Discovery of potent and selective small-molecule PAR-2 agonists

Article abstract of DOI:10.1021/jm800754r

Proteinase activated receptor-2 plays a crucial role in a wide variety of conditions with a strong inflammatory component. We present the discovery and characterization of two structurally different, potent, selective, and metabolically stable small-molecule PAR-2 agonists. These ligands may be useful as pharmacological tools for elucidating the complex physiological role of the PAR-2 receptors as well as for the development of PAR-2 antagonists.

Full text of DOI:10.1021/jm800754r

5490
J. Med. Chem. 2008, 51, 5490–5493
Letters
Chart 1
Discovery of Potent and Selective
Small-Molecule PAR-2 Agonists
Jimmi Gerner Seitzberg,† Anne Eeg Knapp,†
Birgitte Winther Lund,† Sine Mandrup Bertozzi,†
Erika A. Currier,‡ Jian-Nong Ma,‡ Vladimir Sherbukhin,†
Ethan S. Burstein,‡ and Roger Olsson*,†
ACADIA Pharmaceuticals AB, Medeon Science Park, Per Albin
Hanssons Va¨g 35, S-205 12 Malmo¨, Sweden, ACADIA
Pharmaceuticals Inc., 3911 Sorrento Valley BouleVard,
San Diego, California 92121
Most research efforts aimed at elucidating the complex
physiological roles of PAR-2 have relied on PAR-2 APs (e.g.,
SLIGRL), including recently more stable variants (e.g., 2-furoyl-
LIGRLO-NH2).6 The susceptibility to proteolytic degradation
of these peptides constitutes a major limitation for in vivo
applications. To date, no nonpeptidic agonists of PAR-2
receptors have been reported,7 which has hampered attempts
to fully explore the function of the PAR-2 receptor. The
discovery of metabolically stable small-molecule PAR-2 ago-
nists would provide useful pharmacological tools for elucidating
the complex physiological functions of PAR-2 receptors.
Herein we report the discovery and initial SAR of potent and
selective nonpeptidic small-molecule PAR-2 agonists. A chemi-
cal library containing more than 250000 small-molecule drug-
like compounds was screened for agonist activity at the human
PAR-2 receptor using the cell-based functional assay R-SAT,8
and a number of active compounds were identified. The hits
could be divided into two different chemical classes exemplified
by compound 1 (AC-55541),9 displaying full agonism and nM
activity at PAR-2 (pEC50 6.7 and 81% efficacy), and the partial
agonist 2 (AC-98170)10 (pEC50 5.2 and 30% efficacy) (Chart
1). The structure class that included the most potent and
efficacious hit 1 was selected for further exploration.
A focused library around 1 was made by reacting a number
of aromatic aldehydes or ketones with the parent hydrazide.
From this library, an initial structure activity relationship (SAR)
was established which revealed that the methyl substituent on
the hydrazone was essential for activity whereas hydrogen or
higher alkyl groups had a detrimental effect. Although a wide
range of substituents in the 3-position of the aryl hydrazone of
compound 1 were beneficial for the activity at PAR-2, any
substituent in either the 2- or 4-position led to reduced activity.
A variety of heteroaromatic moieties were also attached but all
led to reduced activity compared to 1.11
In addition to the good metabolic stability of 1 in human
and rat microsomes (Clint h/rat 6/19 µL/min·mg), the compound
also complies with the Lipinski rule-of-five. However, com-
pound 1 had low solubility not only in phosphate buffer solution
but also in other mediums such as organic solvents, which
hampered the further use of this compound. Furthermore, the
lack of reasonable building blocks made the synthesis of
analogues of 1 quite demanding. These shortcomings led us to
investigate the possibilities with the second structural class
identified during the initial R-SAT screen. A hit-to-lead
optimization effort of 2 was initiated by first taking advantage
of the structural similarities between compounds 1 and 2.
A molecular overlay of compounds 1 and 2 was performed
using the flexible alignment procedure in MOE.12 In addition
to substitution in the aryl hydrazone part, the model suggested
ReceiVed June 19, 2008
Abstract: Proteinase activated receptor-2 plays a crucial role in a wide
variety of conditions with a strong inflammatory component. We present
the discovery and characterization of two structurally different, potent,
selective, and metabolically stable small-molecule PAR-2 agonists.
These ligands may be useful as pharmacological tools for elucidating
the complex physiological role of the PAR-2 receptors as well as for
the development of PAR-2 antagonists.
Proteinase activated receptor-2 (PAR-2a) is considered to be
an attractive target for drug discovery.1 Physiologically, PAR-2
receptors are thought to fulfill a critical role in mediating
nociceptive and inflammatory responses.2,3 Therefore, drugs
which target PAR-2 have the potential to treat a wide variety
of disorders. Currently PAR-2 antagonists are of particular
interest due to their potential for relieving inflammatory
symptoms in rheumatoid arthritis. Paradoxically, PAR-2 agonists
may also have therapeutic value. Although PAR-2 agonists
would be expected to exacerbate most nociceptive and inflam-
matory processes, they have been suggested to have a protective
role in certain settings. For example, PAR-2 agonists may have
a therapeutic role as gastric cytoprotective agents or as mediators
of airway smooth muscle relaxation.4,5 The development of
small-molecule ligands that selectively target PAR-2 will help
the full assessment of PAR-2 receptors as therapeutic targets.
PAR-2 receptors belong to a subfamily of four G-protein
coupled receptors (PAR-1, PAR-2, PAR-3, and PAR-4) and is
widely expressed throughout the body including the CNS,
cardiovascular, gastrointestinal, and pulmonary systems. The
PARs are activated by tethered peptide ligands exposed by
enzymatic proteolytic cleavage of the extracellular amino-
terminus. PAR-1, PAR-3, and PAR-4 are activated by the
protease thrombin, whereas PAR-2 is activated by trypsin and
a variety of other proteases. In addition, PAR-1, PAR-2, and
PAR-4 can be activated by soluble peptides derived from their
tethered ligands (henceforth PAR activating peptides, or PAR
APs).
* To whom correspondence should be addressed. Phone: +46406013400.
Fax: +46406013401. E-mail: roger@acadia-pharm.com.
†
ACADIA Pharmaceuticals AB.
ACADIA Pharmaceuticals Inc.
‡
a Abbreviations: PAR, proteinase activated receptor; PAR-AP, PAR-
activating protein; SLIGRL, Ser-Leu-Ile-Gly-Arg-Leu-NH2; 2-furoyl-
LIGRLO-NH2, 2-furoyl-Leu-Ile-Gly-Arg-Leu-Orn-NH2; SFLLRN, Ser-Phe-
Leu-Leu-Arg-Asn-NH2; AYPGKF, Ala-Tyr-Pro-Gly-Lys-Phe-NH2.
10.1021/jm800754r CCC: $40.75
2008 American Chemical Society
Published on Web 08/23/2008

Letters
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 18 5491
Table 1. In Vitro Activities of PAR-2 Agonists Using R-SATa
Figure 1. Molecular overlay of 1 and 2.
Scheme 1. Synthesis of 3a
a Reaction conditions: (i) H2NNH2, MW (120 °C, 20 min), 56%; (ii)
3-BrC6H4Ac, EtOH/AcOH 9/1, MW (120 °C, 600 s), 71%.
that substituents ortho- or meta- in the pyrrolidinone aromatic
part should be tolerated as should substituents on the pyrroli-
dinone nitrogen (Figure 1).
This led to the synthesis of 3 (AC-264613),10 which was
easily prepared in two steps from commercially available ester
4 (Scheme 1).
Compound 3 was highly potent at the PAR-2 receptor (entry
5, Table 1) and consolidated the new structure class that could
be systematically modified to further explore the SAR and
potentially incorporate desirable physicochemical properties.
The first set of analogues, prepared by the reaction of various
aromatic ketones with hydrazide 5 (Scheme 1), were aimed at
verifying that the SAR correlated with the results obtained
previously with the 1 series (compounds 6-9, Table 1).
Analogues with substituents in the aromatic part of the pyrro-
lidinone were addressed by the synthesis of a focused library.
The synthesis is outlined in Scheme 2, exemplified with the
3-bromo analogue 10. The first step is a Michael addition
between diethyl malonate and (E)-1-bromo-3-(2-nitrovinyl)ben-
zene using sodium ethoxide as base. The crude Michael product
11 was reduced by Raney nickel using ammonium formate,13
whereby the pyrrolidinone 12 was spontaneously formed as the
thermodynamically more stable trans-isomer via an in situ
cyclization.14 The hydrazide 13 was formed by adding hydrazine
to the ethyl ester 12 under microwave conditions. Finally, the
hydrazide 13 was condensed with 3-bromoacetophenone, yield-
ing the desired product 10. The synthesis of N-methylated
derivative 14 utilized a standard peptide coupling between acid
15 and hydrazone 16 (Scheme 3).
a It was not possible to separate the E- and Z-isomers and hence all
structures were tested as mixtures of E- and Z-isomers on the imine part.
See Supporting Information for details. R-SAT was performed as described
in ref 8. Values represent the means ( SEM of three-nine independent
experiments.
Scheme 2. Synthesis of 10a
Substitution on the pyrrolidinone nitrogen was investigated
and 19 was prepared by a three-step procedure in 26% yield
(Scheme 4).15
The synthesized compounds were tested at human PAR-2
using R-SAT (Table 1). Initially, we explored the SAR around
the aryl hydrazone part of the lead structure (entries 4-9).
Substitution of the meta-position of the aryl was important for
the activity, both the para- and ortho-substituted analogues 6
a Reaction conditions: (i) NaOEt, 0 °C, 2 h; (ii) (a) RaNi, HCO2NH4,
EtOH, 70 °C, o.n. then rt, 3 days, (b) RaNi, H2, rt, 24 h; (iii) H2NNH2 ·H2O,
MW 140 °C, 20 min, 16% (3 steps); (ii) 3-BrC6H4Ac, EtOH:AcOH 10:1,
50 °C, 16 h, 51%.
and 7 respectively showed more than 50-fold reduced activity
compared to the meta-substituted compound 3 (compare entries

5492 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 18
Scheme 3. Synthesis of 14a
Letters
a Reaction conditions: (i) EDC, HOBT, THF, rt, 58%.
Scheme 4. Synthesis of 19a
Figure 2. Receptor selectivity of PAR-2 agonists. Each compound
was tested in concentration response experiments in R-SAT at PAR-1,
PAR-4, and PAR-2 receptors. Activity of each receptor was verified
with the PAR APs SFLLRN, AYPGKF, and SLIGRL for PAR-1, PAR-
4, and PAR-2, respectively (not shown).
The activities of both compounds 3 and 1 were confirmed in
phosphatidyl inositol (PI) hydrolysis and Ca2+ mobilization
assays. The selectivity toward the other PAR subtypes was
tested, and no activity of either 3 or 1 was observed at PAR-1,
PAR-4 (Figure 2), or PAR-3.17 Furthermore, neither compound
had significant affinity in a profiling effort with more than 30
other targets implicated in nociception and inflammation.
Importantly, the compounds show persistent and dose-dependent
activity in vivo in paw edema and thermal hyperalgesia assays,
characteristic of PAR-2 agonists.17
a Reaction conditions: (i) isobutyraldehyde, p-TsOH, MW 160 °C, 15
min, 62%; (ii) H2NNH2 ·H2O, MW 120 °C, 20 min, 61%; (iii) 3-BrC6H4Ac,
10% AcOH in EtOH, 80 °C, 4 h, 70%.
In conclusion, we have described the discovery and charac-
terization of the first small-molecule nonpeptidic PAR-2 ago-
nists. Both compounds 3 and 1 display significantly higher
activities compared to the PAR-2 AP SLIGRL-NH2. These
compounds are highly selective for PAR-2 over the other PAR
subtypes, they are metabolically stable and have persistent
activity in vivo. Even though the possible therapeutic potential
of PAR-2 agonists remains unclear, PAR-2 has been shown to
play a crucial role in a wide variety of conditions with a strong
inflammatory component. This indicates that PAR-2 could
indeed be a novel target for drug development with both agonists
and antagonists having therapeutic potential. The identification
of potent, selective, and metabolically stable small-molecule
PAR-2 agonists will be useful as pharmacological tools for
elucidating the complex and at times contradictive physiological
functions of the PAR-2 as well as for the development of PAR-2
antagonists.
5, 6, and 7). The 5-bromothiophene compound 8 gave about
the same activity as the para-substituted compound 6 (entries
6 and 8). This was not surprising because the thiophene is
regarded as a phenyl isoster and the 5-bromo substitution would
more closely resemble a para-substituted phenyl than a meta-
substituted one. The structurally more rigid dihydroindene 9
showed a slight decrease in activity compared to compound 3
(entries 5 and 9).
As indicated by the alignment of the hits 1 and 2, substituents
in the 2- and 3-position of the 4-phenyl of the pyrrolidinone
were allowed. In line with the 3-bromo substituted compound
10, several of the synthesized analogues were highly active,
however, none of them displayed increased activity toward
PAR-2 compared to compound 3.16 The N-methyl substituted
compound 14 was inactive, indicating the importance of a
hydrogen bond donor next to the carbonyl. Substituents on the
pyrrolidinone nitrogen should be well tolerated according to the
molecular alignment and, indeed, 19 was one of the more active
compounds in the series.
Acknowledgment. We thank Dr. Martin Carnerup for micro-
somal data and Susanna Henriksson for analytical assistance.
Even though compound 3 was more active than hit 1 and
had a low intrinsic clearance in vitro using liver microsomes
(Clint for compound 3 is human/rat 9/37 µL/min·mg), the
aqueous solubility of the compound was not improved. One
contributing factor causing the low solubility of both compounds
1 and 3 may be the presence of hydrogen bond donor-acceptor
pairs in the molecules. Hence, incorporating a substituent into
either the hydrazide or the pyrrolidinone part of compound 3
would potentially increase solubility. However, these modifica-
tions did not improve the solubility of either 14 nor 19.
Providentially, the use of surfactants like Tween increased the
solubility of compound 3 to a satisfying level (>2 mg/mL in
25% Tween), which makes this compound useful for further
studies.
Supporting Information Available: Experimental details and
characterization data for all compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) De Campo, B. A.; Henry, P. J. Protease-Activated Receptors (PARs)
are Partly Pro-Inflammatory and Partly Anti-Inflammatory: Will PAR
Agonists or Antagonists Participate in Future Drug Therapies. Curr.
Drug Targets 2006, 7 (5), 629–637.
(2) Macfarlane, S. R.; Seatter, M. J.; Kanke, T.; Hunter, G. D.; Plevin,
R. Proteinase-Activated Receptors. Pharmacol. ReV. 2001, 53 (2), 245–
282.
(3) Ossovskaya, V. S.; Bunnett, N. W. Protease-activated receptors:
contribution to physiology and disease. Physiol. ReV. 2004, 84 (2),
579–621.
(4) Henry, P. J. The protease-activated receptor2 (PAR2)-prostaglandin
E2-prostanoid EP receptor axis: A potential bronchoprotective unit in
the respiratory tract? Eur. J. Pharmacol. 2006, 533, 156–177.
(5) D’Agostino, B.; Roviezzo, F.; De Palma, R.; Terracciano, S.; De
Nardo, M.; Gallelli, L.; Abbate, G. F.; D’Aiuto, E.; Russo, M.; Cirino,
G.; Rossi, F. Activation of protease-activated receptor-2 reduces
airways inflammation in experimental allergic asthma. Clin. Exp.
Allergy 2007, 37 (10), 1436–1443.
The enantiomers of 3 were separated using chiral HPLC and
each enantiomer was tested for agonist activity at the PAR-2
receptor. It was found that the (+)-isomer was responsible for
essentially all agonist activity (pEC50 7.5 ( 0.0 and %Eff 91 (
13 vs pEC50 < 5 and %Eff 10 ( 3 for the (-)-isomer).

Letters
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 18 5493
(6) (a) McGuire, J. J.; Saifeddine, M.; Triggle, C. R.; Sun, K.; Hollenberg,
M. D. 2-Furoyl-LIGRLO-amide: A Potent and Selective Proteinase-
Activated Receptor 2 Agonist. J. Pharmacol. Exp. Ther. 2004, 309
(3), 1124–1131. (b) Kawabata, A.; Kanke, T.; Yonezawa, D.; Ishiki,
T.; Saka, M.; Kabeya, M.; Sekiguchi, F.; Kubo, S.; Kuroda, R.; Iwaki,
M.; Katsura, K.; Plevin, R. Potent and Metabolically Stable Agonists
for Protease-Activated Receptor-2: Evaluation of Activity in Multiple
Assay Systems in Vitro and in Vivo. J. Pharmacol. Exp. Ther. 2004,
309 (3), 1098–1107.
(7) A low potency small molecule PAR-2 antagonist was recently
reported. Kelso, E. B.; Lockhart, J. C.; Hembrough, T.; Dunning, L.;
Plevin, R.; Hollenberg, M. D.; Sommerhoff, C. P.; McLean, J. S.;
Ferrell, W. R. Therapeutic promise of proteinase-activated receptor-2
antagonism in joint inflammation. J. Pharmacol. Exp. Ther. 2006, 316
(3), 1017–1024.
(12) Molecular Operating Environment, Version 2004.03, Chemical Com-
puting Group Inc. (http://www.chemcomp.com), 1010 Sherbrooke
Street West, Suite 910, Montreal, Canada H3A 2R7. A number of
conformations were generated for each of these two molecules using
a stochastic search algorithm and the MMFF94 force field. These
conformations were then superimposed in 3D to ensure the
maximum overlap of their van der Waals volumes and pharma-
cophore features, like hydrogen bond donors and acceptors,
aromatic rings, and lipophilic moieties. Each resulting alignment
was assigned a score which quantified both feature overlap and
internal strain energy.
(13) Langlois, M.; Guillonneau, C.; Vo Van, T.; Maillard, J.; Lannoy, J.;
Nguyen, H. N.; Morin, R.; Manuel, C.; Benharkate, M. Synthe`se et
proprie´te´s antide´pressives de derives de l’amino-2 phe´nyl-4 delta
1-pyrroline. Eur. J. Med. Chem. 1978, 13 (2), 161–169.
(8) (a) Burstein, E. S.; Ott, T: R.; Feddock, M.; Ma, J. N.; Fuhs, S.; Wong,
S.; Schiffer, H. H.; Brann, M. R.; Nash, N. R. Characterization of the
Mas-related gene family: structural and functional conservation of
human and rhesus MrgX receptors. Br. J. Pharmacol. 2006, 147 (1),
73–82. (b) Burstein, E. S.; Piu, F.; Ma, J. N.; Weissman, J. T.; Currier,
E. A.; Nash, N. R.; Weiner, D. M.; Spalding, T. A.; Schiffer, H. H.;
Del Tredici, A. L.; Brann, M. R. Integrative Functional Assays,
Chemical Genomics and High Throughput Screening: Harnessing
Signal Transduction Pathways to a Common HTS Readout. Curr.
Pharm. Des. 2006, 12 (14), 1717–1729.
(9) Burstein, E.; Knapp, A. E. PAR-2 Modulating Compounds and Their
Use. Patent WO 2006/127379 A2, 2006.
(10) Olsson, R.; Seitzberg, J. G. PAR-2 Modulating Compounds and Their
Use. U.S. Patent Application no. 2007/0123508, 2007.
(14) Relative stereochemistry was assigned by coupling constant analysis
and compared to literature values. Evans, D. A.; Mito, S.; Seidel, D.
Scope and Mechanism of Enantioselective Michael Additions of 1,3-
Dicarbonyl Compounds to Nitroalkenes Catalyzed by Nickel(II)-
Diamine Complexes. J. Am. Chem. Soc. 2007, 129, 11583–11592.
(15) Cook, G. R.; Stille, J. R. Preparation and 3-Aza-Cope Rearrangement
of N-Alkyl-N-allyl Enamines. J. Org. Chem. 1991, 56, 5578–5583.
(16) Unpublished results.
(17) Gardell, L. R.; Ma, J. N.; Seitzberg, J. G.; Knapp, A. E.; Schiffer,
H. H.; Tabatabaei, A.; Davis, C.; Wong, K. K.; Lund, B. W.; Nash,
N. R.; Gao, Y.; Lameh, J.; Schmelzer, K.; Olsson, R.; Burstein, E. S.
J. Pharmacol. Exp. Ther. 2008, in press.
(11) Unpublished results.
JM800754R