Synthesis, antioxidant evaluation, and quantitative structure-activity relationship studies of chalcones

Article abstract of DOI:10.1007/s00044-010-9342-1

Synthesis, antioxidant activity, and quantitative structure-activity relationship (QSAR) of 25 of chalcone derivatives is reported here. They were synthesized by Claisen-Schmidt reaction and were characterized by FTIR, NMR, and mass spectroscopy. Antioxidant activity is evaluated through four different methods namely, superoxide radical-scavenging, hydrogen peroxide scavenging, reducing power, and DPPH radical-scavenging assays. Generally, compounds with -SCH₃ and -OCH₃ in the para position of the A-ring and -OH in the B-ring were more active than others. In few cases some of the compounds were more active than ascorbic acid or butylated hydroxytoluene. QSAR was developed correlating the antioxidant activity with the structural features of the compounds and the predictive capability of the models was estimated using internal and external validation methods. All the predictions were within the 99% confidence level. Spatial, structural, and lipophilic properties of the compounds determine their antioxidant properties.

Full text of DOI:10.1007/s00044-010-9342-1

MEDICINAL
CHEMISTRY
RESEARCH
Med Chem Res (2011) 20:482–492
DOI 10.1007/s00044-010-9342-1
ORIGINAL RESEARCH
Synthesis, antioxidant evaluation, and quantitative
structure–activity relationship studies of chalcones
•
P. M. Sivakumar P. K. Prabhakar
•
M. Doble
Received: 25 August 2009 / Accepted: 3 March 2010 / Published online: 23 March 2010
ꢀ Springer Science+Business Media, LLC 2011
Abstract Synthesis, antioxidant activity, and quantitative
structure–activity relationship (QSAR) of 25 of chalcone
derivatives is reported here. They were synthesized by
Claisen–Schmidt reaction and were characterized by FTIR,
NMR, and mass spectroscopy. Antioxidant activity is
evaluated through four different methods namely, super-
oxide radical-scavenging, hydrogen peroxide scavenging,
reducing power, and DPPH radical-scavenging assays.
Generally, compounds with –SCH3 and –OCH3 in the para
position of the A-ring and –OH in the B-ring were more
active than others. In few cases some of the compounds
were more active than ascorbic acid or butylated
hydroxytoluene. QSAR was developed correlating the
antioxidant activity with the structural features of the
compounds and the predictive capability of the models was
estimated using internal and external validation methods.
All the predictions were within the 99% confidence level.
Spatial, structural, and lipophilic properties of the com-
pounds determine their antioxidant properties.
Introduction
Superoxide (O•2-), peroxyl (ROO•), alkoxyl (RO•),
hydroxyl (HO•), and nitric oxide (NO•) are the oxygen
centered free radicals and they are also called as reactive
oxygen species (ROS). They are generated in the human
body and they lead to a wide range of diseases (Gutteridge,
1993). The half lives of hydroxyl and alkoxyl radicals are
between nanoseconds and seconds. Nitric oxide, superox-
ide anion, and lipid hydroperoxides show lesser reactivity
when compared to hydroxyl and alkoxyl radicals (Ames
et al., 1993). ROS are beneficial to the human body
(involved in energy production, immune system, cell dif-
ferentiation, signal transduction, and biosynthesis of
important compounds) (Droge, 2002) but they also induce
membrane damage, DNA base oxidation, DNA strand
breakage, chromosomal aberrations, and protein altera-
tions. Some ROS attack unsaturated fatty acids (lipid per-
oxidation) and cause oxidative damage to the cell
membrane. Therefore, oxidative stress is thought to be
closely associated with aging, atherosclerosis, and carci-
nogenesis (the last arising due to irreversible DNA damage,
i.e., mutations) (Olinski et al., 2002). Living cells normally
protect themselves from oxidative damage through enzy-
matic and non-enzymatic defense mechanisms. The former
system includes antioxidant enzymes like superoxide dis-
mutase, catalase, and peroxidases. These enzymes metab-
olize superoxide, hydrogen peroxide, and lipid peroxides
and thus prevent the formation of hydroxyl radical (Fujii
and Taniguchi, 1999). Non-enzymatic defense includes
proteins and small molecules such as glutathione, histi-
dine–peptides, the iron-binding proteins transferrin and
ferritin, dihydrolipoic acid, reduced CoQ10, melatonin,
urate, and plasma protein thiols as radical trapping agents.
In spite of the presence of antioxidant defense systems in
Keywords Chalcones ꢀ EPR ꢀ Antioxidant activity ꢀ
Quantitative structure–activity relationship
Electronic supplementary material The online version of this
article (doi:10.1007/s00044-010-9342-1) contains supplementary
material, which is available to authorized users.
P. M. Sivakumar ꢀ P. K. Prabhakar ꢀ M. Doble (&)
Department of Biotechnology, Indian Institute of Technology
Madras, Chennai 600036, India
e-mail: mukeshd@iitm.ac.in
P. M. Sivakumar
e-mail: sivamedchem@yahoo.co.in
123

Med Chem Res (2011) 20:482–492
483
O
Fig. 1 Natural antioxidants
which contain chalcone
backbone
OH
R1
R4
O
R5
R8
CH3
CH3
R2
R3
R6
R7
OH
HO
C C
O
CH3
O
H2
H
CH2
HC
OH
HO
H3C
CH3
OH
Homobutein
Isoprenylchalcone
Structure of the Chalcone
O
OH
O
OH
OH
O
HO
HO
HO
HO
OH
OH
OH
HO
Isoliquiretigenin
Butein
Naringeninchalcone
the body some of the ROS escape and cause damage. Body
antioxidant systems are also supplemented with repair
antioxidants, which combat the damage with the aid of
some of the proteases, lipases, transferases, and DNA
repairing enzymes (Varma et al., 1995). The formation of
ROS and the antioxidant defense systems is always bal-
anced in the body. Imbalance between these two will result
in damage of the cells and the outcome is known on
oxidative stress. Oxidative stress is involved in various
diseases including cancer, ischemia–reperfusion injury,
atherosclerosis, cardiovascular diseases, inflammation, and
neurodegenerative disorders (Lamson and Brignall, 1999;
Flamm et al., 1978; Green and Shuaib, 2006; Vivekananthan
et al., 2003; Ziakas et al., 2006; Mattson, 2003;
Stavrovskaya and Kristal 2005). Thus, antioxidants have
wide variety of therapeutic applications. This encourages
researchers to search for newer antioxidants from both
natural and synthetic sources. Chalcones are obtained as
natural compounds and are largely distributed in plants,
fruits, and vegetables. They belong to the flavonoid group
of molecules. A wide ranging biological activity have been
observed under in vitro conditions for these compounds
over the past 20 years, including anti-inflammatory, anti-
pyretic, and analgesic (Satyanarayana and Rao, 1993),
anticancer (Shibata, 1994), and cytotoxic (Dhar, 1981).
Apart from these, chalcones have also been used in the
treatment of atherosclerosis (Vaya et al., 1997). They also
prevent LDL oxidation and Alzheimer’s disease, a neuro-
degenerative disorder.1 Our laboratory studies have also
revealed that chalcones act as antitubercular agents against
the resistant Mycobacterium tuberculosis H37Rv (Sivaku-
mar et al., 2007a, b). Natural chalcones which have
exhibited antioxidant activity includes isoliquiretigenin,
isoprenyl chalcone (Vaya et al., 1997), butein, homobutein,
and naringenin chalcone (Kozlowski et al., 2007). Figure 1
shows the natural antioxidants which posses the chalcone
backbone. Some of the synthetic chalcones which have
hydroxyl and methoxy substitutions also are found to
posses antioxidant activity (Yayli et al., 2005). Previous
researchers also found that o-, m-, and p-nitro substituted
azachalcones are also potential antioxidant agents (Yayli
et al., 2006). Thiol substitution is also common in some of
the antioxidant compounds where the thiol is converted to
disulfides during the oxidation (Ou et al., 2002).
In this study, 25 derivatives of chalcones with methoxy,
thiomethyl, dimethylamino substitutions in the A-ring were
synthesized and four different antioxidant activities namely
superoxide radical-scavenging, hydrogen peroxide scav-
enging, reducing power, and DPPH radical-scavenging at
50 lg/ml were experimentally determined under in vitro
condition. Antioxidants either quench the radical, donate a
hydride or an electron. These various actions can be
determined by the four different assays proposed here.
Quantitative structure–activity relation (QSAR) correlating
the activities with the structural features or descriptors of
the compounds were also developed to understand the
importance of certain structural features on their antioxi-
dant activities. Some of the earlier researcher too explored
the quantitative relationship (QSAR) between the antioxi-
dant activities and the physiochemical properties (Farkas
et al., 2004; Durand et al., 2007).
Experimental part
Materials and methods
Ferric chloride, nicotinamide adenine dinucleotide (NADH),
and nitroblue tetrazolium (NBT) were purchased from SRL
Mumbai, India. All the acetophenones, benzaldehydes, and
1,1-diphenyl-2-picryl-hydrazyl (DPPH) were purchased
from Sigma Bangalore, India. Butylated hydroxytoluene
(BHT), potassium ferricyanide [K3Fe(CN)6], and trichloro-
acetic acid (TCA) were purchased from Merck, India, and
ascorbic acid, sodium phosphates, and methionine were
1
http://www.supplementquality.com/efficacy/ashitaba.html (21.02.08).
123

484
Med Chem Res (2011) 20:482–492
H2O2 scavenging assay
purchased from Himedia Laboratories, India. All the other
chemicals used were analytical grade.
The assay to determine the ability of the compounds for
scavenging hydrogen peroxide (H2O2) was reported by
Ruch et al. (1989). Hydrogen peroxide solution (30 mM)
was prepared in a phosphate buffer (200 mM, at a pH of
7.4). To 400 ll of chalcone, 0.3 ml of H2O2 was added and
after 10 min, the absorbance was measured spectrophoto-
metrically (Jasco V550) at 238 nm using a quartz cell
(1 cm). Phosphate buffer without H2O2 was used as the
control. The hydrogen peroxide scavenging activity was
determined as follows:
Synthesis of the compounds
Scheme 1 shows the general procedure for the synthesis of
the chalcones. The chalcones (1–25) with various substi-
tutions were synthesized by the method described by Lin
et al. (1977). 10 mM of benzaldehyde and 10 mM of
acetophenone derivatives were mixed together and stirred
in methanol (10 ml) at room temperature. 2 M of NaOH
solution was added to the above mixture which was
maintained under rapid stirring. In majority of the cases an
almost pale yellowish solid precipitated out which was
washed with 50% alcohol, followed by water and then
dried. The yield in most of the cases was more than 80%.
The synthesized chalcones were characterized by NMR,
mass spectrometry and FTIR spectroscopy.
ðAC ꢁ ASÞ
% H2O2 scavenging activity ¼
ꢂ 100
AC
where AC = absorbance of control, AS = absorbance of
test sample.
Reducing power assay
In vitro evaluation of antioxidant activity
The reducing power assay was performed according to the
method described by Oyaizu (1986). 50 lg/ml concentra-
tion of chalcone was prepared in ethanol. 2.5 ml of sodium
phosphate buffer (0.2 M, at a pH of 6.5) and 2.5 ml of 1%
potassium ferricyanide [K3Fe(CN)6] were added to the
solution. The mixture was incubated for 20 min at 50ꢁC
and 2.5 ml of 10% trichloroacetic acid (w/v) was added to
this and centrifuged (Eppendorf Centrifuge 5810 R) at
3000 rpm for 10 min. To the supernatant (2.5 ml), 2.5 ml
of deionised water and 0.5 ml of 0.1% ferric chloride were
added, and the absorbance was measured in a spectropho-
tometer at 700 nm. The increase in the absorbance of the
reaction mixture is directly proportional to the reducing
power of the compound and the activity is determined as
follows:
The methodologies for the evaluation of the four different
antioxidant activities are explained below. All the experi-
ments were performed in triplicate. Ascorbic acid and BHT
were used as standards for all the studies.
Superoxide radical-scavenging assay
Superoxide radical was estimated based on the method
reported by Martinez et al. (2001) with few modifications.
The assay was based on the capacity of the extracts to
inhibit the reduction of NBT in the riboflavin–light–NBT
system (Beauchamp and Fridovich, 1971). 3 ml of reaction
mixture is prepared containing 50 mM of sodium phos-
phate buffer (at a pH of 7.8), 13 mM methionine, 2 lM
riboflavin, 100 lM EDTA, NBT (75 lM), and 1 ml test
sample (ethanolic chalcone, 50 lg/ml) solution. The
increase in absorbance at 560 nm, after 10 min of illumi-
nation from a fluorescent lamp, due to the formation of
blue formazan was measured and the activity was deter-
mined as follows:
ðAC ꢁ ASÞ
% Reducing power activity ¼
ꢂ 100
AC
where AC = absorbance of control, AS = absorbance of
test sample.
% Superoxide scavenging radical activity
ðAC ꢁ ASÞ
DPPH radical-scavenging assay
¼
ꢂ 100
AC
The free radical-scavenging activity of the chalcones was
as per the methods suggested by Epsin et al. (2000) and Hu
et al. (2007). An aliquot of the chalcone (50 lg/ml) in
ethanol was mixed with 1000 ll of 0.1 mM DPPH (pre-
pared with ethanol) and the mixture was vortexed for
5 min. Then it was incubated in dark for 20 min and then
the absorbance was measured in a spectrophotometer at
517 nm. The percentage reduction in the concentration of
DPPH• was calculated using the following equation:
where AC = absorbance of control, AS = absorbance of
test sample.
O
COCH3
CHO
2MNaOH
MeOH
3 hrs
Scheme 1 Synthesis procedure
123

Med Chem Res (2011) 20:482–492
485
methods include multiple linear regression (MLR), partial
cluster analysis (PCA), partial least square (PLS). Genetic
algorithm (GA) based QSAR uses population of random
models and tests the final and fully constructed models
(Caballero et al. 2007). Validation is an important step
after developing the QSAR. There are two types of vali-
dation methods namely, external and internal. External
validation method is a Leave More Out (LMO) method
which divides the data into training and test sets was used
in our studies. The 25 compounds were divided into
training and tests sets, the former comprising of 21 com-
pounds, and the latter comprising of four compounds. The
models were developed using training set and it was used
to predict the activity of test set within the 99% prediction
limits. Several statistical parameters such r2—squared
correlation coefficient, ra2dj—adjusted squared correlation
coefficient, q2—cross-validated squared correlation coeffi-
cient were obtained based on Leave One Out (LOO)
method for the training set compounds. LOO method is
known as an internal validation method. F-ratio is deter-
mined to estimate the quality of the QSARs. These statis-
tics are calculated by the following equations.
ðAC ꢁ ASÞ
% Reduction in DPPH ¼
ꢂ 100
AC
where AC = absorbance of control, AS = absorbance of
test sample.
DPPH scavenging activity assessment by EPR method
DPPH scavenging activity was performed based on literature
report with minor modification (Yu et al., 2002). The active
chalcone was studied at two different concentrations (namely
at 100 and 250 lg/ml) in ethanol. The compound is mixed
with 0.1 mM DPPH and the reaction with the free radical is
initiated. 0.1 mM DPPH with out the chalcone was taken as
control. EPR spectroscopy was recorded using Varian
E-112X-Q band EPR spectrometer at microwave frequency
of 9.54 GHz, microwave power of 10 mW, modulator
amplitude of 5 G, receiver gain of 10,000 and time of 0.5 s.
QSAR development
QSAR relating the activity with the structural features or
descriptors of the compounds were developed using Ceri-
us2 software (Acceryls, USA). Statistical regression anal-
ysis was carried out with Kyplotꢂ (USA) software. The
structures of the chalcones were sketched and their ener-
gies were minimized using consistent valence force field
(CVFF) available in Cerius2 software. This force field is
ideally suited for modeling small molecules and it is also
called a universal force field (Hobza et al., 1997). Mini-
mum energy conformations of several structures were
determined and the conformation which had the lowest
energy was selected. Two forty-nine different classes of
descriptors namely, topological, thermodynamic, spatial,
electronic, Kier & Hall molecular connectivity, and sub
graph count indexes were calculated for all the minimized
structures. Several papers deal with the explanation of
these structural descriptors (Todeschini et al., 1994;
Todeschini and Consonni, 2000; Karelson, 2000). The
percentage (P) antioxidant activities experimentally esti-
mated as described before for all these 25 compounds were
normalized (for developing the QSAR) as log (P/(100 -
P). QSAR is nothing but a regression relation where
activity is the dependant variable and the descriptors are
the independent variables. Genetic function approximation
(GFA) technique available in Cerius2 was used to identify
the best set of descriptors for the QSARs. GFA generates
multiple numbers of random models when compared to
other standard regression analysis. The importance of GFA
in developing efficient QSAR models was explained by
Rogers and Hopfinger (1994). GFA technique builds
models by selecting the descriptors with more physical
significance and hence offers more insight. The other
r2 = 1 - SSE/TSS where SSE (sum of squares of the
error) is equal to R (ydata - ymodel)2, TSS (total sum of
P
2
squares) is equal to
ðydata ꢁ y~Þ and y~ is the average of
all y data. ydata and ymodel are the observed and model
predicted activity values, respectively. The adjusted r2 will
usually be less than or equal to r2 and it is calculated as
ra2dj ¼ 1 ꢁ ðn ꢁ 1=n ꢁ pÞð1 ꢁ r2Þ where n = total number
of observations and p = total number of terms in the model
(including the constant). Adjusted r2 is an indication of
over fitting of the data and as p increases r2 will increase
but ra2dj will decrease. Cross-validated r2 (q2) is an internal
validation method and is equal to 1 - PRESS/TSS, where
PRESS is the predicted sum of squares of errors based on
leave one data point out method. q2 is an internal validation
method and it is also a measure of the validation power of
QSARs, similar to the method LMO (Kiemele et al. 1997).
QSAR is considered to be good if r2 and ra2dj [ 0:6 and
q2 [ 0.5. A large F value indicates that the developed
regression equation is statistically significant and it is not a
chance occurrence.
Results and discussion
Twenty-five compounds were obtained with a yield greater
than 90%. They were characterized with NMR, mass
spectrophotometer, and FTIR. The spectral details of
compounds 5 and 24 are listed here.2 The antioxidant
2
Selected spectral data for C24: IR (KBr) 1652 cm-1(Carbonyl), 1H
NMR (400 MHz, CDCl3): d 3.84 (s, 3H), d 6.90–6.92 (m, 2H), d 7.14
123

486
Med Chem Res (2011) 20:482–492
activity of a compound is due to several factors and they
are (i) suppress the formation of ROS by inhibiting the
enzymes involved in its generation, (ii) chelate the metal
ion responsible for the production of the free radical, (iii)
scavenge the ROS, and (iv) up regulate or protect the
antioxidant defense mechanism (Pietta, 2000). Hence the
antioxidant potential of any compound is related to its
(i) hydrogen or electron donation capacity, (ii) its ability to
stabilize and delocalize the unpaired electron, and (iii)
potential to chelate transition metal ions (Rice-Evans et al.,
1997). The above three actions are achieved either by the
hydrogen atom or single electron transfer. In the case of
ferric reducing antioxidant power (FRAP), it is due to the
single electron transfer and in the cases of superoxide
radical-scavenging, hydrogen peroxide scavenging, and
DPPH radical-scavenging activities, it is due to the transfer
of hydrogen atom (Ou et al., 2002). Researchers have
reported that the antioxidant activity of the flavonoids is
due to the inhibition of the enzyme responsible for the
superoxide radical production, chelation of the metal ions
and scavenging of ROS (Pietta, 2000). Table 1 shows the
structure, observed scavenging activities, and reducing
power of the 25 chalcones.
positively correlated to the electron delocalization capacity
(Sr) (r = 0.56). The active compounds have higher and
less active compounds have lower Sr (the values are given
in Supplementary material). Superdelocalizibility is also
electron releasing index and other researchers too find a
positive correlation between this descriptor against anti-
oxidant activity (Yamagami et al., 2005).
The QSAR for superoxide radical-scavenging activity is
given by the following relation.
Activity = 1.0898–7.6121*Jurs-FNSA-3 ? 0.4929*
ADME_solubility.
(r2 = 0.62, ra2dj ¼ 0:58, q2 = 0.52, F-ratio = 14.92, std
err = 0.20).
The ADME_solubility is positively correlated with
activity and it is directly related to the water solubility of
the compound (Bednarczyk et al. 2003). Earlier researchers
also found that increase in water solubility of fullerenes
results in the increase in their superoxide radical-scav-
enging activity under in vitro conditions. They also
observed that the antioxidant activity is dependent on other
structural features such as charge, shape, and size, as well
as its lipophilicity and redox behavior (Witte et al., 2007).
Jurs-FNSA-3 is known as fractional negatively charged
partial surface area and it has a negative correlation with
activity. Figure 2 shows the parity plot between the
observed and predicted (based on QSAR) superoxide rad-
ical-scavenging activity for the training and test sets. The
data is listed in Table 2. The values of the descriptors are
tabulated in the Supplementary material. All the prediction
is within the 99% confidence limits.
Superoxide radical-scavenging activity
None of the chalcones tested here are more active than
ascorbic acid with respect to superoxide and hydrogen
peroxide reducing activities where as many chalcones
exhibit higher reducing power and DPPH radical-scav-
enging activity. It is mentioned that flavonoids inhibit the
two enzymes namely, xanthine oxidase and protein kinase
C, which in turn inhibits the production of superoxide
anion. More over, the a,b-unsaturated carbonyl present in
chalcone improves the electron delocalization (Pietta,
2000). In addition, hydroxyl and methoxyl substitutions in
A or B-rings are also important for antioxidant activity
(Cioffi et al., 2003). Compounds 24, 5, 7, 16, and 19 show
high superoxide radical-scavenging activity ([50%) and
majority of these compounds essentially posses nucleo-
philic hydroxyl substitution in their B-ring. Compounds
with phenolic substitution can readily donate proton to
form phenoxy radical which are stable. Compounds 25, 22,
23, and 1 show low activity (\16%) and have electrophilic
substitution in their B-ring. These set of compounds are
Hydrogen peroxide scavenging activity
Compounds 10, 14, and 21 exhibit high activity ([45%).
Many compounds are more active than BHT but none are
more active than ascorbic acid. Compounds 2, 13, and 23
exhibit lowest activity (\5%). The active compounds have
electrophilic substitutions in B-ring and in the case of
inactive compounds either they have no substitutions (2
and 13) or methyl (23) in their ring. A negative correlation
is observed between log P values of these high and least
active compounds against the hydrogen peroxide scav-
enging activity (r = -0.47), which shows that increase in
lipophilicity of the compound will decrease its activity.
The QSAR for the hydrogen peroxide activity is given as
Activity = 1.6671 ? 0.0017*PMI-mag ? 0.4618*
ADME_solubility-3.3552*Shadow-XZfrac.
(r2 = 0.70, ra2dj ¼ 0:65, q2 = 0.55, F-ratio = 13.38, std
Footnote 2 continued
(ddd, 1H, J = 8.0 Hz, J0 = 2.4 Hz, J00 = 0.8 Hz), d 7.36 (dd, 1H,
J = 8.4 Hz, J0 = 7.6 Hz), d 7.38 (d, 1H, J = 15.6 Hz), d 7.54–7.58
(m, 3H), d 7.66 (dd, 1H, J = 2.4 Hz, J0 = 1.6 Hz), d 7.79 (d, 1H,
J = 16 Hz).13C NMR (100 MHz, CDCl3): d 191.00, 161.84, 156.56,
145.45, 139.73, 130.41, 129.83, 127.47, 120.79, 120.33, 119.58,
115.25, 114.45, 55.424. EIMS: [M]? 254.37.
err = 0.27).
Both ADME_solubility and PMI-mag show positive and
Shadow-XZfrac show negative correlation with activity.
As mentioned before ADME_solubility is directly related
to the water solubility and by increasing the water
123

Med Chem Res (2011) 20:482–492
487
Table 1 Structure and experimental antioxidant (all four) activities (in percentage) for 25 chalcones (at 50 lg/ml concentration)
R1
O
R5
R2
R3
R6
R7
B
A
R4
Compound A-ring
no.
R1
B-ring
Superoxide radical-
scavenging assay
Hydrogen peroxide
scavenging activity
Reducing DPPH radical-
power assay scavenging activity
R2
R3
R4
R5 R6 R7
1
NMe2
SMe
SMe
13.78
27.44
12.05
4.26
9.52
19.80
3.17
8.08
10.16
15.88
18.63
21.17
2.64
2
3
NO2 17.40
4.61
4
32.60
53.32
25.35
28.70
29.48
33.28
31.68
43.47
46.35
12.34
17.38
4.54
42.43
51.60
24.34
46.72
23.79
30.95
27.23
21.25
10.11
21.81
18.80
40.61
46.01
33.83
26.53
43.54
28.64
32.82
21.99
23.11
37.47
22.37
56.09
22.57
5
SMe
SMe
SMe
SMe
SMe
SMe
OMe
SMe
OH
6
OMe 17.78
50.23
7
OH
15.37
3.16
8
Me 23.08
9
Cl
41.41
39.69
21.29
30.86
18.59
18.02
28.86
23.22
15.29
5.34
10
11
12
NO2
Cl
Cl
Br
4.07
13
14
15
16
17
18
19
20
21
22
23
24
25
BHT
OMe
0.98
SMe
49.68
22.68
23.83
23.83
5.91
17.99
15.55
28.92
15.69
7.90
OMe OMe OMe
SMe
OH 52.58
OMe OMe Cl
OMe
Cl
32.58
23.07
53.44
38.99
OMe
OMe
NO2
NO2
NO2
OMe OMe
NMe2
35.85
17.45
61.30
32.59
4.54
15.42
18.64
7.00
NO2 34.50
Cl 10.87
NMe2
OMe
OMe
Cl
12.93
1.30
Me 15.41
61.97
OH
42.77
24.32
42.57
78.31
35.69
6.93
OMe
NO2
5.39
79.07
26.43
27.96
Ascorbic
acid
82.93
Note: All the activity values are expressed in percentage
solubility, the hydrogen peroxide scavenging activity of the
compounds can be enhanced. PMI-mag is a spatial
descriptor that is based on the principal moment of inertia
of a molecule. It is an indication of the shape, orientation,
and conformational rigidity of the molecule (Sivakumar
et al., 2007a, b). Shadow-XZfrac is the fraction of the area
of molecular shadow in the XZ plane to area of the
enclosing rectangle (Rohrbaugh and Jurs, 1987). This
descriptor, once again is a measure of the shape and size of
the molecule. Figure 3 shows the parity plot of observed
hydrogen peroxide scavenging activity and model
prediction for the test and training sets. All the predictions
are within the 99% confidence limits.
Reducing power activity
Compounds 5, 7, and 16 show highest reducing power
activity ([46%) and compounds 1, 3, and 12 show lowest
activity (\10%). Many compounds exhibit higher activity
than ascorbic acid and compound 5 is marginally less
active than BHT. All the active compounds have essen-
tially hydroxyl substitution in their B-ring. Phenolic
123

488
Med Chem Res (2011) 20:482–492
1.0
0.5
substitution. The electron delocalizabilty (Sr) is positively
correlated (r = 0.57) with the activity of these six
compounds.
The QSAR for reducing power is given by the following
relation.
0.0
Activity = -1.3499 ? 0.4579*H-bond
-0.5
-1.0
-1.5
-2.0
donor ? 0.0047*energy ? 0.0095*Shadow-XZ.
(r2 = 0.85, ra2dj ¼ 0:82, q2 = 0.76, F-ratio = 31.34, std
err = 0.10).
The mechanism behind the reduction of ferric to ferrous
ion and the antioxidant power is dependent totally on the
transfer of single electron from the latter to the former (Ou
et al., 2002). A hydroxyl substitution in the B-ring helps in
the release of electron from antioxidant chalcones. The
descriptor, H-bond donor, represents the aqueous solubility
and the hydrogen bond forming capability of the com-
pound. It is mentioned before that increase in aqueous
solubility will also result in improved antioxidant activity.
Durand et al. (2007) also found the contribution of func-
tional groups, such as like number of tertiary carbon atoms
-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2
Observed activity
0.0 0.2
0.4
Fig. 2 Parity plot between observed superoxide radical-scavenging
activity and model prediction based on QSAR (filled circle training
set, open circle test set) (dotted lines indicate 99% confidence limits)
compounds are very good H-bond donors (Shen et al.,
2007). In the case of low active compounds the B-ring has
either electron withdrawing groups or do not have any
Table 2 Observed and predicted four antioxidant activities of the 25 chalcone derivatives
Compound
no.
Superoxide radical-scavenging Hydrogen peroxide scavenging Reducing power assay
assay
DPPH radical-scavenging
activity
activity assay
Observed
activity
Predicted
activity
Observed
activity
Predicted
activity
Observed
activity
Predicted
activity
Observed
activity
Predicted
activity
1
-0.7962
-0.4223
-0.6765
-0.3154
0.0577
-0.6588
-0.5324
-0.3090
-0.4866
0.0231
-0.8634
-1.3519
-1.3156
-0.4690
-0.3952
-0.3788
-0.3020
-0.3339
-0.1142
-0.0635
-0.8515
-0.6769
-1.3230
-0.0055
-0.5328
-0.5047
-0.5047
-1.2020
-0.2527
-0.6749
0.1998
-1.2033
-1.0656
-0.2298
-1.7755
-0.3252
-0.2628
-0.2537
-1.0117
-0.6422
0.0563
-0.9777
-0.6075
-1.4853
-0.1326
0.0277
-0.5116
-0.5481
-0.4324
-0.7341
-0.1589
-0.4626
0.0027
-1.0563
-0.9464
-0.7240
-0.6402
-0.5709
-1.5670
-0.7409
-1.4870
-0.5193
-0.7434
-1.2490
-1.3725
-2.0027
-0.6589
-0.7348
-0.3906
-0.7303
-1.0666
-0.7390
-0.6400
-1.1231
-0.8281
-1.8796
-0.2558
-1.1279
-1.5851
-1.3231
-0.6389
-1.7621
-0.6377
-1.1085
-0.6096
-1.3459
-0.8469
-0.6441
-1.6119
-1.2447
-1.6730
-0.7606
-1.3970
-0.5838
-0.6217
-1.0789
-0.9077
-0.9573
-0.9639
-0.8206
-1.5440
-0.1279
-0.9457
2
3
4
5
6
-0.6652
0.0040
-0.4800
-0.0585
-0.7987
-0.4234
-0.3291
-0.3975
-0.4862
-0.4194
-0.5203
-0.4878
0.0523
-0.4927
-0.0570
-0.5056
-0.3485
-0.4269
-0.5690
-0.9491
-0.5546
-0.6355
-0.1650
-0.0694
-0.2914
-0.4423
-0.1129
-0.3965
-0.3110
-0.5499
-0.5222
-0.2224
-0.5404
7
8
-0.5229
-0.1507
-0.1817
-0.5678
-0.3503
-0.6415
-0.6581
-0.3918
0.0448
-0.5442
-0.5384
-0.4914
-0.5956
-0.7555
-0.5555
-0.5290
-0.1367
-0.0171
-0.3328
-0.3332
-0.1589
-0.3719
-0.3037
-0.6407
-0.5237
-0.1393
-0.5099
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
-1.2427
-0.7361
-1.2752
0.0801
-0.2818
-0.5372
-0.4032
-0.8437
-0.4347
-0.5291
-0.2646
-0.4622
-1.0632
-0.4395
-0.4432
-0.3159
-0.5231
0.0598
-0.3600
-0.1414
-0.2487
-0.4697
-0.1924
-0.5680
-0.6498
0.1730
-0.1944
-0.2784
-0.9138
-0.7394
0.2120
-0.3156
-1.3230
-0.1265
-0.4930
-1.2445
-0.1735
123

Med Chem Res (2011) 20:482–492
489
8
6
3
2
4
1
2
0
0
-1
-2
-3
-4
-2
-4
-6
-8
-10
-5
-1.6
-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2
Observed activity
0.0
0.2
-1.5
-1.0
-0.5
0.0
Observed activity
Fig. 4 Parity plot between observed reducing power and model
prediction based on QSAR (filled circle training set, open circle test
set) (dotted lines indicate 99% confidence limits)
Fig. 3 Parity plot between observed hydrogen peroxide scavenging
activity and model prediction based on QSAR (filled circle training
set, open circle test set) (dotted lines indicate 99% confidence limits)
Activity = -4.2001 ? 0.6162*H-bond
and number of aliphatic ketones towards antioxidant
activity.
donor ? 0.0147*Jurs-WNSA-1
?0.6302*ADME_Solubilty_level.
(r2 = 0.68, ra2dj ¼ 0:62, q2 = 0.53, F-ratio = 12.11, std
The energy descriptor represents the conformational
energy of the molecule. Energy shows a positive contri-
bution with the activity. Lower energy conformer is ther-
modynamically stable and hence exhibits lower activity
and vice versa. During electron transfer, the antioxidant
will liberate energy (Cartling and Ehrenberg, 1978). So if
the conformer has high energy, it can easily transfer elec-
tron from the antioxidant molecule to ferric and hence it
has a positive contribution with activity.
err = 0.29).
All the three descriptors show positive correlation with
the activity. Since the DPPH scavenging activity is directly
proportional to the hydrogen atom transfer (hydrogen
donating capacity), the descriptor, hydrogen bond donor,
shows a positive contribution. This observation is proved by
Foti (2007) by using phenolic compounds as antioxidants.
They found that hydrogen bond accepting solvent systems
showed reduced antioxidant activity and hydroxyl solvents
(alcohol) doubled the antioxidant capacity (Foti, 2007).
Jurs descriptor is a function of both the shape and
electronic characteristic of the compound. It provides
information about the atomic partial charges towards sol-
vent accessible surface areas. Jurs-WNSA-1 is one among
the six of the descriptors (WPSA-1, WPSA-2, WPSA-3,
WNSA-1, WNSA-2, and WNSA-3) that provides infor-
mation about surface weighted charged partial surface area
(Stanton and Jurs, 1990).
Shadow-XZ is the area of the molecular shadow in the
XZ plane (Rohrbaugh and Jurs, 1987). This descriptor
provides the information about the shape of the molecule.
Earlier research also has identified that the antioxidant
activity depends on shape of the molecule (Witte et al.,
2007). Figure 4 shows the parity plot of observed reducing
power and model prediction for the test and training data
sets.
DPPH scavenging activity
ADME_solubility_level is an indication of the water
solubility of the molecule and, this in turn is a function of
the number of H-bond donor and acceptors, dipole
moment, number of rotatable bonds, and surface area
(Sivakumar et al., 2007a, b). Molecules with hydrophilic
substitutions will perform better than compound with
hydrophobic substitutions. Figure 5 shows the parity plot
between the observed and predicted DPPH radical-scav-
enging activity. Except for one data, all the predictions are
within 99% prediction limits. Table 3 lists the antioxidant
activities and the class of descriptors used in developing
In this assay, 24, 16, 5, and 9 are the most active com-
pounds ([20%). Among four of these compounds 24 and
16 were found to be more active than ascorbic acid and
BHT. Except for compound 9 the other three compounds
have hydroxyl substitution in the B-ring. The DPPH radical
abstracts one of the hydrogen atoms from the antioxidant
molecule and gets converted into the stable 1,1-diphenyl 2-
picrylhydrazine. Compounds 13, 23, 6, and 8 are the least
active compounds (\3%) and they have electrophilic sub-
stitution in their structure.
The QSAR for the DPPH activity is given below.
123

490
Med Chem Res (2011) 20:482–492
4
2
Compound 24 - 100 micro gm/ml + DPPH at 0.1mM
compound 24 - 250 micro gm/ml + DPPH at 0.1mM
DPPH at 0.1mM
0.3
0.2
0
0.1
-2
-4
-6
0.0
-0.1
-0.2
-0.3
-2.2 -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0
Observed activity
3300
3350
3400
Fig. 5 Parity plot between observed DPPH radical-scavenging
activity and model prediction based on QSAR filled circle training
set, open circle test set) (dotted lines indicate 99% confidence limits)
Wave number [cm-1]
Fig. 6 In vitro EPR spectrum recorded to find antioxidant and free
radical-scavenging activity of compound 24 with DPPH
these QSARs. It is observed that all the four QSARs con-
tain predominantly spatial, structural, electronic, and lipo-
philic descriptors.
BHT and ascorbic acid, in certain assays. The mode of
action of the antioxidant will be either due to the single
electron transfer or due to the hydrogen atom transfer.
Good correlation exists between these active compounds
with their superdelocalizability (Table S1 in Supplemen-
tary material). QSARs were developed with training data
set containing 21 compounds. The developed models were
tested with four compounds in test set. All the statistical
measures like r2, q2, ra2dj, F-ratio are found to be in the
acceptable range. These QSARs show the importance of
spatial, structural, electronic and solubility of the com-
pounds. Among these, spatial descriptors are the major
contributors. Earlier researchers also observed the rela-
tionship between the molecular shape, size, and lipophil-
icity with the antioxidant activity (Yamagami et al., 2005).
All the activity predictions were found to be within the
99% prediction limits. Most of the descriptors selected by
GFA for the QSARs were marginally correlated (r \ 0.38)
with each other (Tables S6–S9) in Supplementary material.
Since chalcone 24 is generally more active than any
other compound in this series, EPR spectroscopy with this
compound at two different concentrations, namely at 100
and 250 lg/ml, with 100 lM of DPPH was carried out.
The free radical concentration, (as measured by EPR)
decreases considerably when, compound 24 is added
(Fig. 6).
Conclusion
This paper discusses the synthesis of 25 chalcones and the
evaluation of their in vitro antioxidant activities. Com-
pounds with hydroxyl (phenolic) substitution in the B-ring
exhibited very good activity. Compounds 5, 7, 16, and 24
are amongst the most active compounds and they have
hydroxyl substitution in the B-ring. A few of the com-
pounds, exhibit better activity than the standard drugs,
Table 3 Descriptors used in QSARs for various activities
Type of antioxidant activity
Descriptors contributing to the QSAR equation and their classes
Superoxide radical-scavenging
activity
Jurs-FNSA-3 (spatial/
electronic)
ADME_solubility (ADME properties)
Hydrogen peroxide scavenging
activity
PMI-mag (spatial)
ADME_solubility (ADME
properties)
Shadow-XZfrac (spatial)
Shadow-XZ (spatial)
Reducing power assay
H-bond donor (electronic)
H-bond donor (electronic)
Energy (thermodynamic)
Jurs-WNSA-1 (spatial)
DPPH radical-scavenging activity
ADME_solubility (ADME
properties)
Class of descriptor is given in parenthesis
123

Med Chem Res (2011) 20:482–492
491
Acknowledgment Authors thank SAIF, IIT Madras for providing
spectral analysis.
Kiemele MJ, Schmidt SR, Berdine RJ (1997) Basic statistics tools for
continuous improvement, vol 7, 4th edn. Air academy press,
Colorado springs, Colorado, p 3
Kozlowski D, Trouillas P, Calliste C, Marsal P, Lazzaroni R, Duroux
JL (2007) Density functional theory study of the conformational,
electronic, and antioxidant properties of natural chalcones.
J Phys Chem A 111:1138–1145
References
Lamson DW, Brignall MS (1999) Antioxidants in cancer therapy;
their actions and interactions with oncologic therapies. Altern
Med Rev 4:304–329
Lin J, Rivett DE, Wilshire JFK (1977) The preparation and photo-
chemical properties of some 1,3-diphenyl-2-pyrazolines contain-
ing a heteroaromatic substituent. Aust J Chem 30:629–637
Martinez AC, Marcelo EL, Marco AO, Moacyr M (2001) Differential
responses of superoxide dismutase in freezing resistant Solanum
curtibolam and freezing sensitive Solanum tuberosum subjected
to oxidative and water stress. Plant Sci 160:505–515
Mattson MP (2003) Excitotoxic and excitoprotective mechanisms:
abundant targets for the prevention and treatment of neurode-
generative disorders. Neuromol Med 3:65–94
Olinski R, Gackowski D, Foksinski M, Rozalski R, Roszkowski K,
Jaruga P (2002) Oxidative DNA damage: assessment of the role
in carcinogenesis, atherosclerosis, and acquired immunodefi-
ciency syndrome. Free Radic Biol Med 33:192–200
Ames BN, Shigenaga MK, Hagen TM (1993) Oxidants, antioxidants,
and the degenerative diseases of aging. Proc Natl Acad Sci USA
90:7915–7922
Beauchamp C, Fridovich I (1971) Superoxide dismutase: improved
assays and an assay applicable to acrylamide gels. Anal Biochem
44:276–287
Bednarczyk D, Ekins S, Wikel JH, Wright SH (2003) Influence of
molecular structure on substrate binding to the human organic
cation transporter, hOCT1. Mol Pharmacol 63:489–498
Caballero J, Zampini FM, Collina S, Fernandez M (2007) Quantita-
tive structure–activity relationship modeling of growth hormone
secretagogues agonist activity of some tetrahydroisoquinoline
1-carboxamides. Chem Biol Drug Des 69:48–55
Cartling B, Ehrenberg A (1978) A molecular mechanism of the
energetic coupling of a sequence of electron transfer reactions to
endergonic reactions. Biophys J 23:451–461
Cioffi G, Escobar LM, Braca A, Tommasi ND (2003) Antioxidant
chalcone glycosides and flavanones from Maclura(Chloropho-
ra)tinctoria. J Nat Prod 66:1061–1064
Dhar DN (1981) The chemistry of chalcones and related compounds,
vol 213. Wiley, New York
Ou B, Huang D, Woodill MH, Flanagan JA, Deemer EK (2002)
Analysis of antioxidant activities of common vegetables
employing oxygen radical absorbance capacity (ORAC) and
ferric reducing antioxidant power (FRAP) assays: a comparative
study. J Agric Food Chem 50:3122–3128
Droge W (2002) Free radicals in the physiological control of cell
function. Physiol Rev 82:47–95
Oyaizu M (1986) Studies on product of browning reaction prepared
from glucose amine. Jpn J Nutr 44:307–315
Durand AC, Farce A, Carato P, Dilly S, Yous S, Berthelot P, Chavatte
P (2007) Quantitative structure–activity relationships studies of
antioxidant hexahydropyridoindoles and flavonoid derivatives.
J Enzyme Inhib Med Chem 22:556–562
Epsin JC, Soler-Rivas C, Wichers HJ (2000) Anthocyanin-based
natural colorants: a new source of antiradical activity for
foodstuff. J Agric Food Chem 48:648–656
Farkas O, Jakus J, Heberger K (2004) Quantitative structure–
antioxidant activity relationships of flavonoid compounds.
Molecules 9:1079–1088
Flamm ES, Demopoulos HB, Seligman ML, Poser RG, Ransohoff J
(1978) Free radicals in cerebral ischemia. Stroke 9:445–447
Foti MC (2007) Antioxidant properties of phenols. J Pharm Pharma-
col 59:1673–1685
Fujii J, Taniguchi N (1999) Down regulation of superoxide dismu-
tases and glutathione peroxidase by reactive oxygen and nitrogen
species. Free Radic Res 31:301–308
Green AR, Shuaib A (2006) Therapeutic strategies for the treatment
of stroke. Drug Discov Today 11:681–693
Gutteridge JM (1993) Free radicals in disease processes: a compi-
lation of cause and consequence. Free Radic Res Commun
19:141–158
Hobza P, Kabelac M, Sponer J, Mejzlik P, Vondrasek J (1997)
Performance of empirical potentials (AMBER, CFF95, CVFF,
CHARMM, OPLS, POLTEV), semiempirical quantum chemical
methods (AM1, MNDO/M, PM3) and ab initio Hartree-Fock
method for interaction of DNA bases. Comparison with nonem-
Pietta PG (2000) Flavonoids as antioxidants.
63:1035–1043
Rice-Evans CA, Miller NJ, Paganga G (1997) Structure–antioxidant
activity relationships of flavonoids and phenolic acids. Trends
Plant Sci 2:152–159
Rogers D, Hopfinger AJ (1994) Application of genetic function
approximation to quantitative structure–activity relationships
and quantitative structure–property relationships. J Chem Inf
Comput Sci 34:854–866
Rohrbaugh RH, Jurs PC (1987) Molecular shape and the prediction of
high-performance liquid chromatographic retention indexes of
polycyclic aromatic hydrocarbons. Anal Chim Acta 199:99–109
Ruch JR, Crist KA, Klaunig JE (1989) Effects of culture duration on
hydrogen peroxide-induced hepatocyte toxicity. Toxicol Appl
Pharmacol l00:45l–464
Satyanarayana K, Rao MNA (1993) Antiinflamatory, analgesic, and
antipyretic activities of 3-[4-[3-(4-dimethylaminophenyl)-1-
oxo]-2-propenyl]sydnone. Indian Drugs 30:313–318
Shen L, Ji HF, Zhang HY (2007) How to understand the dichotomy of
antioxidants. Biochem Biophys Res Commun 362:543–545
Shibata S (1994) Anti-tumorigenic chalcones. Stem Cells 12:44–52
Sivakumar PM, Geetha Babu SK, Doble M (2007a) QSAR studies on
chalcones and flavonoids as anti-tuberculosis agents using
genetic function approximation (GFA) method. Chem Pharm
Bull 55:44–50
J Nat Prod
Sivakumar PM, Seenivasan SP, Kumar V, Doble
M (2007b)
Synthesis, antimycobacterial activity evaluation, and QSAR
studies of chalcone derivatives. Bioorg Med Chem Lett
17:1695–1700
pirical beyond-Hartree-Fock results.
18:1136–1150
J
Comput Chem
Hu CC, Lin JT, Lu FJ, Chou FP, Yang DJ (2007) Determination of
carotenoids in Dunaliella salina cultivated in Taiwan and
antioxidant capacity of the algal carotenoid extract. Food Chem
109:439–446
Stanton DT, Jurs PC (1990) Development and use of charged partial
surface area structural descriptors in computer-assisted quanti-
tative structure–property relationship studies. Anal Chem
62:2323–2329
Karelson M (2000) Molecular Descriptors in QSAR/QSPR. Wiley-
Interscience, New York
Stavrovskaya IG, Kristal BS (2005) The powerhouse takes control of
the cell: is the mitochondrial permeability transition a viable
123

492
Med Chem Res (2011) 20:482–492
therapeutic target against neuronal dysfunction and death? Free
Radic Biol Med 38:687–697
Todeschini R, Consonni V (2000) Handbook of molecular descrip-
tors. Wiley-VCH, Weinheim
Todeschini R, Lasagni M, Marengo E (1994) New molecular
descriptors for 2D- and 3D-structures. J Chemom 8:263–273
Varma SD, Devamanoharan PS, Morris SM (1995) Prevention of
cataracts by nutritional and metabolic antioxidants. Crit Rev
Food Sci Nutr 35:111–129
Yamagami C, Akamatsu M, Motohashi N, Hamada S, Tanahashi T
(2005) Quantitative structure activity relationship studies for
antioxidanthydroxybenzalacetones by quantum chemicaland3-
D-QSAR (CoMFA) analyses. Bioorg Med Chem Lett 15:2845–
2850
Yayli N, Ucuncu O, Aydin E, Gok Y, Yasar A, Baltaci C, Yildirim N,
Kucuk M (2005) Intramolecular 4p photocyclization of chalco-
noid-like compounds in solution and antimicrobial activities.
J Photochem Photobiol A 169:229–234
Vaya J, Belinky PA, Aviram M (1997) Antioxidant constituents from
licorice roots: isolation, structure elucidation and antioxidative
capacity toward LDL oxidation. Free Radic Biol Med 23:302–
313
Vivekananthan DP, Penn MS, Sapp SK, Hsu A, Topol EJ (2003) Use
of antioxidant vitamins for the prevention of cardiovascular
disease: meta-analysis of randomised trials. Lancet 361:2017–
2023
Yayli N, Ucuncu O, Yasar A, Kucuk M, Yayli N, Akyuz E, Karaoglu
SA (2006) Synthesis and biological activities of N-alkyl
derivatives of o-, m-, and p-nitro (E)-4-azachalcones and
stereoselective photochemistry in solution, with theoretical
calculations. Turk J Chem 30:505–514
Yu L, Haley S, Perret J, Harris M, Wilson J, Qian M (2002) Free
radical scavenging properties of wheat extracts. J Agric Food
Chem 50:1619–1624
Witte P, Beuerle F, Hartnagel U, Lebovitz R, Savouchkina A, Sali S,
Guldi D, Chronakis N, Hirsch A (2007) Water solubility,
antioxidant activity and cytochrome C binding of four families
of exohedral adducts of C60 and C70. Org Biomol Chem
5:3599–3613
Ziakas GN, Rekka EA, Gavalas AM, Eleftheriou PT, Kourounakis PN
(2006) New analogues of butylated hydroxytoluene as anti-
inflammatory and antioxidant agents. Bioorg Med Chem
14:5616–5624
123