Synthesis, characterization, DNA-binding study and anticancer properties of ternary metal(ii) complexes of edda and an intercalating ligand

Article abstract of DOI:10.1039/b709269e

A series of ternary metal(ii) complexes {M(phen)(edda); 1a (Cu), 1b (Co), 1c (Zn), 1d (Ni); H₂edda = N,N′-ethylenediaminediacetic acid} of N,N′-ethylene-bridged diglycine and 1,10-phenanthroline were synthesized and characterized by elemental analysis, FTIR, UV-visible spectroscopy and magnetic susceptibility measurement. The interaction of these complexes with DNA was investigated using CD and EPR spectroscopy. MTT assay results of 1a-1c, screened on MCF-7 cancer cell lines, show that synergy between the metal and ligands results in significant enhancement of their antiproliferative properties. Preliminary results from apoptosis and cell cycle analyses with flow cytometry are reported. 1c seems to be able to induce cell cycle arrest at G₀/G₁. The crystal structure of 1a is also included. The Royal Society of Chemistry.

Full text of DOI:10.1039/b709269e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Synthesis, characterization, DNA-binding study and anticancer properties of
ternary metal(II) complexes of edda and an intercalating ligand†‡
a
a
a
b
c
d
Chew Hee Ng,* King Chow Kong, Sze Tin Von, Pauline Balraj, Paul Jensen, Eswary Thirthagiri,
e
e
Hirokazu Hamada and Makoto Chikira
Received 19th June 2007, Accepted 27th September 2007
First published as an Advance Article on the web 24th October 2007
DOI: 10.1039/b709269e
A series of ternary metal(II) complexes {M(phen)(edda); 1a (Cu), 1b (Co), 1c (Zn), 1d (Ni); H edda =
2
ꢀ
ꢀ
N,N -ethylenediaminediacetic acid} of N,N -ethylene-bridged diglycine and 1,10-phenanthroline were
synthesized and characterized by elemental analysis, FTIR, UV-visible spectroscopy and magnetic
susceptibility measurement. The interaction of these complexes with DNA was investigated using CD
and EPR spectroscopy. MTT assay results of 1a–1c, screened on MCF-7 cancer cell lines, show that
synergy between the metal and ligands results in significant enhancement of their antiproliferative
properties. Preliminary results from apoptosis and cell cycle analyses with flow cytometry are reported.
1c seems to be able to induce cell cycle arrest at G
0
/G . The crystal structure of 1a is also included.
1
1
. Introduction
varying the type of metal ion. The selection of edda as a second
ligand may enhance the affinity of the ternary complexes towards
DNA through the formation of hydrogen bonding between the
edda and the DNA. Unlike square planar platinum(II) complexes,
the six-coordinate M(phen)(edda) complexes are likely to bind
to polyanionic DNA via non-covalent outer-sphere coordination.
Furthermore, neutral metal complexes should be more lipophilic
than the corresponding ionic metal complexes and their greater
uptake by cells can lead to increased cytotoxicity.
Since the discovery of cisplatin for cancer treatment, numerous
transition metal complexes have been synthesized and screened
for their anticancer properties. One strategy in such efforts is to
synergise the beneficial effect of the ligand and the activity of the
metal to produce a complex with enhanced activity. In the present
post genomic era, another strategy focusses on the interaction
1
2
of metal drugs with proteins that regulate apoptosis. Here,
selection of organic compounds of known therapeutic action and
linking them to the metal has been successfully applied, resulting
in platinum and ruthenium anti-metastatic complexes. A recent
theoretical study, using clustering analysis and self-organising
maps, of more than 1000 metal containing compounds with
potential anticancer properties concluded that their cytotoxicity
is determined by the identity of the metal and the organic ligand,
and the target-specificity can be accomplished by the right metal–
Thus far, most anticancer metal complexes containing
4–9
1
,10-phenanthroline are cationic.
These include tris(1,10-
4
phenanthroline)lanthanum(III) trithiocyanate, platinum(II) com-
plexes with 1,10-phenanthroline and acyclovir or penciclovir,
5
platinum(II) or palladium(II) complexes of 1,10-phenanthroline
6
and
a
series of amino acids, tris(1,10-phenanthroline)-
7
8,9
chromium(III), and copper(II) complexes. Coyle, et al. recently
reported possible differences in cell death mechanism in mam-
malian cells incubated with neutral ternary metal complexes
3
ligand combination.
We have been systematically experimenting with such com-
binations. The present investigation involves the synthesis of a
series of neutral ternary transition metal complexes (Co, Ni, Cu,
Zn) with a moderately cytotoxic intercalating bidentate ligand,
(
Cu(II), Mn(II) and Ag(I)) of 1,10-phenanthroline and malonic
10
acid. [Cu(phen)
2
(mal)]·2H
2
O induced apoptosis while metal-free
phen, and the Mn(II) and Ag(I) ternary complexes did not induced
apoptosis but merely caused non-specific DNA degradation. The
present paper details the synthesis and characterization of our
set of neutral ternary metal complexes, together with their DNA-
binding study and anticancer properties towards breast cancer cell
line, MCF-7.
1
,10-phenanthroline, and a non-toxic tetradentate amino acid,
ꢀ
N,N-ethylene-bridged diglycine (more commonly called N,N -
ethylenediaminediacetic acid, H edda), to evaluate the effect of
2
a
Faculty of Engineering and Science, Universiti Tunku Abdul Rah-
man, 53300, Kuala Lumpur, Malaysia. E-mail: ngch@mail.utar.edu.my;
Fax: +60 3 4107 9803; Tel: +60 3 4107 9802
2. Results and discussion
b
Molecular Pathology Unit, Institute for Medical Research, 50588, Kuala
Lumpur, Malaysia
2.1 Synthesis and characterization
c
School of Chemistry, University of Sydney, NSW, 2006, Australia
d
The ternary complexes were prepared either by first reacting
freshly prepared metal hydroxide with N,N -ethylenediamine-
diacetic acid (H edda) and then finally by adding a methanolic or
ethanolic aqueous solution of 1,10-phenanthroline (phen) or by an
in situ method where the sequence of added reactants is H edda,
NaOH, metal(II) salt and alcoholic aqueous phen. The ternary
metal(II) complexes of the edda and phen precipitated out on the
Cancer Research Initiatives Foundation, 47500, Kuala Lumpur, Malaysia
ꢀ
e
Department of Applied Chemistry, Chuo University, Kasuga 1-13-27,
Bunkyo-ku, Tokyo, 112-8551, Japan
2
†
CCDC reference number 636037. For crystallographic data in CIF or
other electronic format see DOI: 10.1039/b709269e
Electronic supplementary information (ESI) available: Plots of % cell
viability from MTT assay of phen and Cu(edda) on MCF-7. See DOI:
0.1039/b709269e
2
‡
1
This journal is © The Royal Society of Chemistry 2008
Dalton Trans., 2008, 447–454 | 447

View Article Online
Table 1 UV-visible spectral data of aqueous solutions of 1a–1d
−
1
−1
a
−1
−1
a
−1
−1
−1
−1
Complexes
k
1a/nm (e/M cm
)
k
1b/nm (e/M cm
)
k
2
/nm (e/M cm
)
k
3
/nm (e/M cm
)
1
1
1
1
a (Cu)
b (Co)
c (Zn)
d (Ni)
225 (21 191)
226 (28 878)
227 (32 015)
227 (32 047)
269 (18 489)
269 (25 124)
267 (24 730)
270 (29 694)
682 (94)
564 (11)
926 (48)
942 (6)
915 (10)
a
1
k values: p–p* transition bands of phen.
same day or upon slow evaporation of the solution mixtures. FTIR
spectra of these complexes show the presence of asymmetrically
and symmetrically coordinated carboxylate stretching frequencies
−
1
11
at ∼1600 and ∼1400 cm respectively, and of coordinated phen
−
1 12
ligand peaks at ∼(1517–1522), 852 and ∼727 cm . Elemental
analysis of the isolated complexes, 1, reveals a general formula,
[
M(phen)(edda)]·xH
The copper(II) complex, 1a, has a magnetic moment of 1.91 l
while the cobalt(II), 1b, and nickel(II), 1d, complexes have leff values
of 4.89 and 3.06 l respectively. The observed magnetic moments
are in the range observed for other octahedral copper(II),
2
O.
B
B
1
3,14
13–18
13–18
cobalt(II)
than their respective spin-only magnetic moments, l
.24 and 2.83 l due to significant orbital angular momentum
and nickel(II)
complexes. These values are higher
s
, of 1.73,
4
B
contributions. The diamagnetic zinc(II) complex, 1c, may be
octahedral as evidenced by the very close similarity of its FTIR
spectrum with those of 1b and 1d. Based on this magnetic property,
these four complexes can be classified as being paramagnetic or
diamagnetic octahedral complexes.
The ternary complexes seem to be formed from the substitution
of the cis-coordinated water molecules of the initially-formed
[
M(edda)(H
2
O) ] by phen. This is evidenced by comparison of
2
ꢀ
Fig. 1 Molecular structure of (N,N -ethylenediaminediacetato)(1,10-
phenanthroline)copper(II) pentahydrate, 1a. Selected bond lengths ( A˚ ):
the visible spectra of the aqueous solutions of 1 (Table 1)
prepared above and the aqueous solution mixtures (data not
included) obtained by adding phen to previously synthesized cis-
a-[M(edda)(H
spectra are identical.
Cu(1)–O(1) 2.3957(12), Cu(1)–N(4) 2.0263(13), Cu(1)–N(2) 2.0328(13),
Cu(1)–N(1) 2.0369(13), Cu(1)–N(3) 2.0429(13), Cu(1)–O(3) 2.3140(11).
19
2
O)
2
] (M = Cu, Ni, Co) complexes. Their visible
N(2)-containing fragment (of the phen ligand), and this results
in significant deviation from ideal planarity of the equatorial
ligated atoms. The other metal complexes (1b, 1c and 1d), with
2
.2 Crystal structure of 1a
A suitable crystal of 1a was obtained by allowing the reaction
mixture, which was heated in a water bath at 55 C, to cool
◦
symmetric filling of the e
orbitals (dz2 , dx2−y2 ) are free from the
g
overnight. The crystal structure of 1a (Fig. 1) confirms the
octahedral geometry around the copper atom. The coordinated
carboxylate oxygen atoms (O(1) and O(3)) of the tetradentate edda
ligand are trans to each other, with a O(1)–Cu(1)–O(3) angle of
above Jahn–Teller distortion and their ethylenediamine chelate
rings should be conformationally different and their N(1)–N(2)–
N(3)–N(4) equatorial atoms should be more planar, as is the case
19
for the equatorial N
2
O
2
atoms of [Co(edda)(H
2
2
O) ]. Thus, a
◦
˚
1
74.02(4) and an average axial Cu–O bond length of 2.35 A.
change in the type of metal(II) ion modulates subtle changes in
the conformation of the octahedral geometry of the complexes.
As each edda moiety has H-donor (N(3), N(4)) and H-acceptor
(O(2), O(4)) atoms for H-bonding, the above two types of
chelate ring conformation may modulate the possible H-bonding
interaction of the edda moieties with adjacent base pairs of DNA
if the chelated 1,10-phenanthroline moeity intercalates between
the DNA base pairs. The resultant conformational change moves
the hydrogen bonding sites of the edda moiety away or towards
the DNA. Besides the above influence on H-bonding interactions
with DNA due to the presence or absence of tetragonal distortion,
these octahedral complexes can be classified into two types based
on the nature of the distortion of the octahedral geometry.
◦
The bite angle of the ethylenediamine chelate ring is 86.03(5) in
contrast to those of the glycinate rings which are 75.25(5) and
◦
◦
7
6.95(5) . The corresponding glycinate rings of the octahedral di-
ꢀ
aqua[N,N -ethylenedi(N-hydroxymethylglycinato)]copper(II), de-
rived from the N-hydroxymethylation of Cu(edda), have a bigger
◦
20
bite angle (79.5(1) ). These, together with other angles about the
copper atom, show severe distortion of the octahedral geometry.
Interestingly, the tetragonal elongation of the axial bonds (Cu–
O) in 1a lifts up the N(3) atom and pulls down the N(4) atom
of the ethylenediamine chelate ring, resulting in an envelope
conformation of this chelate ring, such that the plane of the N(4)–
N(3)-containing portion is severely tilted from that of the N(1)–
4
48 | Dalton Trans., 2008, 447–454
This journal is © The Royal Society of Chemistry 2008

View Article Online
2
.3 DNA binding study
One of the modes of action of anticancer metal-based compounds
is the targeting of the DNA of cancer cells. Thus, some basic
interaction studies of metal complex–DNA are of relevance. The
CD spectra (Fig. 2(a) ) of the series of complexes 1a–1d with
40 lM B-form calf thymus DNA (CT DNA) in 20 mM Hepes
buffer at pH 7, containing 30 mM NaCl, still shows a positive
band at ∼276 nm due to base stacking and a negative band at
∼
245 nm due to DNA helicity. At a DNA : complex mole ratio
of 1 : 1, the complexes 1a (Cu), 1b (Co) and 1c (Zn) caused
significant enhancement of the DNA band at 276 nm while 1d
(
Ni) produced a minor increase in molar ellipticity. No significant
shift in the position of the band due to DNA helicity is observed for
all complexes, suggesting retention of the B-form conformation.
The above enhancement of molar ellipticity at ∼276 nm suggests
partial intercalation of 1a, 1b and 1c, and this is also observed
for similar ternary complexes, [Cu(phen)(L-threo)(H
2
O)](ClO )
4
and lathanum(III) complexes containing the phen ligand or its
21,22
derivative.
◦
Fig. 3 EPR spectra of 1a (a) in frozen solution at −150 C, (b) in
◦
DNA-pellet at −150 C, and (c) in A-form DNA fiber at room temperature.
U is the angle beween the fiber axis and the static magnetic field.
interaction of the copper complex 1a with highly oriented DNA
fibre. The EPR spectrum of 1a in frozen aqueous solution at pH 7
(
Fig. 3(a)) showed a typical pattern characteristic of a tetragonal
Cu(II) complex with g
ꢁ
= 2.241 and A
ꢁ
= 17.0 mT. The peaks
◦
observed in the EPR spectrum of the DNA pellet at −150
C
(
ꢁ ꢁ
Fig. 3(b) ) correspond to the g and A signals observed for the
frozen solution. However, the EPR spectra of 1a in DNA pellet
and in A-form DNA fiber at room temperature (Fig. 3(c)) were
broadened considerably, indicating that the copper coordination
sphere was deformed inhomogenously on the DNA. The observed
small U dependence of the EPR spectra of the A-form DNA fiber
indicates that the complex is oriented randomly on the fiber.
In contrast to the inhomogenous broadening of the spectra
of 1a in DNA pellet or in A-form DNA fiber, the EPR line
shape changed dramatically after the conformational change of
the DNA fiber from A- to B-form in a humid atmosphere at
ambient temperature (Fig. 4(a) ). The conspicuous U dependence
of the EPR spectra of 1a in B-form DNA fiber indicates that 1a
was reoriented stereo-specifically with respect to the double helical
Fig. 2 Circular dichroism spectra of CT DNA without and with
M(phen)(edda) at 1 : 1 ratio (a) and at 1 : 4 mole ratio (b).
When the DNA : complex ratio was increased to 1 : 4 (Fig. 2(b)),
the difference in enhancement of the band at 276 nm is more
pronounced, indicating a difference in the extent of intercalation
by the various metal complexes. The order of increasing interca-
lation is 1d < 1b < 1a < 1c. The zinc complex, 1c, caused the
most significant red shift and enhancement of the band due to
DNA helicity at 245 nm. A new positive band at 232 nm also
appeared for 1c. Surprisingly, the increased concentration of 1a
caused a slight decrease in intensity of the DNA helicity band at
◦
DNA axis. The intense four line peaks observed at U = 0 indicate
that one of the g tensor axes (g
oriented along the DNA fiber axis. The considerable decrease in
the A value from that observed in frozen solution (A
suggests that the tetragonal coordination structure of 1a changed
considerably in the B-form DNA fiber. It should be noted that
additional signals (the arrows in Fig. 4(a)) were observed at the
ꢁ
= 2.20 and A = 12.0 mT) was
ꢁ
ꢁ
= 17.0 mT)
◦
low-field end of the spectrum of U = 0 and at the high-field end
◦
2
45 nm instead of further enhancement as observed for 1b and
of the spectrum at U = 90 . These signals correspond to another
1c. The nickel complex, 1d, seems not to significantly effect either
species bound with a different coordination structure.
◦
DNA band (245 nm and ∼275 nm). Thus, the above CD analysis
revealed that changing the type of metal ion in 1 affects its binding
interaction with DNA.
To elucidate further understanding of the above DNA binding
differences, we performed an X-band EPR spectral study on the
Freezing of the B-form fiber at −150 C caused another
dramatic change in the EPR line shapes (Fig. 4(b)). The intense
◦
four line signals observed at U = 0 at room temperature were
replaced with broad signals over the range 290–330 mT and with
several weak peaks in the range 250–285 mT. It has been observed
This journal is © The Royal Society of Chemistry 2008
Dalton Trans., 2008, 447–454 | 449

View Article Online
It is obvious that the two axially coordinating carboxyl oxygen
atoms of the edda moiety observed in the crystal structure
prevented deeper intercalation of the phenanthroline moiety and
that the interaction of the carboxyl groups with DNA caused
in return the rearrangement of the coordination sphere of the
complex on the DNA. It is also possible that solvated water
molecules around the carboxyl groups in 1a exerted stress on the
copper coordination sphere when the DNA fibers were frozen to
form micro icebergs in the fibers, resulting in the randomization
of the orientation of the copper coordination planes.
Though a part of 1a binds to DNA retaining the initial
ꢁ
tetragonal structure on the DNA as judged by the g signals in the
EPR spectra of DNA-pellet, one can reasonably presume that the
rhombic distortion of 1a on the DNA is the key to understanding
the function of this complex.
2
.4 Antiproliferative and cell cycle analysis
The antiproliferative activity of M(phen)(edda) complexes on
breast cancer cell line MCF-7 was assessed using an MTT assay.
The IC50 values of the Cu complex 1a, Co complex 1b and Zn
complex 1c were obtained by plotting the cell viability against
concentration of the complexes (Fig. 5, Table 2). Due to an initial
solubility problem, MTT assay was not done for 1d. From the
plots, the average IC50 values (from two pooled results) were 3.8 lM
for 1a, 11.4 lM for 1b and 7.0 lM for 1c.
Fig. 4 EPR spectra of 1a in B-form DNA fiber. (a) At room temperature,
◦
(
b) −150 C.
that the freezing of B-form DNA fibers sometimes results in the
randomization of the orientation of the complexes due to the
23,24
formation of icebergs in the DNA fibers.
The intense peak
The trend in the antiproliferative ability of the complexes is 1a >
that emerged at 295 mT (marked as ꢀ in Fig. 4) indicates that
the complex was distorted toward rhombic symmetry. Though
the line shapes observed at low temperature did not change with
U so conspicuously as those at room temperature, the intensity
of the peaks marked with asterisks decreased with increasing U,
indicating that the species has still some preferred orientation on
the DNA fiber. On the other hand, the intensity of the peaks
marked with a filled circle (᭹) does not change with U, indicating
1
c > 1b. MTT assay carried out by incubating the cells with 1,10-
phenanthroline gave an IC50 value of 74 lM (Table 2; ESI as Data
‡). The enhanced IC50 values for the metal complexes suggest that
1
there is synergy between the cytotoxic phen ligand with these metal
cations. In other words, chelation of the phen ligand to these metal
cations enhanced its cytotoxicity. Such cytotoxic enhancement
resulting from chelation of phen to metal cation in binary or
ternary complexes is seldom investigated, as far as we know.
One such reported synergy involves the octahedral [La(phen)]
complex. Ranford et al. also found that incorparating phen into
the copper(II) complexes of 3,5-disubstituted salicylates enhanced
their in vitro antitumor properties. A similar synergic metal–
that the species is randomly oriented on the fiber. The g
values estimated for latter species are very similar to the g
values estimated for the species in frozen solution and DNA
ꢁ
and A
ꢁ
3+
and
4
ꢁ
A
ꢁ
pellet. Interestingly, the magnetic field of the peak observed at the
26
◦
lowest end of the spectra at U = 0 in Fig. 3(a) almost coincides
ligand combination is also reported for a set of dicopper(II)
complexes where the IC50 values of the complexes for MCF-7 are
in the range 24–54 lM while both the copper(II) chloride and the
with those of the peaks at the lowest end of the spectra of frozen
solution and DNA-pellet. Therefore, these signals can be assigned
as those of the species bound to the DNA keeping the tetragonal
structure observed in the crystal or in frozen solution.
27
various non-polypyridyl proligands have IC50 > 100 lM. In con-
trast, the recent discovery of organic copper complexes as a new
class of proteosome-inhibitors and apoptosis inducer for human
cancer cells reported the occurrence of the reverse effect for the
chosen metal–ligand combination. Here, the 8-hydroxyquinoline
slightly inhibits the proteosomal chymotrypsin-like activity of the
2
+
It has been shown that ternary complexes of [Cu(phen)] with
various amino acids bind to DNA with several different binding
25
modes. One is intercalative and the other is groove bound.
In some cases, the amino acid was replaced by coordinating
groups in DNA. However, it has never been observed that the
amino acid ternary complexes were deformed so much toward
rhombic form upon binding to DNA. It is evident that the edda
moiety in 1a was preserved at the copper coordination site on the
28
Table 2 IC50 values for 1a–1c on MCF-7
Complexes
1
2
Average
2
+
DNA, because no signals from the dissociated [Cu(phen)] (g
ꢁ
=
−
1
2
.29, A
ꢁ
= 0.017 cm ) were detected in the EPR spectra. These
1a (Cu)
3.5
11.4
7.7
4.0
11.3
6.3
3.8
11.4
7.0
1
1
b (Co)
results indicate that 1a binds to DNA, partially intercalating the
phenanthroline plane parallel to the DNA base-pair plane with
the coordination structure of the phen–edda moiety twisted to
form rhombic species. This partial intercalation is correlated by
the earlier CD data.
c (Zn)
a
phen
74.3
a
Cu(edda)
395.5
a
ESI Data 1 and Data 2.
4
50 | Dalton Trans., 2008, 447–454
This journal is © The Royal Society of Chemistry 2008

View Article Online
Fig. 5 Viability of MCF-7 cells in the presence of M(phen)(edda) complexes (Cu,1a; Co, 1b; Zn, 1c). Each complex induced a dose-dependent decrease
in cell viability. Plots 5(a) and 5(b) are from two independent sets of experiment.
CuBr
Devereux et al. found that incorporating phen into a binary
complex Cu(BZA) (benzoate, BZA), to form the water-insoluble
ternary complex [Cu(BZA) (phen)(H O)] greatly enhances its IC50
values from >200 and 130 lM to 9.5 and 21.3 lM for the Hep-
2
by forming the bis(8-hydroxyquinoline)copper(II) complex.
2
2
2
29
G
2
and A-498 cancer cell lines respectively. However, IC50 value
of this ternary complex is not significantly higher than that of
the metal-free phen (IC50 values for Hep-G and A-498 are 4.1
2
and 5.8 lM respectively) tested for the same cell lines (96 h
incubation time). However, our results show that the IC50 values
of the tested ternary M(phen)(edda) complexes (1a, 1b, and 1c)
are significantly enhanced compared to that of metal-free phen
for the MCF-7 cell line (72 h incubation time). The choice of the
second ligand may be crucial as we found that Cu(edda), phen
and 1a have IC50 values of 395.5, 74.3 and 3.8 lM. In fact, on-
going investigations substantiate the significance of choosing an
amino acid or its derivative as the second ligand and this will
be reported subsequently. In contrast, almost all the nine square
planar ternary platinum(II) complexes of phen and various amino
acids, [Pt(phen)(aa)]NO
3
tested (except [Pt(phen)(pro)]NO
3
), are
6
found to be not cytotoxic towards a human leukemia cell line.
Complex geometry may be a significant contributing factor in
ensuring a synergic effect in the rational design of anticancer
ternary metal complexes.
Fig. 6 Percentage of necrotic cells (UL), late apoptosis cells (UR), viable
cells (LL) and early apoptotic cells (LR) for untreated MCF-7 cells and
cells treated with 1a, 1b and 1c.
Investigation into the nature of the cell death mechanism
by incubating MCF-7 cells with the respective complexes by
using fluorescence activated cell sorting (FACS) analysis for both
apoptosis and cell cycle arrest is in progress. Preliminary results
with concentration of complexes lower than their respective IC50
values (except for 1c) are presented herein. FACS of induction of
apoptosis was undertaken for MCF-7 cells treated with cisplatin
as positive control, and the complexes 1a, 1b and 1c (Fig. 6).
Untreated cells showed a majority of viable healthy cells with 86%
untreated control (7%). MCF-7 cells treated with 2.8 lM 1a (Cu),
with 5.3 lM 1b (Co) and with 9.2 lM 1c (Zn) for 72 h showed
22%, 14% and 22% of the cell population at late apoptotic stage
respectively. The percentage of necrotic cells in the above MCF-7
cells treated with 1a, 1b and 1c were 5%, 4% and 7% respectively
compared with 1% for untreated cells. Thus, cell death induced by
the metal complexes 1a, 1b and 1c is mainly via apoptosis. The
total % cell death (apoptosis and necrosis) induced for 1a, 1b and
1c is 31%, 22% and 35% respectively.
−
1
population after 72 h incubation. Cells treated with 10 lg ml
of cisplatin for 72 h as positive control exhibited an increased
number of cells at late apoptotic stage (76%) compared to an
The cell cycle progression data of MCF-7 cells treated with
1a, 1b and 1c were obtained using a FACS analysis of DNA
This journal is © The Royal Society of Chemistry 2008
Dalton Trans., 2008, 447–454 | 451

View Article Online
Fig. 7 FACS analysis of cell cycle arrest of untreated MCF-7 cells and those treated with nocodazole (positive control), 1a, 1b and 1c. Percentage of
cells in sub G /G , G /G , S and G /M were determined after staining with propidium iodide (PI).
0 1 0 1 2
content (Fig 7). Cells in the G
DNA while those in the G2 and M phases have duplicated 4N
DNA. Those in the S phase are in the process of replicating their
0
/G
1
phase have unreplicated 2N
induced expression of p53 and p21 besides inducing a selective
4
decrease in cyclin B
1
. The octahedral La(III) complex exhibits
very promising results on >60 tumor cell models with low lM
values, and low toxicity test on rats. Further work is being carried
out to investigate further the molecular mechanism of cell death of
the M(phen)(edda) complexes, and confirm the above preliminary
observations.
DNA while those in the sub-G
0
/G phase have hypodiploid DNA
1
(
less than 2N). Our results showed that untreated MCF-7 cells
exhibited asynchronous cell cycle profile following 72 h in culture.
When the cells were treated with 1a, 1b and 1c at concentrations
lower their respective IC50 values, the cell cycle profiles of the
complexes obtained were quite similar, except that for 1c (Zn).
3
. Conclusions
For 1c, the cell population at sub-G
from 18% to 41% after 72 h treatment while the G
remained high at 45% compared to untreated cells which had
a corresponding population of 58% at G /G . Increases in cell
population at sub-G /G phase had been taken as evidence of
cells entering apoptosis.
phases; their combined S and G
from 20% to 14%, indicating blockage of cell progression into
S and G /M phases, and thus implying a certain proportion of
cells were arrested at G /G . Accumulation of cells at G /G has
0
/G
1
(hypodiploid) increased
0
/G population
1
The synthesized complexes are found to have octahedral geome-
tries, and they can be classified according to magnetic property
or tetragonal distortion. Binding of these complexes to DNA
is via partial intercalation of the phenanthroline moiety into
the base pairs, as evidenced by CD and EPR data. The edda
moeity prevents deeper intercalation of the phen of 1a and this
intercalation into DNA in turn results in rhombic distortion; this
may similarly occur for the other complexes. The CD data suggest
that the extent of intercalation of the octahedral M(phen)(edda)
complexes varies with the type of metal(II) ion and this seems to be
related to the absence or presence of Jahn–Teller distortion. The
enhanced cytotoxicity of the complexes towards MCF-7 cancer
cells, as evidenced by their IC50 values and in comparison with the
IC50 value of unchelated 1,10-phenanthroline, suggests synergic
combination of the two types of ligand and metal ion used.
Compared to the IC50 value (1.42 lM) of cisplatin for MCF-
0
1
0
1
4,30,31
There were no distinct S and G
2
/M
2
/M phases significantly dropped
2
0
1
0
1
been associated with activation of cell cycle checkpoint due to
4
DNA damage. In addition, 1b may also cause cell cycle arrest
at G
4% when the cells were treated with 1b for 72 h. Thus, both 1b
and 1c induced apoptosis as a major cell death pathway and cell
cycle arrest at G /G as a minor pathway. This mode of action is
in contrast to that of cisplatin which induces cell cycle arrest at
0
/G
1
as there is increase in G
0
/G
1
population from 58% to
6
0
1
32
33
G
2
/M phase. However, it seems that the complex 1a induced cell
death mainly via apoptosis. In this respect, the mode of action
of 1a is similar to the anticancer octahedral [Cu(phen) (mal)]
mal = malonic acid) complex. On the other hand, the ability
of 1c (Zn complex) to induce cell cycle arrest at G /G can be
complex which also
7, those of 1a–1c indicate significant synergic effect. It seems
that the choice of the second ligand is important in designing
effective anticancer ternary metal complexes. Besides such synergic
effect, the type of metal ion can affect the mode of action of these
anticancer metal complexes as exemplified by the ability of the
zinc complex, 1c, to induce cell cycle arrest.
2
10
(
H
2
0
1
3
+
compared to the octahedral [La(phen) ]
3
4
52 | Dalton Trans., 2008, 447–454
This journal is © The Royal Society of Chemistry 2008

View Article Online
4
. Experimental
with a Perkin Elmer 2400 CHN analyser. UV-visible spectroscopic
measurement of the aqueous solutions of the complexes was
carried out on a Perkin Elmer Lambda 40 spectrophotometer.
Circular dichroism spectra of the complexes, in 20 mM Hepes
buffer pH 7 containing 30 mM NaCl, were measured using
a JASCO J-820 CD spectrophotometer. Each sample solution
4.1 Synthesis of complexes
The metal(II) complexes of 1,10-phenanthroline and N,Nꢀ-
ethylenediaminediacetic acid were prepared either (i) by
reacting freshly prepared metal(II) hydroxide with N,Nꢀ-
ethylenediaminediacetic acid, followed by addition of methanolic
or ethanolic aqueous solutions of 1,10-phenanthroline, or (ii) by
an in situ method where the sequence of added reactants was
(3000 lL, 3 ml) was obtained by adding the required volume
of stock solution of calf thymus DNA (Calbiochem, Merck) and
metal complex dissolved in the above Hepes buffer, and finally
topping up the solution with an additional volume of the same
buffer. Two sets of complex–DNA solutions were used: the ratio
of the complex : DNA = 40 lM : 40 lM (1 : 1) in each solution
in one set of solutions while the mole ratio for the second set is
H
2
edda, NaOH, metal(II) salt and alcoholic aqueous phen. The
following is given as an example.
Freshly prepared Cu(OH) was obtained by reacting cold
aqueous CuSO O (0.54 g, 0.002 mol) with cold aqueous
2
4
·5H
2
160 lM : 40 lM (4 : 1).
NaOH (0.17 g, 0.004 mol). The solid hydroxide was added to a
suspension of N,N -ethylenediaminediacetic acid in about 30 ml.
of water. The resultant blue solution was filtered and a methanolic
solution of 1,10-phenanthroline (0.43 g, 0.002 mol) was then
added. The reaction mixture was heated in a water-bath at 55 C
ꢀ
4
.3 Preparation of DNA-pellets and fibers
Salmon testes DNA was dissolved in 20 mM NaCl to produce a
solution of approximately 1 mM DNA base pair concentration.
Subsequently, the complex solution was added dropwise to the
◦
before cooling overnight. Blue crystals suitable for X-ray structure
analysis were obtained on the following day. Yield, 0.47 (46%).
◦
DNA, and stirred for approximately 12 h at 4 C. The resulting
Anal. calc. for 1a, CuC18
H
18
N
4
O
4
·5H
2
O: C, 42.55; H, 5.56; N,
solutions were adjusted to a pH of 7.0 with sodium hydroxide
and hydrochloric acid throughout the experiments. The DNA-
pellet was obtained by ultracentrifugation of the mixed solution.
The DNA fibers were prepared from the pellet as previously
−
1
1
3
1.03%. Found: C, 42.12; H, 5.20; N, 11.22%. FTIR (m/cm ):
391 vs (broad, b), 2921 vs, 1590 vs, 1523 m, 1460 m, 1430 vs, 1397
vs, 1352 m, 1328 vs, 1264 s, 1226 s, 1147 w, 1110 m, 1090 m, 1056
w, 1033 m, 987 m, 912 w, 852 s, 816 m, 725 s, 626 s, 581 s.
23,24,34
reported.
Anal. calc. for 1b, CoC18
H
18
N
4
O
4
·3H
2
O: C, 46.26; H, 5.18; N,
−
1
11.99%. Found: C, 45.79; H, 4.67; N, 11.78%. FTIR (m/cm ):
3401 vs (b), 3275 vs, 2946 s, 1648 vs, 1595 vs, 1517 m, 1455 m,
1424 s, 1366 vs, 1314 s, 1289 m, 1281 m, 1225 m, 1136 m, 1114 w,
1061 m, 1014 s, 951 s, 899 m, 870 m, 852 s, 828 m, 784 s, 728 s,
641 s, 618 s, 571 s, 537 m, 506 m, 422 m.
4
.4 EPR spectra
X-Band (9.2 GHz) EPR spectra at ambient temperature and
◦
−150 C were measured on a JEOL RE-2X spectrometer with 100
kHz field modulation of 0.5 mT. The magnetic field was calibrated
with an NMR field meter EFM-2000 (ECHO Electronics LTD).
The microwave frequency was measured on an Anritsu MF2412A
frequency counter. The EPR spectra were measured at different
angles of U between the fiber axis and the static magnetic field.
Anal. calc. for 1c, ZnC18
H
18
N
4
O
4
·3.5H
2
O: C, 44.78; H, 5.22;
−
1
N, 11.60%. Found: C, 44.84; H, 4.61; N, 11.60%. FTIR (m/cm ):
3370 vs (b), 3277 vs, 2964 w, 1654 s, 1594 vs, 1520 m, 1425 s, 1373 s,
1313 s, 1256 w, 1227 w, 1139 m, 1092 m, 1063 w, 1014 m, 949 m,
902 w, 868 m, 825 w, 785 m, 728 s, 621 m, 582 m, 421 m.
Anal. calc. for 1d, NiC18
H
18
N
4
O
4
·3H
2
O: C, 46.28; H, 5.18; N,
4.5 Antiproliferative screening
−
1
1
3
1
1
5
1.99%. Found: C, 46.04; H, 4.80; N, 11.85%. FTIR (m/cm ):
401 vs (b), 3275 vs, 2946 m, 1648 vs, 1595 vs, 1517 m, 1457 m,
424 s, 1366 vs, 1314 s, 1290 m, 1281 m, 1225 m, 1136 m, 1115 w,
061 m, 1014 s, 951 s, 899 m, 870 m, 828 m, 784 s, 640 m, 618 m,
71 m, 537 m, 506 m, 422 m.
All tissue culture reagents were obtained from Sigma and Life
Technologies Inc., Gaitherburg, MD. The oestrogen receptor
positive human breast adenocarcinoma cell line MCF-7 was
cultivated for a minimum of two passages after thawing prior to ex-
perimentation. The cells were grown in RPMI 1640 supplemented
with 10% FBS, 0.025 M Hepes, 0.024 M sodium bicarbonate,
¯
Crystal data for 1a. CuC18
H
28
N
4
O
9
, M = 507.98, triclinic, P1,
−
1
◦
˚
50 units ml penicillin G/streptomycin sulfate at 37 C in 5%
CO
a = 9.897(1), b = 11.233(2), c = 11.851(2) A, a = 65.609(2),
◦
3
˚
2
.
b = 70.409(2), c = 73.771(2) , V = 1115.1(3) A , Z = 2, D =
c
−
3
−1
MTT assay was used to test the cytotoxicity of metal complexes.
The growing MCF-7 tumour cells were seeded into a 96-well plate
1
.513 g cm , l (Mo-Ka) = 1.036 mm , T = 150(2) K, blue prism,
Bruker CCD-1000 area detector diffractometer, 11229 reflections
5
at a density of 2.3–3.0 × 10 cells per well and incubated in a
collected, 5197 independent measured reflections (Rint = 0.0365),
2
medium containing the complexes at concentration ranging from
F refinement, R
638 independent observed absorption-corrected reflections [I >
1
= 0.0267 (I > 2r(I)), wR = 0.0759 (all data),
2
0.5–50 lM for 72 h. A 0.25% trypsin EDTA solution was used to
4
2
◦
detach the cells from the surface. Every experiment was conducted
in triplicate wells. Into each well 20 ll of MTT (5 mg ml ) was
added. The plates were then incubated at 37 C for 4 h to allow
r(I), hmax = 28.2892 ], 325 parameters.
−
1
CCDC reference number 636037.
◦
MTT to form formazan crystals. The crystals were then solubilized
using DMSO for a few minutes. The absorbance of each well was
measured in a microplate reader Dynatech MRX at 570 nm with
a reference wavelength of 630 nm. The percentage of cell viability
was calculated with the formula: Average A570 value for live cell
4
.2 Characterization of complexes
The infrared spectra of the complexes were recorded on a Perkin-
Elmer FT-IR spectrometer in the frequency range 4000–400 cm
as KBr discs. The C, H and N microanalysis was carried out
−
1
This journal is © The Royal Society of Chemistry 2008
Dalton Trans., 2008, 447–454 | 453

View Article Online
(
test)/average A570 value for live cell (control) × 100. The IC50
4 P. Heffeter, M. A. Jakupec, W. K o¨ rner, S. Wild, N. G. Von Keyserlingk,
L. Elbling, H. Zorbas, A. Koryneuska, S. Knasm u¨ ller, H. Sutterl u¨ ty, M.
Micksche, B. K. Keppler and W. Berger, Biochem. Pharmacol., 2006,
values were obtained from the graphs drawn.
7
1, 426.
4
.6 Apoptosis and cell cycle studies with flow cytometry
5
N. Margiotta, A. Bergamo, G. Sava, G. Padovano, E. de Clercq and G.
Natile, J. Inorg. Biochem., 2004, 98, 1385.
V. X. Jin and J. D. Ranford, Inorg. Chim. Acta, 2000, 304, 38.
Apoptosis and cell cycle analysis were determined using a Becton
Dickinson flow cytometer FACS Calibur . For both analyses,
6
TM
7 K. Balamurugan, R. Rajaram, T. Ramasami and S. Narayanan, Free
Radical Biol. Med., 2002, 33, 1622.
5
−1
MCF-7 cells were plated at a density of 3 × 10 cells ml in
0 mm petri dishes and allowed to recover for 24 h. To analyze cell
cycle phase distribution, the cells were treated with 1a (2.8 lM),
b (5.3 lM) and 1c (9.2 lM) for 72 h and washed with PBS. Cells
were harvested through trypsinization and resuspended in RPMI
640 medium. Floating cells were included in the analysis. After
8
H. Jhou, C. Zheng, G. Zou, D. Tao and J. Gong, Int. J. Biochem. Cell
Biol., 2002, 34, 678.
6
9 S. Zhang, Y. Zhu, C. Tu, H. Wei, Z. Yang, L. Lin, J. Ding, J. Zhang and
Z. Guo, J. Inorg. Biochem., 2004, 98, 2099.
0 B. Coyle, P. Kinsella, M. McCann, M. Devereux, R. O’Connor, M.
Clynes and K. Kavanagh, Toxicol. in Vitro, 2004, 18, 63.
1 M. McCann, M. Geraghty, M. Devereux, D. O’Shea, J. Mason and
L. O’Sullivan, Met.-Based Drugs, 2000, 7, 185 and reference quoted
therein.
1
1
1
1
centrifugation for 10 min at 1500 rpm, the pellets were fixed in
ice-cold 70% ethanol, stored at 4 C. The fixative was removed
by centrifugation. Fixed cells were resuspended in 1 ml of PBS,
◦
1
2 M. Bernal, J. A. Garc ´ı a-V a´ zquez, J. Romero, C. G o´ mez, M. L. Dur a´ n,
A. Sousa, A. Sousa-Pedrares, D. J. Rose, K. P. Maresca and J. Zubieta,
Inorg. Chim. Acta, 1999, 295, 39.
−
1
treated with 15 lL of RNase A (15 lg ml ) and stained with 10 lL
of propidium iodide (Sigma Chemical Co.) for 15 min at room
temperature. The cell cycle profiles were determined on FACS by
using Cell Quest software. A total of 10 000 events were collected
and the cells were properly gated for analysis. The histogram of
DNA content (x-axis, PI-fluorescence) versus counts (y-axis) was
displayed.
After treatment of cells with 1a, 1b and 1c at the respective
concentrations (same as above), the cells were washed with PBS,
trysinized and resuspended in RPMI 1640 medium. Flow cells
were included in the apoptosis analysis. After centrifugation for
13 C. M. Sharaby, Spectrochim. Acta, Part A, 2005, 62, 326.
1
4 M. S. Masoud, S. A. Abou El-enein, M. E. Ayad and A. S. Goher,
Spectrochim. Acta, Part A, 2004, 60, 77.
1
5 S. Chandra and U. Kumar, Spectrochim. Acta, Part A, 2005, 61, 219.
16 M. El-Behery and H. El-Twigry, Spectrochim. Acta, Part A, 2007, 66,
2
8.
1
1
7 S. Chandra and L. K. Gupta, Spectrochim. Acta, Part A, 2004, 60, 1563.
8 H. Hadadzadeh, M. M. Olmstead, A. R. Rezvani, N. Safari and H.
Saravani, Inorg. Chim. Acta, 2006, 359, 2154.
19 C. H. Ng, H. K. Alan Ong, C. W. Kong, K. W. Tan, R. N. Z. R. A.
Rahman, B. M. Yamin and S. W. Ng, Polyhedron, 2006, 25, 3118.
2
0 C. H. Ng, S. B. Teo, S. G. Teoh, H. K. Fun and J.-P. Declercq, Inorg.
Chim. Acta, 2002, 340, 81.
1
0 min at 1500 rpm, the pellets were gently resuspended in 100 lL
of binding buffer (Hepes, NaCl, CaCl in 6.5 : 22.3 : 1 ratio),
lL of Annexin V-FITC (Fharmigen, USA), 5 lL propidium
21 S. Zhang, Y. Zhu, C. Tu, H. Wei, Z. Yang, L. Lin, J. Ding, J. Zhang and
Z. Guo, J. Inorg. Biochem., 2004, 98, 2099.
2 Z.-M. Wang, H.-K. Lin, S.-R. Zhu, T.-F. Liu and Y.-T. Chen, J. Inorg.
Biochem., 2002, 89, 97.
3 W. Harada, T. Nojima, A. Shibayama, H. Ueda, H. Shindo and M.
Chikira, J. Inorg. Biochem., 1996, 64, 273.
4 M. Chikira, M. Inoue, R. Nagane, W. Harada and H. Shindo, J. Inorg.
Biochem., 1997, 66, 131.
5 M. Chikira, Y. Tomizawa, D. Fukita, T. Sugisaki, N. Sugawara,
T. Yamazaki, A. Sasano, H. Shindo, M. Palaniandavar and W. E.
Antholine, J. Inorg. Biochem., 2002, 89, 163.
2
2
2
2
2
5
−
1
iodide (200 lg ml in PBS; Sigma Chemical Co.) in a round
bottom polystyrene tube (Becton Dickinson) for 15 min at room
temperature. 200 lL of binding buffer was added before analysis
by flow cytometry. The cells were similarly analyzed on FACS by
using Cell Quest software. A total of 10 000 events were acquired
and the cells were properly gated for analysis. The cell populations
were displayed as a dot plot divided into four quadrants with
Annexin V-FITC fluorescence (x-axis) versus propidium iodide
fluorescence (y-axis).
2
2
2
6 J. D. Ranford, P. J. Sadler and D. A. Tocher, J. Chem. Soc., Dalton
Trans., 1993, 33931.
ˇ
7 Z. Tr a´ ni cˇ ek, M. Malo nˇ , Z. Sindel a´ rˇ , K. Dole zˇ al, J. Rol cˇ ´ı k, V. Kry sˇ tof,
M. Strnad and J. Marek, J. Inorg. Biochem., 2001, 84, 23.
8 K. G. Daniela, P. Gupta, R. H. Harbach, W. C. Guida and Q. P. Dou,
Biochem. Pharmacol., 2004, 67, 1139.
Acknowledgements
29 M. Devereux, D. O. Shea, A. Kellett, M. McCann, M. Walsh, D. Egan,
C. Deegan, K. Kedziora, G. Rosair and H. M u¨ ller-Bunz, J. Inorg.
Biochem., 2007, 101, 881.
This project is funded by a SAGA research grant from the
Malaysian Academy of Science and the Ministry of Science, Tech-
nology and Innovation of Malaysia (C/4/02). M.C. acknowledges
financial support from Chuo University Personal Research Grant.
3
0 C. C. Sprenger, M. E. Vail, J. Simurdak and S. R. Plymate, Oncogene,
2
000, 21, 140.
3
1 T.-C. Wang, L.-L. Chen, P.-J. Lu, C.-H. Wong, C.-H. Liao, K.-C. Tsiao,
K.-M. Chang, Y.-L. Chen and C.-C. Tzeng, Bioorg. Med. Chem., 2005,
1
3, 6045.
3
3
2 A.-R. Ghezzi, M. Aceto, C. Cassino, E. Gabano and D. Osella, J.,
J. Inorg. Biochem., 2004, 98, 73.
3 E. Pantoja, A. Gallipoli, S. van Zutphen, S. Komeda, D. Reddy, D.
Jaganyi, M. Lutz, D. M. Tooke, A. L. Spek, C. Navarro-Ranninger and
J. Reedijk, J. Inorg. Biochem., 2006, 100, 1955.
References
1
P. Christofis, M. Katsarou, A. Papakyriakou, Y. Sanakis, N. Katsaros
and G. Psomas, J. Inorg. Biochem., 2005, 99, 2197.
2
3
P. J. Dyson and G. Sava, Dalton Trans., 2006, 1929.
34 M. Chikira, S. Suda, T. Nakabayashi, Y. Fujiwara, T. Ejiri, M.
Yoshikawa, N. Kobayashi and H. Shindo, J. Chem. Soc., Dalton Trans.,
1995, 1325.
R. Huang, A. Wallqvist and D. G. Covell, Biochem. Pharmacol., 2005,
6
9, 1009.
4
54 | Dalton Trans., 2008, 447–454
This journal is © The Royal Society of Chemistry 2008