Discovery and characterization of a potent and selective EP4 receptor antagonist

Article abstract of DOI:10.1016/j.bmcl.2015.05.091

Abstract EP₄ is a prostaglandin E₂ receptor that is a target for potential anti-nociceptive therapy. Described herein is a class of amphoteric EP₄ antagonists which reverses PGE₂-induced suppression of TNFα prod

Full text of DOI:10.1016/j.bmcl.2015.05.091

Bioorganic & Medicinal Chemistry Letters xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and characterization of a potent and selective EP receptor
4
antagonist
⇑
Matthew A. Schiffler , Srinivasan Chandrasekhar, Matthew J. Fisher, Anita Harvey, Steven L. Kuklish,
Xu-Shan Wang, Alan M. Warshawsky, Jeremy S. York, Xiao-Peng Yu
Lilly Research Laboratories, A Division of Eli Lilly and Company, Indianapolis, IN 46285 USA
a r t i c l e i n f o
a b s t r a c t
Article history:
EP4 is a prostaglandin E2 receptor that is a target for potential anti-nociceptive therapy. Described herein
Received 20 April 2015
Revised 27 May 2015
Accepted 29 May 2015
Available online xxxx
is a class of amphoteric EP
human whole blood. From this class, a potent and highly bioavailable compound (6) has been selected for
potential clinical studies. EP
this compound are included.
4 2
antagonists which reverses PGE -induced suppression of TNFa production in
4
binding and functional data, selectivity, and pharmacokinetic properties of
Ó 2015 Elsevier Ltd. All rights reserved.
Keywords:
4
EP antagonist
Prostaglandin pathway
Dog bioavailability
Amphoteric
Prostaglandin E
2
(PGE
2
) is a prostanoid implicated in a variety
1
2
of disease states including inflammation, cancer, and bronchitis.
PGE operates through a family of rhodopsin-like 7TM G-protein
coupled receptors known as EP1–4
reported in 1994 by Lydford and McKechnie. It has been demon-
strated in rats that EP is upregulated in dorsal root ganglia during
CFA-induced peripheral inflammatory events, and that both
2
3
.
4
The EP receptor was first
4
4
administration of an EP
decrease pain sensitivity in a thermal withdrawal latency model.
Therefore, EP
4
antagonist and knockdown of EP
4
5
4
antagonism is hypothesized to constitute a potential
pain management therapy for patients suffering from inflamma-
6
tory conditions such as osteoarthritis. Related therapeutic
approaches include inhibition of COX-2 or mPGES-1;8 however,
7
selective EP
by not directly interrupting PGE
mal function of the EP
4
antagonism could carry reduced gastrointestinal risk
synthesis, thereby enabling nor-
2
9
2
receptor. Thus, widespread interest has led
10
to the discovery of varied EP
4
antagonists. Notably, CJ-023423
1
1
12
13
4
Figure 1. Structures of known EP antagonists.
(
1), PGN-1531 (2), and ER-886046 (3) (Fig. 1) have been stud-
ied extensively.
PGE (4, Fig. 2) is the endogenous ligand of EP
known EP antagonists contain a carboxylic acid or another func-
2
4 2
. Like PGE , most
4
tional group of similar acidity. Through our own studies, we dis-
covered 5 (Table 1), which contained not only a carboxylic acid,
Abbreviations: CFA, complete Freund’s adjuvant; CHO, Chinese hamster ovary.
Corresponding author. Tel.: +1 317 433 4994.
E-mail address: schifflerma@lilly.com (M.A. Schiffler).
⇑
2
Figure 2. Prostaglandin E .
http://dx.doi.org/10.1016/j.bmcl.2015.05.091
960-894X/Ó 2015 Elsevier Ltd. All rights reserved.
0
Please cite this article in press as: Schiffler, M. A.; et al. Bioorg. Med. Chem. Lett. (2015), http://dx.doi.org/10.1016/j.bmcl.2015.05.091

2
M. A. Schiffler et al. / Bioorg. Med. Chem. Lett. xxx (2015) xxx–xxx
Table 1
SAR of (phenoxyethyl)piperidine EP
4
antagonistsa
c
HWB IC50c
Compd MW of neutral form
clogD @ pH
7.4
hEP
4
filter binding K
i
hEP
IC50
4
c
filter binding
hEP
IC50
4
c
antagonism functional
b
(g/mol)
1
491.6
3.08
449 nM Â/Ä 1.85 (8)
689 nM Â/Ä 1.74 (8)
11.7 nM Â/Ä 1.98 (6)
1614 nM Â/Ä 2.13
(
120)
5
6
382.4
396.5
0.30
0.73
126 nM Â/Ä 1.10 (3)
40.6 nM Â/Ä 1.72
203 nM Â/Ä 1.19 (3)
18.1 nM Â/Ä 1.69 (3)
899 nM Â/Ä 2.73 (4)
68.8 nM Â/Ä 1.64 (12) 5.62 nM Â/Ä 1.70 (10)
126 nM Â/Ä 2.26 (20)
(
12)
7
8
9
1
410.5
408.5
421.5
414.5
1.24
0.89
0.80
1.08
129 nM Â/Ä 1.14 (3)
44.8 nM Â/Ä 1.45 (2) 77.8 nM Â/Ä 1.22 (2)
372 nM Â/Ä 1.08 (3) 621 nM Â/Ä 1.07 (3)
40.4 nM Â/Ä 1.43 (5) 71.8 nM Â/Ä 1.31 (5)
232 nM Â/Ä 1.06 (3)
14.4 nM Â/Ä 1.62 (3)
5.42 nM Â/Ä 1.76 (5)
35.8 nM Â/Ä 1.41 (2)
2.87 nM Â/Ä 1.66 (7)
291 nM Â/Ä 1.88 (9)
586 nM Â/Ä 1.14 (2)
1209 nM Â/Ä 1.74 (2)
121 nM Â/Ä 2.35 (11)
0
a
b
c
All compounds prepared as a single stereoisomer. Absolute stereochemistry as depicted.
Calculated using Marvin and calculator plugin freeware (www.chemaxon.com, ChemAxon Kft, Budapest, Hungary).
Expressed as geometric mean times or divided by the geometric sample standard deviation, with number of determinations in parentheses.
but also a (phenoxyethyl)piperidine.14 The latter moiety, which
contains a basic amine, is not found in any other currently known
Table 2
In vitro properties of 6
1
5
Resultsa
>17.5
Ki = 1.21
IC50 = 1.63
class of EP
Compounds were compared to known EP
of three assays. Affinity for the receptor was tested using a filter
4
antagonists.
Assay
4
antagonist 1 in a set
hEP
1
filter binding
K
i
l
l
M, IC50 >25
M Â/Ä 2.32,
M Â/Ä 2.31 (6)
M, IC >25 M (5)
lM (3)
hEP2 filter binding
3
binding assay with [ H]PGE
HEK293 cells stably expressing the human EP
PGE -stimulated cAMP antagonist assay in HEK293 cells was used
to determine functional activity. Then, the compounds were
evaluated for their ability to reverse the inhibition of PGE on
TNF production (a known downstream effect of activating the
EP
2
as the radioligand in recombinant
l
1
6
hEP3 filter binding
CYP1A2, CYP2B6, CYP2C19, CYP2C8,
CYP2C9, CYP2D6, CYP3A4 inhibition
Ki >12.7
l
l
50
4
receptor.
A
IC50 >10 lM for all
2
17
a
Expressed as geometric mean times or divided by the geometric sample stan-
2
dard deviation, with number of determinations in parentheses.
a
1
8
4
receptor) in LPS-stimulated human whole blood (HWB). In
each of these assays, 5 compared favorably to 1. Additionally, the
molecular weight and clogD of 5 were both lower than the corre-
sponding properties of 1. At neutral pH, 5 was predicted to exist
measurable activity was observed in antagonist mode, but not in
agonist mode. Thus, 6 was found to be a relatively weak EP antag-
2
1
9
predominantly as a zwitterion (acid pK
a
4.0, amine pK
a
7.5). For
onist as well as a potent EP antagonist. With the expectation that
4
indications such as osteoarthritic pain, rapid absorption could be
a desirable factor because it would theoretically result in quick
onset. To the extent that its amphoteric character could provide
diverse formulation options to facilitate absorption,20 5 was judged
a worthy starting point for SAR optimization.
Incorporation of a benzylic methyl group (6) increased the
potency in all three of the above assays; notably, the IC50 in
HWB was an order of magnitude lower than that of 1.
Homologation of the benzylic substituent (7) did not improve the
potency, nor did incorporation of a second substituent in the form
of a 1,1-cyclopropane (8). Addition of a cyano substituent at the
para position of the phenoxy ring (9) had a detrimental effect on
potency; however, 4-fluoro analogue 10 performed similarly to 6
in all three assays. Although 6 and 10 had similar potency profiles,
the observed selectivity would be sufficient to obviate significant
competing pharmacology, 6 was characterized further.
In an in vitro microsomal mPGES-1 assay, 6 displayed no inhibi-
2
tion up to 62.5 lM. Therefore, by not interrupting PGE synthesis, 6
was expected to exhibit distinct pharmacology from an mPGES-1
inhibitor. Antagonist 6 was tested in vitro against several CYP
enzymes (Table 2) and found to have no inhibitory activity up to
10
l
M. It was thus concluded that the risk of drug-drug interac-
M against
tions with 6 was low. Additionally, 6 was tested at 10
l
a panel of 13 receptors, four ion channels, one transporter, and
one enzyme, and showed no significant activity on any of the tar-
2
1
3
gets. Against hERG in a [ H]-astemizole binding assay, 6 had no
activity up to 100 M.
A synthesis of 6 was achieved from commercially available chi-
l
2
2
6
was chosen for further characterization due to its slightly lower
ral building blocks 11 and 12 as shown in Scheme 1. Protection of
the amine of 11 gave 13, which was subjected to palladium-cat-
alyzed carbonylation. The resulting ester 14 was deprotected to
give the amine coupling partner 15. Alkylation of the amine of
12 yielded 16, which was saponified to reveal carboxylic acid 17.
BOP-promoted coupling of 17 and 15 provided amide 18 as a single
diastereomer, which was saponified and acidified to give
hydrochloride salt 6.
molecular weight and lipophilicity.
Compound 6 was tested in filter binding assays to assess its
selectivity against the other known EP receptors (Table 2). No
detectable binding was observed with either EP
binding was observed with EP , it was approximately 25-fold
weaker than the EP binding of 6. At a 1
cAMP EP functional assays in human recombinant CHO cells,
1 3
or EP . Although
2
16
4
lM concentration in
2
Please cite this article in press as: Schiffler, M. A.; et al. Bioorg. Med. Chem. Lett. (2015), http://dx.doi.org/10.1016/j.bmcl.2015.05.091

M. A. Schiffler et al. / Bioorg. Med. Chem. Lett. xxx (2015) xxx–xxx
3
Based in part on its potency, selectivity, oral bioavailability, and
duration of exposure, 6 was selected for further studies and even-
tually chosen for potential clinical development.
In conclusion, a series of EP antagonists was identified which
4
contained both acidic and basic functionality. SAR optimization
led to 6, which displayed an IC50 of 126 nM Â/Ä 2.26 for inhibition
2
of PGE -induced TNFa reduction in an ex vivo LPS-stimulated
human whole blood assay. In vitro assays indicated that 6 was
selective for EP versus other EP receptors, mPGES-1, and a panel
4
of other targets. In dog, 6 was rapidly absorbed, highly bioavailable,
and slowly cleared. These observations contributed to the selection
of 6 as a clinical candidate.
Acknowledgments
We wish to express our appreciation to Brian Getman for partic-
ipating in generating the data for the EP biochemical assays and
4
Jennifer Weller and Dr. Daniel Mudra for critical ADME support.
We thank Joseph Haas for important assistance with statistical
analysis. Furthermore, we gratefully acknowledge many valuable
scientific discussions with Dr. María Blanco and Tatiana Vetman,
and we thank Dr. Beverly Heinz for advice and support with the
biochemical assays and with the manuscript.
Scheme 1. Synthesis of 6. Reagents and conditions: (a) Boc
9%; (b) CO (105 psi), Pd(OAc) , dppf, CH OH, Et N, CH CN, 85 °C, 72%; (c) HCl, 1,4-
dioxane, 97%; (d) b-bromophenetole, K CO , DMF, 100 °C, 64%; (e) NaOH(aq), THF,
5 °C, then HCl(aq), 81%; (f) BOP, DMF, Et N, rt, 86%; (g) NaOH(aq), CH OH, THF,
then HCl, 1,4-dioxane, 66%.
2 3 2 2
O, Et N, CH Cl , 0 °C,
9
2
3
3
3
2
3
6
3
3
Supplementary data
Compound 6 was soluble in FasSIF and FedSIF to concentrations
greater than 2.0 mg/mL, and had a solubility of 1.32 mg/mL in SGF.
Correspondingly, in vivo ADME studies revealed that 6 was rapidly
absorbed and highly bioavailable as the hydrochloride salt. In dog
Supplementary data (CEREP panel data) associated with this
article can be found, in the online version, at http://dx.doi.org/10.
1
016/j.bmcl.2015.05.091.
(Fig. 3), the tmax was 15 min, and the plasma concentration curve
after oral dosing was of a biphasic nature that closely resembled
that seen after IV dosing. These observations indicated that absorp-
tion was very rapid. The bioavailability calculated from a 10 mg/kg
PO dose and a 1 mg/kg IV dose was 81% Â/Ä 1.20 (n = 3), and the
clearance was 5.2 mL/(min kg) Â/Ä 1.15 (n = 3). After a 10 mg/kg
References and notes
1
2
.
.
Nakanishi, M.; Rosenberg, D. W. Semin. Immunopathol. 2013, 35, 123.
Sastre, B.; del Pozo, V. Mediators Inflamm. 2012. http://dx.doi.org/10.1155/
2012/645383.
PO dose, the plasma concentration of 6 remained above 1 lM in
3. Jones, R. L.; Giembycz, M. A. Br. J. Pharmacol. 2009, 158, 104.
4
5
.
.
Lydford, S. J.; McKechnie, K. Br. J. Pharmacol. 1994, 112, 133.
Lin, C.-R.; Amaya, F.; Barrett, L.; Wang, H.; Takeda, J.; Samad, T. A.; Woolf, C. J. J.
Pharmacol. Exp. Ther. 2006, 319, 1096.
all subjects for at least 12 h and above 100 nM for at least 24 h.
The combination of the ex vivo HWB IC50 and the plasma con-
centrations observed over the time course of oral dosing in dog
6. St.-Jacques, B.; Ma, W. Pain 2013, 154, 313.
7. Zarghi, A.; Arfaei, S. Iran J. Pharm. Res. 2011, 10, 655.
suggested that 6 could be a viable compound to test EP
4
antago-
8
9
.
.
Chang, H. H.; Meuillet, E. J. Future Med. Chem. 1909, 2011, 3.
Hoshino, T.; Tsutsumi, S.; Tomisato, W.; Hwang, H.-J.; Tsuchiya, T.; Mizushima,
T. J. Biol. Chem. 2003, 278, 12752.
nism as a method of treating inflammatory pain in the clinic.
1
1
0. Borriello, M.; Stasi, L. P. Pharm. Pat. Analysis 2013, 2, 387.
1. Nakao, K.; Murase, A.; Ohshiro, H.; Okumura, T.; Taniguchi, K.; Murata, Y.;
Masuda, M.; Kato, T.; Okumura, Y.; Takada, J. J. Pharmacol. Exp. Ther. 2007, 322,
686.
1
1
2. Sutton, J.; Clark, D. E.; Higgs, C.; de Groot, M. J.; Harris, N. V.; Taylor, A.; Lockey,
P. M.; Maubach, K.; Woodrooffe, A.; Davis, R. J.; Coleman, R. A.; Clark, K. L.
Bioorg. Med. Chem. Lett. 2014, 24, 2212.
3. Bender, A. T.; Spyvee, M.; Satoh, T.; Gershman, B.; Teceno, T.; Burgess, L.;
Kumar, V.; Wu, Y.; Yang, H.; Ding, Y.; Akare, S.; Chen, Q. Am. J. Transl. Res. 2013,
5, 92.
1
1
4. Schiffler, M. A.; York, J. S. U. S. Patent 8,962,659, February 24, 2015.
5. During the course of the studies described in this Letter, a report of amphoteric
benzylamine EP4 antagonists was published. See: Borriello, M.; Pucci, S.; Stasi,
L.P.; Rovati, L. PCT Int. Appl. WO 2013004290, January 10, 2013.
6. Davis, T. L.; Sharif, N. A. Br. J. Pharmacol. 1919, 2000, 130.
1
1
7. Wilson, R. J.; Rhodes, S. A.; Wood, R. L.; Shield, V. J.; Noel, L. S.; Gray, D. W.;
Giles, H. Eur. J. Pharmacol. 2004, 501, 49.
1
1
2
8. Murase, A.; Taniguchi, Y.; Tonai-Kachi, H.; Nakao, K.; Takada, J. Life Sci. 2008,
8
2, 226.
9. Calculated using Marvin and calculator plugin freeware (www.chemaxon.com,
ChemAxon Kft, Budapest, Hungary).
0. Varma, M. V.; Gardner, I.; Steyn, S. J.; Nkansah, P.; Rotter, C. J.; Whitney-Pickett,
C.; Zhang, H.; Di, L.; Cram, M.; Fenner, K. S.; El-Kattan, A. F. Mol. Pharm. 2012, 9,
1
199.
Figure 3. Pharmacokinetics of 6 in beagle dog after 10 mg/kg oral dosing (n = 3,
open diamonds) and 1 mg/kg IV dosing (n = 3, filled diamonds). Diamonds represent
geometric mean values, and error bars represent geometric sample standard
deviations.
21. See Supporting information for tabulated data.
22. See Ref. 14 for experimental details and syntheses of all analogues 5–10.
Please cite this article in press as: Schiffler, M. A.; et al. Bioorg. Med. Chem. Lett. (2015), http://dx.doi.org/10.1016/j.bmcl.2015.05.091