Exploring the DNA binding/cleavage, cellular accumulation and topoisomerase inhibition of 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinone Mannich bases and their platinum(II) complexes

Article abstract of DOI:10.1016/j.jinorgbio.2012.10.007

Several chlorido and amino Pt^{2 +} complexes of 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinone Mannich bases HL exhibiting moderate to high cytotoxicity against cancer cell lines were studied in order to investigate their modes of DNA binding, in vitro DNA strand breaks, mechanism of topoisomerase (Topo I) inhibition and cellular accumulation. DNA model base studies have shown that complex 1a [Pt(HL1)Cl₂] was capable of binding covalently to 9-ethylguanine (9-EtG) and 5′-GMP. ¹H NMR and mass spectrometry studies have shown that both chlorides were substituted by 9-EtG ligands, whereas 5′-GMP was able to replace only one chlorido ligand, due to steric hindrance. The chlorido Pt^{2 +} complexes [Pt(HL)Cl₂] highly accumulate in prostate (PC-3) and melanoma (MDA-MB-435) cell lines, being able to induce DNA strand breaks in vitro and inhibit Topo I by a catalytic mode. On the other hand, the free 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinones HL and the amino Pt^{2 +} complexes [Pt(L^-)(NH₃)₂]NO₃ neither cause DNA strand breakage nor exhibit strong DNA interaction, nevertheless the latter were also found to be catalytic inhibitors of Topo I at 100 μM. Thus, coordination of the Mannich bases HL to the PtCl₂ fragment substantially affects the chemical and biophysical properties of the pro-ligands, leading to an improvement of their DNA binding properties and generating compounds that cleave DNA and catalytically inhibit Topo I. Finally, the high cytotoxicity exhibited by the free (uncomplexed) 2-hydroxy-3- (aminomethyl)-1,4-naphthoquinones might be associated with their decomposition in solution, which is not observed for the Pt^{2 +} complexes.

Full text of DOI:10.1016/j.jinorgbio.2012.10.007

Journal of Inorganic Biochemistry 119 (2013) 54–64
Contents lists available at SciVerse ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Exploring the DNA binding/cleavage, cellular accumulation and topoisomerase
inhibition of 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinone Mannich bases
and their platinum(II) complexes
a,1
b
a
a
c
Amanda P. Neves , Michelle X.G. Pereira , Erica J. Peterson , Ralph Kipping , Maria D. Vargas ,
b
c
a,
Floriano P. Silva-Jr , J. Walkimar M. Carneiro , Nicholas P. Farrell ⁎
a
Department of Chemistry, Virginia Commonwealth University, 1001 West Main Street, Richmond, VA 23284, USA
Laboratório de Bioquímica de Proteínas e Peptídeos, Fundação Oswaldo Cruz, Instituto Oswaldo Cruz, Av. Brasil, 4365, Manguinhos, Rio de Janeiro, RJ, CEP: 21040-360, Brazil
Instituto de Química, Universidade Federal Fluminense, Campus do Valonguinho, Centro, Niterói, RJ, CEP: 24020-141, Brazil
b
c
a r t i c l e i n f o
a b s t r a c t
Several chlorido and amino Pt2+ complexes of 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinone Mannich
bases HL exhibiting moderate to high cytotoxicity against cancer cell lines were studied in order to investigate
their modes of DNA binding, in vitro DNA strand breaks, mechanism of topoisomerase (Topo I) inhibition and cel-
Article history:
Received 12 June 2012
Received in revised form 8 October 2012
Accepted 8 October 2012
Available online 15 October 2012
2
lular accumulation. DNA model base studies have shown that complex 1a [Pt(HL1)Cl ] was capable of binding
covalently to 9-ethylguanine (9-EtG) and 5′-GMP. 1H NMR and mass spectrometry studies have shown that
both chlorides were substituted by 9-EtG ligands, whereas 5′-GMP was able to replace only one chlorido ligand,
Keywords:
due to steric hindrance. The chlorido Pt2+ complexes [Pt(HL)Cl
] highly accumulate in prostate (PC-3) and mel-
anoma (MDA-MB-435) cell lines, being able to induce DNA strand breaks in vitro and inhibit Topo I by a catalytic
Naphthoquinone
2
2
+
Pt
complexes
2
+
DNA binding
DNA strand breaks
mode. On the other hand, the free 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinones HL and the amino Pt
complexes [Pt(L−)(NH ) ]NO neither cause DNA strand breakage nor exhibit strong DNA interaction, neverthe-
3
2
3
less the latter were also found to be catalytic inhibitors of Topo I at 100 μM. Thus, coordination of the Mannich
bases HL to the “PtCl ” fragment substantially affects the chemical and biophysical properties of the
2
pro-ligands, leading to an improvement of their DNA binding properties and generating compounds that cleave
DNA and catalytically inhibit Topo I. Finally, the high cytotoxicity exhibited by the free (uncomplexed)
2
-hydroxy-3-(aminomethyl)-1,4-naphthoquinones might be associated with their decomposition in solution,
2
+
which is not observed for the Pt complexes.
©
2012 Elsevier Inc. All rights reserved.
1
. Introduction
naphthoquinones and anthraquinones, their DNA interactions (inter-
calation or alkylation) might also contribute to their cytotoxicity.
Moreover, naphthoquinones have been shown to act as topoisomer-
ase inhibitors, an activity that may also be involved in their cytotoxic
mechanism [10–12].
Natural and synthetic quinone derivatives have been widely in-
vestigated for cancer therapy [1–3]. Many drugs containing a quinone
moiety, such as Mitomycin C and Doxorubicin have received clinical
approval for cancer treatment [1,4,5]. In general, the biological activ-
ity of quinones derives from the ability of the quinone moiety to
accept one or two electrons to form semiquinone and cathecol radi-
cals. These species are re-oxidized by molecular oxygen (O ) generat-
2
ing highly reactive oxygen species (ROS), such as superoxide anion
radical, which play a key role in mutagenesis and carcinogenesis
We have been working extensively on the design of novel
naphthoquinone derivatives [11–15], including Mannich bases and
their Cu2 complexes [16,17], which present interesting bactericidal
activity [17]. The Mannich bases derived from 2-hydroxy-1,4-
naphthoquinone (lawsone) were first synthesized in the 1940s and
their antimalarial and molluscicidal activities have been investigated
[18–21]. Recently, we have reported the synthesis and cytotoxicity
of unprecedented Pt2 complexes of 2-hydroxy-3-(aminomethyl)-1,4-
naphthoquinone Mannich bases (Scheme 1), with high to moderate ac-
tivity against several tumor cell lines [22] (SI (Supplementary Informa-
+
[6,7]. The production of ROS has been commonly related to DNA
+
strand break production and to the oxidative stress observed in
cells [8,9]. Especially for polycyclic quinonoid compounds, such as
tion), Table S1). The search for novel Pt-based drugs [23–25] has been a
⁎
Corresponding author. Tel.: +1 804 828 6320; fax: +1 804 828 8599.
E-mail address: npfarrell@vcu.edu (N.P. Farrell).
Present address: Departamento de Química, Universidade Federal Rural do Rio de
2+
priority since the discovery of cisplatin [26]. Most Pt
complexes act
2
+
1
similarly to cisplatin, through the binding of the [Pt(amine)
2
]
moiety
Janeiro, BR-465, Km 7 Seropédica, RJ, CEP: 23890-000, Brazil.
to the DNA, forming covalent adducts that inhibit replication and
0162-0134/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jinorgbio.2012.10.007

A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
55
5
0 mM NaCl, pH 7.4 and 2 mM phosphate, pHs 7.0 and 5.8) and TAE
(
40 mM Tris, 20 mM acetic acid, 1 mM EDTA, pH 8.0).
1
H NMR spectra were recorded on a Varian Mercury 300 MHz
7 7
NMR spectrometer in DMF-d (N,N-dimethylformamide-d ) and
mixture of D O/DMF-d (3:1). Chemical shifts were referenced
2
7
to 3-trimethylsilylpropionic acid, sodium salt, used as internal stan-
dard. 1 Pt NMR spectra were recorded on a Varian Mercury NMR
95
195
spectrometer using a 10 mm broadband probe.
referenced to Na [PtCl
2 6
6
Pt spectra were
]. The frequency for 1 Pt nuclei was set at
4.32 MHz. Mass spectra were obtained in a Micromass QTof-2 in-
95
strument. Samples, whose final concentrations were 1 μg/μL, were
dissolved in 1:1 methanol:water and injected using direct infusion
method with the flow rate of 0.5μL/min.
2.2. Synthesis
The 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinones HL1–HL3
2
+
derivatives, chlorido 1a–3a and amino 1b–3b Pt
complexes were
synthesized as described in the literature [22]. The DMF solutions
(
(
(
(
5 mM) of compounds [Pt(HL1-3)(DMF)
2
](NO
3
)
2
and [Pt(NH
3 2
)
Scheme 1. Chlorido 1a–3a, aqua 1w–3w and amino 1b–3b Pt2+ complexes of 2-hydroxy-
-(aminomethyl)-1,4-naphthoquinones HL1–HL3 studied in this work.
DMF) ](NO , were obtained by reacting a 100 mM AgNO
2
)
3 2
3
solution
3
0.4 mL, 0.04 mmol) with 1a–3a (0.02 mmol) and with cisplatin
6 mg, 0.02 mmol), respectively, in DMF (3.6 mL) at 50 °C for 24 h.
The resulting AgCl precipitate was filtered off. [Pt(HL1-3)(H
(NO 1w–3w and [Pt(NH (H O) ](NO cw were obtained by di-
lution of the DMF solutions with water to a final concentration of 95:5
(water:DMF). The water solution (5 mM) of [Pt(HL1)(H O) ](NO
1w was also prepared using the same procedure above, except for
the use of H O instead of DMF, for the gel mobility shift assays.
2 2
O) ]
transcription, causing cellular apoptosis and necrosis [27]. However,
3
)
2
)
3 2
2
2
3 2
)
2
+
other Pt
derivatives such as the trinuclear BBR3464 and TriplatinNC
show different DNA binding modes with long-range interstrand
crosslinks and non-covalent interactions – electrostatic and hydrogen
bonding – respectively [28]. Lately, the attachment of biologically rele-
2
2
3 2
)
2
2
+
vant molecules to Pt , e.g. intercalators [29], tamoxifen [30] and porphy-
rins [31] have been successfully explored as an alternative to combine
two different active fragments in one molecule. In the present work, we
describe a series of biophysical and cellular studies on the previously
published [22] 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinone Pt2+
complexes (1a–3a, 1b–3b) and the respective pro-ligands HL1–HL3
2.3. Molecular modeling
Density functional calculations were carried out using the
Gaussian03W molecular orbital package [34]. Geometries were fully
optimized using the B3LYP functional [35] with the LANL2DZ basis
set [36]. The initial geometry was taken from the crystallographic
structure of 1a [22] after replacement of one or two chlorine ions by
the 9-EtG base.
(
Scheme 1). We have also investigated the interactions between the
cis-[Pt(HL1)Cl ] 1a complex and the DNA model bases 9-ethylguanine
9-EtG) and 5′-GMP. We show that the incorporation of Pt
2
2
+
(
into 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinones causes signifi-
cant changes in the biophysical properties of the pro-ligands, strongly af-
fecting their cytotoxicity and mechanism of action.
2.4. DNA model base studies
2
.4.1. Reaction of cis-[Pt(HL1)Cl
2
] 1a with 9-ethylguanine (9-EtG)
For the H NMR, cis-[Pt(HL1)Cl ] (5 mg, 0.0083 mmol) was
allowed to react with 9-EtG (3 mg, 0.0166 mmol) in an NMR tube
using 0.7 mL of DMF-d at room temperature. A 6-fold excess of 1a
1
2
2
. Experimental section
7
195
1
195
2
.1. Materials and methods
and 9-EtG was used for
Pt NMR in 2 mL of DMF-d
7
.
H and
Pt
NMR spectra were recorded over time. After 72 h, an aliquot of the
solution from the NMR tube was evaporated and re-suspended in a
Reagents and solvents were used as purchased without further
purification. Calf thymus DNA (CT-DNA) and ethidium bromide
EtdBr) were obtained from Sigma-Aldrich. 5′-GMP (Aldrich) was
mixture of MeOH/H
analysis.
2
O (1:1) in order to perform a mass spectrometry
(
employed as a disodium salt. Supercoiled plasmid DNA (pUC19)
was isolated from E. coli strain XLI-blue with the use of Qiagen DNA
purification kits. All stock solutions of the investigated compounds
were prepared by dissolving the powdered materials in appropriate
amounts of DMF or DMSO. Deionized double distilled water and ana-
lytical grade reagents were used throughout. CT-DNA stock solution
2.4.2. Reaction of cis-[Pt(HL1)Cl
The reaction of cis-[Pt(HL1)Cl
(8.2 mg, 0.02 mmol) was carried out in the NMR tube in a mixture
of D O/DMF-d (3:1, total 1 mL) at room temperature. A 4-fold excess
of 1a and 5′-GMP was used for
Pt NMR were recorded over time and mass spec-
trometry was performed after analysis of an aliquot from the NMR
tube after 72 h of reaction.
2
] 1a with 5′-GMP
2
] (6 mg, 0.01 mmol) and 5′-GMP
2
7
195
2 7
Pt NMR in D O/DMF-d (3:1, total
2 mL). 1H and
195
was prepared by dissolving the solid material in NaClO
0 mM. Then, the solution was dialyzed for 24 h at 4 °C. The resulting
somewhat viscous solution was clear and particle-free. The CT-DNA
4
buffer
1
–1
concentration in terms of nucleotides L was determined spectropho-
–1
–1
tometrically by employing an extinction coefficient of 6600 M cm
2.5. Measurement of melting curves
–
1
(
nucleotide) at 260 nm [32]. The stock solution was stored at 5 °C
until it was used. EtdBr was dissolved in 10 mM of phosphate buffer
pH 7.4) and its concentration was determined assuming a molar ex-
100 μM nucleotides of CT DNA were incubated with compounds
(
HL1, 1w, 1b and cw at various binding ratios (r
tides]=0, 0.01, 0.03, 0.05, 0.075 and 0.1) in 0.01 M NaClO
DMF for 72 h in the dark. The data were recorded on a JASCO V-550
i
=[Drug]/[nucleo-
–1
–1
tinction coefficient of 5600 L mol cm at 480 nm [33]. Buffers used
4
with 5%
and their abbreviations are: phosphate buffers (10 mM phosphate,

56
A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
UV-visible (UV–vis) spectrophotometer by measuring the absorbance
at 260 nm. Temperature was varied from 40 to 98.5 °C increasing at
the rate of 0.5 °C per min. Melting points were estimated as the tem-
perature when 1/2 of absorbance increase was reached.
K at 1 mg/mL followed by another incubation at 37 °C for 30 min.
Then, 2 μL of 10X gel loading dye were added to DNA samples, followed
by DNA extraction with 24:1 chloroform isolamyl alcohol (CIA). The sam-
ples were immediately vortexed, centrifuged and the blue aqueous phase
was analyzed by electrophoresis on a 1% agarose gel without EtdBr at
30 V for 17 h at room temperature in TAE buffer followed by staining
in 0.5 μg/mL of EtdBr to allow viewing of the samples under UV light.
2
.6. Fluorescence measurements
Measurements were performed on a Varian Cary Eclipse Fluorim-
eter in wellplates of 96 wells. Fluorescence measurements of DNA
modified by the compounds in the presence of EtdBr were performed
at an excitation wavelength of 590 nm, and the emitted fluorescence
was analyzed at 600 nm. Compounds HL1–HL3, 1w–3w, 1b–3b and
cw were first incubated with CT-DNA (r =0.1) for 72 h in water/
i
DMF (95:5) at 37 °C. The EtdBr solution was then added to the
drug/DNA solutions at 25 °C at various ratios of EtdBr/nucleotides
2.8.2. DNA intercalation (unwinding) assay
The unwinding effects of compounds were measured using a DNA
Unwinding Kit from TopoGEN, Inc. (Columbus, OH) with some adap-
tations. This kit includes the relaxation buffer (described above), con-
trol intercalator (m-AMSA; lyophilized), 10% SDS and 10X gel loading
dye. To the relaxed plasmid (prepared as described in the section
above) or to the supercoiled plasmid (control of the unwinding reac-
tion) were added the different test compounds, control compounds
or blank solution to a final concentration of 0.1 mM (final volume=
20 μL). Samples were incubated for 20 min at room temperature
and then, 10 times more enzyme (2 units) were added to the reaction
which was incubated for another 30 min at 37 °C. The reaction was
terminated by addition of SDS and proteinase K, followed by DNA ex-
traction with CIA. Sample analysis was carried out by electrophoresis
on a 1% agarose gel without EtdBr, as described in the section above.
After staining with 0.5 μg/mL of EtdBr, the gel was run one more time
at 100 V for 1 h at room temperature in TAE buffer. This last electro-
phoresis step makes it possible to see the open circle DNA band.
(
r=0, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.08, 0.10, 0.15, 0.20 and
0
.25). The final pH of the solutions was adjusted to 7.4 with phos-
3
−
phate buffer whose final concentration of PO
4
and NaCl were, re-
spectively, 10 and 50 mM.
2
.7. Gel mobility shift assays
A
gel mobility shift assay [37] was employed incubating
supercoiled plasmid DNA (pUC19; 1 μg/15μL) with the compounds
overnight at 37 °C in 2 mM phosphate buffer at pHs 7.0 and 5.8.
The solutions were loaded onto 1% agarose gels in TAE buffer at
1
00 V for 1.5 h. The gels were then stained with EtdBr, followed by
photography on Polaroid 667 film with UV transillumination. In
order to determine the r values ([Pt]/[nucleotides]) for 1w, 2 μg of
b
DNA was incubated with the drug in 30 μL of 2 mM phosphate buffer
2.9. Cellular platinum accumulation
at 37 °C overnight. The DNA was precipitated and washed with EtOH
7
0% to remove the unreacted drug, and then resuspended in the buff-
er. An aliquot of the solution containing 1 μg of DNA was loaded onto
% agarose gels in TAE buffer at 100 V for 1.5 h. The amount of Pt
present in the other aliquots was determined by a Varian ICP-MS.
Cellular accumulation of 1a–3a, 1b–3b and cisplatin was mea-
6
sured in MDA-MB-435 and PC-3 cells. 1×10 cells were seeded in
2
1
60 mm tissue culture dishes (30 000 per cm ). After overnight incu-
bation at 37 °C, the cells were treated with the platinum compounds
for 2, 4 and 6 h at the final concentration of 20 μM at 0.5% DMSO. The
cell monolayers at the end of the incubation with the Pt complexes
were washed twice with 1X PBS (phosphate buffered saline),
trypsinized and harvested into appropriate media. Cell suspensions
were centrifuged, the pellets were digested with 12 M HCl and the
platinum content was analyzed by ICP-MS. All experiments were
made in triplicate.
2
2
.8. Topoisomerase I inhibition studies
.8.1. DNA relaxation assay
Human Topoisomerase I (Topo I) was obtained from TopoGEN,
Inc. (Columbus, OH); proteinase K from Tritirachium album was
obtained from Sigma-Aldrich, Inc. (St. Louis, MO) and was dissolved
in DNase free water. Gel electrophoresis buffer (TAE) was used in
this assay.
The inhibitory effects of the compounds on human Topo I were mea-
sured using an Eukaryotic Topoisomerase I Drug Screening Kit (TopoGEN,
Inc.). This kit includes the supercoiled plasmid DNA (pBR322, 250 ng),
the 10X gel loading dye, the relaxation buffer (10 mM Tris buffer
pH 7.9, 1 mM EDTA, 0.15 M NaCl, 0.1% BSA, 0.1 mM spermidine and 5%
glycerol), the control inhibitor (Camptothecin; lyophilized) and 10%
SDS. All compounds were dissolved in DMSO immediately prior to test-
ing. For the initial screening, the substances were tested at a fixed concen-
tration of 0.1 mM. This assay concentration was chosen based on the
effective concentration of the camptothecin standard (0.1 mM), as
recommended by the kit manufacturer (TopoGen). Active compounds
were tested at varying concentrations, provided that the dilution in the
assay was at least 20 times (5% of DMSO). To ensure that the DMSO
does not interfere in the experiment, different concentrations of DMSO
were tested, showing no interference in the enzyme function in a
DMSO concentration range of 0–5%. Supercoiled plasmid DNA (pBR322,
2.10. Comet assay
Single-cell gel electrophoresis was performed using a commercially
available kit (Trevigen — Gaithersburg, MD). Compounds HL2, 2a, 2b
and cisplatin were incubated with 1×105 melanoma cells (MDA-
MB-435) at 100 μM (1% DMSO). After 4 h, cells were harvested and em-
bedded in low-melting-point agarose at a volume of 1:10 on micro-
scope slides. After solidification in the refrigerator for 30 min, the cells
were lysed in the dark for 30 min at 4 °C. The slides were then incubat-
ed in an alkaline electrophoresis solution (NaOH, pH>13) in the dark
for 20 min at room temperature, which were then washed twice with
1X TBE buffer (pH 7.0). Electrophoresis was performed for 10 min at
25 V in 1X TBE buffer (pH 7.0). After electrophoresis, the slides were
immersed in 70% ethanol for 5 min, and air dried at room temperature.
The slides were then stained with SYBR green dye observed with a fluo-
rescence microscope (Nikon ECLIPSE E600). A total of 70 randomly cho-
sen cells were scored per sample using the CometScore freeware
obtained from TriTek Corporation [38]. Calculations were based on
two parameters: Tail moment (Tail length×% DNA in tail/100) and
Olive moment [(Tail mean−Head mean)×% DNA in tail/100] which
indicate separate parameters of DNA damage [39]. The analyses were
repeated in triplicate.
250 ng) was incubated with human Topo I (an amount sufficient to
achieve full relaxation at 37 °C for 30 min in relaxation buffer) in the ab-
sence or presence of compounds (final volume is 20 μL). Camptothecin
(0.1 mM) was used as positive control. The order of reagent addition
was buffer, test compound, Topo I and finally DNA. The reaction was
stopped by addition of 2 μL of 10% SDS solution and 1 μL of proteinase

A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
57
Fig. 1. Numbering scheme of the atoms for 1a, 9-EtG and 5′-GMP.
3
. Results and discussion
9-EtG carbonyl group. Isomer cis-4 (SI, Fig. S1) is less stable than isomer
−
1
trans-4 in the gas phase by about 23.9 kJ mol . The most stable of the
different possible geometrical conformations for the di-substituted
compound 5 (SI, Fig. S2), which were generated from the structures of
trans-4 and cis-4, is the one that exhibits a hydrogen bond as in
trans-4 and the carbonyl groups of the two 9-EtG bases oriented in op-
posite directions (head-to tail-conformer) [42].
3
.1. Interaction of chlorido complex 1a with 9-EtG and 5′-GMP: DNA
model base studies
Based on the fact that DNA is considered the main target of
Pt-based drugs [27], the binding of platinum complexes with model
DNA nucleobases, such as 9-ethylguanine and with 5′-GMP has
been commonly investigated [40] (Fig. 1). Platinum complexes of
3.1.1. Interaction with 9-EtG
the type [PtA
donor carrier ligand and G
for Pt-DNA interactions [41].
The possible reactions between 1a and the DNA nucleobases are
summarized in Scheme 2. Because the reactions have been carried
out with a two fold excess of the amine (9-EtG or 5′-GMP), and con-
sidering the acidic character of the naphthoquinone hydroxyl group,
2
G
2
] (A
2
=2 monodentate or a bidentate nitrogen
Complex 1a was allowed to react with 9-EtG in DMF-d , at 25 °C at
the 2:1 molar ratio of 9-EtG:1a. The H NMR spectra were recorded
7
1
2
=two guanines) are the simplest models
over time (Figs. 2 and 3).
The 1H NMR spectrum of the mixture 15 min. after mixing
showed the characteristic H8 and CH peaks of unreacted 9-EtG, at δ
2
8.09 and 4.11 ppm, respectively (Fig. 2). Careful monitoring of these
peaks and those of the coordinated aminomethylnaphthoquinone at
δ 9.41 (H14) and 5.82 (H11) allowed the observation of different
platinum-9-EtG species formed in the reaction.
[17–19] these species may be in equilibrium with their respective
deprotonated forms [22].
These reactions can give rise to two different mono-adducts, due
to the asymmetric nature of the complex: the DNA nucleobase (Nu)
can substitute the chloride ligand trans or cis to the pyridyl group,
forming [Pt(HL)(Nu)Cl]Cl (Nu=9-EtG trans-4 or cis-4 and Nu=
After 24 h, the appearance of a small quartet in the CH region, at δ
4.27 (Fig. 2), and of a small doublet at δ 5.95 (Fig. 3) at slightly higher
2
frequency to the H11 doublet of the starting material 1a evidenced
formation of a mono-adduct, possibly trans-4, with the 9-EtG trans
to the pyridyl group, because of the presence of a small doublet at δ
9.30 in the pyridyl group H14 region (SI, Fig. S3a). After 72 h, this ad-
duct remained the most abundant product in solution. Two additional
peaks were observed in the H11 region, one partially hidden under
the doublet resonance of trans-4, which was attributed to the
bis-adduct 5, and the other, at δ 5.86, attributed to the second
mono-adduct cis-4 (Fig. 3). Similar pattern was observed in the δ
9.70–9.20 region, where the corresponding peaks due to the H14 of
the starting material (δ 9.40 ppm), mono-adducts cis-4, trans-4 and
bis-adduct 5 (SI, Fig. S3a) were also observed. The H8 signals of the
coordinated 9-EtG in the reaction products were found between
8.80 and 8.60 ppm (SI, Fig. S3b). Small changes were also observed
5′-GMP trans-6 or cis-6). In addition, the di-substituted products,
[
[
[
Pt(HL1)(Nu)
Pt(L1 )(Nu)
2
]Cl
2
(Nu=9-EtG
5
and 5′-GMP 7) and possibly
−
−
2
]Cl, analogous to the known cis-[Pt(L )(NH
3
2
) ]NO
3
22], can have several conformers in solution, differing with respect
to the relative orientations of the nucleobases [41–43].
The geometries of the mono-adducts of 9-EtG, trans-4 and cis-4,
and of four possible conformers of the bis-adduct 5 have been opti-
mized using DFT in vacuum (SI, Figs. S1 and S2). The X-ray structure
of the dichloro species 1a [22] has been used as a starting point to
generate the initial structures of the reaction products. The structure
of the most stable mono-adduct isomer trans-4 (SI, Fig. S1) is stabi-
lized by a hydrogen bond between the NH hydrogen and O(6) of the
Scheme 2. Reaction scheme for the interactions of complex 1a with 9-EtG and 5′-GMP.

58
A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
Fig. 2. Time-dependent changes in the 1H NMR spectrum of cis-[Pt(HL1)Cl
product, (■) unreacted 9-EtG and (●) unreacted 1a.
2
7
] (1a) upon addition of 2 eq. of 9-EtG in DMF-d , at room temperature. Peaks are marked as (*) reaction
in the naphthoquinone H5′-H8′proton chemical shifts, in the δ
.4–7.8 region, as the result of the deprotonation processes depicted
in Scheme 2.
The 195Pt NMR spectrum of the reaction mixture, recorded after
2 h, exhibited the peak at δ −2133 due to the starting material 1a,
besides two small peaks, at δ −2228 and −2305 ppm, characteristic
of the Pt nucleus in an N Cl environment [40], which were attributed
7
7
3
to the mono-adducts trans-4 and cis-4 (SI, Fig. S4). Due to its low con-
centration in solution, the bis-adduct 5 could not be detected in this
spectrum. However, the ESI-MS spectrum of the reaction mixture
7
2 h after mixing confirmed the formation of both the mono- and
bis-adducts, showing peaks at m/z of 745 and 888, respectively (SI,
Fig. S5).
These observations indicate that although chloride substitution
trans to the pyridyl group occurs faster than to the amine group,
both isomers trans-4 and cis-4 are formed. The fact that trans-4 re-
mains in solution is an indication of its stability, in accordance with
the DFT calculations. Mono-adduct cis-4 probably reacts faster to
give the bis-adduct 5. It should be noted that even after 3 days the
peaks corresponding to unreacted 1a remained visible in the 1NMR
spectrum. By measuring the integrals of the H11 signals of the
starting material 1a and the products it was found that about 35% of
1
a was left unreacted, possibly because part of the 9-EtG base is in-
volved in the deprotonation processes.
In order to study the effect of the steric hindrance of the DNA
model base, we also investigated the reactivity of the chlorido com-
plex 1a towards 5′-GMP, as described below.
3
.1.2. Interaction with 5′-GMP
2
Complex 1a was allowed to react with 5′-GMP (1:2) in a D O/
7
DMF-d (3:1) mixture, at room temperature. Also in this case, the
1
195
reaction was monitored by H and
Pt NMR spectroscopy and
mass spectrometry.
1
The H NMR spectra and the possible reaction products are illustrat-
ed in SI, Fig. S6 and Scheme 2, respectively. The interaction of 1a with
5
′-GMP was conveniently monitored by the changes in the resonance
of the H8 proton of 5′-GMP, at δ 7.99. The spectra taken over time
SI, Fig. S7) show that this peak decreased in intensity concomitantly
to the appearance of two new peaks, at δ 8.56 and 8.66 ppm, assigned
Fig. 3. Spectra of the reaction between compound 1a and 2 equiv. of 9-EtG after 15 min,
4, 48 and 72 h. The signals are due to H11 of the aminomethylnaphthoquinone ligand of
the starting complex 1a (see spectrum taken after 15 min.) and reaction products trans-4,
cis-4 and 5.
2
(

A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
59
to the two geometric isomers of the mono-adduct [Pt(HL1)(5′-GMP)Cl]
DNA was incubated with compounds 1w and 1b in 2 mM phos-
phate buffer (pHs 5.8 and 7.0) for 24 h in the absence of any external
reagent or light (Fig. 5). Cisplatin was used for comparison. Com-
pound 1w induced DNA strand breaks in phosphate buffer at different
pHs (Fig. 5a and b), whereas complex 1b does not show any clear
DNA cleavage (Fig. 5c and d). Increasing concentrations of 1w from
0 to 150 μM (Fig. 5a and b, lanes 1–3) induces DNA single-strand
breaks, as seen by the disappearance of the SC form and the conse-
quent increase in the OC form. However, a little decrease in the
DNA intensity is seen at 300 μM for 1w and from 20 μM for cisplatin,
which may be due to DNA precipitation (Fig. 5).
Cl, trans-6 and cis-6, as described above for the reaction of 1a with
1
9
-EtG. The H NMR spectrum recorded after 72 h exhibits the same sig-
nals, suggesting that no further substitution has occurred.
1
95
The
Pt NMR spectrum taken after 48 h also confirmed forma-
tion of the two mono-adducts trans-6 and cis-6, as it exhibited two
signals, at δ −2384 and −2400, in addition to the resonance due to
the starting material 1a, at δ −2180 (SI, Fig. S8). ESI-MS of the reac-
tion mixture after 72 h shows the peak at m/z=973 corresponding
to the mono-adduct 6 (SI, Fig. S9). The peak of the bis-adduct
containing two 5′-GMP ligands was not detected in this spectrum
(
SI, Fig. S9).
The cleavage of DNA by 1w was found to depend on the pH, as
shown by the complete disappearance of SC DNA at 300 μM at pH 5.8
(Fig. 5b, lane 4), whereas at the same concentration at pH 7.0
(Fig. 5a, lane 4), a little amount of SC DNA is still present. The cleavage
efficacy of this compound at low pH is an interesting result, considering
the acidic nature of solid tumors compared to the normal tissues [49].
Thus, although complex 1a seems to interact efficiently with 9-EtG
to give bifunctional adducts, formation of uniquely mono-adducts
with 5′-GMP is most probably due to the presence of the bulky
3
prevent formation of similar DNA bis-adducts exhibited by cisplatin.
The rates of substitution of Pt complexes are influenced significantly
by the presence of a heterocycle trans to the leaving group and in
some cases direct Cl substitution may occur [44] — it is also possible
that some of the kinetic differences observed for 9-EtG and 5′-GMP
-aminomethyl-1,4-naphthoquinone ligand, which may sterically
b
At pH 5.8, the amount of bound Pt per nucleotide (r values) is
higher than at pH 7.0 which indicates that the enhanced cleavage ef-
−
fect of 1w at pH 5.8 is due to higher levels of Pt-DNA binding. This
−
pH-dependence might be associated with the formation of [Pt(L
)
−
−
−
are due to varying proportions of leaving group (H
2
O, Cl ) caused
(H
2
O)(OH)] and/or [Pt(L )(OH)
2
]
species from deprotonation of
by differing solvent conditions for the purine and mononucleotide
respectively.
In order to investigate further the mode of interaction of the plat-
inum complexes and the HL pro-ligands with DNA, a series of studies
were undertaken, as described below.
H
2
O ligands, which have less ability to bind to DNA than the aquated
species.
These results can be correlated with the reduction potentials
of the compounds which are in the−1.2 to −1.0 V range for the
amino Pt2 complexes and the pro-ligands, and about −0.6 V for
the chlorido Pt2 complexes [22]. The latter derivatives, therefore,
are more easily reduced than the former [22]. Since reduction of the
naphthoquinone ring is the first step of the oxidative stress mecha-
nism, the production of reactive oxygen species involved in the DNA
cleavage should be closely related to the redox potential of the
naphthoquinone derivatives.
+
+
3
.2. Biophysical studies
In these experiments, the aqua species 1w–3w and cw (Scheme 1)
were used in order to facilitate the binding of Pt2 to the CT-DNA.
Compounds that exhibit strong DNA interactions, such as
intercalators, can affect the melting temperature of CT DNA by
+
~
15 °C [45]. HL1 and 1b exhibit only small ΔT
=0.1, indicative of weak DNA interaction
SI, Fig. S10). The complex 1w, however, shows a more significant
m
values of approxi-
3.4. Topoisomerase I inhibition studies
mately −3 °C at
(
ΔT
to that caused by cisplatin, and together with our results described
above on the interaction with nucleobases, may suggest covalent
binding, as described for this drug previously [46].
The binding of EtdBr to DNA by intercalation is substantially
blocked by the formation of platinum-DNA adducts, which results in
loss of fluorescence intensity as compared to the free DNA-EtdBr
complex [47]. The ability of compounds HL1–HL3, 1w–3w, 1b–3b
and cw to inhibit the intercalation of EtdBr into CT-DNA was exam-
ined (Fig. 4). The compounds decrease the intensity of EtdBr-DNA
fluorescence in the following order: 1w–3w>1b–3b≥HL1–HL3.
Although aquated cisplatin cw is a more efficient inhibitor than 2w
and 3w, compound 1w was overall the most efficient inhibitor of
EtdBr intercalation (70% at r=0.25). It is possible that due to lower
r
i
In higher eukaryotes, DNA topoisomerases I (Topo I) are essential
enzymes whose main role is to relieve DNA supercoiling (torsional
tension) ahead of replication and transcription complexes [50,51].
Naphthoquinone compounds e.g. β-lapachone, have been investigated
as topoisomerase inhibitors and, in view of the importance of Topo I as
a molecular target for anticancer drug development, we next investigated
the 2-hydroxy-3-(aminomethyl)-1,4-naphthoquinone pro-ligands and
their Pt2 complexes for Topo I inhibition.
m
i
value of −5 °C at r =0.1 (SI, Fig. S10). This effect is quite similar
+
3.4.1. DNA relaxation assay
Assays separating the supercoiled plasmid (SC) from the fully re-
laxed plasmid (Rel, including the open-circular or nicked form whose
accumulation is induced by Topo I poisons such as campthotecin) and
the partially relaxed topoisomers (Rn) (SI, Fig. S11) can provide qualita-
tive or semi-quantitative information on the Topo I inhibitory potencies
of tested compounds by comparing DNA band intensities of samples
and controls [52]. Topo I inhibitory activities of compounds HL1–HL3,
1
steric hindrance, the shorter carbon chain of 1w (R =Bu) facilitates
better DNA binding compared to 2w and 3w.
1
a–3a and 1b–3b, are summarized in Table 1.
3
.3. Formation of DNA strand breaks
Under the assay conditions, most Pt2 complexes were as active
+
as the CPT standard, inhibiting DNA relaxation by Topo I, except for
the chlorido Pt2 complexes 1a and 2a, which presented slightly
higher activity at the tested concentration of 100 μM (SI, Fig. S12).
The pro-ligands HL1–HL3 were inactive at 100 μM (SI, Fig. S12). All
active compounds were further tested at varying concentrations
(Fig. 6). A clear dose-response behavior was observed for complexes
1b (starting from 500 μM), 2a and 3b (both starting from 100 μM).
The chlorido-complex 2a was confirmed as the most potent inhibitor,
being able to produce detectable inhibition at concentrations down to
+
Quinone derived compounds can generate reactive oxygen species
ROS) through quinone redox cycling and, consequently, induce DNA
(
strand breaks in vitro and apoptosis [9,48]. For this reason, the ability of
compounds 1w and 1b to induce DNA cleavage was investigated. The
induction of DNA strand breaks can be monitored using supercoiled
plasmid DNA and visualized by gel electrophoresis [9]. When treated
with a cleavage agent, supercoiled DNA (SC) is converted into the
open-circle form (OC) by single-strand breaks and to linear DNA by
double-strand breaks. Extensive double-strand breaks lead to DNA
degradation.
2
+
25 μM. Hence, both amino and chlorido Pt
complex types were
able to inhibit Topo I-promoted DNA relaxation.

60
A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
enzyme (false positive). Re-supercoiling, in this case, is due to the change
in DNA linking number that accompanies relaxation by the enzyme [55].
Conversely, specific Topo I inhibitors (e.g. CPT) should not induce
re-supercoiling due to negligible DNA unwinding on binding [56].
It was found that complexes 1a–3a, and 1b and 2b do not stabilize
the cleavage complex, because the intensity of the OC DNA band pro-
duced by the action of Topo I on plasmid DNA is not enhanced (see
the CPT standard in SI, Fig. S13 for comparison of the behavior of a
known Topo I poison). Hence, this analysis confirms that compounds
1
a–3a, 1b and 2b are all true inhibitors of human Topo I enzyme and
demonstrate their action as catalytic inhibitors. On the other hand,
complex 3b caused major unwinding of DNA, as deduced from the
strong SC DNA band observed after treatment of the 3b/relaxed plas-
mid DNA mixture with Topo I (Fig. 7). Hence, for 3b further analyses
are necessary to determine its inhibitory activity over human Topo I.
Complex 3b most likely binds DNA in non-intercalative fashion since
our results also demonstrated the poor ability of 3b to inhibit EtdBr
intercalation between DNA base pairs (Fig. 4).
Fig. 4. Fluorescence of EtdBr in the presence of DNA with the compounds HL1–HL3,
1w–3w, 1b–3b and cw after 72 h incubation.
Noteworthy, among the pro-ligands, HL1 shows a weak, but still
significant, capacity to unwind DNA under the conditions of these
assays (SI, Fig. S14), indicating that Pt complexation is responsible
for the improved DNA intercalating activity seen in these compounds.
These results are in accordance with the other DNA binding assays
reported above showing that the Pt2 complexes interact more
strongly with DNA than the pro-ligands. Finally, the weak DNA inter-
calating activity of HL1 may be the actual reason behind the apparent
Topo I inhibitory activity observed for this compound at 250 μM.
The pro-ligand HL1 was re-tested at 125 μM as a negative control
for inhibition, but at the highest concentration of 250 μM it only
weakly inhibited DNA relaxation (Fig. 6). This result indicates that
the pro-ligands may either have a weak inhibitory activity for them-
selves or may be intercalating DNA at higher concentrations (see un-
winding assay below).
+
3
.4.2. DNA unwinding assay
Because relaxation assays alone fail to distinguish between Topo I in-
hibition, poisoning and DNA binding [53], we also employed a DNA
unwinding assay [54] to assess such distinction. In a simple relaxation
assay under the right conditions, DNA binders will cause re-supercoiling
of plasmid DNA after initial relaxation by Topo I and subsequent removal
of the compound and enzyme, thus appearing to have inhibited the
3.5. Cellular platinum accumulation
Cellular accumulation is an important factor governing the cytotox-
icity of platinum compounds. The inhibition of drug accumulation and
increase of efflux have been directly related to mechanisms of acquired
Fig. 5. Investigation of DNA strand breaks on pUC19 plasmid DNA in 2 mM phosphate buffer after 24 h incubation with (a) 1w and (c) 1b at pH 7.0 and (b) 1w and (d) 1b at pH 5.8
and (e) cisplatin at pH 7.0. Lane 1: DNA control, lanes 2–4: DNA+75, 150, 300 μM compounds. r values ([Pt]/[nucleotides]) for 1w, from 2 to 4, respectively, are: (a) 0.016, 0.023,
b
and 0.033; (b) 0.027, 0.063 and 0.104. The corresponding percentages of DNA, determined by densitometry, are indicated for complex 1w. Due to the lack of solubility in water, HL1
was not tested for DNA strand breaks.

A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
61
Table 1
Topoisomerase
same trend, where drug 2a has exhibited the highest accumulation
I
inhibitory activity of 3-(amino-
−17
and the lowest IC50 value (164×10
mol Pt/cell after 6 h; IC50 =
methyl)naphthoquinone derivatives and correspond-
2
+
a
6.4 μM, Fig. 8a). Complexes 1a, 2a, and 3a show much higher Pt accu-
mulation compared with cisplatin (4.4, 17.4 and 14-fold, respectively,
after 6 h in MDA-MB-435 and 6.2, 16.7 and 23-fold, respectively, after
6 h in PC-3) reflected in better cytotoxicity.
The advantages to cytotoxicity conferred by covalent bond forma-
tion are shown by the fact that even though the charged complexes
ing Pt complexes.
Compounds
Activity
HL1
HL2
HL3
−
−
−
1
2
3
1
2
3
a
a
a
b
b
b
++++
++++
+++
+++
+++
+++
2
b and 3b have considerable accumulation after 6 h in MDA-MB-435
−
17
(56.4 and 144×10
mol Pt/cell, respectively; Fig. 8a), they are still
essentially inactive against this cell line, whereas 1a exhibits similar ac-
−17
cumulation (54.7×10
IC50=19.7 μM).
Compounds 2a and 3a exhibit moderate cytotoxicity against PC-3
mol Pt/cell; Fig. 8a), but moderate activity
(
The compound's activity is represented by a scale
ranging from a “−” sign, meaning total absence of in-
hibitory activity (observed intensity of relaxed DNA
bands is equal or higher than the Topo I control reac-
tion without inhibitor), to “+” , “++”, “+++” and
up to “++++” sign, the latter meaning an inhibitory
activity equal or above the CPT standard.
and accumulate to levels approximately 25-fold higher than cisplatin,
which is not strongly cytotoxic against this cell line under these con-
ditions. Furthermore, the increase of the carbon chain length of R1
was found to strongly enhance the Pt accumulation of the compounds
which can be clearly correlated to their cytotoxicity. The relative con-
tributions of cellular accumulation and DNA damage to overall cyto-
toxicity and the comparison with cisplatin merits further studies
which are in progress.
a
All compounds were tested at 100 μM with 1%
DMSO, except for HL1 and 1b which were tested at
125 μM with 1.25% DMSO.
3
.6. Cellular DNA strand breakage
resistance to cisplatin [57]. Fig. 8 compares the cellular accumulation of
a–3a, 1b–3b and cisplatin upon treatment of melanoma and prostate
1
The comet assay, using TBE electrophoresis after alkali unwinding,
cancer cell lines, MDA-MB-435 and PC-3.
was used to detect DNA strand breaks that occur during drug treat-
ment in cells [58]. Assays were performed using MDA-MB-435 cells
after incubation with HL2, 2a and 2b at 100 μM in 1% DMSO for 4 h.
These compounds, with R =n-heptyl, were chosen due to the low
IC50 in melanoma cell line and high accumulation of 2a. Cisplatin
Enhanced drug accumulation parallels the R1 carbon chain length
(
a: n-Bu; b=n-heptyl and c=n-decyl) for both cell lines, which may
1
be due to the lipophilicity of the compounds. This effect is particularly
striking in the case of the cationic complexes whose accumulation in-
creased over 100 times (for the PC-3 cell line) and 40 times (for the
was used for comparison. H O2 was used as a positive control. The
levels of DNA cleavage are illustrated in Fig. 9.
2
1
MDA-MB-435 cell line), after a 6 h treatment, as the R group changes
from n-Bu to n-decyl, 1b–3b respectively. Concomitantly, 3b exhibits
the lowest IC50 against the PC-3 cell line.
For the MDA-MB-435 cell line, accumulation of the covalently bind-
ing drugs increases in the order 1ab3ab2a and the IC50 follows the
Significant levels of DNA breakage were induced by 2a, whereas HL2,
2b and cisplatin showed a similar profile to the untreated control (Fig. 9).
Even though one of the mechanisms of action of naphthoquinone deriv-
atives has been described as the production of reactive oxygen species
Fig. 6. Effect of the selected compounds on the inhibition of the supercoiled DNA relaxation promoted by Topo I. Supercoiled DNA (pBR322, 250 ng) was incubated with Topo I
10 U) in the presence of different concentrations of compounds and then detected on an agarose gel without ethidium bromide. Lane A, supercoiled DNA without enzyme;
(
lane B, relaxed DNA Marker; lane C, supercoiled DNA incubated with Topo I alone; lane D, supercoiled DNA with Topo I in the presence of 100 μM CPT. The thin arrow indicates
the supercoiled (SC) DNA band, the large arrow indicates the fully relaxed plasmid (Rel) band and the brackets indicate the topoisomers bands (Rn).

62
A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
Fig. 7. Effect of complex 3b on the inhibition of SC DNA relaxation and unwinding ac-
tivity to differentiate simple DNA binders from actual Topo I intercalative poisons or
catalytic inhibitors. “SC”: supercoiled DNA without enzyme (SC DNA band is indicated
by the solid thin arrow). “Relaxed DNA”: standard of circular DNA relaxed by human
Topo I, showing the fully relaxed plasmid (Rel, indicated by the large arrow) and the
topoisomers bands (Rn). “Topo I”: control for the enzyme activity without inhibitors.
Fig. 8. Platinum accumulation in: a) MDA-MB-435 cells; b) PC-3 cells. Cellular plati-
num accumulation was measured by ICP-MS after 2 (left column), 4 (middle column)
and 6 h (right column) of treatment at 20 μM of the indicated compounds. As deter-
mined previously, the IC50 (μM) for each compound measured after 72 h incubation
is indicated [22]. Experiments were performed in triplicate; error bars represent stan-
dard deviation values.
“CPT”: camptothecin, the standard for Topo I poisoning activity. “m-AMSA”: the stan-
dard for DNA intercalation. “ctr. unwinding”: control for the unwinding reaction
where supercoiled DNA is used in place of relaxed DNA to ensure that the tested com-
pound concentration would not inhibit the enzyme action, which is necessary to unveil
the compound unwinding activity. Top panel: analysis was carried out at 1% agarose
gel without ethidium bromide. Bottom panel: detection of nicked DNA, also called open
circle (OC) DNA (dotted arrow) after running gel for additional 1 h with 0.5 μg/mL
ethidium bromide added.
the formation of uniquely mono-adducts with 5′-GMP, and therefore
1a might interact similarly with DNA.
Chlorido Pt2 complexes [Pt(HL)Cl
+
] also induce DNA strand
2
breaks in vitro in phosphate buffer that is increased with the decrease
of the pH, which keeps the naphthoquinone species protonated, thus
favoring DNA binding. The ability of 2a to produce intracellular DNA
strand breaks in MDA-MB-435 cells was confirmed by comet assay
which showed significant levels of DNA breakage for this compound,
whereas HL2 and 2b were ineffective. The ability of the chlorido com-
plexes to produce DNA cleavage may be associated with their higher
reduction potential (around −0.6 V), compared to the pro-ligands
and amino Pt2 complexes (from −1.2 to −1.0 V) [22].
(
ROS) resulting in DNA damage [6,7,48], the Mannich bases (HL) studied
in this work do not provoke DNA strand breaks under the conditions
studied (SI, Fig. S15).
DNA damage induced by the covalent complex 2a is consistent with
its capacity to induce strand breaks in plasmid DNA (SI, Fig. S16). Com-
bined with high accumulation levels in MDA-MB-435 cells (Fig. 8a), the
results explain why 2a exhibits the highest cytotoxicity against melano-
ma, among all tested compounds [22] (SI, Table S1).
+
The pro-ligands exhibit neither strong DNA binding nor Topo I
inhibition. Yet, in direct cytotoxicity assays, they are more active
against most tested cells than their respective Pt2 complexes. We
attribute this unexpected behavior to the intracellular decomposition
of the Mannich bases to the unsaturated quinone methide observed
in solution (data not shown). Indeed, quinone methides are known
DNA alkylating agents [59]. Further studies are in progress to investi-
gate whether these compounds alkylate DNA, as proposed.
+
4
. Conclusions
These studies examined the mechanism of cytotoxic action of
naphthoquinone-platinum(II) complexes. While quinone-based drugs
show significant biological activity and DNA-binding in their own right,
the results presented here show that coordination and stabilization of
the quinone structure can effect marked changes in DNA reactivity.
The results of Topo I inhibition (HL1–HL3b1b–3b≤1a–3a) does
not follow the general cytotoxic trend (1b–3bb1a–3abHL1–HL3)
[22], which indicates that inhibition of this enzyme is not the mecha-
nism of action of the tested compounds at the cell level.
2
The binding properties of [Pt(HL1)Cl ] 1a have shown that complex
1
a is capable of forming the mono-adducts trans-4 and cis-4 and the
bis-adduct 5 with 9-EtG. In contrast, with the more sterically demand-
ing 5′-GMP, 1a can only form mono-adducts (trans-6 and cis-6) under
the same conditions. Considering that the related dichlorido complexes
containing the 2-pyridylmethylamine or 1-(2-pyridyl)ethylamine li-
gands interact efficiently with d(GpG) [42] to give bifunctional adducts,
it is clear that the steric bulk of the HL1 ligand in 1a is responsible for
Further contributions to cytotoxicity occur through cellular accu-
mulation. Accumulation studies confirm that increase of the R1 car-
bon chain length causes significant enhancement of cellular
accumulation of all complexes, consequently contributing to the
highest cytotoxicity being exhibited by the n-heptyl and n-decyl
derivatives.

A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
63
[
6] L. Flowers, S.T. Ohnishi, T.M. Penning, Biochemistry 36 (1997) 8640–8648.
[
7] J.L. Bolton, M.A. Trush, T.M. Penning, G. Dryhurst, T.J. Monks, Chem. Res. Toxicol.
13 (2000) 135–160.
[
8] J.M. Matés, F.M. Sánchez-Jiménez, Int. J. Biochem. Cell Biol. 32 (2000) 157–170.
[
9] A. Srinivasan, H.-J. Lehmler, L.W. Robertson, G. Ludewig, Toxicol. Sci. 60 (2001)
9
2–102.
10] J. Sperry, I. Lorenzo-Castrillejo, M.A. Brimble, F. Machín, Bioorg. Med. Chem. 17
2009) 7131–7137.
11] A. Esteves-Souza, D.V. Figueiredo, A. Esteves, C.A. Câmara, M.D. Vargas, A.C. Pinto,
A. Echevarria, Braz. J. Med. Biol. Res. 40 (2007) 1399–1402.
12] A.S. Cunha, E.L.S. Lima, A.C. Pinto, A. Esteves-Souza, A. Echevarria, C.A. Camara,
M.D. Vargas, J.C. Torres, J. Braz. Chem. Soc. 17 (2006) 439–442.
13] A.I. Francisco, A. Casellato, A.P. Neves, J.W.D. Carneiro, M.D. Vargas, L.C. Visentin,
A. Magalhães, C.A. Câmara, C. Pessoa, L.V. Costa-Lotufo, J.D.B.M. Filho, M.O. de
Moraes, J. Braz. Chem. Soc. 21 (2010) 169–178.
[
[
[
[
(
[
14] A.I. Francisco, T.P. Fragoso, M.D. Vargas, J.W.D. Carneiro, A. Casellato, F.C. Silva, V.F.
Ferreira, J.P. Barbosa, C. Pessoa, L.V. Costa-Lotufo, J.D.B.M. Filho, M.O. Moraes, A.S.
Mangrich, J. Braz. Chem. Soc. 21 (2010) 1293–1302.
[
[
[
15] A.S. Cunha, M.D. Vargas, C.R. Gattass, A.C. Pinto, C.A. Camara, A.S. Esteves, E.L.S.
Lima, Oncol. Rep. 20 (2008) 225–232.
16] A.P. Neves, K.C.B. Maia, M.D. Vargas, L.C. Visentin, A. Casellato, M.A. Novak, A.S.
Mangrich, Polyhedron 29 (2010) 2884–2891.
17] A.P. Neves, C.C. Barbosa, S.J. Greco, M.D. Vargas, L.C. Visentin, C.B. Pinheiro, A.S.
Mangrich, J.P. Barbosa, G.L. da Costa, J. Braz. Chem. Soc. 20 (2009) 712–727.
18] M.T. Leffer, R.J. Hathaway, J. Am. Chem. Soc. 70 (1948) 3222–3223.
19] C.E. Dalgliesh, J. Am. Chem. Soc. 71 (1949) 1697–1702.
20] N.M.F. Lima, C.S. Correia, P.A.L. Ferraz, A.V. Pinto, M.C.R.F. Pinto, A.E.G. Santana,
M.O.F. Goulart, J. Braz. Chem. Soc. 13 (2002) 822–829.
[
[
[
[
[
21] A. Baramee, A. Coppin, M. Mortuaire, L. Pelinski, S. Tomavo, J. Brocard, Bioorg.
Med. Chem. 14 (2006) 1294–1302.
22] A.P. Neves, G.B. Silva, M.D. Vargas, C.B. Pinheiro, L.C. Visentin, J.D.B. Marinho Filho,
A.J. Araújo, L.V. Costa-Lotufo, C. Pessoa, M.O. Moraes, Dalton Trans. 39 (2010)
10203–10216.
[
23] P. Marqués-Gallego, M.A. Gamiz-Gonzalez, F.R. Fortea-Pérez, M. Lutz, A.L. Spek, A.
Pevec, B. Kozlevcarc, J. Reedijk, Dalton Trans. 39 (2010) 5152–5158.
[24] R. Schobert, B. Biersack, A. Dietrich, S. Knauer, M. Zoldakova, A. Fruehauf, T.
Mueller, J. Med. Chem. 52 (2009) 241–246.
[25] H.M. Coley, J. Sarju, G. Wagner, J. Med. Chem. 51 (2008) 135–141.
[26] B. Rosenberg, L. VanCamp, J.E. Trosko, V.H. Mansour, Nature 222 (1969) 385–386.
[27] L. Kelland, Nat. Rev. Cancer 7 (2007) 573–584.
Fig. 9. Comet assay: MDA-MB-435 cells were incubated with 100 μM of each
compound for 4 h at 37 °C. As a positive control, 100 μM H was incubated for
0 min at 4 °C. a) Parameters of DNA damage, Tail moment and Olive moment± standard
deviation are indicated for each. b) Each figure is a representative image of 3 independent
repeats.
2 2
O
3
[
28] J.B. Mangrum, N.P. Farrell, Chem. Commun. 46 (2010) 6640–6650.
[
29] C.L. Smyre, G. Saluta, T.E. Kute, G.L. Kucera, U. Bierbach, ACS Med. Chem. Lett. 2
(
2011) 870–874.
Finally, we have shown that the incorporation of Pt2+ into
-hydroxy-3-(aminomethyl)-1,4-naphthoquinones causes significant
[30]
S. Top, E.B. Kaloun, A. Vessieres, G. Leclercq, I. Laïos, M. Ourevitch, C. Deuschel,
M.J. McGlinchey, G. Jaouen, ChemBioChem 4 (2003) 754–761.
2
[
31] C. Lottner, K.-C. Bart, G. Bernhardt, H. Brunner, J. Med. Chem. 45 (2002) 2064–2078.
changes in the biophysical properties of the pro-ligands, strongly
affecting their cytotoxicity and mechanism of action. Structural
changes, such as switch of the ligands from amino to chloride in the
Pt complexes and increase of the carbon chain length markedly en-
hance DNA interaction and accumulation in the cells.
[32] F. Zsila, Z. Bikádi, M. Simonyi, Org. Biomol. Chem. 2 (2004) 2902–2910.
[
[
33] D. Suh, J.B. Chaires, Bioorg. Med. Chem. 3 (1995) 723–728.
34] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian,
J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J.
Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth,
P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C.
Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V.
Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham,
C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen,
M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03, Revision B.02, Gaussian, Inc.,
Pittsburgh PA, 2003..
2
+
Acknowledgments
The authors thank the Brazilian agencies ‘Conselho Nacional de
Desenvolvimento Científico e Tecnológico’ (CNPq), ‘Coordenação de
Aperfeiçoamento de Pessoal de Nível Superior’ (CAPES) and ‘Fundação
de Amparo à Pesquisa do Estado do Rio de Janeiro’ (FAPERJ) for fi-
nancial support. FAPERJ-PRONEX (E-26/171.512.2010) is acknowl-
edged. M. D. Vargas and A. P. Neves are recipients of CNPq research
fellowships. This paper is in part supported by grants to NF from
National Institutes of Health (RO1 CA78754).
[35] A.D. Beck, J. Chem. Phys. 98 (1993) 5648–5652.
[36] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299–310.
[
[
[
37] M.V. Keck, S.J. Lippard, J. Am. Chem. Soc. 114 (1992) 3386–3390.
38] CometScore, vol. 1.5, TriTek Corp., Sumerduck, VA, 2006.
39] U.C.S. Yadav, K.V. Ramana, Y.C. Awasthi, S.K. Srivastava, Toxicol. Appl. Pharmacol.
227 (2008) 257–264.
[
[
40] S. Roy, J.A. Westmaas, F. Buda, J. Reedijk, J. Inorg. Biochem. 103 (2009) 1278–1287.
41] R. Ranaldo, N. Margiotta, F.P. Intini, C. Pacifico, G. Natile, Inorg. Chem. 47 (2008)
Appendix A. Supplementary data
2820–2830.
Supplementary data to this article can be found online at http://
dx.doi.org/10.1016/j.jinorgbio.2012.10.007.
[42] V.P. Munk, C.I. Diakos, L.T. Ellis, R.R. Fenton, B.A. Messerle, T.W. Hambley, Inorg.
Chem. 42 (2003) 3582–3590.
[
43] E. Pantoja, A. Gallipoli, S. van Zutphen, D.M. Tooke, A.L. Spek, C. Navarro-Ranninger,
J. Reedijk, Inorg. Chim. Acta 359 (2006) 4335–4342.
References
[44] N. Summa, W. Schiessl, R. Puchta, N.E. van Hommes, R. van Eldik, Inorg. Chem. 45
2006) 2948–2959.
(
[
[
1] C. Asche, Mini-Rev. Med. Chem. 5 (2005) 449–467.
2] E.L. Bonifazi, C. Ríos-Luci, L.G. León, G. Burton, J.M. Padrón, R.I. Misico, Bioorg.
Med. Chem. 18 (2010) 2621–2630.
3] R.C. Montenegro, A.J. Araujo, M.T. Molina, J.D.B.M. Filho, D.D. Rocha, E. Lopez-Montero,
M.O.F. Goulart, E.S. Bento, A.P.N.N. Alves, C. Pessoa, M.O. de Moraes, L.V. Costa-Lotufo,
Chem. Biol. Interact. 184 (2010) 439–448.
4] M. Tetef, K. Margolin, C. Ahn, S. Akman, W. Chow, L. Leong, R.J. Morgan, J. Raschko,
G. Somlo, J.H. Doroshow, Invest. New Drugs 13 (1995) 157–162.
5] K.R. Hande, Biochim. Biophys. Acta 1400 (1998) 173–184.
[45] I. Saha, M. Hossain, G.S. Kumar, Phys. Chem. Chem. Phys. 12 (2010) 12771–12779.
[46] R. Zaludová, V. Kleinwachter, V. Brabec, Biophys. Chem. 60 (1996) 135–142.
[47] A. Harris, Y. Qu, N. Farrell, Inorg. Chem. 44 (2005) 1196–1198.
[48] R.J. McKallip, C. Lombard, J. Sun, R. Ramakrishnan, Toxicol. Appl. Pharmacol. 247
(2010) 41–52.
[49] X. Zhang, Y. Lin, R.T. Gillies, J. Nucl. Med. 51 (2010) 1167–1170.
[50] Y. Pommier, Chem. Rev. 109 (2009) 2894–2902.
[51] Y. Pommier, E. Leo, H.-L. Zhang, C. Marchand, Chem. Biol. (2010) 421–433.
[52] C. Bailly, Methods Enzymol. 340 (2001) 610–623.
[
[
[

64
A.P. Neves et al. / Journal of Inorganic Biochemistry 119 (2013) 54–64
[
[
[
[
53] R. Palchaudhuri, P.J. Hergenrother, Curr. Opin. Biotechnol. 18 (2007) 497–503.
54] M.R. Webb, S.E. Ebeler, Biochem. J. 384 (2004) 527–541.
55] A.D. Bates, A. Maxwell, DNA Topology, Oxford University Press, 2005.
56] R.E. McKnight, A.B. Gleason, J.A. Keyes, S. Sahabi, Bioorg. Med. Chem. Lett. 17
[57] J. Zisowsky, S. Koegel, S. Leyers, K. Devarakonda, M.U. Kassack, M. Osmak, U.
Jaehde, Biochem. Pharmacol. 73 (2007) 298–307.
[58] P.L. Olive, J.P. Banath, Nat. Protoc. 1 (2006) 23–29.
[59] D. Verga, M. Nadai, F. Doria, C. Percivalle, M. Di Antonio, M. Palumbo, S.N. Richter,
M. Freccero, J. Am. Chem. Soc. 132 (2010) 14625–14637.
(2007) 1013–1017.