DNA binding and cytotoxicity of two Cu(II) complexes containing a Schiff base ligand along with 1,10-phenanthroline or imidazole as a coligand

Article abstract of DOI:10.1016/j.ica.2018.04.017

We report the synthesis, characterization and biological activities of the compounds [CuL(o-phen)] (1) and [CuL(Imz)] (2), (H₂L = o-HOC₆H₄C(H)=NC₆H₄OH-o, o-phen = 1,10-phenanthroline, and Imz = imidazole). Both the compounds 1 and 2 have been characterized by X-ray crystallography. A four-line EPR pattern originating from the interaction of the unpaired electron with the central ^63/65Cu nucleus (I = 3/2) with A_iso values of 7.3 and 8.5 mT, respectively for 1 and 2 in DMF at RT suggests their monomeric nature in solution. These compounds undergo irreversible oxidation–reduction. Biological studies show intercalative DNA binding and remarkable cell cytotoxicity as well as anticancer activity. IC₅₀ values of 1 and 2 for the human lung cancer A549 cell line (0.67 and 0.59 μM, respectively) and for the breast cancer MCF7 cell line (6.30 and 8.88 μM, respectively) are found to be very promising. Compounds 1 and 2 appear to be more potent for the human lung cancer A549 cell line than some anticancer drugs tested for this cell line. Most importantly, 1 and 2 are found to be remarkably less toxic for the human lung embryonic normal cell line L132 as evident from the cell viabilities in presence of these two compounds.

Full text of DOI:10.1016/j.ica.2018.04.017

Accepted Manuscript
Research paper
DNA binding and cytotoxicity of two Cu(II) complexes containing a Schiff base
ligand along with 1,10-phenanthroline or imidazole as a coligand
Manjuri K. Koley, Sidhali Uday Parsekar, Natarajan Duraipandy, Manikantan
Syamala Kiran, Babu Varghese, Periakaruppan T. Manoharan, Aditya P. Koley
PII:
S0020-1693(17)31717-6
https://doi.org/10.1016/j.ica.2018.04.017
ICA 18208
DOI:
Reference:
To appear in:
Inorganica Chimica Acta
Received Date:
Revised Date:
Accepted Date:
10 November 2017
26 February 2018
10 April 2018
Please cite this article as: M.K. Koley, S.U. Parsekar, N. Duraipandy, M.S. Kiran, B. Varghese, P.T. Manoharan,
A.P. Koley, DNA binding and cytotoxicity of two Cu(II) complexes containing a Schiff base ligand along with 1,10-
phenanthroline or imidazole as a coligand, Inorganica Chimica Acta (2018), doi: https://doi.org/10.1016/j.ica.
2
018.04.017
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

DNA binding and cytotoxicity of two Cu(II) complexes containing a Schiff base
ligand along with 1,10-phenanthroline or imidazole as a coligand
a,
a
b
Manjuri K. Koley *, Sidhali Uday Parsekar , Natarajan Duraipandy , Manikantan
b
c
d,
Syamala Kiran , Babu Varghese , Periakaruppan T. Manoharan * and Aditya P.
e,
Koley *
a
Department of Chemical Engineering, Birla Institute of Technology and Science-
Pilani, K.K. Birla Goa Campus, Goa 403726, India
b
Biological Materials Laboratory, CSIR-Central Leather Research Institute, Adyar,
Chennai 600 020, India
c
Sophisticated Analytical Instrument Facility, Indian Institute of Technology-Madras,
Chennai 600 036, India
d
Department of Chemistry, Indian Institute of Technology-Madras, Chennai 600 036,
India
e
Department of Chemistry, Birla Institute of Technology and Science-Pilani, K.K.
Birla Goa Campus, Goa 403 726, India
_
_
____________________________________________________________________
__________

2
Abstract
We report the synthesis, characterization and biological activities of the compounds
[
CuL(o-phen)] (1) and [CuL(Imz)] (2), (H L = o-HOC H C(H)=NC H OH-o, o-phen
2
6
4
6
4
=
1,10-phenanthroline, and Imz = imidazole). Both the compounds 1 and 2 have been
characterized by X-ray crystallography. A four-line EPR pattern originating from the
6
3/65
interaction of the unpaired electron with the central
Cu nucleus (I = 3/2) with Aiso
values of 7.3 and 8.5 mT, respectively for 1 and 2 in DMF at RT suggests their
monomeric nature in solution. These compounds undergo irreversible oxidation-
reduction. Biological studies show intercalative DNA binding and remarkable cell
cytotoxicity as well as anticancer activity. IC values of 1 and 2 for the human lung
5
0
cancer A549 cell line (0.67 and 0.59 M, respectively) and for the breast cancer MCF7
cell line (6.30 and 8.88 M, respectively) are found to be very promising. Compounds
1
and 2 appear to be more potent for the human lung cancer A549 cell line than some
anticancer drugs tested for this cell line. Most importantly, 1 and 2 are found to be
remarkably less toxic for the human lung embryonic normal cell line L132 as evident
from the cell viabilities in presence of these two compounds.
_
______________________________
Keywords: Cu(II) complex; EPR and electronic spectra; DNA binding; MTT assay;
Cytotoxicity and anticancer activity
_
*
______________________________
Corresponding authors. Email: manjuri@goa.bits-pilani.ac.in (M. K. Koley);
ptm@iitm.ac.in (P.T.

3
Manoharan); koleyap@yahoo.co.in (A.P. Koley)
1
. Introduction
To explore the possibility of finding new metallodrugs that may serve as anticancer
agents [1 13], there have been increasing efforts in recent years to search for or design
and develop metal complexes that show DNA binding property, nucleic acid cleavage
activity as well as cytotoxicity. It is known that the nature of the ligand present in the
metallodrugs plays a critical role in binding of the metal complex to DNA [14] which
is an important target for anticancer drugs and also the metal ion may play an
important role in the efficacy of the metal complex as a drug. For example, Cu(II)
being a redox active metal ion may promote transformation of DNA or protein through
an oxidative pathway. Very recently we have reported [15] a few Cu(II)/Zn(II)
complexes with a carbohydrazone ligand along with 1,10-phenanthroline as a coligand
which were found to be very promising and appeared to be more potent than some
anticancer drugs tested for the human lung cancer A549 cell line and for the breast
cancer MCF7 cell line as revealed from their IC values. In our continuing effort in
5
0
finding metal complexes with suitable donor ligands with additional heterocyclic N-
donors [15, 16], that are capable of effectively binding to DNA, we have synthesized
two Cu(II) complexes, one five-coordinate and another four-coordinate, containing the
Schiff base ligand N-salicylidene-o-aminophenol (H L = o-HOC H C(H)=NC H OH-
2
6
4
6
4
o) and 1,10-phenanthroline or imidazole as the coligand. Here, we report their spectral

4
(
UV-visible, IR, and EPR) and electrochemical behavior as well as their biological
activities such as DNA binding, cytotoxicity and anticancer activity.
. Results and discussion
We have studied the reaction of Cu(OAc) H O with the Schiff base N-salicylidene-o-
2
2
2
aminophenol (o-HOC H C(H)=NC H OH-o) in methanol in presence of either 1,10-
6
4
6
4
phenanthroline or imidazole as the coligand to prepare the complexes 1 and 2,
respectively as shown in Scheme 1.
Scheme 1. Synthesis of the ligand and the complexes.
The IR spectrum of the free Schiff base ligand clearly displays the ν(C=N) band at
1
1
1
630 cm [17]. This is found to be shifted to 1608 cm in the IR spectra of both the
complexes (figures not shown) indicating the coordination of the imine N to the metal

5
ion, apart from the deprotonation and coordination of both the phenolate O . This is
confirmed from the crystal structures of both 1 and 2.
These are moderately soluble in methanol and acetonitrile but readily soluble in DMF
and DMSO. The ESI-MS spectra (in the positive-ion mode) for these complexes have
been recorded in CH CN and the results are presented in Fig. S1 and Fig. S2
3
+
(
Supplementary material). Complex 1 showed peaks at m/z 454.30 [M] and m/z
+
+
4
3
91.30 [(M+K)-2H] while compound 2 showed peaks at m/z 343.01 [M] and m/z
+
44.81 [M+2H] , respectively.
The electronic spectra of these two compounds in the visible region look very similar
Figs. S3 and S4 of the Supplementary material), but the intensities of the bands are
found to be widely different. Compound 1 displays a weak broad d-d band centered
(
-
1
-1
around 650 nm (ε = 144.9 M cm ) while compound 2 displays the weak ligand filed
-1
-1
band centered 565 nm (ε = 45.4 M cm ). Apart from the weak ligand field transition
[15, 16], compound 1 exhibits strong LMCT in the visible region. The charge transfer
-
1
-1
band at 434 nm (ε = 10284 M cm ) for 1 containing 1,10-phenanthroline as the
-
1
-1
coligand is found to be highly intense compared to that at 440 nm (ε = 221.9 M cm )
for 2 with imidazole as the coligand. Also the overall intensity for all the bands for 1 is
much higher than that for 2, showing the differences in their electronic structures.
2
.1. Crystal structures of [CuL(o-phen)] 3.5H O (1)and [CuL(Imz)] (2)
2
The structure of 1 and 2 have been determined by X-ray crystallography. Fig. 1 shows
the ORTEP representation of the molecule of 1 with 40% probability ellipsoids [18].
The crystal for 1 is indexed in monoclinic system with space group C2/c and lattice

6
parameters a = 21.070(3) Å, b = 20.629(3) Å, c = 14.1300(16) Å, α = 90 ,
28.704(3) , γ = 90 The bond length and bond angles are given in Table S1 of the
Supplementary material.
=
1
Fig. 1. (A) ORTEP diagram of the compound 1 and (B) packing of the molecules in
the unit cell. Bond lengths [Å] and angles [°] for 1: N(1)-Cu(1) = 2.287(5); N(2)-
Cu(1) = 2.016(5); N(3)-Cu(1) = 1.904(12); N(3')-Cu(1) = 1.981(17); O(1)-Cu(1) =
1
7
.915(4); O(2)-Cu(1) = 1.933(4); N(3)-Cu(1)-O(1) = 100.8(5); N(3)-Cu(1)-O(2) =
8.2(5); O(1)-Cu(1)-O(2) = 162.51(18); O(1)-Cu(1)-N(3') = 76.5(6); O(2)-Cu(1)-N(3')

7
=
101.9(6); N(3)-Cu(1)-N(2) = 167.6(4); O(1)-Cu(1)-N(2) = 91.41(17); O(2)-Cu(1)-
N(2) = 90.85(19); N(3')-Cu(1)-N(2) = 167.2(6); N(3)-Cu(1)-N(1) = 97.7(3); O(1)-
Cu(1)-N(1) = 100.14(16); O(2)-Cu(1)-N(1) = 97.29(17); N(3')-Cu(1)-N(1) = 100.0(4);
N(2)-Cu(1)-N(1) = 77.79(18).
The Cu atom is five-coordinated. The geometry of coordination is approximately
square pyramidal with O1N3O2N2 atoms forming the basal plane for Cu1. The apical
N(1)-Cu(1) bond length is found to be much longer (2.287(5) Å) than the other two
Cu-N distances (N(2)-Cu(1) = 2.016(5) and N(3)-Cu(1) = 1.904(12)). All the
hydrogens could be located in the difference Fourier map. However, it was decided to
fix the hydrogens at geometrically meaningful positions. These atoms were refined by
riding model. The packing of molecules in the unit cell shows water molecules
mediated H-bonding interactions stabilizing the lattice. Compound 2, on the other
hand, crystallizes in orthorhombic system with space group Pca2 and lattice
1
parameters a = 9.3659(6) Å, b = 17.4585(11) Å, c = 18.1049(13) Å, α = 90 , = 90 ,
γ = 90 . The bond length and bond angles are given in Table S2 of the Supplementary
material.
Fig. 2 shows the ORTEP representation of the molecule with 40% probability
ellipsoids [18] and the packing of the molecules. There are two molecules in the
asymmetric unit. Cu atoms of both molecules (Cu1 and Cu2) have near square-planar
coordination with two nitrogen atoms trans- to each other, one imine N of the Schiff
base ligand and the tertiary N of the imidazole, and two phenolate O of the Schiff base
ligand. The Cu(1)-N(2) bond length is slightly longer than that with imine N(1). As

8
expected, Cu(1)-O(1) is found to be longer than Cu(1)-O(2) bond length showing the
difference in formation of five-membered ring for the former versus six-membered
ring in the latter case. The bond angles clearly show the molecule is not completely
planar. The mean plane calculation of the basal square plane coordination shows the
coordinating N and O atoms to have small tetrahedral distortion (from the ideal square
plane). The packing of molecules in the unit cell shows imidazole mediated H-bonding
interactions as well as the intermolecular -stacking interactions of aromatic rings of
the Schiff base are operating in stabilizing the lattice.

9
Fig. 2. (A) ORTEP diagram of the compound 2 and (B) packing of the molecules in
the unit cell. Bond lengths [Å] and angles [°] for 2: N(1)-Cu(1) = 1.943(11); N(2)-
Cu(1) = 1.953(7); O(1)-Cu(1) = 1.927(6); O(2)-Cu(1) = 1.890(7); N(4)-Cu(2) =
1
.926(9); N(5)-Cu(2) = 1.965(7), O(3)-Cu(2) = 1.932(6); O(4)-Cu(2) = 1.888(6);
O(2)-Cu(1)-O(1) = 169.6(3); O(2)-Cu(1)-N(1) = 94.7(4); O(1)-Cu(1)-N(1) = 84.4(3);
O(2)-Cu(1)-N(2) = 89.9(3); O(1)-Cu(1)-N(2) = 92.3(3); N(1)-Cu(1)-N(2) = 171.3(4).
2
.2. EPR results
The EPR spectra at X-band frequency for the compounds 1 and 2 in powder state as
well as in DMF solution were recorded at room temperature (RT), and the DMF
frozen glass spectra were recorded at liquid nitrogen temperature (LNT) and are
shown in Fig. 3. In general, both the compounds containing Cu(II) as the central metal
ion display strong spectra since they are EPR sensitive with its characteristic g-values
and hyperfine splitting in the frozen state and at RT when the solid compounds are
dissolved in DMF medium. It is noteworthy that RT spectra of both solids (Figs. 3a
and b) exhibit excessively broad lines for reasons of strong dipolar interactions
between the adjacent Cu(II) centres as indicated by the nature of packing in Figs. 1
and 2 above. The EPR of their DMF solutions as shown in Figs. 3c and d are nothing
6
3,65
but a simple 4-line spectra due to I=3/2 from
Cu nuclei with no exhibition of
resolution from the two different isotopes but with a strong m dependent line width.
I
The isotropic g –value for compounds 1 and 2 are 2.1 and 2.085 with the respective
isotropic hyperfine splitting 7.3 and 8.5 mT as shown in Table 1. These values are
characteristic of approximately planar or tetragonal crystal field environment.

1
0
The most notable results, however, come from their frozen solution spectra indicating
their differences in structure where the compound 1 is axial in character and 2 has a
rhombic symmetry. The axial character of compound 1 is confirmed by the
observation of only two anisotropic g-values viz. g = 2.21 and g =2.09 (Fig. 3e)
║
┴
63,65
along with the resolved hyperfine parallel component (A ) due to
Cu and with no
║
resolution of the same in the perpendicular section due to A being less than its line
┴
width. This justifies the crystallographic prediction of its axial character with the
planar unit of Cu(O1N3O2N2) along with the long apical bond of N(1)-Cu(1). On the
other hand, the EPR of compound 2 ( Fig.3f) is clearly rhombic in nature with g =
1
1
2
.21 , g = 2.09, and g = 1.97 with g working out to 2.09, very close to that of g
22
33
av
iso
being 2.085 which again justifies the rhombic character of the molecule as recognized
by crystallography. Similar anisotropy is noted in the corresponding hyperfine
values, A = 19.9 mT, A ≈ 7.0 mT , and A < line width, an average of which
1
1
22
33
makes A ≈ 9.62 mT with the assumption of A ≈ 2.0 mT (less than the line width)
av
33
which is fairly close to the observed A of 8.5 mT. The g values calculated from
iso
av
the corresponding polycrystalline values at RT slightly differ from those of gav
observed for the DMF frozen solution suggesting that DMF may have played a minor
role in disturbing the ligand environment of the molecule when brought into solution.
However, it is gratifying to see that EPR conforms to the crystallographic structures of
both compounds, one being axial and the other being rhombic in their character. All
observed EPR parameters are presented in Table 1.

1
1
Fig. 3. X-band EPR spectra of the compounds. (a)-(b) are RT spectra for 1-2 powder
samples, (c)-(d)
are RT solution spectra for 1-2 in DMF, and (e)-(f) are LNT frozen glass spectra for 1-
in DMF, respectively.
2
Table 1. EPR parameters for compounds 1 and 2 derived from the spectra at RT and
LNT.
Compound/state/temp.
ν (GHz) g value
HFS values
mT)
(
1
- DMF frozen glass at 9.138
g║ = 2.21
g┴ = 2.09
g = 2.11
Δpp ≈ 4.4
to 3.9mT
g║ = 2.14
A = 20.67
║
LNT
A < linewidth
┴
1
- DMF solution at RT
9.434
9.447
A = 7.30
iso
1
- Powder at RT
Not resolved

1
2
g┴ = 2.06
gzz = 2.21
gyy ≈ 2.09
gxx = 1.97
g = 2.085
2
- DMF frozen glass at
9.137
9.433
9.448
A = 19.99
zz
LNT
A ≈ 7.0
yy
A < linewidth
xx
2
2
- DMF solution at RT
A = 8.50
iso
Δ ≈ 4.6 to
pp
4
.13 mT
- Powder at RT
g = 2.11
Not resolved
11
g = 2.07
22
g = 2.05
33
2
.3. DNA binding studies
DNA binding studies were performed using UV-Vis spectroscopy. Electronic
absorption titration experiments were performed with incremental amounts of calf
thymus DNA (CT-DNA) in 10 mM Tris–HCl buffer (pH 7.54) by keeping the
-4
individual complex concentration constant at 2.5 10 M. Absorption spectrum of the
complex 1 and 2 in the absence and presence of CT-DNA has been shown in Fig. 4.
On addition of CT-DNA both the complexes show a decrease in molar absorptivity
(hypochromism of 47.5% for 1 and 82.5 % for 2) of the π → π* absorption band
indicating strong binding of the complexes to DNA. These changes are typical of
complexes that bind to DNA by intercalation mode, which involves strong stacking

1
3
interaction between aromatic chromophore and the DNA base pairs [19]. To determine
the quantitative comparison of the DNA binding ability, we determined the intrinsic
binding constant, K , of the complexes for binding with calf thymus (CT-DNA) by
b
using the following Eq. (1).
(1)
where, [DNA] is the concentration of CT-DNA used, and ε , ε and ε correspond to
a
f
b
apparent extinction coefficients for the complex i.e., Abs/[complex] in the presence of
DNA, in the
absence of DNA and to fully bound DNA, respectively. A plot of [DNA]/(ε − ε )
a
f
versus
DNA] yields a slope = 1/(ε − ε ), and the intercept = 1/K (ε − ε ), respectively. The
[
b
f
b
b
f
binding constant K is calculated from the ratio of the slope to the intercept. The K
b
b
5
1
2
5
1
2
values are 1.53 10 M for 1 (R = 0.99770 for 12 points) and 3.13 10 M for 2 (R
=
0.99867 for 9 points), respectively. These values are comparable to that reported for
5
1
a classical intercalator such as ethidium bromide (K = 1.40 × 10 M ) in 25 mM Tris-
b
HCl buffer (pH 7.33) [20]. The high percentage (82.5 %) of hypochromism and the
higher K value for 2 compared to 1 indicates its higher binding propensity which
b
arises due to the four coordinate square planar structure of 2, which allows the
intercalating ligand to insert into the DNA base pair effectively [21-22], and this
enhanced stacking interaction results in increased binding ability. It is to be noted that

1
4
K of this neutral complex 2 is higher than that for the reported four coordinate
b
cationic compound [Cu(pabt)(Imz)]ClO containing imidazole along with the chelating
4
NNS donor ligand [16]. Table S3 (Supplementary material) shows a comparison of Kb
values with some more analogous compounds reported in literature.
4
Fig. 4. (A) Change in electronic absorption spectra of 1 [2.5 10 M] upon titration
with CT-DNA (0–40 μM) dissolved in 10 mM Tris–HCl buffer (pH 7.54). (B) Change
4
in electronic absorption spectra of 2 [2.5 10 M] upon titration with CT-DNA (0–25
μM) dissolved in 10 mM Tris–HCl buffer (pH 7.54). Arrow shows the decrease in
absorbance with respect to an increase in the concentration of CT-DNA. The figure in
inset shows the linear fit of [DNA]/(ε − ε ) versus [DNA]. The binding constant (K )
a
f
b
was calculated using the Eq. (1).
.4. Viscosity measurements
2
As supporting evidence to the spectroscopic analyses of DNA intercalation exhibited
by both the compounds, viscosity measurements on CT-DNA by varying the
concentration of added complex were also conducted. Normally it is observed that the
relative viscosity of CT-DNA solution increases on interaction with substrates that

1
5
bind by intercalative mode. This is because insertion of metal complex between the
DNA base pairs causes the base pairs to separate which increases the overall length of
DNA, thus resulting in an increase in viscosity of DNA. The effect of addition of
increasing amount of complex 1 and 2 on the relative viscosity depicted in Fig. 5
shows steady increase in the viscosity of the DNA which clearly suggests intercalative
mode of binding of the complexes to DNA. It is also to be noted that the increase in
viscosity is more pronounced in case of 2. The results support the spectral studies
which suggest that complex 2 binds more strongly to CT-DNA.
2.0
1.8
1.6
1.4
1.2
1.0
1
2
+ CT DNA
+ CT DNA
0.0
0.2
0.4
0.6
0.8
1.0
1.2
[Complex]/[DNA]
Fig. 5. Effect of increasing amount of complex 1 (red circles) and 2 (green squares) on
the relative viscosity of CT-DNA (100 μM) at room temperature in 10 mM Tris.HCl
buffer (pH7.46).
2
2
.5. Results on biological activity
.5.1. Cytotoxicity
The cytotoxicity profiling of the compounds 1 and 2 were evaluated by MTT assay
using A549 lung cancer cell line, MCF7 breast cancer cell line, L132 human lung

1
6
embryonic normal cell line and human keratinocyte HaCaT normal cell line. The cell
viability was determined by the reduction of yellow MTT into purple formazan
product by the enzymatic activity of mitochondrial dehydrogenase in the live cells. A
concentration dependent effect on the cell viability was observed (Fig. 6) for both
cancerous cell lines (MCF7 and A549) as well as normal cell lines (HaCaT, L132).
Results indicate that even at as low concentration as 0.25 μM, both the compounds 1
and 2 are found to be toxic to A549 cells as indicated from their respective cell
viability of 88% and 67%, while toxicity is comparatively much less in normal cells
HaCaT at the same dose (cell viability is 97% for 1 and 91% for 2), and almost no
toxicity was observed for L132 lung normal cells that were treated with 0.25 μM of 1
and 2 (Fig. 6). The viability of A549 cells treated with 0.5 μM of 1 and 2 falls to 69.5
%
and 54%, respectively, as compared to cell viability of 93% in case of L132 cells
treated with same dose of 1 or 2. Also 1 shows similar results in normal cell HaCaT as
indicated by 91% cell viability at 0.5 μM dose. Similar observation were found when
A549 cells were treated with 1 μM dose of 1 and 2, cell viability was reduced to 38%
for both compounds, whereas the compounds appear to be much less toxic to L132
normal cells showing cell viability of 78% and 86%, respectively, on treatment with 1
μM dose of 1 and 2. Though cell viability seems fairly moderate (50%) when normal
HaCaT cells are treated with 1 μM dose of 1, compound 2 proved to be highly toxic to
HaCaT cells at the same 1 μM dose. Again remarkable difference in cell viability is
observed when compared the effect of the 2.5, 5 and 10 μM dose of compounds 1 and
2
on lung cancer A549 cell lines versus L132 lung normal cells (Fig. 6). In MCF7 cell
line, compound 1 is found to be more toxic than 2 for the entire concentration range

1
7
tested, but found to be much less toxic as compared to that for A549 cells. However,
both 1 and 2 are found to be very toxic toward HaCaT cells for their concentrations of
.5 μM and higher, but at 1.0 μM concentration, 1 is found to be remarkably less toxic
2
than 2 as revealed from their cell viabilities of 49.5% and 14.1% for 1 and 2,
respectively (Fig. 6).
Fig. 6. MTT assay for the compounds 1 and 2. Panel A. A549 cell line, panel B.
MCF7 cell line, panel C. L132 cell line, and panel D. HaCaT cell line.
The plot of cell viability versus concentrations of the compounds 1 and 2 for all these
cell lines is shown in Fig. S5 (Supplementary material). The IC values for 1 and 2
50

1
8
calculated from this plot for both the cancer cell lines A549 and MCF7 and for the two
normal cell lines are presented in Table 2.
Table 2. IC values for the compounds 1and 2 calculated from the plot of cell
5
0
viability versus concentration in the MTT assay.
Cell line
IC (µM) for
5
0
Compd. 1 Compd.
2
A549
0.67
6.30
7.60
0.99
0.59
8.88
8.95
0.60
MCF 7
L132
HaCaT
It is to be noted here that the IC values for A549 cells are better than the IC value
50
50
reported for cisplatin (4.13 µM for A549). On the other hand, the IC values for 1 and
5
0
2
for MCF7 cell line are found to be slightly higher than that reported for cisplatin
(3.92 µM for MCF7) and much lower than that reported for carboplatin (36.65 µM for
MCF7) [3c]. It is to be noted here that recently we have reported [15] a few
Cu(II)/Zn(II) complexes with a carbohydrazone ligand along with 1,10-phenanthroline
as a coligand which were found to be very promising anticancer agents tested for the
human lung cancer A549 cell line and for the breast cancer MCF7 cell line as revealed
from their IC values. However, those Cu(II) compounds were found to be very toxic
5
0
for the normal L132 and HaCaT cell lines also [15]. In contrast, the two compounds
under present study are found to be much less toxic toward normal L132 cell lines,

1
9
suggesting their suitability as anticancer agents specifically for human lung cancer
A549 cell line.
In order to assess the significance of our results, the IC values and K values of both
5
0
b
1
and 2 have been compared with those of some recently reported copper (II)
complexes for A549 and MCF7 cell lines and presented in Table S4 of the
Spplementary material. The comparison reveals that 1 and 2 exhibit higher value of
intrinsic binding constant K as well as exhibits better cytotoxic response in A549 cell
b
line (as indicated by their lower IC values) than most of the reported copper
50
complexes. Two of the Cu complexes reported by Nagababu and coworkers [1a] show
similar low IC values in A549 cell line, however, these complexes are reported to be
50
cytotoxic toward both cancerous and normal cell lines. Cu(II) complexes reported by
Lu et. al [3c] do exhibit lower IC values in A549 as well as MCF7 cell lines,
50
however, they have not reported cytotoxic effect on any normal cell line. The
complexes 1 and 2 in this study have been tested in both cancerous and noncancerous
cell lines and showed remarkable difference in cytotoxicity in lung cancer A549 cell
line as compared to human lung normal cell line L132, suggesting their potential to
serve as anticancer therapeutic agents. To compare the relative antiproliferative effect
of 1 and 2 on cancerous versus normal cells, their therapeutic index (TI) has been
calculated from the ratio of the IC values of the complexes in normal cells to that in
50
cancer cells and has been presented in supporting information Table S5.

2
0
2
.5.2. Dual-Fluorescence Viability: AO/PI Dual Staining assay for A549 and L132
cells
This dual staining assay using the fluorescence microscopy is used to analyze the
induction of cell apoptosis in presence of the compounds 1 and 2, with the help of
acridine orange (AO) and propidium iodide (PI) nucleic acid binding dyes. AO is
permeable to both live and dead cells, staining all nucleated cells to generate green
fluorescence. On the other hand, PI enters only in dead, dying, and necrotic nucleated
cells with compromised membranes and stains them all generating red fluorescence.
But when cells are stained with both AO and PI together, all dead nucleated cells
fluoresce red and all live nucleated cells fluoresce green due to Förster resonance
energy transfer (FRET). The following morphological changes were observed (Figs. 7
and 8) in the treated cells using this staining method: (a) the viable cells with highly
organized nuclei fluoresce green as seen from all the control in Figs. 7 and 8; (b) early
apoptotic cells with nuclear condensation, that emit orange-green fluorescence,
observed in the treatment with 2.5 and 5 M concentrations of the respective
compounds; (c) late apoptotic cells with highly condensed or fragmented chromatin
and fluoresce orange to red, observed in the treatment with 2.5 and 5 M
concentrations of the respective compounds; and (d) necrotic cells fluoresce red with
no indication of chromatin fragmentation as clearly evident (Figs. 7 and 8) in the
cases of A549 cells treated with 5 M of each of 1 and 2, while such red fluorescence
is not observed for L132 (Figs. 7 and 8) cells with 1-5 M concentrations for 1 and 2
suggesting no or very little DNA damage at these concentrations. These results are

2
1
consistent with the cell viability observed from their MTT assay (Fig. 6), confirming
the marked differences in the cell death in A549 and L132 cells treated with both 1
and 2, and these morphological changes indicate that both the complexes might have
induced cell death via both apoptosis and necrosis [3e, 23].
Fig. 7. AO/PI staining results for compound 1 with A549 and L132 cell lines. Scale
bar: 250 m.

2
2
Fig. 8. AO/PI staining results for compound 2 with A549 and L132 cell lines. Scale
bar: 250 m.
It is to be pointed out here that though the DNA binding experiments carried out in
vitro showed that the two compounds under study are capable of DNA intercalation, it
does not imply that the cell apoptosis observed are originating from this intercalation,
because it will largely depend on the structural integrity of these complexes after
entering the cell membrane. Though this possibility cannot be completely ruled out at
this stage, there could be other possibilities that they may undergo reduction to Cu(I)
in the reducing environment after entering the cell membrane [23e-h] followed by
ligand dissociation (as reflected from their electrochemical behavior) and may, in turn,
either interact with proteasome [2d] causing apoptosis or may combine with dioxygen
generating reactive oxygen species (ROS) which may cause oxidative cleavage of
DNA polynucleotide chains leading to apoptosis [1b]. This could be the reason for
these two compounds behave differently from other reported compounds [1b, c]. It
should be mentioned here in this context that cancer cells are more sensitive to
proteasome inhibition than normal cells [1b], at the same time it is also true that
cancer cells in general have increased levels of reactive oxygen species (ROS)
compared to normal cells. Thus, it is equally probable that the generated Cu(I) species
with a changed ligand environment may either cause proteasome inhibition leading to
cell death or escalate the level of ROS to a toxic level that results in oxidative stress-
induced cell death. This is the most plausible explanation we can give with the
currently available experimental evidences.

2
3
It is to be noted here that the cytotoxicity study of the free Schiff base ligand N-
salicylidene-o-aminophenol was also carried out using the human lung cancer A549
cell line with different concentrations (2.5 µM, 5 µM, 7.5 µM and 10 µM) of the free
ligand for 24 hours. No cell death was observed with 2.5 µM concentration of the free
ligand. The cell viabilities were found to be nearly 96.5 % with 5 µM, 95.5 % with 7.5
µM and about 91 % with 10 µM of the free ligand, respectively (Fig. S6,
Supplementary material), suggesting that the free ligand is capable of causing about 9-
1
0% cell death in the presence of 10 µM concentration of the free ligand. Thus, the
toxicity observed for the compounds 1 and 2 (e.g., with the concentrations 5 µM and
0 µM, respectively) may have some small contribution from the free Schiff base
1
ligand, provided the compounds undergo ligand dissociation after entering the cell
membrane, otherwise the observed toxicity should be accounted for the complex as a
whole.
2
.6. Electrochemical results
The electrochemical behaviour of compounds 1and 2 have been studied in DMF
containing 0.1 M [N(n-Bu) ]ClO at a platinum working electrode using cyclic
4
4
voltammetry (CV) in dinitrogen atmosphere. The cyclic voltammetric behaviour of
both the compounds are found to be independent of scan directions, and yield
oxidation and reduction peaks almost in similar positions irrespective of the scan
direction (Fig. 9, panels A and B for 1, and panels C and D for 2, respectively). For an
initial negative scan compound 1 shows (Fig. 9) a broad reduction wave near 1.25 V,
reversal of this scan yields three successive oxidations waves at 0.78, +0.839 and

2
4
+
+
1.17 V, respectively, further reversal of the scan yields reduction waves at +0.73 and
0.08 V, when scanned within 1.5 to +1.5 V. Compound 2 shows a broad reduction
wave near 1.53 V (versus Ag/AgCl), a weak oxidation peak at 1.20 V followed by
another oxidation peak near 0.0832 V is observed on scan reversal. Further positive
scan shows oxidation wave near +0.88 V, reversal of this scan yielded three successive
reduction waves at +1.26, +0.778 and +0.3168 V, respectively. The reduction peak at
0
.0832 V is coupled to the oxidation wave at +0.3168 V indicating a quasi-reversible
redox process. Thus the absence of any reversible or nearly reversible Cu(II)/Cu(I)
redox couple in the cyclic voltammograms of these complexes suggests that they
undergo irreversible oxidation and reduction involving both metal center and ligand
center, and any of them does not retain the same geometry in both these oxidation
states as it is known that only a reversible or nearly reversible Cu(II)/Cu(I) redox
couple retain almost same geometry in both these oxidation states because the redox
potentials of Cu(II)/Cu(I) systems depend on the relative thermodynamic stabilities of
the two oxidation states in a given ligand environment [16, 17].

2
5
−3
−1
Fig. 9. Cyclic voltammogram of the compounds (1.25 × 10 mol L ) in DMF
−1
containing 0.1 M [N(n-Bu) ]ClO at a scan rate of 100 mV s . Panels A and B for 1,
4
4
panels C and D for 2, for initial negative and positive scans, respectively.
3
. Conclusions
Both the compounds efficiently bind to DNA by intercalation mode with binding
5
1
5
1
constants 1.53 10 M and 3.13 10 M , respectively, as revealed from their DNA
binding study with calf thymus DNA. The compounds 1 and 2 exhibit in vitro
cytotoxicity against human breast cancer MCF7 cell line and human lung cancer A549
cell line and they are found to be highly potent for A549 cell line in particular as
revealed from their IC values. At the same time they are found to be much less toxic
5
0
for the L132 human lung embryonic normal cell line. AO/PI dual staining results

2
6
strongly suggest the induction of both apoptotic and necrotic cell death pathways for
the anticancer activity of these complexes.
4
. Experimental
4
.1. Chemicals
o-Aminophenol, imidazole and salicylaldehyde were obtained from Aldrich. Calf
thymus DNA was obtained from Sigma-Aldrich (USA), N,N-dimethylformamide
(DMF, GR), absolute ethanol, methanol (GR), DMSO, 1,10-phenanthroline
monohydrate and Cu(OAc) H O (GR) were obtained from Merck.
2
2
4
.2. Preparation of N-salicylidene-o-aminophenol (H L). This Schiff base of o-
2
aminophenol and salicylaldehyde was prepared following the reported method [16b].
Yield 75%. Anal. Calc. for C H NO : C, 73.22; H, 5.20; N, 6.56. Found: C, 73.12;
13
11
2
H, 5.21; N, 6.54%.
4
4
.3. Preparation of the complexes
.3.1. [CuL(o-phen)] (1). 1,10-phenanthroline (o-phen) monohydrate (0.398 g, 0.002
mol) in methanol (10 mL) was slowly added to a solution of Cu(OAc) H O (0.399 g,
2
2
o
0
.002 mol) in methanol (20 mL, 23 C) during 30 min with stirring at room
temperature. Then this reaction mixture was slowly added to a solution of N-
o
(
salicyldene)-o-aminophenol (0.426 g, 0.002 mol) in methanol (30 mL, ~ 25 C)
during 30 min with stirring at RT. A dark blue compound separated from the solution.
The solid was filtered, washed well with methanol, then with water and finally again
with methanol and dried in vacuo. This compound was recrystallized from
CH OH:CH CN (1:1) at RT while suitable single crystals were obtained. This
3
3
compound has been characterized by X-ray crystallography. This is to be noted here

2
7
that on long standing (or in vacuum) these crystals lose their water of crystallization
and become amorphous. Anal. Calc. for C H N O Cu: C, 65.99; H, 3.77; N, 9.24.
25
17
3
2
+
Found: C, 66.01; H, 3.80; N, 9.23%. ESI-MS in CH CN: m/z 454.30 [M] , m/z 491.30
3
+
[
(M+K)-2H] .
4
.3.2. [CuL(Imz)] (2). Imidazole (0.136 g, 0.002 mol) in methanol (10 mL) was slowly
added to a solution of Cu(OAc) H O (0.398 g, 0.002 mol) in methanol (20 mL, ~ 5
2
2
o
o
C) during 20 min with stirring at room temperature (~ 23 C) while a dark blue
solution was obtained. Then this solution was slowly added to a solution of N-
o
(
salicyldene)-o-aminophenol (0.426 g, 0.002 mol) in methanol (30 mL, ~ 25 C)
during 30 min with stirring at RT. A green compound separated from the solution. The
solid was filtered, washed well with methanol and dried in vacuo. This compound was
recrystallized from CH CN:CH OH (1 : 1) at RT while suitable single crystals were
3
3
obtained. This compound has also been characterized by X-ray crystallography.
Anal. Calc. for C H N O Cu: C, 56.05; H, 3.82; N, 12.26. Found: C, 56.01; H, 3.87;
1
6
13
3
2
+
+
N, 12.31%. ESI-MS in CH CN: m/z 343.01 [M] , m/z 344.81 [M+2H] .
3
4
.4. Physical measurements
Microanalyses (C, H, N) were performed in a Perkin-Elmer 240C elemental analyzer.
Mass spectra were recorded on a Perkin Elmer (USA) Flexer SQ 300 MS Mass
Spectrometer operating in ESI mode. Infrared spectra were measured with a Shimadzu
IR Affinity - 1 FT-IR spectrometer using KBr pellet. Electronic absorption spectra
were recorded on a Jasco V-570 UV/VIS/NIR spectrophotometer using a pair of

2
8
matched quartz cell of path length of 1 cm. Electron paramagnetic resonance spectra
were recorded on a JEOL, Japan Model: JES - FA200 ESR spectrometer with X and Q
band. Room temperature solution EPR spectra were recorded using an aqueous cell
and frozen glass spectra were recorded in liquid nitrogen using a quartz dewer.
Electrochemical measurements were done with the help of a Bioanalytical system CV-
2
7 electrochemical analyzer and a BAS model X-Y recorder at 298K under dinitrogen.
A standard three electrode cell consisting of a platinum working electrode, a platinum
auxiliary electrode and a Ag/AgCl reference electrode was used. Tetrabutylammonium
perchlorate ([N(n-Bu) ]ClO ) was used as supporting electrolyte.
4
4
4
4
.5. X-ray crystallography
.5.1. [CuL(o-phen)] 3.5H O (1)and [CuL(Imz)] (2)
2
Blue crystals of complex 1 were obtained from the slow evaporation of a methanol-
acetonitrile (1:1) solution of 1 at room temperature while green crystals for 2 were
obtained from the slow evaporation of a methanol-acetonitrile (2:1) solution of 2 at
room temperature. A blue crystal of 1 with approximate size of 0.35 0.25 0.20
mm or a green crystal of 2 with approximate size of 0.4 0.2 0.2 mm was mounted
on a Bruker Axs Kappa Apex2 diffractometer equipped with graphite-monochromated
[Mo K , = 0.71073 Å] radiation. The unit cell parameters (Table 3) were
determined by the method of difference vectors using reflections scanned from three
different zones of the reciprocal lattice. The intensity data were measured using ω and
°
φ scan with frame width of 0.5 . The frames integration and data reduction were
performed using Bruker SAINT-Plus (Version 7.06a) software [24]. The multi-scan

2
9
absorption corrections were applied to the data using SADABS software. The crystal
of 1 is indexed in monoclinic system with space group C 2/c and lattice parameters a =
2
9
1.070(3) Å, b = 20.629(3) Å, c = 14.1300(16) Å, α = 90 ,
= 128.704(3) , γ =
0 . On the other hand, the crystal of 2 is indexed in orthorhombic system with space
group 'P c a 2 ' and lattice parameters a = 9.3659(6) Å, b = 17.4585(11) Å, c =
1
1
8.1049(13) Å, α = 90 , = 90 , γ = 90 .
SIR-92 program [25] was used for solving the structure. Structure was refined using
SHELXL-2014 (Sheldrick, 2014) program [26]. Successive difference fourier map
showed the positions of all hydrogen atoms. However, the hydrogen positions were
geometrically fixed and refined through riding model. The full matrix structure
2
2
2
refinement was carried out through minimization of the function∑(w(F
F )) ,
o
c
2
2
1
2
2
where w = [
(F )+(0.1406P) ] for 1 and w = [
(F ) + (0.0449P) +
o
o
1
2
2
2
1
.2789P] for 2, and P = (F + 2F )/3, F is the measured intensity (i.e., intensity
o
c
o
2
observed) and F is the intensity calculated. The final residual factors were R = 0.06
c
and wR = 0.2326 for 1 and R = 0.05 and wR = 0.1452 for 2, respectively. The largest
3
difference map peak was 0.605 e/Å for 1 and the largest difference map peak was
3
1
.193 e/Å for 2.
Table 3. Crystal data and structure refinement for complexes 1 and 2.
_
_
____________________________________________________________________
__________________
Empirical formula
Formula weight
C H CuN O
C H CuN O
2
5
24
3
5.50
16 13
3
2
518.01
342.83