Substituted phenanthrene imidazoles as potent, selective, and orally active mPGES-1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.10.033

Phenanthrene imidazole 3 (MF63) has been identified as a novel potent, selective, and orally active mPGES-1 inhibitor. This new series was developed by lead optimization of a hit from an internal HTS campaign. Compound 3 is significantly more potent than the previously reported indole carboxylic acid 1 with an A549 whole cell IC₅₀ of 0.42 μM (50% FBS) and a human whole blood IC₅₀ of 1.3 μM. It exhibited a significant analgesic effect in a guinea pig hyperalgesia model when orally dosed at 30 and 100 mg/kg.

Full text of DOI:10.1016/j.bmcl.2007.10.033

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 17 (2007) 6816–6820
Substituted phenanthrene imidazoles as potent, selective,
and orally active mPGES-1 inhibitors
*
´
Bernard Coˆte, Louise Boulet, Christine Brideau, David Claveau, Diane Ethier,
´
´
Richard Frenette, Marc Gagnon, Andre Giroux, Jocelyne Guay, Sebastien Guiral,
´ ´
´
´
Joseph Mancini, Evelyn Martins, Frederic Masse, Nathalie Methot, Denis Riendeau,
Joel Rubin, Daigen Xu, Hongping Yu, Yves Ducharme and Richard W. Friesen
Merck Frosst Centre for Therapeutic Research, 16711 Trans Canada Hwy, Kirkland, Que., Canada H9H 3L1
Received 21 August 2007; revised 4 October 2007; accepted 8 October 2007
Available online 17 October 2007
Abstract—Phenanthrene imidazole 3 (MF63) has been identified as a novel potent, selective, and orally active mPGES-1 inhibitor.
This new series was developed by lead optimization of a hit from an internal HTS campaign. Compound 3 is significantly more
potent than the previously reported indole carboxylic acid 1 with an A549 whole cell IC50 of 0.42 lM (50% FBS) and a human whole
blood IC50 of 1.3 lM. It exhibited a significant analgesic effect in a guinea pig hyperalgesia model when orally dosed at 30 and
100 mg/kg.
ꢀ 2007 Elsevier Ltd. All rights reserved.
CO2H
Prostaglandin E2 (PGE2) is widely recognized as a key
mediator in fever, pain, and inflammatory response.1
The biosynthesis of PGE2 involves the release of arachi-
Arachidonic Acid (AA)
PGH Synthase
donic acid (AA) from membrane glycerophospholipids
followed by several enzymatic transformations (Fig. 1).
First, AA is converted to PGH2 by the cyclooxygenases
NSAID, Coxib
(COX-1/COX-2)
O
PGI
Synthase
TX
CO2H
PGI2
(Prostacyclin)
TXA2
(COX-1/COX-2) and then, this intermediate prostanoid
is further metabolized by downstream specific synthases
to PGE2, PGI2, PGD2, PGF2a, and thromboxane A2
(TXA2). Non-steroidal anti-inflammatory drugs (NSA-
IDs) and selective COX-2 inhibitors (coxibs) intervene
in this prostaglandin pathway at the level of PGH syn-
thase. These drugs exert their anti-inflammatory and
analgesic properties mainly through PGE2 and PGI2
suppression.2 As a potential way to improve the current
therapeutic window of NSAIDs and coxibs, we investi-
gated a new mechanism for selective PGE2 modulation
through downstream PGE synthase inhibition. To this
end, the inducible microsomal PGE synthase-1
(mPGES-1) represents a potential novel target for drug
development in the treatment of inflammatory diseases.3
Supporting this hypothesis are the observations that
mice deficient in mPGES-1 are protected from fever4
Synthase
O
OH
PGF
PGD
Synthase
PGH2
Synthase
PGE
Synthase
mPGES-1
inhibitors
PGF2α
PGD2
O
CO2H
HO
OH
PGE2
Figure 1. The prostaglandin pathway.
and chronic inflammation,5 and are not predisposed to
thrombogenesis or hypertension.6 Herein we report the
identification of the first orally active mPGES-1 inhibi-
tor in a model of hyperalgesia.7
We recently reported that the indole carboxylic acid 1 is a
potent and selective mPGES-1 inhibitor (Fig. 2).8 Despite
its high intrinsic inhibitory potency (IC50 = 0.003 lM) on
Keywords: mPGES-1; PGE2; Inflammation; Prostanoid.
*
Corresponding author. Tel.: +1 514 428 2683; fax: +1 514 428
8670; e-mail: bernard_cote@merck.com
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.10.033

ˆ ´
B. Cote et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6816–6820
6817
CH3
hits, the JAK kinase12 inhibitor azaphenanthrenone 2a
was identified as a potent lead structure (mPGES-1
IC50 = 0.14 lM). Further testing of analogs of 2a from
our sample collection revealed a strong preference for
an ortho-substituted phenyl ring at the R position (Table
1).13 It was also found that the active analogs tend to be
substantially less shifted in the A549 whole cell assay
(50% FBS) when compared to indole carboxylic acid 1.
F
CH3
Cl
N
CO2H
Cl
F
O
N
N
H
F
HN
1
2a
SAR studies and lead optimization were conducted with
the initial objective of dialing out the JAK kinase activ-
ity of 2 while maintaining the mPGES-1 potency of this
new class of inhibitors. Internal data from the JAK ki-
nase program revealed that the pyridone moiety was
essential for good potency. The azaphenanthrenone
template was therefore replaced by the phenanthrene
ring system through the simple assemblage of phenan-
threnequinone and 2-chloro-6-fluoro benzaldehyde to
produce phenanthrene imidazole 4 (Scheme 1).14 As pre-
sented in Table 2, compound 4 is intrinsically four times
more potent than the lead compound 2a, maintains
good cellular activity, and is still less shifted than the
acid 1 in presence of proteins. When PGF2a inhibition
was measured in the A549 whole cell assay (2% and
50% FBS), the phenanthrene imidazole 4 was found to
be at least 10-fold selective for the inhibition of PGE2.
Cl 6'
NC
N
N
H
NC
9'
3 (MF63)
Figure 2. mPGES-1 inhibitors.
the recombinant human mPGES-1 enzyme,9 a 2000-fold
shift was observed with 1 for PGE2 inhibition in the IL-
1b stimulated A549 whole cell assay10 in the presence of
50% fetal bovine serum (FBS, see Table 1).
This poor whole cell potency precluded the use of this
compound for in vivo proof-of-concept studies in mod-
els of inflammation. Moreover, efforts to improve the
cellular activity of analogs of indole carboxylic acid 1
were unsuccessful and a search for a new chemical class
of inhibitors was undertaken. An HTS campaign was
conducted11 using the enzyme assay and among several
Using the phenanthrene imidazole template as a refer-
ence (Table 3, compounds 4 and 5), we then investigated
the minimal structural requirements for mPGES-1 activ-
ity. In general, alteration of the phenanthrene motif for
compounds 6, 7, 9,15 10, and 11 resulted in a significant
decrease in inhibitory activity. Substitution of the imid-
azole (12) or replacement by other heterocycles (13–15)
was also detrimental for mPGES-1 inhibition.
Table 1. mPGES-1 inhibition IC50 for azaphenanthrenones 2
Having recognized the importance of the ortho-substitu-
ent of the biaryl system, we next oriented our SAR
investigation toward the optimization of this critical
portion of the molecule (Table 4). The phenyl ring was
F
N
R
O
N
H
HN
2
Cl
Cl
c
Compound
R
IC50 (lM)
O
O
N
a
mPGES-1a
inhibition
A549, 50%
FBSb PGE2
OHC
+
N
H
F
F
1
0.003
0.14
>10
>10
>10
>10
>10
0.53
2.5
5.8
1.6
—
2a
2b
2c
2d
2e
2f
2Cl, 6F-phenyl
H
n-Propyl
4
Scheme 1. Reagents and condition: (a) NH4OAc, AcOH, 100 ꢁC.
—
—
tert-Butyl
c-Pentyl
Phenyl
—
—
Table 2. Comparative profile of in vitro activity and selectivity for
inhibitors 1, 2a, and 4
2g
2h
2i
2Cl-phenyl
3Cl-phenyl
4Cl-phenyl
4MeO-phenyl
2Br-phenyl
2CF3-phenyl
2,6-Di-Cl-phenyl
0.56
5.3
0.82
—
Compound
mPGES-1
IC50 (lM)
A549 whole cell, IC50a(lM)
2% FBS 50% FBS
PGE2 PGF2a PGE2 PGF2a
0.41
>10
0.33
0.56
0.10
2j
2k
2l
0.65
5.1
9.8
1
0.003
0.14
0.036
0.27
0.21
0.35
2.4
1.1
5.2
5.8
1.6
2.6
>50
19
2m
2a
4
a For assay conditions, see Ref. 9.
b For assay conditions, see Ref. 10.
c Values are means of at least two experiments.
>30
a Values are means of at least two experiments.

ˆ ´
B. Cote et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6816–6820
6818
Table 3. Structural modifications of the phenanthrene imidazole motif
Table 3 (continued)
Compound
Structure
mPGES-1
inhibition
Compound
Structure
mPGES-1
inhibition
a
a
IC50 (lM)
IC50 (lM)
Cl
Cl
Cl
N
S
N
14
15
>10
4
0.036
0.087
N
H
F
Cl
N
0.71
5
N
H
N
H
a Values are means of at least two experiments.
Cl
Cl
N
6
7
6.6
2.0
N
H
Table 4. SAR at the ortho-position of the biaryl system
N
R1
N
N
H
F
N
H
R2
Cl
N
16
8
0.073
Compound
R1
R2
IC50 (lM)
N
H
F
mPGES-1
A549 whole cell
assaya 50% FBS
Cl
PGE2
PGF2a
Cl
N
5
Cl
Br
H
0.087
0.051
0.075
0.036
0.047
0.012
0.009
0.045
0.019
0.017
0.021
1.0
1.6
6.4
9
>10
8.0
16a
16b
4
H
17.1
26.7
>30
>50
>30
>30
>30
>30
>30
>30
N
CN
Cl
H
3.0
2.2
H
Cl
F
16c
16d
16e
16f
16g
16h
16i
Cl
Cl
Cl
CN
Br
F
1.1
Cl
CN
CN
Br
0.76
0.33
2.7
N
10
N
H
Br
1.6
F
CN
CN
Br
F
0.57
1.1
Cl
F
a Values are means of at least two experiments.
N
11
1.3
N
H
mono- and bis-substituted with several groups such as
F, Cl, Br, and CN. We first observed that, although po-
tent, mono-substituted analogs 5, 16a, and 16b were
generally less selective in the cellular assay. Among sev-
eral bis-substituted analogs, the optimal pattern was ob-
tained with R1 = R2 = CN (16e) where the intrinsic and
the cellular activity as well as the selectivity (PGE2/
PGF2a) were significantly improved.
Cl
N
N
12
>10
F
CH3
After optimizing the ortho-substituents of the biaryl sys-
tem, we next studied the effect of substitution on the
phenanthrene core of the molecule. In vitro potency
could be favorably modulated by the simple incorpora-
tion of a chloro substituent at the 60 position as exempli-
fied with the inhibitor 3 (Fig. 2). Profile comparison is
Cl
N
O
13
0.70
F

ˆ ´
B. Cote et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6816–6820
6819
Table 5. Comparative data on potency and selectivity of mPGES-1
inhibitors 1 and 3
HO
O
N
H2N
a
Enzyme or cell assay
IC50 (lM)
1
3
b
a
Human mPGES-1
Rat mPGES-1
0.003
0.13
N/A
>1
0.001
>30
0.0009
>30
3.0
Guinea pig mPGES-1
Human mPGES-2
TX synthase
18
19
17
0.95
5.8
A549 cells, PGE2, 50% FBS
A549 cells, PGF2a, 50% FBS
Human whole blood, PGE2
Human whole blood, TXB2
Human JAK2
0.42
>50
1.3
Cl
Cl
>50
>40
>40
N/A
N/A
O
O
d
e
c
>40
0.1
Human JAK3
>10
a Values are means of at least two experiments.
21
20
Cl
Cl
presented in Table 5 for the indole carboxylic acid 1 and
the phenanthrene imidazole 3. Both compounds have
similar intrinsic mPGES-1 activity with IC50s of 0.003
and 0.001 lM, respectively. They are selective against
mPGES-216 and thromboxane synthase. Compound 3
is 14-fold more potent than the initially reported inhib-
itor 1 with an IC50 of 0.42 lM in the A549 whole cell as-
say (50% FBS). Most importantly, the phenanthrene
imidazole 3 is the first reported mPGES-1 inhibitor to
demonstrate potency in a human whole blood (HWB)
assay. When freshly collected blood was stimulated with
LPS, compound 3 selectively inhibited the production of
PGE2 with an IC50 of 1.3 lM with no concomitant
TXB2 inhibition. It is worth mentioning that this
HWB IC50 is comparable to the ones of marketed cox-
ibs.17 With this markedly improved potency and a suit-
able pharmacokinetic profile,18 compound 3 may serve
as a good tool for in vivo proof-of-concept studies.19
Gram quantities of this new potent mPGES-1 inhibitor
were prepared according to Scheme 2.
Br
Br
NC
NC
N
N
f
N
H
N
H
3
22
Scheme 2. Reagents and conditions: (a) H2NOHHCl, EtOH, reflux,
59%, pyridine; (b) PPA, 100 ꢁC, concd HCl, MeOH, reflux, 92%; (c)
CuCl2, t-BuONO, MeCN, 65 ꢁC, 52%; (d) CrO3, AcOH, 100 ꢁC, 88%;
(e) 2,6-dibromobenzaldehyde, NH4OAc, AcOH, 100 ꢁC, 85%; (f)
CuCN, DMF, 80 ꢁC, 80%.
zyme activity was also observed for mice. To
demonstrate in vivo efficacy, a LPS-induced hyperalge-
sia model in guinea pig was therefore developed7 know-
ing the ability of compound 3 to inhibit this enzyme
(guinea pig IC50 = 0.9 nM). When orally dosed at 30
and 100 mg/kg, inhibitor 3 prevented the hyperalgesic
response in guinea pig in a dose-dependent manner with
a complete blockade of hyperalgesia at 100 mg/kg.21
This constitutes the first reported example of in vivo effi-
cacy for a selective mPGES-1 inhibitor in a pre-clinical
model. A full description of this in vivo pharmacological
profile will be published elsewhere.7
The synthesis of 3 starts with commercially available 3-
acetylphenanthrene 17. This methyl ketone was refluxed
in ethanol in the presence of hydroxylamine hydrochlo-
ride to afford the corresponding oxime 18 in 59% yield.
Phenanthreneamine 19 is obtained by a Beckmann rear-
rangement with polyphosphoric acid at 100 ꢁC followed
by amide hydrolysis under acidic conditions in an over-
all yield of 92%. Conversion of the amine to the aryl
chloride 20 was performed with tert-butyl nitrite in the
presence of copper (II) chloride in 52% yield.20 Phenan-
threne 20 was then treated with chromium (VI) oxide in
acetic acid at 100 ꢁC to afford the corresponding qui-
none 21 in 88% yield. The core of the molecule was
assembled using the same transformation described
above and phenanthrene imidazole 22 was obtained in
85% yield. Finally, the cyano groups were incorporated
by reacting the dibromo precursor 22 with CuCN in
DMF at 80 ꢁC to afford the phenanthrene imidazole bis-
nitrile 3 in 80% yield.
In summary, phenanthrene imidazole 3 (MF63) has been
identified as a new potent, selective, and orally active
mPGES-1 inhibitor. This new series was developed by
lead optimization of an internal HTS hit. Compound 3
shows in vitro superiority over the previously reported in-
dole carboxylic acid 1, both in the A549 whole cell assay
(50% FBS, IC50 = 0.42 lM) and in the human whole
blood assay (IC50 = 1.3 lM). Finally, in a guinea pig
hyperalgesia model, inhibitor 3 demonstrated a signifi-
cant analgesic effect.
References and notes
As disclosed in Table 5, the mPGES-1 inhibitor 3 is not
active against the rat enzyme which precluded the use of
well-established rat inflammatory pain models. This ma-
jor discrepancy between the human and the rodent en-
1. Funk, C. D. Science 2001, 294, 1871.
2. (a) Portanova, J. P. et al. J. Exp. Med. 1996, 184, 883; (b)
Pulichino, A. M. et al. J. Pharmacol. Exp. Ther. 2006,
319, 1043; (c) Murata, T. et al. Nature 1997, 388, 678.

ˆ ´
B. Cote et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6816–6820
6820
3. (a) Jakobsson, P. J.; Thoren, S.; Morgenstern, R.; Sam-
13. SAR of the 4-substituted analogs such as 2i showed that
although potent, these inhibitors were not selective for
PGE2 inhibition.
14. Compound 4 was assayed against JAK kinases: JAK2
IC50 = 0.9 lM; JAK3 IC50 > 10 lM.
uelsson, B. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 7220;
(b) 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; (c)
Murakami, M.; Kudo, I. Prog. Lipid Res. 2004, 43, 3.
4. Engblom, D. Nat. Neurosci. 2003, 6, 1137.
5. Trebino, C. E. Proc. Natl. Acad. Sci. U.S.A. 2003, 100,
9044.
6. Cheng, Y.; Wang, M.; Yu, Y.; Lawson, J.; Funk, C. D.;
FitzGerald, G. A. J. Clin. Invest. 2006, 116, 1391.
7. Xu, Daigen, et al., submitted for publication.
8. Riendeau, D. et al. Bioorg. Med. Chem. Lett. 2005, 15,
3352.
15. For a rigorous SAR analysis, compound 9 should be
compared with 16c.
16. Tanikawa, N.; Ohmiya, Y.; Ohkubo, H.; Hashimoto, K.;
Kangawa, K.; Kojima, M.; Ito, S.; Watanabe, K. Bio-
chem. Biophys. Res. Commun. 2002, 291, 884.
17. Etoricoxib HWB IC50 = 1.1 lM (PGE2) Riendeau, D.
et al. J. Pharmacol. Exp. Ther. 2001, 296, 558.
18. Compound 3 is bioavailable in rat (F = 68%) with a CMax
of 12 lM when dosed orally at 20 mg/kg.
9. Mancini, J. A.; Blood, K.; Guay, J.; Gordon, R.; Claveau,
D.; Chan, C. C.; Riendeau, D. J. Biol. Chem. 2001, 276,
4469.
10. Thoren, S.; Jakobsson, P. Eur. J. Biochem. 2000, 267,
6428.
19. Although some residual JAK2 activity was measured,
compound 3 is still 100· more active against mPGES-1.
Since this inhibitor is also shifted >1000· in the whole
blood assay, we believe that the JAK2 residual activity
should not interfere in the in vivo model.
´
11. Masse, F.; Guiral, S.; Fortin, L. J.; Cauchon, E.; Ethier, D.;
20. Eirin, A.; Fernandez, F.; Gonzalez, C.; Lopez, C. Org.
Prep. Proc. Int. 1992, 245, 509.
21. At 30 mg/kg, a 50% blockade of the hyperalgesia was
measured relative to vehicle treated animal and the no-
LPS (saline) group.
Guay, J.; Brideau, C. J. Biomol. Screen. 2005, 106, 599.
12. Thompson, J. E. et al. Bioorg. Med. Chem. Lett. 2002, 12,
1219, For compound 2a, JAK2 IC50 = 0.2 nM; JAK3
IC50 = 0.6 nM..