Identification of differential anti-neoplastic activity of copper bis(thiosemicarbazones) that is mediated by intracellular reactive oxygen species generation and lysosomal membrane permeabilization

Journal of Inorganic Biochemistry 152 (2015) 20 –37

Contents lists available at ScienceDirect

Journal of Inorganic Biochemistry

journal homepage: w ww.elsevier.com/locate/jinorgbio

Identiﬁcation of differential anti-neoplastic activity of copper

bis(thiosemicarbazones) that is mediated by intracellular reactive

oxygen species generation and lysosomal membrane permeabilization

⁎

Christian Stefani, Zaynab Al-Eisawi, Patric J. Jansson, Danuta S. Kalinowski , Des R. Richardson

Molecular Pharmacology and Pathology Program, Department of Pathology and Bosch Institute, The University of Sydney, Sydney, NSW 2006, Australia

a r t i c l e i n f o

a b s t r a c t

Article history:

Bis(thiosemicarbazones) and their copper (Cu) complexes possess unique anti-neoplastic properties. However,

their mechanism of action remains unclear. We examined the structure–activity relationships of twelve

bis(thiosemicarbazones) to elucidate factors regarding their anti-cancer efﬁcacy. Importantly, the alkyl substitu-

tions at the diimine position of the ligand backbone resulted in two distinct groups, namely, unsubstituted/

monosubstituted and disubstituted bis(thiosemicarbazones). This alkyl substitution pattern governed their:

(1) Cu^II/Iredox potentials; (2) ability to induce cellular ⁶⁴Cu release; (3) lipophilicity; and (4) anti-proliferative

activity. The potent anti-cancer Cu complex of the unsubstituted bis(thiosemicarbazone) analog, glyoxal bis(4-

methyl-3-thiosemicarbazone) (GTSM), generated intracellular reactive oxygen species (ROS), which was

attenuated by Cu sequestration by a non-toxic Cu chelator, tetrathiomolybdate, and the anti-oxidant, N-acetyl-

L-cysteine. Fluorescence microscopy suggested that the anti-cancer activity of Cu(GTSM) was due, in part, to

lysosomal membrane permeabilization (LMP). For the ﬁrst time, this investigation highlights the role of ROS

and LMP in the anti-cancer activity of bis(thiosemicarbazones).

Received 29 June 2015

Accepted 5 August 2015

Available online 17 August 2015

Keywords:

Copper

Bis(thiosemicarbazone)

Reactive oxygen species

Lysosome

Lysosomal membrane permeabilization

1. Introduction

Importantly, excess Cu is a potent oxidant and results in the generation

of cytotoxic ROS in cells [15]. Considering the key role of Cu in both

Copper (Cu) is an essential trace element that plays a key role in the

biochemistry of all living organisms [1]. Its unique electronic structure,

existing in both an oxidized (Cu^II) and reduced state (Cu^I), allows it to

serve as a co-factor for enzymes that are fundamental for cellular

growth and development [1]. Neoplastic cells have a higher require-

ment for Cu, as it plays an important role in promoting physiological

and malignant angiogenesis [2]. Thus, Cu is essential in the de novo for-

mation of blood vessels that enable tumor growth, invasion and metas-

tasis [2]. Further underscoring the importance of Cu in tumor growth is

the presence of elevated Cu levels in the serum and tumors of rats and

humans [3–5]. Elevated Cu levels in cancer patients have been observed

in a wide spectrum of tumors, including: breast [3], cervical [6], ovarian

[6], lung [7], prostate [8], colorectal cancer [9] and leukemia [10]. Most

strikingly, Cu levels were found to correlate with cancer stage and/or

progression [3,9].

promoting angiogenesis and the generation of ROS, the development

of Cu targeting agents has become a promising anti-cancer strategy [15].

Thiosemicarbazones are one such class of promising anti-cancer

agents that have attracted extensive interest [16–21]. In particular, the

di-2-pyridylketone thiosemicarbazone (DpT) series (e.g., di-2-

pyridylketone 4,4-dimethyl-3-thiosemicarbazone; Dp44mT; Fig. 1A)

has demonstrated marked and selective anti-tumor activity in vitro

and in vivo against a variety of human tumor xenografts in mice [19,

22]. The mechanism of action of Dp44mT is partly dependent on the for-

mation of a redox active iron complex that generates cytotoxic ROS,

leading to damage of essential biomolecules [22–24]. More recently,

the Cu(Dp44mT) complex was identiﬁed to exhibit superior intracellu-

lar oxidative properties and anti-cancer efﬁcacy relative to both the li-

gand alone and its Fe^IIIcomplex [25]. Additionally, the Cu(Dp44mT)

complex was shown to target lysosomal integrity, leading to lysosomal

membrane permeabilization (LMP), the redistribution of the lysosomal

protease, cathepsin D, to the cytosol and ultimately results in apoptosis

[26]. Tumor cell invasion and metastasis involves changes in lysosomal

trafﬁcking and increased expression of cathepsins, and it has been

suggested that this may sensitize tumor cells to LMP and lysosomal-

targeting anti-cancer agents [27,28].

Neoplastic cells also differ from their normal counterparts in terms

of their redox metabolism [11]. The environment of the tumor is often

characterized by increased metabolic activity [12], hypoxia [13] and en-

hanced intracellular reactive oxygen species (ROS) generation [14].

⁎

Corresponding authors.

Bis(thiosemicarbazones) are another family of ligands that have also

attracted considerable interest due to their broad pharmacological

E-mail addresses: danutak@med.usyd.edu.au (D.S. Kalinowski),

d.richardson@med.usyd.edu.au (D.R. Richardson).

http://dx.doi.org/10.1016/j.jinorgbio.2015.08.010

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

21

Fig. 1. (A) Line drawings of the chemical structures of: di-2-pyridylketone 4,4-dimethyl-3-thiosemicarbazone (Dp44mT), desferrioxamine (DFO), neocuproine (Neo), Triapine (3-AP), and

tetrathiomolybdate (TM). (B, C) Line drawings of the chemical structures of members of the bis(thiosemicarbazone) series of ligands and their copper complexes. Throughout the article

the unsubstituted/monosubstituted ligands/complexes are indicated in red, while the disubstituted and disubstituted cyclic ligands are denoted in green. (For interpretation of the refer-

ences to color in this ﬁgure legend, the reader is referred to the web version of this article.)

efﬁcacy [29–32]. Their anti-tumor activity was ﬁrst demonstrated in

1958, after oral administration of glyoxal bis(thiosemicarbazone)

(GTS; Fig. 1B) was found to signiﬁcantly reduce Sarcoma 180 tumor

burden in Swiss brown mice [30]. Following the synthesis and evalua-

tion of numerous analogs, kethoxal bis(thiosemicarbazone) emerged

as a promising candidate that consistently increased the lifespan of

Swiss brown mice with L1210 leukemia [29]. The anti-cancer activity

of these ligands was later conﬁrmed in numerous in vitro and in vivo

studies [31–33].

with Cu has been shown to enhance their ability to inhibit DNA

synthesis in sarcoma 180 ascites cells [32], as well as their anti-tumor

activity against Walker 256 carcinoma in rats [35]. Coordination of Cu

by bis(thiosemicarbazones) leads to the formation of a neutral Cu-

^II[bis(thiosemicarbazone)] complex, which is capable of rapidly enter-

ing cells [36,37]. Once inside the cell, it is proposed that intracellular

reduction of the neutral Cu^II[bis(thiosemicarbazone)] complex to a

charged Cu^Icomplex occurs, resulting in intracellular trapping [20].

The reduced Cu^Icomplex is then believed to either dissociate, or be

re-oxidized via a redox-dependent process to the neutral Cu^IIcomplex,

which can be released from the cell [38]. This model is supported by

density functional theory [39] and X-ray crystallographic [40] studies

Although not fully elucidated, the mechanism of action of these

bis(thiosemicarbazones) is believed to be dependent on Cu coordina-

tion, particularly Cu^II[32,34]. Co-treatment of bis(thiosemicarbazones)

22

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

of the ligands and their Cu^IIcomplexes in cell-free investigations. How-

ever, to our knowledge, this reduction and dissociation, leading to intra-

cellular trapping, has not been demonstrated in cells.

concentration in DMSO. The supporting electrolyte was Et₄NClO₄

(0.1 M) and the solutions were purged with nitrogen prior to measure-

ment. Partition coefﬁcients of the free ligands were determined by

ChemBioDraw v.11.0.1. (PerkinElmer, Waltham, MA, USA) using

Crippen's fragmentation procedure [51].

Recent interest in bis(thiosemicarbazones) has centered around the

hypoxia selectivity of Cu^IIdiacetylbis(thiosemicarbazone) (Cu[ATSM];

Fig. 1B) and its potential as a radiopharmaceutical for imaging hypoxic

tissues [37,41]. In fact, Cu[ATSM] has demonstrated potential in

assessing tumor hypoxia, and thus, prognosis in cervical cancer [42]

and rectal carcinoma [43] by positron emission tomography. In fact,

the low oxygen tension of hypoxic cells is proposed to stabilize the

charged Cu^Icomplex, resulting in intracellular accumulation and hypox-

ia selectivity [20]. However, despite this new medical application, their

precise intracellular mechanism of action remains elusive.

Imaging of ﬂuorescent analogs of bis(thiosemicarbazones) has

proved an attractive strategy for the study of their in vitro subcellular lo-

calization [44–47]. A ﬂuorescent pyrene conjugated derivative of

Cu(ATSM) has revealed localization into distinct punctuate structures

that partially co-localized with lysosome/autophagic structures in

HeLa and M17 neuroblastoma cells [48]. In contrast, Cu(ATSM) analogs

with a ﬂuorescent napthenequinone backbone were dispersed evenly in

the cytoplasm of HeLa cells [46]. However, the signiﬁcance of these lo-

calization studies on the mechanism of action of the unconjugated com-

plexes is unclear and requires further investigation.

2.2. General synthesis of ligands

The ligands were synthesized by the following common procedure, ex-

empliﬁed by the synthetic route used for glyoxalbis(thiosemicarbazone)

(GTS). Thiosemicarbazide (10 mmol) was dissolved in ethanol

(10 mL) and the appropriate diketone (5 mmol) was dissolved in etha-

nol (5 mL) and the two solutions then mixed. Glacial acetic acid (5–6

drops) was added and the mixture gently reﬂuxed for 2 to 5 h. The mix-

ture was cooled to room temperature and allowed to stand at 4 °C over-

night to ensure complete precipitation. The product was ﬁltered off and

washed with distilled water (2 × 10 mL) and ethanol (10 mL) and dried

in vacuo.

2.2.1. Glyoxalbis(thiosemicarbazone) (GTS)

Pale yellow powder (yield: 82.1%). Anal. Calc. for C₄H₈N₆S₂: C, 23.5;

H, 4.0; N, 41.1; S, 31.4%. Found: C, 23.6; H, 4.1; N, 41.2; S, 31.2%. ¹H NMR

(DMSO-d₆): 11.68 (s, 2H), 8.30 (s, 2H), 7.88 (s, 2H), 7.70 (s, 2H). MS

(ESI⁺) m/z 205.3 [M + H]⁺, 227.3 [M + Na]⁺, 243.4 [M + K]⁺.

In the present study, we examined a series of bis(thiosemicarbazones)

and their Cu^IIcomplexes (Fig. 1B, C) for their anti-proliferative activity

in SK-N-MC neuroepithelioma and mortal MRC-5 ﬁbroblast cells. This

series of bis(thiosemicarbazone) ligands and complexes were synthe-

sized to investigate the effect of different structural features on electro-

chemical and intracellular behavior, including their ability to affect the

cellular retention of ⁶⁴Cu. The bis(thiosemicarbazone) ligands vary in

their alkyl substitution pattern at their diimine backbone (R₁and R₂)

and their terminal amines (R₃; Fig. 1B, C). This substitution pattern at

R₁and R₂, resulted in two major groups of ligands, namely: the

unsubstituted/monosubstituted group and the disubstituted group (de-

noted in red and green, respectively, throughout the study; Fig. 1B, C).

These two groups of ligands had distinct chemical and biological

activity that was linked to their Cu^II/Iredox potentials, Cu mobilization

activity and lipophilicity. The unsubstituted/monosubstituted

bis(thiosemicarbazones) that were less lipophilic and had less negative

Cu^II/Iredox potentials resulted in cellular ⁶⁴Cu accumulation and greater

anti-proliferative efﬁcacy relative to the disubstituted group, that were

more lipophilic and had more negative redox potentials. Furthermore,

the Cu complex of the unsubstituted bis(thiosemicarbazone) analog,

glyoxal bis(4-methyl-3-thiosemicarbazone) (GTSM), that exhibited

potent anti-cancer activity, demonstrated the ability to mediate intracel-

lular ROS generation and LMP. For the ﬁrst time, we demonstrate the

anti-proliferative activity of the unsubstituted bis(thiosemicarbazone)

ligand, GTSM, is linked with the ability of the resultant Cu complex to

redox cycle and mediate LMP.

2.2.2. Glyoxalbis(4-methyl-3-thiosemicarbazone) (GTSM)

Yellow powder (yield: 85%). Anal. Calc. for C₆H₁₂N₆S₂: C, 31.0; H, 5.2;

N, 36.2; S, 27.6%. Found: C, 31.0; H, 5.4; N, 36.4; S, 27.9%. ¹H NMR

(DMSO-d₆): 11.74 (s, 2H), 8.49 (q, 2H), 7.71 (s, 2H), 2.95 (d, 6H). MS

(ESI⁺) m/z 233.4 [M + H]⁺, 255.4 [M + Na]⁺, 271.4 [M + K]⁺.

2.2.3. Pyruvaldehydebis(thiosemicarbazone) (PTS)

Pale yellow powder (yield: 74.3%). Anal. Calc. for C₅H₁₀N₆S₂: C, 27.5;

H, 4.6; N, 38.5; S, 29.4%. Found: C, 27.2; H, 4.8; N, 38.2; S, 29.6%. ¹H NMR

(DMSO-d₆): 11.66 (s, 1H), 10.38 (s, 1H), 8.35 (d, 2H), 7.91 (s, 2H), 7.65

(s, 1H), 2.13 (s, 3H). MS (ESI⁺) m/z 241.3 [M + Na]⁺, 257.4 [M + K]⁺.

2.2.4. Pyruvaldehydebis(4-methyl-3-thiosemicarbazone) (PTSM)

Pale yellow powder (yield: 78.5%). Anal. Calc. for C₇H₁₄N₆S₂: C, 34.1;

H, 5.7; N, 34.1; S, 26.0%. Found: C, 34.0; H, 5.9; N, 34.0; S, 26.3%. ¹H NMR

(DMSO-d₆): 11.72 (s, 1H), 10.36 (s, 1H), 8.48 (q, 2H), 7.65 (s, 1H), 2.99

(d, 6H), 2.15 (s, 3H). MS (ESI⁺) m/z 247.4 [M + H]⁺, 269.4 [M + Na]⁺,

285.4 [M + K]⁺.

2.2.5. Diacetylbis(thiosemicarbazone) (ATS)

Pale yellow powder (yield: 80.5%). Anal. Calc. for C₆H₁₂N₆S₂: C, 31.0;

H, 5.2; N, 36.2; S, 27.6%. Found: C, 31.0; H, 5.4; N, 36.2; S, 27.4%. ¹H NMR

(DMSO-d₆): 10.21 (s, 2H), 8.41 (s, 2H), 7.85 (s, 2H), 2.16 (s, 6H). MS

(ESI⁺) m/z 233.4 [M + H]⁺, 255.4 [M + Na]⁺, 271.4 [M + K]⁺.

2. Materials and methods

2.2.6. Diacetylbis(4-methyl-3-thiosemicarbazone) (ATSM)

Pale yellow powder (yield: 82.2%). Anal. Calc. for C₈H₁₆N₆S₂: C, 36.9;

H, 6.2; N, 32.3; S, 24.6%. Found: C, 36.7; H, 6.3; N, 32.1; S, 24.5%. ¹H NMR

(DMSO-d₆): 10.22 (s, 2H), 8.37 (q, 2H), 3.01 (d, 6H), 2.20 (s, 6H). MS

(ESI⁺) m/z 283.4 [M + Na]⁺, 299.4 [M + K]⁺.

All reagents were obtained commercially and used without

further puriﬁcation. The chelators, Dp44mT and 3-aminopyridine-2-

carboxaldehyde thiosemicarbazone (3-AP; Fig. 1A), were prepared

and characterized according to previously described methods [23,49,

50]. All synthesized compounds were ≥95% purity.

2.2.7. 2,3-Pentanedionebis(thiosemicarbazone) (CTS)

Pale yellow powder (yield: 82.8%). Anal. Calc. for C₇H₁₄N₆S₂: C, 34.1;

H, 5.7; N, 34.1; S, 26.0%. Found: C, 34.3; H, 5.9; N, 34.0; S, 25.9%. ¹H NMR

(DMSO-d₆): 10.35 (s, 1H), 10.22 (s, 1H), 8.42 (s, 2H), 7.80 (d, 2H), 2.84

(q, 2H), 2.14 (s, 3H), 0.89 (t, 3H). MS (ESI⁺) m/z 269.4 [M + Na]⁺, 285.4

[M + K]⁺.

2.1. Physical methods

¹H NMR (400 MHz) spectra were acquired using a Bruker Advance

400 NMR spectrometer with DMSO-d₆as the solvent and internal refer-

ence (Me₂SO: ¹H NMR δ 2.50 ppm and ¹³C NMR δ 39.5 ppm vs. TMS).

Cyclic voltammetry was performed using a BAS100B/W potentiostat. A

glassy carbon working electrode, an aqueous Ag/AgCl reference and Pt

wire auxiliary electrode were used. All complexes were at ca. 2 mM

2.2.8. 2,3-Pentanedionebis(4-methyl-3-thiosemicarbazone) (CTSM)

Yellow powder (yield: 91%). Anal. Calculated for C₉H₁₈N₆S₂: C, 39.4;

H, 6.6; N, 30.6; S, 23.4%. Found: C, 39.1; H, 6.9; N, 30.7; S, 23.2%. (DMSO-

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

23

d₆): 10.35 (s, 1H), 10.22 (s, 1H), 8.33 (q, 2H), 3.02 (d, 6H), 2.91 (q, 2H),

2.3.6. Cu(ATSM)

2.19 (s, 3H), 0.89 (t, 3H).MS (ESI⁺) m/z 297.4 [M + Na]⁺, 313.4

Yield: 67%. Anal. Calc. for CuC₈H₁₄N₆S₂·0.25H₂O: C, 29.5%; H, 4.5%; N,

[M + K]⁺.

25.8%; Found: C, 29.9%; H, 4.3%; N, 25.6%. MS (ESI⁺) m/z 322 [M + H]⁺.

2.3.7. Cu(CTS)

2.2.9. 3,4-Hexanedionebis(thiosemicarbazone) (DTS)

Yield: 70%. Anal. Calc. for CuC₇H₁₃N₆S₂Cl·0.25MeCN: C, 25.4%; H,

Pale yellow powder (yield: 84.3%). Anal. Calc. for C₈H₁₆N₆S₂: C, 36.9;

H, 6.2; N, 32.3; S, 24.6%. Found: C, 37.1; H, 6.5; N, 32.0; S, 24.4%. ¹H NMR

(DMSO-d₆): 10.37 (s, 2H), 8.42 (s, 2H), 7.75 (s, 2H), 2.82 (q, 4H), 0.87

(t, 6H). MS (ESI⁺) m/z 261.4 [M + H]⁺, 283.4 [M + Na]⁺, 299.4

[M + K]⁺.

3.9%; N, 24.7%; Found: C, 25.5%; H, 3.6%; N, 24.8%. MS (EI) m/z 308 [M]⁺.

2.3.8. Cu(CTSM)

Yield: 78%. Anal Calc for CuC₉H₁₇N₆S₂Cl·0.75MeCN: C, 31.3%; H,

4.8%; N, 23.5%; Found: C, 31.1%; H, 4.6%; N, 23.6%. MS (ESI⁺) m/z 336

[M + H]⁺.

2.2.10. 3,4-Hexanedionebis(4-methyl-3-thiosemicarbazone) (DTSM)

Yellow powder (yield: 91.6%). Anal. Calc. for C₁₀H₂₀N₆S₂: C, 41.6; H,

7.0; N, 29.1; S, 22.2%. Found: C, 41.4; H, 7.2; N, 29.0; S, 22.4%. ¹H NMR

(DMSO-d₆): 10.35 (s, 2H), 8.29 (q, 2H), 3.02 (d, 6H), 2.89 (q, 4H), 0.88

(t, 6H). MS (ESI⁺) m/z 289.5 [M + H]⁺, 311.4 [M + Na]⁺, 327.4

[M + K]⁺.

2.3.9. Cu(DTS)

Yield: 64%. Anal Calc for CuC₈H₁₄N₆S₂Cl·0.25MeCN: C, 27.7%; H,

4.3%; N, 23.8%; Found: C, 27.4%; H, 4.0%; N, 23.4%. MS (ESI⁺) m/z 322

[M + H]⁺.

2.3.10. Cu(DTSM)

2.2.11. 1,2-Cyclohexanedionebis(thiosemicarbazone) (CyTS)

Dark yellow powder (yield: 80.5%). Anal. Calc. for C₈H₁₄N₆S₂: C, 37.2;

H, 5.5; N, 32.5; S, 24.8%. Found: C, 37.0; H, 5.7; N, 32.3 S, 25.0%. ¹H NMR

(DMSO-d₆): 10.60 (s, 1H), 9.99 (s, 1H), 8.39 (d, 2H), 8.01 (s, 2H), 2.52

(td, 4H), 1.66 (t, 4H). MS (ESI⁺) m/z 259.4 [M + H]⁺, 281.4

[M + Na]⁺, 297.4 [M + K]⁺.

Yield: 59%. Anal Calc for CuC₁₀H₁₈N₆S₂: C, 34.3%; H, 5.2%; N, 24.0%;

Found: C, 34.4%; H, 5.1%; N, 23.9%. MS (ESI⁺) m/z 350 [M + H]⁺.

2.3.11. Cu(CyTS)

Yield: 73%. Anal Calc for CuC₈H₁₂N₆S₂Cl·0.66MeCN: C, 29.2%; H,

3.9%; N, 24.3%; Found: C, 29.1%; H, 3.9%; N, 24.4%. MS (ESI⁺) m/z 320

[M + H]⁺.

2.2.12.

1,2-Cyclohexanedionebis(4-methyl-3-thiosemicarbazone)

(CyTSM)

2.3.12. Cu(CyTSM)

Dark yellow/orange powder (yield: 87.2%). Anal. Calc. for

C₁₀H₁₈N₆S₂: C, 41.9; H, 6.3; N, 29.3; S, 22.4%. Found: C, 41.9; H, 6.4; N,

29.2; S, 22.1%. ¹H NMR (DMSO-d₆): 10.57 (s, 1H), 10.01 (s, 1H), 8.48

(q, 6H), 2.95 (d, 2H), 2.54 (td, 4H), 1.65 (t, 4H). MS (ESI⁺) m/z 287.4

[M + H]⁺, 309.4 [M + Na]⁺, 325.3 [M + K]⁺.

Yield: 79%. Anal Calc for CuC₁₀H₁₆N₆S₂·0.25EtOH: C, 35.1%; H, 4.9%;

N, 23.4%; Found: C, 34.7%; H, 4.6%; N, 23.5%. MS (ESI⁺) m/z 348

[M + H]⁺.

2.4. Biological studies

2.3. General synthesis of [Cu^II(L)] complexes

2.4.1. Cell culture

Chelators were dissolved in DMSO as 10 mM stock solutions and di-

luted in medium containing 10% fetal calf serum (Commonwealth

Serum Laboratories, Melbourne, Australia) so that the ﬁnal

[DMSO] b 0.5% (v/v). At this ﬁnal concentration, DMSO had no effect

on proliferation, as shown previously [52]. Human SK-N-MC

neuroepithelioma cells and mortal human MRC5 ﬁbroblasts

(American Type Culture Collection, Manassas, VA) were grown by stan-

dard procedures [22,52] at 37 °C in a humidiﬁed atmosphere of 5% CO₂/

95% air in an incubator (Forma Scientiﬁc, Marietta, OH).

The Cu complexes were prepared by the following general method

[14,20,36], exempliﬁed by the preparation of Cu(GTS): GTS (1 mmol)

was dissolved in EtOH (10 mL). Copper chloride (1 mmol), dissolved

in EtOH (5 mL), was added and the reaction mixture gently reﬂuxed

for 3 h. The mixture was cooled to room temperature and left to stir

overnight. The dark red/brown powder was collected by ﬁltration,

washed with EtOH (2 × 10 mL) and diethyl ether (10 mL) and dried

in vacuo. The product was then recrystallized from EtOH/H₂O or AcCN/

H₂O.

2.4.2. Effect of the chelators on cellular proliferation

2.3.1. Cu(GTS)

The effect of the chelators and complexes on cellular proliferation

was determined by the [1-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tet-

razolium] (MTT) assay using standard techniques [52,53]. The SK-N-MC

neuroepithelioma and MRC-5 ﬁbroblast cell lines were seeded in 96-

well microtiter plates at 1.5 × 10⁴and 1.0 × 10⁴cells/well, respectively,

in medium containing chelators or complexes at a range of concentra-

tions (0.0015–25 μM). The cells were incubated at 37 °C in a humidiﬁed

atmosphere containing 5% CO₂and 95% air for 72 h. After this incuba-

tion, 10 μL of MTT (5 mg/mL) was added to each well and further incu-

bated for 2 h/37 °C. After solubilization of the cells with 100 μL of 10%

SDS-50% isobutanol in 10 mM HCl, the plates were read at 570 nm

using a scanning multi-well spectrophotometer. The inhibitory concen-

tration (IC₅₀) was deﬁned as the chelator concentration necessary to re-

duce the absorbance to 50% of the untreated control. Using this method,

absorbance was shown to be directly proportional to cell counts, as

shown previously [52]. To examine the role of Cu chelation and redox

stress on anti-proliferative activity, SK-N-MC cells were incubated for

72 h/37 °C with chelators or complexes at a range of concentrations

(0–6.25 μM) in the presence of tetrathiomolybdate (TM; 5 μM), N-

acetyl-^L-cysteine (NAC; 5 mM) or buthionine sulfoximine (BSO;

Yield: 72%. Anal. Calc. for CuC₄H₇N₆S₂Cl: C, 15.9%; H, 2.2%; N, 27.8%;

Found: C, 16.3%; H, 2.0%; N, 27.5%. MS (EI) m/z 266 [M]⁺.

2.3.2. Cu(GTSM)

Yield: 67%. Anal. Calc. for CuC₆H₁₀N₆S₂·0.25H₂O: C, 24.2%; H, 3.6%; N,

28.2%; Found: C, 24.6%; H, 3.5%; N, 28.1%. MS (ESI⁺) m/z 295 [M + H]⁺.

2.3.3. Cu(PTS)

Yield: 63%. Anal. Calc. for CuC₅H₈N₆S₂·0.25CH₃CN·H₂O: C, 21.4%; H,

3.5%; N, 28.4%; Found: C, 21.6%; H, 2.9%; N, 28.6%. MS (EI) m/z 280 [M]⁺.

2.3.4. Cu(PTSM)

Yield: 71%. Anal Calc for CuC₇H₁₂N₆S₂·0.25H₂O: C, 26.9%; H, 4.0%; N,

26.9%; Found: C, 27.3%; H, 3.9%; N, 26.7%. MS (ESI⁺) m/z 309 [M + H]⁺,

330 [M + Na]⁺.

2.3.5. Cu(ATS)

Yield: 59%. Anal Calc for CuC₆H₁₀N₆S₂·0.5H₂O: C, 23.8%; H, 3.7%; N,

27.7%; Found: C, 24.2%; H, 3.3%; N, 27.4%. MS (ESI⁻) m/z 294 [M]⁻.

24

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

100 μM) [54,55]. Cells were incubated with control medium, NAC, or TM

for 10 min/37 °C prior to incubation with chelators or their respective

copper complexes, while BSO was incubated for 16 h/37 °C prior to

the addition of ligands or their copper complexes.

to examine co-localization with lysosomes as previously reported

[59–61]. Cells were plated on cover slips (1 × 10⁵cells/mL) and allowed

to grow overnight. To examine the role of redox stress in LMP, cells were

pre-treated with TM (5 μM), NAC (5 mM) and/or BSO (100 μM) [54,55,

62]. Cells were pre-incubated with NAC or TM for 10 min/37 °C, or with

BSO for 16 h/37 °C, prior to chelator incubation. Following washing, the

cells were pre-incubated with CuCl₂(5 μM) for 10 min/37 °C followed

by incubation with GTSM (2 μM) and incubated for 1 h/37 °C followed

by paraformaldehyde ﬁxation (4%/15 min) and digitonin perme-

abilization (100 μM/10 min). After blocking with 5% bovine serum albu-

min, immunocytochemistry was performed with anti-cathepsin D

(Abcam, USA) and anti-Lamp2 (Abcam, USA) antibodies. The cells

were examined with a Zeiss Axio Observer Z1 ﬂuorescence microscope

(Zeiss, Germany) equipped with FITC, Texas Red and DAPI ﬁlters. Im-

ages were captured with an AxioCam camera and AxioVision Rel. 4.7

Software (Zeiss, Germany). Quantiﬁcation of ﬂuorescence was per-

formed with the image processing and analysis software, ImageJ v1.48.

2.4.3. Bis(thiosemicarbazone)-mediated ⁶⁴Cu efﬂux from cells prelabeled

with ⁶⁴Cu

The ability of chelators to mobilize ⁶⁴Cu (ANSTO, Sydney, Australia)

from prelabeled cells were performed by standard methods [26,56].

Brieﬂy, cells were prelabeled for 1 h/37 °C with ⁶⁴Cu (10 μCi/mL;

⁶⁴CuCl₂), washed 4 times on ice and reincubated with medium (control)

or medium containing chelators (25 μM) for 1 h/37 °C. Radioactivity was

measured in the cells and supernatant using a γ-scintillation counter

(Wallac Wizard 3; Perkin Elmer).

2.4.4. Release of ⁶⁴Cu from cells prelabeled with the

⁶⁴Cu[bis(thiosemicarbazone)] complexes

Complexes were prepared by adding equimolar equivalents of ⁶⁴Cu

(10 μCi/mL; ⁶⁴CuCl₂; ANSTO) and bis(thiosemicarbazone) chelator.

SK-N-MC cells were incubated with the complexes (25 μM) for

1 h/37 °C, washed 4 times on ice, reincubated for 1 h/37 °C in control

media, and the percentage of ⁶⁴Cu remaining cell associated assessed

[26]. Radioactivity was measured in the cells and supernatant using a

γ-scintillation counter (Wallac Wizard 3; PerkinElmer).

2.5. Statistical analysis

Experimental data were compared using Student's t-test. Results

were expressed as mean SD (number of experiments) and considered

to be statistically signiﬁcant when p b 0.05.

3. Results and discussion

2.4.5. Intracellular ROS measurements

Intracellular ROS generation was measured using 2′,7′-

dichlorodihydroﬂuorescein diacetate (H₂DCF-DA) [22,57]. H₂DCF-DA

is hydrolyzed by intracellular esterases to the membrane impermeable

analog, 2′,7′-dichlorodihydroﬂuorescein (H₂DCF), which leads to its ac-

cumulation within the cytosol. Cellular oxidants localized to the cytosol

oxidize non-ﬂuorescent H₂DCF to the ﬂuorescent product,

dichloroﬂuorescein (DCF) [22,57]. SK-N-MC cells were incubated with

30 nM H₂DCF-DA for 30 min/37 °C and then washed twice with ice-

cold PBS. The cells were then treated with either the positive control,

H₂O₂(50 μM) for 30 min/37 °C, GTSM (2 or 25 μM), or Cu(GTSM)

(5 μM) for 1 h/37 °C. To examine the effect of Cu and Cu chelation, a

10 min/37 °C pre-incubation of cells with CuCl₂(5 μM) or TM (5 μM)

was used, respectively, prior to further chelator or complex incubation.

Cells were collected for ﬂow cytometric assessment and intracellular

ROS was detected as an increase in green cytosolic DCF ﬂuorescence

with a FACS Canto ﬂow cytometer (Becton Dickinson, Lincoln Park, NJ,

USA). In these studies, 10,000 events were acquired for every sample.

Data analysis was performed using FlowJo software v7.5.5 (Tree Star

Inc., Ashland, OR).

3.1. Preparation and characterization

All ligands were prepared by high yielding Schiff base condensation

reactions following

a previous synthetic procedure for related

bis(thiosemicarbazones) [20]. These compounds were synthesized by

reﬂuxing the appropriate thiosemicarbazide with the 1,2-diketone in a

2:1 molar ratio in an acidic EtOH solution. The resulting

bis(thiosemicarbazones) were found to be only sparingly soluble in

water. However, they exhibited greater solubility in polar aprotic sol-

vents such as DMF, MeCN and DMSO. The ¹H NMR, mass spectrometry

and combustion analyses were consistent with their proposed struc-

tures (Fig. 1B). The Cu^IIcomplexes were prepared by reﬂuxing CuCl₂

with the appropriate ligand in a 1:1 molar ratio in ethanol [20]. The

complexes showed greater water solubility than the ligand, and were

very soluble in polar organic solvents, such as DMF and DMSO.

3.2. Electrochemistry

The electrochemical properties of the Cu complexes of

bis(thiosemicarbazones) have important ramiﬁcations on their anti-

proliferative effects [14,20,36]. Previous structure–activity studies

[20,36,39,40] examining the cellular uptake of Cu(ATSM) and related

complexes have demonstrated that changes in backbone alkylation

can dramatically alter the biological properties of the resultant Cu

complex. This effect is believed to be largely due to changes to the

redox potential of the Cu complex [14,20,36]. Previous studies have

noted a redox potential-dependent rate of reduction, with more nega-

tive potentials favoring slower reduction rates, while more positive

redox potentials resulted in faster reduction rates [20,36,37]. In addi-

tion, computer modeling studies examining the competition between

dissociation and re-oxidation of the Cu complexes demonstrated that

Cu complexes with more negative redox potentials favored re-

oxidation, whereas those with more positive potentials favored dissoci-

ation [20]. Thus, we investigated the electrochemical properties of the

Cu complexes of all synthesized ligands by cyclic voltammetry and the

results are summarized in Table 1 and presented in Fig. 2A. In these

studies, 0.1 M Et₄NClO₄in DMSO was chosen as the solvent of choice be-

cause of the low aqueous solubility of some of the compounds.

2.4.6. Assessment of lysosomal membrane permeability

Distribution of acridine orange (AO; Sigma-Aldrich) was used to de-

termine LMP as previously described and was quantiﬁed with ImageJ

v1.48 (Wayne Rasband NIH, USA) [26,58]. Brieﬂy, cells were incubated

for 15 min/37 °C with AO (20 mM) then washed 3 times with PBS.

The cells were then incubated for 1 h/37 °C with chelators (2 to

25 μM) with or without a 10 min/37 °C pre-incubation with CuCl₂

(5 μM) prior to chelator addition. To examine the role of redox stress

in LMP, cells were pre-incubated with NAC (5 mM) or TM (5 μM) for

10 min/37 °C prior to treatment, while BSO (100 μM) was pre-

incubated for 16 h/37 °C [54,55]. Samples were examined with a Zeiss

Axio Observer.Z1 ﬂuorescence microscope (Zeiss, Oberkochen,

Germany) equipped with FITC and Texas Red ﬁlters. Images were cap-

tured with an AxioCam camera and AxioVision Rel. 4.7 Software (Zeiss).

2.4.7. Immunoﬂuorescence studies of lysosomal membrane

permeabilization

Lysosomal permeability was examined using immunoﬂuorescence

by following lysosomal cathepsin D release as previously reported

[58,59]. Lysosome-associated membrane protein 2 (Lamp2) was used

All synthesized Cu complexes exhibited reversible, one electron Cu^II/I

couples (Fig. 2A). Ligands with no or one alkyl group at R₁and R₂(i.e.,

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

25

Table 1

unsubstituted and monosubstituted ligands, respectively; Fig. 1B; denot-

ed in red) include Cu(GTS), Cu(GTSM), Cu(PTS) and Cu(PTSM). These

complexes exhibited more positive Cu^II/Iredox potentials between

−0.43 and −0.51 V (Table 1; Fig. 2A). Of these Cu complexes, Cu(GTS)

and Cu(GTSM) possessed the most positive Cu^II/Iredox potential of

−0.43 V (Table 1). The addition of a methyl group at R₁(i.e., Cu(PTS)

and Cu(PTSM)) resulted in more negative Cu^II/Iredox potentials (−0.50

to −0.51 V; Table 1). Interestingly, previous studies examining the

redox active DpT series of thiosemicarbazones have observed Cu^II/I

redox couples in a similar range as these Cu[bis(thiosemicarbazones)]

derived from unsubstituted or monosubstituted ligands [24].

Partition coefﬁcients (Log P_calc) and Cu^II/Iredox potentials of the ligands and Cu complexes,

respectively. Log P_calcvalues were calculated using the program ChemBioDrawUltra

v11.0.1 using Crippen's fragmentation procedure [51].

Ligand

Partition coefﬁcient

(Log P_calc

Unsubstituted/monosubstituted

Cu^II/Iredox potential

(V vs. NHE)

)

GTS

GTSM

PTS

0.45

0.84

0.53

1.45

−0.43

−0.50

−0.51

PTSM

Disubstituted

ATS

ATSM

CTS

CTSM

DTS

DTSM

CyTS

All complexes with alkyl groups at the R₁and R₂sites (i.e., disubsti-

tuted ligands; Fig. 1B, C; denoted in green), including Cu(ATS),

Cu(ATSM), Cu(CTS), Cu(CTSM), Cu(DTS), Cu(DTSM), Cu(CyTS) and

Cu(CyTSM), demonstrated completely reversible one electron reduc-

tions (Cu^II/I) between −0.58 and −0.62 V (Table 1). Alteration of the

length of the substituents on the diimine backbone, from methyl to

ethyl, did not appreciably affect Cu^II/Iredox potentials (Table 1). The in-

corporation of a six-membered ring into the diimine backbone in CyTS

0.65

1.48

1.34

2.69

1.69

2.34

2.15

2.78

−0.59

−0.58

−0.59

−0.60

−0.62

CyTSM

Fig. 2. (A) Cyclic voltammograms of 2 mM solutions of: [Cu(DTSM)], [Cu(CTSM)], [Cu(ATSM)], [Cu(PTSM)], and [Cu(GTSM)] (from top to bottom), showing the impact of the diimine back-

bone alkyl substituents on the Cu^II/Iredox potential. Sweep rate 10 mVs⁻¹, solvent 0.1 M Et₄NClO₄in DMSO. (B) The relationship between the Cu^II/Iredox potential and the anti-

proliferative activity (IC₅₀) of Cu[bis(thiosemicarbazones)].

26

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

and CyTSM (Fig. 1C) resulted in the lowest (most negative) Cu^II/Iredox

potential obtained (−0.62 V; Table 1).

ligands was signiﬁcantly (p b 0.001) greater than DFO (IC₅₀: 22.7

1.6 μM), but was signiﬁcantly (p b 0.01) less than that of Dp44mT

(IC₅₀: 0.004 0.001 μM; Table 2). Importantly, these unsubstituted or

monosubstituted ligands demonstrated signiﬁcantly (p b 0.01) greater

anti-proliferative activity than 3-AP (IC₅₀: 0.36 0.03 μM; Table 2). In

contrast, the analogs with alkyl substituents at both the R₁and R₂

sites on the diimine backbone (i.e., ATS, ATSM, CTS, CTSM, DTS, DTSM,

CyTS and CyTSM), demonstrated very poor anti-proliferative effects

against SK-N-MC cells (IC₅₀: ≥11.15 μM; Table 2).

In summary, the Cu[bis(thiosemicarbazones)] complexes can be di-

vided into two groups: (1) those derived from unsubstituted or

monosubstituted ligands with more positive Cu^II/Iredox couples

(−0.43 to −0.51 V; Table 1); and (2) those derived from disubstituted

ligands with more negative Cu^II/Iredox couples (−0.58 to −0.62 V;

Table 1). These results suggest that the Cu complexes derived from

unsubstituted or monosubstituted ligands, with more positive Cu^II/I

redox couples, may be more easily and rapidly reduced to form the

charged Cu^I(L)⁻complex relative to Cu complexes of the disubstituted

ligands. The differences in the electrochemical behavior of the

Cu[bis(thiosemicarbazone)] complexes from unsubstituted and

monosubstituted versus disubstituted ligands will probably result in

important differences in their ability to mediate ROS formation and in-

duce anti-proliferative activity (examined below).

3.3.2. Anti-proliferative activity of the copper complexes [Cu^II(L)] against

cancer cells

Previous studies have shown that complexation of

bis(thiosemicarbazone) ligands with metals results in marked changes

in biological activity [32,68,69]. To assess the importance of Cu chelation

on the anti-proliferative activity of the synthesized ligands, their Cu^II

complexes were prepared and their anti-proliferative activity was

assessed against SK-N-MC neuroepithelioma cells (Table 2). The Cu

complex of Dp44mT, [Cu(Dp44mT)]Cl, was also included as a control,

as its potent anti-proliferative activity has previously been character-

ized [24–26].

3.3. Biological studies

3.3.1. Anti-proliferative activity of the bis(thiosemicarbazone) ligands

against cancer cells

The ability of the bis(thiosemicarbazone) series to inhibit cellular

proliferation was assessed using SK-N-MC neuroepithelioma cells, as

the effect of other thiosemicarbazones and related aroylhydrazones on

these cells has been extensively examined [23,52,63]. Hence, they are

a well established model that enables comparisons to our previous

studies. These ligands were compared to a number of relevant positive

controls, including: (1) Dp44mT (Fig. 1A), a thiosemicarbazone with

potent anti-proliferative activity and chelation efﬁcacy that has been ex-

tensively studied as an anti-cancer agent [22,64]; (2) desferrioxamine

(DFO; Fig. 1A), used for the treatment of Fe overload [64]; and (3) 3-

AP (Fig. 1A) that has been investigated as an anti-cancer agent in clinical

trials [65–67]. In these studies, SK-N-MC cells were treated with various

concentrations (0.0015–25 μM) of the ligands and incubated for

72 h/37 °C. The concentration of the ligands that reduced cellular prolif-

eration to 50% of the untreated control (IC₅₀value) was determined and

the results are presented in Table 2.

Similarly to their respective ligands, the Cu^IIcomplexes derived from

unsubstituted or monosubstituted ligands (i.e., Cu(GTS), Cu(GTSM),

Cu(PTS) and Cu(PTSM)), possessed potent anti-proliferative activity

against SK-N-MC cells (IC₅₀: 0.009–0.016 μM; Table 2). However, the

anti-proliferative activity of these complexes was signiﬁcantly

(p b 0.01) less than that of Cu(Dp44mT), which resulted in an IC₅₀of

0.004

0.001 μM (Table 2). Although Cu(GTSM) (IC₅₀: 0.009

0.001 μM; Table 2) showed signiﬁcantly (p b 0.05) increased anti-

proliferative activity relative to the ligand alone (GTSM; 0.020

0.004 μM; Table 2), the other Cu^IIcomplexes derived from unsubstituted

or monosubstituted ligands (i.e., Cu(GTS), Cu(PTS), or Cu(PTSM)), dem-

onstrated comparable anti-cancer activity in comparison to their corre-

sponding free ligands (i.e., GTS, PTS and PTSM, respectively; Table 2).

In contrast, the Cu^IIcomplexes of the disubstituted ligands (i.e.,

Cu(ATS), Cu(ATSM), Cu(CTS), Cu(CTSM), Cu(DTS), Cu(DTSM),

Cu(CyTS) and Cu(CyTSM)), displayed signiﬁcantly (p b 0.01) increased

anti-proliferative activity (IC₅₀: 0.25–2.07 μM) compared to their corre-

sponding free ligands (Table 2). In fact, Cu complexation resulted in a

greater than a 6-fold increase in anti-cancer activity. All Cu complexes

of the disubstituted ligands had signiﬁcantly (p b 0.01) greater anti-

proliferative activity relative to DFO (Table 2). Excluding the three

All bis(thiosemicarbazone) ligands derived from glyoxal and

pyruvaldehyde (i.e., the unsubstituted or monosubstituted ligands),

namely: GTS, GTSM, PTS and PTSM (Fig. 1B) demonstrated potent

anti-proliferative activity against SK-N-MC cells (IC₅₀

: 0.017–

0.021 μM; Table 2). The anti-proliferative activity of this subset of

Table 2

IC₅₀(μM) values of bis(thiosemicarbazone) chelators and their Cu complexes at inhibiting the growth of SK-N-MC neuroepithelioma cells and mortal MRC-5 ﬁbroblasts as determined by

the MTT assay after a 72 h/37 °C incubation. Results are mean SD (3 experiments). The p values were determined using Student's t-test and compare the activity of the ligand or Cu

complex in normal and neoplastic cells. NS; not signiﬁcant.

Ligand

SK-N-MC IC₅₀(μM)

MRC-5 IC₅₀(μM)

p value

Therapeutic Index

Ligand

Cu^(II)(L)

Ligand

Cu^(II)(L)

Ligand

Cu^(II)(L)

Ligand

Cu^(II)(L)

Controls

DFO

Dp44mT

3-AP

22.7 1.6

0.004 0.001

0.36 0.03

–

N12.5

2.19 0.07

N12.5

–

0.004 0.001

–

Unsubstituted/monosubstituted

GTS

GTSM

PTS

0.021 0.002

0.020 0.004

0.017 0.003

0.016 0.005

0.009 0.001

0.012 0.001

0.016 0.001

8.12 0.48

6.15 0.66

4.73 0.75

4.91 0.74

0.30 0.04

0.27 0.05

0.44 0.05

0.65 0.08

p b 0.01

p b 0.05

387

308

278

289

19

30

37

41

PTSM

Disubstituted

ATS

ATSM

CTS

CTSM

DTS

DTSM

CyTS

N12.5

11.2 0.82

N12.5

0.68 0.13

0.46 0.17

0.25 0.04

0.76 0.04

0.51 0.04

1.83 0.08

1.56 0.13

2.07 0.23

N12.5

2.60 0.30

2.13 0.11

1.27 0.22

2.09 0.12

1.56 0.21

3.46 0.23

3.74 0.28

4.79 0.40

–

p b 0.05

–

N1.1

–

4

5

3

2

CyTSM

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

27

least active complexes, (Cu(DTSM), Cu(CyTS) and Cu(CyTSM)), all other

Cu complexes of the disubstituted compounds demonstrated compara-

ble anti-proliferative activity to the ligand, 3-AP (IC₅₀: 0.36 0.03 μM;

Table 2).

reincubated with medium (control) or medium containing the chela-

tors (25 μM) for 1 h/37 °C. The efﬂux of ⁶⁴Cu into the medium and the

levels of ⁶⁴Cu remaining in the cells were then assessed. The release of

⁶⁴Cu mediated by the bis(thiosemicarbazone) ligands was compared

to Dp44mT and the well characterized Cu chelators, TM (Fig. 1A), and

neocuproine (Neo; Fig. 1A) [26].

In summary, Cu complexation greatly enhanced the anti-

proliferative activity of disubstituted ligands. On the other hand, the

anti-cancer effects of the Cu complexes of the unsubstituted or

monosubstituted ligands were either enhanced or comparable to the

free ligand. These ﬁndings suggest that Cu complexation by

bis(thiosemicarbazones) plays a critical role in their anti-proliferative

activity. Additionally, examining the correlation between anti-

proliferative activity of the Cu complexes and their Cu^II/Iredox poten-

tials (R²= 0.75174; Fig. 2B), demonstrated the existence of two groups,

namely: (1) the unsubstituted and monosubstituted group with high

anti-proliferative activity and Cu^II/Iredox potentials clustered between

approximately −0.4 and −0.5 V (red symbols; Fig. 2B); and (2) the di-

substituted group, which demonstrated more negative Cu^II/Iredox po-

tentials and decreased anti-proliferative efﬁcacy (green symbols;

Fig. 2B).

The three controls, TM, Neo and Dp44mT all signiﬁcantly (p b 0.001–

0.01) reduced cellular ⁶⁴Cu release to 55 6%, 11 2% and 59 7% of

the control, respectively (Fig. 3A). Similar results demonstrating the

ability of Dp44mT and TM to inhibit ⁶⁴Cu release from SK-N-MC cells

has been reported previously [26]. The unsubstituted or

monosubstituted bis(thiosemicarbazone) ligands (i.e., GTS, GTSM, PTS

and PTSM) also led to signiﬁcantly (p b 0.001) decreased cellular ⁶⁴Cu

release relative to control medium (control) alone, reducing it to 16–

36% of the control (Fig. 3A). The cellular release of ⁶⁴Cu mediated by

GTS (16 0.5%) and GTSM (20 1%) was similar to the Cu chelator,

Neo, but signiﬁcantly (p b 0.01) less than TM or Dp44mT. In contrast,

three of the disubstituted bis(thiosemicarbazone) compounds (i.e.,

CTSM, DTSM and CyTSM) were signiﬁcantly (p b 0.01) more effective

at mediating ⁶⁴Cu mobilization relative to the control medium, increas-

ing ⁶⁴Cu release to 146–161% of the control (Fig. 3A). In contrast, their

terminal amine (R₃) unsubstituted counterparts (i.e., CTS, DTS and

CyTS) and the diacetyl bis(thiosemicarbazone) ligands (i.e., ATS and

ATSM) did not signiﬁcantly (p N 0.05) alter ⁶⁴Cu release relative to the

control (Fig. 3A).

3.3.3. Anti-proliferative activity against mortal cells

For a compound to be considered as a possible anti-tumor agent, it

must exhibit potent anti-proliferative activity against neoplastic cells,

and not markedly affect the proliferation of normal cells. To determine

whether these compounds exhibited selectivity towards neoplastic

cells, their effect on the proliferation of mortal MRC-5 ﬁbroblasts was

assessed (Table 2). Their selectivity was measured using a calculated

in vitro therapeutic index (Table 2). This parameter represents the

ratio of the IC₅₀values of normal versus neoplastic cells (i.e., IC₅₀MRC-

5/IC₅₀SK-N-MC; Table 2), with higher values representing greater se-

lectivity against cancer cells.

The unsubstituted or monosubstituted ligands, GTS, GTSM, PTS and

PTSM, were signiﬁcantly (p b 0.01) more effective in neoplastic cells

than normal MRC-5 ﬁbroblasts, suggesting an appreciable therapeutic

index (Table 2). In fact, these ligands displayed IC₅₀values between

4.73 and 8.12 μM in MRC-5 cells, with therapeutic indices ranging

from 278 to 387 (Table 2). In contrast, therapeutic indices of the disub-

stituted ligands could not be determined, or were N1.1 (i.e., for CTSM),

due to their low anti-proliferative activity against both SK-N-MC and

MRC-5 cells.

Generally, the unsubstituted or monosubstituted chelators possessing

high anti-proliferative activity (i.e., GTS, GTSM, PTS and PTSM; Table 2)

demonstrated a decrease in intracellular ⁶⁴Cu release relative to the con-

trol (i.e., an accumulation of intracellular ⁶⁴Cu). In contrast, the disubsti-

tuted ligands with poor anti-proliferative activity (Table 2)

demonstrated comparable or enhanced ⁶⁴Cu release relative to the con-

trol in SK-N-MC cells (Fig. 3A). This observation is in agreement with a

study demonstrating the cellular accumulation of ⁶⁴Cu complexes of

unsubstituted or monosubstituted bis(thiosemicarbazones), such as

GTS, GTSM, PTS and PTSM, and the release of ⁶⁴Cu complexes of disubsti-

tuted ligands, including ATS and ATSM, using mouse mammary EMT6

tumor cells [20].

Plotting ⁶⁴Cu release from cells against their one electron Cu^II/Iredox

potentials yielded a correlation (R²= 0.7861) and demonstrated the

existence of

2 groups, namely: (1) the unsubstituted and

Pre-complexation of the bis(thiosemicarbazone) ligands with Cu re-

sulted in an increase of anti-proliferative activity in MRC-5 cells relative

to the ligands alone (Table 2). This corresponded to a 2.6–27-fold

increase in anti-proliferative activity against MRC5 cells upon Cu

complexation. Importantly, the anti-proliferative efﬁcacy of the Cu^II

bis(thiosemicarbazone) complexes were signiﬁcantly (p b 0.05) de-

creased in normal MRC-5 cells relative to neoplastic SK-N-MC cells.

The Cu complexes of GTS, GTSM, PTS and PTSM displayed appreciable

therapeutic indices, ranging between 19 and 41, while the Cu

complexes of the disubstituted ligands were less selective, leading to

therapeutic indices between 2 and 5 (Table 2).

Collectively, the disubstituted ligands and their Cu complexes were

not particularly selective. In contrast, the unsubstituted or

monosubstituted ligands and to a lesser extent their Cu complexes pos-

sessed very high selectivity owing to their potent anti-proliferative ac-

tivity against neoplastic cells. Thus, GTS, GTSM, PTS and PTSM and

their Cu complexes have appreciable therapeutic indices in targeting

cancer cells over normal cells and show promise as potential anti-

cancer agents.

monosubstituted bis(thiosemicarbazones) (green; Fig. 3B); and

(2) the disubstituted bis(thiosemicarbazones) (red; Fig. 3B). The en-

hanced lipophilicity of the terminal amine (R₃) methylated disubstitut-

ed analogs (i.e., CTSM, DTSM and CyTSM; Log P_calc2.34–2.78; Table 1)

may have facilitated the passive diffusion of the ligand and their com-

plexes through the plasma membrane lipid bilayer relative to their ter-

minal amine unsubstituted analogs (i.e., CTS, DTS and CyTS; Log P_calc

1.34–2.15; Table 1), resulting in enhanced cellular ⁶⁴Cu release

(Fig. 3A). In fact, a plot of the ⁶⁴Cu release vs. Log P_calcvalues yielded a

correlation coefﬁcient (R²) of 0.7536 (Fig. 3C). Collectively, these results

suggest the importance of the lipophilicity of these agents and also their

Cu^II/Iredox potentials on their ability to mobilize cellular ⁶⁴Cu. Indeed,

the ligands that formed Cu complexes with more biologically accessible

reduction potentials and that were relatively more hydrophilic (i.e., GTS,

GTSM, PTS and PTSM; Log P_calc0.45–1.45; Table 1), led to less cellular

⁶⁴Cu release (Fig. 3A) and potent anti-proliferative activity (Table 2).

3.3.5. Release of ⁶⁴Cu from cells prelabeled with the

⁶⁴Cu[bis(thiosemicarbazone)] complexes

As a comparison to cellular labeling using ⁶⁴Cu alone (Fig. 3A), and to

further assess the ability of the bis(thiosemicarbazone) ligands to

mobilize intracellular ⁶⁴Cu, SK-N-MC cells were prelabeled with

bis(thiosemicarbazone) ⁶⁴Cu-complexes (1:1) and the release exam-

ined (Fig. 3D). Brieﬂy, in these studies, SK-N-MC cells were incubated

for 1 h/37 °C with the preformed ⁶⁴Cu[bis(thiosemicarbazone)] com-

plexes (25 μM). Then the cells were washed and re-incubated with

3.3.4. Bis(thiosemicarbazone)-mediated ⁶⁴Cu Efﬂux from cells prelabeled

with ⁶⁴Cu

To assess the role of Cu chelation in the anti-neoplastic activity of

these bis(thiosemicarbazone) analogs, their ability to mobilize cellular

⁶⁴Cu from pre-labeled SK-N-MC cells was assessed (Fig. 3A). Brieﬂy,

cells were prelabeled with ⁶⁴Cu for 1 h/37 °C, washed, and then

28

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

Fig. 3. The effect of bis(thiosemicarbazone) series ligands on: (A) ⁶⁴Cu cellular mobilization from SK-N-MC neuroepithelioma cells prelabeled with ⁶⁴Cu; (B) the relationship between ⁶⁴Cu

cellular mobilization and Cu^II/Iredox potential; and (C) the relationship between ⁶⁴Cu cellular mobilization and lipophilicity (Log P_calc). The effect of bis(thiosemicarbazone) series ligands

on: (D) release of ⁶⁴Cu from SK-N-MC cells prelabeled with bis(thiosemicarbazone) ⁶⁴Cu-complexes formed in situ; and (E) the relationship between ⁶⁴Cu cellular release in (D) and Cu^II/I

redox potentials. Results are mean SD (3 experiments). *, versus control, p b 0.05; **, versus control, p b 0.01; ***, versus control, p b 0.001.

media for 1 h/37 °C. The percentage of ⁶⁴Cu released from cells was then

quantiﬁed, and expressed as a percentage of that found for cells incubat-

ed with medium alone (i.e., control). The release of ⁶⁴Cu from cells after

labeling with the ⁶⁴Cu[bis(thiosemicarbazone)] complexes were

compared to the preformed ⁶⁴Cu complexes of Dp44mT, TM and Neo,

as relative controls (Fig. 3D).

demonstrated poor anti-proliferative activity (Table 2), showed signiﬁ-

cantly (p b 0.05) increased ⁶⁴Cu release compared to the ⁶⁴Cu

complexes of the unsubstituted or monosubstituted ligands (Fig. 3D)

that displayed potent anti-cancer effects (Table 2). However, the ⁶⁴Cu

complexes of the disubstituted ligands, ATS, ATSM, CTS and DTS, dem-

onstrated signiﬁcantly (p b 0.01–0.05) decreased ⁶⁴Cu release compared

to the control. The ⁶⁴Cu complexes of the remaining disubstituted li-

gands, namely, CTSM, DTSM, CyTS and CyTSM, mediated comparable

levels of ⁶⁴Cu release relative to the control (Fig. 3D). Similarly to the

cellular ⁶⁴Cu mobilization results above (Fig. 3A,B), a correlation (R²=

0.8229) was observed between ⁶⁴Cu release (Fig. 3D) and their Cu^II/I

redox potentials (Fig. 3E). Hence, irrespective of cellular prelabeling

with ⁶⁴Cu or ⁶⁴Cu-bis(thiosemicarbazone) complexes, a similar trend

in ⁶⁴Cu release was observed with the unsubstituted/monosubstituted

and disubstituted analogs, with again these two groups demonstrating

distinctly different activity.

The preformed ⁶⁴Cu complexes of TM, Neo and Dp44mT, resulted in

signiﬁcantly (p b 0.001) decreased ⁶⁴Cu release that was 50%, 19% and

39% of the control, respectively (Fig. 3D). Cells incubated with the

⁶⁴Cu complexes of GTS, GTSM, PTS and PTSM all demonstrated signiﬁ-

cantly (p b 0.001) decreased levels of ⁶⁴Cu release (20–27%) relative

to the control, and this was also signiﬁcantly (p b 0.01–0.05) less than

the release of ⁶⁴Cu complexes of TM and Dp44mT (Fig. 3D). In fact, the

⁶⁴Cu complexes of GTS, GTSM and PTS mediated comparable levels of

cellular ⁶⁴Cu release to that of ⁶⁴Cu-Neo (Fig. 3D). Additionally, the

⁶⁴Cu complexes of all disubstituted ligands, which previously

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

29

Previous structure–activity relationship studies have proposed that

bio-reduction of the complexed Cu^IIto Cu^Iby intracellular reductants

was critical for their cellular accumulation [38]. Upon reduction, the

formed Cu^Icomplex (e.g., Cu(GTSM)⁻) was trapped within the cell,

owing to its negative charge and it was suggested that the fate of the

Cu^Icomplex was dependent on its ability to be re-oxidized [37]. For ex-

ample, the Cu^Icomplexes with more negative Cu^II/Iredox potentials,

such as Cu(ATSM)⁻, were readily oxidized to regenerate the neutral

Cu^IIcomplex (e.g., Cu^II(ATSM)), which can diffuse out of the cell [37].

In contrast, Cu^Icomplexes with more positive Cu^II/Iredox potentials,

such as Cu(GTSM)⁻, were not readily oxidized, resulting in the intracel-

lular trapping of Cu [20]. Hence, the observed ⁶⁴Cu accumulation by

complexes with more positive (i.e., more readily reduced) Cu^II/Iredox

potentials (e.g., Cu(GTSM)) is in agreement with the proposed model

of redox potential-dependent intracellular reductive trapping [20].

(Fig. 4A and B; Table 3) of two highly potent unsubstituted or

monosubstituted bis(thiosemicarbazones), namely GTSM and PTSM,

relative to a disubstituted ligand with low anti-proliferative efﬁcacy,

ATSM (Table 2).

Incubation of SK-N-MC cells with TM resulted in an IC₅₀of N6.25 μM

(Table 3). Thus, TM alone at 5 μM did not result in any signiﬁcant cyto-

toxicity. Interestingly, in the presence of TM, the anti-proliferative activ-

ity of GTSM or PTSM was markedly decreased, leading to an increase in

the IC₅₀from 0.017–0.019 μM to N6.25 μM (Fig. 4A, Table 3). This alter-

ation represented a N329-fold decrease in the anti-proliferative activity

of GTSM and PTSM. In contrast, the anti-proliferative activity of ATSM

(IC₅₀N 6.25 μM) was greatly enhanced in the presence of TM, leading

to an IC₅₀value of 0.07 μM (Table 3). This represented a N89-fold in-

crease in the anti-proliferative activity of ATSM (Fig. 4A; Table 3). This

latter ﬁnding suggested that the addition of TM led to a synergistic in-

teraction with ATSM in regards to cytotoxicity via an unknown

mechanism.

Co-incubation of TM with the pre-formed Cu complexes of GTSM or

PTSM also resulted in a marked and signiﬁcant (p b 0.01) decrease in

their anti-proliferative activity from 0.037–0.039 μM to 0.19 μM,

which represented a 4.9 to 5.1-fold increase in their IC₅₀values

(Fig. 4B; Table 3). A smaller, but signiﬁcant (p b 0.05), 1.6-fold decrease

in anti-proliferative activity was also observed for Cu(ATSM) upon co-

incubation with TM viz., from 0.31 μM to 0.50 μM (Table 3).

3.3.6. Anti-neoplastic activity of potent bis(thiosemicarbazones) is

mediated by copper complexation

To further investigate the role of Cu chelation in the anti-neoplastic

activity of the bis(thiosemicarbazones), the ability of the well-known

Cu chelator, TM (Fig. 1A) [62], to modulate the anti-proliferative activity

of the bis(thiosemicarbazones) and their Cu complexes was examined.

Notably, TM is a Cu speciﬁc chelator with low toxicity that forms Cu

complexes [62], and markedly decreases the cytotoxicity of other potent

thiosemicarbazones [26]. In these studies, we assessed the effect of TM

(5 μM) on the anti-proliferative activity of the ligands and Cu complexes

These data suggest the importance of chelatable Cu in the anti-

proliferative activity of the GTSM and PTSM ligands. It can be speculated

Fig. 4. Anti-proliferative activity of: (A) bis(thiosemicarbazone) ligands, GTSM, PTSM and ATSM, and (B) their copper complexes, Cu(GTSM), Cu(PTSM) and Cu(ATSM), with or without a

10 min/37 °C pre-incubation with the non-toxic copper chelator, TM (5 μM). Anti-proliferative activity of the: (C) bis(thiosemicarbazone) ligands, GTSM, PTSM and ATSM, and (D) their

copper complexes, Cu(GTSM), Cu(PTSM) and Cu(ATSM), with or without a pre-incubation with NAC (5 mM) for 10 min or BSO (100 μM) for 16 h/37 °C. Results are mean

SD

(3 experiments).

30

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

Table 3

IC₅₀(μM) values of bis(thiosemicarbazone) chelators and their Cu complexes at inhibiting the growth of SK-N-MC neuroepithelioma cells in the presence of TM, NAC or BSO. Results are

mean SD (3 experiments). NS; not signiﬁcant.

Compound

IC₅₀(μM)

+ TM

+ NAC

+ BSO

IC₅₀(μM)

p value

IC₅₀(μM)

p value

IC₅₀(μM)

p value

TM

N6.25

GTSM

PTSM

ATSM

Cu(GTSM)

Cu(PTSM)

Cu(ATSM)

0.019 0.004

0.017 0.003

N6.25

0.037 0.005

0.039 0.002

0.31 0.027

N6.25

0.07 0.005

0.19 0.01

0.50 0.04

–

N6.25

1.14 0.03

0.68 0.03

0.74 0.07

–

0.014 0.001

0.014 0.002

0.26 0.05

0.015 0.001

0.020 0.001

0.12 0.02

NS

–

p b 0.05

p b 0.01

p b 0.05

p b 0.01

p b 0.05

that TM may not only be able to compete with GTSM and PTSM for cel-

lular Cu, but may also remove Cu from their pre-formed Cu complexes

to inhibit their anti-proliferative activity. Thus, Cu chelation by these

bis(thiosemicarbazones) is an important mechanism of their anti-

cancer activity.

(Fig. 4D; Table 3). In fact, the anti-proliferative activity of Cu(GTSM),

Cu(PTSM) and Cu(ATSM), were signiﬁcantly (p b 0.05) enhanced in

the presence of BSO, with IC₅₀values decreasing from 0.037–0.31 μM

to 0.015–0.12 μM (Fig. 4D, Table 3). This corresponded to a 2–2.6-fold

increase in anti-proliferative activity.

In summary, augmentation of GSH levels through incubation of cells

with the anti-oxidant, NAC, resulted in a decrease in the anti-

proliferative activity of the bis(thiosemicarbazones). This result was

observed for both the free ligands and pre-formed Cu complexes and

was most prominent for the more potent ligands, GTSM and PTSM. In

addition, attenuation of the ability of the cell to manage redox stress

by depleting cellular GSH with BSO potentiated the anti-proliferative

activity of the free ligands and their Cu complexes. These ﬁndings sug-

gest that bis(thiosemicarbazones) and their Cu complexes mediate

their anti-proliferative activity through the promotion of ROS

generation.

3.3.7. Anti-proliferative activity is modulated by intracellular glutathione

levels

The electrochemical studies reported in Table 1 illustrate the poten-

tial ability of Cu complexes of bis(thiosemicarbazones) to cycle between

the oxidized (Cu^II) and reduced (Cu^I) states at potentials accessible to

biological oxidants and reductants. In fact, the complexes with the

most positive redox potentials (i.e., Cu(GTS), Cu(GTSM), Cu(PTS) and

Cu(PTSM); Table 1) also possessed the highest anti-proliferative activity

(Table 2), and resulted in decreased ⁶⁴Cu release from cells (Fig. 3). The

presence of excess cellular Cu (both endogenous and as exogenous Cu

complexes) has previously been shown to be a potent oxidant, resulting

in the generation of ROS in cells [11,26]. Thus, we assessed whether the

anti-proliferative activity of the bis(thiosemicarbazones) and their Cu

complexes was related to their ability to generate oxidative stress. To

examine this, SK-N-MC cells were incubated for 72 h/37 °C with increas-

ing concentrations of GTSM, PTSM and ATSM, or their pre-formed Cu

complexes either: (1) in the presence of NAC (5 mM); or (2) with a

16 h/37 °C pre-incubation with BSO (100 μM) or media alone (Fig. 4C

and D). Notably, NAC is a well characterized anti-oxidant [54,55] and

is a precursor of glutathione (GSH), that has been shown to prevent

the redox-induced cytotoxicity of thiosemicarbazones (e.g., Dp44mT)

and their Cu complexes [26]. Conversely, BSO has been demonstrated

to increase the oxidative stress of Dp44mT by reducing cellular levels

of GSH through the inhibition of γ-glutamylcysteine synthetase, the

enzyme required for the ﬁrst step of GSH synthesis [26,70].

The addition of the anti-oxidant, NAC, led to a marked and signiﬁ-

cant (p b 0.01) decrease in the anti-proliferative activity of GTSM and

PTSM, increasing their IC₅₀values from 0.017–0.019 μM, respectively,

to N6.25 μM (Fig. 4C, Table 3). This represented a N329-fold decrease

in the anti-proliferative activity of GTSM and PTSM in the presence of

NAC. The IC₅₀of ATSM was not achieved in the absence or presence of

NAC (N6.25 μM; Table 3), although NAC slightly decreased the anti-

proliferative efﬁcacy of ATSM relative to ATSM alone (Fig. 4C). A

16 h/37 °C pre-incubation of SK-N-MC cells with BSO prior to the addi-

tion of GTSM or PTSM resulted in a slight, but not signiﬁcant (p N 0.05),

increase in their anti-proliferative activity (Fig. 4C; Table 3). However,

BSO pre-incubation resulted in a marked N24-fold decrease in the IC₅₀

of ATSM relative to ATSM alone (i.e., a decrease in IC₅₀from N6.25 μM

to 0.26 μM; Table 3; Fig. 4C).

3.3.8. Copper complexation is required for bis(thiosemicarbazone)-

mediated ROS generation

To further examine the relevance of ROS generation on the anti-

proliferative activity of bis(thiosemicarbazones), and their Cu

complexes, we examined their ability to catalyze the oxidation of the

non-ﬂuorescent redox probe, H₂DCF-DA, to its ﬂuorescent counterpart,

DCF, using SK-N-MC cells (Fig. 5). Importantly, DCF is a well-

characterized probe for assessing intracellular redox stress [22,57,71].

The SK-N-MC cells were incubated with GTSM (2 or 25 μM) or

Cu(GTSM) (5 μM) for 1 h/37 °C with or without a 10 min/37 °C pre-

incubation with CuCl₂(5 μM) and/or TM (5 μM). The concentrations

of chelators and complexes chosen represent the highest concentrations

that could be used without inducing excessive toxicity. For example,

pre-incubation of cells with CuCl₂(5 μM), prior to incubation with

GTSM (25 μM), resulted in loss of cell membrane integrity and the

DCF signal (data not shown). Thus, after pre-incubation with CuCl₂

(5 μM), cells were incubated with GTSM (2 μM) to ensure cellular viabil-

ity. In parallel studies, hydrogen peroxide (H₂O₂; 50 μM) was used as a

positive control [22]. The results were assessed by ﬂow cytometry

(Fig. 5A, B) and were quantiﬁed and expressed as a percentage of the

control (Fig. 5C).

The positive control, H₂O₂, led to a signiﬁcant (p b 0.01) increase in

DCF ﬂuorescence to 294

pre-incubation of cells with CuCl₂(5 μM) alone did not signiﬁcantly

(p N 0.05) alter DCF ﬂuorescence (94 7%) relative to control cells

17% of the control (Fig. 5C). In contrast,

(Fig. 5C). Incubation of cells with GTSM (25 μM) alone mediated a

small, but signiﬁcant (p b 0.05), increase in intracellular DCF ﬂuores-

cence to 112 10% of the control (Fig. 5C).

Incubation with NAC also had a pronounced effect on decreasing the

anti-proliferative activity of the Cu[bis(thiosemicarbazone)] complexes

(Fig. 4D, Table 3). In fact, incubation of the Cu(GTSM), Cu(PTSM) and

Cu(ATSM) complexes with NAC demonstrated a signiﬁcant (p b 0.01–

0.05) decrease in their anti-proliferative activity, resulting in a 31-, 17-

and 2.4-fold increase in their IC₅₀values, respectively (Table 3). Pre-

incubation with BSO to deplete cellular GSH led to an increase in anti-

proliferative activity of all the Cu[bis(thiosemicarbazone)] complexes

Incubation with GTSM (2 μM) after CuCl₂(5 μM) pre-incubation led

to an increase in DCF ﬂuorescence intensity, resulting in the peak being

shifted to the right relative to control cells (red shading; Fig. 5A). Quan-

tiﬁcation of these results demonstrated a substantial and signiﬁcant

(p b 0.01) increase in DCF ﬂuorescence to 188

16% of control

(Fig. 5C). Moreover, DCF ﬂuorescence from SK-N-MC cells incubated

with GTSM (2 μM) after preincubation with CuCl₂(5 μM) was also sig-

niﬁcantly (p b 0.01) increased relative to incubation with the GTSM

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

31

Fig. 5. Intracellular redox stress was measured using DCF in SK-N-MC cells and assessed with ﬂow cytometry. Cells were incubated with medium alone (control) or: (A) GTSM with a 10

min/37 °C pre-incubation with Cu (5 μM) and/or TM (5 μM); (B) Cu(GTSM), with or without a 10 min/37 °C pre-incubation with TM (5 μM); (C) Geometric mean of obtained peaks. Results

are mean SD (3 experiments). *, versus control, p b 0.05; **, versus control, p b 0.01; ***, versus control, p b 0.001.

ligand (25 μM) alone, despite the greater than ten-fold higher concen-

tration of the ligand. To investigate the effect of Cu chelation on the

ability of the bis(thiosemicarbazones) to mediate ROS generation,

cells were also pre-incubated with TM. Pre-incubation of cells with

TM and Cu led to a signiﬁcant (p b 0.01) decrease in GTSM (2 μM) and

complex, Cu^II(Dp44mT) [26]. This redox active Cu complex was found

to accumulate in lysosomes, inducing apoptosis by disrupting lysosome

integrity [26]. This mechanism of cell death has not been reported

for bis(thiosemicarbazones), and considering the reliance of

bis(thiosemicarbazones) anti-proliferative activity on cellular retention

of Cu (Fig. 3A, D) and Cu complexation (Figs. 4A, B; 5A–C), we examined

whether these agents could also target the lysosome. These studies

were initially performed using the lysosomotropic ﬂuorophore, AO

(Fig. 6), which becomes concentrated in lysosomes and results in a

granular red ﬂuorescence, while AO in the cytosol or nucleus

emits a diffuse green ﬂuorescence [26,72,73]. A reduction in granu-

lar red ﬂuorescence combined with an increased diffuse cytosolic

green ﬂuorescence, indicates relocation of AO from the lysosomes

to the cytosol, following changes in lysosomal membrane perme-

ability [58]. To examine the ability of GTSM to affect AO localiza-

tion, SK-N-MC cells were incubated with GTSM for 1 h/37 °C with

or without a 10 min/37 °C pre-incubation with Cu (5 μM). Addition-

ally, these cells were also pre-incubated with TM (5 μM) or NAC

(2 mM) for 10 min/37 °C, and/or BSO (100 μM) for 16 h/37 °C,

prior to a 1 h incubation with GTSM. The red ﬂuorescence intensity

was quantiﬁed with image processing and analysis software,

ImageJ v1.48 (Fig. 6).

Cu (5 μM)-mediated DCF ﬂuorescence to 105

10% of the control

(Fig. 5C).

The pre-formed Cu complex, Cu(GTSM), was also able to mediate a

signiﬁcant (p b 0.01) increase in DCF ﬂuorescence which was demon-

strated by a right shift in DCF ﬂuorescence intensity (red shading;

Fig. 5B) and an increased in DCF ﬂuorescence to 174 6% of the control

(Fig. 5C). Importantly, pre-incubation with TM signiﬁcantly (p b 0.01)

decreased the Cu(GTSM)-mediated DCF ﬂuorescence to 107 16% of

the control (Fig. 5C). This observation suggested that TM was able to re-

move Cu from the pre-formed complex.

Collectively, these results using DCF demonstrate the ability of GTSM

and its Cu complex to catalyze the formation of intracellular ROS. Impor-

tantly, this effect was reversed in the presence of the Cu chelator, TM,

demonstrating the importance of Cu chelation in GTSM-mediated ROS

generation.

3.3.9. Bis(thiosemicarbazones) target lysosomal integrity

Our previous studies revealed that the lysosome is an important

cellular target in the anti-cancer activity of the Cu thiosemicarbazone

Control SK-N-MC cells showed a granular red ﬂuorescence consis-

tent with AO concentrated in lysosomes (Fig. 6A(a)) [74]. No signiﬁcant

32

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

Fig. 6. (A) Acridine orange (AO) lysosomal stability study in SK-N-MC cells to assess lysosomal membrane permeabilization (LMP). (A) Cells were incubated with GTSM (2 or 25 μM) for

1 h/37 °C, with or without a Cu (5 μM) pre-incubation for 10 min/37 °C, and with pre-incubation with TM (5 μM) or NAC (5 mM) over 10 min/37 °C; or BSO (100 μM) for 16 h/37 °C.

(B) Quantiﬁcation of AO (red ﬂuorescence) using the image processing and analysis software, ImageJ v1.48. Results are typical of 3 experiments. **, versus control, p b 0.01. (For interpre-

tation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

(p N 0.05) alteration in red ﬂuorescence was observed upon incubation

of cells with either TM, NAC, BSO, BSO + NAC, or CuCl₂(Fig. 6A(b)–(f),

and 6B). Incubation of cells with GTSM alone resulted in a marked

Upon pre-incubation with the Cu chelator, TM (Fig. 6A(h)), or the

anti-oxidant, NAC (Fig. 6A(i)), the ability of GTSM to mediate the reloca-

tion of AO to the cytosol was signiﬁcantly (p b 0.01) inhibited relative to

GTSM alone (Fig. 6A(g)). In fact, TM or NAC signiﬁcantly (p b 0.01) in-

creased red ﬂuorescence in GTSM-treated cells to 93% and 86% of the

control, respectively, and this was comparable to control cells

(Fig. 6B). Pre-incubation with BSO (Fig. 6A(j)), prior to the addition of

GTSM, led to a signiﬁcant (p b 0.01) decrease in red ﬂuorescence

(19 8%; Fig. 6B) relative to the control (Fig. 6A(a)), or GTSM alone

and signiﬁcant (p b 0.01) decrease in red ﬂuorescence to 49

3%

(Fig. 6A(g), B) relative to the control (Fig. 6A(a), B). This was accom-

panied by an increase in cytosolic green ﬂuorescence consistent with

relocation of AO from the lysosomes to the cytosol, indicating that

GTSM caused lysosomal membrane permeabilization (LMP;

Fig. 6A(g)).

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

33

(Fig. 6A(g)), and resulted in marked changes in cellular morphology

(i.e., a clear decrease in cell size denoting possible damage). Co-

incubation of cells with GTSM, BSO and NAC (Fig. 6A(k)) resulted in a

signiﬁcant (p b 0.01) increase in red ﬂuorescence (56 10%; Fig. 6B)

relative to cells incubated with GTSM and BSO alone (Fig. 6A(j)).

Hence, NAC decreased the pronounced effect of BSO and GTSM and

led to an increase in cell size, suggesting less toxicity.

ﬂuorescence to 21 6% of the control cells (Fig. 7B). In contrast, incuba-

tion of MRC-5 cells with GTSM (2 μM) + CuCl₂(5 μM) did not result in a

signiﬁcant (p N 0.05) change in red ﬂuorescence (137 6%; Fig. 7A(h);

Fig. 7B) and was comparable to the control (Fig. 7A(e)).

The investigation above in Fig. 7 is consistent with the cellular

proliferation studies where the bis(thiosemicarbazones) and their Cu

complexes showed selective anti-cancer activity against neoplastic

SK-N-MC cells relative to mortal MRC-5 cells (Table 2). Together,

these studies suggest that the mechanism of this selectivity of

bis(thiosemicarbazones) towards neoplastic cells is related to suscepti-

bility to lysosomal injury. It can be speculated that this could potentially

be due to: (1) the greater metabolic turnover in tumor cells relative to

mortal cells [11]; (2) enhanced intracellular ROS in cancer cells [15];

and (3) the abnormal autophagic pathways present in neoplastic cells

that are dependent on the lysosome [13,77].

Pre-incubation of cells with CuCl₂(5 μM), prior to the addition of

GTSM (25 μM; i.e., denoted as GTSM + CuCl₂), led to a complete loss

of cells (data not shown) demonstrating the pronounced toxicity of

the combination, probably due to enhanced formation of the potently

cytotoxic, Cu–GTSM complex (Table 2). Thus, in these studies, the con-

centration of GTSM was decreased to 2 μM to obtain viable cells for ex-

amination. Despite the lower GTSM concentration, incubation with

GTSM + CuCl₂resulted in a signiﬁcant (p b 0.01) decrease in red ﬂuo-

rescence (19 5% vs. control; Fig. 6B), and the near disappearance of

red granular ﬂuorescence (Fig. 6A(l)). Simultaneous with this decrease

in red ﬂuorescence, was an increase in green ﬂuorescence, consistent

with the redistribution of AO from lysosomes to the cytosol following

the loss of lysosomal integrity (Fig. 6A(l)). Pre-incubation of cells with

TM (Fig. 6A(m)) or NAC (Fig. 6A(n)) signiﬁcantly (p b 0.01) increased

red ﬂuorescence to 88 10% and 83 9% of the control, respectively

(Fig. 6B). In fact, the level of red ﬂuorescence observed in

GTSM + CuCl₂-treated cells that received TM or NAC pre-treatment

was comparable to the control (Fig. 6A(a)) and there was only minimal

redistribution of AO to the cytosol (Fig. 6A(m), (n); Fig. 6B). This obser-

vation was consistent with TM or NAC preventing the ability of

GTSM + CuCl₂in disrupting lysosomal stability.

3.3.10. Co-localization of Lamp2 and cathepsin D

To further examine the effect of bis(thiosemicarbazones) and their

Cu complexes on the lysosome, we examined the effect of incubating

SK-N-MC cells with GTSM (2 μM) and CuCl₂(5 μM) on the intracellular

distribution of a major lysosome membrane glycoprotein [60], namely

Lamp2, and the lysosomal protease, cathepsin D (Fig. 8) [58,78]. Lyso-

some permeabilization leads to the redistribution of cathepsin D from

lysosomes into the cytosol and results in the initiation of lysosome-

dependent apoptotic signaling pathways [59,79]. Cathepsin D release

is an early event in the apoptotic cascade, preceding destabilization of

mitochondria, and the subsequent release of cytochrome c and caspase

activation [78].

A 16 h/37 °C pre-incubation of cells with BSO, prior to GTSM and Cu

treatment (i.e., GTSM + CuCl₂+ BSO), led to marked morphological

changes consistent with apoptotic budding (Fig. 6A(o)), a characteristic

sign of cell death [75]. This effect was accompanied by an almost com-

plete and signiﬁcant (p b 0.01) loss of red ﬂuorescence (9 8%) relative

to the control (Fig. 6B), suggesting extensive LMP [76]. Relative to

GTSM + CuCl₂+ BSO alone (Fig. 6A(o)), the addition of NAC prevented

the apoptotic changes (Fig. 6A(p)) and resulted in a signiﬁcant

(p b 0.01) increase in granular red ﬂuorescence (50 11%; Fig. 6B) rel-

ative to BSO alone. Collectively, these results demonstrate that GTSM is

able to mediate LMP in SK-N-MC cells. This effect can be potentiated by

BSO that depletes cells of the anti-oxidant, GSH [70], or prevented by

NAC that supplements cellular GSH [55]. Importantly, the effect of

GTSM on LMP is potentiated by the addition of Cu and prevented by

the non-toxic Cu chelator, TM. Hence, together with the results from

Fig. 5, it can be deduced that a redox active GTSM–Cu complex can in-

duce LMP.

To elucidate the mechanisms involved in the selectivity of the

bis(thiosemicarbazones) towards neoplastic cells (Table 2), mortal

MRC-5 ﬁbroblast cells were compared to SK-N-MC neuroepithelioma

cells using the lysosomotropic ﬂuorophore, AO (Fig. 7). As shown in

Fig. 6 examining SK-N-MC cells, both SK-N-MC (Fig. 7A(a)) and MRC-

5 cells (Fig. 7A(e)) incubated with control medium only (control)

showed a granular red ﬂuorescence consistent with AO concentrated

in lysosomes. Incubation of SK-N-MC cells (Fig. 7A(b)) or MRC-5 cells

(Fig. 7A(f)) with CuCl₂alone resulted in no signiﬁcant (p N 0.05) effect

relative to the respective controls (Fig. 7A(a) and (e)). Examining SK-

N-MC cells, GTSM (25 μM) led to a signiﬁcant (p b 0.01) decrease in

In these studies, SK-N-MC cells were incubated in 3 successive steps:

(1) for 10 min/37 °C with medium alone (control), or medium contain-

ing TM (5 μM), or NAC (2 mM); or alternatively, medium alone (con-

trol), or medium containing BSO (100 μM) for 16 h/37 °C; (2) medium

alone or medium containing CuCl₂(5 μM); and (3) GTSM (2 μM) for

1 h/37 °C. Immunoﬂuorescence microscopy was then used to examine

the effects of GTSM in the presence of CuCl₂on the localization of

Lamp2 (red) and cathepsin D (green; Fig. 8A). To stain nuclei, 4′,6-

diamidino-2-phenylindole (DAPI; blue) was used and was not marked-

ly affected by any of the treatments (Fig. 8A). Green ﬂuorescence inten-

sity was also quantiﬁed using the image processing and analysis

software, ImageJ v1.48 (Fig. 8B).

Untreated SK-N-MC control cells stained for Lamp2 and cathepsin D

displayed a granular/vesicular pattern consistent with the labeling of

lysosomes (Fig. 8A(b), (c)). Importantly, the overlay of Lamp2

(Fig. 8A(b)) and cathepsin D (Fig. 8A(c)) demonstrated a yellow punc-

tate pattern (Fig. 8A(d)) consistent with their co-localization [59]

and suggested the presence of intact lysosomes (Fig. 8A(d)). Previ-

ous studies have demonstrated that while Lamp2 is bound within

the lysosomal membrane as it is an integral membrane protein

[60], cathepsin D is a soluble enzyme within the lysosome [78] and

can be released into the cytosol upon LMP, providing an appropriate

marker for this event [59].

Incubation of SK-N-MC cells with CuCl₂alone also revealed a similar

pattern of punctate vesicles, showing co-localization of Lamp2 and

cathepsin D (Fig. 8A(h)). In fact, no signiﬁcant (p N 0.05) difference in

green ﬂuorescence was observed upon incubation with CuCl₂relative

to the control (Fig. 8B). However, GTSM in the presence of CuCl₂led to

a marked morphological alteration in Lamp2 (Fig. 8(j)) and cathepsin

D (Fig. 8(k)) ﬂuorescence relative to the respective controls (Fig. 8(b),

(c)). In fact, the vesicular distribution of staining in Fig. 8(b) and

(c) disappeared, and was replaced by homogenous clumps of staining

(Fig. 8(j), (k)), suggesting marked damage. In fact, incubation with

GTSM and CuCl₂led to a signiﬁcant (p b 0.001) loss of green cathepsin

red ﬂuorescence (56

12%) relative to the control (Fig. 7A(c) cf.

7A(a); Fig. 7B). In contrast, using mortal MRC5 cells, GTSM (25 μM) re-

sulted in a slight, but not signiﬁcant (p N 0.05) increase in red ﬂuores-

cence to 146

32% (Fig. 7A(g)) relative to the control (Fig. 7A(e);

Fig. 7B). These results demonstrate that neoplastic SK-N-MC cells

were more sensitive to LMP relative to mortal MRC5 cells.

Additionally, co-treatment of SK-N-MC cells with GTSM

(2 μM) + CuCl₂(5 μM) led to an almost complete loss of red ﬂuores-

cence (Fig. 7A(d)) relative to the control (Fig. 7A(a)). This observation

corresponded to a marked and signiﬁcant (p b 0.01) decrease in red

D ﬂuorescence to 47%

8% of the control (Fig. 8B). Importantly, the

overlay of Lamp2 and cathepsin D revealed a reduction in their co-

localization, leading to a decrease of yellow ﬂuorescence (Fig. 8A(l))

relative to the control (Fig. 8A(d)). This observation indicates the

34

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

Fig. 7. (A) Acridine orange (AO) lysosomal stability assay comparing the sensitivity of SK-N-MC neuroepithelioma cells versus MRC-5 lung ﬁbroblasts to lysosomal membrane perme-

abilization (LMP). (A) Cells were incubated with GTSM (2 or 25 μM) for 1 h/37 °C, with or without a 10 min/37 °C pre-incubation with CuCl₂(5 μM); and (B) Quantiﬁcation of AO (red

ﬂuorescence) with the image processing and analysis software, ImageJ v1.48. Results are typical of 3 experiments. **, versus control, p b 0.01. (For interpretation of the references to

color in this ﬁgure legend, the reader is referred to the web version of this article.)

redistribution of cathepsin D from the lysosome to the cytosol following

LMP after GTSM and CuCl₂treatment (Fig. 8A(l)). These results strongly

support the previously described AO studies examining SK-N-MC cells

(Figs. 6, 7) and suggests the damaging effects of GTSM and CuCl₂on ly-

sosome integrity (Fig. 9).

Pre-incubation of cells with the GSH synthesis inhibitor, BSO, prior to

incubation with CuCl₂and GTSM incubation, also resulted in a pro-

nounced reduction in punctuate, vesicular staining of Lamp2 and cathep-

sin D (Fig. 8A(n), (o)) relative to the respective controls (Fig. 8A(b), (c))

and reduced co-localization of Lamp2 and cathepsin D when overlaid

(Fig. 8A(p)) versus the control (Fig. 8A(d)). Additionally, a signiﬁcant

(p b 0.001) decrease in cathepsin D green ﬂuorescence was also observed

(37 12% of control; Fig. 8B). In contrast, pre-incubation of cells with the

anti-oxidant, NAC, prior to CuCl₂and GTSM incubation, helped to prevent

the loss of vesicular staining of Lamp2 (Fig. 8A(r)) and cathepsin D

(Fig. 8A(s)) relative to the respective controls (Fig. 8A(b), A(c)). Addition-

ally, the overlay of Lamp2 and cathepsin D demonstrated a yellow

punctate pattern consistent with their co-localization (Fig. 8A(t))

and suggested the presence of intact lysosomes. Moreover, the levels

of cathepsin D associated green ﬂuorescence (84

6% of control)

upon pre-incubation of NAC prior to CuCl₂and GTSM incubation

was only slightly reduced compared to the control (Fig. 8B).

As observed with NAC, pre-incubation of cells with the Cu chelator,

TM, led to similar protection from the GTSM and CuCl₂-induced loss of

vesicular Lamp2 (Fig. 8A(v)) and cathepsin D (Fig. 8A(w)) staining,

and in fact, their co-localization was maintained (Fig. 8A(x)), relative

to the respective controls (Fig. 8(A)b–d). These observations were also

was in good agreement with the AO studies in Fig. 6. In summary,

these ﬁndings suggest that the cytotoxicity of GTSM is mediated

through the formation of redox active Cu complexes that generate intra-

cellular redox stress that can be prevented by supplementation of GSH

through NAC or chelation of Cu using TM. This redox stress results in

LMP resulting in the release of cathepsin D. A schematic of this mecha-

nism is shown in Fig. 9.

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

35

Fig. 8. (A) Immunoﬂuorescence study examining the redistribution of soluble cathepsin D protein (green) from the lysosomes (labeled with Lamp 2 (red)) to the cytosol. SK-N-MC cells

were pre-incubated with medium alone or CuCl₂(5 μM), and/or BSO (100 μM) for 16 h, or NAC (5 mM) or TM (5 μM) for 10 min/37 °C. Cells were then incubated with GTSM (2 μM) for

1 h/37 °C. Nuclei are labeled with DAPI (blue). (B) Quantiﬁcation of cathepsin D (green ﬂuorescence) with image processing and analysis software, ImageJ v1.48. Results are typical of 3

experiments. **, p b 0.01; ***, versus control, p b 0.001. (For interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

4. Conclusion

intracellular mechanism of action. Knowledge of their mechanisms of

action is crucial to develop effective chemotherapeutics with high selec-

tivity and minimal toxicity. In the current investigation, we examined

the relationship between bis(thiosemicarbazone) structure, electro-

chemical behavior and anti-proliferative activity, and propose a novel

mechanism of bis(thiosemicarbazone)-induced cytotoxicity.

Bis(thiosemicarbazones) and their Cu complexes have been the

focus of a variety of medical applications due to their wide range of

pharmacological effects, including their anti-cancer activity [29–32].

Despite extensive study, there is an incomplete understanding of their

36

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

investigations are consistent with a redistribution of cathepsin D from

lysosomes to the cytosol following LMP. Signiﬁcantly, cathepsin D re-

lease into the cytosol triggers apoptosis through their interaction with

well-documented downstream effectors [27,78].

Signiﬁcantly, these results suggest that bis(thiosemicarbazones),

such as GTSM, behave in

a similar manner to that of the

thiosemicarbazone, Dp44mT, to induce their anti-cancer activity by

forming redox active Cu complexes that target the lysosome to mediate

LMP (Fig. 9) [26]. However, distinct differences in the structure-activity

relationships of both classes of ligands are obvious [23,63,80]. Impor-

tantly, aromatic and aliphatic substitutions at the imine carbon and ter-

minal N4 atom that result in enhanced lipophilicity played a critical role

in increasing the anti-proliferative activity of thiosemicarbazones [23,

63,80]. In contrast, disubstitution of the diimine backbone of the

bis(thiosemicarbazones) was detrimental to their anti-cancer activity.

This suggests that the inductive effects of these substitutions on the

Cu^II/Iredox potentials, rather than changes in their lipophilicity, played

a more signiﬁcant role in their anti-proliferative activity.

In summary, we propose a mechanism for bis(thiosemicarbazone)-

induced cytotoxicity involving copper complexation and intracellular

ROS generation, leading to LMP, the redistribution of cathepsins, and

the initiation of lysosome-centered cell death pathways. Hence, lyso-

somal targeting of cytotoxic drugs, such as some members of the

thiosemicarbazone class, is an important new therapeutic strategy

that deserves further investigation.

Fig. 9. Schematic diagram demonstrating the mechanisms of action of GTSM. GTSM is able

to form redox active Cu complexes that target the lysosome to induce lysosomal mem-

brane permeabilization, leading to the redistribution of cathepsins to the cytosol to induce

apoptosis.

Synthesis of a series of bis(thiosemicarbazone) ligands and their Cu

complexes revealed the importance of the alkyl substitutions at the

diimine positions (R₁and R₂) of the ligand backbone and yielded two

major groups, namely, the unsubstituted/monosubstituted and disub-

stituted bis(thiosemicarbazones), that had distinct electrochemical

and biological activity. The unsubstituted/monosubstituted ligands

(GTS, GTSM, PTS and PTSM), which demonstrated less negative Cu^II/I

redox potentials and were more hydrophilic, exhibited potent anti-

cancer activity with an appreciable in vitro therapeutic index and de-

creased ⁶⁴Cu release from cells. In contrast, the disubstituted ligands

(ATS, ATSM, CTS, CTSM, DTS, DTSM, CyTS, CyTSM), which displayed

more negative Cu^II/Iredox potentials and were more hydrophobic, dem-

onstrated poor anti-proliferative activity and greater ⁶⁴Cu mobilization

relative to their unsubstituted/monosubstituted counterparts. Signiﬁ-

cantly, the alkyl substitution pattern at the diimine positions governed

their: (1) Cu^II/Iredox potentials; (2) ability to mediate cellular ⁶⁴Cu re-

lease; (3) lipophilicity; and (4) anti-proliferative activity.

5. Conﬂict of interest

D.R.R. is a stakeholder in the companies, Oncochel Therapeutics LLC,

USA, and Oncochel Therapeutics Pty Ltd, Australia, that are developing

the thiosemicarbazone, DpC, for the treatment of cancer. D.R.R. also con-

sults for Oncochel Therapeutics LLC and Pty Ltd.

6. Author contributions

C.S. performed studies and wrote the manuscript; Z.A. performed

studies and prepared ﬁgures; P.J. designed studies and wrote the man-

uscript; D.S.K. and D.R.R. designed studies and wrote the manuscript

(participated as equal senior authors).

7. Abbreviations

Importantly, the anti-proliferative activity of these agents could be

inhibited by either: (1) Cu sequestration, using the non-toxic Cu chela-

tor, TM; or (2) supplementation of GSH levels, using NAC. Conversely,

GSH depletion using BSO potentiated their anti-proliferative effects.

These studies suggested the role of Cu chelation, Cu^II/Iredox cycling

and generation of ROS in their mechanism of anti-cancer activity. This

conclusion was supported by ﬂow cytometric studies with the redox

probe, DCF, which revealed signiﬁcant cellular ROS generation by

GTSM, which was dependent on the availability of chelatable Cu.

Morphological studies using the lysosomotropic agent, AO, indicated

that the lysosome was a potentially important target of the potent anti-

cancer bis(thiosemicarbazone), GTSM. Incubation of cells with GTSM in

the presence of supplemental Cu led to the disappearance of AO-stained

lysosomes, consistent with a relocation of AO from the lysosomes to the

cytosol, following LMP. Importantly, preservation of lysosomal integrity

was observed after supplementation of cellular GSH using NAC and Cu

sequestration by TM. Conversely, lyosomal injury increased upon the

depletion of the anti-oxidant, GSH, using the GSH synthesis inhibitor,

BSO.

3-AP

AO

ATS

ATSM

BSO

CTS

3-aminopyridine-2-carboxaldehyde thiosemicarbazone

acridine orange

diacetyl bis(thiosemicarbazone)

diacetyl bis(4-methyl-3-thiosemicarbazone)

buthionine sulfoximine

2,3-pentanedione bis(thiosemicarbazone)

2,3-pentanedione bis(4-methyl-3-thiosemicarbazone);

1,2-cyclohexanedione bis(thiosemicarbazone)

CTSM

CyTS

CyTSM 1,2-cyclohexanedione bis(4-methyl-3-thiosemicarbazone)

DCF

DFO

Dp44mT di-2-pyridylketone 4,4-dimethyl-3-thiosemicarbazone

DTS

DTSM

GSH

GTS

dichloroﬂuorescein

desferrioxamine

3,4-hexanedionebis(thiosemicarbazone)

3,4-hexanedionebis(4-methyl-3-thiosemicarbazone)

glutathione

glyoxal bis(thiosemicarbazone)

glyoxal bis(4-methyl-3-thiosemicarbazone)

GTSM

H₂DCF-DA 2′,7′-dichlorodihydroﬂuorescein diacetate

H₂O_2,

LMP

NAC

Neo

PTS

PTSM

hydrogen peroxide

lysosomal membrane permeabilization

N-acetyl-^L-cysteine;

These results observed using AO were further supported by

immunoﬂuorescence studies examining the intracellular distribution

of the well characterized lysosomal markers, Lamp2 and cathepsin D.

Importantly, a decrease in co-localization of Lamp2 and cathepsin D

was observed with GTSM incubation in the presence of Cu. These

neocuproine

pyruvaldehyde bis(thiosemicarbazone)

pyruvaldehyde bis(4-methyl-3-thiosemicarbazone)

C. Stefani et al. / Journal of Inorganic Biochemistry 152 (2015) 20–37

37

[38] R.I. Maurer, P.J. Blower, J.R. Dilworth, C.A. Reynolds, Y. Zheng, G.E. Mullen, J. Med.

Chem. 45 (2002) 1420 –1431.

[39] T.C. Castle, R.I. Maurer, F.E. Sowrey, M.J. Went, C.A. Reynolds, E.J.L. McInnes, P.J.

ROS

TM

reactive oxygen species

ammonium tetrathiomolybdate

Blower, J. Am. Chem. Soc. 125 (2003) 10040 –10049.

[40] P.J. Blower, T.C. Castle, A.R. Cowley, J.R. Dilworth, P.S. Donnelly, E. Labisbal, F.E.

Sowrey, S.J. Teat, M.J. Went, Dalton Trans. (2003) 4416 –4425.

[41] P.J. Blower, J.R. Dilworth, R.I. Maurer, G.D. Mullen, C.A. Reynolds, Y. Zheng, J. Inorg.

Biochem. 85 (2001) 15 –22.

Acknowledgments

This work was supported by a Project Grant from the National

Health and Medical Research Council (NHMRC) Australia to D.R.R.

[1021607] and D.S.K. [1048972]; a NHMRC Senior Principal Research

Fellowship to D.R.R. [1062607]; a NHMRC RD Wright Fellowship to

D.S.K. [1083057]; and a Helen and Robert Ellis Fellowship to D.S.K.

from the Sydney Medical School Foundation of The University of

Sydney. P.J.J. (10/CDF/2-15) thanks the Cancer Institute NSW for Fellow-

ship support. C.S. is the grateful recipient of an Australian Postgraduate

Award from the University of Sydney.

[42] F. Dehdashti, P.W. Grigsby, J.S. Lewis, R. Laforest, B.A. Siegel, M.J. Welch, J. Nucl. Med.

49 (2008) 201 –205.

[43] D.W. Dietz, F. Dehdashti, P.W. Grigsby, R.S. Malyapa, R.J. Myerson, J. Picus, J. Ritter,

J.S. Lewis, M.J. Welch, B.A. Siegel, Dis. Colon Rectum 51 (2008) 1641 –1648.

[44] S.I. Pascu, P.A. Waghorn, T.D. Conry, B. Lin, H.M. Betts, J.R. Dilworth, R.B. Sim, G.C.

Churchill, F.I. Aigbirhio, J.E. Warren, Dalton Trans. (2008) 2107 –2110.

[45] S. Lim, K.A. Price, S.F. Chong, B.M. Paterson, A. Caragounis, K.J. Barnham, P.J. Crouch,

J.M. Peach, J.R. Dilworth, A.R. White, P.S. Donnelly, J. Biol. Inorg. Chem. 15 (2010)

225 –235.

[46] S.I. Pascu, P.A. Waghorn, B.W. Kennedy, R.L. Arrowsmith, S.R. Bayly, J.R. Dilworth, M.

Christlieb, R.M. Tyrrell, J. Zhong, R.M. Kowalczyk, D. Collison, P.K. Aley, G.C. Churchill,

F.I. Aigbirhio, Chem. Asian. J. 5 (2010) 506 –519.

References

[47] D. Dayal, D. Palanimuthu, S.V. Shinde, K. Somasundaram, A.G. Samuelson, J. Biol.

Inorg. Chem. 16 (2011) 621 –632.

[1] Z.L. Harris, J.D. Gitlin, Am. J. Clin. Nutr. 63 (1996) S836 –S841.

[2] V.L. Goodman, G.J. Brewer, S.D. Merajver, Endocr. Relat. Cancer 11 (2004) 255 –263.

[3] H.W. Kuo, S.F. Chen, C.C. Wu, D.R. Chen, J.H. Lee, Biol. Trace Elem. Res. 89 (2002)

1 –11.

[48] K.A. Price, P.J. Crouch, S. Lim, B.M. Paterson, J.R. Liddell, P.S. Donnelly, A.R. White,

Metallomics 3 (2011) 1280 –1290.

[49] P.V. Bernhardt, L.M. Caldwell, T.B. Chaston, P. Chin, D.R. Richardson, J. Biol. Inorg.

Chem. 8 (2003) 866 –880.

[4] S. Apelgot, J. Coppey, A. Fromentin, E. Guille, M.F. Poupon, A. Roussel, Anticancer Res.

[50] M.C. Liu, T.S. Lin, A.C. Sartorelli, J. Med. Chem. 35 (1992) 3672 –3677.

[51] A.K. Ghose, G.M. Crippen, J. Chem. Inf. Comput. Sci. 27 (1987) 21 –35.

[52] D.R. Richardson, E.H. Tran, P. Ponka, Blood 86 (1995) 4295 –4306.

[53] D.R. Richardson, K. Milnes, Blood 89 (1997) 3025 –3038.

[54] O.I. Aruoma, B. Halliwell, B.M. Hoey, J. Butler, Free Radic. Biol. Med. 6 (1989)

593 –597.

[55] H.U. Simon, A. Haj-Yehia, F. Levi-Schaffer, Apoptosis 5 (2000) 415 –418.

[56] J. Liu, A. Hajibeigi, G. Ren, M. Lin, W. Siyambalapitiyage, Z. Lium, E. Simpson, R.W.

Parkey, X. Sun, O.K. Oz, J. Nuclear, Med. 50 (2009) 1332 –1339.

[57] O. Myhre, J.M. Andersen, H. Aarnes, F. Fonnum, Biochem. Pharmacol. 65 (2003)

1575 –1582.

6 (1986) 159 –164.

[5] S. Taysi, F. Akcay, C. Uslu, Y. Dogru, I. Gulcin, Biol. Trace Elem. Res. 91 (2003) 11 –18.

[6] A. Chan, F. Wong, M. Arumanayagam, Ann. Clin. Biochem. 30 (1993) 545 –549.

[7] M.C. Linder, J.R. Moor, K. Wright, J. Natl. Cancer Inst. 67 (1981) 263 –275.

[8] F.K. Habib, T.C. Dembinski, S.R. Stitch, Clin. Chim. Acta 104 (1980) 329 –335.

[9] S.K. Gupta, V.K. Shukla, M.P. Vaidya, S.K. Roy, S. Gupta, J. Surg. Oncol. 46 (1991)

178 –181.

[10] U. Carpentieri, J. Myers, L. Thorpe, C.W. Daeschner III, M.E. Haggard, Cancer Res. 46

(1986) 981 –984.

[11] A. Gupte, R.J. Mumper, Cancer Treat. Rev. 35 (2009) 32 –46.

[12] T. Wang, Z.J. Guo, Curr. Med. Chem. 13 (2006) 525 –537.

[13] F. Tisato, C. Marzano, M. Porchia, M. Pellei, C. Santini, Med. Res. Rev. 30 (2010)

708 –749.

[58] H.J. Yu, Y.F. Zhou, S.E. Lind, W.Q. Ding, Biochem. J. 417 (2009) 133 –139.

[59] P.J. Jansson, T. Yamagishi, A. Arvind, N. Seebacher, E. Gutierrez, A. Stacy, S. Maleki, D.

Sharp, S. Sahni, D.R. Richardson, J. Biol. Chem. 290 (2015) 9588 –9603.

[60] E.L. Eskelinen, Mol. Asp. Med. 27 (2006) 495 –502.

[14] C.R. Kowol, P. Heffeter, W. Miklos, L. Gille, R. Trondl, L. Cappellacci, W. Berger, B.K.

Keppler, J. Biol. Inorg. Chem. 17 (2012) 409 –423.

[61] T. Yamagishi, S. Sahni, D.M. Sharp, A. Arvind, P.J. Jansson, D.R. Richardson, J. Biol.

[15] D. Trachootham, J. Alexandre, P. Huang, Nat. Rev. Drug Discov. 8 (2009) 579 –591.

[16] 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) 5271 –5294.

[17] R.A. Finch, M.C. Liu, S.P. Grill, W.C. Rose, R. Loomis, K.M. Vasquez, Y.C. Cheng, A.C.

Sartorelli, Biochem. Pharmacol. 59 (2000) 983 –991.

[18] J.S. Casas, M.S. Garcia-Tasende, J. Sordo, Coord. Chem. Rev. 209 (2000) 197 –261.

[19] M. Whitnall, J. Howard, P. Ponka, D.R. Richardson, Proc. Natl. Acad. Sci. U. S. A. 103

(2006) 14901 –14906.

Chem. 288 (2013) 31761 –31771.

[62] Q. Pan, C.G. Kleer, K.L. van Golen, J. Irani, K.M. Bottema, C. Bias, M. De Carvalho, E.A.

Mesri, D.M. Robins, R.D. Dick, G.J. Brewer, S.D. Merajver, Cancer Res. 62 (2002)

4854 –4859.

[63] 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) 3716 –3729.

[64] D.S. Kalinowski, D.R. Richardson, Pharmacol. Rev. 57 (2005) 547 –583.

[65] J.E. Karp, F.J. Giles, I. Gojo, L. Morris, J. Greer, B. Johnson, M. Thein, M. Sznol, J. Low,

Leuk. Res. 32 (2008) 71 –77.

[20] J.L.J. Dearling, J.S. Lewis, G.E.D. Muller, M.J. Welch, P.J. Blower, J. Biol. Inorg. Chem. 7

(2002) 249 –259.

[66] B. Ma, B.C. Goh, E.H. Tan, K.C. Lam, R. Soo, S.S. Leong, L.Z. Wang, F. Mo, A.T. Chan, B.

Zee, T. Mok, Investig. New Drugs 26 (2008) 169 –173.

[67] M.J. Mackenzie, D. Saltman, H. Hirte, J. Low, C. Johnson, G. Pond, M.J. Moore, Investig.

New Drugs 25 (2007) 553 –558.

[68] D. Palanimuthu, S.V. Shinde, K. Somasundaram, A.G. Samuelson, J. Med. Chem. 56

(2013) 722 –734.

[21] S.N. Pandeya, D. Sriram, G. Nath, E. DeClercq, Eur. J. Pharm. Sci. 9 (1999) 25 –31.

[22] J. Yuan, D.B. Lovejoy, D.R. Richardson, Blood 104 (2004) 1450 –1458.

[23] D.R. Richardson, P.C. Sharpe, D.B. Lovejoy, D. Senaratne, D.S. Kalinowski, M. Islam,

P.V. Bernhardt, J. Med. Chem. 49 (2006) 6510 –6521.

[24] P.J. Jansson, C.L. Hawkins, D.B. Lovejoy, D.R. Richardson, J. Inorg. Biochem. 104

(2010) 1224 –1228.

[69] P.S. Donnelly, A. Caragounis, T. Du, K.M. Laughton, I. Volitakis, R.A. Cherny, R.A.

Sharples, A.F. Hill, Q.X. Li, C.L. Masters, K.J. Barnham, A.R. White, J. Biol. Chem. 283

(2008) 4568 –4577.

[25] P.J. Jansson, P.C. Sharpe, P.V. Bernhardt, D.R. Richardson, J. Med. Chem. 53 (2010)

5759 –5769.

[26] D.B. Lovejoy, P.J. Jansson, U.T. Brunk, J. Wong, P. Ponka, D.R. Richardson, Cancer Res.

[70] H. Tapiero, D.M. Townsend, K.D. Tew, Biomed. Pharmacother. 57 (2003) 386 –398.

[71] P. Ubezio, F. Civoli, Free Radic. Biol. Med. 16 (1994) 509 –516.

[72] M. Karlsson, T. Kurz, U.T. Brunk, S.E. Nilsson, C.I. Frennesson, Biochem. J. 428 (2010)

183 –190.

[73] E. Gutierrez, D.R. Richardson, P.J. Jansson, J. Biol. Chem. 289 (2014) 33568 –33589.

[74] H. Erdal, M. Berndtsson, J. Castro, U. Brunk, M.C. Shoshan, S. Linder, Proc. Natl. Acad.

Sci. U. S. A. 102 (2005) 192 –197.

[75] C.B. Thompson, Science 267 (1995) 1456 –1462.

[76] S. Aits, M. Jaattela, J. Cell Sci. 126 (2013) 1905 –1912.

[77] N. Chen, V. Karantza-Wadsworth, Biochim. Biophys. Acta 1793 (2009) 1516 –1523.

[78] P. Benes, V. Vetvicka, M. Fusek, Crit. Rev. Oncol. Hematol. 68 (2008) 12 –28.

[79] P. Boya, G. Kroemer, Oncogene 27 (2008) 6434 –6451.

71 (2011) 5871 –5880.

[27] N. Fehrenbacher, L. Bastholm, T. Kirkegaard-Sorensen, B. Rafn, T. Bottzauw, C.

Nielsen, E. Weber, S. Shirasawa, T. Kallunki, M. Jaattela, Cancer Res. 68 (2008)

6623 –6633.

[28] T. Kallunki, O.D. Olsen, M. Jaattela, Oncogene 32 (2013) 1995 –2004.

[29] F.A. French, B.L. Freedlander, Cancer Res. 18 (1958) 172 –175.

[30] F.A. French, B.L. Freedlander, Cancer Res. 18 (1958) 1290 –1300.

[31] V.C. Barry, M.L. Conalty, J.F. Osulliva, Cancer Res. 26 (1966) 2165 –2170.

[32] B.A. Booth, A.C. Sartorelli, Nature 210 (1966) 104 –109.

[33] H.G. Petering, G.J. VanGiessen, J.A. Crim, H.H. Buskirk, J. Med. Chem. 9 (1966)

420 –425.

[34] D. Kessel, R.S. McElhinney, Mol. Pharmacol. 11 (1975) 298 –309.

[35] J.A. Crim, H.G. Petering, Cancer Res. 27 (1967) 1278 –1285.

[36] J.L.J. Dearling, J.S. Lewis, D.W. McCarthy, M.J. Welch, P.J. Blower, Chem. Commun.

(1998) 2531 –2532.

[80] D.B. Lovejoy, D.M. Sharp, N. Seebacher, P. Obeidy, T. Prichard, C. Stefani, M.T. Basha,

P.C. Sharpe, P.J. Jansson, D.S. Kalinowski, P.V. Bernhardt, D.R. Richardson, J. Med.

Chem. 55 (2012) 7230 –7244.

[37] Y. Fujibayashi, H. Taniuchi, Y. Yonekura, H. Ohtani, J. Konishi, A. Yokoyama, J. Nucl.

Med. 38 (1997) 1155 –1160.

Article Doi

Identification of differential anti-neoplastic activity of copper bis(thiosemicarbazones) that is mediated by intracellular reactive oxygen species generation and lysosomal membrane permeabilization

DOI: 10.1016/j.jinorgbio.2015.08.010

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.jinorgbio.2015.08.010

Full text of DOI:10.1016/j.jinorgbio.2015.08.010

Products guided by the article

R&D Labs maybe for 18667-57-7

Relevant to this article

Hot Product