European Journal of Medicinal Chemistry 42 (2007) 431e439

http://www.elsevier.com/locate/ejmech

Original article

Comparative study of copper(II)ecurcumin complexes as superoxide

dismutase mimics and free radical scavengers
Atanu Barik a, Beena Mishra a, Amit Kunwar a, Ramakant M. Kadam b, Liang Shen c,
Sabari Dutta d,*, Subhash Padhye d, Ashis K. Satpati e,
Hong-Yu Zhang c, K. Indira Priyadarsini a,**
a
Radiation and Photochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai-400085, India
b
Radiochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai-400085, India
c
Shandong Provincial Research Center for Bioinformatic Engineering and Technique, Shandong University of Technology, Zibo 255049, PR China
d
Department of Chemistry, Pune University, Pune-411007, India
e
Analytical Chemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai-400085, India
Received 10 September 2006; received in revised form 18 November 2006; accepted 20 November 2006
Available online 30 November 2006

Abstract

Two stoichiometrically different copper(II) complexes of curcumin (stoichiometry, 1:1 and 1:2 for copper:curcumin), were examined for their
superoxide dismutase (SOD) activity, free radical-scavenging ability and antioxidant potential. Both the complexes are soluble in lipids and
DMSO. The formation constants of the complexes were determined by voltammetry. EPR spectra of the complexes in DMSO at 77 K showed
that the 1:2 Cu(II)ecurcumin complex is square planar and the 1:1 Cu(II)ecurcumin complex is distorted orthorhombic. Cu(II)ecurcumin
complex (1:1) with larger distortion from square planar structure shows higher SOD activity. These complexes inhibit g-radiation induced lipid
peroxidation in liposomes and react with DPPH acting as free radical scavengers. One-electron oxidation of the two complexes by radiolytically
generated azide radicals in Tx-100 micellar solutions produced phenoxyl radicals, indicating that the phenolic moiety of curcumin in the
complexes participates in free radical reactions. Depending on the structure, these two complexes possess different SOD activities, free radical
neutralizing abilities and antioxidant potentials. In addition, quantum chemical calculations with density functional theory have been performed
to support the experimental observations.
Ó 2006 Elsevier Masson SAS. All rights reserved.

Keywords: Cu(II)ecurcumin complex; Superoxide radicals; Free radicals; Density functional theory

1. Introduction
Abbreviations: DFT, density functional theory; DMSO, dimethyl sulfox-
ide; DPPH, 2,20 -diphenyl-1-picryl hydrazyl; EA, electron affinities; ECP, ef- Curcumin is a naturally occurring phytochemical found
fective core potential; EPR, electron paramagnetic resonance; NHE, normal in the rhizomes of Curcuma longa or turmeric, used for
hydrogen electrode; SOD, superoxide dismutase; TBARS, thiobarbituric
acid reactive substances; TD-DFT, time-dependent density functional theory;
centuries in a variety of pharmaceutical applications [1,2]
TE, total electronic energy; TEAP, tetraethyl ammonium perchlorate. including treatment for arthritis [3], as an anti-inflammatory
* Corresponding author. Present address: Department of Biochemistry and agent [4] and as an orally available treatment for diabetes
Molecular Biology, School of Medicine, Wayne State University, 540 E.Can- [5]. Curcumin fed amyloid-infused rats have shown signifi-
field, Detroit, MI-48201, United States. Tel.: þ1 313 993 9113; fax: þ1 313 cantly reduced levels of amyloid plaques [6]. Curcumin is
577 2765.
** Corresponding author. Fax: þ91 22 25505151. a potent antioxidant [7], inhibiting lipid peroxidation [8]
E-mail addresses: sdutta@med.wayne.edu (S. Dutta), kindira@barc.gov.in and effectively scavenging superoxide radical [9], the hy-
(K. Indira Priyadarsini). droxyl radical and nitrogen dioxide among other reactive

0223-5234/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2006.11.012
432 A. Barik et al. / European Journal of Medicinal Chemistry 42 (2007) 431e439

oxygen species [10]. Curcumin is well absorbed, both in vi- 2. Results and discussion
tro [11] and in vivo [12] and has exceedingly low toxicity
index [13]. It is a diferuloyl methane having two o-methoxy 2.1. Spectral and electrochemical
phenolic OH groups attached to the a,b-unsaturated b-dike- properties of the complexes
tone (heptadieneedione) moiety, which can form chelates of
the type 1:1 and 1:2 with copper, iron and other transition The normalized ground state UVeVIS spectra of the com-
metals [14,15]. This property of binding of curcumin to plexes in DMSO are given in Fig. 2(a) and (b) for 1:1 and 1:2
metals like iron and copper is considered as one of the use- complexes, respectively. The 1:1 Cu(II)ecurcumin complex
ful requirements for the treatment of Alzheimer’s disease exhibits absorption maximum at 426 nm and two shoulders
[6,16]. There also appears some correlations between drugs at 410 nm and 450 nm with extinction coefficients of
for Alzheimer’s disease and the metal complexes acting as 96700 M1 cm1 at 426 nm and the 1:2 Cu(II)ecurcumin
SOD mimics [15c,16]. complex exhibits maximum at 370 nm with extinction
Several metallocomplexes of curcumin have been synthe- coefficient of 22125 M1 cm1. The two complexes were in-
sized, characterized and evaluated for various biological soluble in neutral water. The 1:1 Cu(II)ecurcumin complex,
activities. Anti-arthritic properties of a five coordinate however, could be solubilised in alkaline water, while the
curcumin gold complex, Au(cur)2Cl, in which curcuminate 1:2 Cu(II)ecurcumin complex formed a precipitate. In
is bidentate were assessed in an adjuvant-induced rat polyar- DMSO, the 1:2 Cu(II)ecurcumin complex is more soluble
thritis model [17]. Greatly reduced paw swelling was seen than the 1:1 Cu(II)ecurcumin complex and both the com-
after three weeks of Au(cur)2Cl, 30 mg/kg/day by injection. plexes are soluble in lipids and Tx-100 micellar solutions.
Three manganese complexes with curcumin or one of two Aqueous solutions of lipids and micelles are micro heteroge-
related compounds, diacetylcurcumin and 4-(4-hydroxy-3- neous systems and provide hydrophobic sites to solublise
methoxy-phenyl)-1-[7-(4-hydroxy-3-methoxy-phenyl)-[1,4]di- water insoluble compounds [24].
azepam-5-ylidene]but-3-en-2-one [18], were evaluated for in The electrochemical profiles of the complexes were re-
vitro antioxidant properties and superoxide dismutase activity corded in DMSO solvent against Ag/AgCl reference electrode
with IC50 values for the former in the range 6.3e26.3 mM (voltammograms not shown). The E1/2 values in the voltam-
and for the later 8.9e29.99 mM. All the three complexes mograms for 1:1 and 1:2 complexes were found to be at
were also tested in vivo for their potential as neuroprotective 0.16 V and 0.20 V, respectively. The value for 1:1 agrees
agents in vascular dementia; Mn(cur)(OAc) showed signifi- with that reported in our earlier reference [23]. After correct-
cant protective effects in a transient ischemia/repurfusion ing the reference electrode as 0.222 V, the potentials for Cuþ2/
mouse model of neuronal damage [18]. Among a series of Cuþ redox couple for 1:1 and 1:2 complexes were calculated
metal curcuminoids, the Cu(curcumin)2 complexes were to be 0.38 and 0.42 V vs. NHE, respectively. The data indi-
most cytotoxic in cultured L929 cells [19] and showed sig- cates that the reduction potential values of the complexes
nificant reduction in solid tumor volume in ascites tumor- are not very much different and are within the range of
bearing mice. Syntheses of a series of 1,7-diarylheptanoids compounds that are expected to show SOD activity [25].
and their VO(IV), Co(II), Ni(II) and Cu(II) complexes were To estimate the formation constant for the equilibrium
reported [20]. The Cu(II) coordination complexes of these between Cu(II) and curcumin as given in Eq. (1), electrochem-
ligands had IC50 values of 6.6e11.1 mM in Erlich ascites ical scanning of Cu(II) in copper salt solution in DMSO was
tumor cells [20]. Testing for biological activity of the curcu- carried out in presence of different concentrations of
minoid ligands and various metal complexes including both curcumin.
in vivo administrations of compounds (i.p.) to assess antidia-
Kf
betic efficacy and in vitro cell studies of cytotoxic potentials Cu2þ þ p Curcumin # CuðCurcuminÞp ð1Þ
have been extensively studied by Orvig et al. [21]. Recently
Orvig et al. have reported vanadyl, gallium and indium where Kf is the formation constant of the said complex and p is
curcumin complexes for medicinal applications, [22] corrob- the coordination number or the number of molecules of curcu-
orating the importance of curcumin’s free phenolic OH min attached to the copper ion. The shift in the half wave po-
groups for scavenging oxidants and correlated with reduced tential (DE1/2) of the Cu(II) in the absence and in the presence
cytotoxic potential. of different concentrations of curcumin was plotted against the
One of the interests of our group has been to develop logarithmic concentration of curcumin, according to Eq. (2).
new transition metalecurcumin complexes as SOD mimics.
Recently we reported that the 1:1 complex of copper with 0:0591 0:0591
curcumin exhibits SOD activity and in vitro antioxidant DE1=2 ¼  log Kf  p log C ð2Þ
n n
activity [23]. In this paper we compare some of these re-
sults of 1:1 complex along with additional new experiments Here n and C correspond to the number of electrons trans-
with 1:2 complex of copper(II) and curcumin (Fig. 1) to un- ferred in the redox process and concentration of the ligand
derstand the effect of stoichiometry and structural changes (curcumin), respectively [26]. The data could be fitted to a lin-
on their SOD activity, free radical reactions and antioxidant ear plot, and the slope confirmed the formation of either 1:1
activity. complex or 1:2 complex and the intercept gave the value
 A. Barik et al. / European Journal of Medicinal Chemistry 42 (2007) 431e439 433

HO OH
H
H3CO C
OCH3
4 3
H3CCOO OH2 4O O3
5 Cu 5 Cu
1O O2 1O O2
H3CO OCH3 H3CO OCH3
6 C 9 6 C 9
7 7
H8 H8
HO OH HO OH

1:1 complex 1:2 complex

Fig. 1. Molecular structures of Cu(II)ecurcumin complexes (1:1 and 1:2).

of log Kf/0.059. For the estimation of Kf for 1:1 Cu(II)e oxygen atoms [23]. To recognize this asymmetry in the or-
curcumin complex, 500 mM copper acetate in DMSO was ti- thorhombic g tensor more distinctly, we performed careful
trated with 1 mM to 10 mM curcumin and by fitting the data EPR simulations. The results confirm that the 1:1 complex
to a linear plot (Fig. 3a) according to Eq. (2), Kf was deter- has distorted square planar structure around Cu2þ ion and
mined to be 3.7  1014. Similar exercise was carried out in the values of orthorhombic g tensor have been estimated
the case of 1:2 complex, where 500 mM copper chloride solu- to be g1 ¼ 2.325, g2 ¼ 2.071, g3 ¼ 2.062, A1 ¼ 155 G. The
tion was titrated with 1 mM to 10 mM curcumin and by fitting values of A2 and A3 were unresolved as they are within
the data to a linear plot (Fig. 3b) according to Eq. (2) Kf was the line widths. Diaz et al. have reported a good correlation
determined to be 3.9  1015. These results therefore indicate between f factor ( gjj/Ajj, where Ajj is expressed in cm1) and
that both 1:1 and 1:2 Cu(II)ecurcumin complexes are consid- SOD like activity of copper(II) complexes [27]. A f factor
erably stable in the medium studied and 1:2 complex was value smaller than 135 cm is obtained for square planar
shown to have higher stability as compared to that of 1:1 Cu(II) complexes, and this value increases with increasing
complex. tetrahedral distortion. The f value for Cu, Zn SOD is
The EPR spectra for both 1:1 and 1:2 Cu(II)ecurcumin 160 cm, indicating a tetrahedral distortion from square pla-
complexes were recorded in DMSO at 77 K as given in nar geometry and is one of the features that enhance the
Fig. 4a and b, respectively. In the same figure, the EPR catalytic activity of the enzyme. From the above EPR data
spectra, simulated by using Bruker Simfonia software are the f values for 1:1 and 1:2 complexes were determined to
superimposed with the dotted line. The EPR spectrum of be 150 ( g1/A1), and 135 cm ( gjj/Ajj), respectively. Therefore
1:2 Cu(II)ecurcumin complex showed axial symmetry 1:1 Cu(II)ecurcumin complex exhibiting appreciable square
( gjj ¼ 2.295, gt ¼ 2.0852, Ajj ¼ 170 G) which is associated planar distortion is expected to show high SOD-like activity.
with square planar coordination of four equivalent oxygen These EPR results have been further supported by quantum
atoms around copper(II) ions. Earlier we reported that the chemical calculation.
EPR spectrum of 1:1 complex shows slightly orthorhombic
g tensor, which could be due to asymmetric coordination
around Cu2þ ion, as Cuþ2 is coordinated to four inequivalent -0.5
-0.5 (b)
-0.6
E1/2

-0.7
1.0

-0.6 (a) -0.8

E1/2

0.8
-5 -4 -3 -2
log C
Absorbance

(a)
0.6

-0.7
(b)
0.4

0.2
-5 -4 -3 -2
log C
0.0
300 350 400 450 500 550 600 650 Fig. 3. Variation in the half wave potential of Cuþ2 ion in presence of different
Wavelength, nm concentrations of curcumin, (a) 500 mM Cu(OCOCH3)2 and (b) 500 mM
CuCl2. Solid lines represent linear fit for the data according to Eq. (2). Exper-
Fig. 2. Normalized absorption spectrum (a) and (b) of 1:1 and 1:2 Cu(II)e imental conditions: copper salts dissolved in DMSO with 0.1 M TEAP as
curcumin complex, respectively in DMSO. supporting electrolyte, saturated with nitrogen, scan rate 100 mV/s.
434 A. Barik et al. / European Journal of Medicinal Chemistry 42 (2007) 431e439

4.0 denotes the same system with the addition of Cu(II)ecomplex

and ‘‘Blank’’ denotes the system without the enzyme xanthine
2.0 oxidase. This %I when plotted against the concentration of
(b) the complex as shown in Fig. 5 indicates linearity in the
0.0 concentration range studied, from which the IC50 value (the
g
concentration at which the absorbance is half its initial value)
-2.0 was determined to be 6.7 mM and 68 mM for 1:1 and 1:2 com-
A.U.

2.0
1.0
plexes, respectively. The IC50 value for curcumin under the
g1 (a) same condition was found to be 86.8 mM, which is slightly
-4.0 0.0
-1.0 higher than that of the 1:2 complex, but 1:1 complex shows
g2
-6.0 -2.0 10 times more SOD activity as compared to both 1:2 complex
g
-3.0 g3 and curcumin. The IC50 values for SOD activity for the two
-8.0
-4.0 complexes estimated by NBT and cytochrome c methods,
2500 3000 3500 4000 although differ significantly indicate same trend that 1:1 com-
2500 3000 3500 4000 plex is more active than 1:2 complex.
H, Gauss The rate constants for the scavenging of superoxide radical
Fig. 4. EPR spectra of (a) 1:1 and (b) 1:2 Cu(II)ecurcumin complexes in with Cu(II)ecurcumin complexes were calculated according
DMSO at 77 K. Solid lines represent experimental spectra and dashed lines to the competition kinetics method for the above competing
represent the simulated spectra. reactions given in Eqs. (3) and (4), using the Eq. (6) given
below:
2.2. Superoxide scavenging reactions of the complexes    
A0 k2 ½complex
1 ¼ ð6Þ
Earlier we reported the superoxide scavenging ability of 1:1 A k1 ½cytochrome c
complex using NBTþ2 as a standard, monitoring NBTþ
formed by the reduction of NBTþ2 with superoxide radical where A0 and A are the maximum absorbance values at 550 nm
generated by xanthine/xanthine oxidase system and the IC50 in the absence and presence of the complex, respectively.
value for 1:1 complex was reported to be 0.17 mM [23], under Slope of the linear plot for [(A0/A)1] vs. [complex]/[cyto-
the same conditions, the 1:2 complex gave an IC50 value chrome c], gives k2/k1, using the value of k1 as 5.8 
of w2 mM. After extensive literature search, it was found 105 M1 s1 for the reaction of superoxide radical with cyto-
that NBTþ2 is not very selective to superoxide radicals and chrome c [29], k2 was estimated and values for 1:1 and 1:2
may get reduced by xanthine oxidase even under anaerobic complexes are listed in Table 1. The rate constant determined
conditions and thereby introduces errors in the estimations
of SOD activity. Therefore we employed a more sensitive
method using cytochrome c as standard to estimate their com- [1:2 Complex], M
10 20 30 40 50 60
parative SOD activity [28].
0 2 4 6
When 1:1 and 1:2 complexes were added to the reaction 50 1.2
medium comprising of xanthine, xanthine oxidase, EDTA, 80
(d)
cytochrome c (Fe3þ), in Tris buffer, there is a competition be-
(Ao/A-1)

0.8
40
tween cytochrome c (Fe3þ) and the complexes to react with (c)
0.4 (a) 60
superoxide radical as given in Eqs. (3) and (4). % Inhibition
% Inhibition

30 0.0

k1
0.0 0.2 0.4 0.6
cytochrome c Fe3þ þ O,
2 / cytochrome c Fe
2þ
ð3Þ [Cytochrome C]/
40
[Complex] (b)
20
k2
Complex þ O,
2 / Product ð4Þ 20
10
In presence of the complex, the DA/min at 550 nm due to
cytochrome c (Fe2þ) was found to be increased, and the final
0 0
absorbance at 550 nm decreased with the increasing 1 2 3 4 5 6
concentration of the complex. The percentage inhibition [1:1 Complex], M
(%I) of the Cu(II)ecurcumin complex was calculated accord-
Fig. 5. Percentage inhibition (%I ) of superoxide radicals formed by xanthine/xan-
ing to Eq. (5). thine oxidase enzyme by different concentrations of the complexes assayed by
cytochrome c (Fe2þ) absorption at 550 nm, (a) in presence of 1:1 Cu(II)e
DA=minðUninhibitedÞ  DA=minðInhibitedÞ curcumin complex and (b) in presence of 1:2 Cu(II) ecurcumin complex. Inset
%I ¼  100 ð5Þ
DA=minðUninhibitedÞ  DA=minðBlankÞ shows competition kinetic plots for the reaction of superoxide radical with
cytochrome c in presence of (c) 1:1 Cu(II)ecurcumin complex and (d) 1:2
Here, ‘‘Uninhibited’’ denotes the system consisting of xan- Cu(II)ecurcumin complex. (Experimental conditions: 50 mM xanthine, 10 mU/
thine, xanthine oxidase, EDTA and cytochrome c, ‘‘Inhibited’’ ml of xanthine oxidase, 9.52 mM cytochrome c, and 600 mM EDTA, 30% DMSO).
 A. Barik et al. / European Journal of Medicinal Chemistry 42 (2007) 431e439 435

Table 1
Comparative properties of Cu(II)ecurcumin complexes (1:1 and 1:2)
Property Complex (1:1) Complex (1:2)
a
Absorption maximum (DMSO) 426 nm , shoulders at 410 nm and 450 nm 370 nm
Absorption maximum (2% Tx-100) 426 nm 380 nm
Extinction coefficient at absorption maximum 97000  200a 22125  100
(DMSO) (M1 cm1)
Cu(II)/Cu(I) potential vs. NHE 0.38  0.02 V 0.42  0.03 V
Formation constant (Kf) 3.7  1014 3.9  1015
EPR spectra in DMSO at 77 K g1 ¼ 2.325, A1 ¼ 155 G, g2 ¼ 2.071, g3 ¼ 2.0620 gjj ¼ 2.295, Ajj ¼ 170 G, gt ¼ 2.0852
k (O2) þ compound, competition kinetics 7.1  0.1  105 1.04  0.17  105


with cytochrome c (M1 s1)

IC50 value (xanthine/xanthine oxidase assay) 6.7 mM 68.0 mM
Inhibition of lipid peroxidation in liposomes, 98% 15%
10 mM, dose 210 Gy
k (DPPH) þ compound, DMSO (M1 s1) 156  12a (1.1  0.1)  103
k (N3) þ compound, Tx-100 (M1 s1) (2.1  0.4)  109a (4.4  1.1)  108


Absorption maximum of phenoxyl radical 510e520 nma 520e530 nm

Electron affinity (kcal/mol) 73.61 40.67
OeH bond dissociation energy (kcal/mol) 88.51a 87.83
OeH proton dissociation enthalpy (kcal/mol) 322.28 327.81
Gas phase ionization potential (kcal/mol) 149.35a 142.65
a
From Ref. [23].

for curcumin by this method (4.06  0.6  104 M1 s1) As the 1:2 complex is insoluble in water, it was dissolved in
agrees well with that estimated earlier by pulse radiolysis aqueous Tx-100 solutions. The transient spectrum obtained on

[30]. The 1:1 complex shows nearly seven times higher rate reaction of N3 with 1:2 complex recorded 40 ms after the pulse,
constant for superoxide radical as compared to the 1:2 is given in Fig. 7. The spectrum matches qualitatively with that

complex. obtained on reaction of N3 with 1:1 complex and curcumin, in-
dicating that the reaction produces similar phenoxyl radicals
2.3. Inhibition of lipid peroxidation by the complexes which are generated by one-electron oxidation followed by
proton loss from curcumin phenolic OH group [31]. The rate

In order to test the antioxidant activity of the complexes, constants for the reaction of N3 radicals with 1:2 Cu(II)e
we studied their ability to inhibit lipid peroxidation in liposo- curcumin complex was estimated to be 4.4  1.1  108
mal solutions. Peroxidation of lipid is a measure of damage to M1 s1 by following the observed pseudo first order rate con-
the membrane lipids caused by the attack of reactive oxygen stant for the formation of the transient at 500 nm as a function
species. Inhibition of lipid peroxidation by any external agent
is used to evaluate its antioxidant capacity [24].
Fig. 6 shows bar graph for the comparative TBARS forma- 10.0
tion in liposomes in absence and presence of 10 mM of 1:1 and Blank
1:2 Cu(II)ecurcumin complexes on irradiation with g-radia- Curcumin
TBARS, (nmoles/mg liposomes)

tion at absorbed doses of 210, 420 and 630 Gy. The results 8.0 1:2 complex
confirm that both the complexes inhibit lipid peroxidation at 1:1 complex

all the doses. However, 1:1 Cu(II)ecurcumin complex inhibits

peroxidation to a greater extent than either the 1:2 Cu(II)e 6.0
curcumin complex or curcumin, therefore it is expected to
be a better antioxidant.
4.0
2.4. Free radical-scavenging ability of the complexes

Earlier [23] we reported the free radical kinetics for the re- 2.0

actions of 1:1 complex with azide radicals (N3 radicals) and
DPPH radicals using pulse radiolysis and stopped-flow spec-
 0.0
trometer. While N3 radicals react by electron transfer, the re- 210 420 630
action with DPPH occurs by hydrogen atom abstraction Dose, Gy
[31,32]. Generally hydrogen atom transfer is slower than
Fig. 6. g-Radiation induced lipid peroxidation in liposomes (1 mg/ml) in the
that of an electron transfer process. The same experiments absence and presence of curcumin, 1:2 and 1:1 Cu(II)ecurcumin complex
were carried out with 1:2 complex and compared with 1:1 [substrate ¼ 10 mM] at different absorbed doses. Lipid peroxidation was
complex. assayed by reduction in formation of TBARS.
436 A. Barik et al. / European Journal of Medicinal Chemistry 42 (2007) 431e439

0.006 Table 2
Selected structural parameters (bond length in Å and dihedral angle in degree)
1.5

kobs x 10-5, s-1

of curcumin and the complexes (1:1 and 1:2)
1.0 Complex (1:1) Complex (1:2) Curcumin
0.004 O1eCu5 (H5) 1.89a 1.94 1.04
ΔAbsorbance/Gy

0.5
O2eCu5 (H5) 1.92a 1.94 1.55
0.0 O3eCu5 2.01a 1.94 e
0.1 0.2 0.3
O4eCu5 1.95a 1.94 e
[1:2 complex], mM
O1eC6 1.32 1.32 1.36
0.002 C6eC7 1.42 1.42 1.40
C7eH8 1.09 1.09 1.08
C7eC9 1.41 1.42 1.44
C9eC2 1.33 1.32 1.30
:O1eO2eO3eO4 5.36 22.39 e
0.000 The numbering of the atoms corresponds to that shown in Fig. 1.
a
400 450 500 550 600 650 700 From Ref. [23].
Wavelength, nm

Fig. 7. Difference absorption spectrum of transient obtained by reaction of curcumin complex are different (Table 2), while those of the

0.1 mM 1:2 Cu(II)ecurcumin complex with N3 radical at pH 7; (experimental
1:2 Cu(II)ecurcumin complex are identical (1.94 Å, Table
conditions, N2O saturated aqueous solutions of 1:2 complex containing 30 mM
Tx-100 and 0.1 M sodium azide were pulse radiolysed, dose/pulse: 13 Gy). In- 2), which is consistent with the EPR-derived results where or-
set shows variation in the observed rate constant as a function of concentration thorhombic symmetry was observed due to coordination to
of 1:2 Cu(II)ecurcumin complex. four non-equivalent oxygens in the 1:1 Cu(II)ecurcumin com-
plex, while square planar coordination is observed around the
copper in the 1:2 Cu(II)ecurcumin complex. In addition, the
of concentration as given as inset of Fig. 7, the value reported dihedral angle of O1eO2eO3eO4 of the 1:2 Cu(II)ecurcu-
for 1:1 complex under similar conditions is given in Table 1. min complex (22.39 ) is much larger than that of the 1:1
As observed earlier, the transients in all the cases do not react Cu(II)ecurcumin complex (5.36 ), which likely results from
with oxygen and therefore not converted into any other new the steric repulsion between the methoxy groups in both
radicals [23]. curcumins of the 1:2 Cu(II)ecurcumin complex (Fig. 8).
The reaction of DPPH with 1:2 complex was studied in Moreover, as shown in Table 2, copper coordination has little
DMSO. The absorptionetime plots for the decay of DPPH influence on the geometry of parent curcumin, except for the
(10 mM) in presence of the complex (250 mM to 1 mM) at binding site. As the B3LYP/LANL2DZ method can provide
517 nm could be fitted to single exponential function to obtain quite accurate structural parameters of Cu(II) complexes,
the observed rate constant. In presence of different concentra- such as bond length, with a error of less than 0.01 Å [15c],
tions of the complex, the observed rate constant was plotted as the theoretical geometries of the complexes (1:1 and 1:2
a function of the concentration of the complex and from the Cu(II)ecurcumin) are of significance in comparing with other
slope, the bimolecular rate constants for the reaction of 1:2 SOD mimics.
Cu(II)ecurcumin complexes with DPPH was determined to To evaluate the ability of the complexes (1:1 and 1:2 Cu(II)e
be 1.1  103 M1 s1. The bimolecular rate constant for the curcumin) to scavenge superoxide anion and DPPH radical, we
reaction of DPPH with 1:1 complex was earlier reported to calculated their electron affinities (EAs), OeH proton dissoci-
be 1.6  102 M1 s1. The reaction of these complexes with ation enthalpies (PDEs), OeH bond dissociation enthalpies

N3 radicals and DPPH radicals indicate that the two complexes (BDEs) and ionization potentials (IPs), which have been recog-
are able to participate in both electron transfer and hydrogen nized as appropriate theoretical parameters to measure the elec-
atom transfer reactions. This may be mainly from the phenolic tron-accepting, proton-donating, H-atom-donating and
OH groups of curcumin which are intact and unaffected by the electron-donating abilities, respectively of antioxidants [34].
chelation with copper. All these above estimated parameters The lower the parameters are higher the antioxidant activities.
for the two complexes are listed in Table 1. According to the calculated results (Table 1), the EA of 1:1
Cu(II)ecurcumin complex is w30 kcal/mol lower than that of
2.5. Quantum chemical calculations 1:2 complex, which agrees well with the lower reduction po-
tential of the former than the latter. The OeH PDE of the
The physicochemical properties of the complexes (1:1 and 1:1 Cu(II)ecurcumin complex is 5.53 kcal/mol lower than
1:2 Cu(II)ecurcumin) have been studied employing the DFT that of the 1:2 counterpart, while the OeH BDE of the 1:1
calculations, which have been successfully used in determin- Cu(II)ecurcumin complex is about 0.68 kcal/mol higher
ing the metalechelating modes [15c,33] and in exploring the than that of the 1:2 Cu(II)ecurcumin complex, in line with
radical-scavenging mechanisms of antioxidants as well [34]. the experimental observation that the former is more active
Table 2 lists selected structural parameters of curcumin and in scavenging superoxide anion radical (through proton trans-
the complexes (1:1 and 1:2 Cu(II)ecurcumin). It can be seen fer or electron transfer), but less active in scavenging DPPH
that the four Cu(II)eO bond lengths of the 1:1 Cu(II)e radical (through H-atom transfer) than the later. In addition,
 A. Barik et al. / European Journal of Medicinal Chemistry 42 (2007) 431e439 437

Fig. 8. Stereo views of optimized structures of Cu(II)ecurcumin complex (1:2). The arrows indicate the steric repulsion between the methoxy groups of both
curcumins.

as known to us, the azide radical-scavenging ability can be From these studies it appears that the increasing flexibility
characterized by their IP and the lower the IP higher the activ- of the complex to accommodate the reduced cuprous species
ity. However, the different azide radical-scavenging activity of in tetrahedral or linear environments during superoxide radical
the complexes (1:1 and 1:2 Cu(II)ecurcumin) (Table 1) cannot reaction plays a crucial role in the overall antioxidant and
be elucidated by their difference in IP (Table 1), which may SOD activity. The EPR spectral results indicate appreciable
arise from the fact that solvent effect is not considered during distortion from the square planar geometry for 1:1 complex.
the calculation. From this it can be concluded that the 1:1 complex would
In brief, a large part of physicochemical properties of be able to undergo and sustain the distortion from square pla-
Cu(II)ecurcumin complexes (1:1 and 1:2) can be elucidated nar geometry to the distorted tetrahedral one during its reac-
by DFT calculations, which not only provide deeper insights tion with superoxide radical. This allows for the compound
into the experimental findings but also display the potential to remain intact and undergo many redox cycles and hence
of theoretical methods in antioxidant study. as an efficient antioxidant. The 1:2 complex on the other
hand is planar but rigid and hence cannot undergo the distor-
3. Conclusions tions and therefore is less powerful antioxidant. Thus the 1:1
Cu(II)ecurcumin complex having distortion from square pla-
In the recent years, metal-chelating properties of curcumin nar geometry is expected to exhibit a better SOD activity and
and its implications in use of curcumin as a multipotent agent is also a good free radical scavenger.
to combat oxidative stress and Alzheimer’s disease and also as
a probable SOD mimic is gaining a lot of significance. Curcu-
min forms both 1:1 and 1:2 complexes with copper(II). We 4. Experimental
were able to synthesize and characterize these two complexes.
EPR spectra confirm the structures of the complexes as dis- 4.1. Chemicals
torted orthorhombic and symmetric square planar, respec-
tively. Reduction potential of the reversible Cuþ2/Cuþ Curcumin, xanthine, xanthine oxidase, cytochrome c, SOD,
couple was found to be w0.4 V which is in within the range DPPH, Tx-100 (Sigma/Aldrich) were purchased from the local
expected for SOD mimics [25]. However, the SOD activity market. AR grade, cupric chloride (CuCl2$2H2O) and cupric
of the complexes studied by xanthine and xanthine oxidase as- acetate (Cu(OOCCH3)2$2H2O) and spectrograde dimethyl
say indicated that the 1:1 Cu(II)ecurcumin complex is nearly sulfoxide (DMSO) from Spectro Chem. India, Mumbai were
ten times more potent as SOD mimic than the 1:2 Cu(II)ecur- used as received. All the other reagents were of the best purity
cumin complex. Similarly the rate constant for the scavenging available. Absorption spectrometric studies were carried out
of superoxide radical of the 1:1 complex has been found to be using JASCO Ve530 spectrophotometer.
seven times higher than that of the 1:2 Cu(II)ecurcumin com- The synthesis of 1:1 and 1:2 complexes between copper(II)
plex. The complexes were also tested for their ability to inhibit ion and curcumin has been reported in our earlier papers and
radiation induced lipid peroxidation in liposomes, where the described briefly here [15a,23]. The 1:1 Cu(II)ecurcumin
1:1 Cu(II)ecurcumin complex was found to exhibit inhibition complex was synthesized by mixing equimolar ratio of copper
of lipid peroxidation to a greater extent than the 1:2 Cu(II)e acetate and curcumin in dry ethanol and refluxed for 3 h under
curcumin complex. The two complexes show similar electron nitrogen atmosphere. 1:2 Cu(II)ecurcumin complex was
and hydrogen atom transfer reactions with free radicals and synthesized by mixing methanolic solution of curcumin and
produce the phenoxyl radicals similar to those produced aqueous solution of CuCl2$H2O under reflux for 3 h in the
from curcumin. Factors like steric hindrance, OeH bond dis- stoichiometry of 1:2 (copper:curcumin). The precipitated com-
sociation energies and ionization potential are responsible for plexes were filtered, washed with cold water and ethanol and
this. Theoretical calculations support these observations. dried in vacuum.
438 A. Barik et al. / European Journal of Medicinal Chemistry 42 (2007) 431e439

4.2. Analytical experiments Pulse radiolysis studies were carried out using 7 MeV elec-
tron pulses (50 ns) from a linear electron accelerator coupled
EPR spectra were recorded on Bruker ESP 300 spectrome- with absorption detection [40]. The dose per pulse was close
ter, operated at X-band frequency (9e10 GHz) using 100 kHz to 12e14 Gy. One-electron specific oxidants and azide radi-
cals (E(N3/N

field modulation. DPPH sample was used as a field marker. 3 ) ¼ 1.33 V vs. NHE) were generated by the ra-
The spectra were recorded at 77 K using a liquid nitrogen diolysis of N2O purged aqueous solutions containing 0.1 M
Dewar insert. sodium azide according to the procedure reported in Ref. [41].
Cyclic voltammetry was carried out using Eco-Chemie po-
tentiostat AUTOLAB-20 cyclic voltameter attached with VA 4.3. Theoretical
663 stand. For the electrochemical estimation of Cuþ2/Cuþ
couple, solutions of the above complexes in DMSO were The theoretical calculation procedures are as follows. The
scanned employing three-electrode system of Ag/AgCl elec- full geometry optimization for each molecule was performed
trode as the reference electrode, glassy carbon electrode as in vacuo by employing hybrid density functional theory
the working electrode, and platinum rod as the counter elec- (DFT) [42] and B3LYP functional [43]. The standard dou-
trode. The electrochemical setup has been calibrated using ble-z basis set was used for all light elements, while for
Cdþ2 solution. For the estimation of formation constant of metals, nonrelativistic effective core potential (ECP) was em-
the complexes, electrochemical scanning was carried out taking ployed. The valence basis set used in connection with the ECP
the Cu(II) salt solution, both in the presence and absence of is essentially of double-z quality (the LANL2DZ basis set). As
varying concentration of curcumin in an electrochemical cell the molecules are rather large, the B3LYP/LANL2DZ method
comprising of a hanging mercury drop electrode, a Ag/AgCl failed to give zero point vibrational energy and thermal correc-
electrode as reference and a platinum rod as the counter elec- tion to energy. However, according to previous studies
trode. The voltammograms were recorded at room temperature [15c,44], the physicochemical parameters derived from total
in DMSO solvent using 0.1 M tetraethyl ammonium perchlorate electronic energy (TE) are fairly accurate in a relative sense.
(TEAP) as the supporting electrolyte. The solution was kept All the calculations were accomplished by Gaussian 03 pro-
under nitrogen purging to avoid interference from oxygen. gram package [45].
Superoxide radical was generated by enzymatic reaction of
xanthine (50 mM) with xanthine oxidase (10 mU/ml) in pres- Acknowledgements
ence of Tris buffer (pH 7.4) and 600 mM EDTA according to
the procedure given in reference [35]. The superoxide radical The authors would like to express their sincere thanks to.
generated by this method was allowed to react with cyto- Dr. Tulsi Mukherjee and Dr S. K. Sarkar, for the encourage-
chrome c (Fe3þ) (9.5 mM) to produce reduced cytochrome c ment and support. Dr. Shen and Dr. Zhang thank Dr. Hong-
(Fe2þ), absorbing at 550 nm. The change in absorbance per Fang Ji for her help in DFT calculation. This work was
unit time DA/min was monitored up to 300 s, where DA is partially supported by National Key Project for Basic
the difference in absorbance at 550 nm. The concentration of Research (2003CB114400) and National Natural Science
xanthine oxidase is adjusted such that DA/min is w0.025. Foundation of China (30100035 and 30570383).
Lipid peroxidation studies were also carried out in phos-
phatidylcholine liposomes (2:1 mixture of phosphatidylcho-
line and cholesterol). Liposomes were prepared according to References
the procedure given in Ref. [36]. Liposomes suspended in
[1] I. Chattopadhyay, K. Biswas, U. Bandyopdhyay, R.K. Banerjee, Curr.
pH 7.4 phosphate buffers with and without substrate were Sci. 87 (2004) 44e53.
exposed for different doses of g-radiation from 60Co g-source [2] (a) O.P. Sharma, Biochem. Pharmacol. 25 (1976) 1811e1812;
with a dose rate of 42 Gy min1 as measured by standard (b) R.N. Chopra, J.C. Chopra, K.L. Handa, L.D. Kapur, Indigenous
Fricke dosimeter [37], and the lipid peroxidation was esti- Drugs of India, second ed. Dhur, Calcutta, 1958.
mated by TBARS at 532 nm (3532 ¼ 1.56  105 M1 cm1) [3] S.D. Deodhar, R. Sethi, R.C. Srimal, Indian J. Med. Res. 71 (1980)
632e634.
as given in earlier Refs. [38,39]. All the above experiments [4] (a) T.S. Rao, N. Basu, H.H. Siddiqui, Indian J. Med. Res. 75 (1982)
were repeated twice to confirm their reproducibility, where- 574e578;
ever the applicable standard deviations of different experimen- (b) R.C. Srimal, B.N.J. Dhawan, J. Pharm. Pharmacol. 25 (1973)
tal data have been given as errors. 447e452.
The kinetics of DPPH reactions with the complexes was [5] N. Arun, N. Nalini, Plant Foods Hum. Nutr. 57 (2002) 41e52.
[6] L. Baum, A. Ng, J. Alzheimers Dis. 6 (2004) 367e377.
studied using SX-18 MV stopped-flow spectrometer from [7] A.C.P. Reddy, B.R. Lokesh, Food Chem. Toxicol. 32 (1994) 279e283.
Applied Photophysics Ltd., UK in the single mixing mode. [8] S. Daniel, J.L. Limson, A. Dairam, G.M. Watkins, S.J. Daya, Inorg.
Here the two mixing syringes contain DMSO solutions of Biochem. 98 (2004) 266e275.
DPPH and Cu(II)ecurcumin complexes separately and the [9] M.K. Unnikrishnan, M.N.A. Rao, Mol. Cell. Biochem. 146 (1995)
time-dependent absorption changes at 517 nm due to DPPH 35e37.
[10] Y. Sugiyama, S. Kawakishi, T. Osawa, Biochem. Pharmacol. 52 (1996)
after mixing were monitored in presence of different concen- 519e525.
trations of the complex [31,32]. Three independent runs were [11] V. Ravindranath, N. Chandrasekhara, Toxicology 16 (1980) 259e265.
analysed to obtain average observed rate constant. [12] V. Ravindranath, N. Chandrasekhara, Toxicology 20 (1981) 251e257.
 A. Barik et al. / European Journal of Medicinal Chemistry 42 (2007) 431e439 439

[13] V. Ravindranath, N. Chandrasekhara, Toxicology 22 (1982) 337e344. [29] G.R. Hodges, M.J. Young, T. Paul, K.U. Ingold, Free Radic. Biol. Med.
[14] (a) J.P. Annaraj, K.M. Ponvel, P. Athappan, S. Srinivasan, Transit. Metal 29 (2000) 434e441.
Chem. 29 (2004) 722e727; [30] K.I. Priyadarsini, Free Radic. Biol. Med. 23 (1997) 838e843.
(b) O. Vajragupta, P. Boonchoong, L.J. Berliner, Free Radic. Res. 38 [31] K.I. Priyadarsini, D.K. Maity, G.H. Naik, M. Sudheer Kumar,
(2004) 303e314; M.K. Unnikrishnan, J.G. Satav, H. Mohan, Free Radic. Biol. Med. 35
(c) Y. Sumanont, Y. Murakami, M. Tohda, O. Vajragupta, K. Matsumoto, (5) (2003) 475e484.
H. Watanabe, Biol. Pharm. Bull. 27 (2004) 170e173. [32] S.S. Kumar, K.I. Priyadarsini, K.B. Sainis, Redox Rep. 7 (2002) 35e40.
[15] (a) S. Dutta, A. Murugkar, N. Gandhe, S. Padhye, Metal Based Drugs 8 [33] P.J. Stephens, F.J. Devlin, C.F. Chabalowski, M.J. Frisch, J. Phys. Chem.
(2001) 183e187; 98 (1994) 11623e11627.
(b) M. Borsari, E. Ferrari, R. Grandi, M. Saladini, Inorg. Chim. Acta 328 [34] H.Y. Zhang, Curr. Comput. Aided Drug Des. 1 (2005) 257e273.
(2002) 61e68; [35] J.M. McCord, I. Fridovich, J. Biol. Chem. 244 (1969) 6049e6053.
(c) H.F. Ji, H.Y. Zhang, Bioorg. Med. Chem. Lett. 15 (2005) 21e24; [36] B.N. Pandey, K.P. Mishra, Radiat. Phys. Chem. 54 (1999) 481e489.
(d) Y. Jiao, J. Wilkinson IV, E. Christine Pietsch, J.L. Buss, W. Wang, [37] J.W.T. Spinks, R.J. Woods, An Introduction to Radiation Chemistry, third
R. Planalp, F.M. Torti, S.V. Torti, Free Radic. Biol. Med. 40 (2006) ed. Weily Int. Sci., 1990, p. 95.
1152e1160. [38] N. Sreejayan, M.N.A. Rao, K.I. Priyadarsini, T.P.A. Devasagayam, Int. J.
[16] H.Y. Zhang, FEBS Lett. 579 (2005) 5260e5264. Pharmacol. 151 (1997) 127e130.
[17] K.K. Sharma, S. Chandra, D.K. Basu, Inorg. Chim. Acta 135 (1987) [39] S.M. Khopde, K.I. Priyadarsini, S.N. Guha, J.C. Satav, P. Venkatesan,
47e48. M.N.A. Rao, Biosci. Biotechnol. Biochem. 64 (2000) 503e509.
[18] O. Vajragupta, P. Boonchoong, H. Watanabe, M. Tohda, N. Kammasud, [40] S.N. Guha, P.N. Moorthy, K. Kishore, D.B. Naik, K.N. Rao, Proc. Indian
Y. Sumanont, Free Radic. Biol. Med. 35 (2003) 1632e1644. Acad. Sci. (Chem. Sci.) 99 (1987) 261e271.
[19] K. Krishnankutty, P. Venugopal, Synth. React. Inorg. Met. Org. Chem. 28 [41] P. Neta, R.E. Huie, J. Phys. Chem. Ref. Data 17 (1988) 1027e1284.
(1998) 1313e1325. [42] (a) P. Hohenberg, W. Kohn, Phys. Rev. 136B (1964) 864e871;
[20] K. Krishnankutty, P. Venugopal, Synth. React. Inorg. Met. Org. Chem. 33 (b) W. Kohn, L.J. Sham, Phys. Rev. 140A (1965) 1133e1138.
(2003) 343e358. [43] (a) C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785e789;
[21] (a) K.H. Thompson, B.D. Liboiron, Y. Sun, K.D.D. Bellman, (b) A.D. Becke, J. Chem. Phys. 98 (1993) 1372e1377;
I.A. Setyawati, B.O. Patrick, V. Karunaratne, G. Rawji, J. Wheeler, (c) P.J. Stephens, F.J. Devlin, C.F. Chabalowski, M.J. Frisch, J. Phys.
K. Sutton, S. Bhanot, C. Cassidy, J.H. McNeill, C. Orvig, J. Inorg. Bio- Chem. 98 (1994) 11623e11627.
chem. 8 (2003) 66e74; [44] H.F. Ji, H.Y. Zhang, Chem. Res. Toxicol. 17 (2004) 471e475.
(b) K.H. Thompson, K. Böhmerle, E. Polishchuk, C. Martins, P. Toleikis, [45] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J. Tse, V. Yuen, J.H. McNeill, C. Orvig, J. Inorg. Biochem. 98 (2004) J.R. Cheeseman, J.A. Montgomery Jr., T. Vreven, K.N. Kudin,
2063e2070. J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone,
[22] K. Mohammadi, K.H. Thompson, B.O. Patrick, T. Storr, C. Martins, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson,
E. Polishchuk, V.G. Yuen, J.H. McNeill, C. Orvig, J. Inorg. Biochem. H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
99 (2005) 2215e2225. M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene,
[23] A. Barik, B. Mishra, L. Shen, H. Mohan, R.M. Kadam, S. Dutta, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, V. Bakken, C. Adamo,
H.Y. Zhang, K.I. Priyadarsini, Free Radic. Biol. Med. 39 (2005) J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin,
811e822. R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma,
[24] B. Halliwell, J.M.C. Gutteridge, Free Radicals in Biology and Medicine, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski,
second ed. Clarendon Press, Oxford, 1989. S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick,
[25] Z. Durackova, A.M. Mendiola, T.M. Sevilla, A. Valent, Bioelectrochem. A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui,
Bioenerg. 48 (1999) 109e116. A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu,
[26] J.J. Lingane, Chem. Rev. 29 (1941) 1e35. A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox,
[27] A. Diaz, R. Cao, A. Fragoso, I. Sanchez, Inorg. Chem. Commun. 2 T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara,
(1999) 358e360. M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong,
[28] C. Auclair, M. Torres, J. Hakim, FEBS Lett. 89 (1978) 26e28. C. Gonzalez, J.A. Pople, Gaussian 03, Gaussian, Inc., 2003.