Discovery of 3-(4-bromophenyl)-6-nitrobenzo[1.3.2]dithiazolium ylide 1,1-dioxide as a novel dual cyclooxygenase/5-lipoxygenase inhibitor that also inhibits tumor necrosis factor-α production

Article abstract of DOI:10.1016/j.bmc.2009.12.008

In the present study we have discovered compound 1, a benzo[1.3.2]dithiazolium ylide-based compound, as a new prototype dual inhibitor of cyclooxygenase (COX) and 5-lipoxygenase (5-LOX). Compound 1 was initially discovered as a COX-2 inhibitor, resulting indirectly from the COX-2 structure-based virtual screening that identified compound 2 as a virtual hit. Compounds 1 and 2 inhibited COX-1 and COX-2 in mouse macrophages with IC₅₀ in the range of 1.5-18.1 μM. Both compounds 1 and 2 were also found to be potent inhibitors of human 5-LOX (IC₅₀ = 1.22 and 0.47 μM, respectively). Interestingly, compound 1 also had an inhibitory effect on tumor necrosis factor-α (TNF-α) production (IC₅₀ = 0.44 μM), which was not observed with compound 2. Docking studies suggested the (S)-enantiomer of 1 as the biologically active isomer that binds to COX-2. Being a cytokine-suppressive dual COX/5-LOX inhibitor, compound 1 may represent a useful lead structure for the development of advantageous new anti-inflammatory agents.

Full text of DOI:10.1016/j.bmc.2009.12.008

Bioorganic & Medicinal Chemistry 18 (2010) 597–604
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Discovery of 3-(4-bromophenyl)-6-nitrobenzo[1.3.2]dithiazolium ylide
1,1-dioxide as a novel dual cyclooxygenase/5-lipoxygenase inhibitor
that also inhibits tumor necrosis factor-a production
Chien-Shu Chen a,b, Chen-Ming Tan a, , Chiung-Hua Huang a, , Ling-Chu Chang c, Jih-Pyang Wang d,
Fong-Chi Cheng e, Ji-Wang Chern a,f,
*
a School of Pharmacy, National Taiwan University, Taipei 100, Taiwan
b School of Pharmacy, China Medical University, Taichung 404, Taiwan
c Institute of Medicine, Chung Shan Medical University, Taichung 402, Taiwan
d Department of Education and Research, Taichung Veterans General Hospital, Taichung 407, Taiwan
e MDS Pharma Services-Taiwan, Taipei 112, Taiwan
f Department of Life Science, National Taiwan University, Taipei 106, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
In the present study we have discovered compound 1, a benzo[1.3.2]dithiazolium ylide-based compound,
as a new prototype dual inhibitor of cyclooxygenase (COX) and 5-lipoxygenase (5-LOX). Compound 1 was
initially discovered as a COX-2 inhibitor, resulting indirectly from the COX-2 structure-based virtual
screening that identified compound 2 as a virtual hit. Compounds 1 and 2 inhibited COX-1 and COX-2
Received 7 November 2009
Revised 26 November 2009
Accepted 2 December 2009
Available online 6 December 2009
in mouse macrophages with IC50 in the range of 1.5–18.1
to be potent inhibitors of human 5-LOX (IC50 = 1.22 and 0.47
1 also had an inhibitory effect on tumor necrosis factor- (TNF-
l
M. Both compounds 1 and 2 were also found
M, respectively). Interestingly, compound
) production (IC50 = 0.44 M), which was
l
a
Keywords:
Virtual screening
Benzo[1.3.2]dithiazolium ylide 1,1-dioxide
Cyclooxygenase
Lipoxygenase
a
l
not observed with compound 2. Docking studies suggested the (S)-enantiomer of 1 as the biologically
active isomer that binds to COX-2. Being a cytokine-suppressive dual COX/5-LOX inhibitor, compound
1 may represent a useful lead structure for the development of advantageous new anti-inflammatory
agents.
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
duction by selective COX-2 inhibitors might account for their ad-
verse cardiovascular effects.8
Nonsteroidal anti-inflammatory drugs (NSAIDs) such as aspirin,
ibuprofen, and indomethacin are widely used for the treatment of
pain and inflammation. These drugs block arachidonic acid metab-
olism by inhibiting the cyclooxygenase (COX) enzymes. The COX
enzymes, COX-1 (a constitutive isoform) and COX-2 (an inducible
isoform), are responsible for cytoprotective effects and for inflam-
matory effects, respectively.1 Classical NSAIDs may cause gastroin-
testinal (GI) side effects due to indiscriminate inhibition of both
COX isoforms.2 In recent years, selective COX-2 inhibitors, such
as celecoxib,3 rofecoxib,4 and valdecoxib5 (Chart 1), have been
developed as a new generation of NSAIDs with diminished GI side
effects. However, rofecoxib and valdecoxib were withdrawn from
the market because of an increased risk of cardiovascular compli-
cations (e.g., myocardial infarction and stroke).6,7 Because COX-2
In addition to the COX pathway, arachidonic acid is also metab-
olized via the lipoxygenase (LOX) pathway. For example, 5-lipoxy-
genase (5-LOX) is the key enzyme in the synthesis of leukotrienes
from arachidonic acid. Leukotrienes such as LTB4 are potent medi-
ators of inflammation and allergic reactions.9 Thus, 5-LOX inhibi-
tors represent potential therapeutic agents for the treatment of
inflammatory and allergic diseases.10 Recently, dual inhibitors of
COX and 5-LOX have emerged as attractive safe alternatives to
selective COX-2 inhibitors.11 It has been proposed that inhibition
of the COX enzymes by NSAIDs may shunt arachidonic acid metab-
olism to the 5-LOX pathway.12 This shunting mechanism leads to
increased formation of leukotrienes, which are likely to cause side
effects. Early studies suggest that leukotrienes may contribute to
the pathogenesis of NSAID-induced GI mucosal damage.13 Indeed,
increased synthesis of LTB4 has been found in the gastric mucosa
of patients during NSAID treatment.14 Therefore, dual COX/5-LOX
inhibitors can be expected to have a better GI safety profile. A num-
ber of dual COX/5-LOX inhibitors have been developed,11 including
tepoxalin,15 tebufelone,16 and licofelone17 (Chart 1). Tepoxalin is
approved for the treatment of osteoarthritis in dogs in the United
mediates the biosynthesis of prostacyclin,
a vasodilator and
inhibitor of platelet aggregation, the inhibition of prostacyclin pro-
* Corresponding author. Tel.: +886 2 2393 9462; fax: +886 2 2393 4221.
E-mail address: jwchern@ntu.edu.tw (J.-W. Chern).
These authors contributed equally to this work.
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.12.008

598
C.-S. Chen et al. / Bioorg. Med. Chem. 18 (2010) 597–604
H3C
O
N
CF3
O
O
N
N
CH3
H2NO2S
H3CO2S
H2NO2S
Celecoxib
Rofecoxib
Valdecoxib
H3C
Cl
OH
Cl
N
COOH
O
OH
N
N
N
CH3
CH3
HC
H3CO
O
Tepoxalin
Tebufelone
Licofelone
Chart 1. Representative examples of selective COX-2 inhibitors and dual COX/5-LOX inhibitors.
States. Licofelone is the most promising dual COX/5-LOX inhibitor
to date that has been evaluated in Phase III clinical trials and
shown to be effective in the treatment of osteoarthritis with re-
duced GI toxicity compared to classical NSAIDs.18
In this paper, we report the discovery of 3-(4-bromophenyl)-6-
nitrobenzo[1.3.2]dithiazolium ylide 1,1-dioxide (1, Chart 2) as a
novel dual COX/5-LOX inhibitor. Moreover, compound 1 was also
performed using the crystal structure of mouse COX-2 bound to
SC-558, a COX-2 selective inhibitor (PDB code 6COX).28 The SPECS
database (ꢀ76,000 commercially available compounds) was com-
putationally screened against the active site of COX-2 using the
DOCK 3.5 program.29 The 245 top-scoring compounds were visu-
ally inspected for the plausibility of their predicted binding modes,
and 17 virtual hits were selected for biological testing. Four of the
found to inhibit the production of tumor necrosis factor-
a
(TNF-
test compounds showed 21–33% inhibition at 10 lM against hu-
a
). The cytokine TNF- is one of the major mediators of inflamma-
a
man recombinant COX-2. An encouraging result was obtained with
a test compound (SPECS ID number AA-516/31409051, abbrevi-
ated here as AA-9051), which exhibited COX-2 inhibition with an
tion and has been clinically studied as a therapeutic target for
rheumatoid arthritis.19 Compound 1, showing multiple effects on
the inflammatory processes, may represent a useful lead com-
pound to develop advantageous new anti-inflammatory agents.
In this respect, two clinically evaluated dual COX/5-LOX inhibitors,
licofelone and S-2474, also displayed cytokine-modulating proper-
IC50 of 1.7
lM. AA-9051 also inhibited COX-1 derived from human
platelets with an IC50 of 6.7
l
M.
The test compound AA-9051 was purchased according to the
commercial ID number of the virtual hit compound 2 (Chart 2).
This virtual hit was very highly ranked (the fifth) from the DOCK
screening. Superimposition of the docking model of compound 2
(as S-form in the database) and the crystal structure of inhibitor
SC-558 bound to COX-2 is shown in Figure 1. The p-bromophenyl
group in both compounds was located in the same hydrophobic
pocket. The benzo group of compound 2 was overlaid well with
the N1-phenyl ring of SC-558. With respect to polar interactions,
ties on the production of TNF-a
and/or interleukin-1b (IL-1b).20,21
To our knowledge, few studies associated with this type of com-
pounds have been reported,22–26 and prior to this work the only re-
port on the biological activity of such compounds, including
compound 1, was described by Sang et al.26 This class of compounds
was identified as tyrosine kinase inhibitors of the human epidermal
growth factor receptor (HER) family (toward HER1, HER2, and
HER4). In addition, it should be noted that compound 1 can also be
named 1-(4-bromophenyl)-5-nitro-1k4-benzo[1,3,2]dithiazole 3,3-
dioxide, according to the Lambda Convention.
2. Results and discussion
Initially, a structure-based virtual screening approach was used
to search for new COX-2 inhibitors.27 The virtual screening was
Br
Br
Br
5
O2N
H2NO2S
S
S
S
S
S
S
N
N
N
O2N
6
O
O
O
O
O
O
1
2
3
Figure 1. Superimposition of the docking model of compound 2 (S-form, green) and
the crystal structure of inhibitor SC-558 (magenta) bound in the active site of COX-2.
Chart 2. 3-(4-Bromophenyl)benzo[1.3.2]dithiazolium ylide 1,1-dioxides.

C.-S. Chen et al. / Bioorg. Med. Chem. 18 (2010) 597–604
599
the SO2 group of compound 2 could form two hydrogen bonds with
Tyr355. On the other hand, the predicted position of the NO2 group
of compound 2 was not close enough to His90 for hydrogen bond
formation. However, the NO2 group would potentially be expected,
like the SO2 group of SC-558, to form a hydrogen bond with His90
based on the proximity of locations between these two groups and
the mobility of the enzyme structure. Taken together, virtual hit 2
became one of the selected compounds for biological testing.
A preliminary study of the structural modification of compound
2 was carried out by replacing the NO2 group with a SO2NH2 group,
as seen in SC-558 and celecoxib. The corresponding sulfonamide
analog (compound 3, Chart 2) showed only weak inhibition at
mistaken by the chemical supplier. The mass and 1H NMR analyses
of AA-9051 revealed that, compared with compound 2, this com-
pound had identical molecular weight and showed the same
NMR splitting patterns for the three protons of the benzo group.
Therefore, we assumed that AA-9051 might be the 6-NO2 regio-
isomer of compound 2, namely compound 1 (Chart 2). Because at-
tempts to produce suitable crystals of AA-9051 for X-ray
crystallography failed, we decided to synthesize compound 1 as
outlined in Scheme 2 to confirm the above assumption. As ex-
pected, the synthesized compound 1 proved to be identical to
AA-9051 based on TLC, mass, and 1H NMR analyses.
Compounds 1 and 2 were evaluated for their ability to inhibit
COX-1 and COX-2 using the mouse macrophage cell line RAW
264.7 (Table 2). The data showed that in a cellular environment
compound 1 exhibited low micromolar activity against COX-1
10 lM against human recombinant COX-2 (27% inhibition) and hu-
man platelet COX-1 (17% inhibition) (Table 1). Meanwhile, com-
pound 2 (we believed to be AA-9051) was also synthesized in
our laboratories. According to the synthetic rationale (Scheme 1),
such a route would exclusively produce the desired compound 2.
However, it was surprising that the TLC mobility and 1H NMR spec-
trum of the compound prepared were distinct from those of the
purchased compound AA-9051. To address this question, the struc-
ture of the synthesized compound was analyzed by X-ray crystal-
lography and confirmed to be compound 2 as a racemic mixture
(Fig. 2). Accordingly, we concluded that AA-9051 was not identical
to compound 2, and the structural identification of AA-9051 was
(IC50 = 2.1
lM) and COX-2 (IC50 = 4.4
lM). Compound 2 was also
a potent inhibitor of COX-1 (IC50 = 1.5
l
M); however, it was a four-
fold less potent inhibitor of COX-2 (IC50 = 18.1 lM) as compared to
compound 1. The results from cell-based assays indicated that both
compounds 1 and 2 are cell membrane permeable and stable in cell
culture. The ability of compound 1 to inhibit COX-2 was further as-
sessed in THP-1 cells (a human macrophage cell line). Encourag-
ingly, the anti-COX-2 activity of compound 1 in human THP-1
cells was nearly comparable to that of the reference compound
celecoxib (77% vs 83% inhibition at 10 lM).
The emerging trend towards the development of dual COX/5-
LOX inhibitors prompted us to test whether compounds 1 and 2
were able to inhibit 5-LOX. In the cell-based assay for human
Table 1
Inhibitory activity of compounds 1–3 on human recombinant COX-2 and human
platelet COX-1
Br
5
R1
R2
S
S
N
6
O
O
Compd
R1
R2
Inhibitory activity
Human recombinant COX-2
Human platelet
COX-1
1
2
3
H
NO2
SO2NH2
NO2
H
H
IC50 = 1.7
23% @ 10
27% @ 10
l
l
l
M
M
M
IC50 = 6.7
39% @ 10
17% @ 10 l
lM
lM
M
Figure 2. Crystal structures of the (R)- and (S)-enantiomers of compound 2.
O2N
I
O2N
O2N
I
a
b
NH2
SO2NH2
NH2
4
5
6
Br
S
Br
d
c
O2N
O2N
S
S
N
SO2NH2
O
O
7
2
Scheme 1. Synthesis of compound 2. Reagents and conditions: (a) I2, Ag2SO4, EtOH, room temp, 30 min, 95%; (b) (i) HCl/AcOH, NaNO2, CH3CN, <5 °C, 20 min; (ii) SO2, <7 °C,
40 min, then CuCl2, room temp, 16 h; (iii) NH4OH, room temp, 10 h, 22%; (c) 4-bromothiophenol, CuI, neocuproine, t-BuONa, toluene, reflux, 8 h, 55%; (d) Br2, MeOH/H2O,
room temp, 30 min, 56%.

600
C.-S. Chen et al. / Bioorg. Med. Chem. 18 (2010) 597–604
I
NH2
I
a
b
O2N
SO3H
O2N
SO2NH2
O2N
SO3Na
8
9
10
Br
S
Br
c
d
S
S
N
O2N
O2N
SO2NH2
O
O
11
1
Scheme 2. Synthesis of compound 1. Reagents and conditions: (a) (i) HCl, NaNO2, H2O, 0 °C, 30 min; (ii) KI, room temp, 1 h, 82%; (b) (i) SOCl2, DMF, reflux, 4 h; (ii) NH4OH,
room temp, 6 h, 60%; (c) 4-bromothiophenol, CuI, neocuproine, t-BuONa, toluene, reflux, 8 h, 79%; (d) Br2, EtOAc/MeOH/H2O, room temp, 20 min, 54%.
Table 2
Inhibitory activity of compounds 1 and 2 on mouse COXs and human 5-LOX
Br
5
R1
R2
S
S
N
6
O
O
Compd
R1
R2
IC50
Mouse macrophagea
(l
M)
Human leukocyteb
COX-1
COX-2
5-LOX
1
2
H
NO2
NO2
H
2.1
1.5
63.7
56.7
4.4
18.1
0.81
0.21
1.22
0.47
—
NS-398c
LM-4108c
—
a
b
c
Mouse macrophage cell line RAW 264.7.
Human peripheral blood mononuclear leukocytes (PBML).
COX-2-selective reference compounds used in the assays.
5-LOX in peripheral blood mononuclear leukocytes (PBML), com-
pound 1 exhibited 5-LOX inhibitory activity with an IC50 value of
1.22
lM (Table 2). The 5-NO2 analog 2 was 2.6-fold more effective
in inhibiting 5-LOX (IC50 = 0.47
lM) than compound 1. Compound
1, but not compound 2, was considered a lead structure for future
optimization because this compound was shown to be a relatively
potent COX-2 inhibitor and possess an additional anti-inflamma-
tory pharmacological activity related to inhibition of TNF-
duction as discussed below.
a
pro-
Compounds 1 and 2 were tested for their effects on TNF-
a
pro-
duction from LPS-stimulated human THP-1 macrophages. Com-
pound 1 inhibited TNF- production by 92% at 10 M, whereas
a
l
compound 2 lacked such biological activity (Fig. 3A). As shown in
Figure 3B, there was a concentration-dependent suppression of
Figure 3. (A) Effect of compounds 1 and 2, and celecoxib (10
production in LPS-stimulated THP-1 cells. The positive control, dexamethasone, was
tested at 1 M, showing 89% inhibition of TNF- production. Data represent the
mean SD of duplicate determinations. (B) Concentration-dependent inhibition of
TNF- production by compound 1 in THP-1 cells. Data represent the mean SD of
lM) on TNF-a
TNF-
COX-2 selective drug celecoxib, also tested in this assay, showed
no effect on TNF- production. It has been pointed out that NSAIDs,
generally considered anti-inflammatory, might also be pro-inflam-
matory by increasing TNF-
levels.30 For example, aspirin,31
ibuprofen,32 indomethacin,33 and rofecoxib34 have been demon-
strated to up-regulate TNF- production. Moreover, it has been
suggested that TNF- plays a critical role in NSAID-induced gastric
mucosal injury.33 Because compound
could inhibit TNF-
a production by compound 1 with an IC50 of 0.44 lM. The
l
a
a
a
duplicate determinations.
a
production, it not only would be devoid of these potential side
effects but was also expected to have greater anti-inflammatory
efficacy through combination with COX/5-LOX inhibition. The inhi-
a
a
1
a
bition mechanism of compound 1 on TNF-a production remains to

C.-S. Chen et al. / Bioorg. Med. Chem. 18 (2010) 597–604
601
be elucidated, and a possibility through inhibition of p38 kinase35
could be ruled out based on our study, showing no inhibition of
Arg513 (3.1 Å). The NO2 group in compound 1 was shown to be
important for inhibiting COX-2 because the corresponding 6-desni-
p38
a
kinase by compound 1 at 10
l
M.
tro analog displayed no inhibition at 10
binant COX-2.
lM against human recom-
To predict the binding mode of compound 1 in the COX-2 active
site, docking studies were carried out using the program GOLD
(version 3.1.1).36 The GOLD program can perform flexible ligand
docking simulations and thus may allow better prediction of the
binding mode for a compound when compared with the DOCK 3.5
program used in the virtual screening, which performs rigid-body
ligand docking using only a single conformation. In addition, it
was possible that differences in the conformations of active site
residues might be present in the different crystal structures of
COX-2 (free and in complex with various inhibitors). We therefore
attempted to utilize several reported crystal structures of COX-2
for docking simulations in order to find one that would create
the most favorable interactions with compound 1. The four crystal
structures used in the docking studies include mouse COX-2 unli-
ganded (PDB code 5COX) and complexed with inhibitors flurbipro-
fen, indomethacin, and SC-558 (PDB codes 3PGH, 4COX, and 6COX,
respectively).28
We subsequently carried out the docking study of (R)-1 enan-
tiomer with the crystal structure 4COX. The docking solution of
(R)-1 had a lower GOLD fitness score than that of (S)-1 (42.50 vs
48.20). The two enantiomeric forms were predicted to interact
with COX-2 in a completely different manner. It was observed that
instead of the p-bromophenyl moiety, the bicyclic heterocycle of
(R)-1 occupied the hydrophobic pocket and formed two hydrogen
bonds via the SO2 group with Tyr385 and Ser530 (Fig. 5). However,
the polar NO2 group of (R)-1 appeared to be unfavorably located
because it did not make any hydrogen bonds and had unfavorable
contacts with the nonpolar side chains of surrounding residues.
These results contradicted the importance of the role of the NO2
group in compound 1 as an inhibitor of COX-2. Taken together, it
was hypothesized that (S)-1 would be the biologically active enan-
tiomer against COX-2.
Because compound 1 contains a chiral sulfur atom and exists as
a racemate, we sought to predict which enantiomer might be the
biologically active form through the docking studies. Docking stud-
ies were first performed on the (S)-enantiomeric form of com-
pound 1 ((S)-1) in the active site of the four COX-2 crystal
structures because the (S)-enantiomeric form of compound 2 was
initially identified in the virtual screening. The docking solutions
of (S)-1 obtained using the crystal structures 3PGH, 4COX, 5COX,
and 6COX had GOLD fitness scores of 40.82, 48.20, 46.55, and
44.04, respectively. The predicted binding orientations of (S)-1
were similar when docked into the crystal structures 4COX,
5COX, and 6COX. On the other hand, the docking solution obtained
using the crystal structure 3PGH was eliminated because (S)-1 was
located outside the ligand-binding site. The docking solution for
(S)-1 that demonstrated the highest GOLD fitness score (48.20 for
4COX) was chosen to represent the predicted binding mode to
COX-2 (Fig. 4). As indicated in Figure 4, (S)-1 formed two hydrogen
bonds via the SO2 group with Arg120 (3.4 Å) and Tyr355 (2.5 Å).
Similar interactions were observed from the carboxylate group of
indomethacin with the COX-2 enzyme (4COX structure). In addi-
tion, the NO2 group of (S)-1 was inserted into the side pocket of
COX-2 and formed two hydrogen bonds with His90 (3.0 Å) and
3. Conclusions
In conclusion, through structure-based virtual screening, struc-
tural identification, and biological evaluations, we ultimately
determined that compound 1 is a novel dual COX/5-LOX inhibitor
as well as an inhibitor of TNF-
a production. Compound 1, show-
ing multiple actions on the inflammatory pathways, may repre-
sent a useful lead compound for the development of new anti-
inflammatory agents that would have a better GI safety profile
and a broad spectrum of anti-inflammatory activities. Further
studies on the structural modifications of this compound are cur-
rently underway.
4. Experimental
4.1. Virtual screening
The X-ray crystal structure of mouse COX-2 (PDB code 6COX) was
retrieved from the Protein Data Bank (http://www.rcsb.org/pdb) to
prepare the target site for docking calculations. The B-chain struc-
ture of protein and small molecules were removed. The remaining
Figure 4. Predicted binding mode of the (S)-enantiomer of compound 1 in the
active site of COX-2. The dashed lines indicate hydrogen bonds. The hydrophobic
pocket is represented as a molecular surface (magenta).
Figure 5. Docking model of the (R)-enantiomer of compound 1 in the active site of
COX-2. The dashed lines indicate hydrogen bonds. The hydrophobic pocket is
represented as a molecular surface (magenta).

602
C.-S. Chen et al. / Bioorg. Med. Chem. 18 (2010) 597–604
A-chain protein structure was used for the virtual screening per-
formed with the DOCK 3.5 program. The active site region of COX-2
was specified as the target site for ligand docking in virtual screen-
ing. Briefly, a molecular surface around the target site was generated
with the MS program using a 1.4 Å probe radius, and this surface was
used to generate with the SPHGEN program 44 overlapping spheres to
fill the target site. A grid box enclosing the target site was created for
grid calculations with dimensions 21.2 Â 13.8 Â 18.5 Å. The force
filed scoring grids were calculated with the CHEMGRID program using
default AMBER parameters, a distance-dependent dielectric con-
stant of 4r, close contact limits of 2.3 and 2.8 Å for receptor polar
and nonpolar atoms, respectively, a nonbonded cutoff of 10 Å, and
a grid spacing of 0.3 Å. The database for virtual screening was the
SPECS compound collection (2002 release, http://www.specs.net),
which included approximately 76,000 compounds. The three-
dimensional structures of the SPECS compounds were generated
using the CONCORD program (Tripos, Inc., St. Louis, MO). The DOCK 3.5
program performed rigid-body docking using a distance-matching
algorithm. The matching parameters used to run virtual screening
were set as follows: distance_tolerance = 1.5, nodes_maximum = 8,
nodes_minimum = 4, ligand_binsize = 0.4, ligand_overlap = 0.1,
receptor_binsize = 0.4, receptor_overlap = 0.1, and bump_maxi-
mum = 0. TheSPECS databasewas computationally screenedagainst
the active site of COX-2 using the force field scoring function based
on interaction energy. Virtual screening was performed on a Silicon
Graphics Octane workstation with dual 270 MHz MIPS R12000
processors.
For compound selection, the docking models of the 245 top-
ranked compounds (energy scores ranging from À69.2 to À37.8
kcal/mol) were visually inspected using the software Insight II
(Accelrys Inc., San Diego, CA). Together with consideration of chem-
ical diversity, the selection of compounds was assisted by visual
analysis of the docking models with respect to shape fitting and
hydrogen-bonding and hydrophobic interactions. Finally, we se-
lected 17 compounds for COX-2 inhibition assay. The compounds
for testing were purchased from the SPECS company. The SPECS ID
numbers for compounds are listed as follows: AA-516/31409051
(compound 1), AG-205/37204019, AC-907/34122064, AG-205/121
45002, AG-205/36628044, AK-968/12686132, AG-205/12368031,
AE-848/34434063, AG-205/33120044, AG-690/34549025, AK-968/
11109006, AF-399/36911008, AF-615/30368003, AG-205/33680
50, AK-660/36747022, AG-227/37195079, AG-690/34649013.
3PGH, 4COX, 5COX, and 6COX had GOLD fitness scores of 40.82,
48.20, 46.55, and 44.04, respectively. The docking solution for
(S)-1 that demonstrated the highest score (48.20 for 4COX) was
chosen to represent the predicted binding mode to COX-2
(Fig. 4). The crystal structure 4COX was then used in the docking
study of (R)-1 enantiomer. The top-ranked docking model of (R)-
1 (GOLD fitness score: 42.50) is shown in Figure 5.
4.3. Chemistry
4.3.1. Synthesis of 3-(4-bromophenyl)-5-nitrobenzo[1.3.2]dithi-
azolium ylide 1,1-dioxide (2)
4.3.1.1. 2-Iodo-4-nitroaniline (5). 4-Nitroaniline (4, 5.43 g) was
added to a mixture of iodine (10 g) and Ag2SO4 (12.24 g) in EtOH
(100 mL), and the mixture was stirred at room temperature for
30 min. The reaction mixture was filtered, and the filtrate was
evaporated under reduced pressure. The resulting residue was dis-
solved in CH2Cl2, washed with 5% NaOH and H2O, dried (MgSO4),
and evaporated to dryness to give 5 as a yellow solid (95% yield).
Mp 106–107 °C (lit. 107 °C37). 1H NMR (200 MHz, acetone-d6) d
8.44 (d, J = 2.4 Hz, 1H, ArH), 7.99 (dd, J = 2.4, 9 Hz, 1H, ArH), 6.88
(d, J = 9 Hz, 1H, ArH), 6.13 (s, 2H, NH2). 13C NMR (50 MHz, ace-
tone-d6) d 154.60, 138.47, 135.66, 125.91, 112.77, 79.58.
4.3.1.2. 2-Iodo-4-nitrobenzenesulfonamide (6). To a solution of
compound 5 (5.28 g, 20 mmol) in CH3CN (150 mL) cooled to a tem-
perature <5 °C in an ice bath were added concd HCl (16 mL) and
AcOH (8 mL). A solution of NaNO2 (1.66 g) in H2O (3 mL) was then
added slowly via syringe to the acidic solution. After complete
addition, the mixture was stirred for 20 min. SO2 gas was bubbled
into the mixture at a temperature <7 °C for 40 min. To the resulting
mixture was then added a solution of anhydrous CuCl2 (3.4 g,
25 mmol) in H2O (3 mL). When the mixture became deep brown,
the ice bath was removed, and stirring was continued at room tem-
perature for 16 h. The reaction mixture was concentrated to a
small volume and poured into ice water. The yellow solid was col-
lected by filtration to give the sulfonyl chloride derivative. The sul-
fonyl chloride compound was treated with excess ammonia water
in an ice bath, and the mixture was allowed to stir at room temper-
ature for 10 h. The mixture was concentrated to a small volume,
extracted with EtOAc and CHCl3, and evaporated to dryness. Puri-
fication of the crude product by flash column chromatography
(hexane) gave 6 in 22% yield. Mp 203 °C. 1H NMR (400 MHz, ace-
tone-d6) d 8.84 (d, J = 2.4 Hz, 1H, ArH), 8.42 (dd, J = 8.8, 2.4 Hz,
1H, ArH), 8.36 (d, J = 8.8 Hz, 1H, ArH), 7.04 (s, 2H, SO2NH2). 13C
NMR (100 MHz, acetone-d6) d 152.03, 149.72, 137.56, 130.64,
124.12, 92.38. MS (EI) m/z = 328 [M]+. Anal. Calcd for C6H5IN2O4S:
C, 21.97; H, 1.54; N, 8.54. Found: C, 21.88; H, 1.46; N, 8.55.
4.2. Molecular docking studies
Four crystal structures of mouse COX-2 (PDB codes: 3PGH,
4COX, 5COX, and 6COX) were used for docking studies of com-
pound 1. For each crystal structure, the A-chain protein structure
with hydrogen atoms added was utilized in the docking experi-
ments. The 3-D structures of (S)-1 and (R)-1 enantiomers were
built and optimized by energy minimization using the Tripos force
field in the software package SYBYL 6.5 (Tripos, Inc., St. Louis, MO).
Docking experiments were performed using the GOLD program (ver-
sion 3.1.1) on a Silicon Graphics Octane workstation with dual
270 MHz MIPS R12000 processors. The GOLD program utilizes a ge-
netic algorithm (GA) to perform flexible ligand docking simula-
tions. In the present study, for each of the 30 independent GA
runs, a maximum number of 100,000 GA operations were per-
formed on a single population of 100 individuals. Operator weights
for crossover, mutation, and migration were set to 95, 95, and 10,
respectively. Default cutoff values of 2.5 Å for hydrogen bonds
and 4.0 Å for van der Waals distance were used. The GoldScore fit-
ness function was applied for scoring the docking poses using
EXTERNAL_ENERGY_WT = 1.0. Firstly, (S)-1 enantiomer was stud-
ied for docking into each active site of the four COX-2 crystal struc-
tures. The docking solutions obtained using the crystal structures
4.3.1.3. 2-(4-Bromophenylsulfanyl)-4-nitrobenzenesulfona-
mide (7). To a mixture of aryl iodide 6 (0.2 g), CuI (0.014 g), neo-
cuproine (0.016 g), and t-BuONa (0.126 g) in toluene was added 4-
bromothiophenol (0.14 g). The mixture was refluxed under N2 for
8 h. The reaction mixture was neutralized with dil HCl and then
evaporated under reduced pressure. The resulting residue was ex-
tracted with EtOAc (3 Â 15 mL), and the extracts were evaporated
to dryness. The crude product was purified by flash column chro-
matography (hexane/EtOAc 1:1) to give 7 in 55% yield. 1H NMR
(200 MHz, acetone-d6) d 8.20 (m, 2H, ArH), 7.83 (d, J = 2 Hz, 1H,
ArH), 7.72 (m, 2 H, ArH), 7.60 (m, 2H, ArH), 7.12 (s, 2H, SO2NH2).
13C NMR (50 MHz, acetone-d6) d 150.48, 146.36, 140.17, 137.01,
133.99, 131.65, 130.43, 125.08, 124.53, 121.34.
4.3.1.4. 3-(4-Bromophenyl)-5-nitrobenzo[1.3.2]dithiazolium ylide
1,1-dioxide (2). To a solution of 7 (0.06 g) in MeOH/H2O (6 mL/
2 mL) was added bromine (0.015 mL, 0.045 g). The mixture was

C.-S. Chen et al. / Bioorg. Med. Chem. 18 (2010) 597–604
603
stirred at room temperature for 30 min. The precipitated solid was
collected by filtration, washed with H2O, and recrystallized from ace-
tone/hexane to give product 2 in 56% yield. Mp 294–296 °C. 1H NMR
(400 MHz, acetone-d6) d 9.09 (d, J = 1.84 Hz, 1H, ArH), 8.77 (dd, J = 8.4,
1.84 Hz, 1H, ArH), 8.32 (d, J = 8.4 Hz, 1H, ArH), 7.97 (dd, J = 6.7, 2 Hz,
2H, ArH), d 7.82 (dd, J = 6.7, 2 Hz, 2H, ArH). 13C NMR (100 MHz, ace-
tone-d6) d 151.84, 141.51, 137.73, 137.53, 137.08, 134.11, 130.28,
128.43, 124.97, 122.75. MS (EI) m/z = 388 [M+2]+. Anal. Calcd for
C12H7BrN2O4S2: C, 37.22; H, 1.82; N, 7.23. Found: C, 37.30; H, 2.44;
N, 7.13.
(dd, J = 8.6, 1.7 Hz, 1H, ArH), 8.48 (d, J = 8.6 Hz, 1H, ArH), 7.95 (d,
J = 8.7 Hz, 2H, ArH), 7.82 (d, J = 8.7 Hz, 2H, ArH). 13C NMR
(100 MHz, acetone-d6) d 140.90, 138.82, 137.81, 134.22, 130.16,
129.42, 128.52, 128.32, 119.14, 116.69. MS (EI) m/z = 388 [M+2]+.
Anal. Calcd for C12H7BrN2O4S2: C, 37.22; H, 1.82; N, 7.23. Found: C,
37.18; H, 2.08; N, 7.62.
4.4. Human COX-2 and COX-1 inhibition assays
The human COX inhibition assays were performed by MDS
Pharma Services (Taipei, Taiwan) using the test methods num-
bered 118010 and 116020 for COX-2 and COX-1, respectively.
The enzyme sources for COX-2 and COX-1 were human recombi-
nant COX-2 (expressed in insect Sf21 cells) and human platelets,
respectively.
4.3.2. Synthesis of 3-(4-bromophenyl)-6-nitrobenzo[1.3.2]dithi-
azolium ylide 1,1-dioxide (1)
4.3.2.1. 2-Iodo-5-nitrobenzenesulfonic acid (9). 2-Amino-5-
nitrobenzenesulfonic acid, sodium salt (8, 5 g) was dissolved in
H2O (150 mL). To this solution cooled in an ice bath were added
a solution of NaNO2 (1.72 g) in H2O (5 mL) and concd HCl (5 mL).
After stirring at 0 °C for 30 min, KI (4.15 g) was added to the mix-
ture. The ice bath was removed, and the mixture was allowed to
stir at room temperature for 1 h. The reaction mixture was concen-
trated to a small volume, and the solid was collected by filtration to
give 9 in 82% yield. 1H NMR (400 MHz, DMSO-d6) d 8.58 (d,
J = 2.8 Hz, 1H, ArH), 8.18 (d, J = 8.4 Hz, 1H, ArH), 7.81 (dd, J = 8.4,
2.8 Hz, 1H, ArH); 13C NMR (100 MHz, DMSO-d6) d 151.85, 146.88,
142.69, 123.97, 121.85, 103.00.
4.4.1. Human COX-2 inhibition assay
Test compounds were preincubated for 15 min at 37 °C with
purified human recombinant COX-2 (0.11 U) in Tris–HCl buffer
(pH 7.7) containing glutathione (1 mM), phenol (500
hematin (1 M). The enzymatic reaction was initiated by addition
of 0.3 M arachidonic acid. After 5 min incubation at 37 °C, the
lM), and
l
l
reaction was terminated by addition of 1 N HCl. Prostaglandin E2
(PGE2) production was quantified to measure COX-2 activity using
an enzyme immunoassay (EIA) kit (Amersham).
4.3.2.2. 2-Iodo-5-nitrobenzenesulfonamide (10). A mixture of 9
(3.5 g), thionyl chloride (20 mL), and DMF (0.4 mL) was refluxed for
4 h. The reaction mixture was evaporated under reduced pressure,
and the residue was then evaporated from toluene (5 mL) to re-
move residual thionyl chloride. To the sticky residue cooled in an
ice bath was added ammonia water (30 mL) in portions. After stir-
ring at room temperature for 6 h, the reaction mixture was concen-
4.4.2. Human COX-1 inhibition assay
Test compounds were preincubated for 15 min at 37 °C with
human platelets (5 Â 107 cells/mL) in modified HEPES buffer (pH
7.4). The enzymatic reaction was initiated by addition of 100 lM
arachidonic acid. After 15 min incubation at 37 °C, the reaction
was terminated by addition of 1 N HCl. PGE2 production was quan-
tified to measure COX-1 activity using an EIA kit.
trated to
a small volume and then extracted with EtOAc
(3 Â 20 mL). The extracts were evaporated, and the residue was
purified by flash column chromatography (hexane/EtOAc 1:1) to
give 10 in 60% yield. Mp 187 °C. 1H NMR (200 MHz, acetone-d6) d
8.78 (d, J = 2.6 Hz, 1H, ArH), 8.44 (d, J = 8.4 Hz, 1H, ArH), 8.07 (d,
J = 8.4, 2.6 Hz, 1H, ArH), 7.06 (s, 2H, SO2NH2). 13C NMR (50 MHz,
aetone-d6) d 170.50, 147.99, 144.63, 126.95, 123.61, 100.86. MS
(EI) m/z = 328 [M]+. Anal. Calcd for C6H5IN2O4S: C, 21.97; H, 1.54;
N, 8.54. Found: C, 21.79; H, 1.50; N, 8.49.
4.5. Macrophage COX-1 and COX-2 inhibition assays using
mouse RAW 264.7 cells
For macrophage COX-1 and COX-2 assays, mouse RAW 264.7
cells (5 Â 105 cells) in 0.5 mL of DMEM supplemented with 10%
heat-inactivated fetal bovine serum (FBS), 100 U/mL penicillin,
and 100 lg/mL streptomycin were seeded in a 24-well culture
plate and grown at 37 °C under an atmosphere of 5% CO2 overnight.
Cells were washed with phosphate buffered saline (PBS, pH 7.2)
twice. At this stage, in the case of COX-2 assay, the cells were then
4.3.2.3. 2-(4-Bromophenylsulfanyl)-5-nitrobenzenesulfonamide
(11). To a mixture of aryl iodide 10 (0.2 g), CuI (0.014 g), neocupr-
oine (0.016 g), and t-BuONa (0.126 g) in toluene was added 4-
bromothiophenol (0.14 g). The mixture was refluxed under N2 for
8 h. The reaction mixture was neutralized with dil HCl and then
evaporated under reduced pressure. The resulting residue was ex-
tracted with EtOAc (3 Â 15 mL), and the extracts were evaporated
to dryness. The crude product was purified by flash column chroma-
tography (hexane/EtOAc 1:1) to give 11 in 79% yield. 1H NMR
(200 MHz, acetone-d6) d 8.73 (d, J = 2.6 Hz, 1H, ArH), 8.23 (dd,
J = 8.8, 2.6 Hz, 1H, ArH), 7.74 (m, 2 H, ArH), 7.60 (m, 2H, ArH),
7.25 (d, J = 8.8 Hz, 1H, ArH), 7.09 (s, 2H, SO2NH2). 13C NMR
(50 MHz, acetone-d6) d 147.02, 145.68, 141.60, 137.79, 134.23,
130.71, 127.05, 125.13, 123.98, 110.29.
stimulated with 1 lg/mL LPS (serotype 0111:B4) in DMEM med-
ium for 6 h and were washed again with PBS twice. The culture
medium was replaced by DMEM medium without phenol red
and FBS, and then cells were pretreated with test compounds for
1 h at 37 °C before the addition of 30
30 min incubation at 37 °C, the medium was collected for PGE2 as-
say using PGE2 EIA kit according to the manufacturer’s
instructions.
lM arachidonic acid. After
a
4.6. Assay of human 5-LOX inhibition
The human 5-LOX inhibition assay was performed by MDS
Pharma Services (Taipei, Taiwan) using the test method numbered
136,010. The 5-LOX source was human peripheral blood mononu-
clear leukocytes (PBML). Test compounds were preincubated for
15 min at 37 °C with human PBML (5 Â 106 cells/mL) in HBSS buf-
fer (pH 7.4). The enzymatic reaction was initiated by addition of
4.3.2.4. 3-(4-Bromophenyl)-6-nitrobenzo[1.3.2]dithiazolium ylide
1,1-dioxide (1). To a solution of 11 (0.1 g) in EtOAc/MeOH/H2O
(11 mL/10 mL/3 mL) was added bromine (0.008 mL, 0.025 g). The
mixture was stirred at room temperature for 20 min and poured into
H2O (50 mL). The precipitated solid was collected by filtration,
washed with H2O, and purified by flash column chromatography
(hexane/EtOAc 2:3) to give product 1 in 54% yield. Mp 246 °C. 1H
NMR (400 MHz, acetone-d6) d 8.74 (d, J = 1.7 Hz, 1H, ArH), 8.70
the calcium ionophore A23187 (30 lM). After 15 min incubation
at 37 °C, the reaction was terminated by addition of 1 N HCl. After
neutralization with NaOH and centrifugation, the LTB4 concentra-
tion in the supernatant was measured using an EIA kit.

604
C.-S. Chen et al. / Bioorg. Med. Chem. 18 (2010) 597–604
6. Dogne, J. M.; Supuran, C. T.; Pratico, D. J. Med. Chem. 2005, 48, 2251.
4.7. Assay for inhibition of TNF-a production in THP-1 cells
7. Solomon, D. H. Arthritis Rheum. 2005, 52, 1968.
8. McAdam, R. F.; Catella-Lawson, F.; Nardini, I. A.; Fitzgerald, G. A. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 272.
9. Lewis, R. A.; Austen, K. F.; Soberman, R. J. N. Eng. J. Med. 1990, 323, 645.
10. Jampilek, J.; Dolezal, M.; Opletalova, V.; Hartl, J. Curr. Med. Chem. 2006, 13,
117.
11. (a) Kulkarni, S. K.; Singh, V. P. Inflammopharmacology 2008, 16, 1; (b) Leone, S.;
Ottani, A.; Bertolini, A. Curr. Top. Med. Chem. 2007, 7, 265; (c) Charlier, C.;
Michaux, C. Eur. J. Med. Chem. 2003, 38, 645.
12. Martel-Pelletier, J.; Lajeunesse, D.; Reboul, P.; Pelletier, J. P. Ann. Rheum. Dis.
2003, 62, 501.
The human monocytic THP-1 cells (ATCC No. TIB-202) were
grown in RPMI-1640 medium supplemented with 10% FBS, 4.5 g/
L glucose, 10 mM HEPES, 1.0 mM sodium pyruvate, penicillin
(100 U/mL), streptomycin (0.1 mg/mL), amphotericin B (0.25 g/
mL), and -glutamine (2 mM) at 37 °C in humidified 5% CO2. To pre-
pare conditioned medium for the differentiation of THP-1 cells into
macrophages, THP-1 cells (1–2 Â 107 cells/mL) in RPMI-1640 med-
ium supplemented with 10% FBS were treated with 200 ng/mL phor-
bol myristate acetate (PMA, Sigma, USA) for 24 h at 37 °C in
humidified 5% CO2, washed three times with PBS, and incubated
l
l
L
13. Rainsford, K. D. Agents Actions 1993, 39, C24.
14. Hudson, N.; Balsitis, M.; Everitt, S.; Hawkey, C. J. Gut 1993, 34, 742.
15. Argentieri, D. C.; Ritchie, D. M.; Ferro, M. P.; Kirchner, T.; Wachter, M. P.;
Anderson, D. W.; Rosenthale, M. E.; Capetola, R. J. J. Pharmacol. Exp. Ther. 1994,
271, 1399.
16. Weisman, S. M.; Doyle, M. J.; Wehmeyer, K. R.; Hynd, B. A.; Eichhold, T. H.;
Clear, R. M.; Coggeshall, C. W.; Kuhlenbeck, D. L. Agents Actions 1994, 41, 156.
17. Laufer, S. A.; Augustin, J.; Dannhardt, G.; Kiefer, W. J. Med. Chem. 1994, 37, 1894.
18. Alvaro-Gracia, J. M. Rheumatology 2004, 43, i21.
for another 24 h to eliminate the effect of PMA. For TNF-arelease as-
say, THP-1 cells were stimulated with the conditioned medium in a
96-well plate for 72 h to differentiate into macrophages. The differ-
entiated THP-1 cells were preincubated with test compounds at var-
19. Feldmann, M.; Maini, R. N. Nat. Med. 2003, 9, 1245.
iousconcentrations(0.01, 0.1, 1, and 10
were then stimulated with 100 ng/mL LPS (Sigma, USA) for 24 h.
Cell-free supernatants were collected, and TNF- levels were deter-
lM) for 1 h at 37 °C. The cells
20. Laufer, S.; Tries, S.; Pelletier, J. P. Arthritis Rheum. 2001, 44, 2320.
21. Inagaki, M.; Tsuri, T.; Jyoyama, H.; Ono, T.; Yamada, K.; Kobayashi, M.; Hori, Y.;
Arimura, A.; Yasui, K.; Ohno, K.; Kakudo, S.; Koizumi, K.; Suzuki, R.; Kawai, S.;
Kato, M.; Matsumoto, S. J. Med. Chem. 2000, 43, 2040.
22. Wagner, A. W.; Banholzer, R. Chem. Ber. 1963, 96, 1177.
23. Abramovitch, R. A.; Azogu, C. J.; McMaster, I. T. J. Am. Chem. Soc. 1969, 91, 1219.
24. Abramovitch, R. A.; Azogu, C. J.; McMaster, I. T.; Vanderpool, D. P. J. Org. Chem.
1978, 43, 1218.
a
mined using a commercial ELISA kit (Biosource, cat. no. KHC3012).
Acknowledgments
25. Rabai, J.; Kapovits, I.; Jalsovszky, I.; Argay, Gy.; Fulop, V.; Kalman, A.;
Koritsanszky, T. J. Mol. Struct. 1996, 382, 13.
26. Sang, X.; Du, X. K.; Kadow, J. F. Benzo- and Azabenzodithiazole Compounds.
U.S. Patents 7,087,621, 2006 and 0,101,646, 2005.
This work was supported by the National Science Council of the
Republic of China (Grants NSC94-2320-B002-097, NSC95-2320-
B002-035, and NSC96-2320-B002-017). We thank Taiwan Biotech
Co., Ltd. for the gift of celecoxib. Current address for Chien-Shu
Chen is China Medical University.
27. Shoichet, B. K. Nature 2004, 432, 862.
28. Kurumbail, R. G.; Stevens, A. M.; Gierse, J. K.; McDonald, J. J.; Stegeman, R. A.;
Pak, J. Y.; Gildehaus, D.; Miyashiro, J. M.; Penning, T. D.; Seibert, K.; Isakson, P.
C.; Stallings, W. C. Nature 1996, 384, 644.
29. (a) Kuntz, I. D.; Blaney, J. M.; Oatley, S. J.; Langridge, R.; Ferrin, T. E. J. Mol. Biol.
1982, 161, 269; (b) Kuntz, I. D. Science 1992, 257, 1078; (c) Shoichet, B. K.;
Stroud, R. M.; Santi, D. V.; Kuntz, I. D.; Perry, K. M. Science 1993, 259, 1445.
30. Kast, R. E. Biomed. Pharmacother. 2000, 54, 168.
31. Endres, S.; Whitaker, R. E.; Ghorbani, R.; Meydani, S. N.; Dinarello, C. A.
Immunology 1996, 87, 264.
32. Mansilla-Rosello, A.; Ferron-Orihuela, J. A.; Ruiz-Cabello, F.; Garrote-Lara, D.;
Fernandez-Mondejar, E.; Delgado-Carrasco, M. L. J. Surg. Res. 1997, 67, 199.
33. (a) Santucci, L.; Fiorucci, S.; Giansanti, M.; Brunori, P. M.; Di Matteo, F. M.;
Morelli, A. Gut 1994, 35, 909; (b) Appleyard, C. B.; McCafferty, D. M.; Tigley, A.
W.; Swain, M. G.; Wallace, J. L. Am. J. Physiol. 1996, 270, G42.
34. Nalbant, S.; Akmaz, I.; Kaplan, M.; Avsar, K.; Solmazgul, E.; Sahan, B. Clin. Exp.
Rheumatol. 2006, 24, 361.
35. Kumar, S.; Boehm, J.; Lee, J. C. Nat. Rev. Drug Disc. 2003, 2, 717.
36. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. J. Mol. Biol. 1997, 267,
727.
References and notes
1. Hla, T.; Neilson, K. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 7384.
2. Cryer, B.; Feldman, M. Am. J. Med. 1998, 104, 413.
3. Penning, T. D.; Talley, J. J.; Bertenshaw, S. R.; Carter, J. S.; Collins, P. W.; Docter,
S.; Graneto, M. J.; Lee, L. F.; Malecha, J. W.; Miyashiro, J. M.; Rogers, R. S.; Rogier,
D. J.; Yu, S. S.; Anderson, G. D.; Burton, E. G.; Cogburn, J. N.; Gregory, S. A.;
Koboldt, C. M.; Perkins, W. E.; Seibert, K.; Veenhuizen, A. W.; Zhang, Y. Y.;
Isakson, P. C. J. Med. Chem. 1997, 40, 1347.
4. Prasit, P.; Wang, Z.; Brideau, C.; Chan, C. C.; Charleson, S.; Cromlish, W.; Ethier,
D.; Evans, J. F.; Ford-Hutchinson, A. W.; Gauthier, J. Y.; Gordon, R.; Guay, J.;
Gresser, M.; Kargman, S.; Kennedy, B.; Leblanc, Y.; Leger, S.; Mancini, J.; O’Neill,
G. P.; Ouellet, M.; Percival, M. D.; Perrier, H.; Riendeau, D.; Rodger, I.; Zamboni,
R., et al Bioorg. Med. Chem. Lett. 1999, 9, 1773.
5. Talley, J. J.; Brown, D. L.; Carter, J. S.; Graneto, M. J.; Koboldt, C. M.; Masferrer, J.
L.; Perkins, W. E.; Rogers, R. S.; Shaffer, A. F.; Zhang, Y. Y.; Zweifel, B. S.; Seibert,
K. J. Med. Chem. 2000, 43, 775.
37. Jensen, A. E.; Knochel, P. J. Organomet. Chem. 2002, 653, 122.