Discovery of a small molecule RXFP3/4 agonist that increases food intake in rats upon acute central administration

Article abstract of DOI:10.1016/j.bmcl.2019.02.013

The relaxin family peptide receptors have been implicated in numerous physiological processes including energy homeostasis, cardiac function, wound healing, and reproductive function. Two family members, RXFP3 and RXFP4, are class A GPCRs with endogenous

Full text of DOI:10.1016/j.bmcl.2019.02.013

Accepted Manuscript
Discovery of a Small Molecule RXFP3/4 Agonist That Increases Food Intake
in Rats Upon Acute Central Administration
Brian DeChristopher, Soo-Hee Park, Linh Vong, Derek Bamford, Hyun-Hee
Cho, Rohit Duvadie, Allison Fedolak, Christopher Hogan, Toshiyuki Honda,
Pramod Pandey, Olga Rozhitskaya, Liansheng Su, Elizabeth Tomlinson, Iain
Wallace
PII:
S0960-894X(19)30086-1
https://doi.org/10.1016/j.bmcl.2019.02.013
BMCL 26293
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
21 September 2018
4 January 2019
10 February 2019
Please cite this article as: DeChristopher, B., Park, S-H., Vong, L., Bamford, D., Cho, H-H., Duvadie, R., Fedolak,
A., Hogan, C., Honda, T., Pandey, P., Rozhitskaya, O., Su, L., Tomlinson, E., Wallace, I., Discovery of a Small
Molecule RXFP3/4 Agonist That Increases Food Intake in Rats Upon Acute Central Administration, Bioorganic &
Medicinal Chemistry Letters (2019), doi: https://doi.org/10.1016/j.bmcl.2019.02.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Discovery of a Small Molecule RXFP3/4 Agonist That Increases
Food Intake in Rats Upon Acute Central Administration
Brian DeChristopher,* Soo-Hee Park, Linh Vong, Derek Bamford, Hyun-Hee Cho, Rohit Duvadie,
Allison Fedolak, Christopher Hogan, Toshiyuki Honda, Pramod Pandey, Olga Rozhitskaya, Liansheng
Su, Elizabeth Tomlinson, and Iain Wallace
Novartis Institutes for BioMedical Research, 250 Massachusetts Avenue, Cambridge, MA 02139.
KEYWORDS: RXFP3, RXFP4, Relaxin-3, Insulin-like peptide
ABSTRACT: The relaxin family peptide receptors have been implicated in numerous physiological processes including energy
homeostasis, cardiac function, wound healing, and reproductive function. Two family members, RXFP3 and RXFP4, are class A
GPCRs with endogenous peptide ligands (relaxin-3 and insulin-like peptide 5 (INSL5), respectively). Polymorphisms in relaxin-3
and RXFP3 have been associated with obesity, diabetes, and hypercholesterolemia. Moreover, central administration of relaxin-3 in
rats has been shown to increase food intake, leading to body weight gain. Reported RXFP3 and RXFP4 ligands have been restricted
to peptides (both endogenous and synthetic) as well as a low molecular weight positive allosteric modulator requiring a non-
endogenous orthosteric ligand. Described here is the discovery of the first potent low molecular weight dual agonists of RXFP3/4.
The scaffold identified is competitive with a chimeric relaxin-3/INSL5 peptide for RXFP3 binding, elicits similar downstream
signaling as relaxin-3, and increases food intake in rats following acute central administration. This is the first report of small
molecule RXFP3/4 agonism.
The relaxin family peptide receptors (RXFPs) are G protein-coupled receptors with peptide ligands that play roles in diverse
physiological processes, including those controlled by central nervous, cardiovascular, reproductive, and metabolic systems.1 This
array of associated activities makes the RXFPs potential therapeutic targets for neuroscience, cardiovascular, and metabolic
indications. Among these receptors, RXFP3 has been linked to appetite regulation,2,3 arousal, stress responses, anxiety,4 memory,5
and depression.6 RXFP4, the most closely related family member to RXFP3, has been linked to insulin secretion and appetite.7,8 The
endogenous peptide ligands of RXFP3 and RXFP4 are relaxin-3 and insulin-like peptide 5 (INSL5), respectively. These complex
peptides, which contain two amino acid chains linked by both inter- and intra-chain disulfide bridges, have been used extensively to
establish the involvement of RXFP3 and RXFP4 in the various physiological processes detailed above.9,10
Despite the attractiveness of RXFP3 and RXFP4 as therapeutic targets, ligands of these receptors have been limited to the
endogenous peptides or their analogs.11,12,13 In the case of RXFP3, a positive allosteric modulator requiring a non-endogenous relaxin-
3 analog for function has also been reported.14 Whereas significant progress has been made in the understanding of structure activity
relationships of relaxin-3 and INSL5 as well as the design of simplified agonist and antagonist peptide analogs (with varying degrees
of RXFP3/4 selectivity), there remains a need to identify small molecule ligands for these targets. Such chemical matter could
potentially improve access to specific tissues of interest, diversify routes of administration, improve pharmacokinetic properties, and
simplify synthetic access relative to peptide ligands. Toward this end, a high throughput screen was initiated to identify low molecular
weight compounds capable of phenocopying relaxin-3 both in vitro and in vivo. Detailed herein is the identification of a hit scaffold
capable of agonizing both RXFP3 and RXFP4 in vitro as well as the discovery of a potent compound showing a relaxin-3-like
signaling profile in vitro that also increases food intake in rats following acute intracerebroventricular (ICV) administration.

Figure 1. Low molecular weight RXFP3/4 ligands with their corresponding EC50 values for cAMP assays in RXFP3 and RXFP4-
overexpressing cells, respectively. For the RXFP3 assay, n = 26 (compound 1), n = 3 (compound 2), n = 2 (compound 3), and n = 11
(compound 4). For the RXFP4 assay, n = 10 (compound 1), n = 2 (compound 2), n = 1 (compound 3), and n = 2 (compound 4). EC50 values
represent an average of separate experiments when n > 1.
Compound 1 (figure 1)15 was identified from a 70 thousand compound screen using a homogeneous time resolved fluorescence
(HTRF) assay of inhibition of forskolin-stimulated cAMP accumulation in RXFP3-overexpressing CHO-K1 cells (for 1, RXFP3 EC50
= 2.7 M, Emax = 97% of relaxin-3). Subsequently, compound 1 was found to bind competitively with [125I]-R3/I5 (a chimeric peptide
consisting of the relaxin-3 B-chain and the INSL5 A-chain)16 to RXFP3 with a Ki value of 1.7 µM (figure 2). This ability to compete
with [125I]-R3/I5 for RXFP3 binding is shared by relaxin-3 (figure 2). Additional testing revealed that 1 was capable of inhibiting
cAMP accumulation in RXFP4-overexpressing CHO-K1 cells (EC50 = 58 nM, Emax = 99% of relaxin-3), indicating that this scaffold
is capable of RXFP3/4 dual agonism. RXFP3- and RXFP4-overexpressing cells were shown to have comparable receptor expression
levels (see supplemental table 1 in the supporting information).
Medicinal chemistry optimization was undertaken to improve the potency of 1 for expanded in vitro and in vivo validation using
relaxin-3 as a positive control. This optimization identified the indole 5 position as tunable for both RXFP3 and RXFP4 potency, as
evidenced by 5-hydroxy indole-containing compound 2 (figure 1). Replacement of the dichlorophenethyl amine of 1 with a 4-
chlorophenethyl amine (compound 3, figure 1) was also well tolerated. Finally, installation of an ethyl group at the indole 7 position
in combination with the modifications shown in compounds 2 and 3 provided the potent compound 4, which had average cAMP
inhibition EC50 values of 82 nM for RXFP3 (Emax = 105% of relaxin-3) and 2 nM for RXFP4 (Emax = 100% of relaxin-3).
Figure 2. Dose responsive competitive displacement of [125I]-R3/I5 from RXFP3-expressing cell membranes by both relaxin-3 and
compound 1.
Given that the RXFP3 and RXFP4 cAMP EC50 values shown in figure 1 represent averages of separate cellular experiments (n
values shown in the figure 1 caption), the potencies of compound 1, compound 4, and relaxin-3 were compared directly in the same
experiment. In the case of RXFP3 (figure 3A), compound 4 (EC50 = 12.8 nM) was 273-fold more potent than compound 1 (EC50
=
3.5 M). These values are compared to relaxin-3, which had an EC50 = 0.2 nM. For RXFP4 (figure 3B), compound 4 (EC50 = 1.6
nM) was 9-fold more potent than compound 1 (EC50 = 14.6 nM). These values are compared to relaxin-3, which had an RXFP4 EC50
= 0.7 nM. This data further supports the improved cAMP potency of compound 4 relative to compound 1 that was observed in the
aggregate data shown in figure 1 (on average, ~30-fold potency improvement was observed for compound 4 versus compound 1 in
both the RXFP3 and RXFP4 cAMP assays).

Figure 3. Direct comparison of inhibition of forskolin-stimulated cAMP accumulation by relaxin-3, compound 1, and compound 4 in
CHO-K1 cells stably over-expressing RXFP3 (A) and RXFP4 (B).
The observed selectivity of 4 favoring RXFP4 over RXFP3 was confirmed by [35S]GTPS binding experiments using membranes
from CHO-K1 cells overexpressing either RXFP3 or RXFP4 (figure 4). GTPS binding EC50 values were 119 nM for RXFP3 (figure
4A) and 11 nM for RXFP4 (figure 4B), consistent with the trend observed in the cAMP EC50 values. 4 also induced -arrestin
recruitment (EC50 = 4.9 M relative to an EC50 of 83 nM for relaxin-3) in an RXFP3 overexpressing U2OS cell line (figure 5A).
These results indicate that 4 engages both G protein-mediated cAMP signaling and G protein-independent -arrestin recruitment, a
profile shared by relaxin-3.

Figure 4. Dose-responsive GTPS binding induced by both relaxin-3 and compound 4 to CHO-K1 membranes overexpressing RXFP3
(A) or RXFP4 (B).
The activation of MAPKs (ERK1/2, JNK1, and p38-MAPK) was examined following addition of relaxin-3 (100 nM) and 4 (5 M)
to assess functional selectivity of 4. Relaxin-3 and 4 activated ERK1/2 and JNK (figure 5B), but not p38MAPK (data not shown) in
RXFP3-overexpressing CHO-K1 cells. Maximal activation of ERK1/2 and JNK with relaxin-3 and 4 occurred at 2 minutes after
addition. No ERK1/2 or JNK1/2 responses were detected in parental CHO-K1 cells treated with relaxin-3 or 4 (data not shown).
Western blots confirmed that the total JNK or ERK1/2 in treated and untreated RXFP3-overexpressing CHO-K1 cells was not altered.
Based on these results, 4 displays a similar signaling fingerprint compared to relaxin-3.
Figure 5. (A) Comparison of the ability of relaxin-3 and compound 4 to induce -arrestin recruitment activity in U2OS cells stably expressing
human RXFP3. (B) Comparison of the time-course of ERK and JNK phosphorylation following treatment of CHO-K1 cells expressing
human RXFP3 with relaxin-3 (100 nM) or compound 4 (5 M).
Central administration of relaxin-3 has been shown to stimulate food intake in satiated rats.2 To determine if 4 would elicit a
similar effect on food intake, 1 mol of 4 was administered ICV to satiated rats (figure 6). Between 0 and 2 hours post-administration,
relaxin-3 (figure 6A) and 4 (figure 6B) both elicited an increase in food intake (2.9 ± 0.7 grams for relaxin-3 versus 0.8 ± 0.3 grams
for artificial cerebrospinal fluid (aCSF)-treated rats; 3.5 ± 0.6 grams for 4 versus 1.7 ± 0.5 grams for 75% DMSO/25% aCSF-treated
rats). This increase was still observed at 4 hours post-administration of relaxin-3 (4.6 ± 0.9 grams versus 1.7 ± 0.4 grams for aCSF-
treated rats (figure 6A)). For 4, the increase in food intake was still observed at both 4 (7.1 ± 1.1 grams versus 2.5 ± 0.6 grams for
75% DMSO/25% aCSF-treated rats) and 24 (35.3 ± 2.0 grams versus 24.1 ± 2.7 grams for 75% DMSO/25% aCSF-treated rats) hours
post-administration (figure 6B).

Figure 6. Effects of relaxin-3 (A) and compound 4 (B) on food intake in Sprague-Dawley rats following ICV administration; *p <
0.05, **p < 0.01.
The synthesis of 4 is shown in scheme 1. Installation of the ethyl group onto 1-methoxy-4-nitrobenzene 5 was followed by
treatment with vinylmagnesium bromide to form indole 7. Next, Vilsmeier-Haack chemistry was used for installation of the indole
3-position aldehyde to form 8. Removal of the methyl group from the 5-methoxy moiety was then followed by installation of
aminoguanidine 10 (see supporting information for synthesis details) to complete the synthesis of 4.
Scheme 1. Synthesis of compound 4.
NO2
NO2
MeO
b
a
c
In summary, high throughput screening efforts identified the
first known low molecular weight, R3/I5-competitive RXFP3
agonist scaffold capable of inhibiting forskolin-stimulated
cAMP production in both RXFP3- and RXFP4- overexpressing
CHO-K1 cells. The potency of this scaffold was optimized to
provide a tool compound 4 that had a downstream signaling
profile consistent with the endogenous RXFP3 ligand relaxin-3.
Moreover, 4 was capable of stimulating food intake following
ICV administration in rats. This orexigenic effect is consistent
with that observed with relaxin-3 both in this study and
previously.2,3 Compound 4 represents a scaffold that could be
used for further mechanistic studies or target validation around
N
H
OMe
5
OMe
6
7
H
H
O
O
Cl
MeO
HO
NH
e
d
4
+
HN
NH2
N
N
N
H
H
H
10
8
9
: (a) EtMgCl, DDQ, THF, -15 oC, 21%; (b) Vinylmagnesium bromide,
THF, -45 oC, 14%; (c) Oxalyl chloride, DMF, 0-100 oC, 55%; (d) BBr3, DCM, -78 oC to rt;
Reagents and Conditions
(e) TsOH.H2O, MeOH, rt, 59% (two steps).
RXFP3/4 and as a starting point for additional medicinal chemistry.
Corresponding Author

*badechri@gmail.com
Conflicts of Interest
The authors report no conflicts of interest associated with this work.
References
1. Bathgate, R. A. D.; Halls, M. L.; van der Westhuizen, G. E.; Callander, G. E.; Kocan, M.; Summers, R. J. “Relaxin family peptides and their
receptors” Phys. Rev. 2013, 93, 405-480.
2. Ganella, D. E.; Ryan, P. J.; Bathgate, R. A. D.; Gundlach, A. L. “Increased feeding and body weight gain in rats after acute and chronic activation
of RXFP3 by relaxin-3 and receptor-selective peptides: functional and therapeutic implications” Behav. Pharmacol. 2012, 23, 516-525.
3. de Avila, C.; Chometton, S.; Lenglos, C.; Calvez, J. ; Gundlach, A. L.; Timofeeva, E. “Differential effects of relaxin-3 and a selective relaxin-3
receptor agonist on food and water intake and hypothalamic neuronal activity in rats” Behav. Brain Res. 2018, 336, 135-144.
4. Kumar, J. R.; Rajkumar, R.; Jayakody, T.; Marwari, S.; Hong, J. M.; Ma, S.; Gundlach, A. L.; Lai, M. K. P.; Dawe, G. S. “Relaxin’ the brain: a
case for targeting the nucleus incertus network and relaxin-3/RXFP3 system in neuropsychiatric disorders” Br. J. Pharmacol. 2017, 174, 1061-1076.
5. Albert-Gasco, H.; Garcia-Aviles, A.; Moustafa, S.; Sanchez-Sarasua, S.; Gundlach, A. L.; Olucha-Bordonau, F. E.; Sanchez-Perez, A. M. “Central
relaxin-3 receptor (RXFP3) activation increases ERK phosphorylation in septal cholinergic neurons and impairs spatial working memory” Brain
Struct. Funct. 2017, 222, 449-463.
6. Smith, C. M.; Walker, A. W.; Hosken, I. T.; Chua, B. E.; Zhang, C.; Haidar, M.; Gundlach, A. L. “Relaxin-3/RXFP3 networks: an emerging target
for the treatment of depression and other neuropsychiatric diseases?” Front. Pharmacol. 2014, 5, 1-17.
7. Luo, X.; Li, T.; Zhu, Y.; Dai, Y.; Zhao, J.; Guo, Z.-Y.; Wang, M.-W. “The insulinotropic effect of insulin-like peptide 5 in vitro and in vivo”
Biochem. J. 2015, 466, 467-473.
8. Grosse, J.; Heffron, H.; Burling, K.; Hossain, M. A.; Habib, A. M.; Rogers, G. J.; Richards, P.; Larder, R.; Rimmington, D.; Adriaenssens, A. A.;
Parton, L.; Powell, J.; Binda, M.; Colledge, W. H.; Doran, J.; Toyoda, Y.; Wade, J. D.; Aparicio, S.; Carlton, M. B. L.; Coll, A. P.; Reimann, F.;
O’Rahilly, S.; Gribble, F. M. “Insulin-like peptide 5 is an orexigenic gastrointestinal hormone” Proc. Natl. Acad. Sci. USA 2014, 111, 11133-11138.
9. Bathgate, R. A. D.; Lin, F.; Hanson, N. F.; Otvos, L.; Guidolin, A.; Giannakis, C.; Bastiras, S.; Layfield, S. L.; Ferraro, T.; Ma, S.; Zhao, C.;
Gundlach, A. L.; Samuel, C. S.; Tregear, G. W.; Wade, J. D. “Relaxin-3: Improved synthesis strategy and demonstration of its high-affinity interaction
with the relaxin receptor LGR7 both in vitro and in vivo” Biochemistry 2006, 45, 1043-1053.
10. Hossain, M. A.; Bathgate, R. A. D.; Kong, C. K.; Shabanpoor, F.; Zhang, S.; Haugaard,-Jonsson, L. M.; Rosengren, K. J.; Tregear, G. W.; Wade,
J. W. “Synthesis, conformation, and activity of human insulin-like peptide 5 (INSL5)” ChemBioChem 2008, 9, 1816-1822.
11. Patil, N. A.; Rosengren, K. J.; Separovic, F.; Wade, J. D.; Bathgate, R. A. D.; Hossain, M. A. “Relaxin family peptides: structure-activity
relationship studies” Br. J. Pharmacol. 2017, 174, 950-961.
12. Jayakody, T.; Marwari, S.; Lakshminarayanan, R.; Tan, F. C. K.; Johannes, C. W.; Dymock, B. W.; Poulsen, A.; Herr, D. R.; Dawe, G. S.
“Hydrocarbon stapled B chain analogues of relaxin-3 retain biological activity” Peptides 2016, 84, 44-57.
13. Wei, D.; Hu, M.-J.; Shao, X.-X.; Wang, J.-H.; Nie, W.-H.; Liu, Y.-L.; Xu, Z.-G.; Guo, Z.-Y. “Development of a selective agonist for relaxin
family peptide receptor 3” Sci. Rep. 2017, 7, 3230.
14. Alvarez-Jaimes, L.; Sutton, S. W.; Nepomuceno, D.; Motley, S. T.; Cik, M.; Stocking, E.; Shoblock, J.; Bonaventure, P. “In vitro pharmacological
characterization of RXFP3 allosterism: an example of probe dependency” PLoS One 2012, 7, e30792.
15. Giger, R. K. A.; Mattes, H. “Certain aminoguanidine compounds, pharmaceutical compositions containing them and their use in treating
gastrointestinal motility disorders and disorders associated with cephalic pain” US 5510353 A1, 1996.
16. Haugaard-Jonsson, L. M.; Hossain, M. A.; Daly, N. L.; Bathgate, R. A. D.; Wade, J. D.; Craik, D. J.; Rosengren, K. J. “Structure of the R3/I5
chimeric relaxin peptide, a selective GPCR135 and GPCR142 agonist” J. Biol. Chem. 2008, 283, 23811-23818.