Development of Nonpeptidic Inverse Agonists of the Ghrelin Receptor (GHSR) Based on the 1,2,4-Triazole Scaffold

Article abstract of DOI:10.1021/acs.jmedchem.9b02122

GHSR controls, among others, growth hormone and insulin secretion, adiposity, feeding, and glucose metabolism. Therefore, an inverse agonist ligand capable of selectively targeting GHSR and reducing its high constitutive activity appears to be a good cand

Full text of DOI:10.1021/acs.jmedchem.9b02122

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Development of non-peptidic inverse agonists of the
ghrelin receptor (GHSR) based on the 1,2,4-triazole scaffold
khoubaib ben haj salah, Mathieu Maingot, Anne-Laure Blayo, Céline M'Kadmi, Marjorie Damian, Sophie
Mary, Sonia Cantel, Jérémie Neasta, Catherine Oiry, Sylvie Péraldi-Roux, Gimena Fernandez, Guadalupe
Garcia Romero, Mario Perello, Jacky Marie, jean-louis Baneres, Jean-Alain Fehrentz, and Séverine Denoyelle
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b02122 • Publication Date (Web): 03 Sep 2020
Downloaded from pubs.acs.org on September 4, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Development of non-peptidic inverse agonists of the
ghrelin receptor (GHSR) based on the 1,2,4-triazole
scaffold
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
†
†
†
Khoubaib Ben Haj Salah, Mathieu Maingot, Anne-Laure Blayo, Céline M’Kadmi, Marjorie
†
†
†
†
†
†
Damian, Sophie Mary, Sonia Cantel, Jérémie Neasta, Catherine Oiry, Sylvie Péraldi-Roux,
‡
‡
‡
†
Gimena Fernandez, Guadalupe García Romero, Mario Perello, Jacky Marie, Jean-Louis
†
†
,†
Banères, Jean-Alain Fehrentz, Séverine Denoyelle*
†
IBMM, Univ Montpellier, CNRS, ENSCM, Faculty of Pharmacy, 34000 Montpellier, France
Laboratory of Neurophysiology of the Multidisciplinary Institute of Cell, La Plata, 1900
‡
Buenos Aires, Argentina
ABSTRACT
GHSR controls, among others, growth hormone and insulin secretion, adiposity, feeding and
glucose metabolism. Therefore, an inverse agonist ligand capable of selectively targeting GHSR
and reducing its high constitutive activity appears to be a good candidate for the treatment of
obesity-related metabolic diseases. In this context, we present a study that led to the development
of several highly potent and selective inverse agonists of GHSR based on the 1,2,4-triazole
scaffold. We demonstrate that, depending on the nature of the substituents on positions 3, 4 and 5,
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
this scaffold leads to ligands that exert an intrinsic inverse agonist activity on GHSR-catalyzed G
protein activation through the stabilization of a specific inactive receptor conformation. Thanks to
an in vivo evaluation, we also show that one of the most promising ligands not only exerts an effect
on insulin secretion in rat pancreatic islets but also affects the orexigenic effects of ghrelin in mice.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
INTRODUCTION
The ghrelin receptor, a G protein-coupled receptor identified as the growth hormone
1
secretagogue receptor (GHSR), is known for regulating numerous central and peripheral
processes including, among others, growth hormone (GH) secretion, adiposity, feeding and
glucose metabolism. Indeed, administration of its endogenous agonist, a stomach-derived 28-
2
amino acid peptide named ghrelin, induces blood glucose elevation and the reduction of
3
circulating insulin in humans and rodents. Moreover, ghrelin is the most potent orexigenic peptide
hormone known and its administration to humans or rodents potently increases food intake.4
Knowing that GHSR exhibits a high constitutive activity, an inverse agonist compound capable of
selectively targeting the receptor and reducing its constitutive activity appears to be a good
candidate for the treatment of obesity-related metabolic diseases. In an effort to find efficient
GHSR ligands, we previously developed a heterocyclic series that presented high affinity and
selectivity for this receptor. These new ligands, based on the 3,4,5-trisubstituted 1,2,4-triazole
1
’
2’
3’
4’
scaffold (Figure 1), provided four points of diversity (R , R , R and R ) and their synthesis,
starting from an optically pure amino acid, comprised a few steps with moderate to good yields.
We demonstrated that the configuration of the chiral center was retained during the synthetic
process.5-7 An intensive structure activity relationship study, carried out over several hundred
triazole compounds, allowed us to evaluate the ability of these compounds to activate or inactivate
GHSR signaling. Moreover, very interestingly, some of our antagonist compounds were able to
2
ACS Paragon Plus Environment

Page 3 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
antagonize food intake elicited by an agonist such as hexarelin, while they were unable to
antagonize GH secretion elicited by the same agonist. This was the case for JMV2959 (Figure 1,
8
compound 1), a GHSR antagonist which has since been broadly described in the literature to hold
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
potent anti-obesity but also anti-addictive properties.
O
HN
R2'
O
R1'
N
R3'
O
4
R4'
*
5
3
N
N
1
2
H N
3
H
N
N
N
Cl
H
N
N
3,4,5-trisubstituted
,2,4-triazole scaffold
1
1
JMV2959
Figure 1. Structure of the 3,4,5-trisubstituted 1,2,4-triazole scaffold and compound 1 (JMV2959)
9
The first described inverse agonist of GHSR was an analog of substance P (SPA). Even though
1
5
7,9
11
[D-Arg ,D-Phe ,D-Trp ,Leu ]-substance P was able to selectively reduce the constitutive activity
of the receptor, it also acted as a biased ligand toward neuropeptide and chemokine receptors.10-11
However, after several truncations of the original sequence, Holst et al. identified the C-terminal
pentapeptide wFwLL-NH (in its carboxamide version) as the peptide core responsible for the
2
12
inverse agonist activity. Moreover, the presence of a lysine residue at the N-terminus of the
peptide core K-(D-1-Nal)-FwLL-NH led to a specific inverse agonist almost as potent as the
2
1
3
substance P analog. Most recently, our team found that LEAP2 (liver-expressed antimicrobial
1
4
peptide 2), an endogenous ligand described as antagonist of GHSR, behaved as inverse agonist
of the receptor and as competitive antagonist of ghrelin-induced inositol phosphate production and
calcium mobilization.15 We also demonstrated that the entire sequence of LEAP2 was not
necessary for its biological activity. Indeed, the N-terminal part alone, namely the LEAP2 (1-14-
NH ) analog, maintained receptor binding and biological activity, and was able to inhibit ghrelin-
2
induced food intake in mice. In addition to these substance P and LEAP2 peptide analogs, only a
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
few non peptidic inverse agonists of GHSR has been reported in the literature so far (for a review
1
6
see ). In order to overcome the lack of potent small molecules capable of selectively reducing the
constitutive activity of GHSR, we decided to search for inverse agonists based on our 1,2,4-triazole
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1’
2’
3’
4’
scaffold. The first structural modulations with minor changes in the R , R , R and R groups
did not lead to inverse agonists. However, the introduction of a secondary chiral center at position
3
of the triazole ring and the elongation with the Leu-Leu dipeptide sequence allowed us to obtain
1
7
a partial inverse agonist with moderate affinity toward the receptor. In light of these encouraging
results, we decided to carry on our investigation by making structural changes on our initial
antagonist, JMV2959. First, we chose to decorate its primary amine with alkyl or carbonyl groups.
We were able to introduce various substituents including alkanes, alkenes, substituted aromatic
3’
rings and heterocycles. Then, we modulated the R substituent of the 1,2,4-triazole scaffold.
Herein we report the design, the synthesis and the biological evaluation of this new series of
ligands. Having characterized all the compounds by NMR, LC-MS and HRMS, an in vitro SAR
study was performed in order to determine their pharmacological properties and their ability to
reduce the basal activity of GHSR. In order to determine whether this specific 1,2,4-triazole
scaffold was active on endogenously expressed GHSR, we tested one of the most effective ligands
for its ability to modulate the insulinostatic effect of ghrelin in rat isolated pancreatic islets. At last,
we investigated the ability of this lead compound to affect food intake under different experimental
settings in mice.
4
ACS Paragon Plus Environment

Page 5 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULTS AND DISCUSSION
Design and Syntheses. We first designed a set of ligands (compounds 2-8) on which the primary
amine of JMV2959 (compound 1) was substituted by various alkyl groups. Compounds 2-8 were
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
18
synthesized following the same procedure previously described to provide compound 1. Briefly,
the triazole derivative (iv) was obtained in four steps starting from Boc-(D)-Trp-OH as shown in
Scheme 1. First, Boc-(D)-Trp-OH (i) was reacted with 4-methoxybenzylamine (ii) in the presence
of BOP reagent and DIEA in DCM for 2 hours at room temperature. The resulted amide was
transformed into the corresponding thioamide (iii) using the Lawesson’s reagent in DME at 80 °C
for 3 hours. The cyclization into the triazole heterocycle was performed by reacting the thioamide
ˉ
+
(iii) with 3-phenylpropanehydrazide in the presence of silver benzoate (PhCO Ag ) and acetic
2
19
acid (AcOH) in DCM at room temperature for 12 hours. After removal of the Boc protecting
group using a solution of 4N HCl in dioxane at 50 °C for 30 minutes, the structural diversity was
introduced on the triazole derivative (iv) by coupling various N-substituted glycines in the presence
of BOP reagent and DIEA in DCM for 2 hours at room temperature. Depending on the structure
of the N-substituent on the glycine, a last deprotection step was required to afford the final
compounds: a solution of 4N HCl in dioxane at 50 °C for the removal of the Boc and tBu protecting
groups and a solution of diethylamine (Et NH) in DMF at room temperature for the removal of the
2
Fmoc protecting group.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 1. Synthesis of Compounds 1-8a
H
N
O
OH
H2N
BocHN
+
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
i
ii
a, b
H
N
O
H
N
BocHN
S
iii
8
1
FmocHN
c, d
OH
O
BocHN
N
O
OH
e, f
e, d
O
OH
HN
OH
N
N
H
O
Boc
O
7
2
e
N
e, d
HCl.H2N
HO
O
N
N
iv
tBuO
OH
N
N
O
Fmoc
O
e, d, f
e
OH
OH
N
O
N
e
e, d
Boc
O
3
6
5
4
a
Reagents and conditions: (a) BOP, DIEA, DCM, 2 h, 88%; (b) Lawesson’s reagent, DME, 80
ˉ
+
°
C, 3h, 74%; (c) 3-phenylpropanehydrazide, PhCO Ag , AcOH, DCM, 12 h, 81%; (d) 4N
2
HCl/dioxane, 50 °C, 30 min, 100%; (e) BOP, DIEA, DCM, 2 h, 14-86%; (f) Et NH, DMF, 2h,
2
1
00%.
The second set of ligands (compounds 9-35) was also designed starting from compound 1
(Scheme 2). A classical peptide coupling reaction between the primary amine of 1 and a large
2
panel of carboxylic acids allowed us to introduce the structural diversity on substituent R . The
coupling reaction was conducted under basic conditions (DIEA) in the presence of HATU reagent
in DCM for 2 hours at room temperature. In the case of ligands decorated with either an amino
function (compounds 11 and 23) or a hydroxyl group (compounds 21 and 22), the protecting group,
6
ACS Paragon Plus Environment

Page 7 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
required to carry on the coupling reaction, was removed in a final step using a solution of 4N HCl
in dioxane at 50 °C for 30 min.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 2. Synthesis of Compounds 9-35a
O
HN
O
N
R2 OH
O
HCl.H2N
N
+
H
N
N
1
O
HN
O
H
a
R2
N
N
N
H
N
N
O
9, 10, 12-20, 24-35
a, b
11, 21-23
a
Reagents and conditions: (a) HATU, DIEA, DCM, 2 h, 32-78%; (b) 4N HCl/dioxane, 50 °C,
2
3
0 min, 100%. For R , see Table 2.
The last set of ligands (compounds 36-46) included several structural modulations but all were
comprised of a -cyano--phenylethylene moiety (potentially substituted in position para of the
2
3
phenyl group) in place of substituent R (Scheme 3). First, we varied the nature of substituent R .
We decided to introduce either a phenyl group (compounds 36-39), a morpholine group
(compounds 40-42), an indole group (compounds 43 and 44), a piperidine group (compound 45)
or a diethylamine group (compound 46). We used appropriate substituted hydrazides
ˉ
+
(
intermediates v) that we reacted with thioamide (iii) in the presence of PhCO Ag and AcOH in
2
DCM at room temperature for 12 hours. Removal of the Boc protecting group led to triazole
intermediates (vi). At this step of the syntheses, we had the possibility to modulate the spacer
between these triazole intermediates and the 2-cyano-3-phenylacrylyl moiety bearing substituent
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
R . While we kept a glycine residue for most of the ligands, we modified the linker of compounds
3
6-39. For instance, compound 36, which had no spacer, was obtained by directly coupling alpha-
cyano-4-methoxycinnamic acid with intermediate (vi) under the same conditions described above.
The spacer of compound 37 was introduced by using 2-hydroxyacetic acid, which allowed us to
make an ester bond with alpha-cyano-4-methoxycinnamic acid in contrast to the usual amide bond.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
For compounds 38 and 39, we simply replaced the methylene bridge by an ethylene bridge
4
(
compound 38) or an ethylidene bridge (compound 39). At last, diversity of substituent R resulted
from the use of various para substituted alpha-cyanocinnamic acids that were reacted with
deprotected beforehand intermediate (vii) via the same classical peptide coupling conditions
previously described.
8
ACS Paragon Plus Environment

Page 9 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 3. Synthesis of Compounds 36-46a
H
N
O
H
H
R3
a, b
N
N
+
BocHN
H2N
S
O
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
iii
v
O
HN
O
N
N
O
H
N N
CN
HN
O
c
3
6
N
R3
HCl.H2N
d, e
O
N
N
HN
vi
O
CN
O
N
O
O
N
f
BocHN
H
N N
X
OH
O
3
7
O
O
HN
HN
R4
O
CN
O
b, g
H
R3
N
R3
N
N
BocHN
X
N
X
N
H
N
N
H
N N
O
vii
X =
38
39
40-46
a
ˉ
+
Reagents and conditions: (a) PhCO Ag , AcOH, DCM, 12 h, 71-88%; (b) 4N HCl/dioxane, 50
2
°C, 30 min, 100%; (c) alpha-cyano-4-methoxycinnamic acid, HATU, DIEA, DCM, 2 h, 36%; (d)
-hydroxyacetic acid, HATU, DIEA, DCM, 2h, 58%; (e) alpha-cyano-4-methoxycinnamic acid,
EDC, DIEA, DCM, 2 h, 39%; (f) HATU, DIEA, DCM, 2 h, 62%; (g) para substituted alpha-
2
3
4
cyanocinnamic acids, HATU, DIEA, DCM, 2h, 32-72%. For R and R , see Table 3.
All final compounds 2-46 (Figure 2) were purified by reversed-phase preparative HPLC.
Depending on the intricacy of the purification, the purity assessed by analytical reversed phase
C18 HPLC was at least 95% for most compounds and up to 99% for the rest, and the structures of
1
13
the compounds were confirmed by HRMS, H NMR, and C NMR (see supporting information
for more details).
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
H2N
O
H
H
N
H
H
N
N
N
N
N
N
HO
3
4
5
6
7
8
2
H
H
N
H
H
H
H
H
N
H
N
N
N
NC
N
N
NC
NC
H2N
O
O
O
O
O
O
O
10
11
12
13
NC
14
15
9
CN
CN
H
H
N
H
N
H
H
N
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
N
N
O
F
O
O
O
O
O 24
16
1
7
19
20
O
CN
H
H
H
N
N
O
N
O
CN
O
2
5
O
36
37
O
26
O
O
N
O
H
N
HO
O
O
N
N
NH2
H
N
N
H
N
H
N
N
H
N
O
NC
HO
21
O 22
O 23
O
18
HO
CN
CN
CN
CN
H
H
H
N
H
N
N
N
2
7
O
28
O
29
O
30
34
O
N
F
CN
CN
CN
CN
H
H
H
H
N
N
N
N
S
N
H
3
1
O
32
O
33
O
O
O
HN
CN
O
N
H
CN
N
N
H
N
H
CN
O
O
35
38
39
O
H
N
O
HO
CN
CN
CN
H
H
H
N
N
N
O
O
N
N
O
4
0
O
41
O
O
N
42
44
H
N
N
H
N
HO
O
CN
CN
H
H
N
NH
N
O
N
43
O
O
N
H
N
N
H
N
O
O
N
N
CN
H
N
O
N
N
4
5
46
H
N
N
O
Figure 2. Structure of final ligands 2-46. Inverse agonist ligands are framed by dotted lines.
Evaluation of the Ligands in GHSR Expressing HEK293T Cells. Binding and
Pharmacological Activities of the Ligands. We first analyzed in vitro whether the different
compounds directly bound to GHSR stably expressed in HEK293T cells (Tables 1-3). Ki values
were determined from binding competition experiments with fluorescent ghrelin using the
10
ACS Paragon Plus Environment

Page 11 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
homogenous time-resolved fluorescence (HTRF) assay as previously described.20 We also
-5
measured the Inositol Phosphate 1 (IP1) response of each compound (10 M) and calculated it as
-7
a percentage of the maximal stimulation promoted by ghrelin (10 M). We compared this value
with the percentage of the basal response of the receptor (value in square brackets) in order to
determine the pharmacological activity of each compound toward the receptor. The high basal
level of IP1 production in GHSR-expressing cells, representing in our in vitro experiments a range
of 65−75% of the maximal IP1 production promoted by ghrelin, resulted from the high constitutive
activity of GHSR as previously shown.9, 21 A ligand was considered as agonist when it induced a
minimum of 95% stimulation of the receptor compared to ghrelin, as inverse agonist when it
reduced the basal activity of the receptor, and as antagonist when the IP1 response was identical
to the basal level. We defined as “partial” any agonist/inverse agonist that elicited activity below
the maximum level defined as that triggered by a reference compound, usually the natural ligand.
Hence, in the case of agonists, the natural ligand ghrelin was considered as the reference full
agonist that elicits the maximum signal at the interrogated pathway, here IP1 production. To take
into account the experimental error, we considered as partial agonist any compound that induced
a response comprised between the measured basal level and 94% of the maximal effect induced
by ghrelin. The situation was more complex for the inverse agonist activity because of the lack of
any clearly referenced compound. Therefore, in this case, we considered the natural GHSR inverse
agonist LEAP2 as the reference compound that elicits the maximal activity. This peptide decreases
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
5
the basal level of IP1 production by 50% (for more details see ). To be noted, this maximal effect
is closely related to that also triggered by the reference synthetic inverse agonist compound K-(D-
1
-Nal)-FwLL-NH . In our assay, the basal activity of GHSR corresponded to 65-75% of the
2
maximal IP1 production triggered by ghrelin. Hence, a 50% decrease would correspond to a final
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
level of IP1 production of ca. 35%. We thus considered as partial inverse agonist any compound
with a response comprised between 35% and the basal level in the absence of ligand. Parent
compound JMV2959 (ligand 1) was taken as reference. First, we tested the compounds with
alkylated amine (ligands 2-8). As shown in Table 1, all ligands lost some affinity toward the
receptor compared to JMV2959. Steric hindrance did not seem to play a role since ligands 3, 4, 5,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
and 8 presented comparable Ki values while bearing more or less bulky substituents. However,
the presence of a carboxyl group (ligand 6) should be prohibited as it drastically reduced the
affinity of the compound (Ki = 617 nM). Concerning the activity of the ligands, all behaved as
agonists or partial agonists of GHSR. In light of these results, alkylation of the amine function of
compound JMV2959 did not appear as a valid approach to obtain inverse agonists of the receptor.
Therefore, we decided to acylate the amine function with various carboxylic acids offering a wide
2
range of substituents R (Table 2). In terms of affinity, this approach seemed promising as it
lowered the Ki value for most of the ligands. Some ligands even showed Ki values below the
nanomolar range (ligands 14, 30 and 32). Only a cyano group directly linked to the amide bond
prevented the compound to bind to the receptor (ligand 12, Ki = 1.6 μM). It is noteworthy that all
the ligands bearing a -cyanoethylene group substituted in position  by an unsaturated ring such
as phenyl (ligands 27-32), thiophene (ligand 33), pyrrole (ligand 34) or indole (ligand 35), behaved
as inverse agonists of the receptor. The substituent in position para of the phenyl group (ligands
28-32) did not seem to affect the affinity nor the activity of the ligands. As shown in Table 3,
modification of the spacer (Y) between the triazole core and the 2-cyano-3-phenylacrylyl moiety
4
bearing substituent R had an impact on the properties of the ligands. Indeed, shortening the spacer
(ligand 36) or replacing the amide bond by an ester bond (ligand 37) cannot be envisaged since it
3
leads to partial agonists. At last, replacement of a phenyl group by various R substituents such as
12
ACS Paragon Plus Environment

Page 13 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
morpholine, indole, piperidine and diethylamine, led to inverse agonists of the receptor (ligands
4
0-42, 44-46), except ligand 43, which behaved as a partial agonist. For most ligands, these
structural modifications maintained good affinities toward the receptor (Ki = 6.3-14 nM).
Nevertheless, the morpholine substituent seemed to be a little less well tolerated as it lowered
affinities of ligands 40 and 41 (Ki = 67 and 64 nM, respectively).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. Binding and Pharmacological Activity of Ligands 1-8a
O
HN
O
R1
N
N
H
N N
IP-[basal]c
%)
Pharmacological
profile
b
Ligand
R1
Ki (nM)
(
1
2
3
4
5
6
7
-NH
2
.HCl
43  10
65-[65]
100-[76]
96-[76]
82-[61]
88-[72]
80-[72]
88-[72]
antagonist
-NH-CH
-N(CH
-NH-CH
3
96  22
agonist
)
3 2
200  28
300  32
230  31
617  41
136  19
agonist
2
-CH
3
partial agonist
partial agonist
partial agonist
partial agonist
-N(CH
2
-CH
3
)
2
O
NH
HO
NH
N
H2N
8
220  21
85-[72]
partial agonist
a
Values reported in the table were obtained from experiments performed in HEK293T cells
b
stably expressing GHSR. Ki values were determined from HTRF competition binding assays.
Values are expressed as means  SEM of two independent experiments, each experiment being
c
performed in triplicate. Activities were determined by HTRF inositol phosphate (IP) assay. IP
-5
response of each ligand (10 M) is given as percentage of the maximal stimulation promoted by
-7
ghrelin (10 M). The percentage in square brackets refers to the basal response of the receptor also
given as percentage of the maximal ghrelin response.
Table 2. Binding and Pharmacological Activity of Ligands 9-35a
O
HN
O
H
R2
N
N
N
H
N N
O
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
IP-[basal]c
Pharmacological
profile
b
Ligand
R2
Ki (nM)
(%)
9
-H
12  2
79-[62]
93-[62]
85-[64]
76-[67]
partial agonist
partial agonist
partial agonist
partial agonist
10
-CH
-CH
-CN
3
2
10  3
1
1
1
2
-NH
2
130  22
1600  112
1
1
1
3
4
5
4  1
90-[69]
83-[79]
70-[70]
partial agonist
partial agonist
antagonist
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NC
NC
0.45  0.2
3.4  1
O
1
6
28  8
77-[74]
72-[70]
partial agonist
antagonist
NC
CN
17
166  23
O
N
partial inverse
agonist
1
1
8
9
6.6  2
37-[74]
N
NC
115  26
6.3  2
94-[66]
81-[70]
partial agonist
partial agonist
NC
20
N
partial inverse
agonist
2
2
1
2
23  6
35  8
17  7
72-[77]
60-[77]
56-[67]
HO
HO
partial inverse
agonist
O
NH2
partial inverse
agonist
23
2
4
370  36
56  15
79-[71]
74-[71]
partial agonist
antagonist
F
25
2
6
130  26
9  2
95-[69]
25-[77]
19-[74]
22-[77]
12-[77]
agonist
CN
27
inverse agonist
inverse agonist
inverse agonist
inverse agonist
CN
2
2
3
8
9
0
1.0  0.3
3  1
HO
O
CN
CN
CN
0.5  0.2
N
31
2.4  1
22-[70]
inverse agonist
F
CN
3
3
3
2
3
4
0.6  0.2
9.7  2
8.5  2
23-[74]
31-[70]
23-[70]
inverse agonist
inverse agonist
inverse agonist
CN
CN
S
N
H
HN
CN
3
5
6.3 2
19-[74]
inverse agonist
a
Values reported in the table were obtained from experiments performed in HEK293T cells
stably expressing GHSR. Ki values were determined from HTRF competition binding assays.
Values are expressed as means  SEM of two independent experiments, each experiment being
performed in triplicate. Activities were determined by HTRF inositol phosphate (IP) assay. IP
response of each ligand (10 M) is given as percentage of the maximal stimulation promoted by
ghrelin (10 M). The percentage in square brackets refers to the basal response of the receptor also
given as percentage of the maximal ghrelin response.
b
c
-5
-7
14
ACS Paragon Plus Environment

Page 15 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 3. Binding and Pharmacological Activity of Ligands 36-46a
O
HN
R4
CN
Y
N
R3
N
H
N N
O
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
IP-[basal]c
Pharmacol.
profile
b
Ligand
R3
R4
Y
Ki (nM)
(%)
36
-OCH
3
3
122  23
13  3
95-[68]
80-[71]
partial agonist
partial agonist
O
3
3
7
8
-OCH
-OCH
O
O
3
3
N
14  6
3  1
15-[71]
15-[71]
44-[65]
33-[65]
29-[65]
inverse agonist
inverse agonist
H
O
H
N
39
-OCH
-H
O
O
O
O
O
O
partial inverse
agonist
H
N
4
4
4
0
1
2
N
N
67  12
64  11
6.3  2
H
N
-OH
inverse agonist
inverse agonist
H
N
-N(CH
-OH
3 2
)
N
NH
O
O
H
N
43
30  8
82-[65]
7-[71]
partial agonist
inverse agonist
NH
H
N
4
4
-OCH
3
1.4  0.2
O
O
H
N
4
4
5
6
-OCH
-OCH
3
3
14  2
10  3
25-[65]
29-[65]
inverse agonist
inverse agonist
N
N
H
N
a
Values reported in the table were obtained from experiments performed in HEK293T cells
b
stably expressing GHSR. Ki values were determined from HTRF competition binding assays.
Values are expressed as means  SEM of two independent experiments, each experiment being
c
performed in triplicate. Activities were determined by HTRF inositol phosphate (IP) assay. IP
-5
response of each ligand (10 M) is given as percentage of the maximal stimulation promoted by
-7
ghrelin (10 M). The percentage in square brackets refers to the basal response of the receptor also
given as percentage of the maximal ghrelin response.
Pharmacological Activities of the Inverse Agonist Ligands. We selected the ligands that behaved
as inverse agonists and further tested their ability to affect the basal level of IP1 production in
HEK293T cells stably expressing GHSR (Table 4, Figure 3). We determined Emax (maximal
inhibition effect on the basal response promoted by the receptor) and EC values, and compared
5
0
them to reference inverse agonists SPA and K-(D-1-Nal)-FwLL-NH . As shown in Table 4, most
2
of our ligands were at least as potent as the reference compounds. Some were even more potent
than SPA (ligands 28-30, 44). Moreover, ligands described as partial inverse agonists (ligands 21-
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
3) showed less effect on the basal response (Emax values of between 18 and 43% inhibition of the
basal response) than full inverse agonists which were able to reduce the basal response up to more
than 80% (ligands 28-30, 44). As shown in Figure 3, ligands 28, 29, 30 and 44 presented better
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
pharmacological profiles in term of Emax and/or EC values compared to SPA and K-(D-1-Nal)-
50
FwLL-NH2.
Table 4. Pharmacological Activity of Inverse Agonist Ligands 18, 21-23, 27-35, 38-42, 44-46a
IP EC50 b
IP Emaxc
(%)
IP EC50b
(nM)
IP Emaxc
(%)
Ligand
cLogPd
Ligand
cLogPd
5.41
(nM)
SPA
K-(D-1-Nal)-FwLL-NH
66  10
12  3
78
65
55
18
33
43
70
82
83
86
78
76
-
33
34
35
38
39
40
41
42
44
45
46
8.6  4
15.8  6
28  8
3.6  1
1.4  0.6
73  12
42  9
65
73
79
75
78
49
59
64
86
76
69
2
-
4.49
5.59
5.57
5.91
3.13
2.82
3.23
5.44
4.04
3.90
1
2
2
2
2
2
2
3
3
8
1
2
3
7
8
9
0
1
8.2  2
155  32
19  8
4.1
4.85
5.30
4.47
5.49
6.01
5.19
5.34
5.60
5.64
62  12
45  13
21.5  8
27  12
50  18
89  22
14.2  5
6  2
25  9
23  7
35  10
32
a
Values reported in the table were obtained from experiments performed in HEK293T cells
b
stably expressing GHSR. EC values were obtained from concentration-response curves of
inositol phosphate (IP) production (HTRF assay) and are expressed as means  SEM of three
independent experiments, each experiment being performed in triplicate. E
obtained from these plots and are given as percentage inhibition of basal receptor response at a
concentration of 10 M for each compound. cLogP values were calculated using the
MarvinSketch 20.7 software provided by ChemAxon.
5
0
c
values were
max
-5
d
16
ACS Paragon Plus Environment

Page 17 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Concentration-response curves of inverse agonist ligands at promoting IP1 production.
Data are expressed as the percentage of maximal effect of ghrelin on HEK293T cells expressing
GHSR. In this experiment, basal IP1 production of GHSR-expressing cells represents 72.5  1.5%
of the maximal effect triggered by ghrelin. SPA and K-(D-1-Nal)-FwLL-NH described in the
2
literature as inverse agonists are given as references. Data are representative of two independent
experiments, each performed in triplicate.
Activity of Inverse Agonists on the Purified GHSR. GHSR-Mediated G Protein Activation.
We used the purified receptor assembled into lipid nanodiscs in order to evaluate whether the
impact of selected ligands 27, 29, 30 and 44 on GHSR signaling observed in HEK293T cells
resulted from a direct effect on the receptor independently of the cellular context. Note that ligand
2
8 was excluded due to solubility issues, and replaced by ligand 27 (same structure without the
4
methyl substituent R but with a lower cLogP value, see Table 4). The ability of the ligands to
affect GHSR activation was assessed by monitoring GHSR-mediated G protein activation in a
q
GTPγS binding assay (Figure 4). We fixed the maximal GHSR-catalyzed G protein activation with
ghrelin, and we compared our ligands to reference inverse agonist SPA. As expected, the purified
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
apo GHSR assembled into lipid nanodiscs displayed a significant constitutive activity (Figure 4).
Concerning ligands 27, 29, 30 and 44, they were able to significantly reduce G protein activation
q
by the receptor. Ligand 44 has proven to be even more potent than SPA as it almost fully inhibited
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the basal activity of the purified receptor. This indicated that our ligands exerted an inverse agonist
activity on GHSR-dependent basal activation of G . Thus, these data suggest that the inverse
q
agonist activity of our ligands observed in HEK293T cells is directly related to their interaction
with GHSR independently of the cellular context.
Figure 4. Effect of inverse agonist ligands on GHSR-catalyzed G protein activation. BODIPY-
q
FL GTPγS binding to G protein induced by GHSR in lipid nanodiscs in the absence of ligand; in
q
-6
the presence of ghrelin and SPA (all ligands at 10 M); and in the presence of ligands 27, 29, 30
-5
and 44 (all ligands at 10 M). Data are presented as the percentage of maximum BODIPY-FL
fluorescence change measured in the presence of ghrelin, and represent the mean  SEM from
three independent experiments.
18
ACS Paragon Plus Environment

Page 19 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Impact of Inverse Agonists on GHSR Conformation. We previously reported that the
pharmacological properties of GHSR ligands were intricately related to their ability to stabilize
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
22
specific receptor conformations. To assess whether the inverse agonist activity of ligands 27, 29,
3
0 and 44 was indeed associated with their ability to stabilize an inactive conformation similar to
that observed with SPA, we analyzed the conformational features of GHSR using a fluorescence
resonance energy transfer (FRET)-based approach (Figure 5).22 To this end, the purified
monomeric receptor stabilized into nanodiscs was labeled with a fluorescence acceptor (AF488)
and donor (AF350) in the cytoplasmic ends of TM1 and TM6, respectively. As shown in Figure
5, addition of ghrelin or SPA at saturating concentrations induced a significant change in the
FRET-monitored proximity ratio, indicative of a change in the distance between the two
fluorescent probes. These variations were not due to changes in the mobility or orientation of the
fluorophores, as the anisotropy of fluorophores attached to the receptor was not altered by
treatment with ligands. Specifically, the full agonist ghrelin triggered a decrease in the proximity
ratio, whereas the inverse agonist SPA had the opposite effect. As the FRET efficiency is directly
related to the distance between the fluorophores, the increase in the proximity ratio indicates that
SPA binding is associated to a change in GHSR conformation with a decrease in the distance
between the cytoplasmic regions of TM1 and TM6. Such a decrease has been shown to be
2
3
associated to the inactivation process of the ghrelin receptor. All our ligands triggered an increase
in the proximity ratio similar to that observed with SPA, suggesting that these compounds stabilize
similar GHSR conformations. Correlated with GTPγS binding assay results, these data indicate
that our ligands exert an intrinsic inverse agonist activity on GHSR-catalyzed G protein activation
through the stabilization of a specific, inactive receptor conformation.
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Effect of inverse agonist ligands on GHSR conformation. Proximity ratio changes
induced by ghrelin, SPA, and inverse agonist ligands 27, 29, 30 and 44. These changes were
calculated from the FRET signal between the fluorophores in TM1 and TM6 of the purified ghrelin
receptor assembled into lipid nanodiscs (see the Experimental section). The data are representative
of two experiments performed in duplicate, and the error bars represent the SEM.
Evaluation of Ligand 29 on Endogenously Expressed GHSR. In order to perform in vivo
experiments on endogenously expressed GHSR, we initially planned to select ligand 44. Indeed,
this ligand presented not only an excellent pharmacological profile (Ki value of 1.4  0.2 nM, Emax
value of 86% inhibition of the basal level, and EC value of 25  9 nM) but also the best results
5
0
when tested on purified GHSR. Nevertheless, because of its low solubility in water (3.8 mg/L) we
opted for ligand 29. This ligand was selected because (i) it also appeared as one of the most
20
ACS Paragon Plus Environment

Page 21 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
promising compounds in term of pharmacological profile (Ki value of 3  1 nM, Emax value of
83% inhibition of the basal level, and EC value of 27  12 nM), and (ii) it presented an improved
5
0
water solubility (6.5 times more soluble) compared to ligand 44, probably due to the presence of
a hydroxyl substituent on the aryl group (for more details on the water solubility of ligands 29 and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4, see the Supporting Information).
Effect of Ligand 29 in Rat Isolated Pancreatic Islets. To determine whether the substituted 1,2,4-
triazole scaffold was active on native GHSR, we tested the effect of ligand 29 on the modulation
of insulin secretion in rat isolated pancreatic islets (Figure 6). Knowing that ghrelin-evoked GHSR
2
4
activation inhibits glucose-stimulated insulin secretion in isolated islets of Langerhans, we used
this biological system to test ligand 29. Isolated pancreatic islets were therefore incubated with a
stimulating glucose concentration (8.3 mM) in the presence or absence of ghrelin (10 nM) and
ligand 29 (1 μM) and then insulin release was quantified (Figure 6). As previously shown, ghrelin
decreased glucose-stimulated insulin secretion. Importantly, simultaneous treatment with both
ghrelin and ligand 29 blocked the inhibitory action of ghrelin on insulin secretion. These findings
indicate that ligand 29 binds to pancreatic GHSR and prevents its ghrelin-evoked activation.
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Effect of ligand 29 on ghrelin-induced inhibition of insulin secretion in rat pancreatic
islets. Rat pancreatic islets were incubated under basal (2.8 mM glucose) or glucose-stimulated
conditions (8.3 mM glucose) in the presence or absence of ghrelin (10 nM) and ligand 29 (1 µM).
After 1h of incubation, the extracellular medium was collected and insulin concentration was
quantified. The histogram depicts the mean  SEM of insulin secretion expressed in ng/mL/islet.
Two-way ANOVA performed with 8.3 mM glucose data points showed an interaction between
ghrelin and ligand 29 [p < 0.01]. **p < 0.01; ***p < 0.001, Holm-Sidak multiple comparison test,
n = 6.
Effect of Ligand 29 on Food Intake in Mice. First, we tested the effect of K-(D-1-Nal)-FwLL-
NH and ligand 29 on spontaneous food intake during the dark cycle, when mice consume the
2
majority of their food. K-(D-1-Nal)-FwLL-NH was used here as the reference inverse agonist
2
because of its higher specificity compared to SPA. ICV administration of K-(D-1-Nal)-FwLL-NH2
and ligand 29 significantly reduced ad libitum food intake in the early dark phase period, from
22
ACS Paragon Plus Environment

Page 23 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
8:00 h to 23:00 h (Figure 7A) but did not affect overnight food intake, as compared to vehicle
treatment (2623  240, 2179  333 and 2591  197 mg, respectively). Importantly, K-(D-1-Nal)-
FwLL-NH and ligand 29 treatments did not affect body weight nor induce any sickness-like
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
behavior such as a spiky coat, hunched posture, altered breathing rate, labored movements, reduced
activity and/or subdued behavior. Then, we tested the ability of K-(D-1-Nal)-FwLL-NH and
2
ligand 29 to affect ghrelin-induced food intake (Figure 7B). Both GHSR ligands reduced ghrelin-
induced food intake; however, K-(D-1-Nal)-FwLL-NH treatment reduced ghrelin-induced food
2
intake by 67.6  5.6% while ligand 29 treatment reduced ghrelin-induced food intake by 30.3 
4.8%. Thus, these observations suggest that K-(D-1-Nal)-FwLL-NH and ligand 29 behave as
2
GHSR antagonists in vivo. Indeed, since the physiological impact of the constitutive GHSR
activity is uncertain, demonstrating the ability of a given GHSR ligand to display inverse agonist
activity in vivo has proved to be challenging. We have shown that the compensatory hyperphagia
25
that follows a fasting event requires GHSR but occurs in a ghrelin-independent manner. Thus,
we investigated the ability of K-(D-1-Nal)-FwLL-NH and ligand 29 to affect the compensatory
2
hyperphagia in a fasting-refeeding protocol that follows a fasting event (Figure 7C-D). Notably,
K-(D-1-Nal)-FwLL-NH and ligand 29 treatments significantly decreased the cumulative food
2
intake between days 1 and 5 of refeeding (Figure 7C). In particular, K-(D-1-Nal)-FwLL-NH and
2
ligand 29 reduced the compensatory hyperphagia by 20.1  3.5 and 12.4  3.9%, respectively, as
compared to the vehicle treatment (Figure 7D). This in vivo evidence supports the notion that
ligand 29 modulates ghrelin-independent aspects of GHSR activity.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. (A) 5-hour early dark cycle food intake in mice ICV-treated with vehicle, K-(D-1-Nal)-
FwLL-NH or ligand 29. (B) 2-hour food intake in mice ICV-pretreated with vehicle, K-(D-1-Nal)-
2
FwLL-NH or ligand 29 and subsequently ICV-injected with ghrelin. (C) food intake and (D) 5-
2
day cumulative food intake during refeeding of mice ICV-treated with vehicle, K-(D-1-Nal)-
FwLL-NH or ligand 29 during the fasting period. Data represent the mean  SEM and were
2
compared by one-way ANOVA. *p < 0.05, **p < 0.01 and ***p < 0.001 vs. vehicle-treated mice
and ###p < 0.001 vs. K-(D-1-Nal)-FwLL-NH -treated mice, n=33.
2
CONCLUSION
We synthesized a new series of 3,4,5-trisubstituted 1,2,4-triazole derivatives based on antagonist
JMV2959 (compound 1). We modulated the structure by introducing various substituents at
24
ACS Paragon Plus Environment

Page 25 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
different positions on the scaffold and generated 45 new ligands. We carried out an in vitro SAR
study in order to determine the pharmacological properties of the ligands and their ability to reduce
the basal activity of GHSR. We showed that 21 ligands behaved as inverse agonists of GHSR.
Among them, seventeen were at least as potent as the reference compounds K-(D-1-Nal)-FwLL-
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH . More importantly, four ligands (ligands 28-30 and 44) proved to be even more potent than
2
SPA by inhibiting the basal response up to 86%. In order to show a direct interaction between the
ligands and GHSR, we tested the activity of some selected inverse agonists on purified GHSR
assembled into lipid nanodiscs. We proved that these ligands exerted an intrinsic inverse agonist
activity on GHSR-catalyzed G protein activation through the stabilization of a specific inactive
receptor conformation. At last, we investigated the effect of one of the most promising inverse
agonist (ligand 29) on both the modulation of insulin secretion in rat isolated pancreatic islets and
food intake in mice. By directly binding to GHSR, ligand 29 revealed to block the inhibitory action
of ghrelin on insulin secretion. Moreover, ICV administration of ligand 29 reduced ghrelin-
induced food intake in mice, as well as the compensatory hyperphagia that follows a fasting event.
In conclusion, this 3,4,5-trisubstituted 1,2,4-triazole ring has proven to be a promising scaffold for
the development of efficient inverse agonists of GHSR in order to treat obesity-related metabolic
diseases.
EXPERIMENTAL SECTION
General Chemistry Methods. All chemicals for synthesis were purchased from Sigma-Aldrich
Chemical Co., Thermo Fisher GmbH, FlukaTM and Iris Biotech GmbH, and used without further
purification. All solvents were purchased from Sigma Aldrich and Honeywell in gradient grade or
reagent quality. Classical purifications were performed with flash column chromatography using
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
silica gel (VWR, 40-63 μm). HPLC purifications were run on a Waters 4000 preparative apparatus
using a C18 Deltapak column (100 mm x 40 mm, 15μm, 100 Å), with UV detection at 214 nm and
254 nm, at a flow rate of 50 mL/min of a mixture of A, water with 0.1% TFA, and B, acetonitrile
with 0.1% TFA, in gradient mode. Samples for LC-MS analyses were prepared in methanol
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(gradient grade). The LC-MS system consisted of a Waters Alliance 2695 HPLC coupled to a
Water Micromass ZQ spectrometer (electrospray ionization mode, ESI+). All analyses were
carried out using a Phenomenex Onyx reversed-phase column (25 x 4.6 mm). A flow rate of 3
mL/min and a gradient from 0 to 100% of B over 2.5 min were used. Eluent A: water/0.1% FA ;
eluent B: acetonitrile/0.1% FA. UV detection was performed at 214 nm. The purity of the
compounds was assessed by reversed-phase HPCL analyses (UV detection at 214 nm) performed
on a Beckman apparatus (System Gold) using a Chromatolith High Resolution column (RP-18e,
5
0-4.6 mm) with a gradient mode of 100% of A (water with 0.1% TFA) to 100% of B (acetonitrile
1
13
with 0.1% TFA) at a flow rate of 3 mL/min. All purities were found ≥95%. H-NMR and C-
NMR spectra were recorded on Bruker Advance spectrometers at 300, 400 or 500 MHz and 75,
1
00 or 125 MHz, respectively. Chemical shifts (δ) are reported in parts per million (ppm) relative
to the residual peak of the solvent (2.50 and 39.52 ppm for DMSO, 3.31 and 49 ppm for MeOD)
according to the literature.26 Data are reported as follows: chemical shift (δ in ppm), multiplicity
(s=singlet, d=doublet, t=triplet, q=quartet, br s=broad singlet, m=multiplet), coupling constants (J
=
Hz), integration and assignment. HRMS analyses were performed using a microTOF-Q
spectrometer (electrospray ionization mode, ESI+). The lyophilizer from Cryonext (Saint Gely du
Fesc, France) was used with a freezing trap at -90 °C to lyophilize compounds purified by reverse
phase preparative chromatography.
26
ACS Paragon Plus Environment

Page 27 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
General Procedure: Compound 1 (compound 4 in reference 18), triazoles (iv, vi and vii) and
not commercially available hydrazides (v) were synthesized following the same procedures
previously described.18, 27
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General Procedure: Method A (for ligands 2-8). Triazole iv (100 mg, 0.2 mmol, 1eq) was
coupled with carboxylic acids (1.1 equiv), in the presence of BOP (1.1 equiv) and DIEA (pH=9)
for 2 h, in DCM. The mixture was then concentrated in vacuo, and the residue was dissolved in
EtOAc. The organic layer was successively washed with aqueous solutions of 1M KHSO4,
saturated NaHCO , and brine. The organic layer was dried over Na SO , filtered, and condensed
3
2
4
under reduced pressure to yield the desired compound. The compounds that did not require a final
deprotection were purified by preparative HPLC on a C18 column using a water/acetonitrile/TFA
0
.1% gradient.
General Procedure: Method B (for ligands 9-35). Compound 1 (100 mg, 0.18 mmol, 1eq) was
coupled with carboxylic acids (1.1 equiv), in the presence of HATU (1.1 equiv) and DIEA (pH=9)
for 2 h, in DCM. The mixture was condensed under reduced pressure, and the residue was
dissolved in EtOAc. The organic layer was successively washed with aqueous solutions of 1M
KHSO , saturated NaHCO , and brine. The organic layer was dried over Na SO , filtered, and
4
3
2
4
condensed under reduced pressure to yield the desired compound. The compounds that did not
require a final deprotection were purified by preparative HPLC on a C18 column using a
water/acetonitrile/TFA 0.1% gradient.
General Procedure: Method C (for ligand 36). Appropriate triazole vi (100 mg, 1eq) was
coupled with alpha-cyano-4-methoxycinnamic acid (1.1 equiv), in the presence of HATU (1.1
equiv) and DIEA (pH=9) for 2 h, in DCM. The mixture was condensed under reduced pressure,
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 70
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and the residue was dissolved in EtOAc. The organic layer was successively washed with aqueous
solutions of 1M KHSO , saturated NaHCO , and brine. The organic layer was dried over Na SO ,
4
3
2
4
filtered, and condensed under reduced pressure to yield the desired compound that was purified by
preparative HPLC on a C18 column using a water/acetonitrile/TFA 0.1% gradient.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General Procedure: Method D (for ligand 37). Appropriate triazole vi (100 mg, 1eq) was
coupled with 2-hydroxyacetic acid (1.1 equiv), in the presence of HATU (1.1 equiv) and DIEA
(
pH=9) for 2 h, in DCM. The mixture was condensed under reduced pressure and the residue was
dissolved in EtOAc. The organic layer was successively washed with aqueous solutions of 1M
KHSO , saturated NaHCO , and brine. The organic layer was dried over Na SO , filtered, and
4
3
2
4
condensed under reduced pressure. The resulting compound was purified by flash chromatography
on silica gel with a gradient of EtOAc to EtOAc/MeOH 9:1 as eluent and then reacted with alpha-
cyano-4-methoxycinnamic acid (1.1 equiv), in the presence of EDC (1.1 equiv) and DIEA (pH=9)
for 2 h, in DCM. After removal of solvents under reduced pressure, the residue was dissolved in
EtOAc and the organic layer was successively washed with aqueous solutions of 1M KHSO4,
saturated NaHCO , and brine. The organic layer was dried over Na SO , filtered, and condensed
3
2
4
under reduced pressure to yield the desired compound that was purified by preparative HPLC on
a C18 column using a water/acetonitrile/TFA 0.1% gradient.
General Procedure: Method E (for ligands 38-46). Appropriate Boc-protected triazole vii
(100 mg, 1eq) was treated with 4N HCl in dioxane (5 mL) for 30 min at 50 °C. When removal of
the Boc-protecting group was confirmed by HPLC analysis, MeOH was added and the reaction
mixture was condensed under reduced pressure. Additions of MeOH followed by evaporations
were repeated several times in order to remove excess of HCl. The resulting deprotected amine
was coupled with appropriate para-substituted alpha-cyano-cinnamic acid (1.1 equiv), in the
28
ACS Paragon Plus Environment

Page 29 of 70
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
presence of HATU (1.1 equiv) and DIEA (pH=9) for 2 h, in DCM. The mixture was condensed
under reduced pressure, and the residue was dissolved in EtOAc. The organic layer was
successively washed with aqueous solutions of 1M KHSO , saturated NaHCO , and brine. The
4
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
organic layer was dried over Na SO , filtered, and condensed under reduced pressure to yield the
2
4
desired compound that was purified by preparative HPLC on a C18 column using a
water/acetonitrile/TFA 0.1% gradient.
General Procedure: Method F (for ligands 2, 4, 6, 11, 21-23). Removal of the protecting group
was performed at 50 °C with a solution of 4N HCl in dioxane (2 mL) for 30 min. The mixture was
then condensed under reduced pressure, diluted with MeOH, and condensed again under reduced
pressure. Additions of MeOH followed by evaporations were repeated several times in order to
remove excess of HCl. The resulting deprotected compounds were then purified by preparative
HPLC on a C18 column using a water/acetonitrile/TFA 0.1% gradient.
General Procedure: Method G (for ligands 6 and 8). Removal of the Fmoc protecting group
was performed at room temperature with Et NH (10 eq, 1 mmol/ml) in DMF for 2 h. The mixture
2
was then condensed under reduced pressure and the compounds were purified by preparative
HPLC on a C18 column using a water/acetonitrile/TFA 0.1% gradient.
(R)-N-[2-(1H-indol-3-yl)-1-(4-(4-methoxybenzyl)-5-phenethyl-4H-1,2,4-triazol-3-yl)ethyl]-2-
(methylamino)acetamide (2). By methods A and F: 89 mg (0.17 mmol, 85%), orange wax (purity
+
1
HPLC = 99%). LC-MS (ES): t = 1.30 min, m/z 523.3 [M+H] . H NMR (300 MHz, DMSO-d ):
R
6
δ 10.90 (d, J = 1.8 Hz, 1H, NH), 9.35 (d, J = 8.1 Hz, 1H, NH), 8.85-8.71 (m, 2H, NH TFA salt),
2
7
.35 (d, J = 8.1 Hz, 1H, CH aryl), 7.32-7.11 (m, 6H, CH aryl), 7.09-7.02 (m, 2H, CH aryl), 6.91 (t, J
=
7.3 Hz, 1H, CH aryl), 6.71 (br s, 4H, CH aryl), 5.34-5.24 (m, 1H, CH), 5.05 (d, J = 16.9 Hz, 1H,
29
ACS Paragon Plus Environment