Anticancer activity of the thiosemicarbazones that are based on di-2-pyridine ketone and quinoline moiety

Article abstract of DOI:10.1016/j.ejmech.2019.03.027

Thiosemicarbazones (TSC) are a subclass of iron-chelating agents that are believed to have an anticancer activity. The high potential for the application of this compound class can be illustrated by a fact that three TSC have entered clinical trials. The ability to chelate metal ions results in several biochemical changes in the cellular metabolism and growth. An important factor that determines the antitumor activity of TSC is a level of iron regulatory proteins and the antioxidant potential that is specific for each type of cancer cell. However, despite the increasing interest in TSC, their mechanism of anticancer activity is still unclear. For a more effective and rational design, it is crucial to determine and describe the abovementioned issues. In this report, we describe a series of new TSC that are designed on the four main structural scaffolds. The anticancer activity of these compounds was evaluated against a panel of cancer cell lines including colon and breast cancers and gliomas. Special attention was paid to the metal-dependent proteins. The impact of the tested TSC on the cell cycle and redox homeostasis was also determined. These results confirm a p53-independent mechanism of apoptosis.

Full text of DOI:10.1016/j.ejmech.2019.03.027

Accepted Manuscript
Anticancer activity of the thiosemicarbazones that are based on di-2-pyridine ketone
and quinoline moiety
Anna Mrozek-Wilczkiewicz, Katarzyna Malarz, Marta Rejmund, Jaroslaw Polanski,
Robert Musiol
PII:
S0223-5234(19)30245-4
DOI:
https://doi.org/10.1016/j.ejmech.2019.03.027
EJMECH 11197
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 18 December 2018
Revised Date: 7 March 2019
Accepted Date: 11 March 2019
Please cite this article as: A. Mrozek-Wilczkiewicz, K. Malarz, M. Rejmund, J. Polanski, R.
Musiol, Anticancer activity of the thiosemicarbazones that are based on di-2-pyridine ketone and
quinoline moiety, European Journal of Medicinal Chemistry (2019), doi: https://doi.org/10.1016/
j.ejmech.2019.03.027.
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

ACCEPTED MANUSCRIPT
Anticancer activity of the thiosemicarbazones that are based on di-2-pyridine
ketone and quinoline moiety
Anna Mrozek-Wilczkiewicza*, Katarzyna Malarza, Marta Rejmundb, Jaroslaw Polanskib and
Robert Musiolb
aA. Chelkowski Institute of Physics and Silesian Center for Education and Interdisciplinary
Research, University of Silesia, Chorzow, Poland
bInstitute of Chemistry, University of Silesia, Katowice, Poland
*anna.mrozek-wilczkiewicz@us.edu.pl
Keywords: Thiosemicarbazones, anticancer activity, iron chelators, reactive oxygen species,
apoptosis, cell cycle inhibition
ABSTRACT
Thiosemicarbazones (TSC) are a subclass of iron-chelating agents that are believed to
have an anticancer activity. The high potential for the application of this compound class can
be illustrated by a fact that three TSC have entered clinical trials. The ability to chelate metal
ions results in several biochemical changes in the cellular metabolism and growth. An
important factor that determines the antitumor activity of TSC is a level of iron regulatory
proteins and the antioxidant potential that is specific for each type of cancer cell. However,
despite the increasing interest in TSC, their mechanism of anticancer activity is still unclear.
For a more effective and rational design, it is crucial to determine and describe the
abovementioned issues. In this report, we describe a series of new TSC that are designed on
the four main structural scaffolds. The anticancer activity of these compounds was evaluated
against a panel of cancer cell lines including colon and breast cancers and gliomas. Special
attention was paid to the metal-dependent proteins. The impact of the tested TSC on the cell
cycle and redox homeostasis was also determined. These results confirm a p53-independent
mechanism of apoptosis.
1. Introduction
1

ACCEPTED MANUSCRIPT
Thiosemicarbazone derivatives (TSC) are commonly known for their wide biological
activity, which includes their anticancer, antimicrobial and antiviral properties [1]. Interest in
this family of compounds has increased significantly over the years from the very first report
in 1950 to the thousands of papers that were published in 2017. Currently, the anticancer
activity of TSC is being most extensively investigated and several drug candidates are
entering the clinical phase [2]–[4] (NCT02688101, NCT00004213, NCT02433626). A
leading TSC compound, Triapine (3-AP), was carefully investigated in clinical phase I and II
trials. The research indicated that this TSC alone or in a combination with gemcitabine,
cisplatin or radiation effectively inhibits the growth of different types of tumors (NSC
#663249, IND # 68338). Surprisingly, despite such dynamic progress and the many papers
that describe the molecular mechanism of the action of this family of compounds, there are
still many gaps in the complete understanding of their high anticancer activity.
The inhibition of cancer cell proliferation that is caused by TSC is connected with
several factors such as the deactivation of the ribonucleotide reductase (RR) enzyme [5], cell
cycle inhibition [6], reactive oxygen species (ROS) generation [7], the chelation of important
metal ions [8] and its influence on the proteins that are crucial in life and death processes of
cells [9], [10]. The issue is complex enough to explain consistent mechanism of action or even
general remarks on target specificity. On the other hand, individual features of the cancer
types may change the cellular reaction to TSC and the outcome of therapy. Other less
frequently addressed features that affect the landscape of TSC activity are their own structural
features. This is more evident in a wide group of iron chelators in which the nature of the
complexes determines the susceptibility to the Fenton reaction and thus the generation of
ROS. Thiosemicarbazones produce ROS by switching between the different oxidation states
of the metal ions in the redox reaction. For example, iron ions exist in the form Fe2+ and Fe3+
and these forms participate in the redox reaction. The binding of the metal ion by a ligand
(TSC), that does not saturate its coordination sites may allow or even facilitate the redox
cycle. This ability could be changed by the donor atoms that form a coordination bond with
the metal [11]. This is also why fully coordinating chelators such as DFO do not generate
ROS [12], [13]. Lovejoy et al. conducted an experiment with the preincubation of cells with a
strong chelator that binds the metal ions in a cell. After the incubation with the active TSC,
there was no cytotoxic effect, which means that metal chelation is essential for the activity of
TSC [14].
With the expansion of the TSC class, it has become clear that this group is
heterogeneous enough to reveal some distinct variability in the mechanism of action. Prime
2

ACCEPTED MANUSCRIPT
examples are 3-AP and Dp44mT whose different mechanisms were published [15]. The
debate is still ongoing as a topoisomerase inhibitory activity was suggested for Dp44mT [6],
which was previously verified [16]. As we have actively worked in this scientific field for
some time, we found it quite confusing and worth putting more effort into clarifying this
situation. Therefore, in the present work, we selected seventeen newly synthesized TSC (Fig.
1) from four of the most frequently described groups and evaluated their activity profile.
Compounds (1a-e), which have a dipyridyl moiety, are among the most active TSC [17], [18].
A second important group are Triapine and its analogs, which are built on one pyridine ring
(2a-c) [2], [8]. Quinoline (3a-h) and quinazoline (3i, 3j) derivatives have been developed on
the basis of their multicenter chelating ability and have also provided several active agents
[19]–[22].
NH2
N
N
N
N
N
N
N
N
N
N
N
N
N
HN
HN
HN
HN
S
S
NH2
R
S
S
N
R
R
1a; R = -N(Me)2 (Dp44mT)
1b; R = thiomorpholine
1c; R = 3,4-di-Cl
1d; R = 4-NO2
1e; R = 4-CN
3-AP
2a; R = 4-OMe
2b; R = 4-F
2c; R = 4-NO2
N
N
N
N
N
R2
N
N
N
N
N
HN
HN
HN
HN
S
N
S
N
S
N
S
S
N
N
N
N
R1
F
F
F
R
3a
3b; R1 = 4-CN, R2 = H
3h
3i; R = 3,4-di-Cl
3j; R = 4-OMe
3c; R1 = 3,4-di-Cl, R2 = H
3d; R1 = 4-OMe, R2 = H
3e; R1 = 4-NO2, R2 = H
3f; R1 = 4-NO2, R2 = OH
3g; R1 = 4-CF3, R2 = H
Figure 1: Structures of the studied compounds.
2. Results and discussion
3

ACCEPTED MANUSCRIPT
2.1. Chemistry
The synthesis of the TSC was carried out according to known procedures. Compounds
1a-e, which were built on a dipyridyl moiety, were obtained from ketone and appropriate
amines as was described earlier [20]. The pyridine thiosemicarbazones 2a-c were synthesized
from pyridine thiosemicarbazide as reported earlier [8]. The quinoline and quinoxaline
aldehydes were used to form the corresponding derivatives 3a-j.
2.2. Cytotoxicity studies
Iron chelators have gained strong interest for their high level of activity and
multitarget mechanisms of action. After an initial success, 3-AP appeared to cause side effects
such as methemoglobinemia. The next generation of TSC that are di-substituted at a terminal
amine group (N4) were extensively investigated by the Richardson’s group [23]–[25]. Among
them, the derivatives of di-2-pyridylketone were the most active with Dp44mT and an analog
with a cyclohexyl substituent in the amine group (DpC) as prime examples [26]. A series of
derivatives with alkyl and an aryl substituents in the amine group were investigated for DpT
(di-2-pyridine) and BpT (2-benzoylopyridine) thiosemicarbazones [17], [27]. Interestingly,
although an alkyl substitution is beneficial for the activity, their increasing lipophilicity has a
rather detrimental effect [10]. Consequently, larger, but still reasonably lipophilic morpholine
and phenylpiperidine (as in COTI-2) substituents, may result in active compounds [20], [28],
[29].
Our current report combines several structural groups of the TSC family that confirm
this trend. In general, all of the compounds were active against all of the cancer cell lines.
Moreover, their activity was at least two orders of magnitude higher than Triapine. For the
characterization of the antiproliferative activity of all of the newly synthesized derivatives,
1a-e, 2a-c, and 3a-j, we selected six cell lines that represent different human tumors and one
normal fibroblast line. We focused on colon (HCT 116; p53+/+ and p53-/-) and breast cancers
(MCF-7), which are some of the most common type of tumors. Additionally, we also selected
gliomas (U-251, Hs 683), which are particularly dangerous due to their aggressiveness and the
fact that they are difficult to treat. To determine the selectivity index, we selected normal
human dermal fibroblasts (NHDF). In our research, we explored the effect of the p53 tumor
suppressor protein on inhibiting cell growth. It is commonly known that more than 50% of
cancers have a mutation in the TP53 gene. P53 is a cell-protecting protein that plays the role
of a genome guard. Cancers that have a mutation in this protein are more aggressive and are
prone to drug resistance. The prognosis for these tumors is also poor [30]. In our experiments,
4

ACCEPTED MANUSCRIPT
we selected two cell lines that had alterations in TP53 gene. Namely, a colon carcinoma with
a double-deletion HCT 116 – p53-/-, which loss of the p53 function as well as a glioblastoma
U251, which expressed a point missense mutation that changed arginine to histidine at codon
273. The latter mutation has been revealed to be potentially more oncogenic for targeting the
protein to the nucleus [31]. As is presented in Table 1, all of the compounds that were
synthesized showed good activity against all cancer cell lines, which was higher than 3-AP.
The IC50 values varied from 0.01 µM (compound 2c on HCT 116 p53-/-) to values that can be
considered to be inactive (>25 µM). For better clarity, we used a color code map to
distinguish between the highly active (IC50 <1 µM, marked as red), active (IC50 1-6.25 µM,
marked as yellow) and inactive (IC50 >6.25 µM marked as grey) compounds. Their activity
against fibroblasts was used to determine the selectivity indexes (SI) of each cell line. These
data are presented in Table S1 in the Supporting Information.
IC50 < 1 µM
IC50 1 - 6.25 µM
IC50 > 6.25 µM
Table 1: Antiproliferative activity of the studied compounds.
Activity - IC50 [µM]
Comp.
HCT 116
p53+/+
HCT 116
p53-/-
MCF-7
U-251
Hs 683
NHDF
>25
1.12 ± 0.28
1.34 ± 0.34
2.33 ± 0.43
1.48 ± 0.56
1.76 ± 0.29
3-AP
0.00123±
0.00018
0.00110±
0.00006
0.00114±
0.00040
0.00114±
0.00006
0.00396±
0.00070
1a
18.88 ± 1.59
1.23·10-3 ±
0.18·10-3
1.10·10-3 ±
0.06·10-3
1.14·10-3 ±
0.40·10-3
1.14·10-3 ±
0.06·10-3
3.96·10-3 ±
0.70·10-3
(Dp44mT)
0.84 ± 0.19
0.31 ± 0.06
3.20 ± 1.32
0.23 ± 0.15
0.18 ± 0.02
0.19 ± 0.04
0.07 ± 0.02
0.16 ± 0.05
0.20 ± 0.02
0.15 ± 0.04
0.19 ± 0.04
0.28 ± 0.09
0.45 ± 0.19
0.17 ± 0.03
0.56 ± 0.28
0.48 ± 0.19
10.30 ± 1.07
12.87± 0.99
>25
1b
1c
1d
1e
0.015 ±
0.004
0.14 ± 0.02
0.019 ±
0.003
0.034 ±
0.008
12.32 ± 0.37
5.13 ± 1.24
0.17 ± 0.05
0.23 ± 0.08
0.014 ±
0.54 ± 0.10
0.015 ±
0.19 ± 0.05
0.021 ±
6.95 ± 3.65
0.51 ± 0.29
>25
>25
2a
2b
5

ACCEPTED MANUSCRIPT
0.004
0.006
0.006
0.010 ±
0.003
0.021 ±
0.006
0.036 ±
0.007
0.30 ± 0.07
4.68 ± 2.86
13.28 ± 0.56
2c
18.31 ± 0.92
0.56 ± 0.14
0.16 ± 0.05
5.85 ± 0.45
0.43 ± 0.03
0.15 ± 0.06
-
-
-
13.06 ± 1.9
>25
3a
3b
3c
0.23 ± 0.11
0.09 ± 0.01
0.18 ± 0.06
0.47 ± 0.05
6.33 ± 1.61
8.52 ± 2.14
14.74 ± 0.79
0.065 ±
0.004
1.10 ± 0.33
0.03 ± 0.01
0.05 ± 0.01
0.28 ± 0.09
0.28 ± 0.05
3d
5.10 ± 1.79
11.69 ± 1.77
0.14 ± 0.01
0.18 ± 0.04
1.06 ± 0.14
5.47 ± 1.19
0.22 ± 0.02
8.42 ± 1.33
0.07 ± 0.02
0.14 ± 0.05
0.29 ± 0.08
6.27 ± 1.95
0.31 ± 0.09
-
5.54 ± 1.49
-
14.81 ± 0.72
-
3.22 ± 0.77
>25
3e
3f
3g
3h
3i
0.07 ± 0.02
0.07 ± 0.02
0.50 ± 0.14
-
0.17 ± 0.02
0.13 ± 0.04
1.19 ± 0.32
-
0.22 ± 0.06
0.27 ± 0.06
3.62 ± 1.16
-
12.09 ± 0.61
16.66 ± 5.57
21.34 ± 1.34
>25
3j
The greatest activity among the newly synthesized compound was shown by pyridine
derivative 2c (IC50 on HCT 116 p53-/- equals 0.010 µM). The activity dramatically decreased
(30 times) for the wild-type HCT 116. Compound 2c primarily inhibited the proliferation of
the MCF-7 (IC50 = 0.021 µM), and U-251 (IC50 = 0.036 µM) cells while the Hs 683 cells
were rather resistant (IC50 = 4.68 µM). This trend (p53 influence) was observed for all of the
investigated derivatives except 3b, 3c and 3h, which had little or no difference in IC50 against
both colon carcinoma lines. This observation is particularly important because to date almost
no similar findings have been reported for TSC. To the best of our knowledge, TSC iron
chelators have not been distinguished between the wild-type and p53 null cell lines [8], [20],
[22], [32]. Only a few, sporadic quinoline-based compounds that exhibit a preference to HCT
116 p53-/- have been published to date [21]. What is even more interesting is that most of the
compounds presented in Table 1 are also highly active against U251. Only 1c, 3g and 3h
have no preference for the brain tumor cell lines. Thus, this is the first evidence that a change
in the TP53 gene in cancer cells may result in an increasing susceptibility to chemotherapy
with some TSC. The next most active compound was 2b, which also had the greatest effect on
HCT 116 p53-/- (IC50 = 0.014 µM). In this case, along with a prior observation, we observed a
decrease in the activity (12 times) in response to the incubation with HCT 116 p53+/+ as well.
6

ACCEPTED MANUSCRIPT
For this compound, we also observed a considerable inhibition of the proliferation of the
MCF-7 (IC50 = 0.015 µM) and U-251 (IC50 = 0.021 µM) cells and a moderate effect on the Hs
683 cells (IC50 = 0.51 µM). Compound 1c had a very similar pattern of the inhibition of the
proliferation of selected cell lines and, additionally, this compound was the best inhibitor of
the Hs 683 cells (IC50 = 0.17 µM). In the light of that observation it is quite interesting that
gliomas are one of the most treatment-resistant tumors.
Crystal structures show that a metal atom is complexed through three atoms of a
ligand – N (from the pyridine ring) and N and S (from the thiosemicarbazone group) [33]. The
nitrogen from the terminal amine group does not participate in chelating, but it still has an
impact on the cytotoxic activity. Double-alkylation as in Dp44mT, DpC and other highly
active TSC increases the basicity of the N4 atom or the relative lipophilicity of a molecule
[10], [34]. Stacy et al. reported that Dp44mT, but not 3-AP, penetrates the lysosomes and
forms complexes that have a high redox potential with Fe or Cu [35]. Further
permeabilization through the lysosome membrane leads to an initiation of the autophagic
pathway [36], [37]. Moreover, the importance of an electron-withdrawing group at the imine
carbon of the thiosemicarbazone moiety (as in a DpT series) has also been indicated. This has
on influence to lipophilicity, which in addition to its basic character is crucial to being
sequestrated by lysosomes [35].
Conversely, in a recent report of Hager and coworkers, a Triapine derivative
(Me2NNMe2) that was obtained via the double methylation of both amine groups appeared to
be active at nanomolar concentrations [38]. However, the mechanism of action that was
revealed consists of the inhibition of the protein disulphide-isomerase (PDI). Importantly, this
mechanism has been confirmed by in vitro assays, which showed that the Cu-complex of
Me2NNMe2, but not Triapine, effectively inhibits the PDI enzyme. This may suggest that the
dimethylamine group is crucial for the ligand-protein interaction at least for some TSC.
In our series 1 compounds, we observed that the activity can be reversely correlated
with the electron-withdrawing effect of the substituent in the phenylpiperazine moiety. The
derivative 1d with the strongest electron-accepting para-nitro group was the least active
compound in this group. Compound 1b, which is the closest derivative of our reference
Dp44mT (1a), deserves more attention as well. The substitution with a morpholine ring
caused a significant decrease in the activity of 1b against all of the cancer cell lines that were
tested. As was presented by Serda et al. the morpholine and piperazine derivatives of
Dp44mT can be active (IC50 is ten times lower) against both HCT 116 cell lines [20]. In the
current study, thiomorpholine compound 1b was significantly less active than Dp44mT
7

ACCEPTED MANUSCRIPT
against HCT 116 cell lines (700 and 180 times respectively). The detrimental effect of
heterocyclic sulphur was also visible in the quinoline-based compound 3a, which was
completely inactive in this series.
A comparison of the activity of series 1 and 2 led to the conclusion that similar
structural effects promote the cytotoxic activity in both series. The electron-withdrawing
substituents in the phenyl ring are crucial for the activity (compare the methoxy- 2a and
nitro- 2c). However, the pyridine derivatives were generally slightly more active (1d vs. 2c).
In series 3, which was based on the quinoline moiety, we observed the greatest
diversity in biological activity. The thiomorpholine derivative 3a, which opens this family
and compound 3f are the least active in this series. In turn, the conversion of the
thiomorpholine ring into a phenlypiperazine tail significantly improved its activity. Derivative
3d was the fifth most active compound among the 18 investigated TSC. IC50 fluctuated in the
range of 0.031-0.065 µM for MCF-7, U-251 and HCT 116 p53-/-, respectively. The effect of
the p53 protein on the response to treatment was the most noticeable for 3d. This compound
was 35 times more active on the HCT 116 that had a deletion of the TP53 gene than on the
wild type that had an active p53 protein (IC50 = 0.065 vs. 1.10 µM). Similarly, it also had a
high level of activity against the second p53-mutant U251 cell line. The substitution of the
OMe group on 4-CN, 3,4-di-Cl or 4-CF3 caused a loss of selectivity (3d vs. 3b, 3c, 3g, and
3h). Among the series 3 compounds, 3f and 3j (together with 3a) were the least active. This
may indicate that the OH group might be a steric hindrance for metal binding (3e vs. 3f).
Moreover, the addition of the N atom in the quinoline ring had a negative effect on its activity
(3c vs. 3i, 3d vs. 3j). This observation confirms the effect of the electron-accepting effect of
the substituents in the phenylpiperazine fragment that was observed for mono- and dipyridine
TSC. Compounds 3g and 3h, which had an electron-withdrawing group CF3, had a high level
of activity. On the other hand, the non-selective 3d that had an electron-donating methoxy
group was also active (but not against HCT 116 p53+/+). The lower level of activity of the
nitro or cyano derivatives might be explained on the basis of their lower lipophilicity
compared to the trifluoromethyl or dichloro derivatives.
The selectivity indexes for all of the selected cell lines are presented in Table S1
(Supporting Information). Despite their high level of anticancer activity, the calculated SI
were high, which makes the investigated TSC promising candidates for further research.
2.3. Cell cycle assay
8

ACCEPTED MANUSCRIPT
The interesting antiproliferative profiles prompted us to explore the mechanism of
action in more detail. The cytotoxicity results, which were discussed earlier, suggest a p53-
independent mechanism of action. To evaluate this hypothesis, we decided to perform a more
in-depth investigation of three cell lines; HCT 116 p53+/+, MCF-7 and U-251. The most active
compounds in those cells were selected. In the first step, we measured the effect of the tested
compounds on the cell cycle. The results of these experiments are presented in Fig. 2, and
Fig. S1, S2. In all of the cases, we detected the effect of the tested compounds on the cell
cycle phase, which was most pronounced in the HCT 116 cells (Fig. 2). The population of
cells decreased (compared to the control) in the G0/G1 phase after treatment with all of the
tested compounds. The decrease reached more or less 10-20%. The reverse occurred in the
case of the S phase where we observed an increase in the cell population from 17 to 23% (2c).
For the G2/M phase, we detected a similar situation – the highest growth rate was 10% for 3h,
while this was in the range 2-10% for the rest. These observations suggest that investigated
compounds inhibit the cell cycle in the S or G2/M phases.
Figure 2: Effect of the tested TSC (0.5 µM) on regulating the cell cycle in HCT 116 cells.
The histograms show the distribution of the cells in the G0/G1, S and G2/M phases of the cell
cycle for one of the experiments (A). The table shows the mean ± SD percentage of the cells
in the G0/G1, S and G2/M phases of the cell cycle from three independent experiments (B).
9

ACCEPTED MANUSCRIPT
The data were analyzed using one-way ANOVA with a Bonferroni’s post-hoc test: *p<0.05,
**p<0.01, ***p<0.001 compared to the control (C).
In the case of the next analyzed cell lines – MCF-7 and U-251, we observed a similar
but less visible trend (Fig. S1, S2). A decrease in the cell population in G0/G1 phase and an
increase in the S and G2/M phases suggest that all of the tested compounds inhibit
proliferation in the S and G2/M phases.
2.4. Annexin V binding assay
The next step was to evaluate the type of cell death. We performed the assay based on
measurements of the apoptotic cells by counting the fluorescence of FITC conjugated to
Annexin V. Only cells that had a disintegrated cell membrane emitted the green signal. For
this experiment, we selected the same three cell lines as were discussed above – HCT 116,
MCF-7 and U-251 (Fig. 3, and S3, S4). In general, we observed the strongest effect for the 3h
derivative (76.18% of the apoptotic cells for MCF-7, Fig. S3B, C and 64.81% for HCT 116,
Fig. 3B, C). In the HCT 116 cells, a large number of apoptotic cells were observed for all of
the derivatives (60.32-39.39%) and a moderate number for the rest (2a and 3e). These results
correlate well with the IC50 values for these compounds. In MCF-7, we also detected a strong
apoptotic effect (57.60-22.30%) for all of the compounds except 3e. For U-251, the
population of apoptotic cells was smaller and ranged from 30.34 to 13.33%. The smallest
population was observed for 2a and 3e (Fig. S4).
10

ACCEPTED MANUSCRIPT
Figure 3: Evaluation of the induction of apoptosis in the HCT 116 cells after a 48-h treatment
with TSC (0.5 µM). The histograms show the percentage of early and late apoptosis for one of
the experiments (A). The table shows the mean ± SD percentage of the live, early and late
apoptotic cells from three independent experiments (B). The data were analyzed using one-
way ANOVA with a Bonferroni’s post-hoc test: *p<0.05, **p<0.01, ***p<0.001 compared to
the control (C).
2.5. Immunoblotting
As was mentioned above, all of the compounds had a greater biological activity on the
cells that have a mutation or deletion of the TP53 gene. To further evaluate the hypothesis of
a p53-independent mechanism of apoptosis, we measured the expression of the proteins that
are crucial in the proliferation and activation of cell death. The levels of p53, p21, cytochrome
c, cdc2, cyclin D1 and PARP were identified using a Western blot assay on HCT 116 p53+/+
and MCF-7. The results are shown in Figures 4 and 5.
11

ACCEPTED MANUSCRIPT
Figure 4: Effect of the tested TSC (concentration 4 × IC50) on the expression of the cell cycle
and apoptosis proteins in the HCT 116 cells (A). The densitometric analyses of cyclin D1,
cdc2, p21, p53 and cytochrome c were normalized to β-actin. The results are the mean ± SD
of five independent experiments. Statistical differences were analyzed using one-way
ANOVA with a Bonferroni’s post-hoc test: *p < 0.05, **p < 0.01, ***p < 0.001 compared to
the control group (B).
12

ACCEPTED MANUSCRIPT
Figure 5: Effect of the tested TSC (concentration 4 × IC50) on the expression of cyclin D1,
cdc2, p21, p53 and cytochrome c in the MCF-7 cells (A). The densitometric analyses of these
proteins were normalized to β-actin. The results are the mean ± SD of five independent
experiments. Statistical differences were analyzed using one-way ANOVA with a
Bonferroni’s post-hoc test: *p < 0.05, **p < 0.01, ***p < 0.001 compared to the control group
(B). The effect of the TSCs on the cleaving of the PARP protein in MCF-7 cells. The
densitometric analysis of PARP was normalized to α-tubulin (C).
From the biochemical point of view, metal chelators deprive cells of intracellular iron,
and thereby, disrupt the iron metabolism, which may have an effect on multiple target
molecules that are critical in regulating the progression of the cell cycle, cyclins, cyclin-
dependent kinases (CDKs), p53 and p21CIP/WAF1 (Fig. 4, 5). The progression of the cell cycle
in the G1 phase, which is followed by a transition to the S phase, is partially controlled by the
activation of the cyclin D1/cdk4 and cyclin E/cdk2 complexes. In addition, the activity of
cyclin D1 is associated with the p21CIP/WAF1 protein, which plays a crucial role in triggering
various effects on the cell cycle regulation. A decrease in the p21CIP/WAF1 expression may lead
13

ACCEPTED MANUSCRIPT
to the arrest of the G1/S phase cell cycle, thereby affecting the stabilization of the cyclin D1
complex/cdk4. However, on the other hand, by binding iron from the intracellular space, iron
chelators may increase the expression of the p21CIP/WAF1 protein, which leads to the activation
of apoptosis signaling pathways [39]. The S phase in the cell cycle is regulated by cyclin A
and its corresponding cdk. In turn, the cdc2 protein, which is the catalytic subunit of the
heterodimer with cyclin B, is responsible for the transition to the G2/M phase of the cell
cycle. Interestingly, iron chelators cause a marked decrease in cyclin D1, D2 and D3, which
causes the expression of cyclins A and B to also be reduced, but to a lesser extent [17]. This
supports our observation that all of the tested compounds caused cell cycle arrest in the S and
G2/M phases (Fig. 2, S1, S2). The same phenomenon has also been reported elsewhere [7].
The most important regulator of the cell cycle is p53, which is involved in both the G1/S and
the G2/M checkpoints. p53 is engaged in the response to iron depletion, cellular stress, DNA-
damage or the failure of the checkpoints due to the activation of the transcription of many of
the genes that are involved in the arrest of growth, DNA repair and apoptosis when the
damage is irreversible. Several reports have indicated that DFO may up-regulate the p53
expression at the post-transcriptional stage [40]. Another quinoline-related chelator, COTI-2,
has been reported to have a submicromolar activity against cancer cells with p53 mutations
[29]. It is able to bind incorrectly folded p53 proteins, thereby promoting its repair and
reactivation [41]. Interestingly, the restoration of the tumor-suppressor protein leads to a
higher activity (lower IC50) against mutated cells than against their wild-type counterparts
[28]. On the other hand, the ability to reactivate p53 does not necessarily mean a high level of
activity against mutants with a severely damaged gene or on with deletions [42]. It is
noteworthy that HIF1-α is also believed to stabilize p53 and to lead to its up-regulation after
iron depletion [43]. However, in our current experiments, no activation of p53 was observed
in either of the cell lines that were tested (Fig. 4, 5). This, along with the results of the
Annexin V binding assay suggests a p53-independent mechanism of action (Fig. 3, S3, S4).
The results for the cells that had been incubated with the tested compounds differed
for HCT 116 p53+/+ and MCF-7. Quite a high basal level of the p21 was detected (in the
control cells) for the breast cancer and no overexpression was observed (Fig. 5). This result is
in accordance with a previously published report on the p21 level in estrogen receptor positive
breast cancers [44]. However, the 1c and 3e derivatives decreased the level of the p21 protein
to 30 and 50%, respectively (Fig. 5B). There was a significant overexpression of this protein
for all of the tested compounds except 3c for colon cancer (Fig. 4B). The highest effect was
observed for the 3g derivative, in which the p21 level was elevated 4.5-fold. Compounds 1c,
14

ACCEPTED MANUSCRIPT
2a and 2c also had a strong effect on the p21 level and increased it roughly three-fold. On the
other hand, 3h and 1e had only a limited effect. Apparently, there was no relationship
between the activity level and an increase in the p21 expression. The variation in the reactions
to TSC between the cell lines provides insight into the diversity in the regulation mechanisms
for the cell cycle. The basal level of the regulatory proteins and its connection with other
factors such as hormonal activation constitute a cell’s response to changes in the redox
equilibrium. In this regard, the absolute level of p21 does not reflect either cell proliferation or
a vulnerability to cytostatic agents. Iron chelators such as DFO and Dp44mT may increase
the level of the p21 protein via transcriptional factors as was mentioned earlier. However, the
contrary was observed in the case of MCF-7 [45]. Nevertheless, the changes in the p21
protein expression could be a sign of the downstream activation of the apoptotic pathway.
Flow cytometry experiments indicated that the population of apoptotic cells was higher
among the HCT 116 cells for all of the derivatives (Fig. 3).
Another indicator of apoptosis is the release of cytochrome c. In the experiments that
were performed, we observed a decrease in the mitochondrial cytochrome c level in both cell
lines. The exceptions were 1c, 3h, and 3c in the HCT 116 cells (Fig 4B). The highest level of
release was observed in the MCF-7 cell line for 2b (90%), 3h (70%), 2c, 3g and 2a (60%), 3e,
1e and 3c (50%) (Fig 5B). In HCT 116, this down expression was observed to a lesser extent,
but for 2c and 2a (70%), 2b (60%), 3e and 1e (50%), it was still marked. The last target
connected with apoptotic cell dead was the PARP protein. Evidence of apoptotic scenarios is
the detection of a cleaved fragment of the PARP protein with molecular mass 89 kDa. We
observed cleaved PARP only in the MCF-7 cell for 3e and 3g and a small concentration for
3h (Fig. 5C).
These results prompted us to explore two other proteins that are connected with cell
cycle regulation – cdc2 and cyclin D1. In general, we observed a decrease in the expression of
the cdc2 protein in both of the tested cell lines, except for compounds 3c, 2b, 1e and 1c on
HCT 116. For MCF-7, this downregulation was observed for all of the tested compounds,
especially for 3h, 2b, 1e, 1c and 2a. It is interesting that the strongest effect for both cell lines
was observed for 3h (70% for HCT 116 and MCF-7). A similar situation was observed for the
protein cyclin D1 where there was also a decrease for both of the investigated cell lines except
for compounds 2c, 3h and 3g on HCT 116. For MCF-7, this downregulation was particularly
strong for 3h, 1e and 2b (90%), 1c (80%) and 2a (70%). The same was observed for HCT 116
– 3c, 3e and 2b (90%), 1c (80%) and 1e (70%). The most significant difference in the
activation and deactivation of the tested proteins was observed for 3h (downregulation for
15

ACCEPTED MANUSCRIPT
MCF-7, upregulation for HCT 116) and 3c and 3e (downregulation for HCT 116,
upregulation for MCF-7). In general, these results correlate with the cell cycle analysis and
confirm the inhibitory effect of the tested compounds on the cell cycle.
2.6. Generation of ROS
TSC are known for their ability to generate ROS in a Fenton and Haber-Weiss reaction
and this is undoubtedly an important aspect of the cytotoxicity of TSC. The signaling nature
of ROS should be stressed, particularly in cancer cells that have altered redox homeostasis
[7]. An increased level of reactive oxygen species can affect many processes such as RR
inhibition, which is responsible for catalyzing the synthesis of the deoxyribonucleotides
(dNTPs) that are required for DNA synthesis. The activity of RR is determined by subunit R2,
which contains a di-nuclear iron center that is responsible for generating and maintaining a
tyrosyl radical that is required for catalysis. Its destabilization, which is induced by the
generation of ROS by chelators, causes a loss of RR activity and affects the progression of the
cell cycle and cell repair [46]. A combination of redox activity and intercalation is a
prerequisite for the accumulation of DNA damage. ROS are able to induce single lesions on
every base. Changes in the oxidative state have an impact on the progression of the cell cycle
through MnSOD and cdc2 [47]. In our study, the ROS level was quantitatively measured in
both cell lines according to the CellROX protocol (Fig. 6).
16

ACCEPTED MANUSCRIPT
Figure 6: Quantitative measurement of the intracellular ROS level in HCT 116 p53+/+ (A) and
MCF-7 (B) after a 24-h incubation with the tested TSC (1 µM), DOX (5 µM) and antioxidant
– resveratrol (100 µM). The data were normalized to the untreated cells (control). The results
are shown as the mean ± SD of three independent measurements. The data were analyzed
using one-way ANOVA with a Bonferroni’s post-hoc test: *p < 0.05, ** p < 0.01, *** p <
0.001 compared to the control group. Microscopic images of the formation of ROS in the
HCT 116 p53+/+ cells after a 24-h treatment with the tested TSC (concentration 2 × IC50) and a
15-min incubation with hydrogen peroxide (100 µM). The negative control consisted of
untreated cells. Scale bars = 20 µm (C).
Doxorubicin and resveratrol were used as the positive and negative controls, respectively. In
the colon carcinoma, the effect of TSC on the generation of ROS was much higher than in the
case of MCF-7. In general, all of the TSC that were tested had a positive response to the ROS
level. The strongest effects were observed in for compounds 1c, 2b, 2c, 3g and 3h, which
correlates with the activities that are presented in Table 1. In an analogous experiment that
was performed on MCF-7, only 3h was able to generate ROS to a significant level. A possible
explanation might be the fact that MCF-7 has an elevated basal level of ROS and a higher
amount of the main GSH antioxidant [48]. It is worth mentioning that colon and breast
17

ACCEPTED MANUSCRIPT
cancers have a different pattern of ROS, e.g. breast cancers have a higher superoxide radical
but no hydroxyl radical or hypochloric acid [49]. These differences are explained in part by
the estrogen-related redox cycle [50]. Noticeably, in our experiments, the resveratrol, which a
natural antioxidant, had slightly stronger effect on the MCF-7 than HCT 116 cells.
To confirm the importance of metal chelation in ROS production, we performed
experiments in which the metal ions Cu2+ and Fe3+ were added. As is shown in Figure S6
there was no significant difference in the activity of TSC alone or after the addition of Fe3+. In
some cases (1e and 3c) we even observed a slight decrease in the activity. Iron ions must be
reduced to Fe2+ in order to participate in the Fenton reaction [51]. However, in most cases, the
addition of Cu2+ resulted in a substantial improvement of the activity of TSC alone. This
synergistic effect is a commonly known fact [7][15] and the explanation is the participation of
Cu2+ complexes in the Haber-Weiss reaction [11], the great ability of copper to form
complexes with TSC [15] and the ionophoric action of TSC, which increases the
concentration of copper in the cells [52]. Although the concentration of both metals are tightly
regulated in the cell, there are no free copper ions. On the other hand, the storage capacity for
this metal is also lower as requirement for this element is well fulfilled by diet [53]. Finally,
an internalization of copper to lysosomes has been demonstrated by Stacy et al. as main
aspect of overcoming Pgp-mediated drug resistance [35]. For these reasons the addition of
Cu2+ may result in a stronger decrease of the viability of cells.
Additionally we conducted experiments by adding the non-specific chelator EDTA
and we did not observe any significant effect on the cellular proliferation (Fig. S6). TSC are
particularly strong chelators with high affinity to the iron. According to Bernhardt and
coworkers the pyridine based thiosemicarbazones were sufficient to abstract Fe3+ from EDTA
and DFO complexes in a biologically relevant pH [54]. In our experiment addition of EDTA
in the concentration 10 times higher than TSC, has no effect on its activity.
2.7. Intercalation
Higher activity against the p53 mutant and, in particular, the null cell lines may be
connected with intercalation or other damage in DNA [42]. TSC may also increase the small
changes in genome by forming ROS after binding to DNA [7]. To explore this possibility, a
spectroscopic analysis of a mixture of calf-thymus DNA with all of the derivatives was
performed. The results are shown in Table 2 along with doxorubicin, which is well-known
strong intercalator [55]. All of the compounds were able to reduce the absorbance of DNA as
expected. This effect was particularly strong in 2c and 3g-3i and was comparable to the
18

ACCEPTED MANUSCRIPT
doxorubicin. Compounds 3h and 3i were responsible for a strong redshift in the absorption
maxima. The other TSC expressed a moderate but still significant effect with the exception of
3c and 3e where the effect was too small for the intercalators. The same has been observed for
Dp44mT, which is in good agreement with the literature [7]. Interestingly, compound 3c,
which showed no sign of intercalation, also had no selectivity in either of the HCT 116 lines.
Hypochromism is considered to be the result of an interaction between the electronic state of
chelator and the base pairs in DNA, which is a sign of intercalation. Hyperchromism is more
often the effect of a partial uncoiling or groove binding. The absorption spectra of the tested
thiosemicarbazones and CT-DNA are presented in Fig. S7 (Supporting Information).
Table 2: Spectroscopic properties of the tested TSCs that bound to the calf-thymus DNA.
Changes in
%
hypochromism
15.2
Compound
Absorption
∆ε M−1 cm−1
Blue/red shift*
absorbance
268; 326
292
hypochromism
hypochromism
hypochromism
hypochromism
hypochromism
hypochromism
hypochromism
hypochromism
hypochromism
hypochromism
hypochromism
1215.6
1606.7
935.6
8
10
8
1c
10.3
1e
310
11.0
2b
396
24.7
955.6
4
2c
262; 300
380
3.7
1026.7
737.8
2
3c
5.8
2
3e
264; 308
310
25.9
4031.1
3155.6
6433.3
1673.3
3235.6
2
3g
3h
28.7
16
14
2
364
29.6
3i
272; 324
480
8.2
Dp44mT
Doxorubicin
34.2
10
*for the wavelengths of maximum absorption for individual and DNA-bound compounds.
3. Conclusions
Based on our results, it is evident that specific cellular regulatory mechanisms are
crucial for understanding the antiproliferative activity of TSC. For example, the alternative
regulation of the pro- and anti-apoptotic proteins may determine the intrinsic reaction of a cell
to an active dose of a chelator. On the other hand, however, the chemical structure determines
the additional features such as DNA intercalation, ROS generation or ionophoric activity.
These, in turn, may result in the selectivity profile of a drug. In this regard, more data would
be valuable in designing iron chelators for the specific biochemical microenvironment that
TSC deal with.
19

ACCEPTED MANUSCRIPT
4. Experimental session
4.1. Chemistry
4.1.1. General
Microwave reactions were carried out in a Discover® BenchMateTM (CEM)
microwave equipped with 10 mL vessels. Melting point measurements were determined in a
1
Stanford Research Systems OptiMelt (MPA 100). The H and 13C NMR spectra were
recorded on a Bruker Ascend 500 MHz spectrometer at frequencies of 500 MHz and 126
MHz and a Bruker Avance III 400 MHz FT-NMR spectrometer at frequencies of 400 MHz
and 101 MHz using DMSO-d6 as the solvent and TMS as the internal standard. The NMR
solvents were purchased from ACROS Organics. The chemical shifts (δ) are given in ppm
and the coupling constants (J) values are reported in hertz (Hz). The spin multiplicities are
described as s (singlet), d (doublet), dd (double of doublets), t (triplet), q (quartet) and m
(multiplet). The mass spectra were recorded with a Varian 500-MS IT Mass Spectrometer.
Elemental analyses (for C, H, N) were carried out on an automatic Perkin-Elmer 240
microanalyzer (Boston, MA, USA) and were within 0.4% of the calculated values. All of the
evaporations were performed on a rotary evaporator under diminished pressure at 60°C. All of
the reagents and solvents, which were purchased from ACROS Organics, Asta-Tech,
Maybridge, Santa Cruz Biotechnology and Sigma-Aldrich, were used without further
purification. The purity of the thiosemicarbazones was confirmed by HPLC and was at least
95%.
4.1.2. General procedure for the synthesis of the thiosemicarbazides
The mixtures of (1,1'-thiocarbonyl) bis-1H-imidazole (5 mmol) and suitable derivative
of piperazine or thiomorpholine (5 mmol) in methylene chloride (25 ml) were stirred for 24 h
at room temperature. The solutions were extracted three times with distilled water and the
organic phases were dried over MgSO4 and evaporated. The obtained thioketones were
refluxed for 2 hours with hydrazine hydrate (5 mmol). The final thiosemicarbazides were
crystallized from dry methanol.
Thiomorpholine-4-carbothiohydrazide - precursor of 1b and 3a
1
White crystal powder; yield 89%; mp: 160-161
̊
4.03 (m, 8H, CH2), 4.83 (s, 2H, NH2), 9.19 (s, 1H, NH). 13C-NMR (101 MHz, d6-DMSO,
ppm): δ 26.4; 50.8; 182.3.
20

ACCEPTED MANUSCRIPT
4-(3,4-dichlorophenyl)piperazine-1-carbothiohydrazide - precursor of 1c, 3c and 3i
1
White powder; yield 68%; mp: 186-187
̊
4H, CH2), 3.86 (m, 4H, CH2), 4.76 (s, 2H, NH2), 6.93 (m, 1H, Ar-H), 7.14 (s, 1H, Ar-H), 7.40
(d, 1H, J = 9.0 Hz), 9.19 (s, 1H, NH).13C-NMR (101 MHz, d6-DMSO, ppm): δ 47.1; 47.3;
115.6; 116.6; 120.1; 130.9; 132.0; 150.7; 183.0.
4-(4-nitrophenyl)piperazine-1-carbothiohydrazide - precursor of 1d, 2c, 3e and 3f
Yellow powder; yield 87%; mp: 207-208 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.57 (s,
4H, CH2), 3.91 (s, 4H, CH2), 4.97 (s, 2H, NH2), 6.97 (d, 2H, J = 9.3 Hz), 8.08 (m, 2H, Ar-H),
13
9.15 (s, 1H, NH). C-NMR (101 MHz, d6-DMSO, ppm): δ 45.7; 46.7; 112.5; 126.2; 137.3;
154.6; 182.8.
4-(4-cyanophenyl)piperazine-1-carbothiohydrazide - precursor of 1e and 3b
1
Light orange powder; yield 70%; mp: 169-170
̊
3.42 (m, 4H, CH2), 3.89 (m, 4H, CH2), 4.78 (s, 2H, NH2), 7.05 (s, 2H, Ar-H), 7.60 (m, 2H, Ar-
H), 9.15 (s, 1H, NH). 13C-NMR (101 MHz, d6-DMSO, ppm): δ 45.9; 48.9; 98.6; 114.3; 120.5;
133.8; 153.0; 182.9.
4-(4-methoxyphenyl)piperazine-1-carbothiohydrazide - precursor of 2a, 3d and 3j
1
White powder; yield 81%; mp: 194-195
̊
3H, CH3), 3.86 (m, 4H, CH2), 4.11 (m, 4H, CH2), 4.77 (s, 2H, NH2), 6.83 (d, 2H, J = 9.0 Hz),
13
6.90 (m, 2H, Ar-H), 9.17 (s, 1H, NH). C-NMR (101 MHz, d6-DMSO, ppm): δ 19.0; 49.9;
56.5; 114.8; 118.2; 145.4; 153.7; 183.1.
4-(4-fluorophenyl)piperazine-1-carbothiohydrazide - precursor of 2b
Light pink powder; yield 97%; mp: 180-181 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.09
(m, 4H, CH2), 3.87 (m, 4H, CH2), 4.77 (s, 2H, NH2), 6.97 (m, 2H, Ar-H), 7.05 (m, 2H, Ar-H),
13
9.19 (s, 1H, NH). C-NMR (101 MHz, d6-DMSO, ppm): δ 47.6; 49.1; 115.9; 117.9; 148.0;
157.8; 183.0.
4-[4-(trifluoromethyl)phenyl]piperazine-1-carbothiohydrazide precursor of 3g
White powder; yield 56%; mp: 181-182 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.18 (m,
4H, CH2), 3.90 (s, 4H, CH2), 4.77 (s, 2H, NH2), 7.07 (s, 2H, Ar-H), 7.51 (d, 2H, J = 8.4 Hz),
21

ACCEPTED MANUSCRIPT
13
9.18 (s, 1H, NH). C-NMR (101 MHz, d6-DMSO, ppm): δ 46.6; 47.1; 114.3; 118.1; 118.4;
126.7; 153.2; 183.0.
4-[5-(trifluoromethyl)pyridin-2-yl]piperazine-1-carbothiohydrazide precursor of 3h
1
White powder; yield 70%; mp: 206-207
̊
4H, CH2), 3.87 (s, 4H, CH2), 4.76 (s, 2H, NH2), 6.93 (d, 1H, J = 9.1 Hz), 7.82 (s, 1H, Ar-H),
13
8.42 (s, 1H, Ar-H), 9.14 (s, 1H, NH). C-NMR (101 MHz, d6-DMSO, ppm): δ 43.8; 46.9;
49.1; 106.7; 113.7; 135.0; 145.7; 160.3; 183.0.
4.1.3. General procedure for synthesizing the thiosemicarbazones
Two drops of glacial acetic acid as a catalyst were added to the mixtures of
thiosemicarbazides (0.5 mmol) and di(2-pyridyl) ketone, 2-pyridinecarboxaldehyde, 2-
quinolinecarboxaldehyde,
8-hydroxy-2-quinolinecarboxaldehyde
or
2-
quinoxalinecarbaldehyde (0.5 mmol) in ethanol (5 ml). The glass tubes were sealed and
placed into a microwave reactor at 83ºC for 20 minutes (the reactor power did not exceed
50W). The final products were crystallized from dry methanol.
N'-[di(pyridin-2-yl)methylidene]-4-thiomorpholine-1-carbothiohydrazide (1b)
Yellow powder; yield 86%; mp: 223-224 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 2.78 (s,
4H, CH2), 4.28 (s, 4H, CH2), 7.53 (m, 3H, Ar-H), 7.96 (m, 3H, Ar-H), 8.59 (m, 1H, Ar-H),
13
8.82 (m, 1H, Ar-H), 14.65 (s, 1H, NH). C-NMR (126 MHz, d6-DMSO, ppm): δ 51.9; 56.5;
123.6; 124.3; 126.3; 137.5; 148.6; 148.9; 156.0. HRMS (ESI): m/z calculated for C16H18N5S2:
344.1004, found: 344.1004 [M+H]+.
N'-[di(pyridin-2-yl)methylidene]-4-(3,4-dichlorophenyl)piperazine-1-carbothiohydrazide (1c)
Yellow powder; yield 48%; mp: 159-160 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.46 (s,
4H, CH2), 4.15 (s, 4H, CH2), 7.13 (d, 1H, J = 2.6 Hz), 7.43 (d, 1H, J = 9.0 Hz), 7.49 (m, 1H,
Ar-H), 7.61 (m, 2H, Ar-H), 7.98 (m, 4H, Ar-H), 8.61 (d, 1H, J = 3.9 Hz), 8.88 (d, 1H, J = 4.2
13
Hz), 14.66 (s, 1H, NH). C-NMR (126 MHz, d6-DMSO, ppm): δ 47.0; 48.9; 115.2; 116.2;
119.8; 124.2; 125.1; 127.4; 131.0; 132.1; 137.8; 149.1; 150.3; 156.0; 180.6. HRMS (ESI): m/z
calculated for C22H21Cl2N6S: 471.0925, found: 471.0934 [M+H]+.
N'-[di(pyridin-2-yl)methylidene]-4-(4-nitrophenyl)piperazine-1-carbothiohydrazide (1d)
Yellow powder; yield 44%; mp: 194-195
̊

ACCEPTED MANUSCRIPT
4H, CH2), 4.82 (m, 4H, CH2), 6.97 (s, 3H, Ar-H), 7.55 (d, 2H, J = 4.2 Hz), 8.07 (s, 3H, Ar-H),
8.52 (s, 2H, Ar-H), 8.83 (s, 1H, Ar-H), 9.15 (s, 1H, Ar-H), 14.67 (s, 1H, NH). 13C-NMR (126
MHz, d6-DMSO, ppm): δ 45.8; 48.5; 112.3; 113.6; 120.2; 124.2; 126.3; 129.8; 136.9; 148.7;
148.9; 154.6; 182.8. HRMS (ESI): m/z calculated for C22H22N7O2S: 448.1556, found:
448.1550 [M+H]+.
N'-[di(pyridin-2-yl)methylidene]-4-(4-cyanophenyl)piperazine-1-carbothiohydrazide (1e)
1
Light orange powder; yield 33%; mp: 186-187
̊
3.58 (s, 4H, CH2), 4.32 (s, 4H, CH2), 6.84 (d, 2H, J = 8.7 Hz), 7.28 (s, 1H, Ar-H), 7.41 (m,
2H, Ar-H), 7.54 (d, 3H, J = 8.7 Hz), 7.87 (m, 2H, Ar-H), 8.66 (d, 1H, J = 3.5 Hz), 8.78 (d, 1H,
J = 3.8 Hz), 14.83 (s, 1H, NH). 13C-NMR (126 MHz, d6-DMSO, ppm): δ 46.1; 48.7; 99.3;
113.6; 114.9; 120.4; 124.3; 127.3; 133.9; 137.8; 148.9; 153.3; 156.0; 180.5. HRMS (ESI): m/z
calculated for C23H22N7S: 428.1657, found: 428.1648 [M+H]+.
4-(4-methoxyphenyl)-N'-[(pyridin-2-yl)methylidene]piperazine-1-carbothiohydrazide (2a)
Brown powder; yield 36%; mp: 161-162 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 2.72 (m,
4H, CH2), 3.17 (s, 3H, CH3), 3.68 (m, 4H, CH2), 6.84 (s, 2H, Ar-H), 7.31 (m, 1H, Ar-H), 7.40
(m, 1H, Ar-H), 7.55 (m, 1H, Ar-H), 7.67 (m, 1H, Ar-H), 7.84 (m, 1H, Ar-H), 8.55 (d, 1H, J =
4.3 Hz), 10.37 (s, 1H, CH), 15.24 (s, 1H, NH). 13C-NMR (101 MHz, d6-DMSO, ppm): δ 49.7;
55.6; 114.8; 118.2; 123.0; 125.1; 137.1; 145.2; 148.2; 149.5; 153.8; 156.8; 170.8; 182.3.
HRMS (ESI): m/z calculated for C18H22N5OS: 356.1545, found: 356.1545 [M+H]+.
4-(4-fluorophenyl)-N'-[(pyridin-2-yl)methylidene]piperazine-1-carbothiohydrazide (2b)
1
Light brown powder; yield 43%; mp: 162-163
̊
3.17 (s, 4H, CH2), 4.11 (m, 4H, CH2), 7.00 (m, 2H, Ar-H), 7.08 (m, 2H, Ar-H), 7.86 (m, 2H,
Ar-H), 8.21 (s, 1H, Ar-H), 8.57 (dd, 1H, J = 11.3; 4.7 Hz), 11.48 (s, 1H, CH), 15.24 (s, 1H,
13
NH). C-NMR (101 MHz, d6-DMSO, ppm): δ 49.0; 49.4 (d J = 6.2 Hz); 50.3; 115.7; 117.9
(dd, J = 13.3, 7.6 Hz); 120.0; 124.5; 137.3; 144.6; 147.9; 149.9; 153.8; 181.2. HRMS (ESI):
m/z calculated for C17H19FN5S: 344.1345, found: 344.1341 [M+H]+.
4-(4-nitrophenyl)-N'-[(pyridin-2-yl)methylidene]piperazine-1-carbothiohydrazide (2c)
Yellow powder; yield 73%; mp: 213-214 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.69 (s,
4H, CH2), 4.13 (s, 4H, CH2), 6.97 (t, 2H, J = 8.2 Hz), 7.39 (m, 1H, Ar-H), 7.87 (m, 1H, Ar-H),
8.10 (d, 2H, J = 9.1 Hz), 8.22 (s, 1H, Ar-H), 8.59 (d, 1H, J = 4.3 Hz), 11.52 (s, 1H, CH),
23

ACCEPTED MANUSCRIPT
13
15.19 (s, 1H, NH). C-NMR (126 MHz, d6-DMSO, ppm): δ 45.8; 49.4; 112.5; 120.1; 124.6;
126.3; 137.2; 137.3; 144.6; 150.0; 153.8; 154.5; 181.1. HRMS (ESI): m/z calculated for
C17H19N6O2S: 371.1290, found: 371.1280 [M+H]+.
N'-[(quinolin-2-yl)methylidene]-4-thiomorpholine-1-carbothiohydrazide (3a)
Yellow powder; yield 34%; mp: 198-199 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 2.71 (s,
4H, CH2), 3.71 (s, 4H, CH2), 7.61 (d, 1H, J = 6.5 Hz), 7.76 (d, 1H, J = 7.1 Hz), 7.97 (dd, 3H,
13
J = 15.1; 7.9 Hz), 8.14 (s, 1H, Ar-H), 8.35 (s, 1H, CH), 12.32 (s, 1H, NH). C-NMR (126
MHz, d6-DMSO, ppm): δ 25.7; 52.1; 117.4; 127.3; 128.0; 128.4; 129.2; 130.5; 136.9; 142.4;
147.9; 154.2; 163.2. HRMS (ESI): m/z calculated for C15H17N4S2: 317.0895, found: 317.0895
[M+H]+.
4-(4-cyanophenyl)-N'-[(quinolin-2-yl)methylidene]piperazine-1-carbothiohydrazide (3b)
Yellow powder; yield 70%; mp: 202-203 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.58 (s,
4H, CH2), 4.14 (s, 4H, CH2), 7.02 (d, 3H, J = 8.5 Hz), 7.09 (d, 1H, J = 8.5 Hz), 7.64 (m, 2H,
13
Ar-H), 7.81 (m, 2H, Ar-H), 8.03 (m, 2H, Ar-H), 8.37 (s, 1H, CH), 11.67 (s, 1H, NH). C-
NMR (126 MHz, d6-DMSO, ppm): δ 45.8; 49.4; 98.5; 114.1; 114.8; 117.8; 120.4; 128.4;
129.3; 130.4; 133.8; 137.0; 144.0; 148.0; 153.0; 153.9; 173.0; 181.2. HRMS (ESI): m/z
calculated for C22H21N6S: 401.1548, found: 401.1555 [M+H]+.
4-(3,4-dichlorophenyl)-N'-[(quinolin-2-yl)methylidene]piperazine-1-carbothiohydrazide (3c)
Yellow powder; yield 77%; mp: 188-189 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.41 (s,
4H, CH2), 4.12 (s, 4H, CH2), 6.97 (d, 1H, J = 8.8 Hz), 7.18 (s, 1H, Ar-H), 7.44 (d, 1H, J = 8.9
Hz), 7.63 (t, 1H, J = 7.3 Hz), 7.79 (t, 1H, J = 7.6 Hz), 8.02 (m, 3H, Ar-H), 8.35 (s, 1H, CH),
13
8.39 (d, 1H, J = 8.6 Hz), 11.67 (s, 1H, NH). C-NMR (126 MHz, d6-DMSO, ppm): δ 47.6;
50.0; 115.5; 116.5; 117.8; 120.1; 127.7; 128.2; 128.5; 129.3; 130.6; 131.0; 132.0; 137.2;
144.4; 147.9; 150.6; 154.3; 181.2. HRMS (ESI): m/z calculated for C21H20Cl2N5S: 444.0816,
found: 444.0810 [M+H]+.
4-(4-methoxyphenyl)-N'-[(quinolin-2-yl)methylidene]piperazine-1-carbothiohydrazide (3d)
Yellow powder; yield 78%; mp: 161-162 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.17 (m,
4H, CH2), 3.70 (m, 3H, CH3), 4.12 (m, 4H, CH2), 6.86 (d, 2H, J = 9.1 Hz), 6.96 (d, 2H, J =
9.1 Hz), 7.63 (m, 1H, Ar-H), 7.79 (m, 1H, Ar-H), 7.99 (m, 1H, Ar-H), 8.03 (m, 2H, Ar-H),
8.34 (s, 1H, CH), 8.39 (d, 1H, J = 8.7 Hz), 11.65 (s, 1H, NH). 13C-NMR (126 MHz, d6-
24

ACCEPTED MANUSCRIPT
DMSO, ppm): δ 50.2; 50.5; 55.7; 114.8; 117.7; 118.2; 127.6; 128.2; 128.5; 129.3; 130.5;
137.2; 144.3; 145.4; 147.9; 153.7; 154.3; 181.2. HRMS (ESI): m/z calculated for
C22H24N5OS: 406.1702, found: 406.1695 [M+H]+.
4-(4-nitrophenyl)-N'-[(quinolin-2-yl)methylidene]piperazine-1-carbothiohydrazide (3e)
Yellow powder; yield 65%; mp: 206-207 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.72 (m,
4H, CH2), 4.17 (m, 4H, CH2), 7.01 (d, 1H, J = 8.9 Hz), 7.63 (m, 1H, Ar-H), 7.79 (m, 1H, Ar-
H), 8.01 (m, 2H, Ar-H), 8.05 (d, 2H, J = 8.7 Hz), 8.11 (d, 2H, J = 9.4 Hz), 8.35 (s, 1H, CH),
13
8.38 (d, 1H, J = 8.7 Hz), 11.70 (s, 1H, NH). C-NMR (126 MHz, d6-DMSO, ppm): δ 45.8;
49.6; 112.5; 117.9; 126.3; 127.7; 128.2; 128.5; 129.3; 130.5; 137.1; 137.2; 144.4; 147.9;
154.3; 154.5; 181.1. HRMS (ESI): m/z calculated for C21H21N6O2S: 421.1447, found:
421.1459 [M+H]+.
N'-[(8-hydroxyquinolin-2-yl)methylidene]-4-(4-nitrophenyl)piperazine-1-carbothiohydrazide
(3f)
Yellow powder; yield 78%; mp: 204-205 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.74 (m,
4H, CH2), 4.14 (m, 4H, CH2), 6.98 (t, 2H, J = 10.2 Hz), 7.12 (m, 2H, Ar-H), 7.42 (m, 2H, Ar-
H), 8.08 (m, 2H, Ar-H), 8.29 (d, 1H, J = 8.7 Hz), 8.38 (s, 1H, CH), 9.84 (s, 1H, OH), 11.74 (s,
1H, NH). 13C-NMR (126 MHz, d6-DMSO, ppm): δ 45.8; 49.6; 112.4; 112.6; 117.8; 118.6;
126.3; 128.6; 129.2; 137.0; 137.3; 138.6; 144.3; 152.2; 153.6; 154.5; 181.0. HRMS (ESI): m/z
calculated for C21H21N6O3S: 437.1396, found: 437.1387 [M+H]+.
4-[(4-trifluoromethyl)phenyl]-N'-[(quinolin-2-yl)methylidene]piperazine-1-
carbothiohydrazide (3g)
Yellow powder; yield 83%; mp: 207-208 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.52 (s,
4H, CH2), 4.15 (s, 4H, CH2), 7.09 (d, 2H, J = 8.0 Hz), 7.55 (d, 2H, J = 7.9 Hz), 7.63 (t, 1H, J
= 7.2 Hz), 7.79 (t, 1H, J = 7.1 Hz), 8.02 (m, 3H, Ar-H), 8.35 (s, 1H, CH), 8.39 (d, 1H, J = 8.6
13
Hz), 11.68 (s, 1H, NH). C-NMR (126 MHz, d6-DMSO, ppm): δ 46.8; 49.9; 114.3; 117.8;
118.4; 124.4; 126.7 (d, J = 3.7 Hz); 127.7; 128.2; 128.5; 129.3; 130.6; 137.2; 144.4; 147.9;
153.1; 154.3; 181.2. HRMS (ESI): m/z calculated for C22H21F3N5S: 444.1470, found:
444.1473 [M+H]+.
4-[5-(trifluoromethyl)pyridin-2-yl]-N'-[(quinolin-2-yl)methylidene]piperazine-1-
carbothiohydrazide (3h)
25

ACCEPTED MANUSCRIPT
1
Yellow crystal powder; yield 68%; mp: 202-203
̊
3.85 (s, 4H, CH2), 4.13 (s, 4H, CH2), 6.97 (d, 1H, J = 8.9 Hz), 7.63 (t, 1H, J = 7.2 Hz), 7.79 (t,
1H, J = 7.4 Hz), 7.86 (d, 1H, J = 8.6 Hz), 8.02 (m, 3H, Ar-H), 8.35 (s, 1H, CH), 8.38 (m, 1H,
Ar-H), 8.46 (s, 1H, Ar-H), 11.67 (s, 1H, NH). 13C-NMR (126 MHz, d6-DMSO, ppm): δ 44.1;
49.9; 106.7; 113.8; 114.0; 117.8; 127.6; 128.2; 128.5; 129.3; 130.5; 135.0 (d, J = 3.1 Hz);
137.1; 144.4; 145.7(m); 147.9; 154.3; 160.3; 181.4. HRMS (ESI): m/z calculated for
C21H20F3N6S: 445.1422, found: 445.1411 [M+H]+.
4-(3,4-dichlorophenyl)-N'-[(quinoxalin-2-yl)methylidene]piperazine-1-carbothiohydrazide
(3i)
Yellow powder; yield 96%; mp: 177-178 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.41 (m,
4H, CH2), 4.13 (m, 4H, CH2), 6.96 (m, 1H, Ar-H), 7.17 (d, 1H, J = 2.7 Hz), 7.43 (dd, 1H, J =
9.0; 2.5 Hz), 7.87 (m, 2H, Ar-H), 8.10 (m, 2H, Ar-H), 8.35 (s, 1H, CH), 9.38 (s, 1H, Ar-H),
13
11.82 (s, 1H, NH). C-NMR (126 MHz, d6-DMSO, ppm): δ 47.5; 50.1; 115.5; 116.5; 120.1;
129.5; 130.9; 131.0; 131.2; 132.0; 132.3; 141.8; 141.9; 142.1; 143.3; 149.0; 150.5; 181.1.
HRMS (ESI): m/z calculated for C20H19Cl2N6S: 445.0769, found: 445.0788 [M+H]+.
4-(4-methoxyphenyl)-N'-[(quinoxalin-2-yl)methylidene]piperazine-1-carbothiohydrazide (3j)
Yellow powder; yield 90%; mp: 162-163 ̊
C; 1H-NMR (400 MHz, d6-DMSO, ppm): δ 3.18 (m,
4H, CH2), 3.70 (s, 3H, CH3), 4.14 (m, 4H, CH2), 6.86 (m, 2H, Ar-H), 6.97 (m, 2H, Ar-H),
7.87 (m, 2H, Ar-H), 8.10 (m, 2H, Ar-H), 8.35 (s, 1H, CH), 9.37 (s, 1H, Ar-H), 11.78 (s, 1H,
13
NH). C-NMR (126 MHz, d6-DMSO, ppm): δ 50.2; 50.6; 55.7; 114.8; 118.2; 129.5; 130.9;
131.2; 136.8; 141.8; 141.9; 142.0; 143.3; 145.3; 149.1; 153.8; 181.1. HRMS (ESI): m/z
calculated for C21H23N6OS: 407.1654, found: 407.1673 [M+H]+.
4.2. Biological assays
4.2.1. Cell culture
The human colon cancer cell line HCT 116 p53+/+ and the human breast carcinoma
cell line MCF-7 were obtained from ATCC. The HCT 116 that have a p53 deletion (p53-/-)
were kindly provided by prof. M. Rusin from the Maria Sklodowska-Curie Memorial Cancer
Center and Institute of Oncology in Gliwice, Poland. The glioma cell lines U-251 and Hs 683
were kindly provided by prof. G. Kramer-Marek from The Institute of Cancer Research in
London, England. The normal human fibroblast cell lines NHDF were obtained from
PromoCell. The cells were grown as monolayer cultures in Dulbecco’s modified Eagle’s
26

ACCEPTED MANUSCRIPT
medium combined with the antibiotics penicillin/streptomycin (Gibco) (1% v/v) in 75 cm2
flasks (Nunc). DMEM for HCT 116, MCF-7, U-251 and Hs 683 were supplemented with
12% heat-inactivated fetal bovine serum (Sigma) and for NHDF with 15% non-inactivated
fetal bovine serum (Sigma). The cells were cultured under standard conditions at 37°C in a
humidified atmosphere at 5% CO2. All of the cell lines were subjected to routine mycoplasma
testing using the PCR technique with specific Mycoplasma primers in order to ensure that
there was no contamination.
4.2.2. Cytotoxicity studies
The cells were seeded in 96-well plates (Nunc) at a density of 5,000 cells/well (HCT
116, MCF-7, U-251, Hs 683) and 4,000 cells/well (NHDF) and incubated at 37°C for 24 h.
The assay was performed following a 72-h incubation with varying concentrations of the
compounds that were being tested. Then, 20 µL of the CellTiter 96®AQueous One Solution-
MTS (Promega) was added to each well (with 100 µL DMEM without phenol red) and
incubated for 1 h at 37°C. The optical densities of the samples were analyzed at 490 nm using
a multi-plate reader (Synergy 4, Bio Tek). The results are expressed as a percentage of the
control and were calculated as the inhibitory concentration (IC50) values (using GraphPad
Prism 7). The IC50 parameter was defined as the concentration of a compound that was
necessary to reduce the proliferation of cells to 50% of the untreated control. Each individual
compound was tested in triplicate in a single experiment with each experiment being repeated
three or four times.
4.2.3. Impact of copper and iron ions on cellular proliferation
The HCT 116 cells were seeded in 96-well plates (Nunc) at a density of 5,000
cells/well and incubated at 37°C for 24 h. The next day, freshly prepared solutions of the
tested thiosemicarbazone derivatives, 1c, 1e, 2a-c, 3c, 3e, 3h, at varying concentrations were
added to the plate and incubated for the next 72h. Additionally, a 20 µM solution of CuSO4 or
FeCl3, was added into wells with the tested compounds. The controls that were supplemented
with these metal solutions did not affect the cell viability (>95 %). An MTS assay was
performed as was described above. The results are expressed as a percentage of the control.
Each individual compound was tested in triplicate in a single experiment with each
experiment being repeated three times.
4.2.4. Effect of an EDTA chelator on cellular proliferation
27

ACCEPTED MANUSCRIPT
The HCT 116 cells were seeded in 96-well plates (Nunc) at a density of 5,000
cells/well and incubated at 37°C for 24 h. The next day, freshly prepared solutions of the
tested compounds, 1c, 1e, 2a-c, 3c, 3e, 3h, at IC50 concentrations were added to the plate and
incubated for the next 72h. Additionally, freshly prepared solutions of EDTA at one, five and
ten-fold IC50 concentrations that corresponded to the TSC were added into each well. The
controls that were supplemented with EDTA did not affect the cell viability (>95 %). An
MTS assay was performed as was described above. The results are expressed as a percentage
of the control (untreated cells). Each individual compound was tested in triplicate in a single
experiment with each experiment being repeated three times.
4.2.5. Reactive oxygen species formation – microscope images
The HCT 116 p53+/+ cells were seeded into eight-well chambers (Lab-Tek) at a density
of 0.5·105 cells/well and incubated at 37°C. After 24 h, freshly prepared solutions of the
tested compounds, 3c, 3d, 3e (two-fold IC50 concentration), were added. The next day, the
solutions of the tested compounds were removed and the cells were washed with Phosphate
Buffered Saline (PBS) after which 5 µM of CellROX® Green Reagent (Molecular Probes™)
was added. After 30 min of incubation at 37°C, the cells were washed with PBS and then
DMEM without phenol red was added. The observation was performed using an inverted
fluorescence microscope (IX81, Olympus) that was equipped with a CO2 incubator using a
485 nm excitation laser and a 520 nm emission filter.
4.2.6. Quantitative measurement of the level of ROS
The HCT 116 p53+/+ and MCF-7 cells were seeded onto black 96-well plates
(Corning) at a density of 9,000 cells/well and incubated at 37°C. After an overnight
incubation, the solutions of the tested compounds, 1c, 1e, 2a-c, 3c, 3e, 3g-h (1 µM) and DOX
(5 µM), were added and incubated for 24 h. In addition, solutions of 100 µM H2O2 (positive
control) for 20 min and 100 µM resveratrol (negative control) for 24 h were added to the cells.
The generation of ROS was measured using a CellROX® Green Reagent (Molecular
Probes™). Additionally, the quantity of cells in each well was determined using Hoechst
33342 (Molecular Probes™). The solutions of the tested compounds were removed and 100
µL of CellROX Green Reagent and Hoechst 33342 at a final concentration of 5 µM were
added to each well. Then, the cells were incubated for 30 min at 37°C. The fluorescence was
measured using a multi-plate reader (Synergy 4, Bio Tek) at a 485 nm excitation and a 520
nm emission for the CellROX Green Reagent and a 345 nm excitation laser and a 485 nm
28