Molecular characterization, biological activity, and in silico study of 2-(3,4-dimethoxyphenyl)-3-(4-fluorophenyl)-6-methoxy-4H-chromen-4-one as a novel selective COX-2 inhibitor

Article abstract of DOI:10.1016/j.molstruc.2014.10.004

The present study aimed to characterize and investigate 2-(3,4-dimethoxyphenyl)-3-(4-fluorophenyl)-6-methoxy-4H-chromen-4-one (22) as a novel selective COX-2 inhibitor. The data collected from the single X-ray crystallographic analysis and in silico study

Full text of DOI:10.1016/j.molstruc.2014.10.004

Accepted Manuscript
Molecular characterization, biological activity, and in silico study of 2-(3,4-
dimethoxyphenyl)-3-(4-fluorophenyl)-6-methoxy-4H-chromen-4-one as a nov-
el selective COX-2 inhibitor
Kamal Rullah, Mohd Fadhlizil Fasihi Mohd Aluwi, Bohari M Yamin, Mohd
Syukri Baharuddin, Nor Hadiani Ismail, Hilwan Yuda Teruna, Syed Nasir
Abbas Bukhari, Ibrahim Jantan, Juriyati Jalil, Khairana Husain, Lam Kok Wai
PII:
S0022-2860(14)01005-9
DOI:
http://dx.doi.org/10.1016/j.molstruc.2014.10.004
MOLSTR 20999
Reference:
Journal of Molecular Structure
To appear in:
Received Date:
Revised Date:
Accepted Date:
6 August 2014
2 October 2014
2 October 2014
Please cite this article as: K. Rullah, M.F.F. Aluwi, B.M. Yamin, M.S. Baharuddin, N.H. Ismail, H.Y. Teruna, S.N.A.
Bukhari, I. Jantan, J. Jalil, K. Husain, L.K. Wai, Molecular characterization, biological activity, and in silico study
of 2-(3,4-dimethoxyphenyl)-3-(4-fluorophenyl)-6-methoxy-4H-chromen-4-one as a novel selective COX-2
inhibitor, Journal of Molecular Structure (2014), doi: http://dx.doi.org/10.1016/j.molstruc.2014.10.004
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Molecular characterization, biological activity, and in silico study of 2-(3,4-
dimethoxyphenyl)-3-(4-fluorophenyl)-6-methoxy-4H-chromen-4-one as a
novel selective COX-2 inhibitor
Kamal Rullaha,e, Mohd Fadhlizil Fasihi Mohd Aluwia, Bohari M Yaminb, Mohd Syukri
Baharuddinc, Nor Hadiani Ismailc,d, Hilwan Yuda Terunaf, Syed Nasir Abbas Bukharia,
∗
Ibrahim Jantana, Juriyati Jalila, Khairana Husaina, Lam Kok Waia
aDrugs and Herbal Research Centre, Faculty of Pharmacy, Universiti Kebangsaan Malaysia, Jalan Raja Muda
Abdul Aziz, 50300 Kuala Lumpur, Malaysia
bSchool of Chemical Sciences and Food Technology, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor,
Malaysia
cAtta-ur-Rahman Institute for Natural Product Discovery (RiND), Universiti Teknologi MARA, Kampus Puncak
Alam, 42300 Bandar Puncak Alam, Shah Alam, Selangor, Malaysia
dFaculty of Science, Universiti Teknologi MARA, 50400 Shah Alam, Selangor, Malaysia
eSekolah Tinggi Ilmu Farmasi Riau, Universitas Riau, Kampus Bina Widya Km 12.5, Simpang baru-Pekanbaru,
Indonesia 28293
fDepartment of Chemistry, FMIPA-Universitas Riau, Kampus Bina Widya Km 12.5, Simpang baru-Pekanbaru,
Indonesia 28293
Abstract
The present study aimed to characterize and investigate 2-(3,4-dimethoxyphenyl)-3-(4-
fluorophenyl)-6-methoxy-4H-chromen-4-one (22) as a novel selective COX-2 inhibitor. The
data collected from the single X-ray crystallographic analysis and in silico study provide
important insights on the molecular conformation and the binding interactions that are
responsible for the COX-2 selectivity.
Keywords: chromone, single-crystal XRD, prostaglandin E2, COX-2 inhibitor, docking
simulation
∗Corresponding author Tel.: +60 392 897 031

E-mail: david_lam_98@yahoo.com (K.W. Lam)
1. Introduction
Nonsteroidal anti-inflammatory drugs (NSAIDs) are useful therapeutics for the treatment
of pain and inflammation, particularly arthritis. The main drug target of NSAIDs is
cyclooxygenase (COX), which bio-catalyzes the first committed step in arachidonic-acid
metabolism [1]. Most NSAIDs not only selectively inhibit the production of prostaglandins
(PGs) at the sites of inflammation but also other PGs which serve important functions in other
parts of the body, a factor which accounts for some of the toxicity of these agents. The most
frequent complication associated with NSAIDs usage can be found in the gastrointestinal tract
(GIT). GI bleeding, ulceration, perforation, and obstruction are a significant cause of
morbidity and mortality in patients who are treated with these agents [2, 3]. To counter these
limitations, novel scaffolds with selective COX-2 or mPGES-1 inhibition need to be designed
and evaluated [4].
Since COX-2 enzyme, also known as PGH synthase was cloned in 1992 [5], DuPont
company had developed a compound namely DuP-697 (1), that was potent in many anti-
inflammatory assays and did not have the ulcerogenic effects like any conventional NSAIDs
in rats at a single dose up to 400 mg/kg [6]. DuP-697 eventually became the major building-
block for the synthesis of many other new COX-2 inhibitors. Celecoxib and rofecoxib, the
first two COX-2 inhibitors that reached the market, were generally based on DuP-697.
Celecoxib (2) (Celebrex®) is also the first and only COX-2 inhibitor available in the United
States. However, rofecoxib (4) (Vioxx®) and valdecoxib (6) (Bextra®) were known to increase
the risk of heart attacks and strokes with long term use, therefore, both drugs were withdrawn
from the market in September 2004 and April 2005 [7-11]. This eventually led to the
development of new selective COX-2 inhibitors that could act effectively without exerting
major side effects associated with the traditional NSAIDs [12]. The solving of the 3D
structures of an unliganded murine COX-2 and SC-558-protein complex (3) by Kurumbail et
al. (1996) finally shed some light on the structural basis for the selective inhibition of COX-2
which also demonstrate some of the conformational changes associated with time-dependent
inhibition [13]. The development of new generation of selective COX-2 inhibitors typically
relied on the structural modifications of coxib drugs, for example DFU (5), etoricoxib (7),
parecoxib (8), and MPO-0029 (9) [14-16] (Figure 1).
Previous study had shown that 7-methanesulfonylamino-6-phenoxychromone (10), a
synthetic chromone-based exhibited significant potency in the rat models of carrageenan-

induced edema (CPE) and adjuvant-induced arthritis (AA) [17]. Further structural activity
relationship study conducted on compound 10 led to the discovery of the lead structure, T-614
(11). T-614 not only demonstrated potent anti-arthritic activity in chronic inflammatory
disease models but also low gastro-ulcerogenic liability in oral administration and appeared to
be a good drug candidate [18, 19]. T-614 was then introduced in clinical evaluation with 50
mg/day therapy and was found to be as effective as methotrexate (MTX) therapy in 24 weeks.
This signifies a new option for the therapy of patients with active rheumatoid arthritis (RA)
[20]. The presence of chromone scaffold in some natural and synthetic compounds without
the presence of sulfonyl or sulfonamide moiety had also been reported for possessing good
COX inhibition. For example, stellatin (12) and eugenin (13), isolated from
Dysophyllastellata were reported to be selective towards COX-2 inhibition [21, 22]. On the
other hand, a series of chrysin derivatives bearing dichloro group at positions 3’ and 4’ (e.g
compounds 14-16) had been prepared and were found to exhibit good inhibitory activity on
prostaglandin production by COX-2 (IC50 = 0.1-0.5 μM) [23]. In our previous study, we have
also observed that the chromone scaffold significantly active in suppressing the PGE2
production secreted by lipopolysaccharide-induced mouse macrophage cells (RAW 264.7)
[24].
Conventionally, preserving the sulfonyl (R-SOO-R) or sulfonamide (R-SOO-NH2)
functional group is an important criterion in the design of inhibitors with potent COX-2
inhibition and high COX-1/COX-2 selectivity (e.g compounds 1-11). On the possible
downside, long-term use of COX-2 traditional inhibitors bearing these moieties might be
implicated with the increase in cardiovascular side effects such as heart attack and stroke
incidents. Another distinctive characteristic of these inhibitors is the presence of a five-
membered ring in the central core structure. Despite the fact that chromone structure has been
well studied in recent years, the ligand-protein complex binding interactions between
chromones and COX-2 have never been disclosed to date. To initiate this study, first the five-
membered ring core was replaced with a chromone ring (Figure 2) and then the synthesized
compound was further evaluated for the COX-2 inhibitory activity and the COX-1/COX-2
selectivity. The compound structural conformation and its possible binding interactions in the
enzyme binding site were investigated using ligand X-ray diffraction technique and
established molecular modeling methods.

2. Result and Discussion
2.1 Chemical synthesis
Figure 3 illustrates the general retrosynthetic strategy adopted in the synthesis of 2,3-
diarylbenzopyran. The preparation of the 2′-hydroxychalcone derivative was carried out via
the base-promoted Claisen-Schmidt condensation (Scheme 1). First, commercially available
2'-hydroxy-4'-methoxy-acetophenone (17) was reacted with 3,4-dimethoxybenzaldehyde (18)
in the presence of KOH in EtOH. Subsequent neutralization with 1M HCl yielded the desired
product, (E)-3-(3,4-dimethoxyphenyl)-1-(2-hydroxy-4-methoxyphenyl)prop-2-en-1-one (19).
The reaction was highly successful with 95% of yield when it was carried out under the
ultrasound irradiation at 40 ºC [25, 26]. Next, cyclization of 19 using iodine in DMSO
furnished the product 2-(3,4-dimethoxyphenyl)-7-methoxy-4H-chromen-4-one (20) in
moderate yield (57%). Several attempts to introduce the bromine group to 20 using
ammonium bromide and ammonium persulphate at room temperature by grinding in a
solvent-free condition via green synthesis were unsuccessful and did not yield the target
product [27]. Several other preparation methods were referred [28-31] to obtain the target
product 3-bromoflavone which is an important intermediate for the C-3 modification, and it
was found that the treatment of 20 with Bu4NBr/PhI(OAc)2 in CH2Cl2 at room temperature
under nitrogen atmosphere successfully produced 3-bromoflavone, 21 in high yield, 98%
[32].
In order to introduce the fluorophenyl group to compound 21, a straightforward and viable
Suzuki reaction was adopted. This method was previously reported by Dyrager et al. (2011),
in which the introduction of various aryl groups to 3-bromochomones and 3-iodochromones
via Suzuki coupling was easily accomplished in the presence of palladium catalyst [31]. In
our case, the introduction of the fluorophenyl group to 21 was easily done using the in situ
generated palladium nanoparticles in PEG-400 solvent under aerobic condition to give the
target product of 2-(3,4-dimethoxyphenyl)-3-(4-fluorophenyl)-6-methoxy-4H-chromen-4-one
(22) in good yield (64%). This oxygen-promoted ligand-free Suzuki coupling reaction had
been previously reported by Han et al. (2007) [33].
2.2 NMR characterization

The full characterization of compound 22 was conducted by one-dimensional (1D), two-
dimensional (2D) NMR, single crystal X-ray structural analysis, and high resolution mass
spectrometry (ESI-HRMS). It was found that all spectra were consistent with the assigned
structures of the synthesized compounds (Supplementary data). Figure 4 shows the
1
comparison of the H-NMR spectra (500 MHz, CDCl3) for the intermediates 19-21 and the
final target compound 22. The proton chemical shift signals at 13.56 ppm and 7.8 ppm are the
key determinants for the presence of hydroxyl group and H unsaturated in chalcone structure
β
1
as displayed in the H NMR for compound 19. These two signals eventually dissolved
following the ring closure to give the flavone backbone ring of compound 20. Addition of the
bromine group at C-3 of 20 resulted in the obvious loss of the singlet peak at δ 6.72 ppm. On
H
the other hand, presence of the additional aromatic proton peaks (4H) around the region of δ
H
7 ppm signified that the coupling reaction between the intermediate (compound 21) and 4-
fluorophenyl boronic acid was successful. Close inspection on the final spectrum revealed a
doublet signal at δ 8.20 (H-6) and a doublet of doublet signal at δ 7.02 (H-7) which belong
H
H
to the central core ring of compound 22. The adjacent protons H-12 and H-13 belong to the 3,
4-dimethoxybenzyl ring were identified as two doublet signals at δ 7.17 and 6.83 ppm with
H
8.5 Hz coupling constant each. Furthermore, a doublet signal of H-9 at δ 6.96 was observed
H
to couple with H-7, displaying a similar coupling constant of 2.4 Hz. H-16 was identified as a
doublet signal at δ 6.77 ppm with 2.1 Hz coupling constant showing that the signal coupled
H
with H-12. The protons belong to the symmetrical 4-fluorobenzyl ring were observed as two
doublet of doublet signals at δ 7.25 and 7.05 ppm with integration of two protons each,
H
respectively.
The ambiguous assignment of desired final product 2-(3,4-dimethoxyphenyl)-3-(4-
fluorophenyl)-6-methoxy-4H-chromen-4-one (22) was further supported by the combined
information retrieved from 1D NMR; DEPT-90, DEPT-135 and (2D) NMR techniques
experiments; HMQC, HMBC, and finally high resolution mass spectrometry (Table 1,
Figure 5, and Supplementary Data). Figure 5 summarizes the HMBC correlations for
compound 22. The 2D NMR analyses (HMQC and HMBC) confirmed the presence of the
4H-chromen-4-one core in compound 22. Finally, no signal corresponding to the CH2 group
was observed in DEPT spectra further supports the structure of 22 (Supplementary data).
2.3 X-ray structural characterization

The characterization of compound 22 was conducted by a single crystal X-ray structural
analysis. The result obtained was important to determine the position of the dimethoxyl group.
The crystallization was carried out in ethanol. The compound crystallized in a monoclinic
system with space group Pī, a = 15.5760(13) Å, b = 8.5079(7) Å, c = 16.3424(13) Å, α = 90º,
β = 114.509(5)º, γ = 90º, Z = 4 and V = 1970.5(3) Å3. The crystal system and refinement data
are shown in Table 2. Figure 6 shows the molecular structure with numbering scheme. The
whole molecule is not planar but the 4-fluorophenyl F1/C2/(C19-C24), 3,4-dimethoxyphenyl
O4/O5/C1/(C11-C17) and 4H-chromen-4-one, (O1/O2/(C1-C11) are planar with maximum
deviation of 0.082(3) Å for C2 atom from the least square plane. The central chromone ring
makes dihedral angles of 50.34(10)ºꢀ and 42.51(6)º with the 4-fluorophenyl and 3,4-
dimethoxyphenyl rings, respectively. The dihedral angle between the 4-fluorophenyl and 3,4-
dimethoxyphenyl is 58.23(11)º. The bond lengths and angles are in normal ranges. The crystal
structure is stabilized by C10-H10C···O2 (1/2+x, 3/2-y, 1/2+z; D···A 3.344(4) Å and D-H···A
159º)ꢀ and C17-H17B···O3 (1/2+x, 3/2-y, -1/2+zl; D···A 3.339(5) Å and D-H···A 139º)
intermolecular hydrogen bonds (Figure 7, Table 3). In addition, there is also C24-H24···π
bond (1+x, 2-y, 2-z) with the distance of H24···Cg2: (C4-C9) and C24···H24···Cg2 angle of
127º.ꢀThe crystallographic data for the structural analysis were deposited with the Cambridge
Crystallographic Data Centre (CCDC) No 1015759.
2.4 Biological activity
Initially, we examined the cytotoxicity effect of compound 22 on RAW 264.7 cells using
MTT assay. Compound 22 did not show any cytotoxic effect on RAW 264.7 cells even at
high concentration of 50 μM. Therefore, any suppressive effect on PGE2 production
demonstrated by compound 22 could not be attributable to cell cytotoxicity. Compound 22
was then evaluated for the PGE2 inhibitory activity using the same cell line, stimulated by
lipopolysaccharide (LPS) at 50 μM concentration. The results summarized in Table 4 and
Figure 8 showed that compound 22 suppressed PGE2 production with an IC50 value of 13.8
μM.
On the other hand, we also tested the effects of compound 22 on the activities of COX-1
and COX-2. The experiments were performed using the ovCOX-1 and human recombinant
COX-2 and were compared with indomethacin, which is a non-selective COX inhibitor, as a
positive control. This primary data showed that compound 22 exhibited moderate inhibition
on COX-2 activity with an IC50 value of 38.6 μM while on the contrary it did not inhibit the

COX-1 activity even at 100 μM concentration (Table 5). From the result, it was obvious that
compound 22 was highly selective in the inhibition of COX-2 over COX-1 with the
selectivity index (SI) value of more than 2.59. Even though the inhibition was much weaker
than the coxibs [34, 35] which was most possibly due to the absence of the sulfonyl or
sulfonamide functional group from the compound, this data is sufficient to substantiate that 22
was highly selective and could be further developed as a potential selective COX-2 inhibitor
(comparison with SI value 0.06481 of indomethacin drug). The reason for the selectivity will
be discussed in the next section. Future studies will address on the SAR study of compound
22 as a potential selective COX-2 inhibitor.
2.5 Docking analysis of coxibs and compound 22
The protein crystal structures of muCOX-2 (6COX: 2.8 Å) [13] and ovCOX-1 (3KK6:
2.75 Å) [36] were retrieved from the Protein Data Bank (PDB) and prepared for the docking
studies. We opted for these two protein crystal structures for two reasons: i) Rimon et al.
(2010) successfully determined how celecoxib could bind to the active site in one of the
COX-1 monomers (3KK6). Since celecoxib is a close analogue to SC-558, this provides us
an opportunity to understand how these two coxibs could bind to COX isoforms, and ii)
compound 22 shares similar structure with the coxibs except the central core ring.
To validate the docking protocol, the co-crystallized ligands were first removed and then
re-docked to the protein active site using the LibDock module. The final results showed that
the Root-Mean-Square Deviation (RMSD) between the co-crystallised ligands with the
redocked ligands were diminutive (<2.0 Å) (Fig. S1 and Table S1, Supplementary data).
This shows that LibDock was able to predict the ligand binding conformation accurately.
Compound 22 was subjected to a single crystal X-ray experiment and the data retrieved was
solved by direct methods and further refined by least-squares methods using SHELXL-97
program. The obtained structural data of 22 and other selected coxibs (compounds 1-9) were
then employed in the docking study using LibDock module. LibDock uses protein site
features referred to as HotSpots. HotSpots consist of two types which are polar and apolar.
The polar Hotspot is preferred by a polar ligand atom (for example a hydrogen bond donor or
acceptor) and an apolar HotSpot is preferred by an apolar atom (for example a carbon atom).
Two general observations can be detailed from the docking analysis: (i) compounds 1-9
exhibit similar alignments as compared to the co-crystallized ligand SC-558 and (ii) the
aromatic ring A bearing the sulfonyl or group and the aromatic ring B of compounds 1-9 lie in

the same plane with similar dihedral angle as the co-crystallized ligand SC-558 (data not
shown here). Unlike the coxibs, compound 22 was predicted to bind in two distinct
conformations (Figure 9).
In the first conformation, compound 22 binds similar to the coxibs. The central chromone
core ring is well positioned in the hydrophobic pocket 3 formed by Met113, Val116, Val349,
Tyr355, Leu359, and Ala527 residues. The chromone ring could also form weak cation-π
interactions with the positively charged guanidium group of Arg120 within distances of 3.9 Å
and 4.3 Å, respectively. It is known that classical NSAIDs interact with the charged Arg120
residue by charge-charge interaction, which appears to be an unusual interaction for the
coxibs [37]. This could be one of the primary reasons that contribute to the selectivity.
Another interesting observation is the position of the methoxyl group of compound 22 which
could be seen to protrude deeply into the cavity as compared to the trifluoromethyl group of
SC-558. Mostly hydrophobic contacts could be observed between the fluorophenyl ring and
the hydrophobic residues in pocket 1, a similar trend in most of the coxibs. On the other hand,
the bulky dimethoxyphenyl ring is found to occupy pocket 2, which is surrounded by
hydrophobic residues of Leu352, Tyr355, Phe518, Val523, and Ser353 which further extends
to the polar residues (His90, Gln192, and Arg513). The methoxyl group engaged with both
the imidazole and guanidium side chains of His90 and Arg513 by hydrogen bonding
interaction within 2.7Å and 3.2Å, respectively. The presence of sulphonamide moiety in
COX-2 inhibitors is highly vital for the activity as it interacts strongly with His90, Gln192,
and Arg513. In marked contrast, the interactions exhibited by the methoxyl group might be
very much weakened due to the presence of methyl group that hinders the direct interactions
between the electronegative oxygen atom and the polar residues.
Based on the docking results obtained, it was possible that compound 22 could adopt
another binding conformation as shown in Figure 9. The chromone ring is bound in the
hydrophobic pocket 2 which is intriguingly similar to the binding of the o-indole ring of
indomethacin as displayed in another COX-2 protein crystal structure (4COX). Additionally,
the carbonyl group of chromone ring forms a 2.2 Å hydrogen bonding with Tyr355. However,
there are no hydrogen bonding interactions observed between the methoxyl group and the
polar residues in this cavity. Substitution of the methoxyl group with a polar group such as
hydroxyl or sulphonamide is highly anticipated in the next study. The binding of
dimethoxyphenyl ring to the hydrophobic pocket 1 is also reinforced by the hydrogen bonding
interactions between the hydroxyl group Tyr385 and backbone amide of Ser530 within

distances of 3.0Å and 2.9Å, respectively. Finally, the fluorophenyl ring make contacts with
the adjacent residues in hydrophobic pocket 3.
The lack of COX-1 inhibition profile as displayed by compound 22 appeared to result from
the inability of the dimethoxylphenyl ring to bind to hydrophobic pocket 2 as shown in
Figure 10. Substitution of valine to isoleucine at position 523 in COX-1 might be the major
culprit behind the selectivity as it blocked the entrance of ring A into the more restricted
hydrophobic pocket 2. Even though Rimon et. al (2010) have shown that coxibs are able to
bind to COX-1 by structural changes in this pocket, the presence of two methoxyl groups in
ring A might be too bulky preventing the binding from taking place. Another important factor
might be the position of His513 which is more distant, preventing the methoxyl group from
making hydrogen bonding interactions with this residue. Instead, LibDock proposed a
different binding conformation for compound 22 as shown in Figure 10.
2.6 Identification of common pharmacophore model
To rationalize the predicted binding conformation of compound 22 to the binding site
COX-2, we attempt to map it to the common pharmacophore built from a set of well-known
selective COX-2 inhibitors (compounds 1-9). The best docked conformation of compounds 1-
9 were then used in the identification of the similar geometrical features using the common
feature pharmacophore generation protocol implemented in Discovery Studio® 3.1 (Accelrys
Inc., San Diego, CA, USA). The process for constructing the pharmacophore in this study is
shown in Figure 11. The four common chemical features identified including 2 aromatic
rings (RA) and 2 hydrogen bond acceptors. The HBA_3 maps the electronegative atom of the
ring C while HBA_4 maps the methanesulfonyl or sulfonamide group of ring A. The ring
aromatic features (RING_AROMATIC_1 and RING_AROMATIC_2) map the aromatic ring
A and ring B attached to the five-membered central core ring. The distances between
RING_AROMATIC_1 and HBA_3 or HBA_4 were found to be 5.053 Å and 4.869 Å,
respectively. The distances between RING_AROMATIC_2 and HBA_3 or HBA_4 were
found to be 5.051 Å and 9.021 Å, respectively. On the other hand, the distance between
HBA_3 and HBA_4 was found to be 9.335 Å. Figure 11 represents the alignment of the
selected coxibs (compounds 1-9) against Hypo1. This alignment was found to exhibit good
match of features of the coxibs with good fit values in the range of 0.6 to 0.9 for the training
compounds. Fit value implemented in Catalyst was used as a measure of how well the ligand
fits the pharmacophore. The higher the fit score, the better the match. The fit value is

summarized in Table 6. When compound 22 was superposed to the pharmacophore, it
exhibited good fit (Fitvalue = ~0.8) as compared to the coxibs. The mapping result highly
correlated with the docking results further support the proposed binding conformation of 22 as
discussed earlier.
3. Experimental
3.1 Chemical synthesis and characterization
All reagents were purchased from Aldrich, Acros and Merck and were used without further
purification. All the solvents used in the syntheses were analysis and synthesis grade. The
solvents used in spectroscopic measurements were spectroscopic grade. Melting points were
determined on a STUART SMP10 melting point apparatus. 1H, 13C NMR, DEPT, HMQC and
HMBC spectra were recorded on a Bruker 500 MHz spectrometer. ESI-HRMS spectra were
recorded on a Bruker micrOTOF Mass Spectrometer.
3.1.1 Synthesis of (E)-3-(3,4-dimethoxyphenyl)-1-(2-hydroxy-4-methoxy phenyl) prop-2-en-
1-one (19)
KOH 40% (15 mL) was added to a stirred solution of 2-hydroxy-4-methoxyacetophenone
(17) (1.66 g, 10 mmol) and 3,4-dimethoxybenzaldehyde (18) (1.66 g, 10 mmol) in absolute
EtOH (30 mL). The mixture was sonicated for 60 min and was then stirred at room
temperature overnight. The dark red solution was acidified with 1M of HCl to form a yellow
precipitate, which was filtered and rinsed with distilled water. The precipitate was subjected
to recrystallization using absolute EtOH to afford 19 as yellow solid, yield 95%, mp. 157-159
oC; 1H NMR (500 MHz, CDCl3) δ 13.56 (s, 1H), 7.87 (t, 2H), 7.46 (d, J = 15.4 Hz, 1H), 7.27
(dd, J = 8.3, 1.8 Hz, 1H), 7.18 (d, J = 1.9 Hz, 1H), 6.93 (d, J = 8.3 Hz, 1H), 6.51 (dd, J = 11.5,
13
2.5 Hz, 2H), 3.99 (s, 3H), 3.96 (s, 3H), 3.88 (s, 3H). C NMR (126 MHz, CDCl3) δ 191.95,
166.81, 166.24, 151.79, 149.50, 144.72, 131.26, 127.99, 123.44, 118.23, 114.31, 111.38,
110.51, 107.78, 101.23, 77.16, 56.17, 55.71. ESI-HRMS: (C18H18O5) calc. [M+Na+]
337.1052, found 337.1078.
3.1.2 Synthesis of 2-(3,4-dimethoxyphenyl)-7-methoxy-4H-chromen-4-one (20)

Iodine (0.01 equiv) was added to a solution of 19 (1 g, 3.2 mmol) in DMSO (10 mL). The
solution was refluxed for 1 h. The mixture was then poured into saturated Na2S2O3. The
aqueous phase was extracted with EtOAc for 3 times. The crude was purified by flash
chromatography (n-Hexane:EtOAc, 3:6 ratio) to afford 20 as pale yellow crystal, yield 52%,
mp. 184-185 oC;1H NMR (500 MHz, CDCl3) δ 8.14 (d, J = 8.8 Hz, 1H), 7.55 (dd, J = 8.5, 2.1
Hz, 1H), 7.38 (d, J = 2.1 Hz, 1H), 6.99 (dt, J = 6.9, 2.6 Hz, 3H), 6.72 (s, 1H), 4.00 (s, 3H),
3.98 (s, 3H), 3.95 (s, 3H).13C NMR (126 MHz, CDCl3) δ 177.76, 164.14, 163.11, 157.93,
152.01, 149.32, 127.03, 124.36, 119.87, 117.74, 114.25, 111.20, 108.88, 106.36, 100.47,
56.11, 56.08, 55.85. ESI-HRMS: (C18H18O5) calc. [M+Na+] 335.0896, found 335.0950.
3.1.3 Synthesis of 3-bromo-2-(3,4-dimethoxyphenyl)-7-methoxy-4H-chromen-4-one (21)
PhI(OAc)2 (500 mg, 1.55 mmol) was suspended in anhydrous CH2Cl2 (10 mL) under
nitrogen atmosphere at room temperature. Bu4NBr (1.55 mmol) in CH2Cl2 (10 mL) was added
into the mixture and stirred at room temperature for 30 min. Finally, solution of 20 (245 mg,
0.77 mmol) in anhydrous CH2Cl2 (5 mL) was added slowly into the mixture followed by
stirring at room temperature for 8 h. The mixture was quenched with saturated aqueous
NH4Cl and the aqueous phase was extracted with CH2Cl2. The organic layer was dried with
MgSO4 anhydrous and evaporated in vacuo. The crude was purified by flash chromatography
(n-Hexane:EtOAc, 3:6 ratio) to afford 21 as white solid, yield 98%, mp. 165-166 oC; 1H NMR
(500 MHz, CDCl3) δ 8.20 (d, J = 8.9 Hz, 1H), 7.52 (dd, J = 8.4, 2.1 Hz, 1H), 7.43 (d, J = 2.1
Hz, 1H), 7.05-7.00 (m, 2H), 6.91 (d, J = 2.3 Hz, 1H), 4.00 (s, 3H), 3.98 (s, 3H), 3.94 (s, 3H).
13C NMR (126 MHz, CDCl3) δ 172.47, 164.42, 161.24, 157.31, 151.35, 148.53, 127.95,
125.23, 123.06, 115.67, 115.08, 112.47, 110.57, 108.73, 99.92, 56.19, 56.06, 55.90. ESI-
HRMS: (C18H16O5) calc. [M+Na+] 415.0001, found 415.0003.
3.1.4 Synthesis of 2-(3,4-dimethoxyphenyl)-3-(4-fluorophenyl)-7-methoxy-4H-chromen-4-
one (22)
An appropriate amount of 4-fluorophenylboronic acid (4 eq) was added to a suspension of
21 (1 eq) in PEG-400 (15 g) followed by addition of Pd(OAc)2 (0.2 equiv) and K2CO3 (2
equiv). The mixture was heated at 60 ºC for 20 h. The reaction mixture was then poured over
brine and extracted with EtOAc (3 times). The combined organic phase was dried with
MgSO4 anhydrous, and evaporated in vacuo. The crude was purified by flash chromatography

(n-Hexane:EtOAc, 3:6 ratio) to afford 22 as white crystals, yield 64%, mp. 197-198 oC. Good
quality crystals were obtained by recystallization from absolute ethanol. 1H NMR (500 MHz,
CDCl3) δ 8.20 (d, J = 8.9 Hz, 1H), 7.25 (dd, J = 8.7, 5.5 Hz, 2H), 7.17 (dd, J = 8.5, 2.1 Hz,
1H), 7.05 (dd, J = 13.5, 8.6 Hz, 2H), 7.02 (dd, J = 8.9, 2.4 Hz, 1H), 6.96 (d, J = 2.3 Hz, 1H),
6.83 (d, J = 8.5 Hz, 1H), 6.77 (d, J = 2.0 Hz, 1H), 3.96 (s, 3H), 3.91 (s, 3H), 3.61 (s, 3H).13C
NMR (126 MHz, CDCl3) δ 176.68, 164.21, 163.19, 160.93, 157.66, 150.61, 148.22, 132.98,
132.92, 129.52, 127.74, 125.35, 122.74, 120.94, 117.27, 115.50, 115.33, 114.47, 112.85,
110.56, 100.11, 55.90, 55.85, 55.63. ESI-HRMS: (C24H19FO5) calc. [M+Na+] 429.1115, found
429.1120.
3.2 X-ray structure determination
Single crystal X-ray experiment of 22 was performed on Bruker D-QUEST diffractometer
using graphite-monochromated Mo-Kα radiation (λ=0.71073 Å) (Bruker, AXS Inc., Madison,
WI, USA). Intensity data was measured at 301(2) K by the ω-scan. Accurate cell parameters
and orientation matrix were determined by the least-squares fit of 25 reflections. Intensity
data was collected for Lorentz and polarisation effects. Empirical absorption correction was
carried out using multi-scan. The structure was solved by direct methods and least-squares
refinement of the structure was performed by the SHELXL-97 program. All the non-hydrogen
atoms were refined anisotropically. The hydrogen atoms were placed in calculated positions,
allowing them to ride on their parent C atom with Uiso(H)=xUeq(C) where, x=1.5 for methyl;
1.2 for non-methyl groups, except the hydrogen atoms attached to oxygen atoms were located
from Fourier maps and refined isotropically. A summary of the data collections and details of
the structure refinement is given in Table 2. Crystallographic data for the structural
determination has been deposited with the Cambridge Crystallographic Data Center, CCDC
No 1015759. This information may be obtained free of charge from the Director, CCDC, 12
Union Road, CambridgeCB2, 1EZ, UK (fax: +44-1223336033; e-mail: deposit@ccdc.cam.uk
or http://www.ccdc.cam.ac.uk/const/retrieving.html).
3.3 Biology evaluation
3.3.1 Cell Culture.
The murine macrophages-like cell line (RAW 264.7) from the American Type Culture
Collection (ATCC® TIB-71™, Manassas, USA) were maintained in DMEM supplemented

with 10% FBS, 4.5 g/L glucose, sodium pyruvate (1 mM), L-glutamine (2 mM), streptomycin
(50 μg/mL) and penicillin (50 U/mL) at 37 °C and 5% CO2. When RAW 264.7 cells reached
confluences of 80-90%, the cells were scraped out and centrifuged at 110 × g in 4 °C for 10
min. The cell viability of cultured cells used in the assay was always > 95% as determined by
trypan blue dye exclusion.
3.3.2 Cell stimulation and treatment.
RAW 264.7 (4 × 105 cells/well) were seeded into a tissue culture grade 96-well plate
except for blank and incubated for 2 h at 37 °C, 5% CO2 for cell attachment. The attached
cells were activated with 100 U/mL of recombinant IFN-γ and 5 μg/mL of LPS with or
without presence of synthetic compound at a final volume of 100 μL/well. DMSO was used
as vehicle to add synthetic compound into the culture medium and the final concentration of
DMSO was 0.1% in all cultures. Cells were then incubated at 37 °C, 5% CO2 for 17-20 h. The
level of PGE2 was determined by using PGE2 EIA kit (Catalog No. 500141).
3.3.3 Cell viability.
The cytotoxicity of synthetic compound on cultured cells was determined by assaying the
reduction of MTT reagents to formazan salts. After treatment, the supernatant of 96-wells
plate containing cells were removed and MTT reagents (0.05 mg/mL) were added into each
well. The cells were incubated in 37 °C for 4 h and the formazan salts were dissolved by
adding 100% DMSO. The absorbance was then measured at 570 nm on a SpectraMax Plus
Microplate reader (Molecular Devices Inc., Sunnyvale, CA, USA).
3.3.4 Determination of PGE2.
The cell culture supernatants were collected and analyzed for PGE2 secretion PGE2 EIA
kits (Cayman Chemical, Ann Arbor, MI, USA). The protocols provided by the manufacturers
were followed to the detail. The data was acquired using a SpextraMax Plus microplate reader
(Molecular Device, Sunnyvale, CA, USA). The concentration of PGE2 for each sample was
calculated from their respective standard curves.

3.3.5 Cyclooxygenase Assay.
The effect of 22 on COX-1 and COX-2 was determined by using a COX Inhibitor
Screening Kit (Catalog No. 560131) from Cayman Chemicals, Ann Arbor, Michigan, U.S.A.
The reaction mixture was prepared with 1μM heme and COX-1 (ovine) or COX-2 (human
recombinant) enzymes in 100mM Tris-HCl buffer, and then pre-incubated for 10 min in a
water bath (37 °C) at pH 8.0. The reaction was started by adding 10 μL arachidonic acid (final
concentration in reaction mixture 100 μM). After 2 min, the reaction was terminated by
adding 1M HCl and lastly PGE2 was quantified by ELISA. The compounds were dissolved in
DMSO and diluted to desired concentration with potassium phosphate buffer. After
transferring the compounds to 96 well plate coated with a mouse anti-rabbit IgG, tracer
prostaglandin acetylcholine esterase and primary antibody (mouse anti PGE2) were added and
the plates were kept for overnight. Thereafter, the reaction mixtures were removed and the
wells were washed with 10 mM potassium phosphate buffer containing 0.05% Tween 20.
Ellman’s reagent (200 μL) was then added to all the wells and were incubated at room
temperature for 60 min, until the control wells gave an OD = 0.3-0.8 at 412 nm. A standard
curve with PGE2 was generated from the same plate, which was used to calculate the PGE2
levels produced in the presence of test compounds. The results were expressed as percentage
relative to control (solvent treated samples). All the experiments were performed in triplicate
and values normally agreed within 10% [38].
3.4 Molecular docking
Docking studies were performed with the LibDock protocol under the receptor-ligand
interaction section in Discovery Studio® 3.1 (Accelrys, Inc., San Diego, CA, USA). All of the
3D structures of the compounds were built with ChemBioOffice® 2008 (Perkin Elmer, Inc.,
Waltham, MA, USA). The protein crystal structure of the inhibitor-bound COX-1/COX-2 was
retrieved from the Brookhaven Protein Data Bank (PDB codes: 6COX and 3KK6). Both
proteins were pre-treated before the docking. Hydrogen atoms were added to the protein
structure, and all ionisable residues were set at their default protonation of pH 7.4 while the
ligands were prepared and minimized. For compound 22, the crystal structure obtained from
X-ray crystallographic analysis was used as the starting conformation for the docking study.
During the docking process, the receptor was held rigid while the ligands were allowed to flex
during the refinement. Number of polar or apolar receptor hotspots for conformer matching

starting were set at 500 with the docking tolerance 0.25 Å. The conformations generate of the
ligands was set 500 within this relative energy threshold 20 kcal/mol.
3.5 Pharmacophore modeling
Pharmacopore modelling was performed by Discovery Studio® 3.1 with the auto
pharmacophore generation based on bioactive conformation protocol. Pharmacophore models
were generated based on the Catalyst® protocol. The ligand feature types that were considered
in this step included Hydrogen bond acceptors (HBAs), Hydrogen bond donor (HBDs),
hydrophobic features (HY), negative ionizable features (NI), positive ionizable features (PI)
and rings aromatic (RA).
4. Conclusion
A hypothetical model of binding interactions between compound 22 bearing a chromone
scaffold structure and human recombinant COX-2 protein is proposed in this study. With the
successful structural elucidation of compound 22 using simple spectroscopic tools followed
by the retrieved atomic coordinates from X-ray diffraction analysis and biological results, it
can be concluded that the compound inhibited the PGE2 synthesis by selectively binds to
COX-2 over COX-1 protein. The major factor that contributes to the selectivity mainly due to
the presence of chromone ring and bulky dimethoxyl groups as predicted by the molecular
modeling programs (Fig. Graphical Abstract). The important information obtained from
this study will be utilized in the next design of more potent and selective COX-2
inhibitors.
Acknowledgments
This work was financially supported by Science Fund (02-01-02-SF00665), Ministry of
Science, Technology & Innovation, Malaysia. Authors also thank University Kebangsaan
Malaysia (UKM) for the funds provided under the Research University Grant (GUP) (UKM-
GUP02011-014) and UKM-DIP-2012-11.

Supplementary Material
Supplementary data associated with this article can be found, in the online version. These
data include MOLfiles and InChIKeys of the most important compounds described in this
article
References
[1] D.L. Simmons, R.M. Botting, T. Hla, Pharmacological Reviews, 56 (2004) 387-437.
[2] G. Singh, D. Ramey, D. Morfeld, H. Shi, H.T. Hatoum, J.F. Fries, Archives of Internal Medicine, 156
(1996) 1530-1536.
[3] F.K.L. Chan, Nat Clin Pract Gastroenterol Hepatol, 3 (2006) 563-573.
[4] A.A. Romanovsky, Am J Physiol Regul Integr Comp Physiol, 288 (2005) R1098-1099.
[5] T. Hla, K. Neilson, Proceedings of the National Academy of Sciences, 89 (1992) 7384-7388.
[6] K.R. Gans, W. Galbraith, R.J. Roman, S.B. Haber, J.S. Kerr, W.K. Schmidt, C. Smith, W.E. Hewes, N.R.
Ackerman, Journal of Pharmacology and Experimental Therapeutics, 254 (1990) 180-187.
[7] G.A. FitzGerald, New England Journal of Medicine, 351 (2004) 1709-1711.
[8] J.A. Mitchell, T.D. Warner, Nat Rev Drug Discov, 5 (2006) 75-86.
[9] D. Bishop-Bailey, J.A. Mitchell, T.D. Warner, Arteriosclerosis, Thrombosis, and Vascular Biology, 26
(2006) 956-958.
[10] R.S. Bresalier, R.S. Sandler, H. Quan, J.A. Bolognese, B. Oxenius, K. Horgan, C. Lines, R. Riddell, D.
Morton, A. Lanas, M.A. Konstam, J.A. Baron, New England Journal of Medicine, 352 (2005) 1092-1102.
[11] J.-M. Dogné, C.T. Supuran, D. Pratico, Journal of Medicinal Chemistry, 48 (2005) 2251-2257.
[12] R.J. Flower, Nat Rev Drug Discov, 2 (2003) 179-191.
[13] R.G. Kurumbail, A.M. Stevens, J.K. Gierse, J.J. McDonald, R.A. Stegeman, J.Y. Pak, D. Gildehaus, J.M.
iyashiro, T.D. Penning, K. Seibert, P.C. Isakson, W.C. Stallings, Nature, 384 (1996) 644-648.
[14] D. Riendeau, M.D. Percival, S. Boyce, C. Brideau, S. Charleson, W. Cromlish, D. Ethier, J. Evans, J.P.
Falgueyret, A.W. Ford-Hutchinson, R. Gordon, G. Greig, M. Gresser, J. Guay, S. Kargman, S. Léger, J.A.
Mancini, G. O'Neill, M. Ouellet, I.W. Rodger, M. Thérien, Z. Wang, J.K. Webb, E. Wong, L. Xu, R.N.
Young, R. Zamboni, P. Prasit, C.C. Chan, British Journal of Pharmacology, 121 (1997) 105-117.
[15] P. Patrignani, M.L. Capone, S. Tacconelli, Expert Opinion on Pharmacotherapy, 4 (2003) 265-284.
[16] K.J. Kim, M.J. Choi, J.-S. Shin, M. Kim, H.-E. Choi, S.M. Kang, J.H. Jin, K.-T. Lee, J.Y. Lee,
Bioorganic & Medicinal Chemistry Letters, 24 (2014) 1958-1962.
[17] T. Inaba, K. Tanaka, R. Takeno, H. Nagaki, C. Yoshida, S. Takano, Chemical & pharmaceutical bulletin,
48 (2000) 131-139.
[18] Y. Aikawa, M. Yamamoto, T. Yamamoto, K. Morimoto, K. Tanaka, Inflammation Research, 51 (2002)
188-194.
[19] K. Tanaka, T. Yamamoto, Y. Aikawa, K. Kizawa, K. Muramoto, H. Matsuno, A. Muraguchi,
Rheumatology, 42 (2003) 1365-1371.

[20] L.-j. Lu, C.-d. Bao, M. Dai, J.-l. Teng, W. Fan, F. Du, N.-p. Yang, Y.-h. Zhao, Z.-w. Chen, J.-h. Xu, P.-g.
He, H.-x. Wu, Y. Tao, M.-j. Zhang, X.-h. Han, X.-f. Li, J.-r. Gu, J.-h. Li, H. Yu, Arthritis Care &
Research, 61 (2009) 979-987.
[21] R. Gautam, S.M. Jachak, V. Kumar, C.G. Mohan, Bioorganic & Medicinal Chemistry Letters, 21 (2011)
1612-1616.
[22] R. Gautam, A. Srivastava, S.M. Jachak, A. Saklani, Fitoterapia, 81 (2010) 45-49.
[23] T.T. Dao, Y.S. Chi, J. Kim, H.P. Kim, S. Kim, H. Park, Bioorganic & Medicinal Chemistry Letters, 14
(2004) 1165-1167.
[24] K. Rullah, M.F.F. Mohd Aluwi, B.M. Yamin, M.N. Abdul Bahari, L.S. Wei, S. Ahmad, F. Abas, N.H.
Ismail, I. Jantan, L.K. Wai, Bioorganic & Medicinal Chemistry Letters.
[25] J.-T. Li, W.-Z. Yang, S.-X. Wang, S.-H. Li, T.-S. Li, Ultrasonics Sonochemistry, 9 (2002) 237-239.
[26] J.-T. Li, S.-X. Wang, G.-F. Chen, T.-S. Li, Current Organic Synthesis, 2 (2005) 415-436.
[27] K. Jakhar, J. Makrandi, Indian Journal of Chemistry, 51 (2012) 770-773.
[28] Y. Hyup Joo, J. Kwan Kim, S.-H. Kang, Synthetic Communications, 32 (2002) 1653-1658.
[29] Y.H. Joo, J.K. Kim, Synthetic Communications, 28 (1998) 4287-4293.
[30] J. Quintin, C. Roullier, S. Thoret, G. Lewin, Tetrahedron, 62 (2006) 4038-4051.
[31] C. Dyrager, L.N. Möllers, L.K. Kjäll, J.P. Alao, P. Dinér, F.K. Wallner, P. Sunnerhagen, M. Grøtli,
Journal of Medicinal Chemistry, 54 (2011) 7427-7431.
[32] H.S. Rho, B.-S. Ko, H.K. Kim, Y.-S. Ju, Synthetic Communications, 32 (2002) 1303-1310.
[33] W. Han, C. Liu, Z.-L. Jin, Organic Letters, 9 (2007) 4005-4007.
[34] M.A.A. El-Sayed, N.I. Abdel-Aziz, A.A.M. Abdel-Aziz, A.S. El-Azab, K.E.H. ElTahir, Bioorganic &
Medicinal Chemistry, 20 (2012) 3306-3316.
[35] J. Kaur, A. Bhardwaj, Z. Huang, E.E. Knaus, Bioorganic & Medicinal Chemistry Letters, 22 (2012) 2154-
2159.
[36] G. Rimon, R.S. Sidhu, D.A. Lauver, J.Y. Lee, N.P. Sharma, C. Yuan, R.A. Frieler, R.C. Trievel, B.R.
Lucchesi, W.L. Smith, Proceedings of the National Academy of Sciences, 107 (2010) 28-33.
[37] C. Jegerschöld, S.-C. Pawelzik, P. Purhonen, P. Bhakat, K.R. Gheorghe, N. Gyobu, K. Mitsuoka, R.
Morgenstern, P.-J. Jakobsson, H. Hebert, Proceedings of the National Academy of Sciences, 105 (2008)
11110-11115.
[38] S.N.A. Bukhari, G. Lauro, I. Jantan, G. Bifulco, M.W. Amjad, Bioorganic & Medicinal Chemistry, 22
(2014) 4151-4161.

Figure(s)
Figure and Scheme
Figure 1.Selected structures of COX-2 inhibitors.

Figure 2. Replacement of the central core structures of known COX-2 inhibitors with a
chromone ring.

O
O
O
O
O
O
X
Suzuki
Halogenation
coupling
Cyclization
O
O
CH3
OH
O
H
OH
Claisen-Schmidt
condensation
Figure 3. Retrosynthetic analysis for the synthesis 2,3-diaryl benzopyran.

1
Figure 4.Comparison H NMR spectrum of compounds 19-22 (500 MHz, CDCl3).

Figure 5. HMBC correlations of compound 22.

Figure 6. ORTEP diagram of compound 22 with thermal ellipsoids drawn at 50%
probability.

Figure 7. Crystal packing of the 22 complex viewed down the b-axis, the dashed lines (---)
show C–H···O(3H) hydrogen bonding.

Figure 8. Effects of 22 on PGE2 production in LPS-stimulated RAW 264.7 cells. The cells
were co-incubated with LPS (5 µg/mL) and different concentrations of 22 ranging from 0.78
to 50 µM. The supernatants were then collected for the measurement of PGE2 using EIA kit.
Compound 22 significantly inhibited PGE2 levels in LPS-stimulated macrophages. The
values are expressed as the means
S.D of three individual samples. ***P<0.001 as
compared with the LPS-treated macrophages; significant differences between groups were
determined using one-way ANOVA test followed by a Dunnett’s multiple comparison test.
#P<0.05 solvent control compared with the LPS-treated cells; significant difference was
determined using unpaired student’s t-test (n.s is non- significant)

SC-558
Compound 22
Compound 22
(Docked conformation A)
(Docked conformation B)
a
b
c
Figure 9. Plausible important binding interactions between SC-558 or compound 22 and the
muCOX-2 binding site depicted in (a) 2D diagram, (b) 3D diagram, and (c) protein
hydrophobicity surface. The atom colouring for SC-558 and compound 22 is in the
following: carbons in grey, oxygen in red, nitrogen in blue, and hydrogen in white. Hydrogen
atoms have been omitted in the 2D diagram for clarity. The green line indicates the classical
and non-classical hydrogen-bonding interactions, while the orange line indicates the π-
interaction, with distance indicated in angstrom, Å. The surface is colored by the
hydrophobicity of the protein residues, from blue for hydrophilic to brown for hydrophobic.

c
a
b
Figure 10. Plausible important binding interactions between compound 22 and the ovCOX-1
binding site depicted in (a) 2D diagram, (b) 3D diagram, and (c) protein hydrophobicity
surface. The atom colouring for compound 22 is in the following: carbons in grey, oxygen in
red, nitrogen in blue, and hydrogen in white. Hydrogen atoms have been omitted in the 2D
diagram for clarity. The green line indicates the classical and non-classical hydrogen-bonding
interactions, while the orange line indicates the π-interaction, with distance indicated in
angstrom, Å. The surface is colored by the hydrophobicity of the protein residues, from blue
for hydrophilic to brown for hydrophobic.

c
b
a
Figure 11. (A) Generation of common feature pharmacophore model based on the structure
of known COX-2 inhibitors (compounds 1-9). (B) Interfeature distance (Å) representation of
Hypo1. Features are color-coded as follows: cyan for hydrophobic (HY), green for hydrogen-
bond acceptor (HBA), and orange for ring aromatic (RA). (C) Compound 22 is superposed
on the pharmacophoric model.

Scheme
Scheme 1. Reagents and condition of synthesis: (a) KOH 40%, EtOH, (i) sonicate 1 h at 40ºC
(ii) stir at room temperature for overnight; (b) I2, DMSO, reflux 1 h; (c) Bu4NBr/PhI(OAc)2,
CH2Cl2, rt 8 h; (d) 4-Fluorophenylboronic acid, Pd(OAc)2, PEG 400, K2CO3, 60ºC 20 h.