Chemico-Biological Interactions 311 (2019) 108789

Contents lists available at ScienceDirect

Chemico-Biological Interactions
journal homepage: www.elsevier.com/locate/chembioint

Unlike reactivity of mono- and binuclear imine-copper(II) complexes toward T

melanoma cells via a tyrosinase-dependent mechanism
Cléia Justino Nunesa, Andréia Hanada Otakeb, Silvina Odete Bustosb, Rodrigo Boni Fazzia,
Roger Chammasb, Ana Maria Da Costa Ferreiraa,*
a
Departamento de Química Fundamental, Instituto de Química, Universidade de São Paulo, São Paulo, 05508-000, SP, Brazil
b
Instituto do Câncer do Estado de São Paulo (ICESP), Faculdade de Medicina, Universidade de São Paulo, São Paulo, 01246-000, SP, Brazil

A R T I C LE I N FO A B S T R A C T

Keywords: The cytotoxicity of a dinuclear imine-copper (II) complex 2, and its analogous mononuclear complex 1, toward
Dinuclear imine copper complexes different melanoma cells, particularly human SKMEL-05 and SKMEL-147, was investigated. Complex 2, a tyr-
Cytotoxicity osinase mimic, showed much higher activity in comparison to complex 1, and its reactivity was verified to be
Melanoma cells remarkably activated by UVB-light, while the mononuclear compound showed a small or negligible effect.
Melanogenesis
Further, a significant dependence on the melanin content in the tumor cells, both from intrinsic pigmentation or
Clonogenic assays
stimulated by irradiation, was observed in the case of complex 2. Similar tests with keratinocytes and mela-
Modes of action
nocytes indicated a much lower sensitivity to both copper (II) complexes, even after exposition to UV light.
Clonogenic assays attested that the fractions of melanoma cells survival were much lower under treatment with
complex 2 compared to complex 1, both with or without previous irradiation of the cells. The process also
involves generation of reactive oxygen species (ROS), as verified by EPR spectroscopy, and by using fluorescence
indicators. Autophagic assays indicated a remarkable formation of cytoplasmic vacuoles in melanomas treated
with complex 2, while this effect was not observed in similar treatment with complex 1. Monitoring of specific
protein LC3 corroborated the simultaneous occurrence of autophagy. A balance interplay between different
modes of cell death, apoptosis and autophagy, occurs when melanomas were treated with the dinuclear complex
2, in contrast to the mononuclear complex 1. These results pointed out to different mechanisms of action of such
complexes, depending on its nuclearity.

1. Introduction mechanism [11]. However, most of these studies focus mononuclear

Antitumor metallodrugs have deserved increasing interest in the last
years, based in different metal ions [1–3], in an effort to surpassing
undesirable side effects, and induced resistance observed with clinically
used pharmaceuticals [4,5]. Among these metallodrugs, those based in
copper showed some motivating and peculiar properties that pointed
out to promising medicinal applications [6]. Copper compounds usually
act as redox active agents against tumors, damaging biomolecules and
organelles through the formation of reactive oxygen species (ROS)
[7–9], inducing apoptosis. On the contrary, some copper and other
metal-based pro-drugs can be activated in the reducing environment of
cancer cells, being more cytotoxic under hypoxia than in aerobic con-
ditions [10]. Instead, some tris(pyrazolyl)borate-copper(I) complexes
were reported as 26S proteasome inhibitors, associated to endoplas-
matic reticulum stress, and acting by a paraptosis-like cell death
copper species [12], active in catalytic cycles of monoelectronic steps.
Binuclear or higher nuclearity complexes seems to act differently,
allowing alternative approaches. Although some dinuclear copper
compounds, for instance with thiosemicarbazone ligands, were re-
ported to act by ROS-mediated pathways targeting mitochondria [13],
others are efficient hydrolytic agents toward phosphate bonds in nu-
cleic acids, mimicking metal-containing nucleases [14,15]. On the other
hand, new dinuclear mixed valence Cu(I,II) complexes with imidazolin-
4-ones derived ligands were reported to be active against different
tumor cells, acting as inhibitors of human telomerase, HIV reverse
transcriptase, and others polymerases, but not by intercalation at DNA
or causing significant DNA cleavage [16]. Also, some dinuclear com-
plexes with ligands derived from naphthalenediol were rationally de-
signed to bind irreversibly to neighboring phosphate groups in the
backbone of DNA structure [17], and its mode of action involves

*
Corresponding author. Instituto de Química, Universidade de São Paulo, Av. Professor Lineu Prestes 748, São Paulo, 05508-000, SP, Brazil.
E-mail address: amdcferr@iq.usp.br (A.M. Da Costa Ferreira).
URL: http://orcid.org/0000-0001-9898-1692 (A.M. Da Costa Ferreira).

https://doi.org/10.1016/j.cbi.2019.108789
Received 7 January 2019; Received in revised form 24 July 2019; Accepted 7 August 2019
Available online 08 August 2019
0009-2797/ © 2019 Elsevier B.V. All rights reserved.
C.J. Nunes, et al.
blocking DNA synthesis, rather than hydrolyzing phosphate bonds.
Therefore, nuclearity of copper complexes can be a determining factor
of its reactivity.
Melanoma is an aggressive cancer, metastatic, with high fatality if
diagnosed late, and frequently resistant to usual therapies [18–20]. It
arises from malignant transformation of melanocytes that are re-
sponsible for the synthesis of melanin, a dark polymeric product of
tyrosine oxidation, leading to pigmentation and consequent protection
of skin cells from the sun light. Since tyrosinase is also associated to
pathways in melanogenesis and shows an elevated activity in malignant
melanomas compared with healthy melanocytes, some potential pro-
drugs were reported to be activated by this enzyme in treatments
against melanomas [21,22].
In a previous work [23], we verified that some imine dinuclear
copper (II) complexes were much more reactive vs. melanomas than
analogous mononuclear ones. All the studied compounds targeted DNA,
by an oxidative mechanism. Also, these studies indicated a correlation
between the content of melanin and the sensitization of the cells,
especially toward the dinuclear complexes. Melanomas with higher
content of melanin (B16F10) were more susceptible to dinuclear com-
pounds than Tm1 cells, with little content of melanin. Herein, we de-
veloped additional investigations on the cytotoxicity of a dinuclear
imine copper (II) complex 2, in comparison to its analogous mono-
nuclear complex 1, toward different melanoma cells, trying to clarify
differences in its reactivity and better elucidate how its mode of action
depends on nuclearity.
2. Experimental section
2.1. Materials and methods
Most of the reagents 2-(acetyl)pyridine (99%), histamine hydro-
chloride (99%), melanin and copper (II) perchlorate hexahydrated
(98%) were purchased from Sigma-Aldrich Chemical Co. The copper
(II) complexes, mononuclear [Cu(apyhist)H2O](ClO4)2 1, and corre-
sponding dinuclear [Cu2(apyhist)2dpam](ClO4)4 2, have been prepared
and characterized in previous works [23]. The bridging ligand 2,2′-bis
(aminomethyl)biphenyl (dpam) was obtained according to a procedure
described in the literature [24]. Hydrochloric acid, and solvents ethanol
(and dimethylsulfoxide were from Merck Chemical Co. All the solutions
were prepared with deionized water from a Millipore instrument. EPR
spectra were registered in a Bruker EMX instrument, operating at X-
band (9.8 GHz), 20 mW power, and 100 kHz modulation frequency.
DMPO (5,5-Dimethyl-1-pyrroline N-oxide) was used as spin scavenger
in some experiments with hydrogen peroxide, and the formed radical-
adducts were detected at room temperature (298 K), using quartz flat
cells from Wilmad. Interactions with melanin were investigated by re-
gistration of EPR spectra in the presence of increasing amounts of
melanin, monitoring the environment around the copper ions, using
quartz flat cell from Wilmad, in DMSO/NaOH aqueous solution, at
room temperature. In this assay, increasing volumes of melanin solution
containing 5 mg melanin (Aldrich, synthetic) dissolved in 1 mL NaOH
solution (25 mM) were added to a metal solution (4 mM for complex 1
or 2 mM for complex 2) in the weight proportions indicated in the
Figures (from 0.5 to 2.0).
2.2. Cells culture
Murine cells Melan-a, Tm1 and Tm5 were cultivated in RPMI 5%
FBS +200 nM PMA (phorbol-1,2-myristate-1,3-acetate), while for
human pigmented cells SKMEL- 05 (B-Raf mutated gene expression)
and SKMEL- 147 (Wild Type) Dulbecco's Modified Eagle Medium
(DMEM) containing 10% fetal bovine serum (FBS) was used. Both
media were supplemented with 2 mM glutamine, 100 μg/mL penicillin,
100 μg/mL streptomycin, at 37 °C, and under 5% CO2.
2
Chemico-Biological Interactions 311 (2019) 108789
2.3. Cells viability
Cells were plated (1 × 104 cells/well) and incubated for 24 h, at
37 °C, and afterwards treated with each copper complex, for more 24 h.
The complexes were initially dissolved in a little amount of DMSO and
diluted, in RPMI or DMEM buffer, to the desired concentration in the
range 5–100 μM. The viability of both murine (Melan-a, Tm1 and Tm5)
and human SKMEL- 05 (ATCC HTB70, with B-Raf mutated gene ex-
pression), and wild type SKMEL- 147 (Cellosaurus, CVCL 3876) mela-
noma cells was carried out by MTT assay (3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide), monitoring the formation of the
blue colored formazan crystals at 570 nm, in a Tecan Infinite M200
microplate instrument (Switzerland). The respective IC50 values were
estimated through the Origin Program, using a dose response curve. In
the irradiated experiments, cells were initially exposed to UVB radia-
tion (at 14 ± 2 mJ/cm2), and subsequently treated with different
concentrations of the metal complexes, usually 20 μM for complex 1,
and 5 or 10 μM for complex 2, during 24 or 48 h, at 37 °C.
2.4. Melanin assays
Melanin content was measured as previously described elsewhere
[25]. Briefly, 24 h after the cells were plated, 1 × 107 cells were col-
lected and lysed in 1 M NaOH solution containing 10% DMSO. The
samples were then boiled for 1 h, and the absorbance of the supernatant
was determined at 475 nm, compared to a standard curve of melanin
(Sigma) in 1.0 M NaOH [26]. Quantification of melanin was also carried
out in melanoma cells, after being washed with HBSS (Hank's balanced
salt solution), exposed to UVB radiation (at 14 ± 2 mJ/cm2), and
subsequently treated with different concentrations of metal complexes,
usually 20 μM for complex 1, and 5 or 10 μM for complex 2, during 24
or 48 h.
2.5. Reactive oxygen species generation monitored by EPR spectroscopy
Formation of reactive oxygen species (ROS) was monitored by EPR
spectroscopy, using DMPO as spin scavenger [27,28]. In these experi-
ments, solutions of each complex (containing up to 10% DMSO) were
mixed to hydrogen peroxide that acts as a reducing agent, and DMPO in
phosphate buffer (pH 7.4, 50 mM), at room temperature, using quartz
flat cells, and tempol (36 μM) as frequency calibrator. Final con-
centrations: [complex] 100 μM; [H2O2] 250 μM; [DMPO] 100 mM.
2.6. Reactive oxygen species in mitochondria
After treatment of the melanoma cells (irradiated or not) with the
copper complexes, superoxide anions formation was selectively ana-
lyzed in mitochondria, through the oxidation of a fluorogenic dye.
Usually, 50 μg of MitoSOX™ (Invitrogen) mitochondrial superoxide in-
dicator was dissolved in 13 μL of dimethylsulfoxide (DMSO) to make a
5 mM MitoSOX™ reagent stock solution, and then diluted in suitable
solvent to get a 3 μM working solution. 30,000 treated cells were har-
vest with 1.0 mL this MitoSox Red working solution and incubated for
30 min, at 37 °C. Mitochondrial superoxide anions were then measured
by cytometry, at 488 nm excitation and 580 nm emission [29].
2.7. Clonogenic assays
These experiments can show the fraction of cells that survived to the
treatment with the metal complex, and still are able to form a colony
[30]. After being treated with complexes 1 or 2 for 48 h at 37 °C, with or
without exposure to radiation, the cells were washed with PBS, tryp-
sinized and counted. Subsequently, about 200 cells were plated
(21 cell/cm2) under culture condition for 10 days. Afterwards, they
were washed with PBS (twice), and finally dyed with 0.1% violet
crystal; the excess of dye was washed, and clones were counted.
C.J. Nunes, et al.
2.8. Autophagy tests
These tests were displayed in SKMEL-05 and SKMEL-147 cells, after
incubation with complex 2 [31]. After the treatment with the complex
for 24 h or 48 h, the cells were washed and counted. Subsequently,
1 × 104 cells/well was incubated with acridin orange (5 μg/mL) for
30 min, at 37 °C and under 5% CO2. After that, the fluorescence was
measured by flow cytometry (Attune). In inhibitory experiments, after
the treatment with the complexes the cells were washed and treated
with chloroquine (5 μM), for 16 h. Subsequently, they were incubated
with acridin orange, in a similar procedure described above.
2.8.1. Assays by Western blot to quantify autophagic proteins and
tyrosinase
SKMEL-05 and SKMEL-147 cells were plated at 1 × 106 cells/well.
Subsequently they were exposed to UVB radiation (14.31 ± 0.28 mJ/
cm2) and treated with the complexes. Then, the medium was aspirated,
cells were washed twice with PBS and proteins extracts were obtained
in lysis buffer containing 0.1 M Tris (pH 7.5), 150 mM NaCl, 1% Triton
X-100, 1 mM phenylmethane sulfonyl fluoride (PMSF, Sigma-Aldrich,
Saint-Louis, MO, USA) and 2 μg/mL aprotinin, followed by centrifuga-
tion at 13,000 g for 15 min.
2.8.2. Western blot analysis
For immunoblot analysis, whole-cell protein lysates were prepared
and analyzed via Western blotting. Proteins were separated by SDS-
polyacrylamide gel electrophoresis and transferred to a PVDF mem-
brane (Merck, Darmstadt, GE). The membrane was blocked with 5%
nonfat dry milk in Tris-buffered saline and incubated with primary
antibodies against: LC3 and β-actin (1:4000, Sigma-Aldrich, Saint-
Louis, MO, USA, AC-74); and tyrosinase (1:200, C-19, Santa Cruz
Biotechnology, SC-7833), and tubulin (1:1000, Calbiochem, USA,
DM1A) overnight at 4 °C. Blots were developed with a peroxidase-
conjugated secondary antibody, and proteins were visualized by en-
hanced chemiluminescence detection system (GE Healthcare). Band
densities were quantified using Image-J software.
3. Results and discussions
3.1. Cytotoxicity toward melanogenic cells
Complexes 1 and 2 (see Fig. 1) have been prepared according to
methods already described [23,32], isolated as perchlorate salts, and
characterized by spectroscopic methods (UV/Vis, IR, EPR) as well as by
elemental analysis [33]. Both have a di-imine ligand (apyhist) co-
ordinated to a copper (II) ion, and complex 2 exhibit two of these metal
centers, bridged by a diamine ligand (diphenyl-2,2′-aminomethyl,
dpam). A related complex [Cu2(apyhist)2Cl2](ClO4)2 had its crystal
Fig. 1. Studied imine-copper (II) compounds, isolated as perchlorate salts: mononuclear species 1 and corresponding dinuclear complex 2.
3
Chemico-Biological Interactions 311 (2019) 108789
structure determined previously by X-ray crystallography [34], and
when dissolved in aqueous solution gives rise to complex 1, due to the
lability of chloride ions.
Our previous studies indicated that these compounds are cytotoxic
towards melanogenic cells, being the dinuclear compound 2, a better
mimic of tyrosinases, more active than the analogous mononuclear
complex 1 [23]. Herein we confirmed and extended those data, ver-
ifying the influence of UVB light that stimulates melanogenesis, on the
increasing damage caused by these compounds in different melanoma
cells.
Initially, the viability of diverse murine cells (Melan-a, Tm1 and
Tm5 lines was monitored), after incubation with the studied copper
complexes. As shown in Fig. S1 in Supplementary data, the mono-
nuclear complex 1 was inactive, while the dinuclear complex 2 was
more toxic to all the lines tested. The observed toxicity of complex 2
depends on the melanin content in the cells. Tm1 cells with no melanin
were almost insensitive to this compound, while tumorigenic Tm5, and
non-tumorigenic pigmented Ma cells with high melanin concentration
were the most affected. Therefore, this order in reactivity correlates
well with the melanin content in these cells, corroborating previous
results [23].
Next, we showed the influence of these metal complexes in the
melanin content in human melanoma cells. We observed that SKMEL-
05 cells present more melanin than SKMEL-147 and both complexes
induce melanin production only in SKMEL-05, especially complex 2, as
shown in Fig. 2.
Afterwards, the viability of these SKMEL-05 and SKMEL-147 cells in
the presence of the studied metal complexes was verified by MTT as-
says. Fig. 3 shows the data for treatment with different concentrations
of each complex, indicating a much higher activity of dinuclear com-
plex 2 compared to that of mononuclear complex 1, slightly toxic up to
100 μM for both cells. The respective determined IC50 values for com-
plex 2 were (22.3 ± 1.9) μM against SKMEL-05, and (20.0 ± 7.0) μM
vs. SKMEL-147 cells respectively, attesting a similar toxicity of this
compound toward both types of cell.
3.2. Influence of irradiation
Experiments to verify the influence of light (UVB radiation, 14 mJ/
cm2) in the cytotoxicity of both complexes 1 or 2 were also performed,
toward SKMEL-05 and SKMEL-147 melanoma cells, after 24 or 48 h
treatment. The percentage of cells in sub G1 phase was monitored after
these treatments. We have studied several concentrations of complex 1
in SKMEL-05 and SKMEL-147 cells and we did not observe statistical
differences in sub-G1 population (Fig. S2, at Supplementary data).
However, treatment with complex 2 increased sub-G1 populations in
both cell lines and this effect were more pronounced after UVB ex-
posure, as shown in Fig. 4. Moreover, we did not observe differences in
C.J. Nunes, et al.
Fig. 2. Influence of complex 1 and complex 2 in the melanin content of 2 × 105
melanoma cells (SKMEL-05 and SKMEL-147), after treatment for 48 h, at 37 °C.
sensitivity of complex 2 between SKMEL-05 and SKMEL-147.
In similar experiments toward keratinocytes (HaCat cells), complex
2 was much less toxic than vs. tumorigenic melanomas, as displayed in
Fig. 4E. Further, the radiation effect observed in this case was quite
negligible.
Complementary photos of cells after 48 h treatment with these
complexes are displayed in Fig. 5, (see also Fig. S3 at Supplementary
data). Evidence of acidic vacuoles formation was observed in these
photos, as signaled by white arrows in those figures.
Usually, these tests were done with previous irradiation, and sub-
sequent treatment with the metal complexes. However, they were also
done with irradiation in the presence of the most active complex 2 and
the results showed that this complex is photochemically unreactive,
since irradiation of the cells before or after its addition gave very alike
results (data not shown).
Further experiments with different epithelial melanoma cells
(CHL01, UACC) confirmed these results, showing an increase in the
toxicity of complex 2 after UVB irradiation in all cases (shown in Fig.
S4, at Supplementary data). Therefore, all the experiments pointed to a
correlation between melanogenesis and the toxicity of the copper
complexes. This correlation was particularly crucial for the dinuclear
complex 2, mimic of tyrosinase, in comparison to its analogous
mononuclear complex 1.
3.3. Reactive oxygen species formation
Since copper ions are redox active and this property can be decisive
to its reactivity, tests to verify the formation of reactive oxygen species
(ROS) catalyzed by both complex 1 and 2, in the presence of hydrogen
peroxide, was monitored by EPR spectroscopy, using DMPO as spin
scavenger. Characteristic EPR spectra of DMPO-OH (4 lines,
aH = aN = 14.9 G) and DMPO-CH3 (6 lines, aH = 23.4 G, aN = 16.3 G)
Fig. 3. Human melanoma cells viability, monitored by MTT assay, after 24 h incubation with complexes 1 or 2 in different concentrations. A) SKMEL-05 cells; B)
SKMEL-147 cells.
4
Chemico-Biological Interactions 311 (2019) 108789
adducts were obtained for both complexes in the presence of hydrogen
peroxide, as displayed in Fig. 6. The methyl radicals are formed in the
reaction of •OH with dimethylsulfoxide, contained in solutions of the
complexes (up to 1% DMSO). Both were able of generating hydroxyl
radicals, capable of causing damage to diverse biomolecules.
These results explain the toxicity of both complexes towards mela-
nomas, as well as its ability in damaging DNA by an oxidative me-
chanism, as reported previously [23].
Additionally, the formation of ROS during the incubation of mela-
noma cells with the complexes studied were carried out in experiments
by flow cytometry using MitoSoxRed as indicator of superoxide ions.
The results attested the formation of significant amounts of O2•- prob-
ably by oxidative stress in mitochondria of both cells tested, after 24 or
48 h treatment with complex 2, as displayed in Fig. 7. This effect was
more significant in the case of SKMEL-05 cells, indicative of an oxida-
tive mechanism. Both complexes were more reactive after UVB irra-
diation (12.8 mJ/cm2), but complex 1 was much less active, showing
the same level of superoxide ions as the control.
New experiments to measure mitochondrial mass using MitoTracker
Green dye were performed. We did not observe any difference in
SKMEL-05 cells after treatment with complex 2 and neither after irra-
diation. On other hand, we verified an increase in mitochondrial mass
in SKMEL-147 cells after complex 2 treatment and after irradiation, as
shown in Fig. S5, at Supplementary data.
3.4. Clonogenic assays
Clonogenic tests were also developed to verify the possibility of the
melanoma cells survive after treatment with the studied complexes.
These tests showed a very low fraction of survival when cells were
treated with dinuclear complex 2 (C2 = 10 μM), in comparison to
analogous complex 1 (C1 = 20 μM), as displayed in Figs. 8 and 9.
The survival fraction in SKMEL-05 and SKMEL-147 cells after
complex 2 treatment were evaluated, showing a decrease in colony
formation even in irradiated cells. In these conditions, SKMEL-05 cell
were more sensitive after complex 2 treatment followed by irradiation,
with (29.7 ± 3.2) % survival for non-irradiated cells, and only
(12.3 ± 1.5) % for irradiated ones compared with SKMEL-147 cells.
For SKMEL-147 cells, the behavior was analogous regarding the
fraction of cells in sub G1 phase after irradiation, with a higher effect
observed in the case of complex 2, similarly to what was observed for
SKMEL-05 cells. However, in this case, a decrease in the fraction of
survival was observed after treatment with each complex, being com-
plex 2 more reactive.
3.5. Interactions of metal complexes with melanin
Melanin can act as good coordinating agent for metal ions as copper
or iron, and after binding to DNA, it can promote the cleavage of DNA
strands by ROS [35]. It can also prejudice the access of repair enzymes
to the caused lesions, therefore improving the extension of damage. As
C.J. Nunes, et al. Chemico-Biological Interactions 311 (2019) 108789

Fig. 4. Effect of UVB irradiation in the cytotoxicity of complexes 1 (20 μM) and 2 (5 or 10 μM) toward melanoma cells SKMEL-05 and SKMEL-147 cells, after
treatment for 24 h (A, B; 14.2 mJ/cm2) or 48 h (C, D; 13.6 mJ/cm2). (E) Keratinocyte cells viability, without exposition or after exposition to UVB radiation (14.0 mJ/
cm2), and subsequent treatment for 48 h with complex 2 up to 10 μM. Sub G1 populations were evaluated by propidium iodide incorporation.

shown in Fig. 10, interactions between melanin, dissolved in NaOH
solution, with the copper complexes were monitored by adding in-
creasing amounts of melanin to each complex, and registering the
corresponding EPR spectra.
The obtained curves show that both complex 1 and 2 give spectra
very similar, with parameters giso = 2.098 and Aiso = 66G for complex
1, and giso = 2.099 and Aiso = 62G for complex 2. By increasing the
melanin concentration, the characteristic hyperfine structure of the
metal becomes less intense, as displayed in Fig. 10A and B (up to blue
curves e), for [CuL]:[melanin] ratio 1:1. Analyzing curves f in Fig. 10A
and B, when this ratio is 1:2, a change from isotropic to axial geometry
around copper is noticed, attesting the binding of the metal ion at co-
ordinating groups of melanin. This coordination would restrict the
complex movement in solution, explaining the different spectra geo-
metries. The determined parameters change to g⊥ = 2.060, g//
= 2.197, and A// = 66G. On the other hand, when the copper-aqua

5
C.J. Nunes, et al.
Fig. 5. Sensitization of melanoma cells to death induced by complex 2 after previous exposure to UVB light. SKMEL-05 and SKMEL-147 cells were exposed to UVB
light and then treated with complex 1 (mononuclear) or complex 2 (dinuclear) for 48 h. White arrows indicate cytoplasmic vacuoles accumulation in SKMEL-05 cells
after complex 2 treatment. C1: 20 μM complex 1; C2: 5 or 10 μM complex 2.
Fig. 6. Reactive oxygen species (ROS) formation, monitored by EPR spectro-
scopy using DMPO as spin scavenger, in solutions of hydrogen peroxide
(250 mM) and complex 1 (C1 = 100 μM), or complex 2 (C2 = 100 μM), in the
presence of DMPO (100 mM).
complex was used the copper hyperfine structure was not observed, as
shown in Fig. 10C.
Both copper complexes seem to coordinate to melanin, showing
different spectroscopic parameters when compared to those obtained by
addition of free copper ions, or aqua-complex. Further, the dinuclear
complex 2 seems to be reduced more efficiently by the melanin than the
mononuclear ones or the aqua complex (see Fig. 10D). The differences
observed in the reactivity of these complexes toward melanoma cells
6
Chemico-Biological Interactions 311 (2019) 108789
could result from its facility on being reduced, triggering the catalytic
oxidative cycle with formation of ROS.
3.6. Assays of autophagy
Cells can be damaged and die by distinct mechanisms in the pre-
sence of antitumor compounds, including autophagy and necroptosis,
in addition to apoptosis [36]. Usually, different types of death overlap,
and the determination of relationships between them can be helpful in
stablishing strategies against cancer [37].
Autophagy is a cellular process involving engulfment of proteins
and organelles in vesicles named autophagosomes, where they are de-
graded by lysosomal proteases [38]. There are some evidence for
aberrant control of autophagy in cancer: it suppresses tumor growth at
early stages, but promotes its growth in advanced stage [39]. Further,
autophagy plays a role in melanogenesis by regulating melanosome
degradation and biogenesis in melanocytes [40]. Therefore, autophagy
modulation could be a therapeutic tool for advanced cancer as mela-
nomas.
Since after treatment with complex 2, it was observed a remarkable
formation of cytoplasmic vacuoles (see Fig. 5), a more accurate ver-
ification of autophagy process was performed. These assays were done
using acridine orange as indicator (see Experimental section), before
and after UVB irradiation of the cells, and for both lineages the same
level of vacuoles was detected in control experiments (~1.8%). Further,
there was no increase in fluorescence when complex 1 was used,
without and after irradiation, similarly to what is observed in control
experiment, as shown in Fig. 11.
On the contrary, a significant increase in the level of vacuoles was
detected for both cells, dependent on the concentration of complex 2.
However, only for SKMEL-147 cells was verified an influence of light,
i.e., formation of more cytoplasmic vacuoles after UVB irradiation. By
C.J. Nunes, et al. Chemico-Biological Interactions 311 (2019) 108789

Fig. 7. Superoxide ion formation (monitored by fluorescence measurements) in mitochondria of SKMEL-05 cells (A) after 24 h, (C) after 48 h; and SKMEL-147 cells,
(B) after 24 h and (D) 48 h treatment at 37 °C, with complex 1 (C1 = 20 μM), or complex 2 (C2 = 10 μM), using MitoSox red as indicator.

treatment with 10 μM of complex 2, an increase to 5% in orange
fluorescence relative to control was verified in experiment without ir-
radiation, while with previous irradiation an 8% fluorescence level was
achieved. In the case of SKMEL-05 cells, the same level of orange
fluorescence was observed for non-irradiated and irradiated cells
(~5%). Therefore, in these autophagy assays, complex 2 behaved dif-
ferently than complex 1, showing a significant formation of cytoplasm
vacuoles that increased by effect of light over SKMEL-147 cells. These
data reveal that autophagy process is probably induced during
treatment of both melanomas with this dinuclear compound, and the
process is especially favored by light in SKMEL-147 cells. Similar effect
was observed by treating breast cancer cells (MCF-7) with some Schiff
base-copper (II) complexes, for 72 h, at a concentration up to 100 μM
[41].
Additional experiments, using chloroquine as inhibitor of autop-
hagy were also performed, and corroborated the previous data for both
melanoma cells (see Fig. S6, at Supplementary data). Chloroquine acts
at late stage of autophagic process. When protonated it diffuses into

Fig. 8. (A) Clonogenic assay performed in 6-well plates with clones from SKMEL-05 control cells (first line) or after treatment with C1 = 20 mM (second line) or C2
= 10 mM (third line). On the left: non irradiated group, on the right: UVB irradiated group (13.9 mJ/cm2). (B) Surviving fraction of SKMEL-05 derived from
clonogenic assay experiment.

7
C.J. Nunes, et al. Chemico-Biological Interactions 311 (2019) 108789

Fig. 9. (A) Clonogenic assay performed in 6-well plates with clones from SKMEL-147 control cells (first line) or after treatment with C1 = 20 mM (second line) or C2
= 10 mM (third line). On the left: non irradiated group, on the right: UVB irradiated group (13.9 mJ/cm2). (B) Surviving fraction of SKMEL-147 derived from
clonogenic assay experiment.

Fig. 10. EPR spectra of melanin solution (0.5 g/mL) in NaOH 25 mM, after addition of increasing amounts of (A) complex 1 (4 mM), (B) complex 2 (2 mM), or (C)
copper aqua complex (4 mM); (D) Decreasing of copper signal in EPR spectra with the addition of increasing amounts of melanin.

lysosomes, where it is deprotonated and can raising the pH of lysosomal
vacuoles, inhibiting its capacity of acidic degradation [42].
When treated with complex 1, there was no change in orange
fluorescence level in both cells, in comparison to control experiments,
in the presence of only chloroquine. However, when treated with
complex 2, a significant increasing in orange fluorescence was observed
for SKMEL-147 cells. With previous irradiation of the cells, the effect of
complex 1 was very little or negligible in both lineages. On the con-
trary, a noticeable decreasing effect was verified when SKMEL-147 cells
were irradiated and treated with complex 2, in contrast with irradiated

8
C.J. Nunes, et al. Chemico-Biological Interactions 311 (2019) 108789

Fig. 11. Fluorescence data in autophagy assays with melanoma SKMEL-05 (A) or SKMEL-147 lineages (B), without (control) and after irradiation (13.0 mJ/cm2),
followed by 48 h treatment with the complexes 1 (C1 = 20 μM), and 2 (C2 = 5 and 10 μM). Acridine orange was used as indicator in flow cytometry measurements.

SKMEL-05 cells that showed a very little increasing in orange fluores-
cence. These data indicated an inhibitory effect of chloroquine only for
complex 2, and this effect was more noticeable in irradiated SKMEL-
147 cells. Complementary photos of the treated cells in these experi-
ments are also displayed in Figs. S7 and S8, at Supplementary data.
Trying to better elucidate possible mechanisms of action, the
monitoring of autophagy-specific proteins LC3 (microtubule-associated
protein light chain 3) and tubulin was performed, in parallel with tyr-
osinase, after 6 h treatment of the melanoma cells with each metal
complex. Tubulin was used to attest the loading of proteins in these
tests. LC3 proteins are necessary for the formation of autophagosomes,
while tubulin is a precursor of microtubules, which are responsible for
the fusion of autophagosomes with lysosomes to produce autolyso-
somes, facilitating autophagy [43,44].
Results, shown in Fig. 12, indicated that tyrosinase increases by
effect of light in SKMEL-05 cells, and on the contrary, appears almost
insensitive in SKMEL-147 cells. LC3 was not very sensitive to irradia-
tion, in both cells used. Also, both complexes 1 and 2 seem not to affect
tyrosinase protein level intensely, without or with irradiation, in both
types of melanomas, as already verified in the viability studies. How-
ever, complex 1 as well as complex 2 influenced remarkably the level of
LC3 in SKMEL-05 cells. Complex 1 decreased LC3 level under irradia-
tion, as attested by the [LC3]/[tubulin] ratio = 0.2, vs. 0.8 in control.
On the contrary, complex 2 modifies this protein level both without or
with irradiation, as detected by the increased [LC3]/[tubulin]
ratio = 0.6 without irradiation, or 1.6 under effect of light. These
changes in LC3 protein level, leads additionally to changes in the
[tyrosinase]/[tubulin] ratio in the presence of complexes, going from
0.6 to 2 in the presence of complex 1 (20 μM), and from 0.8 to 1.5 in the
presence of complex 2 (10 μM), respectively. In contrast, LC3 and tu-
bulin proteins were very little affected in SKMEL-147 cells by the
treatment with metal complexes.
Therefore, the noteworthy increasing in acidic cytoplasm vacuoles
verified in SKMEL-05 cells after treatment with complex 2 (see Fig. 5)
could be explained by its remarkable effect on increasing LC3 protein
amounts, indicating that autophagy is also operating in this case.
4. Conclusions
Our studies showed that the dinuclear copper (II) complex 2 causes
a remarkable sensitization of different melanoma cells, especially in
human SKMEL-05 and SKMEL-147 lineages, in a range of few μM, after
24 or 48 h treatment, and that this effect is significantly stimulated by
UVB radiation. On the contrary, with the analogous mononuclear
complex 1 this sensitization as well as the light effect are small or even
negligible. Non-tumorigenic cells are much less sensitive to both

Fig. 12. Analysis of LC3, tubulin, and tyrosinase proteins, after 6 h treatment of SKMEL-05 and SKMEL-147 cells with complex 1 (C1 = 20 μM) or complex 2
(C2 = 10 μM), in comparison to control (C).

9
C.J. Nunes, et al.
complexes, before and after exposition to radiation.
For complex 2, mimic of tyrosinase, a correlation between the
caused damage and the melanin content in the cells was observed.
Melanoma cells with higher content of melanin, or after exposition to
UVB radiation that induces melanogenesis, were more sensitive to it.
These data pointed to a different mode of action of dinuclear complex 2
against melanomas, in comparison to the analogous mononuclear 1. On
the contrary, this effect was not detected with complex 1, or toward
non-tumor keratinocytes. Therefore, this effect seems to be selective for
complex 2 toward melanoma cells.
Additional clonogenic tests were consistent with these data and
demonstrated that complex 2 affects more severely the capacity of
melanomas to form colonies than its analogous mononuclear complex
1. Also, under irradiation a significant decrease in survival was verified
for both complexes and both type of cells tested, but it was more re-
markable in the case of complex 2. Since under this condition mela-
nogenesis is improved, the formation of ROS is enhanced under the
effect of light, due to the catalytic cycle involving copper, and therefore
inducing apoptosis. This effect is more remarkable with complex 2.
Moreover, autophagy assays using orange acridine as indicator re-
vealed that, differently from complex 1 the dinuclear complex 2 leads
to a significant increase in the level of cytoplasmic vacuoles, which is
indicative of an autophagic death process. Further evidence for an au-
tophagic process is the inhibition of vacuoles formation by chloroquine,
and increasing effect on LC3 protein level, observed only in treatments
with complex 2. In previous work, we have already attested the re-
activity of mononuclear complex 1 and dinuclear complex 2 toward
DNA by oxidative mechanism, causing single and double strand clea-
vages [23]. These new data seem to point to a balance interplay be-
tween different modes of cell death, apoptosis and autophagy, when
melanomas were treated with this complex 2, what is not observed with
the analogous complex 1.
These results also indicated a prominent influence of copper (II)
complex structure on defining its reactivity. Mononuclear imine-copper
(II) compounds are usually able of efficient mono-electronic steps when
acting as antitumor agents [7,8], damaging oxidatively fundamental
targets as DNA or mitochondria, via oxidative mechanism. On the
contrary, analogous dinuclear complexes could work by di-electronic
reactions, similarly to tyrosinase enzymes that are also implicated in the
formation of melanin, giving rise to alternative or additional modes of
action. In this case, irradiation by UVB light by increasing melano-
genesis improved the observed damage. Therefore, dinuclear copper
complexes could constitute a potential therapeutic approach against
melanoma cancer, devising alternative targets and diversified me-
chanisms.
Conflicts of interest
The authors declare no competing financial interest.
Declaration of interests
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgements
Financial support for this work was provided by the Brazilian
agencies: São Paulo State Research Foundation (FAPESP, grant 2011/
50318-1), CNPq (grant 304776/2014-9), NAP-Redoxoma (Pro-Reitoria
de Pesquisa-USP, Proc. 11.1.9352.1.8), and Network CEPID-Redoxoma
(grant 2013/07937-8). C.J.N. is grateful to Coordenação de
Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES, Finance
Code 001) for fellowships during her Ph.D. studies. The authors also
thanks Professor Marius Réglier (Université Aix-Marseille, Marseille,
10
Chemico-Biological Interactions 311 (2019) 108789
France) for providing the bridge ligand dpam as a gift.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.cbi.2019.108789.
References
[1] N.P.E. Barry, P.J. Sadler, Exploration of the medical periodic table: towards new
targets, Chem. Commun. 49 (2013) 5106–5131.
[2] E. Alessio, Thirty years of the drug candidate NAMI-A and the myths in the field of
ruthenium anticancer compounds: a personal perspective, Eur. J. Inorg. Chem.
(2017) 1549–1560.
[3] J.-X. Liang, H.-J. Zhong, G. Yang, K. Vellaisamy, D.-L. Ma, C.-H. Leung, Recent
development of transition metal complexes with in vivo antitumor activity, J. Inorg.
Biochem. 177 (2017) 276–286.
[4] C.P. Tan, Y.Y. Lu, L.N. Ji, Z.W. Mao, Metallomics insights into the programmed cell
death induced by metal-based anticancer compounds, Metall 6 (2014) 978–995.
[5] G. Housman, S. Byler, S. Heerboth, K. Lapinska, M. Longacre, N. Snyder, S. Sarkar,
Drug resistance in cancer: an overview, Cancers 6 (2014) 1769–1792.
[6] M. Wehbe, A.W.Y. Leung, M.J. Abrams, C. Orvig, M.B. Bally, A Perspective – can
copper complexes be developed as a novel class of therapeutics? Dalton Trans. 46
(2017) 10758–10773.
[7] C. Santini, M. Pellei, V. Gandin, M. Porchia, F. Tisato, C. Marzano, Advances in
copper complexes as anticancer agents, Chem. Rev. 114 (2014) 815 862.
[8] G. Filomeni, G. Cerchiaro, A.M. Da Costa Ferreira, A. De Martino, J.Z. Pedersen,
G. Rotilio, M.R. Ciriolo, Pro-apoptotic activity of novel isatin-Schiff base copper(II)
complexes depends on oxidative stress induction and organelles selective damage,
J. Biol. Chem. 282 (2007) 12010–12021.
[9] P.P. Silva, W. Guerra, J.N. Silveira, A.M. da Costa Ferreira, T. Bortolotto,
F.L. Fischer, H. Terenzi, A. Neves, E.C. Pereira-Maia, Two new ternary complexes of
copper(II) with tetracycline or doxycycline and 1,10-phenanthroline and their po-
tential as antitumoral: cytotoxicity and DNA cleavage, Inorg. Chem. 50 (2011)
6414–6424.
[10] N. Graf, S.J. Lippard, Redox activation of metal-based prodrugs as a strategy for
drug delivery, Adv. Drug Deliv. Rev. 64 (2012) 993–1004.
[11] V. Gandim, F. Tisato, A. Dolmella, M. Pellei, C. Santini, M. Giorgetti, C. Marzano,
M. Porchia, In vitro and in vivo anticancer activity of copper(I) complexes with
homoscorpionate tridentate tris(pyrazolyl)borate and auxiliary monodentate
phosphine ligands, J. Med. Chem. 57 (2014) 4745–4760.
[12] T. Wang, Z. Guo, Copper in medicine: homeostasis, chelation therapy and antitumor
drug design, Curr. Med. Chem. 13 (2006) 525–537.
[13] J. Qi, S. Liang, Y. Gou, Z. Zhang, Z. Zhou, F. Yang, H. Liang, Synthesis of four
binuclear copper(II) complexes: structure, anticancer properties and anticancer
mechanism, Eur. J. Med. Chem. 96 (2015) 360–368.
[14] D. Desbouis, I.P. Troitsky, M.J. Belousoff, L. Spiccia, B. Graham, Copper(II), zinc(II)
and nickel(II) complexes as nuclease mimetics, Coord. Chem. Rev. 256 (2012)
897–937.
[15] D. Montagner, V. Gandin, C. Marzano, A. Erxleben, DNA damage and induction of
apoptosis in pancreatic cancer cells by a new dinuclear bis(triazacyclonane) copper
complex, J. Inorg. Biochem. 145 (2015) 101–107.
[16] A.G. Majouga, M.I. Zvereva, M.P. Rubtsova, D.A. Skvortsov, A.V. Mironov,
D.M. Azhibek, O.O. Krasnovskaya, V.M. Gerasimov, A.V. Udina, N.I. Vorozhtsov,
E.K. Beloglazkina, L. Agron, L.V. Mikhina, A.V. Tretyakova, N.V. Zyk, N.S. Zefirov,
A.V. Kabanov, O.A. Dontsova, Mixed valence copper(I,II) binuclear complexes with
unexpected structure: synthesis, biological properties and anticancer activity, J.
Med. Chem. 57 (2014) 6252–6258.
[17] T. Jany, A. Moreth, C. Gruschka, A. Sischka, A. Spiering, M. Dieding, Y. Wang,
S.H. Samo, A. Stammler, H. Bögge, G.F. von Mollard, D. Anselmetti, T. Glaser,
Rational design of a cytotoxic dinuclear Cu2 complex that binds by molecular re-
cognition at two neighboring phosphates of the DNA backbone, Inorg. Chem. 54
(2015) 2679–2690.
[18] R.J. Sullivan, M.B. Atkins, Molecular-targeted therapy in malignant melanoma,
Expert Rev. Anticancer Ther. 9 (2009) 567–581.
[19] A.N. Houghton, D. Polsky, Focus on melanoma, Cancer Cell 2 (2002) 275–278.
[20] M. Rastrelli, S. Tropea, C.R. Rossi, M. Alaibac, Melanoma: epidemiology, risk fac-
tors, pathogenesis, diagnosis and classification, In Vivo 28 (2014) 1005–1012.
[21] E. Buitrago, R. Hardré, R. Haudecoeur, H. Jamet, C. Belle, A. Boumendjel,
L. Bubacco, M. Réglier, Are human tyrosinase and related proteins suitable targets
for melanoma therapy? Curr. Top. Med. Chem. 16 (2016) 3033–3047.
[22] S. Jawaid, T.H. Khan, H.M.I. Osborn, N.A.O. Williams, Tyrosinase activated mela-
noma prodrugs. Anti-cancer Ag, Med. Chem. 9 (2009) 717–727.
[23] C.J. Nunes, B.E. Borges, L.S. Nakao, E. Peyroux, R. Hardré, B. Faure, M. Réglier,
M. Giorgi, M.B. Prieto, C.C. Oliveira, A.M. Da Costa Ferreira, Reactivity of dinuclear
copper(II) complexes towards melanoma cells: correlation with its stability, tyr-
osinase mimicking and nuclease activity, J. Inorg. Biochem. 149 (2015) 49–58.
[24] L. Benmenni, E.H. Alilou, M. Giorgi, M. Pierrot, M. Réglier, Crystal structures and
energy refinement of some 2,2'-disubstituted biphenyl compounds, J. Chem.
Crystallogr. 24 (1994) 345–352.
[25] M.S. Eller, K. Ostrom, B.A. Gilchrest, DNA damage enhances melanogenesis, Proc.
Natl. Acad. Sci. U.S.A. 93 (1996) 1087–1092.
[26] D.-N. Hu, Methodology for evaluation of melanin content and production of
C.J. Nunes, et al.
pigment cells in vitro, Photochem. Photobiol. 84 (2008) 645–649.
[27] N. Khan, C.M. Wilmot, G.M. Rosen, E. Demidenko, J. Sun, J. Joseph, J. O’hara,
B. Kalyanaraman, H.M. Swartz, Spin Traps: in vitro toxicity and stability of radical
adducts, Free Radic. Biol. Med. 34 (2003) 1473–1481.
[28] S. Rohn, L.W. Kroh, Electron spin resonance–a spectroscopic method for de-
termining the antioxidative activity, Mol. Nutr. Food Res. 49 (2005) 898–907.
[29] B.M. Polster, D.G. Nicholls, S.X. Ge, B.A. Relofs, Use of potentiometric fluorophores
in the measurement of mitochondrial reactive oxygen species, Methods Enzymol.
547 (2014) 225–250.
[30] N.A.P. Franken, H.M. Rodermond, J. Stap, J. Haveman, C. van Bree, Clonogenic
assay of cells in vitro, Nat. Protoc. 1 (2006) 2315–2319.
[31] G. Triola, Autophagy: assays and small-molecule modulators, in: H. Waldmann,
P. Janning (Eds.), Concepts and Case Studies in Chemical Biology, first ed., Wiley-
VCH Verlag, Weinheim, Germany, 2014, pp. 63–82.
[32] V.C. Silveira, M.P. Abbott, M. Cavicchioli, M.B. Gonçalves, H.M. Petrilli, L. de
Rezende, A.T. Amaral, D.E.P. Fonseca, G.F. Caramori, A.M. da Costa Ferreira,
Peculiar reactivity of a di-imine copper(II) complex regarding its binding to al-
bumin protein, Dalton Trans. 42 (2013) 6386–6396.
[33] M.A.A. Azzellini, I.A. Bagatin, A.M. Da Costa Ferreira, Di-imine copper(II) com-
plexes as redox mediator and modulator in 2-deoxy-D-ribose oxidative damage,
Redox Rep. 11 (2006) 25–37.
[34] W.A. Alves, S.A. de Almeida Filho, R.H.A. Santos, A.M. Da Costa Ferreira, Diimine
copper(II) complexes as building blocks for microporous catalytic materials, Inorg.
Chem. Commun. 6 (2003) 294–299.
[35] A.A. Suzukawa, A. Vieira, S.M.B. Winnischofer, A.C. Scalfo, P. Di Mascio, A.M. da
Costa Ferreira, J.-L. Ravanat, D. de L. Martins, M.E.M. Rocha, G.R. Martinez, Novel
properties of melanins include promotion of DNA strand breaks, impairment of
repair, and reduced ability to damage DNA after quenching of singlet oxygen, Free
Radic. Biol. Med. 52 (2012) 1945–1953.
11
Chemico-Biological Interactions 311 (2019) 108789
[36] F. Radogna, M. Dicato, M. Diederich, Cancer-type-specific crosstalk between au-
tophagy, necroptosis and apoptosis as a pharmacological target, Biochem.
Pharmacol. 94 (2015) 1–11.
[37] L.A. Booth, S. Tavallai, H.A. Hamed, N. Cruickshanks, P. Dent, The role of cell
signalling in the crosstalk between autophagy and apoptosis, Cell. Signal. 26 (2014)
549–555.
[38] V. Nikoletopoulou, M. Markaki, K. Palikaras, N. Tavernarakis, Crosstalk between
apoptosis, necrosis and autophagy, Biochim. Biophys. Acta 1833 (2013)
3448–3459.
[39] M. Corazzari, G.M. Fimia, P. Lovat, M. Piacentini, Why is autophagy important for
melanoma? Molecular mechanisms and therapeutic implications, Semin. Cancer
Biol. 23 (2013) 337–343.
[40] W.J. Yun, E.-Y. Kim, J.-E. Park, S.Y. Jo, S.H. Bang, E.-J. Chang, S.E. Chang,
Microtubule-associated protein light chain 3 is involved in melanogenesis via reg-
ulation of MITF expression in melanocytes, Sci. Rep. 6 (2016) 19914, https://doi.
org/10.1038/srep19914.
[41] K. Konariková, G.A. Perdikaris, H. Gbelcová, L. Andrezálová, M. Švéda, T. Ruml,
L. Laubertová, S. Režnáková, I. Žitnanová, Autophagy in MCF-7 cancer cells in-
duced by copper complexes, Pharmacol. Rep. 68 (2016) 1221–1224.
[42] M. Redmann, G.A. Benavides, T.F. Berryhill, W.Y. Wani, X. Ouyang, M.S. Johnson,
S. Ravi, S. Barnes, V.M. Darley-Usmar, J. Zhang, Inhibition of autophagy with ba-
filomycin and chloroquine decreases mitochondrial quality and bioenergetic func-
tion in primary neurons, Redox Biol 11 (2017) 73–81.
[43] I. Tanida, T. Ueno, E. Kominami, LC3 and autophagy, Autophagosome And
Phagosome. Methods in Molecular Biology, vol. 445, Humana Press, Springer,
2008, pp. 77–88.
[44] E. Shvets, Z. Elazar, Flow cytometric analysis of autophagy in living mammalian
cells, Methods Enzymol. 452 (2009) 131–141.