Sodium [2′-[(cyclopropanecarbonyl-ethyl-amino)-methyl]-4′-(6- ethoxy-pyridin-3-yl)-6-methoxy-biphenyl-3-yl]-acetate (AM432): A potent, selective prostaglandin D2 receptor antagonist

Article abstract of DOI:10.1016/j.bmcl.2010.12.016

Compound 21 (AM432) was identified as a potent and selective antagonist of the DP₂ receptor (CRTH2). Modification of a bi-aryl core identified a series of tri-aryl antagonists of which compound 21 proved a viable clinical candidate. AM432 shows excellent potency in a human whole blood eosinophil shape change assay with prolonged incubation, a comparatively long off-rate from the DP₂ receptor, excellent pharmacokinetics in dog and in vivo activity in two mouse models of inflammatory disease after oral dosing.

Full text of DOI:10.1016/j.bmcl.2010.12.016

Bioorganic & Medicinal Chemistry Letters 21 (2011) 1036–1040
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Sodium [20-[(cyclopropanecarbonyl-ethyl-amino)-methyl]-40-(6-ethoxy-pyridin-
3-yl)-6-methoxy-biphenyl-3-yl]-acetate (AM432): A potent, selective prosta-
glandin D2 receptor antagonist
⇑
Nicholas Stock , Deborah Volkots, Karin Stebbins, Alex Broadhead, Brian Stearns, Jeffrey Roppe,
Timothy Parr, Christopher Baccei, Gretchen Bain, Charles Chapman, Lucia Correa, Janice Darlington,
Christopher King, Catherine Lee, Daniel S. Lorrain, Pat Prodanovich, Angelina Santini, Jilly F. Evans,
John H. Hutchinson, Peppi Prasit
Amira Pharmaceuticals, 9535 Waples St., Suite 100, San Diego, CA 92121, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Compound 21 (AM432) was identified as a potent and selective antagonist of the DP2 receptor (CRTH2).
Modification of a bi-aryl core identified a series of tri-aryl antagonists of which compound 21 proved a
viable clinical candidate. AM432 shows excellent potency in a human whole blood eosinophil shape
change assay with prolonged incubation, a comparatively long off-rate from the DP2 receptor, excellent
pharmacokinetics in dog and in vivo activity in two mouse models of inflammatory disease after oral
dosing.
Received 11 November 2010
Revised 30 November 2010
Accepted 2 December 2010
Available online 9 December 2010
Keywords:
Prostaglandin D2
CRTH2
Ó 2010 Elsevier Ltd. All rights reserved.
Antagonist
DP2 receptor
Arachidonic acid (AA), once liberated from the phospholipid
membrane wall via the action of phospholipase A2 (sPLA2), serves
as a precursor for a variety of pro-inflammatory enzymes in the
human body, notably the cyclooxygenase enzymes (COX-1/COX-
2) and 5-lipoxygenase (5-LO) enzyme. The action of COX-1/2 on
AA generates the prostaglandin PGH2 which in turn is further
metabolized to a variety of signaling molecules, each with their
own specific set of receptors (Fig. 1). Prostaglandin D2 (PGD2) is
the major product released from activated mast cells and exerts
its biological actions through selective binding to the G-protein
coupled receptors DP1 and DP2 (CRTH2).1 The DP2 receptor is ex-
pressed on basophils, eosinophils and Th2 cells and activation of
DP2 by PGD2 promotes the release of a variety of Th2 cytokines,
eliciting a continued immune response.2 The DP2 receptor has been
implicated in a number of inflammatory conditions such as
asthma, allergic rhinitis and, more recently, chronic obstructive
pulmonary disease (COPD). As such, DP2 poses an interesting target
for pharmacological intervention in these disorders.3
( Fig. 2). As part of our efforts to develop potent, selective DP2
antagonists suitable for clinical development, further SAR of the
biphenyl acetic acid series was undertaken culminating in the dis-
covery of AM432 (21) as
a potential candidate for clinical
development.
We have recently reported the discovery of several biphenyl
acetic acid based DP2 antagonists that demonstrated efficacy in a
murine model of allergic rhinitis, compound 14 and AM1565 (2)
⇑
Corresponding author.
E-mail address: nickstock@gmail.com (N. Stock).
Figure 1. Arachidonic acid processing through the prostaglandin pathway.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.12.016

N. Stock et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1036–1040
1037
Compound 1 possesses a carbamate group, a potential struc-
tural alert with regards to cross-reactivity through serine prote-
ases.8 Furthermore, significant quantities of the secondary amine
9 (from cleavage of the carbamate) were also detected after oral
dosing in rat, indicating the potential for complications in develop-
ment of a carbamate bearing inhibitor (see Fig. 3).
Concurrent work leading to the discovery of compound 2
(AM156) identified the cyclopropyl amide as a replacement for
the benzyl carbamate of 1 which was well tolerated and metabol-
ically stable.5 Further SAR therefore focused on modifications with
the cyclopropyl amide in place of the benzyl carbamate moiety
employing relatively non-polar aromatic substituents on the ‘B’
ring with the aim of minimizing the observed protein shift (Table
2).
Figure 2. Biphenylacetic acid DP2 inhibitors previously reported by our laborato-
ries. hDP2, human DP2 membrane binding assay6; HSA, human serum albumin;
hESC, human whole blood eosinophil shape change assay.7
Synthesis of these compounds is outlined in Scheme 1. Com-
mercially available 5-bromo-2-iodobenzonitrile was reduced to
the aldehyde 10 using standard reduction conditions (DIBAl-H) in
good yield. Reductive amination using sodium cyanoborohydride
in the presence of ethylamine followed by acylation with Boc
anhydride afforded the t-butylcarbamate 11. Cross coupling of
the previously disclosed boronate 124 with the more reactive aryl
iodide under Suzuki9 conditions furnished the biphenyl acetic ester
13 which could be then further transformed to the boronate 14
Compound 1 was discovered through internal screening efforts
and structural optimization, primarily focusing on ring ‘A’ and the
moieties attached to the benzylic amine on ring ‘B’.4 We sought to
optimize potency, off-target activities and pharmacokinetic
parameters of this scaffold through modification of the ‘B’ ring
and the substituents attached. All compounds were screened in a
DP2 membrane binding assay in the absence and presence of
0.2% human serum albumin (HSA) to evaluate intrinsic potency
at the receptor and ascertain the degree of protein shift associated
with these compounds.6 Compounds of interest were further pro-
filed in a human whole blood eosinophil shape change assay (hESC)
whereby activation of the DP2 receptor by PGD2 results in eosino-
phil degranulation and changes in cell morphology which can be
assayed using flow cell cytometry.7
Initial replacement of the tri-fluoromethyl moiety of compound
1 with a variety of groups (Table 1) indicated a wide tolerance for
diverse functionalities. These compounds were prepared as previ-
ously disclosed using the appropriately functionalized ‘B’ ring.4
All the compounds tested showed good intrinsic potency in the
DP2 membrane binding assay. However, in the presence of 0.2%
HSA varying degrees of protein shift were observed, the more polar
functions (amide 3, sulfonamide 4 and carboxylic acid – 8) show-
ing the greatest loss of activity in the presence of serum, presum-
ably due to greater protein binding for these compounds.
Figure 3. Benzyl carbamate cleavage of 1 to afford amine metabolite 9 after oral
dosing in rat.
Table 2
In vitro hDP2 membrane binding and human whole blood eosinophil shape change
(hESC) IC50’s for compounds 18–22a
Table 1
In vitro hDP2 membrane binding IC50’s for compounds 3–8a
Compound
R
hDP2 binding (nM)
hESC (nM)
ꢀHSA
+0.2% HSA
Compound
R
hDP2 binding (nM)
ꢀHSA +0.2% HSA
18
27
37
4
19
20
21
22
7
9
52
60
12
23
5
3
4
7
3
60
90
6
31
5
6
F
1
5
9
7
27
107
18
7
8
12
8
24
a
hDP2 binding values are the average of three experiments; hESC IC50’s are the
254
average of two experiments from separate donors. For compound 21 hDP2 binding
values are the average of five experiments and the hESC IC50 is the average of eight
experiments from separate donors.
a
hDP2 binding values are the average of three experiments.

1038
N. Stock et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1036–1040
Preparation of the desired cyclopropyl amides through reaction
of the secondary amine with cyclopropane carbonyl chloride and
Hünigs base afforded the esters 17. The desired acids (18–22) were
obtained through hydrolysis of the ethyl ester under standard reac-
tion conditions (LiOHꢁH2O, THF, H2O).
These ‘tri-aryl’ antagonists proved intrinsically potent in the
DP2 binding assay, with the quinoline 19, fluoro-pyridine 20 and
ethoxy-pyridine 21 showing the greatest intrinsic binding affinity.
All of these compounds showed a modest shift in the presence of
+0.2% HSA but more notable was the hESC assay results for both
the pyrazole 18 and ethoxy-pyridine 21. The apparent increase in
potency between the protein shifted binding assay and hESC assay
is presumably due to the difference in the expression levels of DP2
receptors in these assays. Pyrazole 18 had very poor exposure in
rat PK (data not shown), however, the ethoxy-pyridine 21 upon
dosing in rat had acceptable pharmacokinetic parameters (Fig. 4).
Furthermore, po (2.5 mg/kg) and iv (2 mg/kg) dosing in dog
showed extremely good PK properties (Cmax = 30
Cl = 0.5 mL/min/kg, AUC = 87 h g/mL).
lM, t½ = 6 h,
l
Due to the structural similarity of the ligands in the arachidonic
acid pathway, the potential for antagonist cross-reactivity between
the various AA metabolite receptors needs to be evaluated. We
therefore counter-screened 21 against a subset of these targets to
ascertain its selectivity (Table 3). Gratifyingly, 21 (AM432) showed
no cross-reactivity against the TP or IP receptors, COX-1 or COX-2
enzymes and minimal activity at the DP1 receptor. Furthermore,
there was no apparent agonist or antagonist activity against the
three PPAR isoforms at concentrations up to 250
21 also showed no significant inhibitory action against the human
CYP450 isoforms tested at concentrations up to 50 M and no time
lM. Compound
l
dependent inhibition of CYP3A4. Compound 21 was also devoid of
any significant activity at the human ether-a-go-go receptor
(hERG) in vitro at a concentration of 10 l
M.10
Scheme 1. (a) DIBAl-H, THF, ꢀ78 °C to rt (88%); (b) (i) Na(CN)BH3, AcOH, EtNH2,
MeOH, rt (81%); (ii) Boc2O, CH2Cl2, rt (99%); (c) boronate 12, Pd(PPh3)4 (cat.), K2CO3,
DME, H2O, 80 °C (66%); (d) Pd(dppf)Cl2ꢁCH2Cl2 (cat.), KOAc, bis(pinacolato)diboron,
dioxane, 85 °C (88%); (e) aryl iodide/bromide, Pd(PPh3)4 (cat.), K2CO3, DME, H2O,
80 °C; (f) 4 N HCl in dioxane, CH2Cl2, rt (99%); (g) (i) cyclopropane carbonyl chloride,
Hünigs base, CH2Cl2, rt (99%); (h) LiOHꢁH2O, THF, H2O, rt (quant).
It can be advantageous, when developing antagonists of
G-protein coupled receptors, such as DP2, to identify compounds
with long off-rates from the receptor. Antagonists with longer
off-rates may display greater clinical efficacy through prolonged
receptor blockade and this potentially mitigates unforeseen poor
pharmacokinetic parameters in human. Utilizing the DP2 radioli-
gandbindingassayandatritiatedformof21withvaryingincubation
times, we determined that 21 was fully bound to the DP2 receptor
after 30 min incubation at 37 °C and had an appreciable off-rate
half-life of approximately 90 min compared to 11 min for PGD2
(Fig. 5). Furthermore, utilizing the human whole blood eosinophil
under standard conditions (Pd(dppf)Cl2, dioxane, KOAc, bis(pinaco-
lato)diboron). A variety of heterocycles could then be cross coupled
with the boronate 14 under Suzuki conditions to yield the tri-aryl
intermediates 15. Removal of the Boc group under standard condi-
tions (4 N HCl in dioxane) afforded the amine hydrochloride 16.
100000
10000
1000
100
10
1
0.1
0.01
0.001
0
4
8
12
16
20
24
Time (hr)
Dog - IV
Dog - PO
Rat - IV
Rat - PO
Figure 4. Rat and dog pharmacokinetics for 21, dose normalized to 1 mg/kg.

N. Stock et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1036–1040
1039
Table 3
female BALB/c mice were sensitized by intraperitoneal administra-
tion of 10 g OVA complexed with Imject Alum on days 0 and 7. On
days 21–25, mice were challenged intranasally with 200 g OVA or
20 L saline and were orally dosed with either 0.5% methylcellu-
lose vehicle or compound 21 (10 mg/kg). The number of sneezes
and nasal rubs were counted during an 8 min period immediately
following OVA challenge on days 21, 23 and 25. OVA challenge in-
creased the number of sneezes and nasal rubs and this was signif-
icantly reduced by 21 (Fig. 6). These data provide further support
for the beneficial effects of DP2 antagonists in murine models of
allergic rhinitis.13
We recently reported on the pharmacological blockade of the
DP2 receptor inhibiting cigarette smoke induced inflammation
in the mouse lung.14 In this model female BALB/c mice were ex-
posed to the smoke of seven cigarettes on days 0–2. Mice were
orally dosed with compounds 21 (10 or 50 mg/kg) or 0.5% meth-
ylcellulose vehicle on days ꢀ1 through 2. On day 3, mice were
euthanized by isoflurane inhalation and bronchoalveolar lavage
and differential cell counts were performed as described previ-
ously.14 Cigarette smoke exposure produced an influx of neutro-
phils, lymphocytes and macrophages into the airways and this
was significantly reduced by 50 mg/kg of 21 (Fig. 7). At 10 mg/
Counter-screen and CYP450 data for 21 (AM432)a
l
l
Assay target
IC50
l
DP1 receptor
TP receptor
IP receptor
COX-1
COX-2
PPAR a, b and c
26.7 lM
>100
>100
>100
>100
>250
l
M
M
M
M
M
l
l
l
l
CYP3A4
CYP2C9
CYP2C19
CYP2D6
CYP1A2
>50
>50
>50
>50
>50
l
M
M
M
M
M
l
l
l
l
CYP3A4 TDI
hERG
0.002
11% @ 10
lM
a
hDP1 binding performed using 3H-PGD2 and
human platelet membranes. hTP binding per-
formed using human platelet membranes and 3H-
SQ-29548. hIP binding performed using hIP/293
membranes and 3H-iloprost.
shape change assay (hESC) with a longer incubation time (4 h), 21
showed a 2.6-fold increase in potency with an IC50 of 1.7 nM
(n = 4). This may be attributable to either the necessity of a longer
incubation in human blood to fully equilibrate with the DP2 receptor
from other blood proteins or the compound’s apparent long off-rate
from the receptor or both.
Given the favorable in vitro profile of 21 and suitable pharma-
cokinetics, it was further profiled in several animal models of
inflammatory disease. PGD2 and its receptor DP2 have been associ-
ated with both allergic rhinitis and asthma in human.11 The former
can be modeled in mice through sensitization of the animals with
ovalbumin (OVA) and subsequent nasal challenge.12 Therefore,
kg, 21 significantly reduced neutrophils and caused
a non-
significant trend toward reduction in lymphocytes and
a
macrophages. These data are consistent with previously pub-
lished reports describing efficacy of novel DP2 antagonists in
murine cigarette smoke exposure models.15
In conclusion, we have identified a potent and selective DP2
antagonist, compound 21 (AM432), with a 4 h hESC potency of
1.7 nM, acceptable pharmacokinetics in two species and efficacy
in both a murine model of allergic rhinitis and a cigarette smoke
induced inflammatory model of COPD.
Sneeze 5 day average
1250
1000
750
##
40
30
20
10
0
*
PGD2
21 (AM432)
500
250
0
PBS
Veh 21 (AM432)
0
30
60
90
120
150
180
Incubation time (minutes)
Nasal rub 5 day average
1200
1000
800
600
400
200
0
PGD2
40
30
20
10
0
21 (AM432)
#
*
PBS
Veh 21 (AM432)
0
30
60
90
120 150 180
Figure 6. Mouse allergic rhinitis data for 21 (AM432) (10 mg/kg) versus PBS treated
unsensitized mice and vehicle treated sensitized mice. Bars represent SEM of n = 7
mice per group. (Top) Number of sneezes, 5-day average; ##P <0.01 versus PBS, ⁄P
<0.05 versus vehicle; Tukey’s post hoc following ANOVA: (bottom) nasal rub counts,
5-day average; #P <0.05 versus PBS, ⁄P <0.05 versus vehicle; Tukey’s post hoc
following ANOVA.
Incubation time (minutes)
Figure 5. On- and off-rate analysis of PGD2 and compound 21 (AM432) at 37 °C.
Off- and on-rate determinations were performed using the hDP2 membrane binding
assay with varying incubation times at 37 °C. (Top) Both PGD2 and 21 (AM432) are
fully bound after 30 min. (Bottom) PGD2 t½ = 11 min, 21 (AM432) t½ = 86 min.

1040
N. Stock et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1036–1040
1.0
0.8
0.6
0.4
0.2
0.0
References and notes
***
1. Pettipher, R.; Hansel, T. T.; Armer, R. Nat. Rev. Drug Disc. 2007, 6, 313.
2. Kostensis, E.; Ulven, T. Trends Mol. Med. 2006, 12, 148.
3. Norman, P. Expert Opin. Invest. Drugs 2010, 19, 947.
4. Scott, J. M.; Baccei, C.; Bain, G.; Broadhead, A.; Evans, J. F.; Fagan, P.;
Hutchinson, J. H.; King, C.; Lorrain, D. S.; Lee, C.; Prasit, P.; Prodanovich, P.;
Santini, A.; Stearns, B. A. Bioorg. Med. Chem. Lett. 2010, accepted for publication.
5. Stebbins, K. J.; Broadhead, A. R.; Correa, L. D.; Scott, J. M.; Truong, Y. P.; Stearns,
B. A.; Hutchinson, J. H.; Prasit, P.; Evans, J. F.; Lorrain, D. S. Eur. J. Pharmacol.
2010, 638, 142.
6. The DP2 radioligand binding assay was performed on membranes from 293 cells
stably expressing human DP2. To measure binding, [3H]-PGD2 was incubated
together with 293(hDP2) membranes in the presence of increasing concentrations
of compounds. Membranes were harvested andwashed usinga Brandel Harvester
and the amount of [3H]-PGD2 that remained bound to the cells was measured on a
TopCount. Theconcentrationofcompounds requiredtoachievea50%inhibitionof
[3H]-PGD2 binding (the IC50) was determined. The binding assay was carried out
both in the absence and presence of 0.2% human serum albumin (HSA) to evaluate
the protein shift associated with the compounds.
**
**
0.06
0.04
0.02
0.00
**
7. Human blood was drawn into EDTA tubes and used within 2 h of draw. One
hundred microliter aliquots of fresh blood were incubated for 15 min at 37 °C
plus or minus test compound in 50% DMSO/water. PGD2 (50 nM final
concentration) or vehicle was added from a 1
were continued for 5 min at 37 °C. The reactions were placed on ice and 250
ice-cold 1:4 diluted Cytofix (BD Biosciences) in PBS immediately added. The
reactions were transferred to FACS tubes and lysed with ammonium chloride
lysing solution at room temperature for 15 min. Tubes were centrifuged and the
l
M stock in PBS and incubations
L of
l
*
*
cells washed once with 3 mL cold PBS before re-suspension in 200 lL of ice-cold
1:4 diluted Cytofix in PBS. Eosinophil shape change was analyzed on a FACS
calibur by analyzing forward light scatter of the autofluorescent cells.
8. Carbamates are often irreversible inhibitors of various enzymes, see for
example: (a) Alexander, J. P.; Cravatt, B. F. Chem. Biol. 2005, 12, 1179; (b) Lin,
G.; Chen, G.-H.; Ho, H.-C. Bioorg. Med. Chem. Lett. 1998, 8, 2747; (c) Hosie, L.;
Sutton, L. D.; Quinn, D. M. J. Biol. Chem. 1987, 262, 260.
9. For a review on the Suzuki–Miyaura reaction, see: Miyaura, N.; Suzuki, A. Chem.
Rev. 1995, 95, 2457.
10. hERG analysis performed by ChanTest corporation using a patch–clamp style
assay.
11. (a) Naclerio, R. M.; Meier, H. L.; Kagey-Sobotka, A.; Adkinson, N. F., Jr.; Meyers, D.
A.; Norman, P. S.; Lichtenstein, L. M. Am. Rev. Respir. Dis. 1983, 128, 597; (b)
Seibert, T.; Depner, M.; Kormann, M.; Knopp, N.; Illig, T.; von Mutius, E.; Kasbech,
M.; Cameron, L. J. Allergy Clin. Immunol. 2008, 121, S81; (c) Shirasaki, H.; Kikuchi,
M.; Kanaizumi, E.; Himi, T. Ann. Allergy Asthma Immunol. 2009, 102, 110.
12. Nakaya, M.; Dohi, M.; Okunishi, K.; Nakagome, K.; Tanaka, R.; Imamura, M.;
Baba, S.; Takeuchi, N.; Yamamoto, K.; Kaga, K. Lab. Invest. 2006, 86, 917.
13. (a) Nomiya, R.; Okano, M.; Fujiwara, T.; Maeda, M.; Kimura, Y.; Kino, K.;
Yokoyama, M.; Hirai, H.; Nagata, K.; Hara, T.; Nishizaki, K.; Nakamura, M. J.
Immunol. 2008, 180, 5680; (b) Shirashi, Y.; Burgess, L. E.; Takeda, K.; Eberhardt,
C.; Wright, D.; Cook, A.; Corrette, C.; Hunt, K.; Clark, C.; Kim, G.; Delisle, K.;
Hayter, L.; Neale, J.; Chantry, D.; Doherty, G.; Gelfand, E. W. Am. J. Respir. Crit.
Care Med. 2009, 179, A5718.
2.0
1.5
1.0
0.5
0.0
*
#
Figure 7. Broncoalveolar lavage (BAL) fluid cell count from mice after cigarette
smoke exposure after 10 and 50 mg/kg po dosing of compound 21 (AM432). (Top)
Neutrophil count; ⁄⁄P <0.01 versus smoke Tukey’s post hoc following ANOVA, ⁄⁄⁄P
<0.001 versus air, Tukey’s post hoc following ANOVA; (middle) lymphocyte count;
⁄P <0.05 versus smoke, Tukey’s post hoc following ANOVA, ⁄⁄P <0.01 versus air
Tukey’s post hoc following ANOVA; (bottom) macrophage count; ⁄P <0.05 versus
air, one-tailed t test, #P <0.05 versus smoke, one-tailed t test.
14. Stebbins, K. J.; Broadhead, A. R.; Baccei, C. S.; Scott, J. M.; Truong, Y. P.; Coate,
H.; Stock, N. S.; Santini, A. M.; Fagan, P.; Prodanovich, P.; Bain, G.; Stearns, B. A.;
King, C. D.; Hutchinson, J. H.; Prasit, P.; Evans, J. F.; Lorrain, D. S. J. Pharmacol.
Exp. Ther. 2010, 332, 764.
15. Sargent, C.; Stinson, S.; Schmidt, J.; Dougall, I.; Bonnert, R.; Paine, S.; Saunders,
M.; Foster, M. In Proceedings of the British Pharmacological Society 2009, 7th
James Black Conference, www.pa2online.org.