Selective inducible microsomal prostaglandin E2 synthase-1 (mPGES-1) inhibitors derived from an oxicam template

Article abstract of DOI:10.1016/j.bmcl.2010.01.060

Here we describe the SAR of a series of potent and selective mPGES-1 inhibitors based on an oxicam template. Compound 13j demonstrated low nanomolar mPGES-1 inhibition in an enzyme assay. In addition, it displayed PGE₂ inhibition in a cell-based assay (0.42 μM) and had over 238-fold selectivity for mPGES-1 over COX-2 and over 200-fold selectivity for mPGES-1 over 6-keto PGF_1α.

Full text of DOI:10.1016/j.bmcl.2010.01.060

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1604–1609
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Selective inducible microsomal prostaglandin E2 synthase-1 (mPGES-1)
inhibitors derived from an oxicam template
a
b
a
a
a
Jane Wang a, , David Limburg , Jeff Carter , Gabriel Mbalaviele , James Gierse , Michael Vazquez
*
a Pfizer Inc. St. Louis Laboratory, PGRD Research, 700 Chesterfield PKWY W, Chesterfield, MO 63017, USA
b Icagen, Inc. 4222 Emperor Blvd., Suite 350 Durham, NC 27703, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Here we describe the SAR of a series of potent and selective mPGES-1 inhibitors based on an oxicam tem-
Received 19 November 2009
Revised 11 January 2010
Accepted 13 January 2010
Available online 25 January 2010
plate. Compound 13j demonstrated low nanomolar mPGES-1 inhibition in an enzyme assay. In addition,
it displayed PGE2 inhibition in a cell-based assay (0.42
lM) and had over 238-fold selectivity for mPGES-1
over COX-2 and over 200-fold selectivity for mPGES-1 over 6-keto PGF1
.
a
Ó 2010 Elsevier Ltd. All rights reserved.
Unmet therapeutic needs in the treatment of inflammation and
arthritis include the need for improved analgesic activity and dis-
ease modification relative to current COX-2 inhibition therapies.
In particular, analgesic agents that display increased maximal
inhibition of pain or utility in a broader population would provide
benefit to aging patients. Development of a new generation of anti-
inflammatory drug with better efficacy through an alternative
mechanism providing a different biochemical profile is a very
urgent and challenging mission for drug discovery scientists.
It has been well established that prostaglandin E2 (PGE2) is an
important mediator in acute and chronic inflammation.1 Reduction
of PGE2 production by the inhibition of COX-2 can remit the signs
and symptoms of pain and edema associated with inflammation.
Microsomal prostaglandin synthase type 1 (mPGES-1) is a member
of the MAPEG family of glutathione transferases, which also in-
cludes FLAP and LTC4 synthase. mPGES-1 is downstream of COX-
2 in the arachidonic acid pathway. It is up-regulated in response
to inflammatory signals and is primarily responsible for the gener-
ation of PGE2 during inflammation. Murine knockout of mPGES-1
demonstrates almost complete reduction of PGE2 from peritoneal
macrophages stimulated with LPS and complete reversal of the
severity and incidence in collagen induced arthritis.2 Therefore, a
selective inhibitor of mPGES-1 would be expected to inhibit PGE2
production induced by inflammation while sparing constitutive
PGE2, prostacyclin (PGI2), and thromboxane production. This selec-
tivity should differentiate mPGES-1 inhibitors from NSAIDS and
COX-2 inhibitors in the treatment of inflammation and arthritis.3
Interest in mPGES-1 as a new therapeutic target is growing and
the number of publications, patent applications, and conference pre-
sentations have increased recently. There are several reported small
molecule mPGES-1 inhibitors in the literature (Fig. 1). MK-886 (1),4 a
FLAP inhibitor with an indole core, displayed mPGES-1 inhibition for
rat (IC50 = 3.2
l
M) and human (IC50 = 1.6
lM), but it also possessed
potent FLAP inhibition (IC50 = 0.026–0.1
lM). Merck scientists have
carried out SAR studies and produced more potent and selective
mPGES-1 inhibitors (2).4 Due to high protein binding and poor cell
permeability, the IC50 of 2 in the cell assay was right shifted and it
did not display reasonable cell activity. Biolipox also published a ser-
ies of patents covering mPGES-1 inhibitors which were based on an
indole scaffold (3).5 These compounds were claimed as potent
mPGES-1 inhibitors (IC50 = 0.062–0.075 lM), but no cellular or
in vivo data were reported. Merck disclosed another series of
mPGES-1 inhibitors (4) derived from a JAK inhibitor,6a which dis-
played sub nanomolar potency (IC50 = 0.001
1000-fold selectivity over other prostaglandin synthases. In the
A549 cells with 50% FBS, it displayed an IC50 = 0.42 M. Biolipox
described other non-indole mPGES-1 inhibitors but they lacked
low nanomolar potency (57a IC50 = 0.39 M; 67b IC50 = 1.3 M; 77c
IC50 = 1.1
M). More recently, Merck reported diarylimidazole 88
as a potent mPGES-1 inhibitor (IC50 = 0.0008 M no other data
l
M),6b and more than
l
l
l
l
l
reported). Koeberle et al. had published a series of pirinxic acid
derivatives8b and the benzo[g]indole-3-carboxylate series8c as novel
mPGES-1/5-LO dual inhibitors with valuable pharmacological
profile.
A series of benzo-thiopyran S-dioxides, exemplified by 9 (Fig. 2)
were discovered by high throughput screeningof the Pfizerchemical
file against the human mPGES-1 enzyme. This series of compounds
demonstrated moderate inhibition against human mPGES-19
Abbreviations: PG, prostaglandin; COX, cyclooxygenase; NSAID, nonsteroidal
anti-inflammatory drug; mPGES-1, microsomal prostaglandin E synthase-1; PGE2,
prostaglandin E2; PGI2, prostaglandin I2; PGF2a, prostaglandin F2a; 6-keto PGF1a,
6-keto prostaglandin F1a.
* Corresponding author. Tel.: +1 636 527 5037; fax: +1 636 247 2086.
E-mail addresses: jl47wang@yahoo.com, jane.1.wang@pfizer.com (J. Wang).
(IC50 = 1.68
lM) and displayed selectivity for mPGES-1 (PGE2) over
a
COX-2 (PGF2 ) in the IL-1-stimulated fetal fibroblast cell assay10
(PGE2 IC50 = 3.4 lM; PGF2 IC50 >90 lM). PGF2 was generated as a
a a
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.060

J. Wang et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1604–1609
1605
Scheme 1. Reagents and conditions: (a) o-xylene, 135 °C, pyridine, 4 Å molecular
sieves, overnight, yield depending on the aniline used.
D), the linker (X), and the linker position on the C ring (see Fig. 3 for
a general structure).
Our initial effort was to evaluate whether we could replaced the
two carbon linker (X) (analog 10) with no linker between the C and
D rings and reduced the 3,4-dichloro to 4-chloro (13a). Biphenyl
13a was prepared by utilizing the chemistry described in Scheme 1
by heating aniline (12a) with ester 11 in o-xylene. Biphenyl 13a
exhibited threefold less potency both in the enzyme and cell assay
compared with compound 10, but it kept a similar cell to enzyme
right shift ratio (threefold). Based on this data, we designed analogs
with a one hetero atom linker using a nitrogen, an oxygen, or a ke-
tone (13b–e) to explore the impact of these atoms to mPGES-1 inhi-
bition. These analogs were prepared by utilizing the chemistry
describedin Scheme1 by heating variousanilines (12b–e) with ester
11 in o-xylene. The phenoxy 13b demonstrated a twofold decrease
enzyme inhibition and a 4.6-fold decrease in cell potency compared
with 13a. The aniline 13c exhibited a twofold improvement in
enzyme potency, but it showed a twofold decrease potency in the
cell compared with 13a. The para substituted keto analog 13d had
similar cell potency compared with 13a, but the meta substituted
keto 13e had greatly decreased mPGES-1 inhibition. This data sug-
gests that human mPGES-1 enzyme inhibition is not very sensitive
to the linker length or the nature of the heteroatoms, both H-bond
donors and acceptors had comparable potency against mPGWS-1.
Enzyme activity varied within a 2-fold range, when the C and D
rings were linked at the para position. We observed that incorpora-
tion of a heteroatom linker has a greater effect on the cellular activity
and ratio of the cellular activity to enzyme activity varied fivefold
(biphenyl 13a), 10-fold (keto 13d), 13-fold (phenoxy 13b), and
28-fold (aniline 13c). This was possibly due to the hydrogen bond
donor nature of the aniline, which decreased cell penetration. This
data also demonstrated that the position of the D ring on the C ring
was very critical for the mPGES-1 inhibition (Table 1).
Figure 1. Structures of mPGES-1 inhibitors.
Figure 2. Structures of early leads.
stable product of PGH2 as described by Mbalaviele et al.10b We ex-
plored further to replace the benzo-thiopyran with dioxobenzo-thi-
azinone (oxicam type) to produce compound 10. We realized that
since oxicams were COX-1 and COX-2 inhibitors, we wantedto make
sure that they were truly mPGES-1 inhibitors and not COX inhibitors
by screening them in our COX assays. None of these oxicam analogs
Since 13a demonstrated good PGE2 inhibition both in the enzyme
and cell and displayed favorable c log P13a (4.59) and c log D13b
(0.710) compared with 10 (c log P = 5.69 and c log D = 1.70). We
sought to further improve the potency and selectivity by varying
the substituents on the D ring of the biphenyl analog (13a) to boost
had COX-1 or COX-2 inhibition less than 50 lM (data not shown).
As described in Scheme 1,11 compound 10 can be prepared by
heating 4-(3,4-dichlorophenylethyl) aniline 1212 with commer-
cially available methyl 4-hydroxy-2H-1,2-benzothiazine-3-carbox-
ylate 1,1-dioxide 11. Compound 10 demonstrated potency both in
the enzyme (mPGES-1 IC50 = 0.11
IC50 = 0.46 M), and it demonstrated selectivity for mPGES-1 over
COX-2 (PGF2 IC50 >68 M). The selectivity over COX-2 in cellular
assay was confirmed using recombinant enzyme assays (hCOX-1
IC50 = 489 M, hCOX-2 IC50 = 467 M). Based on this data, we con-
ducted an extensive SAR exploration on the four rings (A, B, C, and
lM) and cell-based assays (PGE2
l
l
a
l
l
Figure 3. General structure.

1606
J. Wang et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1604–1609
Table 1
Table 2
SAR of mPGES-1 inhibition by variation of the linker between the C and D rings,
SAR of substituents on D ring
compounds 10, 13a to 13e
#
Structure R
mPGES-1
Cell IC50
PGE2/PGF2
l
Mb mPGES-1/
IC50
l
Ma
COX-2
a
#
Structure R1
and R2
mPGES-1
Cell IC50 l
PGE2/PGF2
Mb
mPGES-1/COX-2
selectivity
selectivity
IC50
l
Ma
a
13a
13b
0.29
1.45/33.7
23
13a
13f
13g
13h
13i
13j
13k
13l
R
R
1 = H
0.29
1.45/33.7
23
2 = 4-Cl
R1 = H
0.86
3.44/74.7
2.24/>100
n.d.
22
R
2 = 3-Cl
R1 = H
2 = 4-Me
R1 = H
2 = 4-CN
R1 = H
2 = 4-H
R1 = 3-Cl
0.53
0.15
6.73/70
10
<0.07
3.96
>45
n.d.
14
R
R
13c
13d
4.24/>100
>23
3.13
6.73/92.6
0.42/>100
R
0.016
0.038
0.043
>238
>48
>98
>77
14
R
2 = 4-Cl
R1 = 2-Cl
R2 = 3-Cl
<0.92/44.4
<0.76/>74.8
<0.75/>57.7
6.59/94.7
n.d.
0.11
3.91
1.14/95.6
2.16/95.3
84
44
R
R
1 = 2-Cl
2 = 4-Cl
13m R1 = 3-Me
<0.026
1.00
R
2 = 4-Me
R1 = 2-Me
13n
13o
13e
R
2 = 3-Me
R1 = 3-OMe
R2 = 4-OMe
32.2
n.d.
a
For enzyme assay conditions, see Ref. 9. The IC50 curve was generated based on
duplicated data, which was done n P 2 except 13a n = 1.
For cellular assay conditions, see Ref. 10. The IC50 curve was generated based on
duplicated data, which was done n = 1.
n.d.: no data was provided.
b
a
The IC50 curve was generated based on duplicated data, which was done n P 2
except 13a, 13g, and 13o n = 1.
b
The IC50 curve was generated based on duplicated data, which was done n P 5
for 13j and 13m, n = 3 for 13f, 13k, and 13i, and n = 1 for 13a, 13g, 13i, and 13n.
its potency. We used the same chemistry described in Scheme 1 to
conduct the SAR exploration.
After evaluation of various substituents, we observed that the
nature of the substituents and group positions on the D ring were
important to their mPGES-1 inhibition, (see examples in Table 2).
In general, para chloro (13a) was more potent than meta chloro
(13f) on the D ring, disubstituted analogs (13j, 13k, 13l, 13m)
are more potent mPGES-1 inhibitors than mono substituted ana-
logs (13a, 13f, 13g). Analog (13j) with 3,4-dichloro was more pre-
ferred than that with 2,3-dichloro (13k) and 2,4-dichloro (13l).
Chloro (13a) and methyl (13g) substituents were much more pre-
ferred than hydrogen (13i), cyano (13h), and methoxy (13o) sub-
stituents. Analog 13j demonstrated the greatest inhibition both
in the human mPGES-1 enzyme assay and cell-based assay against
PGE2 production, but did not exhibit any COX inhibition (hCOX-1
IC50 = 118
fold selectivity for COX-2 in the cell.
lM, hCOX-2 IC50 = 263 lM). It also displayed over 238-
Scheme 2. Reagents and conditions: (a) toluene/water/ethanol = 4:2:1, 3 equiv
Na2CO3, Pd(PPh3)4, 85 °C, overnight, yield 50%; (b) o-xylene, 135 °C, pyridine, 4 Å
molecular sieves, overnight, yield 39% and 43%.
With demonstration of low nanomolar potency both in the en-
zyme and cell assays and selectivity for COX-2, we sought to incor-
porate a pyridyl ring as replacement for either the C or D phenyl
ring to further decrease c log P of 13j (5.2). Analogs 17 and 18 were
prepared by the chemistry described in Scheme 2 via a Suzuki
reaction14 to form the dichloro-phenylpyridine amine 16a and
chloro-pyridinylaniline 16b, which were then heated with ester
11 in xylene to provide the desired products 17 and 18. Analogs
19 and 20 were prepared by alternative chemistry (Scheme 3),
since applying the normal chemistry as in Scheme 2 did not lead
to the desired products with the pyridin-2-amine. We used the
same Suzuki reaction to prepare the dichlorophenylpyridine
amines (23a–b), followed by reacting with known starting material
24 to form the amides15 (25 and 26). Analogs 19 and 2016 were
obtained after treating the amides (25 and 26) with lithium i-prop-
oxide in DMSO.
Analogs bearing pyridyl in the C or D ring (18, 19, and 20) exhib-
ited a dramatic decrease in activity, except the 3-pyridyl analog 17
which uniquely retained mPGES-1 activity in both enzyme and cell
assays (Table 3). This data suggested that the biphenyl group was
more preferred for mPGES-1 activity (Table 4).
As incorporation of polar functionality in the C and D rings led
to a decrease in activity, next we wanted to introduce polar groups
in the A ring and to identify a conformationally less constrained B
ring to increase aqueous solubility. First, we incorporated methoxy

J. Wang et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1604–1609
1607
Table 4
A and B rings effect on mPGES-1 inhibition
b
#
Structure R
mPGES-1
Cell IC50
(
l
M)
mPGES-1/COX-
2 selectivity
a
IC50
(l
M)
PGE2/PGF2
a
30
0.69
1.29/>100
>77
Scheme 3. Reagents and conditions: (a) toluene/water/ethanol = 4:2:1, 3 equiv
Na2CO3, Pd(PPh3)4, 85 °C, pyridine, overnight, yield 58%; (b) 2.4 equiv tributyl-
amine, 1.2 equiv 2-chloro-1-methylpyridinium iodide, dichloromethane, reflux for
1 h, yield 75%; (c) DMSO, 3 equiv lithium i-propoxide, 80 °C, yield 10%.
31
32
33
0.098
3.05
0.18
1.45/>100
>69
n.d.
n.d.
Table 3
n.d.
SAR of pyridine as either C or D ring
n.d.
#
Structure R
mPGES-1a IC50
M)
Cell IC50
(lM) PGE2/ c log P
b
(
l
PGF2
a
34
36
38
39
0.28
0.19
5.76
2.04/n.d.
0.67/>100
37.6/>100
n.d.
n.d.
>149
>3
17
0.16
0.69/33.7
n.d.
4.71
18
19
20
12.7
65.6
11.7
3.30
4.50
4.37
n.d.
29.6
n.d.
n.d.
a
The IC50 curve was generated based on duplicated data, which was done n P 2
except 39 n = 1.
b
The IC50 curve was generated based on duplicated data, which was done n P 2
a
The IC50 curve was generated based on duplicated data, which was done n = 1
except 31, 34, and 38 n = 1.
except 17 n = 4.
b
The IC50 curve was generated based on duplicated data, which was done n = 2.
Carboxylic acid 38 was reduced by treating with borane–THF to
obtain alcohol 39 as a minor product.
or hydroxyl groups into the A ring. Compounds (30 and 32) were
prepared (Scheme 4) by heating aniline 12j with esters 27 and
28, and compound 30 and 32 were O-demethylated to form com-
pounds 31 and 33. The N-methyl analog 34 (Scheme 4) was also
prepared by a one step reaction via heating aniline 12j and 29 in
xylene. The B ring saturated analog 36 was obtained (Scheme 5)
by first hydrogenation of ester 11 under 10% Pd on carbon condi-
tions, followed by hydrolysis to form 35. Then 35 was reacted with
aniline 12j to form amide 36 using the same amide formation con-
dition described in Scheme 3. The B-ring opened analogs 38 and 39
(Scheme 6) were prepared by first forming amide 37, which then
was treated with sodium hydroxide to form desired product 38.
The analogs with substituents on the A ring (30, 31, 32, and 33)
did not display improved the mPGES-1 inhibition. For analogs 30
and 32, both having a methoxy group on the A ring, lost activity,
42-fold (30) and 185-fold (32), respectively. With hydroxyl replac-
ing methoxy on the A ring (31 and 33), they displayed better activ-
ity compared with 30 and 32, and only lost sixfold and 11-fold
activity compared with 13j. It seemed that position 6 (30 and 31)
had better toleration than position 7 (32 and 33), and the hydroxyl
group (31 and 33) was a more preferred group than methoxy (30
and 32). Methylated sulfonamide 34 was designed to disrupt the
ring planarity to improve the solubility, but it lost 17-fold
mPGES-1 inhibition in enzyme and fivefold inhibition in the cell

1608
J. Wang et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1604–1609
Scheme 4. Reagents and conditions: (a) o-xylene, 135 °C, pyridine, 44 Å molecule
sieves, overnight, yield depending on the anilines; (b) 5.0 equiv BBr3, dichloroeth-
ane, 60 °C for 2 h, yield 50%.
Scheme 6. Reagents and conditions: (a) 2.4 equiv tributylamine, 1.2 equiv 2-
chloro-1-methylpyridinium iodide, dichloromethane, reflux for 1 h, yield 75%; (b)
THF/MeOH = 4:1, 5 equiv NaOH (2.5 N), rt, 4 h, 1 N HCl, yield 100%; (c) THF, 4 equiv
BH3.THF, room temp, 3 h, added water, 50 °C, 0.5 h, yield 10%.
prostanoids was further demonstrated in IL-1ß-stimulated RASF
cells.18 It also inhibited PGE2 production in the LPS/human whole
blood assay19 with an IC50 around 5
lM. The low inhibition in
the whole blood assay was likely due to the high protein binding
of the molecule.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2010.01.060.
Scheme 5. Reagents and conditions: (a) EtOAc, cat. AcOH, 10% Pd–C, H2 45 psi,
overnight, yield 90%; (b) THF/MeOH = 4:1, 4 equiv NaOH (2.5 N), 50 °C, 2 h, 1 N HCl,
yield 100%; (c) 2.4 equiv tributylamine, 1.2 equiv 2-chloro-1-methylpyridinium
iodide, dichloromethane, reflux for 1 h, yield 87%.
References and notes
1. Kobayashi, T.; Narumiya, S. Prostagland. Other Lipid Mediat. 2002, 68–69, 557.
2. Trebino, C.; Stock, J.; Gibbons, C.; Naiman, B.; Wachtmann, T.; Umland, J.;
Pandher, K.; Lapointe, J.; Saha, S.; Roach, M.; Carter, D.; Thomas, N.; Durtschi,
B.; McNeish, J.; Hambor, J.; Jakobsson, P.; Carty, T.; Perez, J.; Audoly, L. Proc.
Natl. Acad. Sci. U.S.A. 2003, 100, 9044.
and had same aqueous solubility range as 13j (<3
the saturated B-ring analog 36 had a better c log P (4.14) and im-
proved water solubility17 from 3
M (13j) to 25 M, it exhibited
a 10-fold decreased enzyme inhibition and 1.5-fold decrease in
the cell. Analog 36 demonstrated no COX inhibition in the enzyme
lM). Although
l
l
3. (a) Portanova, J. P.; Zhang, Y.; Anderson, G.; Hauser, S.; Masferrer, J.; Seibert, K.;
Gregory, S.; Isakson, P. J. Exp. Med. 1996, 184, 883; (b) Pulichino, A. M.;
Rowland, S.; Wu, T.; Clark, P.; Xu, D.; Mathieu, M.; Riendeau, D.; Audoly, L., et al
J. Pharmacol. Exp. Ther. 2006, 319, 1043; (c) Murata, T.; Ushikubi, F.; Matsuoka,
T.; Hirata, M.; Yamasaki, A.; Sugimoto, Y.; Ichikawa, A.; Aze, Y.; Tanaka, T.;
Yoshida, N.; Ueno, A.; Oh-ishi, S.; Narumiya, S. Nature 1997, 388, 678; (d)
Jakobsson, P. J.; Thoren, S.; Morgenstern, R.; Samuelsson, B. Proc. Natl. Acad. Sci.
U.S.A. 1999, 96, 7220; (e) Murakami, M.; Naraba, H.; Tanioka, T.; Semmyo, N.;
Nakatani, Y.; Kojima, F.; Ikeda, T.; Fueki, M.; Ueno, A.; Oh, S.; Kudo, I. J. Biol.
Chem. 2000, 275, 32783; (f) Murakami, M.; Kudo, I. Prog. Lipid. Res. 2004, 43, 3.
4. Riendeau, D.; Aspiotis, R.; Eithier, D.; Gareau, Y.; Grimma, E.; Guaya, J.; Guirala,
S.; Juteaua, J.; Mancinia, H.; Méthota, N.; Rubina, J.; Friesena, R. Bioorg. Med.
Chem. Lett. 2005, 15, 3352.
5. Olofsson, K.; Suna, E.; Pelcman, B.; Ozola, V.; Katkevics, M.; Kalvins, I. Indoles
Useful in the Treatment of Inflammation. WO 2005/123673 A1, 2005.
6. (a) Chau, A., Côté, B.; Ducharme, Y.; Frenette, R.; Friesen, R. W.; Gagnon, M.;
Giroux, A.; Martins, E.; Yu, H.; Wu, T. 2-(Phenyl or heterocyclic)-1H-
phenanthro[9,10-D]imidazoles as mPGES-1 inhibitors. WO 063466 A1, 2006.;
(b) Côté, B.; Boulet, L.; Brideau, C.; Claveau, D.; Ethier, D.; Frenette, R.; Gagnon,
M.; Giroux, A.; Guay, J.; Guiral, S.; Mancini, J.; Martins, E.; Massé, F.; Méthot, N.;
Riendeau, D.; Rubin, J.; Xu, D.; Yu, H.; Ducharme, Y.; Friesen, R. Bioorg. Med.
Chem. Lett. 2007, 17, 6816.
7. (a) Pelcman, B.; Olofsson, K.; Arsenjans, P.; Ozola, V.; Suna, E.; Kalvins,
I.Thienopyrroles Useful in the Treatment of Inflammation. WO Patent 077412
A1, 2006.; (b) Pelcman, B.; Olofsson, K.; Schaal, W.; Kalvins, I.; Katkevics, M.;
Ozola, V.; Suna, E. Benzoxazoles UseFul in the Treatment of Inflammation. WO
2007/042816 A1, 2007.; (c) Pelcman, B.; Olofsson, K.; Habarova, O.; Kalvins, I.;
Suna, E.; Trapencieris, P.; Andrianov, V. Naphthalene–disulfonamides Useful for
the Treatment of Inflammation. WO 2007/042817 A1, 2007.
8. (a) Juteau, H.; Wu, T.; Ducharme, Y. In 234th National Meeting of the Am. Chem.
Soc. Boston, MA. August, 2007; MEDI 430.; (b) Koeberle, A.; Zettl, H.; Greiner,
C.; Wurglics, M.; Schubert-Zsilavecz, M.; Werz, O. J. Med. Chem. 2008, 51, 8068;
assay (hCOX-1 IC50 >50 lM, hCOX-2 IC50 >50 lM). We determined
that the acidic proton within the dioxobenzothiazinone is very
important to achieve potent mPGES-1 inhibition. In order to im-
prove solubility and further lower the c log P, we opened the B (thi-
azinone) ring to provide carboxylic acid 38 and alcohol 39 to mimic
the oxicam. Unfortunately, analogs 38 and 39 were poor mPGES-1
inhibitors.
Because 13j displayed excellent potency for mPGES-1 and selec-
tivity over COX-2, we wanted to check whether 13j had inhibitory
effects on other prostanoids, such as thromboxane B2 (TXB2) and
6-keto-PGF1 . The inducible nature of mPGES-1 and COX-2 expres-
a
sion made synovial fibroblasts derived from patients with rheuma-
toid arthritis (RASF) the suitable cells for the search of inhibitors
that selectively inhibit mPGES-1 function in converting PGH2 to
PGE2.10b 13j blocked the production of PGE2 from PGH2 (PGE2
IC50 ꢀ0.5
lM), while sparing the production of 6-keto PGF1
a
(a metabolite of PGI2), PGF2 , and TXB2 (all three IC50 >100
lM).
a
In summary, we have identified very potent and selective
mPGES-1 inhibitors which demonstrated better cell activity over
previously reported series. Compound 13j displayed excellent
mPGES-1 inhibition and selectivity over COX-2 in the human fetal
fibroblast cell assay. In addition, the selectivity of 13j over other

J. Wang et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1604–1609
1609
(c) Koeberle, A.; Haberl, E.; Rossi, A.; Pergola, C.; Dehm, F.; Northoff, H.;
16. Herrmann, W.; Geibei, W.; Satzinger, G. DE 3,212,485 A1, 1983.
Troschuetz, R.; Sautebin, L.; Werz, O. Bioorg. Med. Chem. 2009, 17, 7924.
9. Incubation of enzyme with 2 lM PGH2 for 41 s at rt and assessment of PGE2 by
17. An aqueous solution (with 2% DMSO) of the test compound is stirred for 24 h at
room temperature, then filtered. The nitrogen content of the filtered aqueous
solution is measured and, in conjunction with the molecular formula, is used to
calculate the concentration of the solution. True thermodynamic solubility
requires the measurement of the concentration of saturated aqueous solutions
in equilibrium with solid; these measurements are usually laborious and
ELISA. Enzyme is suspended in 100 mM K3PO4 at pH 6.2 buffer and 2.5 mM
glutathione. Compound solubilized in DMSO and added at a 1:9 w/w DMSO to
enzyme ratio. PGH2 (Cayman Chemical, Ann Arbor MI) is diluted 12.8Â in ice
cold 10 mM HCl from a 278 lM stock in acetone. The reaction is begun by the
addition of 1:10 volume of PGH2 to the enzyme inhibitor mixture for a final
concentration of 2 M. The reaction is terminated by the addition of 1:10
relatively expensive. This apparent solubility assay is a high-throughput,
l
relatively low cost test. Data published by Analiza (see their website: http://
www.analiza.com) show a good correlation between data from this high-
throughput assay and traditional shake flask solubility data (r2 = 0.9993, with
standard deviation of <3%). Method: A Hamilton Starlet Liquid Handler or
Analiza’s ADW (Automated Discovery Workstation) robotic platform was used
for all steps of the analysis. 50 mM sodium phosphate buffer, at the indicated
pH, was freshly prepared from NaH2PO4 and Na2HPO4 and filtered. Buffer
volumes of 2.5 mM FeCl2. The reaction is immediately diluted 120Â into ELISA
buffer (according to Cayman Chemical, Ann Arbor MI recipe). PGE2 formed was
calculated from a standard curve of PGE2 by ELISA (Cayman Chemical, Ann
Arbor MI). The% control activity was calculated as the percentage difference
between negative control (100% inhibited with a reference compound) and
enzyme only control. The difference between enzymatic versus non-enzymatic
production of PGE2 are typically 3–4-fold. IC50’s are calculated by 4 parameter
log fit of the % control data.
(294
Millipore polycarbonate filter solubility plate (part# MSSLBPC10) for a final
final DMSO concentration of 2%. The plates were heat sealed with a
lL) was combined with 6 lL of 30 mM DMSO stock solutions in a
10. (a) Human fetal fibroblasts (FF), plated at 3 Â 106 cells/cm2 in DMEM
supplemented with 15% FBS, glutamine, pen/strep and HEPES, were
incubated at lmv37 °C, 5% CO2. Approximately 20–24 h later, media were
refreshed and cells were treated with 1 ng/mL IL-1b for an additional 20–24 h,
after which they were washed once with DMEM without serum. Cells were
then incubated with compounds for 50 min in DMEM without serum then with
a
polypropylene seal. After 24 h of 200 rpm shaking at room temperature (23–
25 °C), the plates were vacuum filtered into deep well receiver plates. Filtrates
were injected into the chemiluminescent nitrogen detector (Antek 8060) for
quantification. Apparent concentration was divided by the number of
nitrogens in the sample (from the molecular formula) to determine the
10
l
M arachidonic acid for 10 min. PGE2 levels were analyzed by ELISA. To
concentration. The results are reported here in lg/mL and lM.
measure PGF2 levels, FF were processed as described above, except that after
incubation with the compounds for 50 min, cells were treated with SnCl2
18. Synovial fibroblasts derived from patients with rheumatoid arthritis (RASF),
were isolated via enzymatic digestions from primary synovial tissues isolated
after knee synovectomy as previously described. RASF were plated at
a
(which converts PGH2 non-enzymatically to PGF2
)
for 10 min before
a
treatment with 10
lM
arachidonic acid for an additional 10 min. In this
8 Â 104 cells/cm2 and cultured for three days in DMEM containing 10
lM L-
assay, a COX-2 inhibitor blocks the production of PGH2 and consequently
neither PGE2 nor PGF2 is synthesized, whereas PGE2 not PGF2 production
glutamine, 25 lM HEPES, 10 units/mL penicillin, 10 lg/mL streptomycin,
supplemented with 10% (v/v) fetal bovine serum (FBS), at 37 °C in 95% air,
5% CO2 atmosphere. To test the effects of the inhibitors on prostaglandin (PG)
production during 24 h incubation, the culture media were replaced with fresh
DMEM containing 1% (v/v) FBS, and cells were treated for 24 h with 1 ng/mL IL-
1b in the presence of vehicle (1% Me2SO, final concentration) or the inhibitors
(1% Me2SO, final concentration). The conditioned media were then collected for
PG analysis. To determine direct effects of the inhibitors on enzyme activity,
cells were treated for 24 h with 1 ng/mL IL-1b as described above. The
conditioned media were removed, and cells were washed twice with serum-
free media. Cells were then treated with the media containing either the
a
a
should be inhibited by a mPGES-1 inhibitor. PGF2 levels were analyzed by
a
ELISA; (b) Mbalaviele, G.; Pauleya, A.; Shaffera, A.; Zweifela, B.; Mathialagana,
S.; Mnicha, S.; Nemirovskiya, O.; Cartera, J.; Giersea, J.; Wang, J.; Vazqueza, M.;
Moore, W.; Masferrer, J. Biochem. Pharmacol., pending.
11. Zia-ur-Rehman, M.; Choudary, J. A.; Ahmad, S.; Siddiqui, H. L. Chem. Pharm. Bull.
2006, 54, 1175.
12. Simons, L. J.; Caprathe, B. W.; Callahan, M., et al Bioorg. Med. Chem. Lett. 2009,
19, 654.
13. (a) This protocol uses the BioByte (www.biobyte.com) program c log P, version
4.3, to calculate the logarithm of the partition coefficient. The partition
coefficient, often taken on the log scale (log P) is defined for neutral
compounds. The reference system is octanol/water. To take ionization into
account distribution coefficients (log D) can be calculated at a selected pH, for
example, 7.4 or 6.5. Often log D values have more physiological meaning than
log P values; (b) Calculated logarithm of the octanol/water distribution
coefficient using ACD pchbat version 9.3. PH (default value = 7.4) IONIZE
(default value = UnSet).
14. Shen, H. C.; Ding, F.; Luell, S.; Forrest, M.; Carballo-Jane, E.; Wu, K.; Wu, T.;
Cheng, K.; Wilsie, L.; Krsmanovic, M.; Taggart, A.; Ren, N.; Cai, T.; Deng, Q.;
Chen, Q.; Wang, J.; Wolff, M.; Tong, X.; Holt, T.; Waters, G.; Hammond, M.; Tata,
J.; Colletti, S. J. Med. Chem. 2007, 50, 6303.
vehicle (1% Me2SO) or the inhibitors (1% Me2SO) for 50 min, and with 10
lM
arachidonic acid for additional 10 min, at 37 °C in 95% air, 5% CO2 atmosphere,
and the conditioned media were collected for PG analysis. For PGF2 studies,
a
2 mg/mL SnCl2 was added to the cells 40 min after addition of the inhibitors (or
10 min before the addition of arachidonic acid). The levels of PGE2, PGF1 and
a
PGF2 in the conditioned media were measured by ELISA.
a
19. Human whole blood collected from healthy human donors in heparinized
tubes was mixed with human head and neck squamous cell carcinoma, 1483
cells. Briefly, 100
well and incubated with the compounds for 15 min at 37 °C in 95% air, 5% CO2
atmosphere, then with 30 M arachidonic acid for 10 min. The assay plates
l
L blood/well (384-well) were mixed with 105 1483 cells/
l
were centrifuged at 930g for 10 min at room temperature, and the plasma was
15. Groundwater, P. W.; Garnett, I.; Morton, A. J. J. Chem. Soc., Perkin Trans. 1 2001,
2781.
removed for quantitation of PGE2 levels by ELISA.