Discovery and SAR of PF-4693627, a potent, selective and orally bioavailable mPGES-1 inhibitor for the potential treatment of inflammation

Article abstract of DOI:10.1016/j.bmcl.2012.11.109

Inhibition of mPGES-1, the terminal enzyme in the arachidonic acid/COX pathway to regulate the production of pro-inflammatory prostaglandin PGE2, is considered an attractive new therapeutic target for safe and effective anti-inflammatory drugs. The discovery of a novel series of orally active, selective benzoxazole piperidinecarboxamides as mPGES-1 inhibitors is described. Structure-activity optimization of lead 5 with cyclohexyl carbinols resulted in compound 12, which showed excellent in vitro potency and selectivity against COX-2, and reasonable pharmacokinetic properties. Further SAR studies of the benzoxazole ring substituents lead to a novel series of highly potent compounds with improved PK profile, including 23, 26, and 29, which were effective in a carrageenan-stimulated guinea pig air pouch model of inflammation. Based on its excellent in vitro and in vivo pharmacological, pharmacokinetic and safety profile and ease of synthesis, compound 26 (PF-4693627) was advanced to clinical studies.

Full text of DOI:10.1016/j.bmcl.2012.11.109

Bioorganic & Medicinal Chemistry Letters xxx (2012) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and SAR of PF-4693627, a potent, selective and orally bioavailable
mPGES-1 inhibitor for the potential treatment of inflammation
⇑
Graciela B. Arhancet , Daniel P. Walker, Sue Metz, Yvette M. Fobian, Steven E. Heasley, Jeffrey S. Carter,
John R. Springer, Darin E. Jones, Michael J. Hayes, Alexander F. Shaffer, Gina M. Jerome, Michael T. Baratta,
Ben Zweifel, William M. Moore, Jaime L. Masferrer, Michael L. Vazquez
Pfizer Worldwide Medicinal Chemistry, Pfizer, Inc. 700 Chesterfield Parkway West, Chesterfield, MO 63017, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online xxxx
Inhibition of mPGES-1, the terminal enzyme in the arachidonic acid/COX pathway to regulate the produc-
tion of pro-inflammatory prostaglandin PGE2, is considered an attractive new therapeutic target for safe
and effective anti-inflammatory drugs. The discovery of a novel series of orally active, selective benzox-
azole piperidinecarboxamides as mPGES-1 inhibitors is described. Structure–activity optimization of lead
5 with cyclohexyl carbinols resulted in compound 12, which showed excellent in vitro potency and selec-
tivity against COX-2, and reasonable pharmacokinetic properties. Further SAR studies of the benzoxazole
ring substituents lead to a novel series of highly potent compounds with improved PK profile, including
23, 26, and 29, which were effective in a carrageenan-stimulated guinea pig air pouch model of inflam-
mation. Based on its excellent in vitro and in vivo pharmacological, pharmacokinetic and safety profile
and ease of synthesis, compound 26 (PF-4693627) was advanced to clinical studies.
Keywords:
Inflammation
m-PGES-1
Benzoxazole
Clinical candidate
PF-4693627
Ó 2012 Elsevier Ltd. All rights reserved.
For patients with inflammatory pain caused by osteoarthritis
(OA) and rheumatoid arthritis (RA), the first line of treatment is
nonsteroidal anti-inflammatory drugs (NSAIDs) and COX-2 inhibi-
tors (COXIBs). The COXIBs are superior to NSAIDs due to their low-
er incidence of GI side effects, but they are no more effective in
symptom remission than NSAIDs and safety issues have surfaced,
including renal toxicity and increased cardiovascular risk. There-
fore, a drug that displays equal or better efficacy for treating OA
and RA relative to NSAIDs and COXIBs, through an alternative bio-
chemical pathway would be an exciting product concept for drug
discovery.
Prostaglandin E2 (PGE2) is an important mediator in acute and
chronic inflammation, pain, and fever and is found in the synovial
fluid of patients with OA and RA.1 The biosynthesis of PGE2 begins
with the cleavage of arachidonic acid (AA) from membrane phos-
pholipids by phospholipase A2, followed by the conversion of AA
to PGH2 by cyclooxygenase (COX), and finally to the production
of PGE2 by PGE synthase (PGES).2 Three forms of PGES have been
reported: cytosolic PGES (cPGES), microsomal PGES-1 (mPGES-1),
and microsomal PGES-2 (mPGES-2).3 Both cPGES and mPGES-2
are constitutively expressed in a variety of tissues, while mPGES-
1 is up-regulated under inflammatory conditions.
tive PGE2, prostacyclin (PGI2), and thromboxane production. This
selectivity should differentiate mPGES-1 inhibitors from NSAIDS
and COX-2 inhibitors in the treatment of inflammation and arthri-
tis. A number of excellent reviews summarizing the biology, phar-
macology of mPGES-1, as well as the progress in the development
of mPGES-1 inhibitors, have recently become available.4
Figure 1 shows examples of mPGES-1 inhibitors recently de-
scribed in the literature. A series of potent and selective mPGES-
1 inhibitors, represented by 1 was obtained by optimization of a
5-lipoxygenase-activating protein (FLAP) inhibitor lead with mod-
erate rat and human mPGES-1 inhibition.5 These compounds are
highly protein shifted and thus they have not shown mPGES-1
inhibition in vivo.
The benzoxazole scaffold has been described in the patent liter-
ature to yield inhibitors of the MAPEG family enzymes including
mPGES-1.6 The most potent compounds against mPGES-1 dis-
closed are 2 (IC50 = 1.3
lM) and 3 (IC50 = 1.5 lM) although no
selectivity and efficacy data was reported.
Phenanthrene imidazole 4 (MF63) was described as a potent
inhibitor of human mPGES-1 and showed potency in a LPS-
stimulated human whole blood assay where it inhibited PGE2
production, IC50 = 1.3 lM. Although 4 is inactive against rat mPGES-
A selective inhibitor of mPGES-1 would be expected to inhibit
PGE2 production induced by inflammation while sparing constitu-
1 for unknown reasons, it is active on guinea pig enzyme and showed
efficacy in a guinea pig model of LPS-induced hyperalgesia.7
In a prior study, we disclosed a novel class of benzoxazoles as
inhibitors of mPGES-1.8 The initial HTS hit (5, Fig. 2), showed mod-
est activity in our enzyme assay (850 nM), but it lacked potency in
⇑
Corresponding author. Tel.: +1 636 926 7471; fax: +1 636 926 7449.
E-mail address: graciela.arhancet@novusint.com (G.B. Arhancet).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.11.109
Please cite this article in press as: Arhancet, G. B.; et al. Bioorg. Med. Chem. Lett. (2012), http://dx.doi.org/10.1016/j.bmcl.2012.11.109

2
G. B. Arhancet et al. / Bioorg. Med. Chem. Lett. xxx (2012) xxx–xxx
X
R
HN
N
O
O
CO2H
F
N
Cl
2, R = CF3
3, R = Me
Cl
1
Cl
NC
NC
H
N
N
4
Figure 1. Scaffolds of reported mPGES-1 inhibitors.
Bn
O
SAR
Cl
Cl
HN
O
N
O
HN
O
N
O
N
N
Parallel chemistry
5
6
Figure 2. Optimization of benzoxazole HTS hit.
our cell-based assay. A parallel chemistry program ensued to study
the SAR of the piperidine and cyclopentane rings of 5. Based on this
work, benzoxazole 6 was identified as a very potent inhibitor
(mPGES-1 IC50 = 18 nM) with equal potency in a fetal fibroblast cell
assay.9 Unfortunately, the compound showed only modest activity
Table 1
Amino cyclohexyl carbinol optimization
Cl
N
R1
O
N
O
in our human whole blood (HWB) cell-based assay (IC50 = 7.5 lM).
Also, while compound 6 had an acceptable in vivo dog pharmaco-
kinetic (PK) profile; in general, the in vivo PK for compounds pos-
sessing a cyclopentane ring on the right-hand side was quite
variable. Herein, we describe the design and synthesis of new
benzoxazoles, focusing on the SAR of the left-hand benzoxazole
and right-cyclopentyl regions of the molecule. Also, we will detail
in vitro and in vivo efficacy and safety data that led to the discov-
ery of PF-04693627, a candidate suitable for human studies.
Our structure–activity relationship (SAR) was initially focused
on improving the activity of lead 5 in our HWB cell assay. It was
hypothesized that the large right-shift of compound 6 on going
from the enzyme to HWB cell-based assay was likely due the com-
pound’s high clogP (4.2). We hypothesized that replacing the ben-
zyloxycyclopentane in compound 6 with a smaller, more polar
group would lead to an increase in potency in the HWB cell assay.
To that end, we focused on compound 7 (Table 1), which was dis-
covered during the hit-to-lead phase of the program.8 While com-
pound 7 was not as potent as compound 6, it possessed a more
favorable HWB cell to enzyme ratio (36 vs 220), presumably due
to the compound’s lower clogP of 1.7. Additional SAR work on
the cyclohexyl ring revealed that moving the hydroxymethyl group
from the 4-position of the cyclohexane ring to the 3-position in
compound 10 resulted in potent activity in both the enzyme and
HWB cell assays. It was further determined that the (1S,3S)-
enantiomer (12) is 75-fold more potent than the corresponding
(1R,3R)-enantiomer (11).10 Interestingly, while the cis-geometry
is preferred in the 1,4-disubstituted cyclohexane series (compare
compound 7 with compound 8), the trans-geometry (10) is
preferred in the 1,3-disubstituted cyclohexane series over the
a
Compound
R1
HN
mPGES-1 IC50 (nM)
7
383
OH
OH
HN
HN
8
9
3620
6750
206
HO
HO
HN
HN
10
11
R
R
4850
65
HO
HN
S
S
12
HO
mPGES-1 enzyme assay as described in Ref. 9. IC50 values represent the con-
centration to inhibit 50% of PGE2 relative to vehicle control. Numbers indicate IC50
values generated from individual 10-point concentration response relationships in
duplicate.
a
Please cite this article in press as: Arhancet, G. B.; et al. Bioorg. Med. Chem. Lett. (2012), http://dx.doi.org/10.1016/j.bmcl.2012.11.109

G. B. Arhancet et al. / Bioorg. Med. Chem. Lett. xxx (2012) xxx–xxx
3
cis-geometry (9).11 While compound 12 displayed a significantly
improved HWB cell to enzyme ratio of 5, it lacked in vivo metabolic
stability in a rat PK model (vide infra).
At this point, we turned our attention toward optimizing the
left-hand benzoxazole region of the molecule with the hope of
identifying compounds with equal or better mPGES-1 potency
and improved rat PK compared to compound 12.
Over 100 analogs possessing substituent modifications to the
benzoxazole core were prepared. A subset of those analogs is de-
tailed in Table 2. The synthesis of benzoxazoles 13–18, and 22 is
shown in Scheme 1. Toward this end, treatment of commercially
available aminophenols 32 with carbon disulfide afforded 2-mer-
captobenzoxazoles 33. Replacement of the thiol in 33 with ethyl
isonipecotate gave rise (80–90%) to piperidinebenzoxazoles 34.
Saponification of the ester in 34 led to carboxylic acids 35.
HBTU-Mediated coupling of acids 35 with (1S,3S)-3-amin-
ocyclohexylmethanol10 afforded amides 13–18 and 22.
The synthesis of analogs 19–21 is shown in Scheme 2. Thus,
bromination of benzoxazole 37 afforded bromide 38. Suzuki-cross
coupling reaction of in situ generated organo-boranes with bro-
mide 38 gave rise (75–80%) to alkyl-substituted benzoxazoles
39.14 Saponification of the ester in 39 and coupling with amine
36 led to compounds 19–21.
trifluoromethylation gave rise to 42.15 Separation of 42 from unre-
acted aryl iodide 41 was not possible. However, hydrogenation of
the crude product mixture converted unreacted iodide into 37,
which allowed clean isolation of 42 via chromatography. Saponifi-
cation of the ester in 42 and coupling with amine 36 led to analog
23.
Difluoromethyl analog 24 was prepared as shown in Scheme 5.
Thus, exposure of 37 to hexamethylenetetramine (HMTA) led to a
69% yield of aldehyde 43.16 Fluorination of the aldehyde with DAST
afforded difluoride 44. Saponification of the ester and coupling
with amine 36 gave rise to analog 24.
Our original screening hit (5) had a chlorine atom at the 5-posi-
tion of the benzoxazole. Initial attempts to replace this substituent
with hydrogen or methyl produced inactive analogs (data not
shown). In general the replacement of chlorine at this position
was met with limited success. For instance, while other small, lipo-
philic, electron-withdrawing groups, such as bromo (13) or trifluo-
romethyl (17) were tolerated at this position, they did not offer any
significant benefit in terms of potency or lower clearance com-
pared to the chloro analog 12. Replacing the chloro substituent
with small, polar substituents, such as cyano (14), ethanesulfonyl
(15) or carboxamido (16) caused
mPGES-1 activity.
a significant right-shift in
6-Aryl and heteroaryl substituted benzoxazoles 25–31 were
prepared as described in Scheme 3. Accordingly, Suzuki cross cou-
pling of bromide 38 with the appropriate aryl or heteroaryl boronic
acid afforded coupled products 40. Treatment of the ester in 40
with sodium hydroxide, followed by subsequent amide coupling
with amine 36 led to analogs 25–31.
Having identified chlorine as an ideal group at C(5), we next
explored substitution at the C(6) position of the benzoxazole
(Table 2). Analogs 18–21 show the effects of placing an alkyl sub-
stituent at C(6). We were pleased to find that all of these analogs
showed enhanced activity in the mPGES-1 enzyme assay. While
the methyl (18), ethyl (19) and isopropyl (20) analogs also showed
enhanced activity in the HWB cell assay, the isobutyl analog had
similar cell activity to compound 12. In general, as the clogP
Analog 23 was prepared as shown in Scheme 4. Toward this
end, iodination of compound 37 gave iodide 41; subsequent
Table 2
Benzoxazole ring optimization
R5
O
N
N
O
HN
S
R6
S
HO
HLM CLintc (mL/min/kg)
1.74 19.7
a
b
Compound R5
R6
mPGES-1 IC50 (nM) HWB IC50 (nM) HWB/mPGES-1 IC50 ratio clogP
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Cl
H
65
340
5.2
Br
CN
H
H
H
H
H
Me
Et
i-Pr
i-Bu
Cl
CF3
CHF2
47
5700
>33,000
>33,000
88
25
20
6
15
16
33
49
3
3
3
3
2
18
8
761
18.5
1.89 25.6
0.55 < 8
0.32
À0.05
9800
>50,000
>50,000
510
338
162
261
761
684
174
336
119
109
240
133
53
632
317
Ethylsulfonyl
Carboxamide
CF3
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
5.8
13.5
8.1
43.5
50.7
42.8
5.3
1.96 22.2
2.24 28.3
2.77 64.6
3.17 46.4
3.70 44.0
2.36 27.1
2.49 26.5
1.94 40.2
3.52 81.9
4.09 37.1
4.09 55.8
3.83 <8
6.9
4-Fluorophenyl
4-Chlorophenyl
3-Chlorophenyl
4-Ethoxyphenyl
4-(Trifluoromethoxy)phenyl
4-Methoxypyridyl
4-(Trifluoromethyl)pyridyl
39.7
36.3
80
44.3
26.5
35.1
39.6
4.41 50.8
2.70 52.0
2.91 68.8
a
mPGES-1 enzyme assay as described in Ref. 9. IC50 values represent the concentration to inhibit 50% of PGE2 relative to vehicle control. Numbers indicate IC50 values
generated from individual 10-point concentration response relationships in duplicate.
b
Lipopolysaccharide (LPS) stimulated human whole blood (HWB) cell assay. IC50 values represent the concentration to inhibit 50% of PGE2 relative to vehicle control, see
Ref. 12 for further details. Numbers indicate IC50 values generated from individual 10-point concentration response relationships in duplicate.
c
Human liver microsomal stability assay. CLint refers to total intrinsic clearance obtained from scaling in vitro half-lives in human liver microsomes.13
Please cite this article in press as: Arhancet, G. B.; et al. Bioorg. Med. Chem. Lett. (2012), http://dx.doi.org/10.1016/j.bmcl.2012.11.109

4
G. B. Arhancet et al. / Bioorg. Med. Chem. Lett. xxx (2012) xxx–xxx
NH2
OH
Cl
Cl
Cl
N
O
N
a
b
SH
N
CO2Et
O
R
R
R
c
32
34
33
Cl
Cl
O
N
O
N
O
d
N
CO2H
N
S
HN
R
R
S
35
13-18, 22
HO
S
H2N
S
OH
36
Scheme 1. Preparation of compounds 13–18 and 22. Reagents and conditions: (a) CS2 (2.0 equiv), KOH (2.1 equiv), EtOH, reflux, 16 h, 80–95%; (b) ethyl isonipecotate
(1.1 equiv), toluene, reflux, 16 h, 80–90%; (c) 2.5 N NaOH (2.0 equiv), MeOH/H2O 9:1, rt, 8 h, 90–98%; (d) (1S,3S)-3-aminocyclohexylmethanol (36), HBTU (1.2 equiv), Et3N
(1.5 equiv), DMF, rt, 16 h, 70–90%.
Cl
Cl
N
N
R
a
b
N
CO2Et
N
CO2Et
+
O
O
Br
37
38
Cl
Cl
O
N
O
N
c, d
N
N
CO2Et
R
R
HN
O
39
19-21
HO
Scheme 2. Preparation of compounds 19–21. Reagents and conditions: (a) Br2 (1 equiv), AcOH, rt, 1 h, 99%; (b) olefin, 2.5 M 9-BBN/THF, rt, 15 min; 38, Pd(dppf)2Cl2 (cat.),
K3PO4, 60 °C, 4 h, 75–80%; (c) 2.5 N NaOH (2.0 equiv), MeOH/H2O 9:1, rt, 8 h, 90–98%; (d) 36, HBTU (1.2 equiv), Et3N (1.5 equiv), DMF, rt, 16 h, 70–90%.
Cl
Ar
Cl
N
N
Replacing the alkyl group at C(6) with a chloro group produced
analog 22. This compound showed potent enzyme activity. With
its lower clogP of 2.4, we had hoped that this compound would
have a desirable cell-to-enzyme ratio. Unfortunately, compound
22 showed a poor cell-to-enzyme ratio of 43 however, substituting
the C(6) chloro group in 22 for trifluoromethyl gave compound 23,
which showed enhanced activity in both the enzyme and cell as-
says. In this case, the lower clogP of compound 23 translated to
a lower cell-to-enzyme ratio of 5. Difluoromethyl compound 24
also showed good enzyme and cell activity although the compound
showed higher in vitro clearance in HLM. In general, it was found
that a chlorine atom at C(5) and a non-polar substituent at C(6)
had a synergistic effect on mPGES-1 activity.
Continuing our search for potent compounds with reduced
clearance, we next considered substituted phenyl rings at the
C(6)-position (25–29, Table 2). We were encouraged to find that
these analogs were significantly more potent in the mPGES-1 en-
zyme assay than many of the alkyl-substituted analogs. Due to
their high clogP’s, many of the phenyl-substituted analogs showed
a high cell-to-enzyme ratios (27–80). Despite their high clogP
(3.5–4.4) and high cell-to-enzyme ratios, many of the aryl-analogs
(25, 26, 28, 29) showed very potent activity in the HWB cell assay
and in general these analogs showed a range of in vitro metabolic
stabilities in HLM.
a
N
N
CO2Et
CO2Et
O
O
Br
40
38
Cl
Ar
O
N
O
b, c
N
HN
25-31
HO
Scheme 3. Preparation of compounds 25–31. Reagents and conditions: (a) aryl
boronic acid (1.1–1.5 equiv), Pd(P(Ph3)4 (cat.); Na2CO3 (2 equiv), DMF, reflux, 8 h,
60–85%; (b) 2.5 N NaOH (2.0 equiv), MeOH/H2O 9:1, rt, 8 h, 90–98%; (c) 36, HBTU
(1.2 equiv), Et3N (1.5 equiv), DMF, rt, 16 h, 70–90%.
Cl
N
O
a, b
c, d
37
N
CO2Et
X
41
, X = I
42, X = CF3
Cl
F3C
O
N
O
N
HN
Pyridyl-substituted analogs 30 and 31 were prepared in an at-
tempt to lower the clogP while maintaining the excellent potency
seen in the phenyl-substituted analogs (Table 2). Unfortunately,
both analogs showed a right shift in enzyme and cell potency,
and both analogs showed high in vitro clearance.
Rat pharmacokinetic (PK) data for compound 12 and a selected
group of potent analogs from Table 2 is shown in Table 3. Com-
pound 12 showed good bioavailability, however, its low volume
of distribution and modest half life resulted in an unfavorable
23
OH
Scheme 4. Preparation of compound 23. Reagents and conditions: (a) NIS
(1.1 equiv), AcOH, rt, 16 h, 91%; (b) methyl 2,2-difluoro-2-(fluorosulfonyl)acetate
(3.0 equiv), CuI (1.5 equiv), DMF, 80 °C, 24 h; H2, 10% Pd/C, MeOH, 24 h, 40%; (c)
2.5 N NaOH (2.0 equiv), MeOH/H2O 9:1, rt, 8 h, 90–98%; (d) 36, HBTU (1.2 equiv),
Et3N (1.5 equiv), DMF, rt, 16 h, 70–90%.
increased from 2.2 to 3.7 for methyl compound 18 to isobutyl ana-
log 21, the cell-to-enzyme ratio also increased from 3 to 51.
Please cite this article in press as: Arhancet, G. B.; et al. Bioorg. Med. Chem. Lett. (2012), http://dx.doi.org/10.1016/j.bmcl.2012.11.109

G. B. Arhancet et al. / Bioorg. Med. Chem. Lett. xxx (2012) xxx–xxx
5
Cl
Cl
N
N
a
b
37
N
CO2Et
N
CO2Et
H
O
O
OHC
F
F
44
43
Cl
O
N
O
N
H
c, d
HN
F
F
24
OH
Scheme 5. Preparation of compound 24. Reagents and conditions: (a) HMTA (2.2 equiv), CF3CO2H, 120 °C, 18 h, 69%; (b) DAST (1.1 equiv), CH2Cl2, À78 °C, 6 h, 80%; (c) 2.5 N
NaOH (2.0 equiv), MeOH/H2O 9:1, rt, 8 h, 90–98%; (d) 36, HBTU (1.2 equiv), Et3N (1.5 equiv), DMF, rt, 16 h, 70–90%.
Table 3
Table 4
PGE2 inhibition
Compound
Rat in vivo pharmacokinetic propertiesa
Compound
Guinea pig carrageenan stimulated air pouch
CL (mL/min/kg) Vdss (L/kg) Half-life (h) MRT (h) F (%)
Dose (mg/kg)
Plasma concd (
lM)
PGE2 (%) inh.
12
19
23
24
26
28
29
17
82
34
60
12
28
14
0.72
2.1
14
1.4
3.0
2.4
6.4
0.51
0.40
19.3
1.4
3.7
1.4
6.7
0.70
0.44
7.22
0.39
4.42
6.26
7.78
66
42
Naproxen
10
10
10
10
153.9
1.6
1.2
60
66
63
52
23
26
29
0.51
59
17
86
Carrageenan-simulated air pouch performed as described in Ref. 17.
a
these data, compound 26 was selected as a clinical candidate. Cur-
rently, 26 is undergoing additional in vitro and in vivo safety stud-
ies in anticipation of human studies.
Clearance (CL), volume of distribution (Vss) and half-life pharmacokinetic data
were determined following a 1.0 mg/kg dose iv in Sprague–Dawley rats. Mean
residence time (MRT) and bioavailability (%F) were determined following a 1.0 mg/
kg po dose in Sprague–Dawley rats.
In summary, we have described a novel class of benzoxazoles as
mPGES-1 inhibitors with excellent potency and selectivity. Careful
modification of the right-hand portion of HTS-hit 5 led to com-
pound 12. Introduction of fluoroalkyl and aryl substituents at the
C(6)-position of the benzoxazole ring led to the discovery of com-
pounds 23, 26, and 29 with a remarkably improved potency in the
HWB assay and favorable PK profiles. These compounds were fur-
ther profiled in a Guinea pig carrageenan-stimulated air pouch
model of inflammation. As a result of its in vitro potency and selec-
tivity, in vivo efficacy, pharmacokinetic properties, and preclinical
safety profile, benzoxazole 26 (PF-4693627) was selected as a clin-
ical candidate for the treatment of inflammation caused by OA and
RA.
Table 5
Compound 26 selectivity against relevant enzymes as described in Ref. 18
Compound
HWB-1483 IC50
mPGES-1 PGDS TXAS 5-LO mPGES-1
0.18 >50 >50 >50 0.006
(
l
M)
Fetal fibroblast IC50 (lM)
COX-2
>10
26
metabolic profile. Substitution with an ethyl group at the C(6)-po-
sition (19) resulted in very high clearance while replacement with
trifluromethyl produced a more metabolically stable compound
(23). Aryl substitution at C(6) produced compounds (26, 28, 29)
with good clearance and improved half-lives compared to 12 while
maintaining good bioavailability.
Acknowledgments
To date, no selective mPGES-1 inhibitors have been reported in
the published literature that are active on the rodent enzyme.
Hence, inflammation models historically used to evaluate NSAIDs
and COX-2 inhibitors could not be utilized. As a result, the carra-
geenan-stimulated air pouch model was established in the guinea
pig in order to assess in vivo efficacy data.17 Compounds 23, 26,
and 29 were tested in the air pouch inflammation model at
10 mg/kg administered orally. The potent NSAID, Naproxen, was
used as a reference. We were pleased to find that all three benzox-
azole compounds showed good inhibition of PGE2 production in
We thank Joe T. Collins for analytical support in the preparation
of compound 36, and Cathleen E. Hanau and Hwang-Fun Lu for the
preparation of valuable intermediates. We thank Pfizer-St. Louis
Chemistry Management for encouraging the publication of this
work.
References and notes
1. (a) Samad, T. A.; Moore, K. A.; Sapirstein, A.; Billet, S.; Allorhorme, A.; Poole, S.;
Bonventre, J. V.; Woolf, C. J. Nature 2001, 410, 471; (b) Funk, C. D. Science 1871,
2001, 294; (c) Kobayashi, T.; Narumiya, S. Prostaglandins Other Lipid Mediators
2002, 68; (d) Goulet, J. L.; Pace, A. J.; Key, M. L.; Byrum, R. S.; Nguyen, M.; Tilley,
S. L.; Morham, S. G.; Langenbach, R.; Stock, J. L.; McNeish, J. D.; Smithies, O. J.
Immunol. 2004, 173, 1321.
this assay while no inhibition of 6-keto-PGF1 was observed
a
(Table 4).
Overall compounds 26 and 29 had the best combination of
in vitro cell potency, PK, in vivo efficacy and pharmaceutical prop-
erties. Furthermore, since the synthesis of compound 26 was more
attractive than compound 29; it was chosen for further profiling.
The selectivity of 26 was profiled against several relevant human
targets in the following assays: (a) HWB/1483 for selectivity
against TXAS, PGDS, 5-LOX, 15-LOX and 12-LOX; (b) Human fetal
fibroblasts for selectivity against COX-2. Results are summarized
in Table 5. The data show that 26 is selective against relevant hu-
man enzymes. Compound 26 was also submitted to a broad panel
screen (CEREP) and no significant findings were observed. Based on
2. Jachak, S. M. Curr. Opin. Investig. Drugs 2007, 8, 411.
3. Brenneis, C.; Coste, O.; Schmidt, R.; Angioni, C.; Popp, L.; Nusing, R. M.; Becker,
W.; Scholich, K.; Geisslinger, G. J. Cell Mol. Med. 2008, 12, 639.
4. Friesen, R. W.; Mancini, J. A. J. Med. Chem. 2008, 51, 4059.
5. Riendeau, D.; Aspiotis, R.; Ethier, D.; Gareau, Y.; Grimm, E. L.; Guay, J.; Guiral, S.;
Juteau, H.; Mancini, J. A.; Methot, N.; Rubin, J.; Friesen, R. W. Bioorg. Med. Chem.
Lett. 2005, 15, 3352.
6. Pelcman, B.; Olofsson, K.; Schaal, W.; Kalvins, I.; Katkevics, M.; Ozola, V.; Suna,
E. PCT Int. Appl. WO 2007042816, 2007.
7. Cote´, B.; Boulet, L.; Brideau, C.; Claveau, D.; Ethier, D.; Frenette, R.; Gagnon, M.;
´
Giroux, A.; Guay, J.; Guiral, S.; Mancini, J.; Martins, E.; Masse, F.; Methot, N.;
Riendeau, D.; Rubin, J.; Xu, D.; Yu, H.; Ducharme, Y.; Friesen, R. W. Bioorg. Med.
Chem. Lett. 2007, 17, 6816–6820.
Please cite this article in press as: Arhancet, G. B.; et al. Bioorg. Med. Chem. Lett. (2012), http://dx.doi.org/10.1016/j.bmcl.2012.11.109

6
G. B. Arhancet et al. / Bioorg. Med. Chem. Lett. xxx (2012) xxx–xxx
8. Kablaoui, N.; Patel, S.; Shao, J.; Demian, D.; Hoffmaster, K.; Berlioz, F.; Vazquez,
Pouches were allowed to develop for 2 days from initial injection. Animals
(6 per group) were fasted with free access to water for 16 to 24 h prior to drug
administration. Compounds or vehicle (0.5% Methelcellulose, 0.025% Tween-
20) were administered by oral gavage (1 mL) 2 h after injection into the pouch
of 2 mL of a 1% suspension of carrageenan (FMC BioPolyer, Philadelphia, PA
Cat# GP209) dissolved in saline. At 6 h post-carrageenan injection, 1.5 mL of
M. L.; Moore, W. M.; Nugent, R. A. http://dx.doi.org/10.1016/j.bbr.2011.03.031..
9. Wang, J.; Limburg, D.; Carter, J.; Mbalaviele, G.; Gierse, J.; Vazquez, M. Bioorg.
Med. Chem. Lett. 2010, 1604.
10. For the synthesis of the amino alcohols contained in compounds 10–12, see:
Walker, D. P.; Heasley, S. E.; MacInnes, A.; Anjeh, T.; Lu, H.-F.; Fobian, Y. M.;
Collins, J. T.; Vazquez, M. L.; Mao, M. K. Synlett 2011, 2959.
100 lM of arachidonate acid (NU-Chek PREP, Inc., Elysian MN) dissolved in
11. A detailed study looking at the conformation analysis of the 1,4- and 1,3-
disubstituted cyclohexane analogs is published in the adjoining paper.
12. IC50 values for inhibition of PGE2 production in the LPS-stimulated human
whole blood assay: Human whole blood was collected from healthy donors in
saline was injected into the pouch and after 10 min fluid was collected and
volume measured. The fluid was centrifuged at 800 g (Beckman Coulter,
Allegra X-15R) for 10 min at 4 °C, and the supernatants were collected for
analysis of PGE2 and 6-keto-PGF1
. PGE2 and 6-keto-PGF1a levels were
a
heparinized tubes and stimulated with 20
l
g/ml LPS in the presence of varying
quantitated by ELISA (Cayman Chemical Company, Ann Arbor, MI) by
multiplying ng/ml by total pouch volume (ml) to normalize for total pouch
ng. Each point represents the mean S.E.M. (n = 5–12). ED50 and ED80 values
for each experiment were calculated using a four-parameter sigmoid model.
Naproxen was used as a reference compound at 10 mg/kg.
concentrations of inhibitor and 1% DMSO. Following incubation overnight (20–
24 h) at 37 °C in 95% air/5% CO2, assay plates were centrifuged and the plasma
removed for quantitation of PGE2 by ELISA. Data were expressed as% inhibition
of PGE2 with 10 lM indomethacin used for maximal inhibition of PGE2.
Inhibitors were evaluated at 10 concentrations in duplicate with threefold
serial dilutions and IC50 values calculated using a 4-parameter logistic model.
Each IC50 value represents the average of n P4 experiments. Standard error of
the IC50 was generally less than 30%.
18. HWB/1483 assay: Human head and neck squamous cell carcinoma, 1483 cells,
were maintained in cultures at 37 °C in 95% air, 5% CO2 atmosphere in High
Glucose D-MEM (Invitrogen, Grand Island, NY) containing extra 10 mM HEPES
buffer (Invitrogen), Penicillin–Streptomycin–Glutamine (Invitrogen), and 15%
FBS (Sigma, St. Louis, MO). On the day of the experiments, 1483 cells were
trypsinized, centrifuged at 400Âg, and the pellets were resuspended with
human whole blood (HWB). HWB was collected from healthy human donors in
10 ml heparinized tubes (Vacutainer tubes, Becton Dickenson, Franklin Lakes,
NJ). The cell mixtures were pre-incubated with compounds for 15 min (at a
final concentration of 1% DMSO) followed by incubation with 30 mM
arachidonic acid. The levels of eicosanoids were determined by LC/MS/MS.
Human fetal fibroblasts assay: Cells were stimulated with 1 ng/ml IL-1b for
24 h. The conditioned media were removed, and cells were washed twice with
serum-free media. Cells were then treated with the serum-free media
13. Assay conditions adopted from published protocols, see: (a) Riley, R. J.;
McGinnity, D. F.; Austin, R. P. Drug Metab. Dispos. 2005, 33, 1304; (b) Obach, R.
S. Drug Metab. Dispos. 1999, 27, 1350; (c) Hosea, N. A.; Collard, W. T.; Cole, S.;
Maurer, T. S.; Fang, R. X.; Jones, H.; Kakar, S. M.; Nakai, Y.; Smith, B. J.; Webster,
R.; Beaumont, K. J. Clin. Pharmacol. 2009, 49, 513.
14. Miyaura, N.; Ishiyama, T.; Ishikawa, M.; Suzuki, A. Tetrahedron Lett. 1986, 6369.
15. Chen, Q.; Wu, S. J. Chem. Soc., Chem. Commun. 1989, 705.
16. Duff, J. C.; Bills, E. J. J. Chem. Soc. 1987, 1932.
17. Use of the animals in these studies was reviewed and approved by the Pfizer
Institutional Animal Care and Use Committee. The animal care and use
program is fully accredited by the Association for Assessment and
Accreditation of Laboratory Animal Care, International. Male Duncan Hartley
guinea pigs (250–300 g), (Charles River Laboratories, Wilmington, MA) were
used in these studies. Animals were acclimated in the animal facility for one
week prior to experimentation. Air pouches were produced by subcutaneous
injection of 5 mL of sterile air into the intrascapular area of the back. The next
day pouches were injected with 3–5 mL of sterile air to maintain pouches.
containing either the vehicle (at
compounds (1% DMSO) for 50 min, and with 10
additional 10 min, at 37 °C in 95% air, 5% CO2 atmosphere. To measure PGF2
a
final concentration 1% DMSO) or the
l
M
arachidonic acid for
a
levels, SnCl2 was added to the cells 40 min after addition of the compounds (or
10 min before addition of arachidonic acid). PG levels in the supernatants were
assayed by ELISA.
Please cite this article in press as: Arhancet, G. B.; et al. Bioorg. Med. Chem. Lett. (2012), http://dx.doi.org/10.1016/j.bmcl.2012.11.109