Discovery of non-peptidergic MrgX1 and MrgX2 receptor agonists and exploration of an initial SAR using solid-phase synthesis

Article abstract of DOI:10.1016/j.bmcl.2009.01.085

A class of small molecules displaying comparable activities with peptide ligands BAM22 and corticostatin-14 at both the human and rhesus monkey MrgX1 and MrgX2 receptors, respectively, was discovered. A comparative study to compare solid-phase and solution-phase chemistries for the efficient synthesis of the active class, tetracyclic benzimidazoles, was undertaken. The solid-phase chemistry was found to be superior both for the synthesis of analogs and for the synthesis of gram quantities.

Full text of DOI:10.1016/j.bmcl.2009.01.085

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1729–1732
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of non-peptidergic MrgX1 and MrgX2 receptor agonists
and exploration of an initial SAR using solid-phase synthesis
Leila Malik a, Nicholas M. Kelly a, Jian-Nong Ma b, Erika A. Currier b, Ethan S. Burstein b, Roger Olsson a,
*
a ACADIA Pharmaceuticals AB, Medeon Science Park, S-205 12 Malmö, Sweden
b ACADIA Pharmaceuticals Inc., 3911 Sorrento Valley Blvd., 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:
A class of small molecules displaying comparable activities with peptide ligands BAM22 and corticosta-
tin-14 at both the human and rhesus monkey MrgX1 and MrgX2 receptors, respectively, was discovered.
A comparative study to compare solid-phase and solution-phase chemistries for the efficient synthesis of
the active class, tetracyclic benzimidazoles, was undertaken. The solid-phase chemistry was found to be
superior both for the synthesis of analogs and for the synthesis of gram quantities.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 5 January 2009
Revised 25 January 2009
Accepted 26 January 2009
Available online 30 January 2009
Keywords:
Mas related
MrgX1
MrgX2
Agonist
SNSR
Sensory-neuron specific receptors
Nociception
Non-peptidergic
Small molecule
Solid-phase
R-SAT
Mas-related G-protein coupled receptors (Mrg receptors), also
known as sensory-neuron specific receptors (SNSRs), are a large
family of over 50 subtypes of orphan receptors expressed almost
exclusively in small diameter sensory neurons.1,2 The physiological
functions of these receptors are unknown but their expression pat-
tern suggests a role in nociception. MrgX1 and MrgX2 are human
receptor subtypes for which peptide ligands have been identi-
fied,2–5 and for which more direct evidence for a role in nociception
exists.6 However, despite their obvious potential as drug targets,
little progress has been made exploiting MrgX1 or MrgX2 receptors
therapeutically. This is partly because rodent orthologs of these
receptors do not exist necessitating the use of higher species such
as rhesus monkey to perform in vivo studies,7,8 and partly because
the known peptide ligands (e.g., BAM22)2 and (e.g., cortistatin-14)3
for MrgX1 and MrgX2 receptors, respectively, activate other recep-
tors complicating target validation. Recently small molecule antag-
onists for the MrgX1 receptor were reported9 but further
development of MrgX receptor selective non-peptidergic agonists
and antagonists will be necessary to fully elucidate the complex
pharmacology of the Mrg family.
Herein we report the discovery and initial structure–activity
relationship (SAR)ofpotent and selectiveMrgX1 and MrgX2receptor
agonists. A high-throughput screen of 250,000 small molecules was
performed using recombinant human MrgX1 receptors expressed
in the mammalian cell-based functional assay receptor selection
and amplification technology (R-SAT).7,10 The high-throughput
screen resulted in the identification of a number of agonist hits from
different structure classes. Subsequent pharmacological profiling
revealed several compounds from one chemical class that also had
agonist activity at human MrgX2 receptors. Two of the compounds
that showed selectivity between MrgX1 and MrgX2 receptors were
selected as starting points for exploring the SAR. Compounds 1 and 2
showed full efficacy at the MrgX1 receptor (pEC50 5.6) and the
MrgX2 receptor (pEC50 6.5), respectively (see Fig. 1).
A solid-phase chemistry protocol was used to resynthesize 1
and 2 and to further synthesize a focused compound library
(Scheme 1).11 The focused library was tested for activity at the
MrgX1 and MrgX2 receptors using R-SAT, a representative set of
compounds is shown in Table 1 (compounds 1, 2, and 12–19).
The first step in the synthesis of tetracyclic benzimidazoles
(Scheme 1) requires attachment of benzhydrylamine (MBHA) resin
to the selected amino acid. Coupling of the N-protected amino acid
to MBHA resin using N,N0-diisopropylcarbodiimide (DIC) and N-
* Corresponding author. Tel.: +46 406013466; fax: +46 406013401.
E-mail address: roger@acadia-pharm.com (R. Olsson).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.085

1730
L. Malik et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1729–1732
O
F
F
HN
HN
NH
i
ii
iii
O
H
O
H
O
N
O
NO2
O
NH
O
NO2
O
NH2
O
NO2
N
N
O
N
N
iv
N
N
N
O
N
NH
H
NH
H
NO2
NH
N
N
O
O
HN
O
v
O
N
2
2
N
20
N
O
CO2H
F
1
2
NH
H
O
NH
Figure 1. Identified MrgX1 and MrgX2 receptor agonists 1 and 2.
O
hydroxybenzotriazole (HOBt) in DMF, followed by deprotection of
the N-Boc group provided the resin bound compound 3. The resin
was then heated in a solution of 4-fluoro-3-nitrobenzoic acid and
diisopropylethylamine (DIEA) in N-methylpyrrolidinone (NMP)
affording compound 4. Coupling of 4 with an amine using DIC
and HOBt in DMF resulted in compound 5 with two diversity
points. Reduction of the nitro group to the amine 6 was achieved
using a 2-M solution of SnCl2 in N-methylpyrrolidinone, then reac-
tion with 5-fluoro-2-nitrobenzaldehyde afforded the desired tri-
substituted benzimidazole 7. Displacement of the fluoro substitu-
ent with appropriate amines followed by reduction of the nitro
group resulted in compound 9. This was then cleaved from the
Scheme 2. Reagents and conditions: (i) 2,2-dimethylpropane-1,3-diol, TsOH (cat)
in toluene, reflux 3 h, 99% yield; (ii) cycloheptanamine, in EtOH, iPr2EtN (1.5 equiv),
MW 180 °C, 20 min, 69% yield; (iii) Na2S2O4 (5 equiv), Na2CO3 (5 equiv), EtOH/H2O,
75% yield; (iv) carboxylic acid 20, HOBt (1.5 equiv), EDCI.HCl (1.5 equiv), Et3N
(2 equiv), DMAP (0.1 equiv). CH3CN, MW 120 °C, 5 min, 28% yield; (v) Fe, 50 °C,
45 min, EtOH/H2O/HCl (4:3:1), 50 °C, 90 min, 12% yield.
and 3). While the synthetic route in Scheme 3 basically follows the
solid-phase protocol, the strategy depicted in Scheme 2 differs
from the solid-phase protocol in having the formation of the benz-
imidazole as the last step. Although both solution-phase methods
gave the target molecules, neither of the methods was superior
to the solid-phase protocol. The major obstacles were the time con-
suming and yield reducing purifications between reaction steps,
which demanded extensive chromatography to yield pure com-
pounds. Maybe not unexpectedly, this shows that a well validated
solid-phase protocol is a powerful method to use. Hence, the solid-
phase protocol was further used in the resynthesis and in the syn-
thesis of gram quantities of material (Scheme 1).
resin utilizing
a mixture with trifluoromethanesulfonic acid
(TFMSA) to form the corresponding non-cyclic benzimidazole 10.
Unfortunately the ring closure did not go to completion during
the cleavage, although this might have been expected from previ-
ous reports.11 Ring closure was finally obtained by heating in the
microwave (MW) for 5 min in concd HCl. All the compounds were
purified by preparative MS–HPLC and verified by NMR and LC/MS.
The overall yield of this solid-phase synthesis protocol was
between 5% and 25% over 10 steps.
The library initially screened included about 500 compounds of
this class with R1 based on the L-amino acids glycine, alanine,
Meanwhile to meet the expected demand for the synthesis of
larger quantities and to speed up the resynthesis, a considerable ef-
fort was put into developing a solution-phase protocol (Schemes 2
valine, isoleucine, phenylalanine, methionine, glutamine, and the
non-endogenous amino acid 2-amino-2-phenylacetic acid. In the
screen, 30 confirmed hits were found (6% hit rate) and all of the
O
NO2
NH
R1
NO2
NH
R1
R1
O
O
O
OH
O
R1
i, ii
iii
iv
O
O
+
H
N
HN
2
NH
R2 NH
HO
NH2
O
NH
NH
MBHA
3
4
5
O
O
NO2
NO2
R2
R2
NH2
NH
N
N
N
N
N
H
N
H
O
v
vi
vii
O
O
O
R2 NH
F
NHR3
NH
R1 NH
R1
R1
N
N
H
H
6
7
8
R3
O
O
NH2
O
NH2
O
R2
R2
N
N
R2
N
N
N
H
N
N
H
viii
ix
x
H
N
NHR3
N
R1
O
NHR3
R1
NH
R1
N
H
NH2
O
9
10
11
Scheme 1. Reagents and conditions: (i) HOBt, DIC, DMF, rt, 12 h; (ii) 55% TFA, DCM, rt, 40 min; (iii) 4-fluoro-3-nitrobenzoic acid, DIEA, NMP, 70 °C, 24 h; (iv) R2NH2, DIC,
HOBt, DMF, rt, 24 h; (v) SnCl2, NMP, rt, 24 h; (vi) 5-fluoro-2-nitrobenzoic acid, NMP, AcOH, 75 °C, 72 h; (vii) R3NH2, NMP, rt, 24 h; (viii) SnCl2, NMP, rt, 24 h; (ix) mixture of 84%
TFA, 8% thioanisole 8% TFMSA, rt, 2 h; (x) concd HCl, MW 100 °C, 5 min, 5–25% overall yield.

L. Malik et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1729–1732
1731
Table 1
In vitro activation of human and rhesus monkey MrgX1 and MrgX2 receptors using R-SAT
O
R3
N
N
R2
H
NH
R1
O
Compound
R1
R2
R3
MrgX1 (human)
MrgX2 (human)
MrgX1 (monkey)
MrgX2 (monkey)
pEC50
Eff (%)
pEC50
Eff (%)
pEC50
Eff (%)
pEC50
Eff (%)
BAM22a
Cortistatin-14
1
2
6.1 0.6
—
5.6 0.3
—
100 14
NA
98 16
NA
—
NAb
100 28
NA
4.5 0.6
—
5.6 0.7
—
50 25
NA
—
NA
86 25
NA
6.3 0.6
—
6.5 0.5
6.1 0.4
—
6.4 0.2
89
NA
3
110 36
50 23
N
12
13
14
15
16
17
18
—
NA
6.4 0.6
6.8 0.4
216 82
87 26
60 19c
NA
—
NA
6.7 0.4
133 66
N
N
N
H
—
NA
N
N
N
N
N
N
N
6.2 1.1
5.9 0.4
7.2 0.5
6.7 0.5
6.5 0.3
6.5 0.6
108 41
94 27
147 20
133 35
135 32
129 27
5.8 0.6
6.0 0.2
6.1 0.0
6.1 0.3
6.1 0.2
6.4 0.3
102
1
—
NA
NH
F
F
N
N
—
57 10
—
NA
HN
F3C
F3C
F3C
F3C
N
HN
6.2 0.4
5.7 0.7
—
103
5
157
0
5.4 0.2
85 32
NH
N
HN
HN
HN
28 19
127 30
126 21
124 30
—
NA
HN
HN
NH2
O
NA
—
NA
O
O
NH
19
5.9 0.4
41 13
5.7 0.2
27
4
N
a
BAM22: bovine adrenal medulla peptide 22 and Cortistatin-14 (Ref. 7) were used as reference ligands and set to 100% Eff at the MrgX1 receptor and the MrgX2 receptor,
respectively.
b
NA: no activity at pEC50 > 5.0. pEC50 and efficacy values are the average of at least two independent experiments SD, where each experiment is carried out with
duplicate determinations for each data point.
c
Max efficacy at concentrations <10 lM.
3
2
1
0
2
1
0
hits had R1 originating from either glycine, or preferably valine or
isoleucine. The glycine analog of compound 12 was 50 times less
potent at the MrgX2 receptor.12 Therefore, R1 was limited by using
the amino acids valine and isoleucine in the new analogs synthe-
sized (Table 1). In this limited set of compounds, two selective com-
pounds 18 and 12 displayed increased activity at the MrgX1 and
MrgX2 receptors, respectively, compared to compounds 1 and 2
(Table 1 and Fig. 2). Both compounds had similar potency as the
peptide reference compounds BAM22 and cortistatin-14. The initial
SAR indicated that R1 should be an alkyl group such as isopropyl or
isobutyl. In general, molecules where R3 equaled aryl (12 and 13) or
R2 equaled non-aryl (2, 12, and 13) provided compounds that were
either inactive or showed weak agonist activity at the MrgX1 recep-
tor. This trend was also seen when the original hits were analyzed.
In addition, although the library contained several molecules where
MrgX2
MrgX1
MrgX1
MrgX2
-10 -9
-8
-7
-6
-5
-10 -9
-8
-7
-6
-5
Compound 12
Compound 18
Figure 2. Receptor selectivity of MrgX receptor agonists. Each compound was
tested in concentration response experiments in R-SAT at human MrgX1 and MrgX2
receptors.

1732
L. Malik et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1729–1732
CO2H
be stable in human liver microsomes (Clint 5 lL/min mg). These
NO2
CO2H
compounds also activated the homologous monkey receptor sub-
types with similar potencies and selectivities as the human sub-
types (Table 1).
Thus, despite the fact that rodent homologues of MrgX1 and
MrgX2 receptors do not exist, these compounds, or analogs of
them, could be used preclinically to probe the physiological func-
tions of MrgX1 and MrgX2 receptors.
In conclusion, a new class of potent and selective non-peptider-
gic MrgX1 and MrgX2 receptor agonists has been identified which
display activities comparable with the reference peptides BAM22
and corticostatin-14.
O
H
N
CO2Me
H
i
ii
NO2
NH
NO2
NH
F
iii
CO2Me
H
CF3
NH2
N
O2
N
O
H
N
CO2Me
F
iv
O
HN
H
CO2Me
H
N
HN
CF3
v
Acknowledgment
CF3
N
N
N
O2
We thank Mrs. Sine Mandrup Bertozzi for analytical assistance.
N
N
H
vi
O
15
CO2Me
H
References and notes
NH
C
F3
1. Dong, X.; Han, S.; Zylka, M. J.; Simon, M. I.; Anderson, D. J. Cell 2001, 106, 619.
2. Lembo, P. M.; Grazzini, E.; Groblewski, T.; O’Donnell, D.; Roy, M. O.; Zhang, J.;
Hoffert, C.; Cao, J.; Schmidt, R.; Pelletier, M.; Labarre, M.; Gosselin, M.; Fortin,
Y.; Banville, D.; Shen, S. H.; Strom, P.; Payza, K.; Dray, A.; Walker, P.; Ahmad, S.
Nat. Neurosci. 2002, 5, 201.
3. Robas, N.; Mead, E.; Fidock, M. J. Biol. Chem. 2003, 278, 44400.
4. Spier, A. D.; De Lecea, L. Brain Res. Rev. 2000, 33, 228.
5. Kamohara, M.; Matsuo, A.; Takasaki, J.; Kohda, M.; Matsumoto, M.; Matsumoto,
S.; Soga, T.; Hiyama, H.; Kobori, M.; Katou, M. Biochem. Biophys. Res. Commun.
2005, 20, 1146.
Scheme 3. Reagents and conditions: (i) isoleucine methylester, MeOH, reflux 72 h;
(ii) DIC, HOBt, 3-trifluoromethyl-benzylamine, DMF, rt, 48 h, 80% yield over two
steps; (iii) EtOH, 5% Pd/C, H2, 30% yield; (iv) 5-fluoro-2-nitrobenzaldehyde, NMP/
AcOH 1:1, MW 70 °C, 10 min, 32% yield; (v) N1,N1-dimethylpropane-1,3-diamine,
NMP, MW 100 °C, 10 min 34% yield; (vi) SnCl2, NMP, rt, after 24 h, HCl, MW 100 °C,
10 min, 22% yield.
6. Chen, H.; Ikeda, S. R. J. Neurosci. 2004, 26, 5044.
both R2 and R3 equaled aryl substituents, including permutations
of, for example, 12, 13, and 14, none of these compounds were iden-
tified as actives at the MrgX1 receptor in the initial screen. MrgX1
receptor agonists were preferred when R2 equals phenylpiperazine
(14 or 15). While most meta-CF3-benzyl amine containing com-
pounds activated both receptors (16, 17, and 19), compound 18
was a potent and selective MrgX1 receptor agonist.
7. Burstein, E. S.; Ott, T. R.; Feddock, M.; Ma, J. N.; Fuhs, S.; Wong, S.; Schiffer, H.
H.; Brann, M. R.; Nash, N. R. Br. J. Pharmacol. 2006, 147, 73.
8. Zhang, L.; Taylor, N.; Xie, Y.; Ford, R.; Johnson, J.; Paulsen, J. E.; Bates, B. Mol.
Brain. Res. 2005, 133, 187.
9. Kunapuli, P.; Lee, S.; Zheng, W.; Alberts, M.; Kornienko, O.; Mull, R.; Kreamer,
A.; Hwang, J. I.; Simon, M. I.; Strulovici, B. Anal. Biochem. 2006, 351, 50.
10. 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. Curr.
Pharm. Des. 2006, 12, 1717.
The agonist activity of compound 12 was also confirmed in Ca2+
mobilization assay.12 In addition compound 12 was also found to
11. Lang, H.; Pei, Y. W.O. Patent 2001,023,392, 2001; Chem Abstr. 2001, 247, 341.
12. Result not shown.