Efficient DNA cleavage induced by copper(II) complexes of hydrolysis derivatives of 2,4,6-tri(2-pyridyl)-1,3,5-triazine in the presence of reducing agents

Article abstract of DOI:10.1002/ejic.200600821

The reaction of 2,4,6-tri(pyridyl)-1,3,5-triazine (ptz) and copper(II) salts in dmf/water (1:1) results in the hydrolysis of ptz and formation of the anions bis(2-pyridylcarbonyl)amide (ptO₂^-) and bis(2-pyridylamine)amide (ptN₂^-), which are found in the complexes [Cu(ptN₂)(OAc)]·3H₂O (1), [Cu-(ptO ₂)(OAc)(H₂O)]·H₂O (2), [Cu(ptN ₂)(for)]·3H₂O (3) (for = formate), [Cu(ptO ₂)(for)(H₂O)] (4), [Cu(ptO₂)(benz)] ·H₂O (5) (benz = benzoate), and [Cu(ptO₂)F(H ₂O)]₂·3H₂O (6). This report includes the chemical and spectroscopic characterization of all these complexes along with the crystal structures of 4-6. The coordination spheres of Cu^II in 4 and 5 are best described as distorted tetragonal square pyramidal for the former and distorted square planar for the latter. The crystal structure of 6 shows the presence of two discrete monomeric [Cu(ptO₂)F(H₂O)] entities in the crystallographic asymmetric unit in which both copper(II) ions have a distorted square-pyramidal coordination geometry. The binding of the complexes to DNA has been investigated with the aid of viscosity and thermal denaturation studies, both of which indicate that the interaction is probably due to the outer-sphere mechanism. The ability of the compounds to cleave DNA has also been tested. Efficient oxidative cleavage was observed in the presence of a mild reducing agent (ascorbate) and dioxygen. Mechanistic studies with reactive oxygen species (ROS) scavengers confirm that hydrogen peroxide, the hydroxyl radical, singlet oxygen-like species, and the superoxide anion are necessary diffusible intermediates in the scission process. A mechanism involving either the Fenton or the Haber-Weiss reaction plus the formation of copper oxene species is proposed for the DNA cleavage mediated by these compounds. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejic.200600821

FULL PAPER
DOI: 10.1002/ejic.200600821
Efficient DNA Cleavage Induced by Copper(II) Complexes of Hydrolysis
Derivatives of 2,4,6-Tri(2-pyridyl)-1,3,5-triazine in the Presence of Reducing
Agents
Joaquín Borrás,*[a] Gloria Alzuet,[a] Marta González-Alvarez,[a] Jose L. García-Giménez,[a]
Benigno Macías,[b] and Malva Liu-González[c]
Keywords: Copper triazine complexes / Oxidative DNA damage / ROS scavengers / Oxidation
The reaction of 2,4,6-tri(pyridyl)-1,3,5-triazine (ptz) and cop-
per(II) salts in dmf/water (1:1) results in the hydrolysis of ptz
and formation of the anions bis(2-pyridylcarbonyl)amide
(ptO2–) and bis(2-pyridylamine)amide (ptN2–), which are
found in the complexes [Cu(ptN2)(OAc)]·3H2O (1), [Cu-
(ptO2)(OAc)(H2O)]·H2O (2), [Cu(ptN2)(for)]·3H2O (3) (for =
formate), [Cu(ptO2)(for)(H2O)] (4), [Cu(ptO2)(benz)]·H2O (5)
plexes to DNA has been investigated with the aid of viscosity
and thermal denaturation studies, both of which indicate that
the interaction is probably due to the outer-sphere mecha-
nism. The ability of the compounds to cleave DNA has also
been tested. Efficient oxidative cleavage was observed in the
presence of a mild reducing agent (ascorbate) and dioxygen.
Mechanistic studies with reactive oxygen species (ROS)
(benz = benzoate), and [Cu(ptO2)F(H2O)]2·3H2O (6). This re- scavengers confirm that hydrogen peroxide, the hydroxyl
port includes the chemical and spectroscopic characteriza-
tion of all these complexes along with the crystal structures
of 4–6. The coordination spheres of CuII in 4 and 5 are best
described as distorted tetragonal square pyramidal for the
former and distorted square planar for the latter. The crystal
structure of 6 shows the presence of two discrete monomeric
[Cu(ptO2)F(H2O)] entities in the crystallographic asymmetric
unit in which both copper(II) ions have a distorted square-
pyramidal coordination geometry. The binding of the com-
radical, singlet oxygen-like species, and the superoxide
anion are necessary diffusible intermediates in the scission
process. A mechanism involving either the Fenton or the
Haber–Weiss reaction plus the formation of copper oxene
species is proposed for the DNA cleavage mediated by these
compounds.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
tion metal ion.[4,5] Moreover, copper is an important struc-
tural metal ion in chromatin,[6–8] which contains about one
copper ion per kilobase.[6] For these reasons, there is an
increased interest in the ability of copper ion to participate
in DNA-damaging reactions in vivo.[8,9]
Introduction
The interaction of reactive oxygen species (ROS) with
DNA is of considerable interest because of the potential
pathobiological significance of ROS-induced DNA damage.
Such damage may be promutagenic and has been linked
to the pathogenesis of human diseases such as cancer and
aging.[1,2]
The transition metal ion catalyzed reduction of hydrogen
peroxide has served as a useful model reaction for generat-
ing ROS. Reduction of H2O2 by copper ion produces highly
reactive DNA-damaging species.[3] Copper ion induces sig-
nificantly more DNA base damage in the presence of H2O2
than does ferrous ion, the other biologically relevant transi-
Kinetic,[9–11] inhibitor,[4,5,9] and sequence-context[12–15]
studies have all suggested that DNA damage induced by
copper ion in the presence of H2O2 occurs site-specifically
at the sites of DNA-associated copper. However, the nature
of the last DNA-oxidizing species produced by the interac-
tion of the DNA-CuI complex with H2O2 remains uncer-
tain. Some investigators have suggested that the hydroxyl
radical is the principal reactive intermediate,[9,16] while
others have proposed less reactive intermediates such as an
oxidocopper complex[11,14,15] or cupryl ion.[10] Regardless of
the true nature of the reaction mechanism, exposure of
DNA to copper has been reported to result in single- and
double-strand breaks, modified bases, abasic sites, and
DNA-protein crosslinks.[17,18]
[a] Departament de Química Inorgànica, Facultat de Farmàcia,
Universitat de València,
Vicent Andrés Estellés s/n, 46100 Burjassot, Spain
Fax: +34-963-544-960
E-mail: Joaquin.Borras@uv.es
[b] Departamento de Química Inorgánica, Facultad de Farmacia,
Universidad de Salamanca
2,4,6-Tri(2-pyridyl)-1,3,5-triazine (ptz; Scheme 1) is of
current interest due to its utility as a spacer for designing
multinuclear metal complexes. Compounds of this family
of 2,4,6-triaryltriazines are usually stable against hydrolysis;
Campus Unamuno, 37007 Salamanca, Spain
[c] SCSIE, RX. Universitat de València,
50 Dr. Moliner, 46100 Burjassot, Spain
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
822
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 822–834

Efficient DNA Cleavage Induced by Copper(II) Complexes
FULL PAPER
indeed, their hydrolysis requires the presence of a concen-
trated mineral acid and temperatures above 150 °C.[19] Ler-
ner and Lippard[20,21] have found that ptz undergoes hydrol-
ysis in the presence of CuII in aqueous media. They have
also reported the crystal structure of a copper(II) complex
with a hydrolysis product of ptz. On the basis of the Cu–N
bond lengths and angles of the carbonyl carbon atoms
within the chelate ring, they suggested that the coordination
of ptz induces an angular strain which allows a nucleophilic
attack at the carbon atoms of the triazine ring by the sol-
vent, which, in turn, results in the hydrolysis of ptz.[20,22]
Kinetic and thermodynamic data have revealed that such a
reaction probably occurs by nucleophilic attack at the tri-
azine ring by OH– or H2O.[23] However, a comparison of
the structural and NMR spectroscopic data of rhodium and
ruthenium complexes with ptz and a ptz derivative have led
to the conclusion that a decrease in electron density at the
carbon atoms of the triazine ring due to the electron-with-
drawing effect (LǞM) of the metal ion is a more predomi-
nant factor than angular strain in the metal-assisted hydrol-
ysis of ptz.[24,25] Very recently, Li et al. have confirmed by
a theoretical study that the hydrolysis of ptz is affected by
the electron-withdrawing effect of the metal ions.[26]
The hydrolysis of ptz (Scheme 1) involves the formation
of several intermediates containing either two imino nitro-
–
gens (ptN2 ), one imino and one carbonyl (ptNO–), and
–
finally two carbonyl groups (ptO2 ). The literature contains
–
reports of metal complexes of the ptO2 ligand, usually de-
scribed as bpca–, which is formed upon deprotonation of
bis(2-pyridylcarbonyl)amine (Hbpca; Scheme 1). Mono-,
di- (homo- and heteronuclear), and trinuclear complexes
have been reported for various divalent and trivalent metal
ions (MII = CuII, FeII, MnII, CoII, NiII, and ZnII; MIII
=
FeIII, RhIII).[27–33] Among these, a ternary complex of cop-
per(II) with one bpca– and another derivative formed dur-
ing the hydrolysis process, namely 2-picolinamide, has been
Scheme 1. Hydrolysis of 2,4,6-tri(2-pyridyl)-1,3,5-triazine.
Results and Discussion
–
described.[28] The ptO2 anion is obtained by means of a
Cannizzaro-type reaction of pyridine-2-carboxaldehyde and
NaNCO in the presence of copper metal ion.[34] Kawijara
et al. have described the ptN2– ligand,[35] which forms when
ptz is mixed with copper(II) acetate at reflux in dry MeOH
in the presence of NaClO4. This ligand links the copper(II)
ions to form a polymer chain.
Synthesis
The reaction of copper salts and ptz in dmf/H2O (4:1)
–
–
resulted in the hydrolysis of ptz to ptO2 or ptN2 anions
and afforded complexes of composition [Cu(ptO2)(X)-
(H2O)], [Cu(ptN2)(X)]·3H2O (X = acetate or formate),
[Cu(ptO2)(benz)]·H2O, and [Cu(ptO2)(F)(H2O)]2·3H2O.
In this paper we describe the crystal structures of cop-
–
per(II) complexes with ptO2 and other new copper com-
–
plexes with ptN2 . The ptz and the ptz-derived ligands pres-
Crystal Structures
ent a planar structure and can interact with the bases of
DNA through stacking, which is one of the well-known
conditions for interaction with DNA in order to give rise
to ROS that then cleave the DNA. In this vein, Thorp et
[Cu(ptO2)(for)(H2O)] (4)
Figure 1 shows an ORTEP view of the molecule with the
al.[36] have reported the DNA-cleavage activity of RuII-ptz atomic numbering scheme. Selected bond lengths and
complexes based on the formation of oxidoruthenium(IV). angles are listed in Table 1. The coordination sphere of cop-
Our group has also published numerous papers about the per(II) is best described as a distorted tetragonal square
nuclease activity of copper(II) complexes in the presence of pyramid, as can be deduced from the τ value of 0.24.[43] The
reducing agents.[37–42] In this paper we assay the nuclease basal positions are occupied by the three nitrogen atoms of
–
activity of the synthesized copper(II) complexes in the pres- the ptO2 ligand and one oxygen atom of the monodentate
ence of reducing agents such as mercaptopropionic acid, formate (for) anion. The apical site is occupied by the oxy-
ascorbic acid, and glutathione.
gen atom from the coordinated water molecule. The Cu–N
Eur. J. Inorg. Chem. 2007, 822–834
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
823

J. Borrás et al.
FULL PAPER
bond lengths, which range from 1.942(3) to 2.022(2) Å, are
The crystal packing is stabilized by intermolecular hydro-
very similar to those found in other previously reported gen bonds. A moderate hydrogen bond is formed between
– [27]
copper(II) complexes with ptO2 .
The Cu–Ofor and the O1W and O3 of the formate anion [O1W···O3 2.792(5) Å,
Cu–Owater bond lengths are 1.933(2) and 2.329(4) Å, respec- O1W–H1A–O3 174.92(4.60)°]. In addition, O1W forms a
tively. The four basal atoms are quasi coplanar, and the moderately strong hydrogen bond with the carbonyl O1
copper ion is displaced 0.106 Å below the plane. The second [O1W···O1 2.944(3) Å, O1W–H1B–O1 137.56(6)°].[44]
oxygen of the formate group is located a long distance away
[Cu(ptO2)(benz)]·H2O (5)
from copper(II) [2.882(4) Å].
An ORTEP view of the complex with the atom number-
ing scheme is shown in Figure 2. Selected bond lengths and
angles are listed in Table 1. The geometry of the copper(II)
ion can be described as distorted square planar. The coordi-
–
nation polyhedron is formed by a ptO2 , which functions
as a tridentate ligand through the nitrogen atoms, and one
carboxylate O atom from the benzoate (benz) ligand. The
three Cu–N bond lengths, which are within the range
1.9268(13)–2.0007(16) Å, are similar to those of the com-
plex described above and to those found for other cop-
per(II) ptO2 complexes.[27] The distortion of the square-
planar geometry can be determined by the dihedral angle
between the CuN2 and CuNO planes. When the dihedral
angle is 0° the geometry is planar while when the angle is
90° the stereochemistry is tetrahedral. In this complex, the
dihedral angle value of 3.48(6)° indicates a very slightly dis-
torted square planar configuration. A distorted square
pyramid could be considered due to fact that the Cu–O4
Figure 1. ORTEP drawing of [Cu(ptO2)(for)(H2O)] (4).
Table 1. Selected bond lengths [Å] and angles [°] for [Cu(ptO2)(for)(H2O)] (4), [Cu(ptO2)(benz)]·H2O (5), and [Cu(ptO2)F(H2O)]2·3H2O
(6).[a]
[Cu(ptO2)(for)(H2O)] (4)
[Cu(ptO2)(benz)]·H2O (5)
[Cu(ptO2)F(H2O)]2·3H2O (6)
Cu(1)–O(2)
Cu(1)–N(2)
Cu(1)–N(1)
Cu(1)–N(1)#1
Cu(1)–O(1W)#2
1.933(2)
1.942(3)
2.022(2)
2.022(2)
2.329(4)
Cu(1)–O(3)
Cu(1)–N(2)
Cu(1)–N(1)
Cu(1)–N(3)
Cu(1)–O(4)
1.9268(13)
Cu(1)–N(2A)
Cu(1)–F(1)
Cu(1)–N(3A)
Cu(1)–N(1A)
Cu(1)–O(1)
Cu(2)–F(2)
Cu(2)–N(2B)
Cu(2)–N(3B)
Cu(2)–N(1B)
Cu(2)–O(2)
1.87(2)
1.933(12)
1.95(3)
1.9279(14)
1.9927(15)
2.0007(16)
2.690(2)
2.02(2)
2.290(10)
1.902(11)
1.94(2)
1.98(2)
2.03(2)
2.283(11)
O(2)–Cu(1)–N(2)
O(2)–Cu(1)–N(1)
N(2)–Cu(1)–N(1)
O(2)–Cu(1)–N(1)#1
N(2)–Cu(1)–N(1)#1
N(1)–Cu(1)–N(1)#1
O(2)–Cu(1)–O(1W)#2
N(2)–Cu(1)–O(1W)#2
N(1)–Cu(1)–O(1W)#2
176.2(3)
98.28(6)
81.46(6)
98.28(6)
81.46(6)
161.35(12)
91.25(19)
92.51(17)
94.08(9)
O(3)–Cu(1)–N(2)
O(3)–Cu(1)–N(1)
N(2)–Cu(1)–N(1)
O(3)–Cu(1)–N(3)
N(2)–Cu(1)–N(3)
N(1)–Cu(1)–N(3)
178.09(6)
99.37(7)
82.13(7)
96.25(6)
82.19(6)
164.01(6)
N(2A)–Cu(1)–F(1)
N(2A)–Cu(1)–N(3A)
F(1)–Cu(1)–N(3A)
N(2A)–Cu(1)–N(1A)
F(1)–Cu(1)–N(1A)
N(3A)–Cu(1)–N(1A)
N(2A)–Cu(1)–O(1)
F(1)–Cu(1)–O(1)
N(3A)–Cu(1)–O(1)
N(1A)–Cu(1)–O(1)
F(2)–Cu(2)–N(2B)
F(2)–Cu(2)–N(3B)
N(2B)–Cu(2)–N(3B)
F(2)–Cu(2)–N(1B)
N(2B)–Cu(2)–N(1B)
N(3B)–Cu(2)–N(1B)
F(2)–Cu(2)–O(2)
173.1(6)
80.0(13)
98.3(10)
82.7(16)
97.8(11)
160.6(13)
91.2(6)
95.5(4)
94.5(7)
94.7(5)
170.2(6)
102.6(10)
77.5(14)
94.1(12)
84.3(13)
160.5(14)
95.5(4)
N(1)#1–Cu(1)–O(1W)#2 94.08(9)
N(2B)–Cu(2)–O(2)
N(3B)–Cu(2)–O(2)
N(1B)–Cu(2)–O(2)
94.2(6)
94.3(7)
94.0(6)
[a] Symmetry transformations used to generate equivalent atoms: #1 x, y, –z + 2; #2 x + 1, y, z.
824
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 822–834

Efficient DNA Cleavage Induced by Copper(II) Complexes
FULL PAPER
distance is 2.690(2) Å. In this case, the carboxylate acts as a crystal structure shows the presence of two discrete mono-
bidentate ligand, in contrast to that observed for [Cu(ptO2)- meric [Cu(ptO2)F(H2O)] entities and three lattice water
(for)(H2O)] (4) and for the previously reported complex molecules in the crystallographic asymmetric unit. Both
[Cu(ptO2)(OAc)(H2O)]·H2O.[27] The four basal atoms devi- copper(II) ions have a distorted square-pyramidal coordi-
ate slightly from planarity, and the copper ion is displaced nation geometry. The equatorial planes are defined by three
–
0.037 Å below the plane.
nitrogen atoms of the ptO2 ligand [Cu–N in the range
1.87(2)–2.02(2) Å for Cu(1) and 1.94(2)–2.03(3) Å for
Cu(2)] and one fluoride anion [Cu(1)–F(1) 1.933(12),
Cu(2)–F(2) 1.902(11) Å]. The apical position is occupied by
the O atom of a water molecule [Cu(1)–O1 2.290(10),
Cu(2)–O2 2.283(11) Å]. The equatorial plane displays a
slight tetrahedral distortion as Cu(1) is displaced 0.1407 Å
below the plane while Cu(2) lies 0.1559 Å above the plane.
These two monomeric [Cu(ptO2)F(H2O)] entities show dif-
ferent trigonality parameters (τ). These different τ values Ϫ
0.24 for Cu(1) and 0.16 for Cu(2) Ϫ are due to crystal pack-
ing effects.
–
The CuN3F plane in each entity is coplanar to the ptO2
ligand [angle between planes of 179.74° in Cu(1) and
178.63° in Cu(2)]. Moreover, both [Cu(ptO2)F(H2O)] enti-
ties are also coplanar (the angle between the Cu1N3F and
Cu2N3F planes is 178.70°).
The crystal structure is stabilized by many hydrogen-
bonding interactions involving the fluoride anions (F1, F2),
–
the oxygen atoms of the ptO2 ligand (O2, O3), and the
Figure 2. ORTEP drawing of [Cu(ptO2)(benz)]·H2O (5).
oxygen atoms of the water molecules (O1W, O2W, and
O3W), some of which are of moderate strength.[44]
The molecule is stabilized by moderate and weak hydro-
gen bonds formed by O1W and two O carbonyl groups
[O1W···O1 3.112(3) Å, O1W–H1A–O1 129.67(2.35)°;
O1W···O2 2.858(2) Å, O1W–H1A–O2 150.18(2.45)°;
O1W···O1 3.112(3) Å, O1W–H1B–O1 101.46(2.35)°].[44]
Spectroscopic Properties
The IR spectra of [Cu(ptN2)(for)]·3H2O (3; violet) and
[Cu(ptO2)(for)(H2O)] (4; blue) are significantly different
(see Figure S1 in the Supporting Information). Thus, while
[Cu(ptO2)F(H2O)]2·3H2O (6)
the presence of a strong band at 1712 cm–1, attributable to
–
Figure 3 displays the structure of complex [Cu(ptO2)- the ν(C=O) vibrations characteristic of the ptO2 ligand,[27]
F(H2O)]2·3H2O with the labeling scheme. Selected in- is a significant feature in the IR spectrum of 4, the IR spec-
teratomic distances and angles are listed in Table 1. The trum of 3 does not show this band but rather shows a band
Figure 3. ORTEP drawing of [Cu(ptO2)F(H2O)]2·3H2O (6).
Eur. J. Inorg. Chem. 2007, 822–834
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
825

J. Borrás et al.
FULL PAPER
at 1666 cm–1 due to the δ(N–H) vibrations. Another domain water molecules and the dissociation of the carboxylate
in which noteworthy differences exist between the IR spec- anions. Moreover, a weak LMCT band at about 400 nm,
tra of both complexes is that between 3500 and 3000 cm–1. similar to that observed in the solid-state spectra, is present,
For example, in the spectrum of 4, a strong band with two which indicates the interaction between the metal ion and
–
peaks is observed at 3500 and 3440 cm–1, attributable to the ptO2 anion. Finally, the UV/Vis spectrum of an aque-
ν(O–H) vibrations of the water molecules, while in the IR ous solution of complex 1, while showing a band at 590 nm
spectrum of 3 two bands appear at 3300 and 3100 cm–1. (ε = 49 –1 cm–1), shows none around 400 nm (see Figure S2
These can most likely be assigned to ν(N–H) and ν(O–H) in the Supporting Information), thus confirming the results
vibrations of the lattice water molecules, respectively. Both in the solid state.
spectra exhibit bands at 1600 and 1350 cm–1 that corre-
spond to the νas(COO–) and νs(COO–) vibrations of the for-
mate group.
EPR Spectroscopy
The interpretation of the IR spectrum of 4 is corrobo-
rated by the corresponding IR spectra of [Cu(ptO2)(benz)]·
H2O (5) and [Cu(ptO2)F(H2O)]2·3H2O (6), which both
show the strong band at 1712 cm–1 and the two peaks at
3500 and 3440 cm–1. In contrast, the IR spectrum of
[Cu(ptN2)(OAc)]·3H2O (1) shows no bands at 3500, 3440,
or 1712 cm–1, but rather at 3300, 3100, and 1660 cm–1, sim-
ilar to those of [Cu(ptN2)(for)]·3H2O (3). In the IR spec-
trum of 5, the bands attributable to the νas(COO–) and
νs(COO–) vibrations of the benzoate group appear at 1600
and 1350 cm–1, respectively.
The X- and Q-band spectra of polycrystalline samples of
the complexes were recorded at room temperature. Powder
samples of 2, 4, 5, and 6 were prepared from the single
crystals. The EPR parameters obtained by simulation with
WINSimphonia[45] are given in Table 2. The Q-band EPR
spectrum of 5 is the axial-only spectrum, with gʈ Ͼ gЌ. This
indicates a mainly copper(II) dx2–y2 orbital ground state.
The other Q-band EPR spectra are rhombic, with gx Ͻ gy
Ͻ gz and R values ranging from 0.07 to 0.38, which suggests
that the unpaired electron is in the dx2–y2 orbital. The
pattern of the EPR spectra is in good agreement with the
square-planar geometry of 5 and the square-pyramidal ge-
ometry of 4 and 6. Interestingly, the better resolution of the
Q-band spectra allowed us to deduce the rhombic nature of
the EPR spectrum of 2, which had been hypothesized to be
axial in a previous paper.[27]
Differences between complexes 3 and 4 are also found in
their reflectance spectra. Thus, while the spectrum of 4
shows an unresolved broad d–d band at about 637 nm, that
of 3 presents the d–d transition at about 565 nm. The same
pattern is observed for the d–d bands of 2 (620 nm) and 1
(565 nm). Complexes 5 and 6 show the d–d band at 582
and 627 nm, respectively. The position of the d–d maximum
thus follows the order 5 Ͻ 2 Ͻ 6 Ͻ 4. The order of the d–
d maximum position for the three latter complexes is in
good agreement with the extent of geometrical distortion
defined by the τ parameter: 0.16 (2), 0.16 and 0.24 (6), and
0.24 (4). In addition to the d–d band, a shoulder appears
at about 420 nm in complexes 2, 4, and 5; this can be as-
signed to an LMCT band. While the spectrum for the com-
It is also interesting to note that the EPR spectra of the
–
complexes with the ptO2 ligands (2 and 4) are different
–
from those of the complexes with the ptN2 ligand (1 and
3; see Figure S3 in the Supporting Information).
Mass Spectral Analysis
The FAB spectra of complexes 2, 4, 5, and 6 show a
plex 1 does not exhibit this band, that of 6 shows a very major peak at m/z 289 that corresponds to the [Cu(ptO2)]+
strong band at 400 nm that can be attributed to an F–Cu fragment due to the dissociation of the carboxylate or F–
LMCT band.
anions. There are, however, other significant peaks in the
As the complexes are soluble in H2O, we also recorded spectra. The FAB spectrum of 2, for example, shows peaks
their UV/Vis spectra in aqueous solution. The spectra of at m/z 349 and 307 attributable to the [Cu(ptO2)(OAc) +
aqueous solutions of complexes 5, 2, and 4 show the same H]+ and [Cu(ptO2) + H2O]+ fragments. The FAB mass spec-
maxima at 646 nm (ε = 45, 91, and 90 –1 cm–1, respec- trum of 6, which is similar to that of 2, shows a peak at m/z
tively), thus suggesting that all three compounds contain 309 that can be attributed to the [Cu(ptO2)F + H]+ frag-
the same chromophore in solution. The shift of the d–d ment. The FAB spectrum of complex 4 displays a peak at
band position with respect to that in the diffuse reflectance m/z 335 that corresponds to the [Cu(ptO2)(for)H]+ entity.
spectra is probably a consequence of the coordination of The FAB spectrum of 5 shows a peak at m/z 411 that is
Table 2. X- and Q-band EPR data for the complexes.
Q-band
gx
X-band
gʈ
Compound
gy
gz
R
gʈ
gЌ
gЌ
[Cu(ptN2)(OAc)]·3H2O (1)
[Cu(ptO2)(OAc)(H2O)]·H2O (2)
[Cu(ptN2)(for)]·3H2O (3)
[Cu(ptO2)(for)(H2O)] (4)
[Cu(ptO2)(benz)]·H2O (5)
[Cu(ptO2)F(H2O)]·H2O (6)
2.045
2.046
2.043
2.044
–
2.062
2.049
2.064
2.064
–
2.217
2.230
2.216
2.238
–
0.11
0.07
0.14
0.17
–
2.060
2.211
–
2.063
2.191
–
2.208
2.060
–
2.19
2.056
–
2.207
2.046
2.047
2.094
2.218
0.38
826
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 822–834

Efficient DNA Cleavage Induced by Copper(II) Complexes
FULL PAPER
assigned to the [Cu(ptO2)(benz) + H]+ fragment. Interest- turation profile of DNA. The melting temperature (Tm) of
ingly, the mass spectrum of 1 presents a peak at m/z 305, CT-DNA upon addition of 3, 5, or 6 was found to be 75.7,
which corresponds to the [Cu(ptN2) + H2O]+ fragment. The 78.5, and 75.6 °C respectively. The complexes thus produce
presence of this peak agrees with the difference in mass be- no increase in Tm upon addition to DNA (Tm = 76.6 °C in
–
–
tween ptO2 (226.06) and ptN2 (224.09).
cacodylate buffer 0.1 , NaCl 2 m, pH 6.0). These results
suggest that the complexes do not effect duplex stabilization
through intercalation between the DNA base-pairs.
Electrochemical Behavior
In summary, both viscosity and Tm measurements indi-
cate that the complexes do not intercalate between the
DNA base-pairs. Most probably, the interaction of the
complexes is mainly electrostatic through the phosphate
backbone of the DNA since the compounds are present in
solution as the positive [Cu(ptO2)]+ or [Cu(ptN2)]+ species.
In the electroactivity domain of acetonitrile/Bu4NClO4,
the cyclic voltammograms of complexes 2 and 5 have the
same shape. A reduction peak is observed around –0.90 mV
(vs. SCE) towards cathodic potentials; this can be attributed
to the reduction of CuII to CuI. The same Ecat value for
both complexes indicates that 2 and 5 give identical species
in acetonitrile solution, probably the [Cu(ptO2)]+ entity,
thus confirming the UV/Vis and FAB data. On the reverse
scan, an anodic peak was found at about –0.2 mV. The vol-
tammogram is clearly irreversible, probably because the CuI
species formed upon reduction of CuII is unstable. The re-
duction peak at –0.90 mV relative to SCE corresponds to
–0.68 mV relative to NHE. Such a value indicates a higher
stabilization of the CuII species with respect to those of CuI.
Attempts to determine the electrochemical behavior of
the complexes in cacodylate buffer, the medium used in the
biological studies, were unsuccessful because a solid precipi-
tated in the electrolytic cell.
Nuclease Activity of Complexes and Mechanistic
Investigations
Oxidative DNA Cleavage Mediated by the Compounds
The ability of complexes 1–6 to cleave DNA was investi-
gated by means of gel electrophoresis with supercoiled
pUC18 in 0.1  cacodylate buffer (pH 6.0) and with ascorb-
ate activation. Control experiments with CuSO4 were also
carried out under the same experimental conditions. The
DNA cleavage activity of the complexes was assayed by the
conversion of supercoiled DNA (SC, form I) to nicked cir-
cular (NC, form II) or linearized DNA (LC, form III). All
the complexes were found to exhibit nuclease activity (Fig-
ure 4). At 6, 18, and 30 µ the compounds present the same
cleavage pattern (lanes 2–4, parts a–f of Figure 4). The fact
that the complexes exhibit similar nuclease activity is not
surprising since they are dissociated in solution into
[Cu(ptO2)]+ or [Cu(ptN2)]+. The [Cu(ptO2)]+ or [Cu-
(ptN2)]+ entities derived from the complexes must therefore
be the active species that mediate DNA-strand scission
since CuSO4 at the highest concentration assayed is only
able to produce a partial conversion of SC DNA into NC
DNA (lane 6, Figure 4, g). Thus, the DNA cleavage activity
of the compounds can be ascribed to the cooperative effect
DNA-Binding Interactions
The mode and propensity of binding of the complexes to
calf-thymus DNA (CT-DNA) were studied with the aid of
various techniques.
Viscosity Measurements
Viscosity measurements were carried out on CT-DNA by
varying the concentration of the complexes. Hydrodynamic
measurements that are sensitive to length changes are re-
garded as the least ambiguous and most critical tests of
binding in solution. One classic intercalation model results
in the lengthening of the DNA helix as the base-pairs are
separated to accommodate the binding ligand, thus leading
to the increase of DNA viscosity. In contrast, complexes
that bind exclusively in DNA grooves by means of partial
and/or nonclassical intercalation under the same conditions
typically cause either a less pronounced change (positive or
negative) of DNA solution viscosity or none at all.[46] The
results obtained with our compounds reveal that the pres-
ence of complexes 3, 5, or 6 has no effect on the relative
viscosity of CT-DNA. This indicates that the interaction of
these complexes follows the general pattern of complexes
that bind in DNA grooves by either a partial or nonclassical
intercalation.
–
–
between CuII and the ptO2 or ptN2 ligands.
Because certain copper(II) complexes are able to bring
about hydrolytic cleavage of DNA, the possibility of a hy-
drolytic mechanism for the DNA damage mediated by these
compounds must be taken into account. In order to clarify
this point, the nuclease activity of complexes 1–6 was as-
sayed without ascorbate activation. No DNA cleavage was
observed, which indicates that a hydrolytic mechanism is
not involved (lane 1, parts a–f of Figure 4).
The nuclease efficiency of a large number of CuII com-
plexes is known to depend on the reducing agent used for
initiating the DNA cleavage. We thus tested the activity of
4 in the presence of various activators [ascorbate, hydrogen
peroxide, 3-mercaptopropionic acid (MPA), and glutathi-
one] at a 50-fold excess relative to the complex concentra-
tion. The results, shown in Figure 5 (A), indicate the follow-
Thermal Denaturation
In order to obtain a better understanding of the interac- ing increasing order of efficiency: hydrogen peroxide Ͻ glu-
tion of the compounds, the binding of complexes 3, 5, and tathione Ͻ MPA Ͻ ascorbate (cf. lanes 8, 10, 12, and 14 in
6 to CT-DNA was studied by examining the thermal dena- Figure 5, A). By way of comparison, the same assays were
Eur. J. Inorg. Chem. 2007, 822–834
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
827

J. Borrás et al.
FULL PAPER
Figure 4. Agarose gel electrophoresis of pUC18 plasmid DNA treated with the complexes or CuSO4 with [ascorbate]:[tested compound]
= 100:1 or without ascorbate. Incubation time: 1 h (37 °C). (a) [Cu(ptN2)(OAc)]·3H2O (1), (b) [Cu(ptO2)(OAc)(H2O)]·H2O (2) (c)
[Cu(ptN2)(for)]·3H2O (3), (d) [Cu(ptO2)(for)(H2O)] (4), (e) [Cu(ptO2)(benz)]·H2O (5), (f) [Cu(ptO2)F(H2O)]2·3H2O (6), (g) CuSO4; a–f:
lane 1: 60 µ complex without ascorbate; lane 2: 6 µ complex with ascorbate 100ϫ; lane 3: 18 µ complex with ascorbate 100ϫ, 30 µ
complex with ascorbate 100ϫ; g: lane 1: λDNA/EcoRI + Hind III Marker; lane 2: supercoiled DNA; lane 3: supercoiled DNA with 3 m
ascorbate; lane 4: 6 µ CuSO4 with ascorbate 100ϫ; lane 5: 18 µ CuSO4 with ascorbate 100ϫ; lane 6: 30 µ CuSO4 with ascorbate
100ϫ.
Figure 5. A) Agarose gel electrophoresis of pUC18 plasmid DNA treated with [Cu(ptO2)(for)(H2O)] (4) in the presence of different
reducing agents. Incubation time: 1 h (37 °C). Lane 1: λDNA/EcoRI + Hind III Marker; lane 2: supercoiled DNA; lane 3: CuSO4 30 µ
+ ascorbate 50ϫ; lane 4: CuSO4 30 µ + H2O2 50ϫ; lane 5: CuSO4 30 µ + MPA 50ϫ; lane 6: CuSO4 30 µ + glutathione 50ϫ; lane 7:
4 18 µ + ascorbate 50ϫ; lane 8: 4 30 µ + ascorbate 50ϫ; lane 9: 4 18 µ + H2O2 50ϫ; lane 10: 4 30 µ + H2O2 50ϫ; lane 11: 4 18 µ
+ MPA 50ϫ; lane 12: 4 30 µ + MPA 50ϫ; lane 13: 4 18 µ + glutathione 50ϫ; lane 14: 4 30 µ + glutathione 50ϫ. B) Agarose gel
electrophoresis of pUC18 plasmid DNA treated with [Cu(ptO2)(benz)]·H2O (5) in the presence of ascorbate and ascorbate/H2O2. Incu-
bation time: 1 h (37 °C). Lane 1: 5 6 µ + ascorbate 50ϫ; lane 2: 5 18 µ + ascorbate 50ϫ; lane 3: 5 30 µ + ascorbate 50ϫ; lane 4: 5
6 µ + ascorbate/H2O2 50ϫ; lane 5: 5 18 µ + ascorbate/H2O2 50ϫ; lane 6: 5 30 µ + ascorbate/H2O2 50ϫ; lane 7: 5 6 µ + ascorbate
100ϫ; lane 8: 5 18 µ + ascorbate 100ϫ; lane 9: 5 30 µ + ascorbate 100ϫ.
828
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 822–834

Efficient DNA Cleavage Induced by Copper(II) Complexes
FULL PAPER
performed with CuSO4. Lanes 3–6 (Figure 5, A) show that cal scavengers (DMSO, tert-butyl alcohol, potassium io-
CuSO4 at 30 µ is not an efficient chemical nuclease in the dide, and urea), singlet oxygen scavengers (azide and
presence of these activating agents. Several authors have DABCO), superoxide radical scavengers (Tiron), and cata-
studied the influence of different reducing agents on the lase. The groove binding preferences were tested in the pres-
cleavage of DNA by copper(II) complexes.[47–50] Sigman et ence of the minor groove binder distamycin and the major
al.,[47] for example, examined the influence of the activating groove binder methyl green. The reduction of CuII to CuI
agent on the nuclease activity of a bis(o-phenanthroline)- during the cleavage process was evaluated with neocupro-
copper(II) complex and deduced that ascorbic acid and 3- ine, a ligand that chelates strongly to copper(I).
mercaptopropionic acid are the most efficient reducing
agents. However, while these same authors found MPA to
be superior to ascorbic acid because it produces less back-
ground cleavage, in our study ascorbic acid is clearly more
efficient than MPA. Chiou et al.[48] have indicated that as-
corbate is more effective in cleaving DNA than other reduc-
ing agents such as MPA and dithiothreitol even though the
latter compounds possess higher reducing powers than as-
corbate. The uniqueness of ascorbate as a reducer in the
scission reaction may depend on its ability to generate hy-
drogen peroxide in the presence of oxygen and metal ions,
whereas other reducing agents with a thiol group are known
to produce superoxide, which rapidly undergoes dismu-
tation in aqueous solution. Detmer et al.[49] have found that
a combination of ascorbate and H2O2 is a more effective
activating agent than ascorbate alone. In order to elucidate
the influence of mixing ascorbate with H2O2, the cleavage
process of 5 was studied with ascorbate/H2O2 (50ϫ) or as-
corbate (50ϫ and 100ϫ) as activators (Figure 5, B). No sig-
nificant differences in DNA damage were found.
A strong inhibition of DNA cleavage mediated by 5 is
observed upon addition of DMSO, tert-butyl alcohol, po-
tassium iodide, or urea, thereby indicating that hydroxyl
radical is involved in the cleavage process (lanes 4–7, Fig-
ure 6, A). The addition of sodium azide or DABCO (lanes 8
and 9, Figure 6, A) also decreases the cleavage efficiency,
1
which indicates that either O2 or a singlet-oxygen-like en-
tity is one of the active oxygen intermediates responsible for
the cleavage. The inhibitory activity of sodium azide can
be ascribed to the affinity of the azide anion for transition
metals.[51] Adding Tiron to the reaction mixture decreases
the DNA damage slightly, which suggests that the superox-
ide anion is involved in the DNA scission (lane 10, Figure 6,
A). Catalase also inhibits the nucleolytic process, which
confirms that hydrogen peroxide is also necessary in the
DNA strand scission (lane 14, Figure 6, A).
Treatment of pUC18 DNA with distamycin prior to the
addition of 5 has no effect on the cleavage reaction medi-
ated by the compound (lane 11, Figure 6, A). In contrast,
prior treatment with methyl green attenuates the DNA
strand scission (lane 12, Figure 6, A). It can thus be inferred
from these results that 5 prefers major groove binding. Fi-
Mechanistic Studies
The mechanism of pUC18 DNA cleavage by the com- nally, neocuproine does not inhibit the DNA damage
pounds was studied using inhibiting reagents. The results (lane 13, Figure 6, A). This is probably due to the strong
for complexes 3 and 5 are shown in Figure 6. The involve- redox stability of the [Cu(ptO2)]+ species, as was deduced
ment of ROS was investigated with standard hydroxyl radi- from the electrochemical studies.
Figure 6. A) Agarose gel electrophoresis of pUC18 plasmid treated with 30 µ [Cu(ptO2)(benz)]·H2O (5) and 1.5 m ascorbate/H2O2 in
the presence of potential inhibitors. Incubation time: 1 h (37 °C). B) Agarose gel electrophoresis of pUC18 plasmid treated with 30 µ
[Cu(ptN2)(for)]·3H2O (3) and 1.5 m ascorbate/H2O2 in the presence of potential inhibitors. Incubation time: 1 h (37 °C). Lane 1: λDNA/
EcoRI + Hind III Marker; lane 2: supercoiled DNA; lane 3: complex without inhibitors; lane 4: complex + DMSO (1 ); lane 5: complex
+ tert-butyl alcohol (1 ); lane 6: complex + IK (0.4 ); lane 7: complex + urea (0.4 ); lane 8: complex + NaN3(100 m); lane 9: complex
+ DABCO (0.4 ); lane 10: complex + Tiron (10 m); lane 11: complex + distamycin (8 µ); lane 12: complex + methyl green (2.5 µL of
a 0.01 mgmL–1 solution); lane 13: complex + neocuproine (75 µ); lane 14: complex + catalase (6.5 units).
Eur. J. Inorg. Chem. 2007, 822–834
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
829

J. Borrás et al.
FULL PAPER
The results for complex 3 (Figure 6, B) indicate that, as
was the case for 5, hydroxyl radical, 1O2, and hydrogen per-
oxide are the active species responsible for the DNA dam-
age (lanes 4–7, 8, 9, and 14, Figure 6, B). However, the
cleavage scission process mediated by this compound is
somewhat different from that of 5. First of all, the superox-
ide anion does not seem to be involved in the DNA scission
(lane 10, Figure 6, B). Also, since neocuproine attenuates
the DNA cleavage in this case (lane 13, Figure 6, B), it
seems likely that the reduction of CuII to CuI takes place,
probably due to a lower stability of [Cu(ptN2)]+ with re-
spect to that of [Cu(ptO2)]+. Moreover, in contrast to 5,
complex 3 does not show major or minor groove binding
preferences (lanes 11 and 12, Figure 6, B).
The involvement of hydroxyl radicals was subjected to
further studies. The hydroxyl release by 5 was studied by
using 2-deoxy--ribose as a radical trap.[52] 2-Deoxy--ri-
bose provides a chemically equivalent model for the oxidat-
ive target in DNA strand scission chemistry. The quantita-
tion of 2-deoxy--ribose under sodium ascorbate/H2O2 ac-
tivation conditions is displayed in Figure 7, where it can
be seen that the calculated production of hydroxyl radicals
increases to about 700 n at 60 min.
On the other hand, Sigman has suggested that bis(o-
phenanthroline)copper(II), the classic nuclease system, fol-
lows a different mechanistic pathway.[54] Thus, reduction of
the copper to cuprous state leads to the generation of super-
–
oxide anion. The O2 then dismutates to H2O2 and di-
oxygen and the hydrogen peroxide reacts with another
equivalent of copper(I) to produce hydroxyl radical-like
species, which may be bound to the metal. This species,
which could be considered analogous to a metal-oxido spe-
cies, is responsible for initiating DNA strand scission.[55]
Bocarsly et al. have compared their proposed mechanism
with that of Cu-phenanthroline. They concluded that the
differences between the two mechanisms are most likely re-
lated to the electronic differences between the ligands. Their
mechanism is related to complexes with ligands that present
σ character, while the mechanism proposed by Sigman is
related to complexes with ligands that present π character
–
–
(phen). Since the ptO2 and the ptN2 ligands in the title
complexes present both σ and π character, both mecha-
nisms can be proposed. This could explain why both hy-
droxyl radical scavengers and scavengers of singlet oxygen
or singlet-oxygen-like species inhibit the DNA damage and
strand breakage.
Conclusions
Ternary copper(II) complexes with 2,4,6-tri(2-pyridyl)-
1,3,5-triazine hydrolytic derivatives have been prepared and
structurally characterized with the aid of X-ray crystal-
lography. The binding of DNA has been studied with the
aid of viscosity and Tm measurements, which indicate that
the complexes do not intercalate between the DNA base
pairs. In contrast, an electrostatic interaction of the com-
plexes with the phosphate backbone of the DNA can be
deduced since the compounds are present in solution as the
cationic [Cu(ptO2)]+ or [Cu(ptN2)]+ species. The complexes
act as efficient chemical nucleases upon ascorbate acti-
vation. The participation of H2O2, hydroxyl radical, super-
oxide anion, and singlet-oxygen-like entities in the scission
process mediated by the compounds suggests a mechanism
pathway involving a Fenton or a Haber–Weiss reaction and
copper(II)-oxene species.
Figure 7. Production of hydroxyl radicals by [Cu(ptO2)(benz)]·H2O
(5) as assayed by the 2-deoxy--ribose/2-thiobarbituric acid system.
In general, two mechanisms can be proposed for CuII
complexes that mediate oxidative DNA cleavage, one in-
volving hydroxyl radicals and another involving metal-ox-
ido species. Bocarsly et al.[49,50,53] have suggested that hy-
droxyl radicals are generated by a variety of chemical and
physical processes. Transition-metal-mediated hydroxyl rad-
ical production is known to occur by a number of routes;
two well-known pathways are the Fenton and the Haber–
Weiss mechanism. In the Fenton mechanism, the metal acts
as a catalyst for hydroxyl radical generation from hydrogen
peroxide:
Experimental Section
Materials and Physical Measurements: Reagents and solvents were
commercially available and used without further purification.
pUC18 was purchased from Roche Diagnostics. Calf-thymus DNA
(CT-DNA) was supplied by Sigma–Aldrich.
Elemental analyses (C,N,H) were performed with a Carlo–Erba
AAS instrument. IR spectra (KBr disks) were recorded with a
Mattson satellite FT-IR spectrometer in the range 4000–400 cm–1.
Diffuse reflectance spectra (nujol mulls) of the complexes were re-
corded with a Shimadzu UV-2101 PC spectrophotometer. UV/Vis
whereas the Haber–Weiss reaction involves the reaction be-
tween superoxide anion and hydrogen peroxide:
830
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 822–834

Efficient DNA Cleavage Induced by Copper(II) Complexes
FULL PAPER
spectra of the complex solutions were recorded with an HP 8453
Unit cell parameters: a = 6.575(5), b = 13.462(5), c = 15.516(5) Å,
spectrophotometer. FAB mass spectra were obtained with a VG V = 1373.4(12) Å3, Z = 4, Dx = 1.701 mgm–3, µ = 1.619 mm–1.
Autospec spectrometer with 3-nitrobenzyl alcohol as matrix. EPR
spectra were recorded at the X- and Q-band frequencies at room
temperature with a Bruker ELEXSYS spectrometer. Electrochemi-
cal measurements were performed with the aid of a Princeton Ap-
plied Research Model 273A potentiostat/galvanostat. Cyclic vol-
tammograms were obtained for acetonitrile solutions of the com-
plexes with [TBA][ClO4] as the supporting electrolyte under argon.
The electrochemical cell employed had a standard three-electrode
configuration, with a platinum working electrode, a platinum wire
counter electrode, and an Ag/AgCl reference electrode.
Multiple observations were averaged, Rmerge = 0.000, resulting in
1533 unique reflections of which 1390 were observed with I Ͼ
2σ(I). Final mosaicity was 0.42. Data completeness was 99.4%. All
non-hydrogen atoms were refined anisotropically. Hydrogen atoms
were located in the difference Fourier map and located geometri-
cally. The final cycle of full-matrix least-squares refinement based
on 1533 reflections and 131 parameters converged to a final value
of R1 [F2 Ͼ 2σ(F2)] = 0.0291, wR2 [F2 Ͼ 2σ(F2)] = 0.0732, R1 (F2)
= 0.0338, wR2 (F2) = 0.0754. The final difference Fourier maps
showed no peaks higher than 0.323 eÅ–3 at 1.56 Å from C32 and
no troughs deeper than –0.513 eÅ–3 at 0.72 Å from Cu1.
[Cu(ptN2)(OAc)]·3H2O (1): An aqueous solution of Cu2Ac4·2H2O
(10 mL; 200 mg, 1 mmol) was added with continuous stirring to
40 mL of a yellow dmf solution of ptz (312 mg, 1 mmol). A dark
green solution was formed immediately; after 30 min it turned blue-
green in color and overnight it became blue with a small amount
of violet precipitate. The precipitate was isolated by filtration,
washed, and dried until constant weight. Yield: 20%.
C14H19CuN5O5: calcd. C 41.94, H 4.77, N 17.47; found C 41.15,
H 4.59, N 17.76.
[Cu(ptO2)(benz)]·H2O
(5):
Violet
crystal,
size
0.8ϫ0.16ϫ0.008 mm, monoclinic, space group P21/a (determined
from the systematic absences). Data collection was performed at
293 K with a Nonius Kappa-CCD single crystal diffractometer,
with Mo-Kα radiation (λ = 0.7173 Å). The crystal–detector distance
was fixed at 30.4 mm, and a total of 175 images were collected by
means of the oscillation method, with a scan angle per frame of
1.7° oscillation and a 25-s exposure time per image. Unit cell di-
mensions were determined from 6088 reflections between θ =
0.998° and 33.142°. Unit cell parameters: a = 10.359(5), b =
10.295(5), c = 16.555(5) Å, β = 99.482(5)°, V = 1741.4(13) Å3, Z =
4, Dx = 1.636 mgm–3, µ = 1.293 mm–1. Multiple observations were
averaged, Rmerge = 0.027, resulting in 6517 unique reflections of
which 4279 were observed with I Ͼ 2σ(I). Final mosaicity was 0.6.
Data completeness was 98.5%. All non-hydrogen atoms were re-
fined anisotropically. Hydrogen atoms were located in the differ-
ence Fourier map and located geometrically. The final cycle of full-
matrix least-squares refinement based on 6517 reflections and 315
parameters converged to a final value of R1 [F2 Ͼ 2σ(F2)] = 0.0381,
wR2 [F2 Ͼ 2σ(F2)] = 0.0887, R1 (F2) = 0.0701, wR2 (F2) = 0.0997.
The final difference Fourier maps showed no peaks higher than
0.43 eÅ–3 at 0.71 Å from C5 and no troughs deeper than 0.56 eÅ–3
at 0.66 Å from Cu1.
Slow evaporation of the filtrate yielded blue, needle-like crystals
that correspond to [Cu(ptO2)(OAc)(H2O)]·H2O (2), which has pre-
viously been reported as [Cu(bpca)(OAc)(H2O)]·H2O.[26]
[Cu(ptN2)(for)]·3H2O (3) and [Cu(ptO2)(for)(H2O)] (4): An aqueous
solution of [Cu2(for)4]·4H2O (10 mL; 225 mg, 1 mmol) was added
to 40 mL of a yellow dmf solution of ptz (312 mg, 1 mmol). A dark
green solution was formed immediately; after 30 min it turned blue-
green in color and overnight it became blue with a small amount
of violet precipitate. The precipitate was isolated by filtration,
washed, and dried until constant weight. Yield: 16%.
C13H16CuN5O5: calcd. C 42.43, H 3.92, N 18.20; found C 42.43,
H 4.14, N 18.12.
Slow evaporation of the filtrate yielded blue, needle-like crystals
that correspond to [Cu(ptO2)(for)(H2O)] (4). Yield: 47%.
C13H11CuN3O5: calcd. C 44.26, H 3.14, N 11.47; found calcd. C
44.10, H 3.09, N 11.90.
[Cu(ptO2)F(H2O)]2·3H2O
(6):
Blue
crystal,
size
¯
0.4ϫ0.25ϫ0.15 mm, triclinic, space group P1. Data collection
was performed at 293 K with a Nonius Kappa2000 single crystal
diffractometer, with Cu-Kα radiation (λ = 1.54184 Å). The crystal–
detector distance was fixed at 35 mm, and a total of 261 images
were collected by means of the oscillation method, with a scan
angle per frame of 2° oscillation and 2-s exposure time per image.
Unit cell dimensions were determined from 1136 reflections be-
tween θ = 0.999° and 47.840°. Unit cell parameters: a = 6.7680(4),
b = 10.2740(5), c = 20.8690(10) Å, α = 98.635(2)°, β = 90.641(2)°,
γ = 107.633(2)°, V = 1364.78(12) Å3, Z = 2, Dx = 1.722 mgm–3, µ =
2.621 mm–1. Multiple observations were averaged, Rmerge = 0.000,
resulting in 1477 unique reflections of which 745 were observed
with I Ͼ 2σ(I). Final mosaicity was 0.65. Data completeness was
59.4%. All non-hydrogen atoms were refined anisotropically. Hy-
drogen atoms were located in the difference Fourier map and lo-
cated geometrically. The final cycle of full-matrix least-squares re-
finement based on 1477 reflections and 407 parameters converged
to a final value of R1 [F2 Ͼ 2σ(F)2] = 0.048, wR2 [F2 Ͼ 2σ(F2)] =
0.122, R1 (F2) = 0.108, wR2 (F2) = 0.162. The final difference Fou-
rier maps showed no peaks higher than 0.36 eÅ–3 at 0.92 Å from
H3B and no troughs deeper than –0.37 eÅ–3 at 0.31 Å from H2A.
[Cu(ptO2)(benz)]·H2O (5): This compound was obtained by a pro-
cedure similar to that described above, but with Cu2(benz)4
(611 mg, 1, mmol), which was obtained either by mixing copper(II)
perchlorate with sodium benzoate or by electrochemical synthesis.
Slow evaporation of the resulting blue solution gave rise to blue,
needle-like crystals after four or five days. Yield: 65%.
C19H15CuN3O5: calcd. C 53.21, H 3.52, N 9.75; found C 52.99, H
3.44, N 9.75.
[Cu(ptO2)F(H2O)]·1.5H2O (6): The synthesis of this compound is
similar to that described above, but with CuF2·2H2O (137 mg,
1 mmol). Blue crystals formed 4 or 5 days after its preparation.
Yield: 22%. C12H13CuFN3O4.5: calcd. C 40.74, H 3.70, N 11.88;
found C 41.50, H 3.38, N 12.03.
X-ray Crystallography
[Cu(ptO2)(for)(H2O)]
(4):
Greenish-blue
crystal,
size
0.2ϫ0.2ϫ0.2 mm, orthorhombic, space group A21am (determined
from the systematic absences). Data collection was performed at
293 K with a Nonius Kappa-CCD single crystal diffractometer,
with Mo-Kα radiation (λ = 0.7173 Å). The crystal–detector distance
was fixed at 45 mm, and a total of 148 images were collected by
means of the oscillation method, with a scan angle per frame of 2°
oscillation and a 10-s exposure time per image. Unit cell dimensions
were determined from 895 reflections between θ = 2.91° and 27.49°.
Data collection for all complexes was performed with the program
Collect.[56] Data reduction and cell refinement were performed with
the programs HKL Denzo and Scalepack.[57] The crystal structure
was solved by direct methods with the program SIR-97.[58] Aniso-
Eur. J. Inorg. Chem. 2007, 822–834
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
831

J. Borrás et al.
Table 3. Crystal and structure refinement data for [Cu(ptO2)(for)(H2O)] (4), [Cu(ptO2)(benz)]·H2O (5) and [Cu(ptO2)F(H2O)]2·3H2O (6).
FULL PAPER
[Cu(ptO2)(for)(H2O)] (4)
[Cu(ptO2)(benz)]·H2O (5)
[Cu(ptO2)F(H2O)]2·3H2O (6)
Empirical formula
Formula weight
C13H11CuN3O5
351.78
C19H15CuN3O5
428.88
C24H26Cu2F2N6O9
707.59
Temperature [K]
Wavelength [Å]
293(2)
0.71069
293(2)
0.71069
293(2)
1.54184
¯
Crystal system, space group
a [Å]
cba, A21am
6.575(5)
monoclinic, P21/a
10.359(5)
triclinic, P1
6.7680(4)
b [Å]
c [Å]
α [°]
13.462(5)
15.516(5)
90.000(5)
10.295(5)
16.555(5)
90.000(5)
10.2740(5)
20.8690(10)
98.635(2)
β [°]
90.000(5)
90.000(5)
1373.4(12)
4, 1.701
1.619
712
99.482(5)
90.000(5)
1741.4(13)
4, 1.636
1.293
876
90.641(2)
107.633(2)
1364.78(12)
2, 1.722
2.621
γ [°]
Volume [Å3]
Z, calculated density [Mgm–3]
Absorption coefficient [mm–1]
F(000)
720
Limiting indices
8ՅhՅ8
–15ՅhՅ15
–15ՅkՅ14
–25ՅlՅ25
10399/6517 [Rint = 0.0272]
6517/0/315
R1 = 0.0381, wR2 = 0.0887
R1 = 0.0701, wR2 = 0.0977
–6ՅhՅ6
–9ՅkՅ9
0ՅlՅ16
2561/1477 [Rint = 0.0000]
1477/3/407
R1 = 0.0476, wR2 = 0.1218
R1 = 0.1079, wR2 = 0.1615
–17ՅkՅ17
–19ՅlՅ20
1568/1533 [Rint = 0.0000]
1533/2/131
R1 = 0.0291, wR2 = 0.0732
R1 = 0.0338, wR2 = 0.0754
Reflections collected/unique
Data/restraints/parameters
Final R indices [I Ͼ 2σ(I)]
R indices (all data)
tropic least-squares refinement was carried out with SHELXL-
97.[59] Geometrical calculations were made with PARST.[60] The
crystallographic plots were made with ORTEP.[61] A summary of
the crystallographic data of the complexes is given in Table 3.
CCDC-615303, -615304, and -615305 (for 4–6, respectively) con-
tain the supplementary crystallographic data for this paper. These
data can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
sisting of 0.25% bromophenol blue, 0.25% xylene cyanole, and
30% glycerol was added. The solution was then subjected to elec-
trophoresis on a 0.8% agarose gel in 0.5ϫ TBE buffer (0.045  tris,
0.045  boric acid, and 1 m EDTA) containing 2 µL per 100 mL
of a solution of ethidium bromide (10 mgmL–1) at 80 V for about
2 h. The gel was photographed on a capturing gel printer plus TDI.
To test for the presence of reactive oxygen species (ROS) generated
during strand scission, various scavengers were added to the reac-
tion mixtures. The scavengers used were Tiron (10 m), DMSO
(1 ), tert-butyl alcohol (1 ), potassium iodide (0.4 ), urea
(0.4 ), sodium azide (100 m), DABCO (0.4 ), and catalase (6.5
units). An assay in the presence of the minor groove binder dista-
mycin (8 µ) and the major groove binder methyl green (2.5 µL of
a 0.01 mgmL–1 solution) was also performed. Neocuproine, a li-
gand that chelates strongly to copper(I), was tested at 75 µ. Sam-
ples were treated as described above.
Viscosity Measurements: Viscosity experiments were carried out
with a Cannon 25 L97 viscometer immersed in a Julabo thermo-
statted water bath maintained at 25.0Ϯ0.1 °C. Solutions of the
complexes (1–10 µ) in cacodylate buffer (pH 6.0) were added to a
calf-thymus DNA (50 µ base pairs) solution in cacodylate buffer.
The flow times were measured with a digital stopwatch. The flow
measurements were performed in triplicate. Data are presented as
(η/η0)1/3 vs. the ratio of the concentration of complex and DNA,
where η is the viscosity of DNA in the presence of the complex
and η0 is the viscosity of DNA alone. Viscosity values were calcu-
lated from the observed flow time of a DNA-containing solution
corrected for the flow time of buffer alone (t0), η = t – t0.
Hydroxyl Radical Assay by 2-Deoxy-D-ribose Degradation: Hy-
droxyl radical production by the complexes was assayed by degra-
dation of 2-deoxy--ribose followed by quantitation of the 2-thiob-
arbituric acid adduct.[62] A 100 µ solution of metal complex was
treated at room temperature with 4 m 2-deoxy--ribose with acti-
vation by 50 m sodium ascorbate and H2O2. Reactions were per-
formed in a 4 mL volume of 0.1  cacodylate buffer (pH 6.0). Reac-
tion solutions were quenched with EDTA (2.4 m final concentra-
tion) and stored in ice for 5 min. Each aliquot was then treated
with 2-thiobarbituric acid (1.1 m final concentration) at 95 °C for
15 min. The chromophore concentration was measured from its
fluorescence intensity at 553 nm (excitation at 532 nm). The fluo-
rescence intensity was used due to its sensitivity and also to the
fact that ascorbate by-products absorb in the visible region and
therefore interfere with visible quantitation of the thiobarbituric
adduct. Fluorescence intensity was converted to concentration with
a standard fluorescence curve constructed from known concentra-
tions of the chromophore, synthesized from malondialdehyde and
2-thiobarbituric acid.[63] Blank reactions lacking a cleavage agent
were included in each run and the blank intensity was subtracted
from each experimental point.
Thermal Denaturation Experiments: DNA-melting experiments
were carried out by monitoring the absorbance (260 nm) of CT-
DNA (100 µ base-pairs) at different temperatures in the absence
and presence of the complexes in a 4:1 DNA/complex ratio. Mea-
surements were performed with an Agilent 8453 UV/Vis spectro-
photometer equipped with a Peltier temperature-controlled sample
cell and driver (Agilent 89090A). The solution containing the ter-
nary complex and CT-DNA in 0.1  cacodylate buffer (2 m NaCl,
pH 6.0) was stirred continuously and heated with a rate of tempera-
ture increase of 1 °Cmin–1. The temperature interval studied
ranged from 25 to 90 °C.
pUC18 DNA Cleavage: A typical reaction with the complexes was
undertaken by mixing 7 µL of 0.1  cacodylate buffer (pH 6.0),
1 µL of pUC18 (0.25 µgµL–1), 6 µL of a solution of the tested com-
plex at increasing concentrations between 20 and 100 µ, and 6 µL
of ascorbate (100-fold molar excess relative to the concentration of
the complex) in cacodylate buffer. The mixtures were allowed to
stand for 1 h at 37 °C, then 3 µL of a quench buffer solution con-
832
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 822–834

Efficient DNA Cleavage Induced by Copper(II) Complexes
FULL PAPER
[28]
[29]
[30]
[31]
[32]
[33]
[34]
A. Cantarero, J. M. Amigo, J. Faus, M. Julve, T. Debaerde-
maeker, J. Chem. Soc., Dalton Trans. 1988, 2033–2039.
J. V. Folgado, E. Martinez-Tamayo, A. Beltrán-Porter, D.
Beltrán-Porter, Polyhedron 1989, 8, 1077–1083.
A. Kamiyama, T. Noguchi, T. Kajiwara, T. Ito, Inorg. Chem.
2002, 41, 507–512.
Supporting Information (see also the footnote on the first page of
this article): IR spectra of [Cu(ptN2)(for)]·3H2O (3) and the [Cu-
(ptO2)(for)(H2O)] (4; Figure S1); UV/Vis spectra of aqueous solu-
tions of [Cu(ptN2)(OAc)]·3H2O (1) and [Cu(ptO2)(OAc)(H2O)]·
H2O (2; Figure S2); EPR spectra of [Cu(ptN2)(for)]·3H2O (3) and
[Cu(ptO2)(for)(H2O)] (4; Figure S3).
T. Glaser, T. Lugger, R. Fröhlich, Eur. J. Inorg. Chem. 2004,
394–400.
B. Casellas, F. Constantino, A. Mandonnet, A. Caneschi, D.
Gatteschi, Inorg. Chim. Acta 2005, 358, 177–185.
S. A. Cotton, V. Franckevicius, M. F. Mahon, L. L. Ooi, P. R.
Raithby, S. A. Teat, Polyhedron 2006, 25, 1057–1068.
S. R. Dey, C. R. Choudhury, S. P. Dey, D. K. Dey, N. Mondal,
S. O. G. Mahalli, K. M. A. Malik, S. Mitra, J. Chem. Res. (S)
2002, 496–499.
Acknowledgments
J. B., G. A., and J. L. G.-G acknowledge financial support from the
Spanish Centro de Investigación Científica y Tecnólogica (CICYT)
(CTQ2004-03735). M. G.-A. wishes to thank the city government
of Valencia (Spain) for a Carmen and Severo Ochoa postdoctoral
fellowship. B. M. acknowledges financial support from the Junta
de Castilla y León (SA056A05).
[35]
[36]
[37]
T. Kawijara, A. Kamiyama, T. Ito, Chem. Commun. 2002,
1256–1257.
N. Gupta, N. Grover, G. A. Neyhart, P. Singh, H. B. Thorp,
Inorg. Chem. 1993, 32, 310–316.
M. González-Álvarez, G. Alzuet, J. Borrás, B. Macías, M.
Del Olmo, M. Liu-González, F. Sanz, J. Inorg. Biochem. 2002,
89, 29–35.
B. Macías, I. García, M. J. Villa, M. González-Álvarez, J.
Borrás, A. Castiñeiras, J. Inorg. Biochem. 2003, 96, 367–374.
M. González-Álvarez, G. Alzuet, J. Borrás, B. Macías, A. Cas-
tiñeiras, Inorg. Chem. 2003, 42, 2992–2998.
M. González-Álvarez, G. Alzuet, J. Borrás, M. Pitié, B. Meun-
ier, J. Biol. Inorg. Chem. 2003, 8, 644–652.
B. Macías, M. V. Villa, F. Sanz, J. Borrás, M. González-Álva-
rez, G. Alzuet, J. Inorg. Biochem. 2005, 99, 1441–1448.
R. Cejudo, G. Alzuet, M. González-Álvarez, J. L. García-Gim-
énez, J. Borrás, M. Liu-González, J. Inorg. Biochem. 2006, 100,
70–79.
A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn, C. G. Ver-
schoor, J. Chem. Soc., Dalton Trans. 1984, 1349–1356.
G. A. Jefrey, An Introduction to Hydrogen Bonding, Oxford
University Press, Oxford, 1997.
WINEPR-Simphonia 1. 25 Bruker Analytic Gmbh, Kalrsruhe,
Germany, 1994–96.
J. Liu, T. Zhang, L. Que, H. Zhou, Q. Zhang, J. Liangnian, J.
Inorg. Biochem. 2002, 91, 269–276.
[1] S. S. Wallace, Int. J. Radiat. Biol. 1994, 66, 579–589.
[2] K. Z. Guyton, T. W. Kensler, Br. Med. Bull. 1993, 49, 523–544.
[3] O. I. Arouma, B. Haliwell, E. Gajewski, M. Dizdaroglu, Bi-
ochem. J. 1991, 273, 601–604.
[4] M. Dizdaroglu, G. Rao, D. A. Beary, R. A. LaBiche, K. J.
Hardy, Arch. Biochem. Biophys. 1991, 285, 317–324.
[5] S. E. Bryant, D. A. Vizard, D. A. Beary, R. A. LaBiche, K. J.
Hardy, Nucleic Acid Res. 1981, 9, 5811–5823.
[6] C. D. Lewis, U. K. Laemli, Cell 1982, 29, 171–181.
[7] A. M. George, S. A. Sabovijev, L. E. Hart, W. A. Cramp, G.
Harris, S. Homsey, Br. J. Cancer 1987, 55, 141–144.
[8] R. Stoewe, W. A. Prutz, Free Rad. Biol. Med. 1987, 3, 97–105.
[9] M. Dizdaroglu, O. I. Arouma, B. Haliwell, Biochemistry 1990,
29, 8447–8451.
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[10] M. Masarwa, H. Cohen, D. Meyerstein, D. L. Hickman, A.
Bakac, J. H. Espenson, J. Am. Chem. Soc. 1988, 110, 4293–
4297.
[11] J. Sagripanti, K. H. Kraemer, J. Biol. Chem. 1989, 264, 1729–
1734.
[12] K. Yamamoto, S. Kawanishi, J. Biol. Chem. 1989, 264, 15435–
15440.
[13] G. R. A. Johnson, N. B. Nazat, R. A. Saadalla-Nazhat, J.
Chem. Soc., Faraday Trans. 1 1988, 84, 501–510.
[14] N. C. Thomas, B. L. Iolcy, A. L. Rheingold, Inorg. Chem. 1988,
27, 3426–3429.
A. Mazumder, C. L. Sutton, D. S. Sigman, Inorg. Chem. 1993,
32, 3516–3560.
S. H. Chiou, N. Ohtsu, K. G. Bensch, in Biological and Inor-
ganic Copper Chemistry (Eds.: D. Karlin, J. Zubieta), Adenine
Press, New York, 1985, pp. 119–123.
[15] S. Chirayil, V. Hegde, Y. Jahng, R. P. Thummel, Inorg. Chem.
1991, 30, 2821–2823.
[49]
[50]
[16] N. Gupta, N. Grover, G. A. Neyhart, P. Singh, H. B. Thorp,
Inorg. Chem. 1993, 32, 310–316.
C. A. Detmer III, F. V. Pamatong, J. R. Bocarsly, Inorg. Chem.
1996, 35, 6292–6298.
C. A. Detmer III, F. V. Pamatong, J. R. Bocarsly, Inorg. Chem.
1997, 36, 3676–3682.
C. J. Burrows, J. G. Muller, Chem. Rev. 1998, 98, 1109–1152.
B. Haliwell, J. M. C. Gutteridge, in CRC Handbook of Methods
for Oxygen Radical Research (Ed.: R. A. Greenwald), CRC
Press, Boca Raton, FL, 1985, p. 177.
F. V. Pamatong, C. A. Detmer III, J. R. Bocarsly, J. Am. Chem.
Soc. 1996, 118, 5339–5345.
D. R. Graham, L. E. Marshall, K. A. Reich, D. S. Sigman, J.
Am. Chem. Soc. 1980, 102, 5419–5421.
D. S. Sigman, Acc. Chem. Res. 1986, 19, 180–186.
COLLECT, Nonius BV, 1997–2000.
DENZO-SCALEPACK, Z. Otwinowski, W. Minor, “Pro-
cessing of X-ray Diffraction Data Collected in Oscillation
Mode”, Methods in Enzymology, Volume 276: Macromolecu-
lar Crystallography, part A, p. 307–326 (Eds.: C. W. Carter Jr,
R. M. Sweet), Academic Press, 1997.
SIR97: A. Altomare, M. C. Burla, M. Camalli, G. L. Cas-
carano, C. Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G.
Polidori, R. Spagna, J. Appl. Crystallogr. 1999, 32, 115–119.
[17] R. M. Berger, J. R. Holcombe, Inorg. Chim. Acta 1995, 232,
217–221.
[18] R. M. Berger, D. D. Ellis II, Inorg. Chim. Acta 1996, 241, 1–4.
[19] E. M. Smolin, L. Rapoport, S-Triazines and Derivatives, Inter-
science, New York, 1959, p. 163.
[20] E. I. Lerner, S. J. Lippard, J. Am. Chem. Soc. 1976, 98, 5397–
5398.
[21] E. I. Lerner, S. J. Lippard, Inorg. Chem. 1977, 16, 1546–1551.
[22] J. Faus, M. Julve, J. M. Amigo, T. Debaerdemaeker, J. Chem.
Soc., Dalton Trans. 1989, 1681–1687.
[23] P. A. Gillard, P. A. Williams, Transition Met. Chem. 1979, 4,
18–23.
[24] P. Paul, B. Tyagi, M. M. Bhadbhade, E. Suresh, J. Chem. Soc.,
Dalton Trans. 1997, 2273–2277.
[25] P. Paul, B. Tyagi, A. K. Bilakhiya, M. M. Bhadbade, G. Rama-
chandaiah, Inorg. Chem. 1998, 37, 5733–5742.
[26] X.-P. Zhou, D. Li, S.-L. Zheng, X. Zhang, T. Wu, Inorg. Chem.
2006, 45, 7119–7125.
[27] J. V. Folgado, E. Coronado, D. Beltrán-Porter, R. Burriel, A.
Fuertes, C. Miratvilles, J. Chem. Soc., Dalton Trans. 1988,
3041–3045.
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
Eur. J. Inorg. Chem. 2007, 822–834
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
833

J. Borrás et al.
FULL PAPER
[59] SHELX97, G. M. Sheldrick, SHELX97. Programs for Crystal
Structure Analysis (Release 97-2), University of Göttingen,
Germany, 1997.
[60] PARST: a) M. Nardelli, Comput. Chem. 1983, 7, 95–97; b) M.
Nardelli, J. Appl. Crystallogr. 1995, 28, 659.
[61] ORTEP3 for Windows: L. J. Farrugia, J. Appl. Crystallogr.
1997, 30, 565.
[62] B. Halliwell, J. M. C. Gutteridge, in CRC Handbook of methods
for Oxygen Radical Research (Ed.: R. A. Greenwald), CRC
Press, Boca Raton, FL, USA, 1985, p. 177.
[63] R. M. Burger, A. R. Berkowith, J. Peisach, S. B. Horwitz, J.
Biol. Chem. 1980, 255, 11832–11838.
Received: September 5, 2006
Published Online: January 12, 2007
834
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 822–834