A Systematic Exploration of Macrocyclization in Apelin-13: Impact on Binding, Signaling, Stability, and Cardiovascular Effects

Article abstract of DOI:10.1021/acs.jmedchem.7b01353

The apelin receptor generates increasing interest as a potential target across several cardiovascular indications. However, the short half-life of its cognate ligands, the apelin peptides, is a limiting factor for pharmacological use. In this study, we systematically explored each position of apelin-13 to find the best position to cyclize the peptide, with the goal to improve its stability while optimizing its binding affinity and signaling profile. Macrocyclic analogues showed a remarkably higher stability in rat plasma (half-life >3 h versus 24 min for Pyr-apelin-13), accompanied by improved affinity (analogue 15, K_i 0.15 nM and t_1/2 6.8 h). Several compounds displayed higher inotropic effects ex vivo in the Langendorff isolated heart model in rats (analogues 13 and 15, maximum response at 0.003 nM versus 0.03 nM of apelin-13). In conclusion, this study provides stable and active compounds to better characterize the pharmacology of the apelinergic system.

Full text of DOI:10.1021/acs.jmedchem.7b01353

Subscriber access provided by UNIV OF DURHAM
Article
A systematic exploration of macrocyclization in apelin-13 –
impact on binding, signaling, stability and cardiovascular effects
Kien TRAN, Alexandre Murza, Xavier Sainsily, David Coquerel, Jérôme Côté,
Karine Belleville, Lounès Haroune, Jean-Michel Longpré, Robert Dumaine, Dany
Salvail, Olivier Lesur, Mannix Auger-Messier, Philippe Sarret, and Éric Marsault
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01353 • Publication Date (Web): 20 Feb 2018
Downloaded from http://pubs.acs.org on February 22, 2018
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.
Journal of Medicinal Chemistry 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 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A systematic exploration of macrocyclization in
apelinꢀ13 – impact on binding, signaling, stability
and cardiovascular effects
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
§
,¥
§,¥
§,¥
#,¥
§,¥
Kien Trân, Alexandre Murza, Xavier Sainsily, David Coquerel, Jérôme Côté, Karine
§
,¥
§,¥
§,¥
§,¥
Belleville,
Lounès Haroune,
JeanꢀMichel Longpré, Robert Dumaine
Dany Salvail,†
#
,¥
#
§,¥,$
*,§,¥,$
Olivier Lesur, Mannix AugerꢀMessier, Philippe Sarret
and Éric Marsault
§
Département de PharmacologieꢀPhysiologie, Faculté de Médecine et des Sciences de la Santé,
Université de Sherbrooke, Sherbrooke, J1H 5N4, Québec, Canada
#
Département de Médecine, Faculté de Médecine et des Sciences de la Santé, Université de
Sherbrooke, Sherbrooke, J1H 5N4, Québec, Canada
† IPS Thérapeutique Inc., Sherbrooke, J1G 5J6, Québec, Canada
¥
Institut de Pharmacologie de Sherbrooke, Sherbrooke, J1H 5N4, Québec, Canada
Author Contributions
$
E.M. and P.S. contributed equally to directing this study.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 52
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
KEYWORDS: APJ receptor, Apelin, GPCR, macrocycle, stability, inotrope, cardiovascular
disease, ringꢀclosing metathesis.
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
ACS Paragon Plus Environment

Page 3 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT
The apelin receptor generates increasing interest as a potential target across several
cardiovascular indications. However, the short halfꢀlife of its cognate ligands, the apelin
peptides, is a limiting factor for pharmacological use. In this study, we systematically explored
each position of apelinꢀ13 to find the best position to cyclize the peptide, with the goal to
improve its stability while optimizing its binding affinity and signaling profile. Macrocyclic
analogues showed a remarkably higher stability in rat plasma (halfꢀlife > 3 h versus 24 min for
Pyrꢀapelinꢀ13), accompanied with improved affinity (analogue 15, K 0.15 nM and t 6.8 h).
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
i
1/2
Several compounds displayed higher inotropic effects ex vivo in the Langendorff isolated heart
model in rats (analogues 13 and 15, maximum response at 0.003 nM versus 0.03 nM of apelinꢀ
1
3). In conclusion, this study provides stable and active compounds to better characterize the
pharmacology of the apelinergic system.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 52
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
INTRODUCTION
Apelin is the endogenous ligand of the apelin receptor (also known as APJ or angiotensin
1
receptorꢀlike 1), which belongs to the G proteinꢀcoupled receptor (GPCR) superfamily. In
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
humans, the APLN gene on chromosome X encodes for the 77ꢀaminoꢀacid precursor,
2,3
preproapelin which is cleaved to generate several isoforms: apelinꢀ36, apelinꢀ17, apelinꢀ13 and
2
,4,5
apelinꢀ13(1ꢀ12) (collectively called hereafter the apelins).
Among them, the pyroglutamate
form of apelinꢀ13 (Pyrꢀapelinꢀ13, referred below as apelinꢀ13) is the most abundant in human
6
plasma. Structurally, apelinꢀ13 is composed of two distant pharmacophoric regions: the Nꢀ
2
3
4
5
7,8
terminal RPRL (Arg ꢀPro ꢀArg ꢀLeu ), essential for binding affinity, and the Cꢀterminal PMPF
1
0
11
12
13
7,9
(Pro ꢀMet ꢀPro ꢀPhe ), important for both binding affinity and signaling. Binding of the
apelins to the APJ receptor triggers several intracellular signaling pathways, mediated by the
Gα , Gα and possibly the Gα proteins, subsequently leading to the recruitment of several
i/o
13
q
1
0–12
signal transduction cascades, such as the activation of PLC/PKCε,
AMPꢀactivated protein
1
3–15
kinase (AMPK)/eNOS,
as well as the regulation of ERK1/2 phosphorylation and
16,17
PI3K/Akt/p70S6 kinase signaling.
Another APJ ligand, ELABELA (ELA, also known as
Toddler, or Apela), was reported in 2013 to exhibit the same affinity for the receptor and to exert
1
8,19
similar signaling profiles as the apelins.
Importantly, both ligands are critical regulators of
9,20–22
cardiovascular development and function.
Existing knowledge collectively makes the APJ receptor a compelling target for the treatment
2
3
of cardiovascular diseases. Physiologically, apelins stimulate cardiac contractility while
23,24
decreasing peripheral resistance and promoting angiogenesis.
A decrease in APJ receptor
density in the left ventricle has also been observed in dilated cardiomyopathy and cardiac
2
5
ischemia patients. Although reports on the levels of apelins and ELA in diverse pathologies
ACS Paragon Plus Environment

Page 5 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
26
may be contradictory, apelinꢀ13 treatment has consistently shown a positive impact on cardiac
contractility and lowering of the mean arterial blood pressure (MABP) in heart failure
2
7,28
patients.
Investigator trials on pulmonary arterial hypertension (PAH) patients confirmed this
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
therapeutic potential, wherein apelin infusions improved cardiac output and lowered pulmonary
29
vascular resistance. Those beneficial effects of apelinꢀ13 and ELA were demonstrated in many
30–32
animal models.
Likewise, our group recently reported their benefits on cardiac and renal
3
3,34
functions as well as survival in a systemic inflammation model of sepsis and septic shock.
While the therapeutic potential of the apelinergic system is promising, the short halfꢀlife of
apelinꢀ13 and other apelins remains a limiting factor for their pharmacological use. Indeed, the
halfꢀlife (t ) of Pyrꢀapelinꢀ13 is around 24 minutes in rat plasma ex vivo and a few minutes in
1
/2
vivo, with major cleavage sites mediated by AngiotensinꢀConverting Enzyme 2 (ACEꢀ2) on the
1
2
13
4
5
5
penultimate position (residues Pro ꢀPhe ) and neprilysin between residues Arg ꢀLeu or Leu ꢀ
6
35–37
Ser .
Few approaches have been used to improve the stability of Pyrꢀapelinꢀ13. Conjugation
of fatty acid derivatives or PEG to the Nꢀterminus of apelin increased halfꢀlife above 24 h in rat
38,39
12
13
plasma.
Replacement of residues adjacent to the cleavage sites of ACE 2 (Pro , Phe ) or
4
5
6
36,38,40
neprilysin (Arg , Leu or Ser ) also contributes to protect against proteolytic degradation.
Macrocyclization is a recognized method to simultaneously improve the proteolytic stability of
peptides and modulate their biological activity, and has been successfully applied against several
41,42
targets including proteases, proteinꢀprotein interaction and GPCRs.
Few macrocyclization
6
12
attempts were reported for the apelins, including cyclization between Ser and Pro , which
43
1
6
improved plasma stability, cyclization between Gln and Ser , which provided Gα ꢀbiased
i
4
4
cyclic analogues MM07 (21, Figure 1), as well as bicyclic antagonist c[CRPRLC]ꢀKHꢀ
45
c[CRPRLC]. More recently, we reported a series of Gα ꢀbiased macrocycles bearing a
i
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 52
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
7
11 46
Tyr(OBn) residue in position 13 (ring closure between residues His and Met ), and Ma et al.
provided the first Xꢀray structure of human APJ receptor in complex with a macrocyclic
6
9 47
analogue of apelinꢀ17 (cyclized between residues Ser and Gly ).
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
To date, despite the above, there has been no systematic probing of the impact of apelinꢀ13
cyclization on its biological properties such as binding, stability, and signaling profile. In the
present study, cyclization was implemented on every position of Pyrꢀapelinꢀ13, using an
48
approach reminiscent of the “rolling loop scan” reported by Reichwein et al., and the impact of
such structural modifications on binding affinity, plasma stability, and signaling profiles of the
resulting analogues was assessed.
ACS Paragon Plus Environment

Page 7 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULTS AND DISCUSSION
Design of macrocycles
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
To build the macrocycles, two amino acids located at the i, i+4 positions of Pyrꢀapelinꢀ13 were
replaced by allylglycine (Scheme 1), then side chains were connected using ringꢀclosing
metathesis (RCM) to generate a series of 17ꢀmembered rings. This ring size incorporates three
endocyclic amino acids corresponding to the i+1, i+2 and i+3 residues. Similar to previous
11
49
2
4
studies, residue Met was replaced by Nle to avoid sulfur oxidation. Since Arg and Arg are
7
known to be crucial for binding, their side chains were conserved in the N ꢀlinked macrocycles.
α
In these cases, cyclization was performed with allyl groups introduced on the alpha amine (Nα)
5
0,51
via FukuyamaꢀMitsunobu allylation,
generating in fine a tertiary amide (analogues 3, 6) (see
Supplementary information Scheme 1S).
Synthesis
Pyrꢀapelinꢀ13 analogues were synthesized manually using the Fmoc strategy on Wang resin
ꢀ1 52,53
(
0.3 mmol.g ).
The first amino acid was loaded on resin via a Mitsunobu reaction using
triphenylphosphine (PPh ) and diisopropyl azodicarboxylate (DIAD), then unreacted benzylic
3
alcohols on the resin were capped with acetic anhydride. The Fmoc group was then removed
using piperidine in N,Nꢀdimethylformamide (DMF) (20/80) and the next amino acid, activated
by [hexafluorophosphate of Oꢀ(7ꢀazabenzotriazolꢀ1ꢀyl)ꢀ1,1,3,3ꢀtetramethyluronium] (HATU)
and diisopropylethylamine (DIPEA) in DMF, was coupled. The full linear sequence was built by
repetition of the above steps. Subsequently, RCM was carried out using the HoveydaꢀGrubbs
54
second generation catalyst (HGꢀII) using the modified protocol of Patgiri et al (Scheme 1).
Briefly, dry resin was swelled with dichloroethane and dry Ar was bubbled through the mixture
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 52
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
for 30 minutes. The HGꢀII RCM catalyst was then added and the reaction mixture heated to 120
o
C for 10 min under microwave irradiation. Finally, the peptide was cleaved from the resin using
a cocktail of trifluoroacetic acid (TFA)/triꢀisopropylsilane (TIPS)/H O (95:2.5:2.5). This step
2
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
was followed by peptide precipitation in tertꢀbutyl methyl ether (TBME) and subsequent
purification using reverseꢀphase preparative HPLC coupled to mass spectrometry.
The FukuyamaꢀMitsunobu secondary amine synthesis was used to generate Nꢀlinked
5
1
macrocycles. Accordingly, the amine group was first nosylated on resin using oꢀnosyl chloride
Supplementary Schemes 1S, 2S). The allyl group was then introduced using allylic alcohol,
(
PPh and DIAD. To achieve a higher yield, RCM was preferentially performed immediately after
3
this step rather than at the end of peptide elongation. Peptides were cleaved and purified as
above.
Macrocyclization by RCM was generally associated with lowꢀtoꢀmedium conversion (15ꢀ
30%). In particular, precursor 5L could not be converted to compound 5 even after 2h
microwave irradiation. For this compound, an alternative approach was implemented via
1
2
cyclization prior to the addition of the Pyr ꢀArg residues. All purified products were
characterized by UPLCꢀMS and highꢀresolution mass spectrometry (HRMS).
Impact of macrocyclization on binding affinity
In addition to macrocycles, biological activity of the corresponding linear precursors (named
with suffix L for linear analogue, Table 1) was assessed to evaluate the respective impact of
residue mutation vs cyclization on ligandꢀreceptor binding and signaling (Figure 1).
ACS Paragon Plus Environment

Page 9 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Several macrocycles possess equivalent or lower affinity for the APJ receptor compared to
their linear analogues (Table 1). Interestingly, these differences vary as a function of the
macrocyclization position. Based on this analysis, the Nꢀ and Cꢀtermini emerge as the most
suitable regions for cyclization [i, i+4; 17ꢀmembered cycle], with compounds 2, 3, 5, 6, 11, 13
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
displaying K values between 1 and 10 nM. It is interesting to note that the APJ receptor is
i
tolerant of modifications to the structure of its ligand in such distant positions, which suggests
that these positions are potential sites to further modulate the structure of the ligand.
Macrocyclic analogues modified at the Nꢀterminus displayed only a marginal decrease in
affinity compared to their linear counterparts (e.g. 2, K 5.7 nM vs 2L, K 2.5 nM; 3, K 4.5 nM
i
i
i
2
4
vs 3L, K 2.9 nM; 5, K 3.0 nM vs 5L, K 1.9 nM). The presence of Arg and Arg turned out to
i
i
i
7,8,53
be important to preserve affinity, consistent with current knowledge.
As a further testimony
2
of the role of Arg residues, analogue 3 possessing an N ꢀallylated Arg residue, with almost the
α
same cyclization position as 4, displayed a 100ꢀfold higher affinity compared to 4 in which the
2
Arg position is replaced by allylglycine (3, K 4.5 nM vs 4, K 509 nM). Likewise, macrocycle 6
i
i
4
4
bearing an N ꢀallylated Arg residue is 600ꢀfold more potent than 7 in which Arg is replaced by
α
allylglycine (6, K 9.1 nM vs 7, K 6074 nM). Compared with the wellꢀstudied macrocyclic
i
i
1
6
44
compound 21, cyclized between Gln and Ser residues (K 48 nM), compounds 2 and 3 in this
i
series bearing similar cyclization showed 5ꢀ to 10ꢀfold improved binding affinities. One possible
6
explanation is that 21 possesses a larger ring incorporating more residues (additional Ser residue
1
compared to 2, and additional Pyr residue compared to 3), which may impact orientation of the
RPRL pharmacophore.
On the other hand, macrocyclization in the central portion of Pyrꢀapelinꢀ13 was detrimental on
binding affinity. Analogues 7 and 8 possessed one order of magnitude lower binding affinity
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 52
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
compared to their respective linear analogues (7, K 6074 nM vs 7L, K 489 nM; and 8, K 41 nM
i
i
i
vs 8L, K 3.9 nM) (Table 1). This is consistent with previous studies whereby Nꢀmethylation of
i
amide bonds induced a significant loss of potency in activity reporter assay using stably
3
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
5
6
7
8
9
0
expressing recombinant human APJꢀGα cell line. Taken together, these results suggest that
16
the central part of Pyrꢀapelinꢀ13 must adopt a wellꢀdefined conformation distinct from that
imposed by the current cyclization approach. A promising macrocyclic template for this position
is AMGꢀ3054 (22) (reported by Ma et al.), which features a larger ring and was nonetheless
6
9
8,47
cyclized between Ser and Gly , albeit on an apelinꢀ17 scaffold.
Among this set of compounds, Cꢀterminallyꢀmodified macrocycles 11ꢀ13 exhibited high
affinity for the APJ receptor, with analogue 13 possessing the highest affinity (K 1.1 nM),
i
similar to its linear congener (13L, K 2.1 nM). Knowing that the Cꢀterminal interacts deeper in
i
47
the transmembrane domain, this data suggests a large binding pocket able to accommodate the
1
0
17ꢀmembered ring. The substitution of the Pro residue by an allylglycine was also detrimental
for ligand binding affinity, irrespective of cyclization site (9, K 188 nM vs 9L, K 41 nM; 10, K
i
i
i
2
68 nM vs 10L, K 194 nM; 14, K 310 nM vs 14L, K 445 nM).
i
i
i
Impact of macrocyclization on signaling
The signaling profile of the most promising analogues (K < 30 nM) was assessed using
i
Bioluminescence Resonance Energy Transfer (BRET) assays in HEK cells expressing the human
APJ receptor, with a particular emphasis on activation of Gα (canonical pathway of the APJ
i1
17
receptor leading to intracellular cAMP reduction), and recruitment of βꢀarrestin2 (associated
ACS Paragon Plus Environment

Page 11 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
55
with receptor internalization). Macrocycles and their linear analogues were divided into two
groups based on their ability to modulate APJ receptor signaling (Figure 2).
The first group consists of most analogues for which cyclization showed a deleterious effect on
receptor signaling (2, 6, 11, 12, 13). When linear compounds displayed similar potencies as Pyrꢀ
apelinꢀ13, cyclization induced a 10ꢀ to 40ꢀfold loss in Gα activation (2, EC Gα 23 nM vs 2L,
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
i1
50
i1
0
.4 nM) and 5ꢀ to 25ꢀfold decrease in βꢀarrestin2 recruitment (2, EC βꢀarr2 834 nM vs 2L, 35
50
nM).
The second group includes analogs where cyclization showed beneficial effects on signaling
properties (compounds 3, 5). While the linear compound 5L induced a significant decrease in
downstream signaling activation compared to Pyrꢀapelinꢀ13, macrocyclization slightly restored
potency on both pathways (5, EC Gα 3.7 nM vs 5L, 9.2 nM; the same for βꢀarrestin2).
5
0
i1
2
3
4
5
Structurally, the Arg ꢀPro ꢀArg ꢀLeu moiety has been altered in analogues 3L and 5L. N ꢀ
α
2
3
allylation was introduced on residue Arg in 3L and residue Pro was replaced by allylglycine in
5
L. Since the Nꢀterminus of apelin was shown by NMR and molecular modelling to be ordered,
5
6,57
with a putative βꢀturn structure,
modifications such as amino acid substitution would be
expected to disrupt this secondary structure and impact intracellular signaling. As a matter of
fact, hydrogen bonding with the backbone is often critical in maintaining peptide secondary
5
8
structure. Despite this, compound 3L, which cannot form a critical Hꢀbond by virtue of N ꢀ
α
allylation, showed a modest decrease in binding affinity and signal transduction. This is in
1
2
agreement with previous studies showing that amide bond methylation between Pyr ꢀArg and
3
4
2+
39
Pro ꢀArg did not affect Ca mobilization. Accordingly, this structural moiety seems to be
3
3
more governed by Pro rather than intramolecular Hꢀbonds. Indeed, replacement of Pro by
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 52
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
allylglycine led to decrease in Gα activation and βꢀarrestin2 recruitment (compound 5L).
i1
Interestingly, macrocyclization on this position either had a neutral impact (compound 3 vs 3L),
or marginally improved potency compared to its linear analogue (compound 5 vs 5L).
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
Impact of macrocyclization on plasma stability
The influence of macrocyclization on plasma stability was then assessed via incubation of
compounds in rat plasma followed by UPLCꢀMS analysis at selected time points (0, 2, 4, 7, 24 h
or 0, 10, 20, 60, 120 min). Halfꢀlives and observed cleavage sites of the most active analogues
are reported in Figure 3 (for other compounds, see Figure 1SꢀSupp. Info.). Previous work on
linear analogues of Pyrꢀapelinꢀ13 showed that substitutions next to the critical proteolytic site
increased peptide stability, however halfꢀlives of those linear analogues were usually less than 2
36
hours. In this study, macrocycles displayed significant improvements in plasma stability, and
the protection offered by cyclization critically depends on their position.
Cyclization in the central and Cꢀterminus portions of the peptide were the most beneficial, with
/2
t1 > 3 h compared to 0.4 h for Pyrꢀapelinꢀ13. Particularly, compounds 6 and 13 possessed t1/2
7
.7 h and 8.6 h, respectively. Compound 13 (t1/2 8.6 h) was also more stable than its linear
counterpart 13L (t1/2 2.4 h). At the same cyclization position, N ꢀlinked macrocyclic analogues
α
displayed even longer halfꢀlives (3, t1/2 2.2 h vs 4, t1/2 0.7 h), (9, t1/2 > 24 h vs 10, t1/2 4.1 h)
(Figure 1SꢀSupp. Info).
Regarding cleavage sites, no endocyclic cleavage was observed. Moreover, most analogues
4
13
were cleaved between Arg and Phe , confirming that macrocyclization in this region protects
the peptide from proteolysis, which substantially improves its stability (Figure 1SꢀSupp. Info.).
The previously reported cleavage sites identified on apelinꢀ13 were also observed in this study,
ACS Paragon Plus Environment

Page 13 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
12
13
except when endocyclic. ACE2ꢀmediated cleavage between Pro and Phe was observed on 3ꢀ
6
7
36
7
, 9, 10; cleavage between Ser and His was observed on 11 and 12, and cleavage between
4
5
5
6
either Arg ꢀLeu (on 8ꢀ10) or Leu ꢀSer (on 9ꢀ11, 14), recently reported to be mediated by
3
6,37
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
neprilysin,
was also observed.
7
8
In the Cꢀterminus, fragments corresponding to cleavage between His ꢀLys (on 4, 5, 12) and
1
0
11
Pro ꢀNle (on 2 ꢀ 5) were also observed, consistently with previously reported cleavage sites of
3
6
apelinꢀ13 analogues. Some new cleavage sites were observed in the current study, such as
8
9
between Lys ꢀGly on macrocycles 3, 7, 13.
In the above analysis, one should keep in mind that the absence of an observable fragment
does not necessarily mean that this fragment was not produced; instead, it may undergo
subsequent cleavage to a smaller fragment. The aforementioned cleavage sites may also exist on
Pyrꢀapelinꢀ13, however were not reported to date. For example, in our previous studies,
replacement of selected residues of Pyrꢀapelinꢀ13 with unnatural amino acids to prevent the
5
6
7
8 36
cleavage between Leu ꢀSer , revealed the presence of another cleavage site between His ꢀLys .
This cleavage site was again found in this series of macrocycles (4, 5, 12).
Exploring the binding pocket with larger macrocycles
To further investigate the properties of the binding pocket, several analogues of 13
9
13
[
macrocycle linked X ꢀX ; 17ꢀmembered ring] were built, in which size and linker type were
modified (Figure 4). Results reveal that enlarging macrocycle size up to 20 atoms with no
9
13
additional endocyclic amino acid [B ꢀX ; 20ꢀmembered ring] improved binding affinity
compared to Pyrꢀapelinꢀ13 (K 0.7 nM). This modification led to the most potent macrocyclic
i
analogues in this study, analogues 15 (K 0.15 nM) and 17 (K 0.10 nM) (Table 2). These also
i
i
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 52
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
exhibit similar properties in terms of Gα signaling (15, EC 1.3 nM; 17, EC 1.7 nM) and βꢀ
i1
50
50
arrestin2 recruitment (15, EC 57 nM; 17, EC 63 nM) as apelinꢀ13 (EC Gα 1.1 nM and
5
0
50
50
i1
EC βꢀarr2 60 nM). Furthermore, compound 15 displays significant improvements in plasma
50
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
stability compared to apelinꢀ13 with t1/2 6.8 h (Figure 3).
The only difference between 15 and 17 is the chiral center on amino acid 13. 15 features an Lꢀ
1
3
13
allylglycine linker in position 13, whereas 17 carries a
Dꢀallylglycine . This modification does
not induce a major difference in affinity between the two compounds, contrasting with the 20ꢀ
13
13
fold decrease observed when substituting LꢀPhe by DꢀPhe on Pyrꢀapelinꢀ13 from a previous
4
9
study. One difference between these two sets of results is the presence of the 2ꢀ
nitrobenzenesulfonamide (oꢀnosyl) on the side chain of position 9, which may provide an
additional energetically favorable interaction with the receptor. Based on the Xꢀray structure of
4
7
the APJ receptor, the space between side chains of residues 9 and 13 is an aromatic cage
2
4
35
85
composed of residues from the Nꢀterminus (Trp ), transmembrane domains I (Tyr ), II (Trp ,
9
3
88
291
Tyr , Tyr ) and VII (Phe ), which creates potential opportunities for πꢀπ interaction with
electronꢀdeficient moieties such as the oꢀnosyl group. Accordingly, removal of the nosyl group
on the same macrocycles results in a 6000ꢀfold decrease (16, K 917 nM) and 2000ꢀfold decrease
i
(
18, K 206 nM) in binding affinity compared to nosylated analogues (15 and 17), respectively.
i
The newly created positive charge of the amine on the linker in 16 and 18 may be contributive to
4
7,49
such loss in this essentially hydrophobic pocket.
Finally, truncation of the Nꢀterminal portion
of 15 and 17 completely abolished their affinity for the APJ receptor (19 and 20, K > 10 000
i
nM), which emphasizes the importance of this portion for ligand binding. Thus, ring enlargement
by itself was beneficial only in the presence of the Nꢀnosyl group.
ACS Paragon Plus Environment

Page 15 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Regarding degradation profile, there are differences in cleavage sites between 13 and 15
4
despite identical macrocyclization position (Figure 3). The cleavage site of neprilysin (Arg ꢀ
5
5
6
7
8
Leu , Leu ꢀSer ) and those between His ꢀLys were identified on 15, whereas the only cleavage
8
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
site identified on 13 was between Lys and the macrocycle core. This suggests that ring size or
linker type may influence the interaction between peptides and proteases. However, the halfꢀlife
of the two peptides is not significantly different and it should be kept in mind that several
fragments may not be identified due to subsequent cleavages.
Effect of macrocyclic analogues on blood pressure and cardiac performance in rats
The Nꢀ and Cꢀterminallyꢀmodified macrocycles 5, 13 and 15, with the highest binding affinity
and potency on the Gα and βꢀarrestin2 pathways in this collection, were tested for their ability
i1
to modulate blood pressure in vivo and to modify cardiac contractility in the Langendorff isolated
heart perfusion assay ex vivo (Figure 5).
Blood pressure was monitored after i.v. bolus administration of two doses (19.5 nmol/kg and
6
5 nmol/kg) of compounds 5, 13 and 15 to anesthetized rats. Our results revealed that
compounds 5 and 15 present similar hypotensive effects to Pyrꢀapelinꢀ13, inducing rapid and
robust blood pressure lowering up to 40 mmHg. As opposed to 5 and 15, compound 13 appears
less potent, inducing blood pressure drop only at the highest dose tested. Given that previous
8
studies suggested a link between hypotension and receptor internalization, we next assessed APJ
internalization after stimulation with either apelinꢀ13, 13 or 15 using enzymeꢀlinked
immunosorbent assay (ELISA) (Figure 6). Results indicate that 13 (EC 469 nM) is 10ꢀfold less
5
0
potent than its analogues 15 or Pyrꢀapelinꢀ13 (EC 45 nM and 24 nM, respectively) to induce
50
receptor internalization, consistent with its lower ability to modulate blood pressure. It was also
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 52
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
noted that the hypertensive phase observed with Pyrꢀapelinꢀ13 (presumably due to baroreflex
activation), was not observed following i.v. administration of compound 15, even at the highest
1
3
dose. Interestingly, although replacement of Phe by alanine in previous studies yielded a
5
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
5
6
7
8
9
0
physiological antagonist devoid of hypotensive effect, analogues 13 and 15, devoid of an
aromatic residue in position 13, still display an agonist profile in vivo.
Despite longer halfꢀlife in vitro, the hypotensive action of 13 and 15 lasted no longer than 5
2
0
minutes in this test, similarly to other APJ receptor agonists. This may tentatively be explained
with several factors. Firstly, internalization and desensitization of the APJ receptor could
55
contribute to reduce the duration of action of agonists. Moreover, although they display higher
stability in plasma ex vivo, their elimination and metabolism in vivo may reduce circulating
concentrations and hence their effect. To assess this latter point, we performed a preliminary
pharmacokinetic analysis in which apelinꢀ13, 13 and 15 were administered to rats i.v. at 3 mg/Kg
(Supp Info Figure 2S). Results of this study show that both 13 and 15 display significantly
higher circulating concentrations and AUC (13, AUC 246 µg/mL.min; 15, 168 µg/mL.min)
compared to Pyrꢀapelinꢀ13 (AUC 0.2 µg/mL.min). While the hypotensive phase ends 2–3 min
after administration for the three compounds, the remaining quantity at 5 min (13, 8.4 µg/mL or
5
1
.6 µM; 15, 4.3 µg/mL or 2.5 µM) are sufficient to saturate the available APJ receptors (13, Ki
.1 nM; 15, K 0.15 nM). This finding suggests that the rapid normalization of blood pressure
i
may not relate to circulating concentrations but other factors, such as receptor desensitization or
6
0
the baroreflex which liberates vasoconstrictor agents such as catecholamines.
To evaluate the ability of new analogs to modulate cardiac function in the Langendorff model,
the heart of SpragueꢀDawley rats was isolated and perfused with different concentrations of
macrocyclic analogues, with doses ranging from 0.001 nM to 0.3 nM. In this experiment, the
ACS Paragon Plus Environment

Page 17 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
61
increase in left ventricular developed pressure is recorded as a measure of inotropic effect.
Results showed that compounds 13 and 15 behave as powerful cardiostimulants. Indeed, both
compounds demonstrated increased potency (with maximum response at a concentration of
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
0
.003 nM as opposed to 0.03 nM for Pyrꢀapelinꢀ13), as well as increased efficacy (compound 13
reaching 217 ± 33 % from baseline compared to 81 ± 11 % from baseline for Pyrꢀapelinꢀ13 at
dose of 0.003 nM, p < 0,001). On the other hand, macrocycle 5 possessed similar potency and
efficacy to Pyrꢀapelinꢀ13.
CONCLUSION
This study reports the first systematic study of the influence of macrocyclization on binding
affinity, signaling profiles, plasma stability, ex vivo and in vivo cardiovascular effects of
macrocyclic analogues of Pyrꢀapelinꢀ13. By moving the site of ring closure along the sequence,
several positions appeared suitable for macrocyclization on the Nꢀ and Cꢀtermini of apelinꢀ13.
1
7ꢀmembered Cꢀterminal macrocycle 13 possesses a comparable binding affinity for the APJ
receptor compared to its linear analogue. Enlarging the ring led to 15 and 17 possessing higher
affinity (15, K 0.15 nM; 17, K 0.10 nM or 5ꢀ to 7ꢀfold higher than Pyrꢀapelinꢀ13). The signaling
i
i
profile of highꢀaffinity analogues was established using BRETꢀbased assays. In term of Gαi1
activation and βꢀarrestin2 recruitment, macrocyclization was detrimental on most positions,
2
3
4
5
except for the nosylated analogues 15 and 17. Furthermore, in the Arg ꢀPro ꢀArg ꢀLeu moiety,
3
whereas linear analogues lost their potency due to Pro replacement (5L), macrocyclization was
beneficial by partially restoring potency for both pathways (5). Regarding plasma stability, the
4
13
sequence between Arg ꢀPhe seems optimal for cyclization, protecting the peptide from
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 52
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
degradation. Accordingly, 13 possesses a longer halfꢀlife in rat plasma (8.6 h) compared to Pyrꢀ
apelinꢀ13 (24 minutes) and its linear analogue 13L (2.4 h). This translated into higher circulating
concentrations in rats in vivo. Finally, the impact of compounds on blood pressure and on the
isolatedꢀperfused heart ex vivo suggest that compounds 13 and 15 are powerful modulators of the
cardiovascular system. Altogether, these new macrocycles represent useful pharmacological
tools to better rationalize the links between signaling profile and physiological action and more
broadly the pharmacoꢀphysiology of the apelinergic system.
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
ACS Paragon Plus Environment

Page 19 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EXPERIMENTAL SECTION
Chemistry
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
Materials. Wang resin (maximum loading 0.5 ꢀ 1.0 mmol/g, particle size 200ꢀ400 mesh),
diisopropylethylamine (DIPEA), trifluoroacetic acid (TFA) and amino acids were purchased
from ChemꢀImpex International (Wood Dale, USA). Polypropylene cartridges 12 mL with 20
µm PE fit were purchased from Applied Separation (Allentown, USA), HATU was ordered from
Matrix Innovation (Québec, Canada). Other reagents and solvent were purchased from Sigmaꢀ
Aldrich (Missouri, USA) and Fisher Scientific (Hampton, USA). The above reagents and
solvents were used as received.
Solid phase synthesis (compounds 1ꢀ20 and 2Lꢀ14L). All peptides were synthesized at 0.1
mmol scale on Wang resin (loading 0.3 mmol/g). For the loading step, resin was first swelled in
4
mL anhydrous tetrahydrofuran (THF) for 5 min. Fmocꢀprotected amino acid (0.3 mmol) and
triphenylphosphine (78 mg, 0.3 mmol) were then added and the mixture was shaken for another
min. A solution of DIAD (60 µL, 0.3 mmol) in 1 mL anhydrous THF was added slowly and
5
the reaction mixture agitated overnight on an orbital shaker at 120 rpm. The resin was then
filtered and washed with a sequence of DMFꢀDCMꢀiPrOHꢀDCMꢀiPrOHꢀDCM (5 mL solvent, 3
min cycles). Unreacted benzylic groups were capped using 5 mL of a solution of acetic
anhydrideꢀDIPEAꢀDCM (4:1:0.2). For amino acid coupling, the Fmoc group was first removed
by treating with 20 % piperidine/DMF (5 mL, 2 x 10 min) and the resin was washed again with
the above washing sequence. The Fmocꢀprotected amino acid (0.5 mmol) was then added as a
solution with HATU (190 mg, 0.5 mmol) and DIPEA (87 µL, 0.5 mmol) in 5 mL DMF, and the
mixture was shaken for 30 min. Upon completion of the reaction, the resin was washed again
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 52
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
with the above washing sequence. Deprotection and coupling steps were repeated to produce the
full linear sequence. Finally, the resin was washed with diethyl ether and dried in a vacuum
desiccator prior to the metathesis step.
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
Ring closing metathesis (compounds 2ꢀ20). Dry resin (0.1 mmol peptide) was placed in a 10
mL microwave tube equipped with a stirring bar, then 3 mL anhydrous 1,2ꢀdichloroethane was
added to swell the resin. This suspension was purged with argon for 30 min. Subsequently, 15
nd
mg (0.023 mmol) of HoveydaꢀGrubbs II generation catalyst was added. The reaction mixture
was agitated, purged with argon for 5 min and submitted to microwave irradiation in a Discover
SP microwave oven (CEM, Matthews, USA) with the following parameters: controlled
o
temperature 120 C, time 10 min, maximum power 300 W. The resin was filtered and washed
with the washing sequence DCMꢀMeOHꢀDCMꢀMeOHꢀDCM (5 mL, 3 min each). Final cleavage
from the resin and simultaneous sideꢀchain deprotections were performed with 4mL of a TFAꢀ
TIPSꢀH O (95:2.5:2.5) mixture for 4 hours, then the crude product was precipitated in 35 mL of
2
o
TBME at 0 C, centrifuged at 2000 rpm for 10 min and isolated as a light brown solid after
removal of the supernatant.
For compound 3, 5, 6, 9, 15ꢀ20, the procedure was slightly modified as described in
supplementary information (Supp. Info Schemes 1S, 2S).
Peptide purification. The crude product was reꢀsuspended in acetic acid 10 % and purified on a
TM
TM
preparative HPLCꢀMS system from Waters (Milford, USA) (column XSELECT CSH Prep
C18 (19 x 100 mm) packed with 5 µm particles, UV detector 2998, MS SQ Detector 2, Sample
manager 2767 and a binary gradient module) using acetonitrile and water + 0.1 % formic acid as
eluents. Pure fractions were lyophilized to give the final product as a white solid. For purity
ACS Paragon Plus Environment

Page 21 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
assessment, compounds were analyzed on a UPLCꢀMS system from Waters (Milford, USA)
(
column Acquity UPLC® CSH™ C18 (2.1 x 50 mm) packed with 1.7 µm particles) with the
following gradient: acetonitrile and water with 0.1% HCOOH (0→0.2 min: 5% acetonitrile;
.2→1.5 min: 5%→95%; 1.5→1.8 min: 95%; 1.8→2.0 min: 95%→5%; 2.0→2.5 min: 5%). All
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
0
compounds have ≥95% purity, except 7L (≥90 % purity) and 9L (≥87%). HRMS of all analogues
were obtained using electrospray infusion on a maXis ESIꢀQꢀTof apparatus from Bruker
(Billerica, USA).
In vitro and in cellulo assays
Materials. High glucose Dulbecco’s Modified Eagle Medium (DMEM), G418, penicillin/
streptomycin, were ordered from Invitrogen Life Technologies (Carlsbad, USA). Fetal bovine
serum (FBS) was obtained from Wisent (SaintꢀJeanꢀBaptiste, Canada) and bovine serum albumin
(BSA) from BioShop (Burlington, Canada). Polyethylenimine (PEI) was obtained from
Polysciences (Warminster, USA).
Cell culture. Human Embryonic Kidney cells (HEK293) stably transfected of YFPꢀtagged
human APJ receptor was cultured in DMEM medium with 10 % of FBS. Cells were incubated at
o
3
7 C, in a humid atmosphere maintaining 5 % CO . G418 (400 µg/mL) was used to maintain
2
selection pressure for receptor expressed cells and penicillin/streptomycin (0.1 %) were added to
prevent bacteria contamination.
Binding experiment. The binding experiments were carried out on cell membranes. For
membrane preparation, HEK293 cells stably expressing the YFPꢀtagged human APJ receptor
were submitted to a freeze ꢀthaw cycle to break down the cells. Lysed cells were gently
transferred to a falcon tube containing 4 mL of resuspension solution (1 mM EDTA and 50 mM
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 52
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
TrisꢀHCl, pH 7.4) and cell membranes were pulled down by centrifugation at 3500 rpm for 15
o
minutes at 4 C. The precipitate was reꢀsuspended in binding buffer (50 mM TrisꢀHCl, 0.2 %
BSA, pH 7.4). Competition binding assay was performed in 96ꢀwell plates where 15 µg of
1
25
75
77
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
membrane proteins were incubated with 0.2 nM of radiolabeled [ I] [Nle , Tyr ] Pyrꢀapelinꢀ
ꢀ5
ꢀ11
ꢀ4
1
1
3 (820 Ci/mmol) and test ligand with concentrations ranging from 10 to 10 nM (or 10 to
ꢀ10
0
nM for compounds with low affinity) in a total volume of 200 µL for 1 h at room
temperature. Unbound ligands were removed by filtration through glass fibber filter (Millipore,
o
preꢀabsorbed of PEI 0.5 % for 2 hours at 4 C), and then washed three times with 120 µL of
binding buffer. The gamma emission was then recorded using γꢀcounter 1470 Wizard from
PerkinElmer (Waltham, USA) (80 % efficiency). Nonꢀspecific binding, determined in the
ꢀ5
presence of 10 nM of unlabeled Pyrꢀapelinꢀ13, was not excess over 5 % of total signal. Results
were plotted on a concentrationꢀresponse curve using GraphPad Prism 7 to calculate the IC50
values, which represent the concentration of testing ligand displaced 50 % of radiolabeled ligand
from the receptor. Dissociation constant K was calculated from the IC value using the Chengꢀ
i
50
62
Prusoff equation and results were showed as mean ± SEM of three independent experiments.
BRET assays. HEK293 cells seeded in T175 flasks were allowed to grow in high glucose
DMEM supplemented with 10% FBS, 100 U/mL penicillin/streptomycin, 2 mM glutamine, and
2
0 mM HEPES at 37˚C in a humidified chamber at 5% CO . All transfections were carried out
2
with polyethylenimine. After 24 h, cells were transfected with the plasmids coding for hAPJ,
Gα ꢀRlucII(91), GFP10ꢀGγ , and Gβ (from cDNA.org) (for BRET based Gα activation assay)
i1
2
1
i1
53,63,64
or coding for hAPJꢀGFP10 and RlucIIꢀβꢀarrestin2 using polyethylenimine.
To perform
BRET assays, cells were transferred into white 96ꢀwell plates BD Bioscience (Mississauga,
Canada) at a concentration of 50 000 cells/well 24 h after transfection and incubated at 37˚C
ACS Paragon Plus Environment

Page 23 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
overnight. Cells were then washed with PBS and 90 µL of HBSS was added in each well. Then,
ꢀ5
ꢀ11
cells were stimulated with analogues at concentrations ranging from 10 M to 10 M for 5 min
at 37˚C (Gα ) or for 30 min at RT (βꢀarrestin2). After stimulation, 5 µM of coelanterazine 400A
i1
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
was added to each well and the plate was read using the BRET filter set of a GeniosPro plate
2
reader (Tecan, Austria). The BRET ratio was determined as GFP10 /RlucII . Data were
em
em
plotted and EC values were determined using GraphPad Prism 7. Each data point represents the
50
mean ± SEM of at least three different experiments each done in triplicate.
Rat plasma stability assay. Plasma was obtained from male rats by centrifugation of blood at
1
3000 rpm during 5 min at 4°C. Synthetic analogues (6 µL of 1 mM aqueous solution) were
o
incubated with plasma (27 µL) at 37 C for 2, 4, 7, 24 hours (or 10, 20, 60, 120 minutes for
compounds having shorter halfꢀlife). Enzymatic reactions were stopped by adding 70 µL of
3
9
acetonitrileꢀethanol (1:1) solution containing nicotinamide 0.5 µM (internal standard). This
o
mixture was then vortexed and centrifuged at 13000 rpm for 20 minutes at 4 C. The supernatant
was filtered through a 4 mm nylon 0.2 µm filter, diluted with 40 µL distilled water and analyzed
using Acquity UPLCꢀMS system class H (CSH C18 column packed with 1.7 µm particles).
Percentage of peptide remaining was plotted on an exponential decay curve using GraphPad
Prism 7 for calculating halfꢀlife. MS spectra at different time points were compared with those at
0
min (rat plasma inactivated by quenching solution before adding compound) to identify
cleavage fragments.
Assessment of internalization by EnzymeꢀLinked Immunosorbent Assay (ELISA)
HEK293 cells were seeded in 24ꢀwell plates preꢀcoated with 0.1 mg/mL polyꢀLꢀLysine (sigma)
at 200 000 cells/well. 48 h postꢀtransfection with the human HAꢀtagged APJ receptor, cells were
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 52
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
ꢀ5
ꢀ11
washed and stimulated with compounds at concentrations ranging from 10 M to 10 M for 30
min at 37˚C in HBSS. Cells were then fixed in 3.7% (v/v) formaldehyde/Trisꢀbuffered saline
(TBS) (20 mM TrisꢀHCl, pH 7.5 and 150 mM NaCl) for 5 min at room temperature. Cells were
then washed twice with TBS and incubated 30 min with TBS containing 1% BSA at room
temperature to block nonꢀspecific binding. A monoclonal antiꢀHAꢀperoxidase antibody (clone
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
3
F10, Roche Applied Sciences) was then added at a dilution of 1:1000 in TBSꢀBSA 1% for 60
min. Following the incubation with the primary antibody, cells were washed twice with TBS and
50 µL of 3,3′,5,5′ꢀtetramethylbenzidine (SigmaꢀAldrich, Canada) were added. The plates were
2
incubated at room temperature and the reaction was stopped using 250 µL of HCl 2N. 200 µL of
the colorimetric reaction was transferred to a 96ꢀwell plate and the absorbance was measured at
4
50 nm. Cells transfected with empty vector were used to determine background. Data were
plotted and EC values were determined by using GraphPad Prism 7. Each data point represents
5
0
the mean ± SEM of at least three different experiments. The optical density (DO) measured at 1
µM for 13 or 15 was normalized to the DO measured at 1 µM for Pyrꢀapelinꢀ13 providing the
percentage of internalization induced by these compounds versus Pyrꢀapelinꢀ13, set at 100% as
the reference.
In vivo and ex vivo assays
Adult male Sprague Dawley rats were purchased from Charles River Laboratories (Stꢀ
Constant, Quebec, Canada) and maintained on a 12 h light/12 h dark cycle with access to food
and water ad libitum. The experimental procedures in this study conformed with the Animal
Care Committee of Université de Sherbrooke and in accordance with policies and directives of
the Canadian Council on Animal Care.
ACS Paragon Plus Environment

Page 25 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Blood pressure test. Male SpragueꢀDawley rats (250ꢀ350g) were anesthetized with an
intramuscular injection of ketamine/xylazine solution (87/13 mg/kg) and placed in supine
position on a thermostatic pad. To measure systemic arterial blood pressure, a catheter (PE 50
filled with heparinized saline) was inserted into the right carotid artery and connected to a Microꢀ
Med transducer (model TDXꢀ300, Calabasas, USA) linked to a MicroꢀMed blood pressure
analyzer (model BPAꢀ100c). Another catheter (PE10) was inserted into the left jugular vein for
bolus injection of vehicle (isotonic saline) followed 5 minutes later by the injection of either Pyrꢀ
apelinꢀ13 or macrocyclic analogues (given at 19.6 or 65 nmol/kg; volume of 0.25 ml over 10 s).
The i.v. catheter was flushed with saline (0.3 ml) immediately after each injection to remove
residual injected substances. Only one analogue was given per experiment.
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
Ex vivo heart contractile function. Hearts from male SpragueꢀDawley rats were excised and
cannulated through the aorta connected to a Langendorff apparatus. Contractile function was
recorded through a salineꢀfilled latex balloon inserted into the left ventricle. Left ventricular
pressure was continuously monitored through a preꢀcalibrated physiological pressure transducer
(
Molecular Devices, Sunnyvale, USA). Thereafter, incremental concentrations of either APLNꢀ
3 or macrocyclic compounds 0.001ꢀtoꢀ0.3 nM, were infused. Data for Left Ventricular
Developed Pressure (LVDP) were collected and processed using the Clampfit 10.2 program
Molecular Devices, Sunnyvale, USA)
1
(
In vivo pharmacokinetic analysis. Male SpragueꢀDawley rats (250ꢀ350g) were anesthetized
with the same protocol described in the blood pressor test. Compounds (dissolved in 350 ꢀL in
normal saline) were administered via the jugular vein at a dose of 3 mg/kg and blood samples
were collected from the carotid artery into EDTAꢀcoated tubes (Sarstedt, Nümbrecht, Germany)
at 1, 5, 10, 15, 30, 60 and 120 min (1, 2, 5, 10, 15 min pour pyrꢀapelinꢀ13) after administration.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 52
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
o
To remove blood cells, samples were immediately centrifuged at 13000 rpm for 5 min at 4 C
and the supernatant (plasma) was transferred to a microtube. To the plasma (100 µL) was added
acetonitrile (200 µL) containing 1.5 µM internal standard. 13 was used as internal standard for
analyses of Pyrꢀapelinꢀ13 and 15 was used as internal standard for analyses of 13. This mixture
was vortexed, centrifuged at 13000 rpm for 20 minutes and filtered through 4 mm nylon 0.2 µm
filter to give a clear solution. LCꢀMS/MS analysis was performed using positive electrospray
ionization (ESI+) source in MultiꢀReactionꢀMonitoring mode on a QTRAP 6500+ mass
spectrometer from Sciex (Washington, USA), equipped with an eksgience HALO ES peptide
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
C18 column (50 mm × 0.5 mm, 2.7 µm). Solvent flow rate was set to 15 µL min and column
temperature was kept at 50 º C. Injection volume was 5 µL. Mobile phase was 0.10% formic
acid/water (A) and 0.10% formic acid/acetonitrile (B), with an elution gradient starting with 5%
of eluent B, increasing to 100% in 4 min and then back to initial conditions in 2 min for a total
run time of 8 min. Optimized parameters were obtained by direct infusion of Pyrꢀapelin 13, 13
ꢀ
1
and 15 analytical standard solutions at 100 ng mL as follows: CUR, 20, CAD, medium, IS 5100
V, TEM, 300º C, GS1, 15, GS2 13. Two daughter traces (transitions) were used. The most
abundant transition was used for quantification, the second most abundant for confirmation. Pyrꢀ
apelin 13, 13 and 15 in the supernatant were analyzed after sample purification. Results are
reported in the Supplementary Information Section.
ACS Paragon Plus Environment

Page 27 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ASSOCIATED CONTENT
Supporting Information
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
Synthetic scheme of N ꢀlinked macrocyclic analogues and Cꢀterminallyꢀmodified macrocyclic,
α
HRMS spectra and analytical UPLCꢀMS spectra, structure, molecular formula strings, halfꢀlife
and degradation profile in rat plasma of compounds described on this article, in vivo
pharmacokinetic analysis of apelinꢀ13, compounds 13 and 15, binding curves of apelinꢀ13,
compounds 13 and 15 on the rat apelin receptor. This content is available free of charge on ACS
Publication website.
AUTHOR INFORMATION
Corresponding Author
Phone: +1 819.821.8000 ext 72433
Fax: +1 819.564.5400
Email: eric.marsault@usherbrooke.ca
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
Financial support from Université de Sherbrooke, the Canadian Institutes of Health Research, the
Natural Sciences and Engineering Research Council of Canada, Canada Foundation for
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 52
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
Innovation and Merck Sharpe & Dohme (donation to the Faculty of Medicine and Health
Sciences of Université de Sherbrooke) is gratefully acknowledged. The Canadian Francophonie
Scholarship Program (PCBF), MITACS and FRQS are also acknowledged for scholarship grants
to K.T., A.M., X.S. and D.C., respectively. M.A.ꢀM. is the recipient of a Heart and Stroke
Foundation of Canada (HFSC) New Investigator award and a research scholar from the Fonds de
la Recherche du Québec en Santé (FRQS). P.S. is the recipient of the Canada Research Chair in
Neurophysiopharmacology of Chronic Pain. E.M. is a member of the FRQNTꢀfunded Proteo
Network and the FRQSꢀfunded Réseau Québécois de Recherche sur le Médicament. The authors
would also like to thank Prof. Michel Bouvier (Institut de Recherche en Immunologie et Cancer,
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
Montréal, Québec, Canada) for the use of human Gα and βꢀarrestin2 biosensors.
i1
ABBREVIATIONS
AUC: area under curve, BRET: Bioluminescence Resonance Energy Transfer, DCM:
dichloromethane, DIAD: diisopropylazodicarboxylate, DIPEA: N,Nꢀdiisopropylethylamine,
DMF: N,Nꢀdimethylformamide, GPCR: G ProteinꢀCoupled Receptor, HATU: Oꢀ(7ꢀ
azabenzotriazolꢀ1ꢀyl)ꢀ1,1,3,3ꢀtetramethyluronium
hexafluorophosphate,
HEK:
Human
Embryonic Kidney, UPLC: Ultra high Performance Liquid Chromatography, HRMS: High
Resolution Mass Spectrometry, iPrOH: isopropanol, LVDP: Left Ventricular Developed
Pressure, MABP: Mean Arterial Blood Pressure, SAR: StructureꢀActivity Relationship, TBME:
tertꢀbutyl methyl ether, TFA: Trifluoroacetic acid, TIPS: Triisopropylsilane.
ACS Paragon Plus Environment

Page 29 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
REFERENCES
(
1) O’Dowd, B. F.; Heiber, M.; Chan, A.; Heng, H. H. Q.; Tsui, L.ꢀC.; Kennedy, J. L.; Shi, X.;
Petronis, A.; George, S. R.; Nguyen, T. A Human Gene That Shows Identity with the Gene
Encoding the Angiotensin Receptor Is Located on Chromosome 11. Gene 1993, 136, 355–
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
3
60.
(
2) Tatemoto, K.; Hosoya, M.; Habata, Y.; Fujii, R.; Kakegawa, T.; Zou, M.ꢀX.; Kawamata,
Y.; Fukusumi, S.; Hinuma, S.; Kitada, C.; Kurokawa, T.; Onda, H.; Fujino, M. Isolation
and Characterization of a Novel Endogenous Peptide Ligand for the Human APJ Receptor.
Biochem. Biophys. Res. Commun. 1998, 251, 471–476.
(
(
(
3) Wang, G.; Qi, X.; Wei, W.; Englander, E. W.; Greeley, G. H. Characterization of the 5′ꢀ
Regulatory Regions of the Rat and Human Apelin Genes and Regulation of Breast Apelin
by USF. FASEB J. 2006, 20, 2639–2641.
4) Shin, K.; Pandey, A.; Liu, X.ꢀQ.; Anini, Y.; Rainey, J. K. Preferential Apelinꢀ13 Production
by the Proprotein Convertase PCSK3 Is Implicated in Obesity. FEBS Open Bio 2013, 3,
3
28–333.
5) Yang, P.; Kuc, R. E.; Brame, A. L.; Dyson, A.; Singer, M.; Glen, R. C.; Cheriyan, J.;
Wilkinson, I. B.; Davenport, A. P.; Maguire, J. J. [Pyr1]Apelinꢀ13(1–12) Is a Biologically
Active ACE2 Metabolite of the Endogenous Cardiovascular Peptide [Pyr1]Apelinꢀ13.
Front. Neurosci. 2017, 11, 1–14.
(6) Zhen, E. Y.; Higgs, R. E.; Gutierrez, J. A. Pyroglutamyl Apelinꢀ13 Identified as the Major
Apelin Isoform in Human Plasma. Anal. Biochem. 2013, 442, 1–9.
ACS Paragon Plus Environment