Nuclease and anti-proliferative activities of copper(ii) complexes of N3O tripodal ligands involving a sterically hindered phenolate

Article abstract of DOI:10.1039/c3dt32659d

Copper(ii) complexes 1²⁺-6 of a series of tripodal ligands involving a N₃O donor set, namely 2-[(bis-pyridin-2-ylmethyl-amino)- methyl]-4-methoxy-phenol (1L), 2-tert-butyl-4-methoxy-6-[bis-pyridin-2-ylmethyl- amino)-methyl]-phenol (2L), 2-tert-butyl-4-methoxy-6-{[(2-pyridin-2-yl-ethyl)- pyridin-2-ylmethyl-amino]-methyl}-phenol (3L), 2-tert-butyl-4-methoxy-6-{[(6- methyl-pyridin-2-ylmethyl)-pyridin-2-ylmethyl-amino]-methyl}-phenol (4L), 2-tert-butyl-4-fluoro-6-{[(6-methyl-pyridin-2-ylmethyl)-pyridin-2-ylmethyl- amino]-methyl}-phenol (5L) and 2-tert-butyl-4-methoxy-6-{bis[(6-methyl-pyridin- 2-ylmethyl)-amino]-methyl}-phenol (6L), respectively, were synthesized. Complexes 1²⁺, 3⁺ and 4⁺ were structurally characterized by X-ray diffraction. The structure of 1²⁺ is dimeric, with an essentially trigonal bipyramidal geometry around the copper(ii) ions and two bridging deprotonated phenolate moieties. The mononuclear complexes 3 ⁺ and 4⁺ contain a square pyramidal copper ion, coordinated in axial position by the phenol moiety. In the water-DMF (90:10) mixture at pH 7.3 all the copper(ii) complexes are mononuclear, mainly under their phenolate neutral form (except 3⁺), with a coordinated solvent molecule. The DNA cleavage activity of the complexes was tested towards the X174 DNA plasmid. In the absence of an exogenous agent 1²⁺ does not show any cleavage activity, 2⁺ and 3⁺ are moderately active, while 4⁺, 5⁺ and 6⁺ exhibit a high nuclease activity. Experiments in the presence of various scavengers reveal that reactive oxygen species (ROS) are not involved in the strand scission mechanism. The cytotoxicity of the complexes was evaluated on bladder cancer cell lines sensitive or resistant to cisplatin. The IC₅₀ values of the complexes 2⁺, 4⁺, 5⁺ and 6⁺ are lower than that of cisplatin (range from 6.3 to 3.1 μM against 9.1 μM for cisplatin). Furthermore, complexes 2⁺, 4⁺, 5⁺ and 6 ⁺ are able to circumvent cisplatin cellular resistance.

Full text of DOI:10.1039/c3dt32659d

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Nuclease and anti-proliferative activities of copper(II)
complexes of N3O tripodal ligands involving a sterically
hindered phenolate†
Cite this: Dalton Trans., 2013, 42, 8468
Nathalie Berthet,a Véronique Martel-Frachet,b Fabien Michel,a Christian Philouze,a
Sylvain Hamman,a Xavier Ronotb and Fabrice Thomas*a
Copper(II) complexes 12+–6 of a series of tripodal ligands involving a N3O donor set, namely 2-[(bis-
pyridin-2-ylmethyl-amino)-methyl]-4-methoxy-phenol (1L), 2-tert-butyl-4-methoxy-6-[bis-pyridin-2-
ylmethyl-amino)-methyl]-phenol (2L), 2-tert-butyl-4-methoxy-6-{[(2-pyridin-2-yl-ethyl)-pyridin-2-ylmethyl-
amino]-methyl}-phenol (3L), 2-tert-butyl-4-methoxy-6-{[(6-methyl-pyridin-2-ylmethyl)-pyridin-2-ylmethyl-
amino]-methyl}-phenol (4L), 2-tert-butyl-4-fluoro-6-{[(6-methyl-pyridin-2-ylmethyl)-pyridin-2-ylmethyl-
amino]-methyl}-phenol (5L) and 2-tert-butyl-4-methoxy-6-{bis[(6-methyl-pyridin-2-ylmethyl)-amino]-
methyl}-phenol (6L), respectively, were synthesized. Complexes 12+, 3+ and 4+ were structurally character-
ized by X-ray diffraction. The structure of 12+ is dimeric, with an essentially trigonal bipyramidal geometry
around the copper(II) ions and two bridging deprotonated phenolate moieties. The mononuclear com-
plexes 3+ and 4+ contain a square pyramidal copper ion, coordinated in axial position by the phenol
moiety. In the water–DMF (90 : 10) mixture at pH 7.3 all the copper(II) complexes are mononuclear,
mainly under their phenolate neutral form (except 3+), with a coordinated solvent molecule. The DNA
cleavage activity of the complexes was tested towards the ϕX174 DNA plasmid. In the absence of an
exogenous agent 12+ does not show any cleavage activity, 2+ and 3+ are moderately active, while 4+, 5+
and 6+ exhibit a high nuclease activity. Experiments in the presence of various scavengers reveal that
reactive oxygen species (ROS) are not involved in the strand scission mechanism. The cytotoxicity of the
complexes was evaluated on bladder cancer cell lines sensitive or resistant to cisplatin. The IC50 values of
the complexes 2+, 4+, 5+ and 6+ are lower than that of cisplatin (range from 6.3 to 3.1 µM against 9.1 µM
for cisplatin). Furthermore, complexes 2+, 4+, 5+ and 6+ are able to circumvent cisplatin cellular resistance.
Received 6th November 2012,
Accepted 5th April 2013
DOI: 10.1039/c3dt32659d
www.rsc.org/dalton
more, patients with drug-resistant breast, colon carcinoma or
lung cancer showed higher serum copper contents than
Introduction
Copper plays an important role in biological systems, as do patients whose tumors respond to treatment.5 These essential
many other metals. It is involved in many primordial processes observations open new prospects in anticancer strategies.
such as dioxygen transport, oxidation processes, signalling, For example it has been recently suggested that the copper
etc. Recently, copper has also been shown to be a cofactor level could be used as a marker of drug resistance.5 On the
essential for the tumour angiogenesis processes, in contrast to other hand, drugs harbouring a biologically relevant metal,
other transition metals.1,2 High serum and tissue levels of such as copper, were proposed to be more effective and less
copper were, for instance, found in many types of human toxic.2,6 They could thus represent a valuable and cheaper
cancer (breast, prostate, colon, lung, and brain).3,4 Further- alternative to cisplatin ([cis-diamminedichloridoplatinum(II)]),
which is so far the most used anti-cancer metallodrug,7 in
spite of severe side-effects and cellular resistance.8 For
example, cisplatin-based chemotherapy is the first-line therapy
aDépartement de Chimie Moléculaire, Chimie Inorganique Redox Biomimétique
(CIRE), UMR CNRS 5250, Université J. Fourier, B.P. 53, 38041 Grenoble cedex 9,
for metastatic bladder cancer with a response rate of approxi-
mately 50%.9 For patients ineligible for cisplatin, there is a
continued need to identify new and more effective drugs.
It has been known for decades that copper complexes could
cleave DNA efficiently in vitro (Fig. 1). In most of the cases they
are active only under specific conditions and DNA cleavage is
France. E-mail: Fabrice.Thomas@ujf-grenoble.fr; Fax: +33 476 514 836;
Tel: +33 476 514 373
bAGing Imaging Modeling, FRE 3405, Université Joseph Fourier, CNRS, EPHE, UPMF,
38706 La Tronche cedex, France
†Electronic supplementary information (ESI) available. CCDC 795941, 795943
and 795944. For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/c3dt32659d
8468 | Dalton Trans., 2013, 42, 8468–8483
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 1 Copper complexes exhibiting a DNA cleavage activity: (a) in the pres-
ence of a reductant or H2O2;11 (b) upon irradiation,13 (c, d, e) without an exter-
nal agent.17,22,23
Fig. 2 Formula of the ligands used in this study.
promoted by reactive oxygen species (ROS).10–14 As a result, the
nuclease activity is observed upon the addition of a reductant
(mercaptan) or a co-oxidant (such as H2O2) to the copper(II)
A prototypical example is the dicopper(II)
complex of the bis(bipyridylamine) ligand shown in Fig. 1a.
proposed later by Ghosh et al. to account for the nuclease
activity of related copper(II) and zinc(II) phenolate complexes.23
In this study, we investigate the coordination chemistry, the
DNA cleavage activity (in the absence of an external agent) and
the anti-proliferative effect against a bladder tumor cell line of
a series of copper(II) complexes involving a sterically hindered
(and thus pro-phenoxyl radical) phenolate group (Fig. 2).
Variations in the phenolic ortho substituent, as well as in the
alkylpyridine arms, are shown to dramatically influence both
DNA cleavage and biological activity.
complex.11,12
The mercaptan reduces the copper(II) center into copper(I),
2−
which then reacts with dioxygen to form reduced ROS (O2
,
O2˙− and/or OH˙) responsible for DNA cleavage. The addition
of H2O2 to similar copper(II) complexes also induces DNA clea-
vage, but in this case metal coordination facilitates deprotona-
tion of hydrogen peroxide, thereby dramatically enhancing its
reactivity.10 An alternative strategy to generate toxic ROS is the
use of copper(II) complexes involving a photo-sensitizer such
as dpq (Fig. 1). Upon irradiation the copper complex depicted
in Fig. 1b 13 forms highly reactive singlet oxygen 1O2, which
damages DNA.13,14
Results and discussion
Synthesis of the ligands and copper complexes
The ligands 1L, 2L, 4L, 5L and 6L (Fig. 2) were obtained by
condensation of the appropriate phenol with bis-pyridin-2-
Copper complexes that cleave DNA without the need for an
external agent are more interesting for an in cellulo use, but ylmethyl-amine derivatives in the presence of formaldehyde.
they are much rarer. Their development is a subject of current 3L was synthesized by reductive amination of 3-tert-butyl-2-
interest.15–23 The nuclease activity of copper(II) complexes in
hydroxy-5-methoxy-benzaldehyde in the presence of (2-pyridin-
the absence of an exogenous agent results from either hydro- 2-yl-ethyl)-pyridin-2-ylmethyl-amine; the complexes were syn-
lytic or oxidative catalysis. Neves et al.15,16 and Palaniandavar thesized by reacting the ligand with one molar equivalent of
et al.17,18 developed several copper complexes (Fig. 1c) able to
Cu(ClO4)2·6H2O in methanol (12+) or acetonitrile (2+–5+).
mediate hydrolytic DNA cleavage: The role of the metal ion is
The copper complex of 1L was isolated in the presence of
here to favour the nucleophilic attack at the DNA phosphates. base in dimeric form. Brown single crystals of 1·(ClO4)2·
The copper complexes of Tambjamine E and Doxycycline are
(CH3CN)4 were grown by slow diffusion of diethyl ether into a
rare examples of oxidative cleavage whereby the ligand itself methanolic solution of the complex (Table 1). The copper com-
reduces Cu(II) into Cu(I), which further reduces O2 into reactive plexes of 2L–5L, namely 2+–5+, were isolated as monomers due
O2˙−.19,20 ROS were also identified as responsible for DNA
to the steric bulk at the ortho position of the phenol. We
cleavage in some other cases but their origin was unclear.21 reported previously the crystal structure of 2·ClO4.24 Blue crys-
Recently, a high DNA cleavage activity has been reported tals of 3·ClO4, 4·ClO4 and 5·ClO4 were grown by slow diffusion
for the copper(II)-phenolate complex depicted in Fig. 1d. This
of diethyl ether into a concentrated solution of the complex in
complex also exhibits a significant anti-proliferative activity acetonitrile. Only the former two gave crystals of sufficient
towards L1210 murine leukemia cancer cell line and quality for an X-ray diffraction analysis. Attempts to crystallize
A2780 human ovarian carcinoma cell line, with IC50 values
the copper complex of 6L in a similar manner were unsuccess-
similar to cisplatin (3.4–10.3 μM).22 Its reactivity towards DNA ful: mixing one equivalent of ligand with the copper salt
was quite original since the proposed reactive species is a Cu(II)- in CH3CN affords a blue solution that quickly turns green and
phenoxyl radical formed during aerobic oxidation of the
then colourless. It is likely that the steric demand of two
phenolate complex.22 Similar phenoxyl radical species were α-methylpyridines decreases the copper binding constant.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8468–8483 | 8469

View Article Online
Paper
Dalton Transactions
Table 1 Crystallographic data for the copper complexes 1·(ClO4)2·(CH3CN)4, 3·ClO4 and 4·ClO4
1·(ClO4)2·(CH3CN)4
3·ClO4
4·ClO4
Formula
M
C48H52Cu2N10O12Cl2
1158.98
C27H34CuN4O10Cl2
709.02
C27H34CuN4O10Cl2
709.02
Crystal system
Space group
a/Å
b/Å
c/Å
α/°
β/°
Monoclinic
P121/c1
11.731(2)
19.464(4)
10.997(2)
90
93.31(3)
90
2506.6(9)
2
Monoclinic
P121/n1
15.395(3)
10.905(2)
18.813(4)
90
90.53(3)
90
3158(1)
Triclinic
P1
ˉ
8.792(2)
9.363(2)
19.591(4)
83.90(3)
78.29(3)
83.83(3)
1564.2(5)
2
γ/°
V/Å3
Z
4
T/K
150
1.536
10.28
293
1.491
9.21
293
1.505
9.30
Dc/g cm−3
μ (cm−1
)
Monochromator
Wavelength
Reflections collected
Independent reflections (Rint
Observed reflections
R
Graphite
MoKα (0.71073 Å)
30 584
10 407 (0.0437)
7267 [I > 2σ(I)]
0.0433
Graphite
MoKα (0.71073 Å)
34 696
6544 (0.0622)
4950 [I > 2σ(I)]
0.0564
Graphite
MoKα (0.71073 Å)
47 781
9782 (0.0381)
7148 [I > 2σ(I)]
0.0457
)
Rw
0.1045
0.1451
0.1081
Since unchelated copper(II) is known to be a strong oxidizer
in CH3CN it may slowly oxidize the phenol into a phenoxyl
radical (well-known green colour), which further decom-
poses.25 This assumption is confirmed by the fact that repla-
cing CH3CN by DMF prevents this phenomenon. The desired
complex 6+ was thus generated in situ in DMF for further analy-
sis by adding one molar equivalent of Cu(ClO4)2·6H2O to the
ligand 6L. Noteworthy, 6+ is mainly in its phenolate form
(see ESI†) in this solvent, due to the basic character of DMF.
Fig. 4 X-ray crystal structures of 3+ and 4+ shown with 30% thermal ellipsoids.
X-ray crystal structures of the copper complexes
H atoms are omitted for clarity, except H1.
The ORTEP views of 12+, 3+, 4+ are shown in Fig. 3 and 4, and
selected bond lengths and angles are listed in Table 2.
In 12+ the two copper atoms have similar pentacoordinate
environments that share the oxygen atoms O1 and O1* from
two ligands. The geometry around the metal ion is intermedi-
ate between trigonal-bipyramidal and square-pyramidal. The
Table 2 Selected bond lengths (Å) and angles (°) for the copper complexes
12+, 3+ and 4+
12+
Cu–O1 2.110(1)
Cu–N2 2.022(2)
O1–Cu–O1* 82.1(1)
O1–Cu–N3 117.5(1)
O1*–Cu–N3 95.5(1)
N2–Cu–N3 140.1(1)
3+
Cu–O1* 1.944(1)
Cu–N3 2.000(2)
O1–Cu–N1 93.4(1)
O1*–Cu–N1 174.5(1)
N1–Cu–N2 82.6(1)
Cu–N1 2.026(2)
O1–Cu–N2 100.8(1)
O1*–Cu–N2 101.2(1)
N1–Cu–N3 84.0(1)
Cu–O1 2.460(4)
Cu–N2 2.001(3)
O1–Cu–O10 170.7(2)
O1–Cu–N3 99.7(2)
N1–Cu–N3 93.8(2)
N2–Cu–N4 92.3(2)
4+
Cu–O1 2.444(2)
Cu–N2 2.016(2)
O1–Cu–O6 176.5(1)
O1–Cu–N3 95.4(1)
N1–Cu–N3 82.3(1)
N2–Cu–N4 92.7(1)
Cu–O10 2.826(6)
Cu–N3 2.008(3)
O1–Cu–N1 87.9(1)
O1–Cu–N4 87.4(2)
N1–Cu–N4 173.3(2)
N3–Cu–N4 91.7(2)
Cu–N1 2.062(3)
Cu–N4 2.014(3)
O1–Cu–N2 92.6(1)
N1–Cu–N2 83.1(2)
N2–Cu–N3 167.3(2)
Cu–O6 2.495(2)
Cu–N3 2.066(2)
O1–Cu–N1 89.5(1)
O1–Cu–N4 91.6(1)
N1–Cu–N4 175.4(1)
N3–Cu–N4 102.0(1)
Cu–N1 2.013(2)
Cu–N4 1.985(2)
O1–Cu–N2 90.0(1)
N1–Cu–N2 82.9(1)
N2–Cu–N3 164.5(1)
Fig. 3 X-ray crystal structures of 12+ shown with 30% thermal ellipsoids. H
atoms are omitted for clarity.
8470 | Dalton Trans., 2013, 42, 8468–8483
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
relative extent of the trigonal-bipyramidal distortion is indi-
The geometry of 4+ is roughly similar to that observed for
cated by an index τ representing the degree of trigonality 3+, with the metal ion lying within an axially elongated octa-
within the structural continuum between square-planar hedral geometry. The copper atom is coordinated in the equa-
and trigonal-bipyramidal structures.26 The τ index calculated torial plane by one tertiary nitrogen N1, two pyridine nitrogens
according to the equation τ = (β − α)/60 (where β = O1*–Cu–N1 N2 and N3 and one acetonitrile nitrogen N4. The phenol
and α = N2–Cu–N3) is 0.58. Assuming that 0 corresponds to a oxygen O1 and one perchlorate oxygen O5 bind weakly in axial
perfect square pyramidal geometry and 1 a perfect trigonal position with Cu–O1 and Cu–O5 bond distances of 2.444(2)
bipyramid, the geometry around the metal center in 12+ could and 2.495(2) Å respectively. The copper–nitrogen bond dis-
be thus better described as distorted trigonal-bipyramidal. tances Cu–N1, Cu–N2, Cu–N3 and Cu–N4 are 2.013(2),
Each copper atom is coordinated to three nitrogen and two 2.016(2), 2.066(2) and 1.985(2) Å respectively. It is noteworthy
oxygen atoms: the pyridine nitrogens N2 and N3, the tertiary that the Cu–N3 bond is the weakest Cu–N bond in the series,
amino nitrogen N1 and two μ-phenolato oxygens O1, O1*. as a result of the incorporation of a steric bulk at the α-posi-
The Cu–O1 bond (2.110(1) Å) is longer than the Cu–O1* one tion of the pyridine.29 The angle between the two planes
(1.944(1) Å) and the O1–Cu–O1* angle (82.1(1)°) is significantly defined by the N1–Cu–N3 and N2–Cu–N4 atoms is rather
different from the Cu–O1–Cu one (97.9(1)°). The CuO1CuO1* small (only 4.8°), in contrast with 2+ and 3+. The weakening of
unit is found to be completely flat, in contrast with butterfly one of the pyridine–metal bonds thus favours an almost
shapes commonly observed in dimers of tripodal ligands invol- square planar equatorial geometry around the copper ion.
ving two phenolato donors (N2O2 donor set).24,27 As a con-
EPR properties of the complexes
sequence the intermetallic distance is longer, the Cu–Cu bond
length being 3.060(1) Å.
The EPR spectra of the copper(II) complexes were recorded in a
water–DMF (90 : 10) mixture due to their low solubility in pure
water. The 100 K EPR spectra recorded at pH 7.3 are shown in
Fig. 5. They all consist of an (S = 1/2) signal with copper hyper-
fine structure, which is typical of mononuclear complexes.
This result is not surprising for complexes 2+–5+ since they
were isolated as monomers in the solid state. Regarding 12+, a
water or DMF molecule likely acts as a fifth ligand, thereby
replacing one μ-phenolate moiety in the copper coordination
sphere (Scheme 1).
In order to confirm this hypothesis we recorded the EPR
spectrum of 12+ in the less coordinating solvent CH3CN. No
signal could be evidenced at 100 K, showing that 12+ cannot be
described as two isolated (S = 1/2) spins. Clearly, the two
copper(II) ions are magnetically exchange coupled. This result
3+ contains a copper atom within an axially elongated octa-
hedral geometry. One tertiary amine N1 and two pyridine N2,
N3 nitrogen donors from the ligand, as well as one nitrogen
N4 from an exogenous acetonitrile molecule coordinate in the
equatorial positions. The axial positions are occupied by two
oxygen donors, one from the phenol moiety (O1) and the
other (O10) from an exogenous perchlorate ion. The copper–
nitrogen bond distances Cu–N1, Cu–N2, Cu–N3 and Cu–N4 are
2.062(3), 2.001(3), 2.008(3) and 2.014(3) Å respectively, whereas
the axial copper–oxygen bond distances Cu–O1 and Cu–O10
are found to be much longer, 2.460(4) and 2.826(6) Å, with a
copper displaced 0.068 Å above the mean plane formed by the
four nitrogen donors, toward the axial O1 atom. The large
axial bonds reflect the very weak donor ability of the phenol
and perchlorate ligands, and the sensitivity of the d9 electronic
configuration to Jahn–Teller distortion. Interestingly,
a
hydrogen bond exists between the phenol hydrogen H1 and
the oxygen of one perchlorate molecule (O1–O4 bond distance
of 2.76 Å). It is noticeable that the two copper–pyridine nitro-
gen bond distances Cu–N2 and Cu–N3 were found to be short,
nearly equivalent in the previously described complex 2+
(1.974(3) and 1.983(3) Å). The longer Cu–N2 and Cu–N3 bond
distances observed in the case of 3+ (2.001(3) and 2.008(3) Å)
likely originate from the formation of an in-plane 6-membered
rather than 5-membered coordination ring.28 The two pyridine
nitrogen atoms N2 and N3 are located 0.153 and 0.138 Å below
the mean plane defined by the N1,N2,N3,N4 atoms, whereas
the tertiary amine and acetonitrile nitrogens are located 0.149
and 0.142 Å above this mean plane, with an angle between the
two N1–Cu–N3 and N2–Cu–N4 planes of 12.8°. In the case of
2+, we have reported that the mean displacement of the nitro-
gens with respect to the four nitrogens plane was only 0.055 Å,
whereas the angle between the two opposite N–Cu–N planes
was 10.2°. Therefore, an increase in the length of the linker
between the tertiary amine nitrogen and one pyridine results
in a marked tetrahedral distortion of the copper ion.
Fig. 5 X-band EPR spectra of 0.5 mM solutions of (a) 12+, (b) 2+, (c) 3+, (d) 4+,
(e) 5+ and (f) 6+ in a water–DMF 90 : 10 mixture at pH 7.3 ([Tris] = 0.05 M,
[NaCl] = 0.02 M). Solid lines represent experimental spectra, and dotted lines
represent simulations using parameters given in the text. T = 100 K, microwave
frequency 9.34 GHz (a, c, d, f) or 9.44 GHz (b, e), power 10 mW, modulation
frequency 100 kHz, amplitude 0.3 mT.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8468–8483 | 8471

View Article Online
Paper
Dalton Transactions
Scheme 1 Dissociation of the dimer 12+ in aqueous solution (L = H2O or DMF).
Table 3 Spin Hamiltonian parameters for mononuclear copper complexesa
g values
gxx, gxy, gzz
ACu (mT)
Axx, Ayy, Azz
Complex
Fig. 6 UV-Vis spectra of 0.5 mM solutions (except (e): 0.25 mM) of the com-
plexes in a water–DMF 90 : 10 mixture at pH 7.3 ([Tris] = 0.05 M, [NaCl] =
0.02 M): (a) 12+, (b) 2+, (c) 3+, (d) 4+, (e) 5+, (f) 6+. T = 298 K, l = 1.000 cm.
12+
2+
3+
4+
5+
6+
2.053, 2.053, 2.248
2.057, 2.057, 2.253
2.055, 2.055, 2.262
2.055, 2.055, 2.255
2.057, 2.057, 2.245
2.053, 2.053, 2.265
1.0, 1.0, 17.0
1.5, 1.5, 16.3
1.0, 1.0, 15.5
1.5, 1.5, 16.1
1.5, 1.5, 17.1
1.0, 1.0, 15.3
Table 4 UV-Vis properties of the complexes at pH 7.3 in water–DMF (90 : 10)
and thermodynamic constants
a Concentration of the complexes: 0.5 mM; medium: water + 0.05 M
Tris HCl buffer pH 7.5 + 10% (vol) DMF. T = 100 K.
Complex
λmax/nm (ε/M−1 cm−1 a
)
pKa (phenol)c
Kb e (M−1
)
d
12+
2+
3+
4+
5+
6+
470 (880), 690br (340)
501 (620)/501 (890)b
496 (970)
500 (880)
468 (1070)
—
7770 1500
2580 300
1940 550
6.98 0.02
5.80 0.03
5.49 0.02
is not surprising if one considers 12+ as dimer. Both the short
intermetallic distance and the favorable orbital overlap lead to
a significant magnetic interaction between the paramagnetic
(S = 1/2) metal ions. The resulting system is diamagnetic or
paramagnetic (S = 1).
From the simulation of the spectra in the water–DMF
(90 : 10) mixture, we obtained the spin Hamiltonian para-
meters listed in Table 3. They are consistent with an essentially
square pyramidal geometry for the copper center. Interestingly,
slight differences of gzz are observed in the complexes, reflect-
ing distortions in the metal ion geometry within the series.
f
—
—
d
f
—
g
521 (950)
4.44 0.04
—
a 0.05 M Tris buffer + 0.02 M NaCl at 298 K. b The first value at pH 7.3
in 0.05 M Tris + 0.02 M NaCl, and the second value at pH 9 in NaClO4
0.1 M medium. c 0.1 M NaClO4. d Precipitation is observed at neutral
pH in the presence of 0.1 M NaClO4. e Binding constant for CT DNA in
a 0.05 M Tris buffer + 0.02 M NaCl at 298 K. f The spectral variations
are too small to determine a binding constant. g Significant DNA
digestion occurs during titration.
whereas a low-energy shoulder is rather indicative of square
pyramidal geometry. The significant overlap between the CT
transitions and the d–d bands precludes such a detailed analy-
Electronic spectra of the complexes
The electronic spectra of the copper complexes were recorded sis. Nevertheless the observed NIR-tailoring, together with the
in a water–DMF (90 : 10) mixture at pH 7.3 (Fig. 6, Table 4). EPR results, is fully consistent with a square pyramidal geome-
The electronic spectrum of 12+ displays two main transitions at try of the copper ion. This assumption is further confirmed by
470 nm (ε = 880 M−1 cm−1) and 690 nm (ε = 340 M−1 cm−1). the close similarity between the spectra of 2+, 3+, 4+, 5+, 6+ and
Based on its low intensity the latter is attributed to d–d tran- some structurally characterized copper(II) phenolate complexes
sitions. The higher energy band, which is absent in the spec- where the metal ion lies in a square pyramidal geometry.28–32
trum of the free ligand, arises from a phenolate-to-copper CT On the basis of the reported ε values for phenolate-to-copper
transition. The spectra of 2+, 3+ and 4+ exhibit similar shapes, CT (ca. 1000 M−1 cm−1) for complexes having the same geome-
with a dominating band at ca. 500 nm and a less intense try24,28,29 it is reasonable to assume that 3+ and 4+ exist almost
shoulder above 700 nm (Table 4). Comparison with the litera- exclusively in their deprotonated phenolate form at pH 7.3
ture data24 shows that the former band arises from a pheno- (in other words, their phenol’s pKa is lower than 6) (Scheme 2).
late-to-copper(II) charge transfer (CT), while the latter is
Of interest is the lower intensity of the LMCT band of 2+,
assigned to d–d transitions. For copper(II) complexes of tri- which suggests that the phenol’s pKa is much higher for this
podal ligands it has been proposed that the shape of the d–d compound (see below). We previously reported the spectrum
band, and especially the presence of a high or low-energy of 2+ in its deprotonated phenolate form in CH3CN, which
shoulder, reflects the geometry of the metal ion.26 A high- is characterized by an absorption maximum at ca. 530 nm.29
energy shoulder is indicative of trigonal bipyramidal geometry, The shift of 30 nm resulting from the change of solvent
8472 | Dalton Trans., 2013, 42, 8468–8483
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
phenolate-to-copper CT band, consistent with protonation of
the phenolate moiety (Scheme 2). The titration data were
refined with SPECFIT software in order to extract the phenol’s
pKas. The computed pKas are found to be 6.98 0.02, 5.80
0.03, 5.49 0.02 and 4.44 0.04 for 2+, 3+, 4+ and 6+ respect-
ively. Noteworthy, the pKa value measured for 2+ is close to
that reported for the axial tyrosine in the copper active site of
Galactose Oxidase (ca. 8)30 and structurally characterized
copper complexes involving axial phenol/phenolates.31
Scheme 2 Acid–base equilibrium for 2+ in H2O–DMF solution (L = H2O or
DMF).
The trend of pKa 2+ > 4+ > 6+ indicates that substitution of
the pyridines by methylpyridines stabilizes the phenolate
forms. This feature likely arises from geometrical rearrange-
ments associated with deprotonation.32 Indeed, phenols are
weak donors that systematically coordinate the copper in axial
position. Exhaustive structural characterizations on copper(II)-
phenolate complexes of N3O tripodal ligands related to 2+
(i.e. involving exclusively two pyridin-2-ylmethyl N-donors)
revealed that the phenolate moiety usually coordinates in axial
position.24,33,34 On the other hand, Shimazaki et al. investi-
gated copper(II)-phenolate complexes of N3O tripodal ligands
involving one or two 6-methylpyridin-2-ylmethyl N-donors.29
They reported that the phenolate group occupies systematically
the equatorial position. This behaviour is ascribed to the steric
hindrance of the methyl group, which forces the methylpyri-
dine to coordinate in an axial position. A noticeable lowering
of pKa is also observed on increasing the size of the chelate
ring involving the pyridine nitrogen, as reflected by the differ-
ence in pKa between 2+ and 3+. Previously, it was found that
the nature of the chelate rings could influence the positioning
of the phenolate (the steric constraint at the copper center
differs on replacing the five-membered chelate rings by six-
membered ones), but the phenomenon is less clear. For
example the copper(II) complex of the 4-nitro-2-{[(2-pyridin-
2-yl-ethyl)-pyridin-2-ylmethyl-amino]-methyl}-phenolate ligand
exhibits an axially bound phenolate,33 in contrast with the
complex of the 2,4-di-tert-butyl-6-{N,N-bis[2-(2-pyrid-2-ylethyl]-
amino]-methyl}-phenolate28 and 4-nitro-2-{N,N-bis[2-(2-pyrid-2-
ylethyl]amino]-methyl}-phenolate ligands,35 where the pheno-
late coordinates in equatorial position. We will therefore limit
our discussion to 2+, 4+ and 6+ and interpret the lower pKas of
4+ and 6+, when compared to 2+, by a structural rearrangement
induced by deprotonation of the phenol moiety.32
The UV-Vis spectra of the copper complexes were recorded
after 10 min incubation with calf thymus DNA (concentrations
up to 500 μM in mole base pairs). On increasing the DNA
concentration a decrease in the intensity of the CT band is
observed for 12+, 2+ and 3+, while very little change was
observed for 4+ and 5+. Noteworthy, we did not observe a sig-
nificant red shift of the CT band. A plot of [DNA]/(εa − εf)
(where εa is the apparent molar extinction coefficient at a given
DNA concentration and εf is the molar extinction coefficient of
the free complex) as a function of DNA concentration gives a
straight line (Fig. 8) for 12+ and 2+ from which the intrinsic
binding constants Kb could be calculated (see the Experimen-
tal section). In the case of 3+ the absorbance change resulting
from DNA binding is smaller than that for 2+ and 3+;
Fig. 7 Evolution of the UV-Vis spectra as a function of pH of a 0.5 mM solution
of 2+ in a water–DMF 90 : 10 mixture ([NaClO4] = 0.1 M): T = 298 K, l =
1.000 cm. Arrows indicate spectral changes upon an increase of the pH from
3.54 to 8.84.
confirms that the coordinating CH3CN molecule (observed in
the X-ray crystal structure) is substituted in the water–DMF
medium (Scheme 2). The spectrum of 5+ resembles that of 4+,
although the CT band is blue-shifted by 32 nm. This behaviour
is easily explained by the weaker electron-donating ability of
the fluorine group, when compared to the methoxy substitu-
ent.24 More surprising, the phenolate-to-copper CT is observed
at 521 nm for 6+ while it appeared at 500 nm for 4+, which
exhibits a similar para phenolate substituent. Such a trend was
also reported by Shimazaki et al. for related copper com-
plexes.29 A possible explanation is a difference in the geometry
of the phenolate complexes, which affects the strength of the
Cu–O bond.29
We investigated the pH-dependence of the UV-Vis spectra
(Fig. 7) in the pH range 3.5–9 in a water–DMF (90 : 10) mixture
(+0.1 M NaClO4). Except for 12+ and 5+, which were found to
be insufficiently soluble at neutral pH in this medium, we
observed perfect agreement between the spectra at pH 7.3 in
the 0.1 M NaClO4 and the 0.05 M Tris media. This shows that
the buffer or supporting electrolyte does not interfere with
complexes 2+, 3+, 4+ and 6+. For these complexes an increase in
pH from 7 to 9 does not induce a significant shift of λmax, indi-
cating that the pKa of the coordinated water molecule is rela-
tively high (>10) or DMF binds instead. In contrast, a decrease
in pH down to 3.5 results in the disappearance of the
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8468–8483 | 8473

View Article Online
Paper
Dalton Transactions
consequently the resolution is poorer and the uncertainty with time, with a progressive decrease of the CT band. This
higher. Based on structural similarities between 12+, 2+ and 3+ suggests that significant DNA digestion occurs during the titra-
we assumed that the dependence of [DNA]/(εa − εf)] vs. [DNA] tion experiment, thus precluding refinement of the data to
was linear in the latter case and treat the data accordingly. The extract binding constants.
binding constants Kb (Table 4) determined from the spectral
change are 7770 1500, 2580 300 and 1940 550 M−1 for Electrochemical properties of the complexes
12+, 2+ and 3+, respectively. As expected, the Kb value of 12+ is
the highest, due to the lack of steric hindrance at the ortho
The electrochemical behaviour of the copper complexes was
investigated both at pH 7.3 (phenolate form) and 3.5 (phenol
form) by cyclic voltammetry in a water–DMF (90 : 10) medium
(+0.1 M NaClO4). The potentials are referenced vs. SCE
(Table 5).
At pH 7.3 all the compounds exhibit an oxidation wave in
the range 0.30–0.65 V (Table 5, Fig. 9). It is found to be irre-
versible for both 12+ and 5+ (Eap = 0.65 and 0.59 V respectively),
while it is reversible for 2+, 3+, 4+ and 6+, i.e. for complexes
involving an electron-rich and o,p-protected phenolate moiety
(E1/2 = 0.30, 0.32, 0.35, 0.36 V, respectively). This behaviour is
strong evidence that the redox process is ligand-centered,
affording phenoxyl radical species. Indeed, the potential
values measured for 2+, 3+, 4+ and 6+ are very similar to those
reported for 2+ in CH3CN (+0.38 V vs. SCE)24 and other related
position of the phenolate group. The Kb values are lower than
those of typical intercalators such as EthBr (Kb = 1.4 × 106 M−1
)
and partially intercalating complexes such as heteroleptic
[Cu(tdp)(dipyrido-[3,2-d2′,3′-f]-quinoxaline)]+ (9 × 105 M−1).17
They are typical of weak DNA binding interactions, as observed
in the [Cu(tdp)(bipyridine)]+ complex (Kb = 7100 M−1) and
derivatives17 and very close to that recently reported for the
copper complex of the [(2-pyridin-2-yl-ethyl)-pyridin-2-
ylmethyl-amino]-methyl ligand (roughly 104).36 The efficient
nuclease 6+ (see below) exhibits a dramatically different behav-
iour. Indeed, the spectra were found to evolve significantly
Fig. 8 Plot of the absorbance at 500 nm as a function of DNA concentration
(expressed in mole base pairs) for a 0.22 mM solution of 2+ in a water–DMF
90 : 10 mixture at pH 7.3 ([Tris] = 0.05 M, [NaCl] = 0.02 M). Inset: plot of [DNA]/
(εa − εf ) as a function of the DNA concentration (εa is the apparent molar extinc-
tion coefficient at a given DNA concentration, and εf is the molar extinction
coefficient of the free complex). T = 298 K, l = 1.000 cm.
Fig. 9 Cyclic voltammetry curves of 0.5 mM solutions of the phenolate com-
plexes in a water–DMF 90 : 10 mixture at pH 7.3: (a) 12+ in 0.05 M Tris contain-
ing 0.02 M NaCl, (b) 2+ and (c) 3+ in 0.1 M NaClO4. Scan rate = 0.1 V s−1
Potentials are given vs. the SCE reference.
.
Table 5 Electrochemical properties of the complexesa
pH = 3.5
pH = 7.3a
Complexes
Redb
Oxc
Redb
Ox
12+
2+
3+
4+
5+
6+
−0.35; −0.31
−0.35; −0.31
−0.20; −0.10
−0.22; −0.14
−0.23; −0.12
−0.06; 0.22
0.85
0.64
0.58
−0.40; −0.32d
−0.56; −0.24
−0.42; −0.29
−0.48; −0.24
−0.35; −0.17d
−0.35; −0.16
Eap = 0.65c,d
E11/2 = 0.30; Epa,2 = 0.65
E11/2 = 0.32; Epa,2 = 0.57
E11/2 = 0.35; Epa,2 = 0.63
Eap,1 = 0.59c,d
e
—
0.64
0.53
E11/2 = 0.36; Epa,2 = 0.56
a In water–DMF (90 : 10) + 0.1 M NaClO4 at 298 K, except 1 and 5 at high pH. All potentials are given vs. the SCE reference. b Epc and Eap are given.
c Irreversible processes, Eap is given. d In water–DMF (90 : 10) medium containing [Tris] = 0.05 M, [NaCl] = 0.02 M at pH 7.3 due to the low
solubility of the complexes in the water–DMF (90 : 10) + 0.1 M NaClO4 mixture at neutral pH. e Ill-defined wave.
8474 | Dalton Trans., 2013, 42, 8468–8483
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
complexes for which the copper(II)-phenoxyl radical form
was electrogenerated and unambiguously identified by spec-
troscopy. A slight increase in oxidation potential is observed
upon substitution of the pyridines by methylpyridines. The
decreased N-donating ability of the methylpyridine thus desta-
bilizes the radical form, as previously observed by Shimazaki
et al.29 When scanning towards the negative region of poten-
tials a metal-centered reduction peak Ecp is observed in the
potential range −0.35 to −0.56 V (Table 5). An enhancement of
the anodic peak current was observed in the reverse scan of
12+ and 2+,37 suggesting that the electron transfer involves
adsorbed copper(I)-phenolate material at nearly the same
potential as the “normal” electron transfer.38 As shown in
Table 4 the cathodic wave shifts to positive potentials in the
order 2+ < 4+ < 6+. The electron density of the copper center
thus decreases with an increase in the number of methylpyri-
dine groups in the ligand.39 A similar anodic shift of the
reduction wave is observed on going from 2+ to 3+ (substitution
of one pyridylmethyl by a longer pyridylethyl arm). The influ-
ence of the ligand arm’s length on the Cu(II)/Cu(I) redox poten-
tial in copper(II) tris(pyridyl)alkylamine complexes has been
investigated recently.40 In this case the difference in oxidation
Fig. 10 Agarose gel electrophoresis patterns of the cleavage reaction of
ϕX174 supercoiled DNA (20 μM base pairs) mediated by the copper complexes
in a phosphate buffer 10 mM, pH 7.2 (+10% DMF) at 37 °C for 1 h. Abbrevi-
ations: NC: nicked circular, L: Linear, SC: supercoiled. (a) Complex concentration
= 300 μM (constant); lane 0 is the DNA control without the copper complexes;
lane 1, complex 12+; lane 2, complex 2+; lane 3, complex 3+; lane 4, complex 4+;
lane 5, complex 5+; lane 6, complex 6+. (b, c, d): Dose dependent activity
(complex concentrations are indicated in μM): (b), 4+; (c), 5+; and (d), 6+.
potential originates from structural rather than electronic 300 μM. The reaction was conducted on ϕX174 DNA (20 μM, in
effects: Cu(II) favours five-membered coordination rings; there- base pairs) in phosphate buffer for 1 h at 37 °C (Fig. 10a).
fore compounds containing longer arms are stabilized in lower
In the absence of complex, DNA was found to be mainly in
oxidation states and the Cu(II)/Cu(I) redox couple increases. the supercoiled form (82%) both at pH 7.2 (Fig. 10a, lane 0)
Finally, the trend of Ecp 4 < 5 very likely reflects the lower elec- and 8.2. Cleavage efficiency of the complexes was next evalu-
tron-donating ability of the fluorine substituent, which ated by subtracting this starting percentage of DNA in SC
decreases the electron density at the metal.
form. In the presence of 300 μM of 12+, no significant change
At pH 3.5 the six compounds exhibit an irreversible oxi- in the proportion of the SC and cleaved forms of DNA was
dation wave above 0.53 V assigned to the oxidation of the observed (lane 1), indicating that this complex is not an
phenol moiety. It is well known that oxidation of phenols efficient nuclease.
occurs at potentials well above those required for oxidation of
Complexes 2+ and 3+ were found to exhibit a moderate DNA
the corresponding phenolates into phenoxyl radicals.41 The cleavage activity, as judged by the appearance of the nicked
metal-centered reduction wave is observed within the potential circular form in 15 and 22% ratio, respectively, at pH = 7.2
range −0.35 to +0.32 V. Again, an abnormally high Iap/Ipc ratio is (lanes 2 and 3, respectively). Similar moderate activities were
observed, suggesting that the reduced complexes easily adsorb reported for copper(II) complexes of bis(pyridyl)alkylamine
at the surface of the electrode. Interestingly, low Epc values were ligands that do not involve an intercalating moiety.42–44 The
observed for 12+ and 2+ (−0.31 V), while intermediate values better activity of 2+ and 3+, when compared to 12+, suggests
were measured for 3+, 4+ and 5+ (−0.20, −0.22, −0.23 respect- that a hydrophobic tert-butyl substituent in ortho position of
ively), and a relatively high potential (−0.06 V) was obtained the phenol enhances the nuclease activity. Since tert-butyl
for 6+. This trend follows that observed for the phenolate com- groups are expected to increase the steric protection and thus
pounds (see above) and thus will not be commented on further. perturb the access to the metal center, the enhanced activity
may result from either electronic effects, stronger hydrophobic
interactions of the compound with DNA and/or differences in
the complex geometry. Finally, we investigated the nuclease
Nuclease activity of the complexes
The ability of the copper(II) complexes to cleave DNA has been activity of these compounds at pH 8.2 since 2 was found to
investigated by gel electrophoresis using ϕX174 RF1 super- exist as a mixture of phenol and phenolate forms at physiologi-
coiled DNA in phosphate buffer (10 mM, pH 7.2 or 8.2 with cal pH. The extent of DNA damage was found to be essentially
10% DMF). Cleavage was evidenced by monitoring the conver- the same within the error margin at both pH (data not shown).
sion of the covalently closed circular supercoiled plasmid DNA Interestingly, the cleavage activity of a copper(II) complex of a
form (SC) into the nicked circular form (NC, single strand tripodal N3O ligand involving a 2-hydroxy-5-methyl-benzal-
breakage) and then to the linear form (L, double strand dehyde moiety was recently investigated by Neves et al. In con-
breakage).
trast with our results they did not detect nuclease activity at
The nuclease activity of the complexes was first investigated pH 7, but observed significant DNA cleavage at pH 8.16 It has
at pH 7.2 and 8.2 using a constant complex concentration of to be emphasized that there are also some examples in the
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8468–8483 | 8475

View Article Online
Paper
Dalton Transactions
literature of a drop in cleavage activity that results from an
increase in pH.45
Complexes involving methylpyridines exhibit a remarkable
nuclease activity: in the case of 4+ total disappearance of the
SC form was observed at pH 7.2 (lane 4) to the detriment of
the formation of 94% of NC and 6% of L forms. Thus 4+ pro-
motes both single strand and double strand DNA cleavage
without the need for external agents. For comparison, the
recently described copper nuclease [Cu(tdp)(dpq)]+ promotes
only single strand cleavage under comparable conditions,17
while the copper(II)-xylylguanidine TACN complex exhibits
a similar activity.46 The nuclease activity of 4+ is slightly lower
than that reported for the heteroleptic complex [Cu(Phen)-
(maltol)]+ 47 and other complexes involving intercalating
groups.18,45 Complexes 5+ and 6+ exhibit a slightly lower
activity (lanes 5 and 6 respectively) at 300 μM since the SC
DNA form was not fully converted to the NC form (17% and
29% of the SC form remains for 83% and 71% of the NC form
produced, for compounds 5+ and 6+, respectively), while
double strand cleavage (L form) was not observed. For the
latter compounds a moderate enhancement of the nuclease
activity was observed at higher pH (full digestion), thus sup-
porting a hydrolytic pathway (data not shown).
Fig. 12 Agarose gel electrophoresis patterns of supercoiled ϕX174 DNA
(20 μM base pairs) incubated with the copper complexes (300 μM; excepted in
lane 1) in a phosphate buffer 10 mM pH 7.2 (+10% DMF) at 37 °C for 1 h; (a)
4+, (b) 5+, (c) 6+. Lane 1, DNA control; lane 2, DNA + complex (300 μM) without
scavenger; lanes 3–11, DNA + complex (300 μM) in the presence of agents: lane
3, NaN3 (100 μM); lane 4, Superoxide Dismutase (0.5 unit); lane 5, Catalase (0.1
unit); lane 6, DMSO (2 μL); lane 7, EtOH; lane 8, mercaptoethanol (0.71 mM);
lane 9, Hoechst 33258 (100 μM); lane 10, NaCl (350 μM); lane 11, EDTA
(10 mM).
significant amounts of double strand breakage could be
evidenced for 500 μM concentrations of 5+ and 6+.
Examination of the time course for the reaction (Fig. 11a–c)
reveals that more than 80% of plasmid is cleaved after two
hours for 4+ and 6+ at 300 μM. The rate constants were found
For the most active compounds 4+, 5+ and 6+ we investi-
gated the dependence of the DNA cleavage on both the concen-
tration of the copper complexes (Fig. 10b–d) and the reaction
time at a given complex concentration (300 μM, Fig. 11a–c).
Not surprisingly, the higher the concentration, the greater the
yield in cleavage product (Fig. 10b–d). For 5+ and 6+ the NC
form could be detected for complex concentrations within the
30–50 μM range, while >90% cleavage is observed in the
300–500 μM range. There is a progressive increase of the reac-
tion yield over a wide concentration range for 5+ and 6+, as
shown in Fig. 10b–d. In contrast, 4+ exhibits a dramatic
enhancement of cleavage activity from 100 μM (no significant
activity) to 300 μM (full digestion of the SC form). Such a dose-
dependent drop in activity is rather uncommon, but it has
been recently reported for the copper(II) complex of a bis
(pyridyl)alkyl-acridine conjugate.48
to be 0.49
0.05 min−1, 0.09
0.01 min−1 and 0.50
0.05 min−1 for 4+, 5+ and 6+, respectively. While 4+ and 6+
exhibit comparable rate constants, 5+, which exhibits a less
electron-donating fluorine substituent, was found to react
much slower with DNA.
In order to get insight into the reaction mechanism we evalu-
ated the effect of various scavengers and exogenous agents on
the cleavage activity (Fig. 12). For all the complexes the
cleavage is fully inhibited by the addition of EDTA (lane 11),
showing that the metal ion is essential for the activity. For the
three compounds neither NaN3 (singlet oxygen scavenger),
superoxide dismutase (superoxide scavenger) nor Calatase
(H2O2 scavenger) inhibit cleavage (lanes 3, 4, 5), indicating
that ROS are not responsible for the activity. DMSO and EtOH
also did not affect the strand scission by 4+ (lanes 6, 7),
showing that the freely diffusible OH˙ species is not respon-
sible for the cleavage. Surprisingly, DMSO was found to
enhance the cleavage activity in the case of 5+, while both
DMSO and EtOH improve cleavage by 6+. We interpret these
unexpected results by a higher solubilization of the complexes
in the presence of these co-solvents.
For 4+, 5+ and 6+ 50% cleavage of the SC form is observed
for concentrations of 120
50 μM. It is noteworthy that
The addition of a reductant (mercaptoethanol) in the
medium enhances DNA cleavage in the case of 6+ (lane 8).
This result may be understood by considering an alternative
oxidative cleavage pathway,17,18 which is initiated by reduction
of the copper(II) into copper(I). This fact is consistent with elec-
trochemical measurements, which showed that 6+ exhibits the
highest Cu(II)/Cu(I) redox potential. Mercaptoethanol slightly
inhibits the activity of 4+, while it does not affect the cleavage
by 5+. The main cleavage pathway, thus, likely remains hydro-
lytic for both complexes.
Fig. 11 Agarose gel electrophoresis patterns of supercoiled ϕX174 DNA
(20 μM base pairs) incubated for different times at 37 °C with 300 μM of the
copper complexes in a phosphate buffer 10 mM pH 7.2 (+10% DMF): (a) 4+, (b)
5+, (c) 6+.
8476 | Dalton Trans., 2013, 42, 8468–8483
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 6 Anti-proliferative activity of the complexesa
IC50 (μM)
In the presence of a minor groove binder (Hoechst 33258)
the cleavage activity of 4+ and 5+ was significantly inhibited
(lane 9). This experiment shows that both complexes bind to
DNA in the minor groove. In addition, excess of sodium chlor-
ide does not inhibit the cleavage reaction (lane 10), which
rules out a simple electrostatic interaction of copper complexes
4+ and 5+ with DNA.
Similarly to previous studies we performed religation assays
with T4 ligase on the ϕX174 DNA plasmid after incubation
(1 h) with the three most active complexes 4+, 5+, 6+ (300 μM)
or restriction nuclease Xho I (blank).22 For complexes 4+, 5+, 6+
the initial band corresponding to the NC form of DNA is still
present after religation, which would at first glance suggest an
oxidative rather than hydrolytic cleavage. Nevertheless, the
latter results should be used with special care since DNA
hydrolysis was also assumed to be responsible for other DNA
damage which could not be enzymatically religated. For
instance hydrolytic cleavage products may not possess strictly
matched ends suitable for religation.50 In addition, the DNA
phosphate group could remain bound to the complex after
hydrolysis, and thus is inaccessible for the ligase.15 Because
the relevance of the latter experiment is not particularly
strong, we still entertain the hydrolytic mechanism as the
most serious option.
Complexes
RT112
RT112-CP
RF
Cisplatin
9.1 1.2
99.1 5.8
3.1 0.2
20.8 2.5
6.3 0.6
3.7 0.2
3.9 0.1
23.8 1.2
ND
5.5 0.4
ND
5.2 0.4
5.0 0.3
4.7 0.5
2.6
ND
1.8
ND
0.8
1.4
1.2
12+
2+
3+
4+
5+
6+
a The concentration resulting in 50% loss of cell viability relative to
untreated cells (IC50) was determined from dose–response curves.
Results represent the means SD of three independent experiments.
RF is the relative ratio between IC50 of the cisplatin-resistant cells
against IC50 of the sensitive cell lines. ND: not determined.
to replace cisplatin in patients where cisplatin therapy failed
due to chemoresistance, the four more active complexes 2+, 4+,
5+ and 6+ were tested on a bladder cancer cell line resistant to
cisplatin (RT112-CP).57 As shown in Table 6, RT112-CP cells
are 2.6 times more resistant to cisplatin than the RT112
parental cell line. Interestingly, all four tested complexes (com-
pounds 2+, 4+, 5+ and 6+) overcame the resistance observed
with cisplatin as expressed by the resistance factors (RF),22
which are lower for our complexes compared to cisplatin. Com-
pared to the copper pyrimol complex (the structure shown in
Fig. 1d)22 the RF values are much lower and even close to 1 for
4+, 5+ and 6+. So, these complexes have not only higher anti-
tumour activity than cisplatin against a bladder cancer cell line,
but also they are very powerful in circumventing the cellular
resistance of RT112-CP cells.
Anti-proliferative activity against bladder tumor cells
Some anticancer agents approved for clinical uses trigger cell
death by damaging DNA. For example, platinum-containing
drugs cause crosslinking of DNA,51 Melphalan and Chloram-
bucil are alkylating agents52,53 while Bleomycin acts by produ-
cing single-strand and double-strands breaks in DNA.54 Since
our molecules induced DNA cleavage, we have carried out
in vitro assays to evaluate the cytotoxic and anti-proliferative
effects of the complexes 12+–6+.
Effect of complexes on the cell cycle progression
In order to identify more potent and safe molecules for the In order to further characterize the mechanism of action of
treatment of bladder cancer, the biological activity of our complexes 2+, 4+, 5+ and 6+, cell cycle analysis was performed
copper complexes was studied on the RT112 cell line, derived to determine the effect of these molecules on cell cycle pro-
from a human bladder tumor. Cells were incubated with gression. RT112 cells were incubated with vehicle alone
various concentrations of the complexes 12+–6+ and their pro- (DMSO) as the control, or with IC50 concentrations of com-
liferation was monitored by a MTT assay.55 Our results indicate plexes. After 24 or 48 h, cells in suspension in the medium
that it was necessary to treat cells for at least 48 h to observe a and attached cells were harvested and analyzed by flow cyto-
massive cell death. The anti-proliferative activities of tested metry. As shown in Fig. 13A, after 24 h the majority of control
compounds after 48 h of treatment, expressed as IC50 values cells (55%) are in G0/G1 phases, whereas the rest of the cells
(the concentration resulting in 50% loss of cell viability relative are approximately equally distributed between the S and
to untreated cells), are summarized in Table 6. Cisplatin, G2 + M phases (26% and 19% respectively). A 24 h treatment
which is currently used in the management of advanced of RT112 cells with complex 2+ results in a significant accumu-
bladder cancer, was applied as a test referential agent.56 lation of cells in the G0/G1 phase (73%), with a concomitant
Complex 12+ presented a low cytotoxic activity against RT112 loss from the S phase (9%). This G0/G1 phase accumulation
cells (IC50 = 99.1 µM), which could be correlated to its inability increases with incubation time, since it reaches 82% after 48 h
to cleave DNA. For complexes 2+ and 3+ there seems to be no against 70% in the control (Fig. 13B). The results obtained
direct correlation between their moderate nuclease activity and with the other complexes (4+, 5+ and 6+) are different. After
their cytotoxicity, since IC50 of complex 3+ is 6.7 times more 24 h of incubation with the IC50 concentration of the mole-
than that of complex 2+. Finally, complexes 4+, 5+ and 6+, cules, the G2 + M accumulation of cells increases from 19% in
which have higher DNA cleavage activity, were about 1.4–2.4 the control to 39% with complex 4+, 34% with complex 5+ and
times more active than cisplatin against the RT112 cell line. In 44% with complex 6+ (Fig. 13A). This is accompanied by a
order to determine if our complexes could be valuable agents decrease of the number of cells in the remaining phases of the
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8468–8483 | 8477

View Article Online
Paper
Dalton Transactions
changes in DNA binding constants do not explain directly the
trend of cleavage activity in the series since 3+ exhibits a lower
affinity for DNA than 12+, although it is more active.
A hydrolytic pathway is often proposed for copper-catalyzed
DNA cleavage in the absence of a reductant or oxidants.
Recently, oxidative pathways were proposed to account for DNA
cleavage by some copper(II) complexes involving sterically
hindered phenolate or aminophenolate coordinating groups
(Fig. 1d). They imply the formation of a reactive phenoxyl
radical by aerobic oxidation of the coordinated phenolate. The
phenoxyl/phenolate potential for the complex shown in Fig. 1d
is +1.11 V vs. Ag/AgCl (roughly +1.07 V vs. SCE), i.e. much
higher than the compounds investigated there.22 Therefore,
from a thermodynamic point of view 12+–6+ might form phe-
noxyl radicals more easily than Reedijk’s complex, rendering
an oxidative mechanism promoted by phenoxyl species plaus-
ible.22 Nevertheless, large differences in chemical stabilities
for the radical species (estimated via the Iap/Ipc ratio of the phe-
noxyl/phenolate redox couple) were observed within our series,
without any evident correlation between DNA cleavage activity
and phenoxyl radical stabilities. In addition, attempts to trap a
phenoxyl radical intermediate by UV-Vis spectroscopy upon
reaction with DNA (phenoxyl radicals exhibit an intense π–π*
transition at around 420 nm)24 were unsuccessful. These
experimental results thus argue against the involvement of a
phenoxyl radical as reactive species for DNA cleavage and
rather support a hydrolytic pathway. Hydrolytic DNA cleavage
may be initiated by substitution of the labile coordinated
solvent by a phosphate group of the DNA backbone, which is
activated by the Lewis acidity of the copper ion. Methylpyridine
N-donors could facilitate this activation and subsequent
nucleophilic attack. A detailed investigation on the cleavage
mechanism is currently under way in our laboratories.
Fig. 13 Cell cycle distribution of RT112 cells treated with complexes 2+, 4+, 5+
and 6+. RT112 cells were treated with vehicle alone (Control, CT) or with the
IC50 value of each compound for 24 h (A) or 48 h (B). Cells were then fixed,
stained with PI and cell cycle was analyzed by flow cytometry. Each value is the
mean SD of three independent determinations.
cell cycle. This G2 + M accumulation is sustained after 48 h of
incubation with complex 4+ (36% against 14% in the control)
and even increased with complexes 5+ and 6+, which led to 43
and 48% respectively of cells in G2 + M phases (Fig. 13B).
Complexes 4+, 5+ and 6+ thus clearly alter the cell cycle pro-
gression. This is most likely owing to DNA damage which
results in the activation of cell cycle checkpoints in dividing
cells. If cells undergo DNA damage in G2, the G2/M check-
point is activated and stops the cell cycle to prevent cells from
starting mitosis.58,59 So, complexes 2+, 4+, 5+ and 6+ promote
cell cycle arrest in G2 + M phases, which probably reflects their
activity on DNA.
The anti-proliferative activity of the four more active
nucleases 2+, 4+, 5+ and 6+ was tested against RT112 bladder
cancer cell lines. The four investigated compounds exhibit IC50
values lower than cisplatin (3.1–6.3 μM vs. 9.1 μM). They are
also able to circumvent the cellular resistance of the RT112CP
cell line, as reflected by RF up to three times lower for 2+, 4+,
5+ and 6+ than for cisplatin. Finally, we show here that the
most active compounds alter cell cycle progression, which
probably reflects their activity on DNA. Further analysis will be
conducted to determine molecular alterations induced by our
complexes and to understand how they could block the cell
Conclusion
The present results point to a very simple way to dramatically cycle at different levels.
enhance the nuclease activity, in the absence of an exogenous
agent, of copper-phenolate complexes involving tripodal
ligands: when at least one pyridine is α-methylated the yield of
the reaction is significantly improved, with 75 to 95% DNA
digestion after 2 hours for complex concentrations of 300 μM
Experimental section
Materials and methods
(vs. <22% for complexes involving exclusively non-methylated All chemicals were of reagent grade and were used without
pyridine donors). We hypothesize that this result is correlated purification. NMR spectra were recorded on a Bruker AM 300
to the weaker N-donating ability of methylpyridines, which (1H at 300 MHz) spectrometer. Chemical shifts are quoted rela-
facilitates exchange of the coordinated exogenous ligand and tive to tetramethylsilane (TMS). Mass spectra were recorded on
modifies significantly the structure of the phenolate complex a Thermofinningen (EI/DCI) or a Bruker Esquire ESI-MS appar-
(equatorial positioning of the phenolate moiety). Noteworthy, atus. For pKa determinations UV/Vis spectra were recorded on
8478 | Dalton Trans., 2013, 42, 8468–8483
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
a Cary Varian 50 spectrophotometer equipped with a Hellma NaCl solution and dried over Na2SO4. The solution was evapor-
immersion probe (1.000 cm path length). The temperature in ated under vacuum to afford yellow oil. Column chromato-
the cell was controlled using a Lauda M3 circulating bath and graphy on silica gel (ethyl acetate–pentane (1/5) yielded 3L
1
the pH was monitored using a Methrom 716 DMS Titrino (263 mg, 65%) as a yellow solid. H NMR (300 MHz, CDCl3): δ
apparatus. A least square fit of the titration data was realized = 1.40 (9H, s), 2.97–3.09 (4H, m), 3.73 (3H, s), 3.83 (2H, s), 3.85
with SPECFIT software. For UV/Vis measurements at neutral (2H, s), 7.03–7.20 (4H, m), 7.52 (1H, td, 3J = 7.6 Hz, 4J =
3
4
3
pH (buffered media) a Perkin Elmer Lambda 2 spectropho- 1.9 Hz), 7.57 (1H, td, J = 7.6 Hz, J = 1.8 Hz), 8.47 (1H, d, J =
tometer was used. X-band EPR spectra were recorded on either 4.8 Hz), 8.52 (1H, d, 3J = 5.9 Hz). 13C NMR (75 MHz, d6-DMSO):
a Bruker ESP 300E or a Bruker EMX Plus spectrometer δ = 29.8 (3 CH3), 35.2 (C), 35.6 (CH2), 54.0 (CH2), 56.0 (CH3),
equipped with a Bruker nitrogen flow cryostat. Spectra were 59.0 (CH2), 59.8 (CH2), 112.0 (CH), 113.3 (CH), 121.7 (CH),
simulated using Bruker SIMFONIA software. Electrochemical 122.6 (CH), 123.1 (C), 123.7 (CH), 123.9 (CH), 136.8 (CH), 136.9
measurements were carried out using a CHI 620 potentiostat. (CH), 138.3 (C), 149.4 (CH), 149.6 (CH), 150.7 (C), 151.8 (C),
Experiments were performed in a standard three-electrode cell 152.1 (C), 158.2 (C), 158.4 (C), 160.0 (C). MS (DCI, NH3/
under an argon atmosphere. A glassy carbon disc electrode isobutane): m/z (%) 406 (100) [M + H]+. M.p. 70 °C. Anal. Calcd
(3 mm diameter) or a platinum disc electrode (1 mm dia- for C25H31N3O2: C, 74.04; H, 7.70; N, 10.36. Found: C, 73.82;
meter), which was polished with 1 mm diamond paste, was H, 7.75; N, 10.31%.
used as the working electrode. An SCE was used as the
reference.
4L. 2-tert-Butyl-4-methoxy-phenol (1.8 g, 10 mmol),
(6-methyl-pyridin-2-ylmethyl)-pyridin-2-ylmethyl-amine (2.13 g,
10 mmol)29 and formaldehyde (0.75 ml of a 37% aqueous solu-
tion) in ethanol–H2O (4 : 6, 10 ml) were refluxed for 24 h. The
Crystal structure analysis
The collected reflections were corrected for absorption reaction mixture was then extracted with dichloromethane,
(SADABS) and solved by direct methods and refined with Olex washed with a saturated NaCl solution and dried over Na2SO4.
software.60 All non-hydrogen atoms were refined with anisotro- The solution was evaporated under vacuum to afford orange
pic thermal parameters. Hydrogen atoms were generated in oil. Column chromatography on silica gel (ethyl acetate–
idealized positions, riding on the carrier atoms, with isotropic pentane (1/1)) yielded 4L (1.82 g, 45%) as a pale pink oil.
thermal parameters.
1H NMR (200 MHz, CDCl3): δ = 1.45 (9H, s), 2.58 (3H, s), 3.73
4
(3H, s), 3.76 (2H, s), 3.83 (2H, s), 3.85 (2H, s), 6.49 (1H, d, J =
Preparation of the ligands and complexes
3.0 Hz), 6.80 (1H, d, 4J = 3.0 Hz), 7.00 (2H, d, 3J = 7.5 Hz),
1L. 4-Methoxyphenol (1.24 g, 10 mmol), bis-pyridin-2- 7.00–7.16 (2H, m), 7.38 (2H, d, 3J = 7.8 Hz), 7.50 (1H, t, 3J =
ylmethyl-amine (1.99 g, 10 mmol) and formaldehyde (0.75 ml 7.54 Hz), 7.61 (1H, td, 3J = 7.8 Hz, 4J = 1.7 Hz), 8.52 (2H, d, 3J =
of a 37% aqueous solution) were refluxed for 24 h in ethanol– 5.1 Hz), 10.40 (1H, s, broad). 13C NMR (75 MHz, CDCl3): δ =
H2O (1 : 1, 12 ml). The reaction mixture was then extracted 24.5 (CH3), 29.9 (3 CH3), 35.4 (C), 56.3 (CH3), 58.2 (CH2), 59.7
with dichloromethane, washed with a saturated NaCl solution (2 CH2), 112.7 (CH), 1313.4 (CH), 120.5 (CH), 122.1 (CH), 122.5
and dried over Na2SO4. The solution was evaporated under (CH), 123.6 (C), 123.9 (CH), 136.9 (CH), 137.1 (CH), 138.2 (C),
vacuum to afford brown oil. Column chromatography on silica 149.3 (CH), 150.8 (C), 151.8 (C), 157.8 (C), 158.4 (C), 158.8 (C).
gel with CH2Cl2–methanol (10/0 to 5/5) as the eluent yielded MS (DCI, NH3/isobutane): m/z (%) 406 (100) [M + H]+. Anal.
1L (0.67 g, 20%) as a yellow oil. 1H NMR (400 MHz, CDCl3): Calcd for C25H31N3O2: C, 74.04; H, 7.70; N, 10.36. Found:
4
δ = 3.74 (2H, s), 3.77 (3H, s), 3.87 (4H, s), 6.64 (1H, d, J = 3.0 C, 72.26; H, 7.63; N, 10.36%.
Hz), 6.74 (1H, dd, 3J = 8.8 Hz, 4J = 3.0 Hz), 6.83 (1H, d, 3J =
5L. This ligand was prepared in a similar way as 4H, by
3
3
8.8 Hz), 7.16 (2H, dd, J1 = 7.3 Hz, J2 = 5.1 Hz), 7.34 (2H, d, using the 2-tert-butyl-4-fluoro-phenol instead of the 2-tert-
3
4
3J = 7.9 Hz), 7.62 (2H, td, J = 7.8 Hz, J = 1.8 Hz), 8.57 (2H, d, butyl-4-methoxy-phenol. Column chromatography on silica gel
3J = 4.9 Hz). 13C NMR (100 MHz, CDCl3): δ = 55.9 (CH3), 57.1 (ethyl acetate–pentane (2/8)) yielded 5L (40%) as a pale yellow
1
(CH2, 59.1 (2 CH2), 114.0 (CH), 115.9 (CH), 116.9 (CH), 122.9 oil. H NMR (400 MHz, CDCl3): δ = 1.44 (9H, s), 2.59 (3H, s),
(CH), 123.3 (CH), 123.4 (C), 136.8 (CH), 149.0 (CH), 151.4 (C), 3.74 (2H, s), 3.83 (2H, s), 3.84 (2H, s), 6.63 (1H, dd, 3J = 8.1 Hz,
152.3 (C), 158.2 (C). MS (DCI, NH3/isobutane): m/z (%) 336.2 4J = 3.1 Hz), 6.90 (1H, dd, 3J = 11.1 Hz, 4J = 3.1 Hz), 7.01 (1H, d,
(100) [M + H]+. M.p. 70 °C. Anal. Calcd for C21H25N3O2: 3J = 7.6 Hz), 7.07 (1H, d, J = 7.6 Hz), 7.13 (1H, dd, J = 7.0 Hz,
3
3
C, 71.77; H, 7.17; N, 11.96. Found: C, 71.73; H, 6.45; N, 4J = 5.0 Hz), 7.36 (1H, d, J = 7.8 Hz), 7.50 (1H, t, J = 7.8 Hz),
3
3
3
4
3
12.42%.
7.60 (1H, td, J = 7.8 Hz, J = 1.8 Hz), 8.53 (1H, d, J = 4.7 Hz),
3L. 3-tert-Butyl-2-hydroxy-5-methoxy-benzaldehyde (208 mg, 10.70 (1H, s, broad). 13C NMR (100 MHz, CDCl3): δ = 24.4
1
mmol), (2-pyridin-2-yl-ethyl)-pyridin-2-ylmethyl-amine (CH3), 29.7 (3 CH3), 35.3 (C), 57.6 (CH2), 59.6 (CH2), 59.8
(213 mg, 1 mmol)35 and glacial acetic acid (0.2 ml) were dis- (CH2), 113.1 (CH), 114.1 (CH), 120.4 (CH), 122.1 (CH), 122.5
solved in ethanol (20 ml). After 30 min stirring, 62 mg of (CH), 123.8 (CH), 124.0 (C), 136.9 (CH), 137.2 (CH), 138.6 (C),
sodium cyanoborohydride (62 mg, 1 mmol) were slowly added 149.3 (CH), 156.9 (C), 157.5 (C), 158.4 (C), 158.6 (C). MS (DCI,
(over 6 hours). The mixture was stirred for 12 hours and then NH3/isobutane): m/z (%) 394 (100) [M + H]+. Anal. Calcd for
neutralized with concentrated HCl (2 M). The reaction mixture C24H28FN3O: C, 73.26; H, 7.17; N, 10.68. Found: C, 74.01; H,
was extracted with dichloromethane, washed with a saturated 7.22; N, 10.52%.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8468–8483 | 8479

View Article Online
Paper
Dalton Transactions
6L. This ligand was prepared in a similar way as 4L, by Determination of the DNA binding constants
using bis-(6-methyl-pyridin-2-ylmethyl)-amine61 instead of
(6-methyl-pyridin-2-ylmethyl)-pyridin-2-ylmethyl-amine. Column
chromatography on silica gel (ethyl acetate–pentane (2/8))
yielded 6L (40%) as a pale yellow oil. 1H NMR (400 MHz,
CDCl3): δ = 1.45 (9H, s), 2.56 (6H, s), 3.73 (3H, s), 3.74 (2H, s),
The DNA binding constants were obtained from titration of
solutions of the complexes with CT DNA. DNA (type I,
fibers, from Sigma Aldrich) was first purified by four succes-
sive extractions with the phenol–CHCl3–isoamyl alcohol
25–24–1 mixture. Then 0.1 volume (with respect to the
aqueous phase) of a 3 M sodium acetate solution and 2
volumes of ethanol were added and the solution was stored
overnight at 253 K. The precipitated DNA was solubilised and
analyzed by UV-Vis to ensure that the A260/A280 ratio was ≥2. In
a typical titration experiment 50 μl of a stock solution of the
complex in DMF (0.5–1 mM) was mixed with 20 μl of phos-
phate buffer (50 mM) at pH 7.3, and 430 μl of an aqueous solu-
tion of CT DNA (concentration ranging from 0 to 0.5 mM) was
added. After the equilibrium was reached (10 min) at 298 K
the UV-Vis spectrum of the solution was recorded. The DNA
binding constants Kb were calculated using the equation:
4
4
3.82 (4H, s), 6.48 (1H, d, J = 3.1 Hz), 6.79 (1H, d, J = 3.2 Hz),
3
3
7.00 (2H, d, J = 7.6 Hz), 7.15 (2H, d, J = 7.6 Hz), 7.49 (1H, t,
3J = 7.9 Hz), 10.40 (1H, s, broad). 13C NMR (100 MHz, CDCl3):
δ = 24.5 (CH3), 29.7 (3 CH3), 35.2 (C), 56.0 (CH3), 58.0 (CH2),
59.7 (2 CH2), 112.6 (CH), 113.3 (CH), 120.4 (CH), 121.9 (CH),
123.6 (C), 137.0 (CH), 138.0 (C), 150.7 (C), 151.8 (C), 157.9 (C),
158.0 (C). MS (DCI, NH3/isobutane): m/z (%) 420 (100)
[M + H]+. Anal. Calcd for C26H33N3O2: C, 74.43; H, 7.93; N,
10.02. Found: C, 74.36; H, 7.81; N, 10.06%.
1·(ClO4)2. 1L (54.6 mg, 0.163 mmol) was dissolved in
methanol (3 mL) and then Cu(ClO4)2·6H2O (60.3 mg,
0.163 mmol in a methanolic solution) and NEt3 (22.8 μl,
0.163 mmol ml) were added. The solution was stirred for
10 min at room temperature, left to evaporate overnight, and
½DNAꢀ=ðεa ꢁ εf Þ ¼ ½DNAꢀ=ðεb ꢁ εf Þ þ 1=Kb
then stored for 10 h at 253 K. A dark brown precipitate of where εa is the apparent molar extinction coefficient of the
1·(ClO4)2 was formed, which was collected by filtration (50 mg, complex at a given DNA concentration, εf is the molar extinc-
31% yield). Slow diffusion of diethyl ether into a concentrated tion coefficient of the free complex, and εb is the molar extinc-
CH3CN solution of the complex affords brown single crystals tion coefficient of fully bound complex. A plot of [DNA]/(εa −
of 1·(ClO4)2·(CH3CN)4. MS (ESI, CH3CN): m/z (%) 396 (100) εf) as a function of DNA concentration gives a straight line
[M0.5 − 2ClO4]+. Anal. Calcd for C40H40Cl2Cu2N6O12: C, 48.30; whose slope is 1/(εb − εf) and the Y-intercept is equal to
H, 4.05; N, 8.45. Found: C, 48.52; H, 4.03; N, 8.35. UV-Vis (λmax 1/Kb(εb − εf). Each binding constant was obtained from a tripli-
[nm] [ε/M−1 cm−1] in CH3CN, 298 K): 466 [1015], 853 [670].
3·ClO4. 3L (170 mg, 0.419 mmol) was dissolved in CH3CN
(3 mL) and then Cu(ClO4)2·6H2O (156 mg, 0.412 mmol) was
added. The solution was stirred for 30 min. Slow diffusion of
diisopropyl ether into the resulting solution affords blue
cate experiment.
Procedure for DNA cleavage experiments
single crystals of 3·ClO4 (237 mg, 80% yield). MS (ESI, CH3CN): ϕX174 RF1 DNA was purchased from Fermentas and was
m/z (%) 467 (100) [M − H − 2ClO4 − CH3CN]+. Anal. Calcd for stored at −20 °C. The typical reaction mixture, containing
C27H34Cl2CuN4O10: C, 45.74; H, 4.83; N, 7.90. Found: C, 45.82; double-stranded DNA and the CuII complexes in a 10 mM
H, 4.83; N, 7.90. UV-Vis (λmax [nm] [ε/M−1 cm−1] in CH3CN, phosphate solution (pH 7.2 or 8.2) with 10% DMF, was incu-
298 K): 573 [230].
bated at 37 °C for the required time, with or without additives.
Complexes 4·ClO4 and 5·ClO4 were prepared in an identical After the incubation period, the reaction was quenched at
manner in 80 and 63% yield respectively. 6+ was prepared −20 °C, followed by the addition of loading buffer (6× loading
in situ in its phenolate form by mixing equimolar amounts of dye solution, Fermentas). The reaction mixture was loaded on
6L and Cu(ClO4)2·6H2O in DMF.
a 0.8% agarose gel in Tris–Boric acid–EDTA buffer (pH 8.2)
4·ClO4. MS (ESI, CH3CN): m/z (%) 467 (100) [M − H − 2ClO4 (0.5× TBE) and electrophoresis was performed at 70 V for 2 h
− CH3CN]+. Anal. Calcd for C27H34Cl2CuN4O10: C, 45.74; H, 30 min–3 h. After DNA migration, the gels were stained by
4.83; N, 7.90. Found: C, 45.50; H, 4.77; N, 7.86. UV-Vis (λmax incubation for 10 min with a 1 μg ml−1 ethidium bromide (EB)
[nm] [ε/M−1 cm−1] in CH3CN, 298 K): 590 [210].
solution and then washed with distilled water. The gels were
5·ClO4. MS (ESI, CH3CN): m/z (%) 455 (100) [M − H − 2ClO4 visualised and the fluorescence was quantified using Imager
− CH3CN]+. Anal. Calcd for C26H31Cl2CuFN4O9: C, 44.80; H, Typhoon 9400 and Image Quant software. The cleavage
4.48; N, 8.04. Found: C, 44.51; H, 4.39; N, 7.90. UV-Vis (λmax efficiency was measured by determining the ability of the
[nm] [ε/M−1 cm−1] in CH3CN, 298 K): 622 [100].
complex to convert the supercoiled DNA (SC) to nicked circular
6+ in DMF solution (phenolate cationic form with a coordi- form (NC) and linear form (L). For kinetic analysis we deter-
nated DMF molecule). MS (ESI, DMF–H2O): m/z (%) 481.2 (30) mined the amounts of SC and NC forms on the agarose gel by
[M − DMF]+, 482.2 (70) [M + H − DMF]+ (this copper(I) form of image processing with imageJ 1.46r freeware (Wayne Rasband
the complex results from the reduction of 6+ in the gas National Institute of Health, USA). The data were then refined
phase).62 EPR (100 K, DMF): gxx = gyy = 2.057, gzz = 2.265, Axx (Cu)
=
by simplex algorithm by considering a mono-exponential
Ayy
= 1.5 mT, Azz
(Cu)
= 15.7 mT; UV-Vis (λmax [nm] function (commercial Bio-kine 4.44 software, Biologic Co., Le
(Cu)
[ε/M−1 cm−1] in DMF, 298 K): 543 [975].
Pont-de-Claix, France).
8480 | Dalton Trans., 2013, 42, 8468–8483
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Enzymatic religation assays
585 44 nm filter. Parameters from 2 × 104 cells were acquired
using Cell Quest Pro software (Becton Dickinson). The percen-
tage of cell cycle distribution in the G1, S and G2 + M phases
was determined using FCS Express 3 software (De Novo
Sofware, Los Angeles, CA).
The supercoiled DNA was treated with the copper complexes at
37 °C and pH 7.2. DNA samples were purified over QIAquick
PCR purification columns (Qiagen) and were used for religa-
tion experiments. Ligation was performed in 25 μL with 2.5 μL
of 10× ligation buffer and 2 units of T4 ligase (Fermentas) for
16 h at 16 °C. Similarly, and for comparison, the supercoiled
DNA was digested using XhoI restriction enzyme (Fermentas) Acknowledgements
and religated using the above procedure, and found 100%
religation.
We thank the Labex Arcane (Université Joseph Fourier,
Grenoble) and the Association Espoir and GEFLUC (Délégation
de l’Isère) for financial support.
Cell culture
Human bladder cancer cell line RT112 was obtained from Cell
Lines Service (Eppelheim, Germany). Cisplatin resistant RT112
cells (RT112-CP) were kindly provided by B. Köberle (Institute
of Toxicology, Clinical Centre of University of Mainz, Mainz,
Germany). RT112 and RT112-CP cells were cultured in RPMI
1640 medium supplemented with 10% (v/v) fetal calf serum
(FCS) and 2 mM glutamine (Invitrogen Life Technologies,
Paisley, UK). Cells were maintained at 37 °C in a 5% CO2-
humidified atmosphere and tested to ensure freedom from
mycoplasma contamination.
Notes and references
1 G. J. Brewer, Exp. Biol. Med., 2001, 226, 665.
2 T. Theophanides and J. Anastassopoulou, Crit. Rev. Oncol.
Hematol., 2002, 42, 57; X. Zhang, C. Bi, Y. Fan, Q. Cui,
D. Chen, Y. Xiao and Q. Ping Dou, Int. J. Mol. Med., 2008,
22, 677; C. Marzano, M. Pellei, F. Tisano and C. Santini,
Anti-Cancer Med. Chem., 2009, 9, 185; S. Tardito and
L. Marchio, Curr. Med. Chem., 2009, 16, 1325.
3 A. Chan, F. Wong and M. Arumanayagam, Ann. Clin.
Biochem., 1993, 30, 545; M. Zowczak, M. Iskra, L. Torlinski
and S. Cofta, Biol. Trace Elem. Res., 2001, 82, 1.
4 H. Mazdak, F. Yazdekhasti, A. Movahedian, N. Mirkheshti
and M. Shafieian, Int. Urol. Nephrol., 2010, 42, 89.
5 S. Majumder, S. Chatterjee, S. Pal, J. Biswas, T. Efferth and
S. K. Choudhuri, Biometals, 2009, 22, 377.
6 S. Adsule, V. Barve, D. Chen, F. Ahmed, Q. P. Dou,
S. Padhye and F. H. Sarkar, J. Med. Chem., 2006, 49, 7242.
7 E. Wong and C. M. Giandomenico, Chem. Rev., 1999, 99,
2451; E. R. Jamieson and S. J. Lippard, Chem. Rev., 1999,
99, 2467; J. Reedijk, Eur. J. Inorg. Chem., 2009, 1303.
8 A. M. Angeles-Boza, P. M. Bradley, P. K. L. Fu, S. E. Wicke,
J. Bacsa, K. M. Dunbar and C. Turro, Inorg. Chem., 2004, 43,
8510; T. Tanaka, K. Yukawa and N. Umesaki, Oncol. Rep.,
2005, 14, 1365; D. Wang and S. J. Lippard, Nat. Rev. Drug
Discovery, 2005, 4, 307.
Cell proliferation assay
Inhibition of cell proliferation by copper complexes was
measured by a MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-
tetrazolium bromide) assay. RT112 and RT112-CP cells were
seeded into 96-well plates (5 × 103 cells per well) in 100 µl of
culture medium. After 24 h, cells were treated with cisplatin
(Sigma-Aldrich, Lyon, France) or complexes 12+–6+ at various
concentrations. In parallel, a control with DMSO (vehicle
alone) at the same dilutions was done. Following incubation
for 48 h, 10 µl of a MTT (Euromedex, Mundolsheim, France)
stock solution in PBS at 5 mg ml−1 was added in each well and
the plates were incubated at 37 °C for 3 h. Plates were then
centrifugated for 5 min at 400 g before the medium was
discarded and replaced with DMSO (100 µl per well) to solubil-
ize water-insoluble purple formazan crystals. After 15 min
under shaking, absorbance was measured on an ELISA reader
(Tecan, Männedorf, Switzerland) at a test wavelength of
570 nm and a reference wavelength of 650 nm. Absorbance
obtained by cells treated with the same dilution of the vehicle
alone (DMSO) was rated as 100% cell survival. Each data point
is the average of three independent experiments.
9 A. Q. Lei, L. Cheng and C.-X. Pan, Expert Rev. Anticancer
Therapy, 2011, 11, 1851.
10 B. Meunier, Chem. Rev., 1992, 92, 1411; D. S. Sigman, Acc.
Chem. Res., 1986, 19, 180; G. Jiang, N. Xiao, P. Shi, Y. Zhu
and Z. Guo, Coord. Chem. Rev., 2007, 1951; M. Pitié and
G. Pratviel, Chem. Rev., 2010, 110, 1018.
Cell cycle analysis
11 L. Li, N. N. Murthy, J. Telser, L. N. Zakharov, G. P. A. Yap,
A. L. Rheingold, K. D. Karlin and S. E. Rokita, Inorg. Chem.,
2006, 45, 7144.
RT112 cells were cultured in 6-well plates (2 × 105 cells per
well) for 24 h before treatment with DMSO or complexes 2+, 4+,
5+ and 6+ for 24 h or 48 h. Trypsinized and floating cells were 12 For recent representative articles see: S. Thyagarajan,
then pooled, fixed with 70% ethanol, washed with PBS and
stained with 20 µg ml−1 propidium iodide (PI) in the presence
of 0.5 mg ml−1 of RNAse (Sigma-Aldrich, Lyon, France). Data
N. N. Murthy, A. A. Narducci Sarjeant, K. D. Karlin and
S. E. Rokita, J. Am. Chem. Soc., 2006, 128, 7003; J. L. Garcia-
Gimenez, G. Alzuet, M. Gonzales-Alvarez, A. Castinheiras,
M. Liu-Gonzalez and J. Borras, Inorg. Chem., 2007, 46,
7178; M. Gonzales-Alvarez, G. Alzuet, J. Borras, L. des
Castillo-Agudo, J. M. Montejo-Bernardo, A. Gutierrez-
acquisitions were performed with
a
FACScan (Becton
Dickinson France, Le Pont-de-Claix, France) equipped with a
488 nm argon laser and PI fluorescence was collected with a
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8468–8483 | 8481

View Article Online
Paper
Dalton Transactions
Rodriguez and S. Garcia-Granda, J. Biol. Inorg. Chem., 2008, 25 F. Michel, F. Thomas, S. Hamman, E. Saint-Aman,
13, 1249; Y. Kou, J. Tian, D. Li, W. Gu, X. Liu, S. Yan, C. Bucher and J. L. Pierre, Chem.–Eur. J., 2004, 10, 4115.
D. Liao and P. Cheng, Dalton Trans., 2009, 2374; A. Arbuse, 26 A. W. Addison, T. N. Rao, J. Reedijk, J. Von Rijn and
M. Font, M. A. Martinez, X. Fontrodona, M. J. Prieto, G. C. Verschoor, J. Chem. Soc., Dalton Trans., 1984, 1349.
V. Moreno, X. Sala and A. Llobet, Inorg. Chem., 2009, 48, 27 D. Zurita, I. Gautier-Luneau, S. Ménage, J. L. Pierre and
11098; K. Ghosh, P. Kumar, N. Tyagi, U. P. Singh,
V. Aggarwal and M. C. Baratto, Eur. J. Med. Chem., 2010, 45,
3770.
E. Saint-Aman, J. Biol. Inorg. Chem., 1997, 2, 46; S. Itoh,
S. Takayama, R. Arakawa, A. Furuta, M. Komatsu, A. Ishida,
S. Takamuku and S. Fukuzumi, Inorg. Chem., 1997, 36,
1407; Y. Shimazaki, S. Huth, A. Odani and O. Yamauchi,
Angew. Chem., Int. Ed., 2000, 39, 1666; F. Thomas,
G. Gellon, I. Gautier-Luneau, E. Saint-Aman and
J.-L. Pierre, Angew. Chem., Int. Ed., 2002, 41, 3044.
13 S. Dhar, M. Nethaji and A. R. Chakravarty, Inorg. Chem.,
2005, 44, 8876.
14 S. Dhar, M. Nethaji and A. R. Chakravarty, Inorg. Chim.
Acta, 2005, 358, 2437; S. Dhar, M. Nethaji and
A. R. Chakravarty, Inorg. Chem., 2006, 45, 11043.
15 M. Scarpellini, A. Neves, R. Horner, A. J. Bortoluzzi,
B. Szpoganics, C. Zucco, R. A. Nome Silva, V. Drago,
28 S. Itoh, M. Taki, H. Kumei, S. Takayama, S. Nagatomo,
T. Kitagawa, N. Sakurada, R. Arakawa and S. Fukuzumi,
Inorg. Chem., 2000, 39, 3708.
A. S. Mangrich, W. A. Ortiz, W. A. C. Passos, M. C. B. De Oli- 29 Y. Shimazaki, S. Huth, S. Hirota and O. Yamauchi, Bull.
veira and H. Terenzi, Inorg. Chem., 2003, 42, 8353; Chem. Soc. Jpn., 2000, 73, 1187.
C. Fernande, G. L. Parrilha, J. A. Lessa, L. J. M. Santiago, 30 C. G. Saysell, C. D. Borman, A. J. Baron, M. J. McPherson
M. M. Kanashiro, F. S. Boniolo, A. J. Bortoluzzi, and A. G. Sykes, Inorg. Chem., 1997, 36, 4520.
N. V. Vugman, M. H. Herbst and A. Horn Jr., Inorg. Chim. 31 F. Michel, F. Thomas, S. Hamman, C. Philouze, E. Saint-
Acta, 2006, 359, 3167. Aman and J. L. Pierre, Eur. J. Inorg. Chem., 2006, 3684.
16 M. C. B. Oliveira, D. Mazera, M. Scarpellini, P. Cardoso 32 F. Michel, S. Hamman, F. Thomas, C. Philouze, I. Gautier-
Severino, A. Neves and H. Terenzi, Inorg. Chem., 2009, 48,
2711.
17 V. Rajendiran, R. Karthik, M. Palaniandavar, H. Stoeckli-
Luneau and J. L. Pierre, Chem. Commun., 2006, 4122;
F. Michel, S. Hamman, C. Philouze, C. Pérez Del Valle,
E. Saint-Aman and F. Thomas, Dalton Trans., 2009, 832.
Evans, V. S. Periasamy, M. A. Akbarsha, B. S. Srinag and 33 H. Adams, N. A. Bailey, I. K. Campbell, D. E. Fenton and
H. Krishnamurthy, Inorg. Chem., 2007, 46, 8208. H. YU He, J. Chem. Soc., Dalton Trans., 1996, 2233.
18 S. Ramakrishnan, V. Rajendiran, M. Palaniandavar, 34 M. Vaidyanathan, R. Viswanathan, M. Palaniandavar,
V. S. Periasamay, M. A. Akbarsha, B. S. Srinag and
T. Balasubramanian, P. Prabhaharan and T. P. Muthiah,
H. Krishnamurthy, Inorg. Chem., 2009, 48, 1309;
Inorg. Chem., 1998, 37, 6418.
R. Loganathan, S. Ramakrishnan, E. Suresh, A. Riyasdeen, 35 H. Adams, N. A. Bailey, C. O. Rodriguez de Barbarin,
M. A. Akbarsha and M. Palaniandavar, Inorg. Chem., 2012,
51, 5512.
D. E. Fenton and Q. Y. He, J. Chem. Soc., Dalton Trans.,
1995, 2323.
19 S. Borah, M. S. Melvin, N. Lindquist and R. A. Manderville, 36 P. Kumar, S. Gorai, M. K. Santra, B. Mondal and D. Manna,
J. Am. Chem. Soc., 1998, 120, 4557; M. S. Melvin, Dalton Trans., 2012, 41, 7573.
J. T. Tomlinson, G. R. Saluta, G. L. Kucera, N. Lindquist 37 M. C. B. De Oliveira, M. Scarpellini, A. Neves, H. Terenzi,
and R. A. Manderville, J. Am. Chem. Soc., 2000, 122, 6333.
20 T. Bortolotto, P. Pereira Silva, A. Neves, E. C. Pereira-Maia
and H. Terenzi, Inorg. Chem., 2011, 50, 10519.
A. J. Bortoluzzi, B. Szpoganics, A. Greatti, A. S. Mangrich,
E. M. de Souza, P. M. Fernadez and M. R. Soares, Inorg.
Chem., 2005, 44, 921.
21 S. S. Bhat, A. A. Kumbhar, H. Heptullah, A. A. Khan, 38 R. H. Wopschall and I. Shain, Anal. Chem., 1967, 39, 1514.
V. V. Gobre, S. P. Gejji and V. G. Puranik, Inorg. Chem., 39 H. Nagao, N. Komeda, M. Mukaida, M. Suzuki and
2011, 50, 545.
K. Tanaka, Inorg. Chem., 1996, 35, 6809.
22 P. U. Maheswari, S. Roy, H. Den Dulk, S. Barends, G. Van 40 M. Schatz, M. Becker, F. Thaler, F. Hampel, S. Schindler,
Wezel, B. Kozlevcar, P. Gamez and J. Reedijk, J. Am. Chem.
Soc., 2006, 128, 710; P. U. Maheswari, S. Barends, S. Özalp-
Yaman, P. De Hoog, H. Casellas, S. J. Teat, C. Massera,
M. Lutz, A. L. Spek, G. P. Van Wezel, P. Gamez and
J. Reedijk, Chem.–Eur. J., 2007, 13, 5213; S. Roy,
R. R. Jacobson, Z. Tyeklar, N. N. Murthy, P. Ghosh,
Q. Chen, J. Zubieta and K. D. Karlin, Inorg. Chem., 2001,
40, 2312; W. Canõn-Mancisidor, E. Spodine, D. Venegas-
Yazigi, D. Rojas, J. Manzur and S. Alavarez, Inorg. Chem.,
2008, 47, 3687.
P. U. Maheswari, M. Lutz, A. L. Spek, H. den Dulk, 41 F. Thomas, O. Jarjayes, C. Duboc, C. Philouze, E. Saint-
S. Barends, G. P. van Wezel, F. Hartlc and J. Reedijk, Dalton
Trans., 2009, 10846.
23 K. Ghosh, P. Kumar, N. Tyagi and U. P. Singh, Inorg. Chem.,
Aman and J.-L. Pierre, Dalton Trans., 2004, 2662.
42 L. Zhu, O. dos Santos, C. W. Koo, M. Rybstein, L. Pape and
J. W. Canary, Inorg. Chem., 2003, 42, 7912.
2010, 49, 7614; K. Ghosh, P. Kumar, V. Mohan, U. P. Singh, 43 S. I. Kirin, C. M. Happel, S. Hrubanova, T. Weyhermueller,
S. Kasiri and S. S. Mandal, Inorg. Chem., 2012, 51, 3343. C. Klein and N. Metzler-Nolte, Dalton Trans., 2004, 1201.
24 A. Philibert, F. Thomas, C. Philouze, S. Hamman, E. Saint- 44 Y. An, S.-D. Liu, S.-Y. Deng, L.-N. Ji and Z.-W. Mao, J. Inorg.
Aman and J. L. Pierre, Chem.–Eur. J., 2003, 9, 3803.
Biochem., 2006, 100, 1586.
8482 | Dalton Trans., 2013, 42, 8468–8483
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
45 Y. Shao, X. Sheng, Y. Li, Z.-L. Jia, J.-J. Zhang, F. Liu and
C. Papandreou and E. Briasoulis, Crit. Rev. Oncol. Hematol.,
G.-Y. Lu, Bioconjugate Chem., 2008, 19, 1840.
2013, DOI: 10.1016/j.critrevonc.2012.12.003, ASAP.
46 L. Tjioe, A. Meininger, T. Joshi, L. Spiccia and B. Graham, 55 D. C. Marks, L. Belov, M. W. Davey, R. A. Davey and
Inorg. Chem., 2011, 50, 4327. A. D. Kidman, Leuk. Res., 1992, 16, 1165.
47 A. Barve, A. Kumbhar, M. Bhat, B. Joshi, R. Butcher, 56 D. J. Gallagher and M. I. Milowsky, Curr. Treat. Options
U. Sonawane and R. Joshi, Inorg. Chem., 2009, 48, 9120. Oncol., 2009, 10, 205.
48 M. Tian, H. Ihmals and E. Brötz, Dalton Trans., 2010, 39, 57 B. Köberle, J. Payne, K. A. Grimaldi, J. A. Hartley and
8195. J. R. Masters, Biochem. Pharmacol., 1996, 52, 1729.
49 C. Liu, S. Yu, D. Li, Z. Liao, X. Sun and H. Xu, Inorg. Chem., 58 J. R. Jackson, A. Gilmartin, C. Imburgia, J. D. Winkler,
2002, 41, 913.
L. A. Marshall and A. Roshak, Cancer Res., 2000, 60, 566;
R. S. DiPoala, Clin. Cancer Res., 2002, 8, 3311.
59 J. H. Houtgraaf, J. Versmissen and W. J. van der Giessen,
Cardiovasc. Revasc. Med., 2006, 7, 165.
50 C. Liu, S. Yu, D. Li, Z. Liao, X. Sun and H. Xu, Inorg. Chem.,
2002, 41, 913.
51 F. M. Muggia and T. Fojo, J. Chemotherapy, 2004, 4, 77.
52 P. Falco, S. Bringhen, I. Avonto, F. Gay, F. Morabito, 60 O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard
M. Boccadoro and A. Palumbo, Expert Rev. Anticancer Ther.,
2007, 7, 945.
53 D. Catovsky, M. Else and S. Richards, Clin. Lymphoma
Myeloma Leuk., 2011, 11, S2.
54 M. Froudarakis, E. Hatzimichael, L. Kyriazopoulou,
K. Lagos, P. Pappas, A. Tzakos, V. Karavasilis, D. Daliani,
and H. Puschmann, J. Appl. Crystallogr., 2009, 42, 339.
61 H. Nagao, N. Komeda, M. Mukaida, M. Suzuki and
K. Tanaka, Inorg. Chem., 1996, 35, 6809.
62 L. Gianelli, V. Amendola, L. Fabbrizzi, P. Pallavicini and
G. G. Mellerio, Rapid Commun. Mass Spectrom., 2001, 15,
2347.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 8468–8483 | 8483