Design, synthesis, biological evaluation and molecular docking of curcumin analogues as antioxidant, cyclooxygenase inhibitory and anti-inflammatory agents

Article abstract of DOI:10.1016/j.bmcl.2005.02.039

Curcuminoids were isolated from Curcuma longa and their pyrazole and isoxazole analogues were synthesized and evaluated for antioxidant, COX-1/COX-2 inhibitory and anti-inflammatory activities. The designed analogues significantly enhance COX-2/COX-1 selectivity and possess significant anti-inflammatory activity in carrageenan induced rat paw edema assay. Pyrazole, isoxazole analogues of curcumin (4 and 7) exhibited higher antioxidant activity than trolox. Molecular docking study revealed the binding orientations of curcumin analogues in the active sites of COX and thereby helps to design novel potent inhibitors.

Full text of DOI:10.1016/j.bmcl.2005.02.039

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1793–1797
Design, synthesis, biological evaluation and molecular docking
of curcumin analogues as antioxidant, cyclooxygenase
inhibitory and anti-inflammatory agents
C. Selvam,a,ꢀ Sanjay M. Jachak,a,* Ramasamy Thilagavathib and Asit. K. Chakrabortib
aDepartment of Natural Products, National Institute of Pharmaceutical Education and Research (NIPER), Sector 67,
SAS Nagar, Punjab 160 062, India
bDepartment of Medicinal Chemistry, National Institute of Pharmaceutical Education and Research (NIPER), Sector 67, SAS Nagar,
Punjab 160 062, India
Received 22 December 2004; revised 14 February 2005; accepted 15 February 2005
Abstract—Curcuminoids were isolated from Curcuma longa and their pyrazole and isoxazole analogues were synthesized and eval-
uated for antioxidant, COX-1/COX-2 inhibitory and anti-inflammatory activities. The designed analogues significantly enhance
COX-2/COX-1 selectivity and possess significant anti-inflammatory activity in carrageenan induced rat paw edema assay. Pyrazole,
isoxazole analogues of curcumin (4 and 7) exhibited higher antioxidant activity than trolox. Molecular docking study revealed the
binding orientations of curcumin analogues in the active sites of COX and thereby helps to design novel potent inhibitors.
Ó 2005 Elsevier Ltd. All rights reserved.
Nonsteroidal anti-inflammatory drugs (NSAIDs) are of
huge therapeutic benefit in the treatment of rheumatoid
arthritis and various types of inflammatory conditions.
The target for these drugs is cyclooxygenase (COX), a
rate limiting enzyme involved in the conversion of ara-
chidonic acid into inflammatory prostaglandins. The
two isozymes of COX involved in prostaglandin biosyn-
thesis are COX-1 and COX-2. COX-1 is known as a
housekeeping enzyme and constitutively expressed in
all tissues, while COX-2 is constitutively expressed only
in kidney, brain and ovaries. COX-2 is increasingly ex-
pressed during inflammatory conditions by pro-inflam-
matory molecules such as IL-1, TNF-a, LPS and
agents such as carrageenan.1–4
The dichloromethane extract of C. longa exhibited sig-
nificant COX-1 inhibitory activity in COX catalyzed
prostaglandin biosynthesis assay in vitro. The CH2Cl2
extract was chemoprofiled and found to contain curc-
uminoids. These results prompted us to investigate
structure activity relationship (SAR) studies on curcu-
min analogues.
Curcumin,
1,7-bis-(4-hydroxy-3-methoxyphenyl)-1,6-
heptadiene-3,5-dione, is found as a major pigment in
the Indian spice turmeric (C. longa, Zingiberaceae).
The rhizome of the C. longa has been used in indigenous
medicine for the treatment of inflammatory disorders
and its medicinal activity has been known since ancient
times. Curcumin is reported to have anti-inflammatory,
antioxidant and anticancer properties.8,9 From the liter-
ature it was found that curcumin was investigated for
COX inhibitory activity using bovine seminal vesicles,
microsomes and cytosol from homogenates of mouse
epidermis showed IC50 value of 2 lM,10 52 lM11 and
5–10 lM,12 respectively. Moreover, one pyrazole ana-
logue of curcumin was synthesized and investigated for
lipoxygenase inhibitory activity13 and the three pyrazole
analogue of curcuminoids were synthesized and investi-
gated for endothelial cell proliferation and cytotoxic
activity.14,15
As a part of our continuing program to discover COX-1
and COX-2 inhibitory compounds from Indian medici-
nal plants,5–7 the rhizome of Curcuma longa was studied.
Keywords: Curcuminoids; Pyrazole analogues; Isoxazole analogues;
COX-1; COX-2; Anti-inflammatory activity; Antioxidant activity.
*
Corresponding author. Tel.: +91 172 2214682; fax: +91 172
2214692; e-mail: sanjayjachak@niper.ac.in
ꢀ Present address: Laboratoire de Chimie et Biochimie Pharmacolog-
´ ´
iques et Toxicologiques, UMR8601-CNRS, Universite Rene Des-
cartes-Paris V, 45 rue des Saints-Peres, 75270 Paris Cedex 06, France.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.02.039

1794
C. Selvam et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1793–1797
The curcuminoids such as curcumin (1), demethoxy cur-
cumin (2) and bisdemethoxycurcumin (3) were isolated
from the dichloromethane extract by silica gel column
chromatography (CH2Cl2–MeOH).16
matography to give 7 (450 mg) and 8 (50 mg). Com-
pound 7 was previously synthesized from vanillin and
3,5-dimethylisoxazole-MeI and not investigated for
any biological activity so far.17
The pyrazole analogues were prepared (Scheme 1) by
treating the CH2Cl2 extract (1 g) with hydrazine hydrate
(5 mL) in acetic acid (50 mL). The reaction mixture was
stirred at room temperature for 7 h. It was washed with
water, extracted with CH2Cl2 and evaporated to obtain
the crude residue. It was purified by column chromato-
graphy on silica gel to give pyrazole derivatives 4
(600 mg), 5 (95 mg) and 6 (90 mg).
It has been found that in inflammatory disorders, exces-
sive free radical generation takes place and several NSA-
IDs and phenolic compounds with anti-inflammatory
activity are reported to act as radical scavengers.18 The
o-methoxyphenolic moiety has been found to be an
essential structural feature for antioxidant property of
curcuminoids.19 Accordingly, antioxidant property of
curcuminoids, pyrazole and isoxazole analogues was
evaluated using DPPH radical scavenging in vitro
assay.20
The isoxazole analogues 7 and 8 were prepared by treat-
ing the CH2Cl2 extract with hydroxylamine hydrochlo-
ride in acetic acid at 85 °C for 6 h (Scheme 1). The
reaction mixture was evaporated to dryness, washed
with water and the residue was purified by column chro-
The radical scavenging activity results are shown in
Table 1. All the compounds demonstrated significant
antioxidant effect. Among them the pyrazole analogue
O
OH
N
NH
NH2-NH2. H2O
CH3COOH
HO
OH
HO
OH
R1
R2
R1
R2
1. R1 = R2 = OCH3
2. R1 = OCH3; R2 = H
3. R1 = R2 = H
4. R1 = R2 = OCH3
5. R1 = OCH3; R2 = H
6. R1 = R2 = H
N
O
NH2-OH. HCl
CH3COOH
HO
OH
R1
R2
7. R1 = R2 = OCH3
8. R1 = OCH3; R2 = H
Scheme 1.
Table 1. Anti-oxidant, COX-1, COX-2 inhibitory and anti-inflammatory activities of curcumin analogues
Compound
Anti-oxidant activitya
% Inhibition (100 lM)a Anti-inflammatory activityb
COX-2 % Inhibition at 75 mg/kg
35.0 3.6 61.4 2.7
IC50 (lM) ARP (1/ED50
)
Stoichiometric valuec H atoms per moleculed COX-1
1
11.06
21.36
36.37
9.70
5.4
2.8
1.7
6.2
3.9
1.9
5.6
3.1
4.5
—
0.37
0.71
1.21
0.32
0.52
1.08
0.36
0.63
0.44
—
2.7
1.4
0.8
3.1
1.9
0.9
2.8
1.6
2.2
—
80.5 2.1
67.8 2.8
41.7 2.9
87.0 2.6
73.9 3.0
44.4 1.5
80.8 3.6
47.4 3.7
—
2
31.6 0.4
33.2 0.8
—
—
3
4
61.0 2.0 68.8 2.9
5
15.45
32.41
10.71
18.96
13.1
43.8 2.9
35.8 2.5
—
—
6
7
8g
58.1 2.1 60.3 2.9
35.0 4.3
—
—
Trolox
Celecoxib
Ibuprofen
—
—
—
77.6 0.6
14.4e
2.05e
71.9 2.2f
54.5 1.4h
—
—
—
265.5e
a n = 3–4.
b n = 6 animals/group.
c Theoretical concentration of antioxidant to reduce 100% of the DPPH.
dMoles reduced DPPH per mole antioxidant.
e IC50 value.
f 50 mg/kg.
g New compound.22,23
h 100 mg/kg.

C. Selvam et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1793–1797
1795
(4) was found to be more potent than curcumin (1). But
the isoxazole analogue 7 was equipotent to curcumin.
tive sites of both COX-1 and COX-2 enzymes. For each
compound the most stable docking model was selected
according to the best scored conformation predicted
by the FlexX scoring function. The complexes were en-
ergy-minimized with a MMFF9427 force field till the
gradient convergence 0.05 kcal/mol was reached. The
distance dependent dielectric function (e = 4r) was used.
Compound 4 exhibited reduction of more than three
molecules of DPPH per molecule of 4, whereas curcumin
reduces less than three molecules of DPPH/molecule. O-
Methoxyphenolic compounds such as curcumin (1), pyr-
azole analogue of curcumin (4) and isoxazole analogue
of curcumin (7) exhibited potent antioxidant activity in
DPPH radical scavenging assay. It is well known that
the ortho-methoxy substitution enhanced the stability
of the phenoxy radical.19 Thus the compounds 1, 4 and
7 showed higher antioxidant activity than the reference
compound trolox, and the other curcumin analogues,
possessing two methoxy groups. Among these three com-
pounds, compound 4 demonstrated better scavenging
activity than the other two compounds; the reason may
be the presence of pyrazole NH in 4.
The three compounds could dock into the active site of
COX-1 successfully. The binding energies of À36.45,
À37.07 and À31.61 kcal/mol were obtained for 1, 4
and 7, respectively. The lower interaction energy ob-
served for 4 rationalizes the tighter binding of pyrazole
analogue (Fig. 2) into the COX-1 active site than that
of the other two compounds. The tight binding can be
explained in terms of extra hydrogen bonding with pyr-
azole NH and Tyr 355. All the three compounds were
involved in the hydrogen bonding with a residue Ser
530. The hydrogen bonding distance between one of
the methoxy group of curcumin with OH of Ser 530
COX-1/COX-2 inhibitory activity was evaluated employ-
ing the COX catalyzed prostaglandin biosynthesis assay
in vitro.5–7,21 All the compounds were investigated for
COX-1 and COX-2 inhibitory activity at 100 lM using
COX catalyzed prostaglandin biosynthesis assay (Table
1). The structure activity relationship studies (SAR) re-
vealed that curcumin to its pyrazole derivative (4) in-
creased the COX-1 activity slightly (80.5–87.0%
inhibition), whereas the COX-2 inhibitory activity in-
creased twofold (35–61.0% inhibition). Thus pyrazole
analogue (4) showed significant enhancement in the selec-
tivity towards COX-2 enzyme (COX-2/COX-1 = 0.70)
compared to curcumin (COX-2/COX-1 = 0.43). Com-
pound 5 exhibited higher COX-2 inhibitory activity
(43.8% 2.9%) than curcumin (35.0% 3.6%). Both
isoxazole analogue 7 and 1 were equipotent towards
COX-1 enzyme. For COX-2 enzyme isoxazole analogue
7 showed significantly increased COX-2 inhibitory activ-
ity as well as COX-2/COX-1 (0.72) ratio.
˚
˚
was found to be 3.703 A (OÁ Á ÁO) 2.791 A (OÁ Á ÁH).
One of the phenyl ring of curcumin was surrounded
by active site amino acid residues Tyr 385, Leu 384,
Phe 518, Met 522 and Ser 530. The heptanoid part
was surrounded by residues Ile 523, Ala 526, Gly 526,
Glu 524, Ser 353, Leu 359, Val 349 and Leu 352. The
second phenyl ring (Fig. 1) was surrounded by Tyr
355, His 90, Leu 357 and Arg 120 and Glu 524. A similar
trend was observed for 4 and 7 complexes.
The hydrogen bonding distance between Ser 530 and
˚
methoxy group of 4 was found to be 3.022 A (OÁ Á ÁO)
˚
2.132 A (OÁ Á ÁH). Another hydrogen bonding between
˚
OH of the second phenyl ring and Arg 83 (3.319 A,
˚
OÁ Á ÁN, 2.406 A, OÁ Á ÁH–N) was observed. NH of the
pyrazole was involved in hydrogen bonding interaction
˚
˚
with Tyr 355 (3.373 A, NÁ Á ÁO, 2.564 A, OÁ Á ÁH) (Fig.
2). The isoxazole analogue 7 orients in a similar fashion
to that of 1 and 4. However, only one hydrogen bond
was observed between the methoxy group and OH of
Both pyrazole 4 and isoxazole 7 analogues demon-
strated better COX-2 inhibitory activity in comparison
with curcumin. Since compounds 1, 4 and 7 exhibited
good COX inhibitory and antioxidant activities, they
were investigated for in vivo anti-inflammatory activity
using carrageenan induced rat paw edema assay at
75 mg/kg (Table 1). Among these three compounds pyr-
azole analogue of curcumin, 4 exhibited the highest
activity (68.8% inhibition).
˚
˚
Ser 530 (3.124 A, O–O, 2.295 A (OÁ Á ÁH) (Fig. 3).
FlexX could dock only curcumin to the active site
of COX-2. However, 4 and 7 were not docked into
the active site by this method. Therefore using
The level of COX-1/COX-2 inhibitory and anti-inflam-
matory activities of compounds 1, 4 and 7, prompted
us to perform molecular docking studies to understand
the ligand–protein interactions and COX-1/COX-2
selectivity in detail. All the calculations were performed
using SYBYL6.924 software installed on SGI octane 2
workstation. The crystal structures of COX-1 and
COX-2 enzymes complexed with indomethacin
[1PGGÆpdb, 4COXÆpdb]25 were used for the docking.
The active site of the enzyme was defined to include res-
˚
idues within a 6.5 A radius to any of the inhibitor atoms.
The FlexX program5,26 is an automated docking pro-
gram, was used to dock compounds 1, 4 and 7 on the ac-
Figure 1. Binding of curcumin into the active site of COX-1.

1796
C. Selvam et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1793–1797
pyrazole/isoxazole rings and Tyr 355 contributed to sta-
bilize the ligand–enzyme complexes. Molecular docking
studies further supported the strong inhibitory activity
of 4 compared to 1 and 7.
In conclusion, eight curcumin analogues including one
new compound were investigated for antioxidant and
COX-1/COX-2 inhibitory activities. SAR studies re-
vealed that replacement of the b-diketo fragment of cur-
cumin by a pyrazole ring significantly enhances COX-2/
COX-1 selectivity. The mixed functioning antioxidant
and COX inhibitory properties may provide superior
anti-inflammatory activity.28 Thus, curcumin analogues
were investigated for antioxidant activity and found to
be good radical scavengers. Molecular docking studies
further helps in understanding the various interactions
between the ligands and enzyme active sites in detail
and thereby helps to design novel potent inhibitors.
Figure 2. Binding of 4 into the active site of COX-1.
References and notes
1. Vane, J. R.; Botting, R. M. Inflamm. Res. 1998, 47, S78–
S87.
2. Ryn, J. V.; Trummlitz, G.; Pairet, M. Curr. Med. Chem.
2000, 7, 1145–1161.
3. Carter, J. S. Expert Opin. Ther. Pat. 2000, 10, 1011–
1020.
4. Beuck, M. Angew. Chem., Int. Ed. 1999, 38, 631–633.
5. Selvam, C.; Jachak, S. M.; Gnana Oli, R.; Thilagavathi,
R.; Chakraborti, A. K.; Bhutani, K. K. Tetrahedron Lett.
2004, 45, 4311–4314.
6. Selvam, C.; Jachak, S. M.; Bhutani, K. K. Phytother. Res.
2004, 18, 582–584.
Figure 3. Binding of 7 into the active site of COX-1.
7. Selvam, C.; Jachak, S. M. J. Ethnopharmacol. 2004, 95,
209–212.
curcumin-COX-2 complex, compounds 4 and 7 were
modeled in the active site of COX-2 enzyme. The car-
bonyl oxygen of curcumin was forming a hydrogen
8. Nadkarni, J. M. In Indian Materia Medica; Popular book
depot: Bombay, 1954; Vol. 1, p 414.
9. Lin, J. K.; Shiau, S. L. Proc. Natl. Sci. Counc. 2001, 25,
59–66.
10. Ammon, H. P. T.; Mack, S. H.; Sabieraj, J. J. Ethnophar-
macol. 1993, 38, 113–118.
11. Flynn, D. L.; Rafferty, M. F.; Boctor, A. M. Prosta.
Leukotr. Med. 1986, 22, 357–360.
˚
bonding interaction with Arg120 (3.562 A, OÁ Á ÁN,
˚
2.78 A, OÁ Á ÁH). Whereas, compounds 4 and 7 did not
produce any hydrogen bonding interactions with
COX-2 enzyme. However, favourable van der Waals
interactions between styryl carbon atoms and the hydro-
phobic residues such as Val 523, Val 116, between meth-
oxy groups of 4 (Fig. 4) and 7 and Ala 516 and between
12. Huang, M. T.; Lysz, T.; Ferraro, T.; Abidi, T. F.; Laskin,
J. D.; Conney, A. H. Cancer Res. 1991, 51, 3813–
3819.
13. Flynn, D. L.; Belliotti, T. R.; Boctor, A. M.; Connor, D.
T.; Kostlan, C. R.; Nies, D. E.; Ortwine, D. F.; Schirier,
D. J.; Sricar, J. C. J. Med. Chem. 1991, 34, 518–
525.
14. Ishida, J.; Ohtsu, H.; Tachiban, Y.; Nakanishi, Y.;
Bastow, K. F.; Nagai, M.; Wang, H. K.; Itokawa, H.;
Lee, K. H. Bioorg. Med. Chem. 2002, 10, 3481–
3487.
15. Shim, J. S.; Kim, D. H.; Jung, H. J.; Kim, J. H.; Lim, D.;
Lee, S. K.; Kim, K. W.; Ahn, J. W.; Yoo, J. S.; Rho, J. R.;
Shin, J.; Kwon, H. J. Bioorg. Med. Chem. 2002, 10, 2987–
2992.
16. Anderson, A. M.; Mitchell, M. S.; Mohan, R. S. J. Chem.
Educ. 2000, 77, 359–360.
17. Lampe, W.; Smolinska, J. Bull. Acad. Polon. Sci. 1958, 6,
481–486.
18. Maffei Facino, R. M.; Cairini, M.; Saibene, L. Arch.
Pharm. Med. Chem. 1996, 329, 457–463.
19. Barclay, L. R. C.; Vinqvist, M. R. Org. Lett. 2000, 2,
2841–2843.
Figure 4. Binding of 4 into the active site of COX-2.

C. Selvam et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1793–1797
1797
20. Torres, J. L.; Lozano, C.; Julia, L.; Sanchez-Baeza, F. J.;
Anglada, J. M.; Centelles, J. J.; Cascante, M. Bioorg. Med.
Chem. 2002, 10, 2497–2509.
23. Selvam, C.; Jachak, S. M.; Thilagavathi, R.; Chakraborti,
A. K. Indian WTO Patent Application No. 1704/DEL/
2004, 2004.
21. Mittal, S.; Malde, A.; Selvam, C.; Arun, K. H. S.; Johar,
P. S.; Jachak, S. M.; Ramarao, P.; Bharatam, P. V.;
Chawla, H. P. S. Bioorg. Med. Chem. Lett. 2004, 14, 979–
982.
24. SYBYL 6.9 Molecular Modelling Software; Tripos Asso-
ciates Inc.: 1699 S. Hanley, St. Louis MI 63144, USA.
25. Abola, E. E.; Berstein, F. C.; Bryant, S. H.; Koetzle, T. F.;
Weng, J. Protein Data Bank. In Crystallographic Dat-
abases—Information Content, Software Systems, Scientific
Applications; Allen, F. H., Berjerhoff, G., Sievers, R. R.,
Eds.; Data Commission of the International Union of
Crystallography: Bonn, 1987; p 171.
26. Chakraborti, A. K.; Thilagavathi, R. Bioorg. Med. Chem.
2003, 11, 3989–3996.
27. Halgren, T. J. Am. Chem. Soc. 1990, 112, 4710–
4723.
22. Spectroscopic data for 8: 1H NMR (CD3OD, 300 MHz): d
3.90 (s, 3H, –OCH3), 6.71 (s, 1H), 6.82 (d, 3H, J = 8.3),
6.93 (m, 2H), 7.06 (d, 1H, J = 8.2), 7.19 (s, 1H), 7.27 (m,
2H), 7.46 (d, 2H, J = 8.5); 13C NMR (CD3OD, 75 MHz):
170.31, 163.92, 159.64, 159.40, 149.34, 148.80, 138.12,
137.87, 136.43, 136.22, 129.92, 129.44, 122.59, 122.34,
116.83, 113.37, 111.37, 110.93, 98.7, 56.49; IR (KBr) Vmax
3441.3, 2921.0, 2851.8, 1575.2, 1511.9, 1269.2 cmÀ1
;
APCI-MS m/z 336.1 (M+1). (C20H17NO4) C, H, N, calcd
71.63, 5.11, 4.18; found 71.59, 5.12, 4.19.
28. Dannhardt, G.; Laufer, S. Curr. Med. Chem. 2000, 7,
1101–1112.