the resistance factor values for the
C1 was roughly 14 times lower than that of Doxorubicin, suggesting
that the Fe complex is not potential multidrug-resistant substrates.
3.5. pBR322 plasmid DNA cleavage activity
Most previously reported metal complexes are capable of cata-
lyzing the cleavage of DNA [54]. We therefore examined whether L
and complex C1 exhibited DNA-cleaving activities using an agarose
gel electrophoresis assay. It can be seen from Fig. 4 that complex C1
could efficiently relax the supercoiled form (Form I) of DNA into an
open circular form (Form II) in a concentration-dependent fashion.
It should be noted that under the same condition, free ligand L and
Fe(NO3)3 did not exhibit any cleavage activity. Thus, the cleavage
properties of the present complexes are attributed to the proximity
of the DNA-bound complexes to the deoxyl ribose rings and the
coordination geometries of complexes, as understood from the
spectral and electrochemical properties [55].
3.7. Possible anticancer mechanism of the Fe(III) complex
Most previously reported metal compounds are assumed to
induce cell death through DNA damage [57]. Therefore, immuno-
blotting analyses were conducted to monitor changes in expression
of biomarkers related to the DNA damage pathway. HepG2 cells
incubated for 48 h with C1 showed a marked increase in expression
of the phosphorylated forms of H2AX (gH2AX), CHK1, CHK2, and
p53 (Ser15) proteins (Fig. 5A and B), indicating that C1 kill cells by
damaging DNA [58e60].
Excess intracellular ROS, including H2O2, Oꢀ2 and HOꢀ, could
cause DNA damage and trigger p53 activation [61e63]. To confirm
the redox activity of C1 complex as suggested by the electro-
chemical and ascorbate oxidation studies, the ability of this com-
plex to catalyze the production of intracellular ROS in HepG2 cells
was assessed using the fluorescent DCF probe and flow cytometry.
Flow cytometry analysis results show that C1 can significantly
3.6. Cytotoxic activity of the Fe(III) complex
Previous studies have illustrated that complexation of thio-
semicarbazones chelators with metal ions can result in marked
changes in biological activity [28,31,32]. To determine the effect of

362
Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
Fig. 7. (A) Representative dot plots of PI and annexinV double staining on the HepG2 cells in the presence of the indicated concentrations of C1 for 24 h. (B) Representative images of
AO/EB double stained HepG2 cells after treatment with C1 at the indicated concentrations for 24 h. (C) Effects of C1 on mitochondrial membrane potential analyzed by JCꢀ1 staining
and flow cytometry. HepG2 cells were treated with vehicle, and C1 at the indicated concentrations indicated for 24 h.
increase the intracellular ROS levels (Fig. 3C). The DCF fluorescence
peaks for each concentration evaluated were quantified. These data
reveal a significant (p < 0.001) increase in H2DCF oxidation to
Cell cycle arrest and apoptosis are two of the most common
cellular response to DNA damage [66,67]. Therefore, flow cyto-
metric analysis was carried out to examine the effects of C1 on
HepG2 cell cycle distribution. As depicted in Fig. 6A, comparison
with untreated control, the percentage of G2/Mꢀphase decreases
to 8.58 1.5% and 2.78 2.3% 13.2 at C1 concentrations of 0.2 and
358 12% and 540 15% of control cells at 0.2 and 0.4
mM,
respectively (Fig. 5D). However, no significant increase in H2DCF
oxidation was observed with the same concentrations of L ligand
and Fe(III) alone over this same incubation period. These results
suggest that C1 complex' marked antitumor activity may due to
their ability to generate cytotoxic ROS during a redox cycle [64]. In
addition, cancer cells exhibit greater ROS stress than normal cells
[65]. Therefore, elevation of the ROS levels by metalꢀdrugs may be
a way to selectively kill cancer cells without causing significant
damage to normal cells.
0.4
mM, respectively. The percentages of cells in the S phase
increased to 38.90 3.3% and 46.20 2.1%, at the same C1 con-
centrations. These results indicated that C1 induced a dose-
dependent Sꢀphase arrest of HepG2 cells. Progression through
the cell cycle is tightly controlled by cyclins complexes and
cyclinꢀdependent kinases (CDKs) at different phases [68]. To
explore the molecular mechanisms of C1-induced Sꢀphase block,

Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
363
chromatin condensation, or apoptoticꢀbody formation were
further examined by AO/EB staining. As shown in Fig. 7B, HepG2
cells treated with C1 display a doseꢀdependent morphological
change. These results provide validation for the apoptosis pathway.
Mitochondrial dysfunction plays an important role in triggering
apoptosis, which is demonstrated by several key events such as the
reduction of mitochondrial membrane potential (Djm), the onset of
mitochondrial permeability transition, and the release of cyto-
chrome c [72]. To establish whether apoptosis induced by C1 is
related to mitochondrial dysfunction, Djm was evaluated by flow
cytometry using JCꢀ1 staining. Treatment of HepG2 cells with C1
cause a decrease the Djm level in a concentrationꢀdependent
manner (Fig. 7C), which confirmed the activation of mitochondria
mediated apoptosis. The Bcl-2 family proteins have been described
as key regulators of Djm [73]. Previously, many studies have
demonstrated that metal complexes induced mitochondria medi-
ated apoptosis in cancer cells through regulation of Bcl-2 family
members [74e77]. In the present study, Western blot analysis
revealed that C1 upregulated the expression of Bax and Bad
(proapoptosis Bcl-2 family proteins), and suppressed the expres-
sion of Bcl-xl and Bcl-2 (prosurvival Bcl-2 family proteins)
(Fig. 8AꢀC). The ratio of Bcl-xl/Bad and Bcl-2/Bax is decreased,
leading to the release of apoptogenic factors, such as cytochrome c.
Subsequently, cytochrome c caused activation of caspase-3, and -9.
These results indicate that C1 can induce mitochondria mediated
apoptosis in HepG2 cells through regulating the expression of Bcl-2
family proteins.
4. Conclusion
In this study, we synthesized an
aꢀNꢀheterocyclic thio-
semicarbazone Fe(III) complex, and investigated its anticancer ac-
tivity and explored the underlying molecular mechanisms. DNA
binding and cleavage experiments indicated that the Fe(III) com-
plex exhibited the high pBR322 DNA cleaving ability and the high
binding affinity toward CT-DNA. The Fe complexꢀinduced DNA
damage and ROS overproduction induces phosphorylation of p53
and upregulate the expression of cyclin A/CDK2, leading to S phase
arrest. On the one hand, the activation of p53 induces the mito-
chondrial dysfunction through regulating the expression of Bclꢀ2
family proteins, and then triggered the mitochondrial release of
apoptogenic factors, like cytochrome c and caused cleavage of
caspase family proteases such as caspase-3 and -9, finally resulting
in apoptosis. On the basis of these results, we suggest that the
thiosemicarbazone Fe complex with multiꢀdeath pathways may be
a candidate for further evaluation as a chemopreventive and
chemotherapeutic agent for human cancers.
Fig. 8. (A) Western blot analysis of cleaved Caspaseꢀ3, cleaved Caspaseꢀ9, Bclꢀ2,
Bclꢀxl, Bad, Bax and cytochrome c in HepG2 cells treated with different concentrations
of C1 for 48 h. (B) Percentage expression levels of Bclꢀ2, Bclꢀxl, Bad, and Bax. (C)
Percentage expression levels of cleaved Caspaseꢀ3, cleaved Caspaseꢀ9 and Cyto-
chrome c. The percentage values are those relative to the control. Results are the
mean SD (n ¼ 3): (*) p < 0.05, (**) p < 0.01, (***) p < 0.001.
Acknowledgements
the expression of S phaseꢀspecific cell cycle regulatory proteins
(major CDK2 and cyclin A) were examined by immunoblotting
analyses. Compared with the vehicleꢀtreated control, CDK2 and
cyclin A increased significantly upon C1 treatment (Fig. 6B and C),
suggesting that activation of cyclin A/CDK2 might play a important
role in C1ꢀmediated growth Sꢀphase arrest.
This work was supported by the Natural Science Foundation of
China (31460232), Natural Science Foundation of Guangxi
(2014GXNSFDA118016), Ministry of Education of China
(CMEMR2015-A01).
Many metal cancer drugs exert their cytotoxic effects through
apoptosis, and we therefore monitored features related to this
pathway [69e71]. Using a dual Annexin V staining/PI flow cytom-
etry assay, we explored the occurrence of apoptosis in HepG2 cells
treated for 24 h with C1. As can be seen from Fig. 7A, C1 can effi-
ciently induce apoptosis in HepG2 cells. Even at a low concentra-
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2016.07.041.
References
tion of 0.2 mM, C1 prompts a large population of cells to undergo
early (18.1%) and late (8.75%) stage apoptosis. To gain more evi-
dence for the apoptotic cell death induced by C1 complex, the
apoptosisꢀinducing properties such as cytoplasmic shrinkage,
[1] K.D. Mjos, C. Orvig, Chem. Rev. 114 (2014) 4540e4563.
[2] L. Kelland, Nat. Rev. Cancer 7 (2007) 573e584.
[3] G. Gasser, I. Ott, N. MetzlerꢀNolte, J. Med. Chem. 54 (2011) 3e25.

364
Y. Gou et al. / European Journal of Medicinal Chemistry 123 (2016) 354e364
[4] M.A. Jakupec, M. Galanski, V.B. Arion, C.G. Hartinger, B.K. Keppler, Dalton
Trans. 2 (2008) 183e194.
[41] M. Lee, A.L. Rhodes, M.D. Wyatt, S. Forrow, J.A. Hartley, Biochemistry 32
(1993) 4237e4245.
[5] P.J. Dyson, G. Sava, Dalton Trans. (2006) 1929e1933.
[42] E.W. Yemeli, G.R. Blake, A.P. Douvalis, T. Bakas, G.O. Alberda van Ekenstein,
P.J. van Koningsbruggen, Chem. Eur. J. (2010), http://dx.doi.org/10.1002/
chem.201002100.
[6] S.M. Schmitt, M. Frezza, Q.P. Dou, Front. Biosci. Sch. Ed. 4 (2012) 375e391.
ꢀ
[7] J.M. Perez, M.A. Fuertes, C. Alonso, C. Navarro-Ranninger, Crit. Rev. Oncol.
Hematol. 35 (2000) 109e120.
[8] C.X. Zhang, S.J. Lippard, Curr. Opin. Chem. Biol. 7 (2003) 481e489.
[9] N.C.N. Andrews, Engl. J. Med. 341 (1999) 1986e1995.
[10] T.C. Iancu, H. Shiloh, A. Kedar, Cancer 61 (1988) 2497e2502.
[11] N.T. Le, D.R. Richardson, Biochim. Biophys. Acta 1603 (2002) 31e46.
[12] D.R. Richardson, E. Baker, Biophys. Acta 1053 (1990) 1e12.
[13] E.H. Morgan, Mol. Asp. Med. (1981) 1e123.
[14] J.W. Larrick, P. Cresswell, J. Supramol. Struct. 11 (1979) 579e586.
[15] H.L. Elford, M. Freese, E. Passamani, H.P. Morris, J. Biol. Chem. 245 (1970)
5228e5233.
[43] V. Amendola, M. Boiocchi, L. Fabbrizzi, L. Mosca, Chem. Eur. J. 14 (2008)
9683e9696.
[44] M. Li, J. Zhou, H. Zhao, C. Chen, J. Wang, J. Coord. Chem. 62 (2009) 1423e1429.
[45] M. Krasowska, A. Kochel, A. Filarowski, CrystEngComm 12 (2010) 1955e1962.
[46] D.R. Richardson, D.S. Kalinowski, V. Richardson, P.C. Sharpe, D.B. Lovejoy,
M. Islam, P.V. Bernhardt, J. Med. Chem. 52 (2009) 1459e1470.
[47] C. Stefani, G. Punnia-Moorthy, D.B. Lovejoy, P.J. Jansson, D.S. Kalinowski,
P.C. Sharpe, P.V. Bernhardt, D.R. Richardson, J. Med. Chem. 54 (2011)
6936e6948.
[48] J. Olmsted, D.R. Kearns, Biochemistry 16 (1977) 3647e3654.
[49] L. Stella, A.L. Capodilupo, M. Bietti, Chem. Commun. (2008) 4744e4746.
[50] M. van de Weert, L. Stella, J. Mol. Struct. 998 (2011) 144e150.
[51] B.L. Fei, Z.X. Huang, W.S. Xu, D.D. Li, Y. Lu, W.L. Gao, Y. Zhao, Y. Zhang, Q.B. Liu,
J. Photochem. Photobiol. B160 (2016) 43e52.
[52] Y. Zhang, J.Y. Wen, M.H. Mahmood, X.L. Wang, B.B. Lv, X. Ying, H. Wang, L.N. Ji,
H.Y. Liu, Luminescence 30 (2015) 1045e1054.
[53] J.L. Li, L. Jiang, B.W. Wang, J.L. Tian, W. Gu, X. Liu, S.P. Yan, J. Inorg. Biochem.
145 (2015) 19e29.
[54] M. Chen, X.Y. Tang, S.P. Yang, H.H. Li, H.Q. Zhao, Z.H. Jiang, J.X. Chen,
W.H. Chen, Dalton Trans. 44 (2015) 13369e13377.
[55] M. Alagesan, N.S.P. Bhuvanesh, N. Dharmaraj, Eur. J. Med. Chem. 78 (2014)
281e293.
[16] L. Thelander, P. Reichard, Annu. Rev. Biochem. 48 (1979) 133e158.
[17] G.Y. Lui, Z. Kovacevic, V. Richardson, A.M. Merlot, D.S. Kalinowski,
D.R. Richardson, Oncotarget 6 (2015) 18748e18779.
[18] E. GumiennaꢀKontecka, M. PyrkoszꢀBulska, A. Szebesczyk, M. Ostrowska,
Curr. Med. Chem. 21 (2014) 3741e3767.
[19] D.J. Lane, T.M. Mills, N.H. Shafie, A.M. Merlot, R. Saleh Moussa, D.S. Kalinowski,
Z. Kovacevic, D.R. Richardson, Biochim. Biophys. Acta 1845 (2014) 166e181.
[20] D.R. Richardson, P.C. Sharpe, D.B. Lovejoy, D. Senaratne, D.S. Kalinowski,
M. Islam, P.V. Bernhardt, J. Med. Chem. 49 (2006) 6510e6521.
[21] Y. Yu, D.S. Kalinowski, Z. Kovacevic, A.R. Siafakas, P.J. Jansson, C. Stefani,
D.B. Lovejoy, P.C. Sharpe, P.V. Bernhardt, D.R. Richardson, J. Med. Chem. 52
(2009) 5271e5294.
[22] D.S. Kalinowski, D.R. Richardson, Pharmacol. Rev. 57 (2005) 547e583.
[23] H. Huang, Q. Chen, X. Ku, L. Meng, L. Lin, X. Wang, C. Zhu, Y. Wang, Z. Chen,
M. Li, H. Jiang, K. Chen, J. Ding, H. Liu, J. Med. Chem. 53 (2010) 3048e3064.
[56] P.V. Bernhardt, J. Mattsson, D.R. Richardson, Inorg. Chem. 45 (2006) 752e760.
[57] K. Suntharalingam, T.C. Johnstone, P.M. Bruno, W. Lin, M.T. Hemann,
S.J. Lippard, J. Am. Chem. Soc. 135 (2013) 14060e14063.
[58] S. Matsuoka, G. Rotman, A. Ogawa, Y. Shiloh, K. Tamai, S.J. Elledge, Proc. Natl.
Acad. Sci. U.S.A. 97 (2000) 10389e10394.
ꢀ
[24] E.A. Enyedy, M.F. Primik, C.R. Kowol, V.B. Arion, T. Kiss, B.K. Keppler, Dalton
Trans. 40 (2011) 5895e5905.
[25] R.W. Brockman, J.R. Thomson, M.J. Bell, H.E. Skipper, Cancer Res. 16 (1956)
167e170.
[59] J.Y. Ahn, J.K. Schwarz, H. PiwnicaꢀWorms, C.E. Canman, Cancer Res. 60 (2000)
5934e5936.
[26] C.M. Nutting, C.M.L. van Herpen, A.B. Miah, S.A. Bhide, J.P. Machiels, J. Buter,
C. Kelly, D. de Raucourt, K.J. Harrington, Ann. Oncol. 20 (2009) 1275e1279.
[27] J.F. Zeidner, J.E. Karp, A.L. Blackford, B.D. Smith, I. Gojo, S.D. Gore, M.J. Levis,
H.E. Carraway, J.M. Greer, S.P. Ivy, K.W. Pratz, M.A. McDevitt, Haematologica
99 (2014) 672e678.
[28] P.J. Jansson, P.C. Sharpe, P.V. Bernhardt, Des R. Richardson, J. Med. Chem. 53
(2010) 5759e5769.
[29] J. Yuan, D.B. Lovejoy, D.R. Richardson, Blood 104 (2004) 1450e1458.
[30] P.V. Bernhardt, P.C. Sharpe, M. Islam, D.B. Lovejoy, D.S. Kalinowski, Des
R. Richardson, J. Med. Chem. 52 (2009) 407e415.
[60] S.Y. Shieh, M. Ikeda, Y. Taya, C. Prives, Cell 91 (1997) 325e334.
[61] I.R. Indran, M.P. Hande, S. Pervaiz, Cancer Res. 71 (2011) 266e276.
[62] K.ꢀR. Park, D. Nam, H.ꢀM. Yun, S.ꢀG. Lee, H.ꢀJ. Jang, G. Sethi, S.K. Cho,
K.S. Ahn, Cancer Lett. 312 (2011) 178e188.
[63] Z. Luo, L. Yu, F. Yang, Z. Zhao, B. Yu, H. Lai, K.H. Wong, S.M. Ngai, W. Zheng,
T. Chen, Metallomics 6 (2014) 1480e1490.
[64] D.B. Lovejoy, P.J. Jansson, U.T. Brunk, J. Wong, P. Ponka, D.R. Richardson,
Cancer Res. 71 (2011) 5871e5880.
[65] D. Trachootham, J. Alexandre, P. Huang, Nat. Rev. Drug. Discov. 8 (2009)
579e591.
[31] D.S. Kalinowski, Y. Yu, P.C. Sharpe, M. Islam, Y.T. Liao, D.B. Lovejoy, N. Kumar,
P.V. Bernhardt, D.R. Richardson, J. Med. Chem. 50 (2007) 3716e3729.
[32] P.V. Bernhardt, P.C. Sharpe, M. Islam, D.B. Lovejoy, D.S. Kalinowski,
D.R. Richardson, J. Med. Chem. 52 (2009) 407e415.
[33] G.M. Sheldrick, SHELXTLV5.1, Software, Reference, Manual, BrukerAXS, Inc,
Madison, W.I. U.S.A, 1997.
[34] M.W. Schmidt, K.K. Baldridge, J.A. Boatz, S.T. Elbert, M.S. Gordon, J.H. Jensen,
S. Koseki, N. Matsunaga, K.A. Nguyen, S.J. Su, T.L. Windus, M. Dupuis,
J. Montgomery, J. Comput. Chem. 14 (1993) 1347e1363.
[66] H. Lai, Z. Zhao, L. Li, W. Zheng, T. Chen, Metallomics 7 (2015) 439e447.
[67] R. Sinha, K. Ei-Bayoumy, Curr. Cancer Drug Targets 4 (2004) 13e28.
[68] J. Bloom, F.R. Cross, Nat. Rev. Mol. Cell Biol. 8 (2007) 149e160.
[69] E.R. Jamieson, S.J. Lippard, Chem. Rev. 99 (1999) 2467e2498.
[70] D. Wang, S.J. Lippard, Nat. Rev. Drug Discov. 4 (2005) 307e320.
[71] U. Fischer, K. Schulze-Osthoff, Cell Death Differ. 12 (Suppl. 1) (2005) 942e961.
[72] C. Wang, R.J. Youle, Annu. Rev. Genet. 43 (2009) 95e118.
[73] S. Cory, J.M. Adams, Nat. Rev. Cancer 2 (2002) 647e656.
[74] N. Wang, Y. Feng, L. Zeng, Z. Zhao, T. Chen, ACS Appl. Mater. Interfaces 7
(2015) 14933e14945.
[35] A.D. Becke, J. Chem. Phys. 98 (1993) 5648e5652.
[36] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B37 (1988) 785e789.
[37] J. Tomasi, B. Mennucci, E. Cances, Theochem 464 (1999) 211e226.
[38] C. Stefani, P.J. Jansson, E. Gutierrez, P.V. Bernhardt, D.R. Richardson,
D.S. Kalinowski, J. Med. Chem. 56 (2013) 357e370.
[75] T. Chen, Y. Liu, W.J. Zheng, J. Liu, Y.S. Wong, Inorg. Chem. 49 (2010)
6366e6368.
[76] Z. Deng, L. Yu, W. Cao, W. Zheng, T. Chen, Chem. Commun. 51 (2015)
2637e2640.
[39] T.B. Chaston, D.R. Richardson, J. Biol. Inorg. Chem. 8 (2003) 427e438.
[40] M. Alagesan, N.S.P. Bhuvanesh, N. Dharmaraj, Dalton Trans. 42 (2013)
7210e7223.
[77] J. Qi, S. Liang, Y. Gou, Z. Zhang, Z. Zhou, F. Yang, H. Liang, Eur. J. Med. Chem. 96
(2015) 360e368.