Design and Characterization of Superpotent Bivalent Ligands Targeting Oxytocin Receptor Dimers via a Channel-Like Structure

Article abstract of DOI:10.1021/acs.jmedchem.6b00564

Dimeric/oligomeric states of G-protein coupled receptors have been difficult to target. We report here bivalent ligands consisting of two identical oxytocin-mimetics that induce a three order magnitude boost in G-protein signaling of oxytocin receptors (OTRs) in vitro and a 100- and 40-fold gain in potency in vivo in the social behavior of mice and zebrafish. Through receptor mutagenesis and interference experiments with synthetic peptides mimicking transmembrane helices (TMH), we show that such superpotent behavior follows from the binding of the bivalent ligands to dimeric receptors based on a TMH1-TMH2 interface. Moreover, in this arrangement, only the analogues with a well-defined spacer length (～25 ?) precisely fit inside a channel-like passage between the two protomers of the dimer. The newly discovered oxytocin bivalent ligands represent a powerful tool for targeting dimeric OTR in neurodevelopmental and psychiatric disorders and, in general, provide a framework to untangle specific arrangements of G-protein coupled receptor dimers.

Full text of DOI:10.1021/acs.jmedchem.6b00564

Subscriber access provided by the University of Exeter
Article
Design and characterization of superpotent bivalent ligands
targeting oxytocin receptor dimers via a channel-like structure
Marta Busnelli, Gunnar Kleinau, Markus Muttenthaler, Stoytcho Borissov Stoev, Maurice
Manning, Lucka Bibic, Lesley A. Howell, Peter J. McCormick, Simona Di Lascio,
Daniela Braida, Mariaelvina Sala, G. Enrico Rovati, Tommaso Bellini, and Bice Chini
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00564 • Publication Date (Web): 15 Jul 2016
Downloaded from http://pubs.acs.org on July 16, 2016
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 free 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 accessible to all
readers and 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 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Design and characterization of superpotent bivalent ligands targeting
oxytocin receptor dimers via a channel-like structure
1
,2
3
4
5
5
Marta Busnelli , Gunnar Kleinau , Markus Muttenthaler , Stoytcho Stoev , Maurice Manning ,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6
6
6
2
2
Lucka Bibic , Lesley A. Howell , Peter J. McCormick , Simona Di Lascio , Daniela Braida ,
1,2
7
2
1*
Mariaelvina Sala , G.Enrico Rovati , Tommaso Bellini and Bice Chini
1
CNR, Institute of Neuroscience, Milan, Italy 20129
2
Department of Medical Biotechnology and Translational Medicine, Università degli Studi di
Milano, Milan, Italy 20129
3
Institute of Experimental Pediatric Endocrinology, CharitéꢀUniversitätsmedizin Berlin, Berlin,
Germany 13353
4
Institute for Molecular Bioscience, The University of Queensland, Brisbane, Australia 4072
5
Department of Biochemistry and Cancer Biology, University of Toledo, Toledo, OH USA
4
3614
6
School of Pharmacy, University of East Anglia, Norwich Research Park, Norwich, UK NR4
7TJ
7
Department of Pharmacological and Biomolecular Sciences, Università degli Studi di Milano,
Milan, Italy 20133
*Corresponding author: Dr. Bice Chini, CNR, Institute of Neuroscience, Via Vanvitelli 32,
2
0129 Milano, Italy. Phone +39ꢀ02ꢀ50316958; Fax +39ꢀ02ꢀ50317132; Eꢀmail: b.chini@in.cnr.it
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 55
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
ABSTRACT
Dimeric/oligomeric states of Gꢀprotein coupled receptors have been difficult to target. We
report here bivalent ligands consisting of two identical oxytocinꢀmimetics that induce a three order
magnitude boost in Gꢀprotein signaling of oxytocin receptors (OTRs) in vitro and a 100ꢀ and 40ꢀ
fold gain in potency in vivo in the social behavior of mice and zebrafish. Through receptor
mutagenesis and interference experiments with synthetic peptides mimicking transmembrane
helices (TMH), we show that such superpotent behavior follows from the binding of the bivalent
ligands to dimeric receptors based on a TMH1ꢀTMH2 interface. Moreover, in this arrangement,
only the analogs with a well defined spacer length (~25 Ǻ) precisely fit inside a channelꢀlike
passage between the two protomers of the dimer. The newly discovered oxytocin bivalent ligands
represent a powerful tool for targeting dimeric OTR in neurodevelopmental and psychiatric
disorders and, in general, provide a framework to untangle specific arrangements of Gꢀprotein
coupled receptor dimers.
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
2
ACS Paragon Plus Environment

Page 3 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Gꢀprotein coupled receptors (GPCRs), the largest eukaryotic family of transmembrane proteins
involved in cellular signal transduction, have been shown to associate to form dimers and/or higher
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
, 2
order oligomers. Despite their promise as new druggable targets, the molecular details of these
complexes have been elusive. Elucidating the number and arrangement of protomers in homoꢀ and
heteroꢀoligomers, the cellular distribution of these different conformations and, most importantly,
their functional relevance in normal and pathological conditions, still represents an extraordinary
3
challenge. Synthetic analogs specifically targeting GPCRs homomers and heteromers not only
constitute a promising new class of therapeutics, but represent powerful tools to contribute to the
enlightenment of how these complexes function.
One strategy to pharmacologically target GPCR dimers is to develop bivalent ligands, single
chemical entities composed of two covalently tethered pharmacophores. Such an approach seeks to
tether two pharmacophores with a spacer that enables the ligands to bind simultaneously into the
4
ꢀ6
two protomers of a dimeric receptor target. There are two general types of bivalent ligands:
heterobivalent, where the two pharmacophores are different, and homobivalent, where the two
pharmacophores are identical. So far, research has primarily focused on the design of heterobivalent
7
ligands to target orthosteric and allosteric binding sites within the same receptor or in heteromeric
receptor complexes. Heterobivalent targeting of heteromeric receptors have been extensively
8
pursued by Portoghese and coꢀworkers who reported the most successful heterobivalent ligands
9, 10
developed to date and which usually include at least one antagonist molecule.
Homobivalent
analogues have also been developed and investigated. However, in the majority of cases,
homobivalent agonist ligands designed to probe dimeric receptors were found to lose activity and
1
1ꢀ13
14, 15
became either partial agonists
or, in some cases, antagonists .
Scattered reports of
homobivalent ligands with increased in vitro and/or in vivo efficacy or potency revealed very short
16
7
linkers, more compatible with their binding to different sites within a single protomer than to the
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 55
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
two orthosteric sites of a dimeric receptor. A recent paper reported homobivalent ligands for the
melanocortin receptor that display increased binding affinity by 14ꢀ to 25ꢀfold, slight increased
cyclic adenosine 3’,5’ꢀmonophosphate (cAMP) signaling potency by 3ꢀ to 5ꢀfold compared to their
monovalent counterparts; in vivo, these ligands significantly decrease feeding but at a dose that was
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
17
similar to monovalent.
In this work we pharmacologically targeted dimeric conformations of the human oxytocin
receptor (OTR), a prototypic class A GPCR that has been shown to form homoꢀ and heterodimers in
18
vitro and in vivo. OTR homodimers and OTR heterodimers with the highly related vasopressin
1
9
V1a and V1b receptors have been demonstrated in HEK 293 cells and evidence for in vivo OTR
2
0
21, 22
heterodimerization with the dopamine D2 receptors and adrenergic β2 receptors
has also
been reported.
23
The oxytocin (OT) system is a key regulator of all the aspects of social behavior and, in
humans, OT facilitates the processing of social information, improves cognitive empathic abilities
2
4
and increases interpersonal trust. Based on these premises, OT administration has been proposed
25, 26
27
and used as a treatment of social deficits in autistic
and schizophrenic patients, representing
28
a domain of great expectations and relevance . However, the short life of the endogenous peptide
and its limited selectivity urges for the development of new potent oxytocinꢀ mimetics for
29
translational studies in vitro and in vivo.
To contribute towards addressing these issues, we
synthesized and characterized a series of homobivalent OT analogues designed to specifically
activate dimeric forms of the human OTR. We tested these compounds for their capability to bind
and activate OTR homodimers and predicted, by molecular modeling, their potential binding modes
at a dimeric OTR constellation. The suggested OTR dimer formation was validated by mutagenesis
studies and single transmembrane interference assays in combination with Gꢀprotein activation.
Finally, we used mouse and zebrafish animal models to study highly potent bivalent agonistic OTR
ligands in vivo.
4
ACS Paragon Plus Environment

Page 5 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULTS
Design strategy and synthesis of homobivalent dOTK ligands
OT (1) is a nonapeptide neurohormone characterized by a disulfide bond between the two
cysteine residues at positions 1 and 6 and a short Cꢀterminal tripeptide (Figure 1). Prior work,
demonstrated that deamination of cysteine at position 1 and substitution of the leucine residue at
position 8 with other amino acids did not affect the pharmacological properties of the peptide
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
30
towards the OTR. In particular, OT deaminated at position 1 and bearing a lysine at position 8,
and indicated here as dOTK (2) (Figure 1), could be conjugated to the bulky group DOTA (1,4,7,10
tetraazacyclododecaneꢀ1,4,7,10ꢀtetraacetic acid) without showing any significant drop in the
3
0
affinity for the OTR. Based on these observations, we used the lysine side chain at position 8 of
dOTK as the conjugation site to generate a series of differently sized bivalent dOTK ligands (4a-e),
joined by alkane spacers of different lengths, from C6 to C14 atoms (Figure 1). Alkane spacers
8
were chosen on the basis of their previous successful use in bivalent ligand design ; moreover,
alkane spacers are flexible and dynamic, their reaction conditions fairly well established and with
high yields, and are commercially available in a whole range of lengths. The chemical structure and
nomenclature of these compounds are schematically reported in Figure 1 and their synthesis in
Supporting Experimental Section.
Mono- and bi-phasic activation of OTR by monovalent and bivalent dOTK ligands
The capability of OTR to form homodimers was confirmed via a Bioluminescence Resonance
31
Energy Transfer (BRET) saturation assay (Figure 2A). In this experiment, HEK293 cells are coꢀ
transfected with a fixed DNA amount of OTR fused to the Renilla reniformis Luciferase (Rluc, the
BRET energy donor) and increasing DNA amounts of OTR fused to the blueꢀshifted variant of
2
Aequorea Victoria green fluorescent protein GFP (the BRET energy acceptor); this allows for the
2
formation of OTRꢀRluc/OTRꢀGFP dimeric complexes in a dose dependent manner, which, by
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 55
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
consequence, leads to a doseꢀdependent increase in the BRET signal, up to a point (a saturation
2
plateau) where all OTRꢀRluc receptors are complexes with OTRꢀGFP receptors. As a control to
rule out the possibility that the observed BRETꢀsignal is not a consequence of stochastic collision
2
events due to receptor over expression, the distantly related CD4 receptor fused to GFP (CD4ꢀ
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
2
2
GFP ) was coexpressed with OTRꢀRluc. As shown in Figure 2B, the OTRꢀRluc and CD4ꢀGFP
pair failed to produce a hyperbolic curve, whereas a robust and saturable BRET signal (BRETmax
2
0
.38 ± 0.04), was obtained for OTR, a result consistent with the formation of OTRꢀRluc/OTRꢀGFP
dimers.
To functionally evaluate monovalent and bivalent ligands, we employed a BRETꢀbased
biosensor that measures the ligand induced G protein activation (Figure 2C). In this biosensor, the
BRET energy donor Rluc is inserted within the Gα subunit amino acid sequence, and the acceptor, a
10
10
variant of the green fluorescent protein (GFP ) is Nꢀterminally fused to the Gγ subunit (GFP ꢀ
2
31, 32, 33
Gγ
2
) .
Ligandꢀinduced OTRꢀGq activation leads to GDP release coupled to a conformational
rearrangement of the heterotrimeric G protein complex, and, by consequence, to a decrease in the
BRET ratio. Using this assay, we previously demonstrated that OT (1) promotes OTR coupling to
several Gꢀprotein isoforms in a concentrationꢀdependent manner, allowing the identification of Gq
34
as being engaged at a lower peptide concentration. Several OT analogs were analyzed for
functional selectivity, biased properties; among these analogs, dOTK (2) was able to activate the Gq
3
4
pathway with a maximal effect similar to OT (1). Here, OT (1) and dOTK (2) full concentrationꢀ
response curves for OTR/Gq activation were generated (Figure 2D). Data were best fitted by a
fourꢀparameters logistic model for monotonic curves, with a resulting EC of 4.0 ± 1.6 nM (SE) for
50
dOTK (2), a value that is comparable with that obtained for OT (1) EC = 2.1 ± 1.3 nM (SE) (Table
5
0
1
). dOTK (2) and OT (1) thus present identical pharmacological properties towards OTR/Gq
activation, validating the choice of the dOTK (2) analog as a parent structure for the bivalent
synthesis. Interestingly, OT (1) and dOTK (2) curves revealed comparable Hill coefficients (ꢀ0.49
6
ACS Paragon Plus Environment

Page 7 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
for OT (1) and ꢀ0.51 for dOTK (2)) statistically different from ꢀ1 (P < 0.05) (Table 1), suggesting
the activation of a nonꢀhomogeneous population of OTRs.
The bivalent ligand series was then evaluated with the Gq BRET biosensor. As shown in
2
Figure 2E, the concentrationꢀresponse curves of dOTK ꢀC6 (4a), dOTK ꢀC12 (4d) and dOTK ꢀ
2
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
C14 (4e) were again monotonic and could be best fitted with a classical logistic model, with a Hill
coefficient different from ꢀ1 (P<0.05) (Table 1). On the contrary, the two bivalent ligands, dOTK2ꢀ
C8 (4b) and dOTK ꢀC10 (4c), promoted Gq activation with a biphasic behavior that could be best
2
fitted by a model for biphasic doseꢀresponse curves assuming a Hill coefficient equal to ꢀ1 for both
2 2
phases. The resulting EC50 for dOTK ꢀC8 (4b) and dOTK ꢀC10 (4c) (Table 1) indicated a first
activation process with an EC50Hi in the picoꢀmolar range (EC50Hi = 0.8 ± 0.7 pM (SE) for dOTK
2
ꢀ
C8 (4b) and EC50Hi = 2.8 ± 1.9 pM (SE) for dOTK ꢀC10 (4c)) and a second activation process with
2
an EC50Lo in the high nanoꢀmolar range (EC50Lo= 67.3 ± 17.9 nM (SE) for dOTK ꢀC8 (4b) and
2
2
EC50Lo= 13 ± 5.5 nM (SE) for dOTK ꢀC10 (4c)).
Statistical analysis of the maximal activation values (Emax) (Table 1), indicated that the Emax
values of the bivalent analogs were not statistically different from that of dOTK (2), suggesting that
dOTK (2) and the bivalent compounds (4a-e) activated the same pool of receptors.
Finally, we assayed the contribution of the spacer by synthesizing the monovalent dOTKꢀC8
(3) analog. As shown in Figure 2D, dOTKꢀC8 (3) generated a monophasic activation curve with a
calculated EC50 of 4.1 ± 4.6 nM (SE) and a Hill coefficient = ꢀ0.49 statistically different from ꢀ1
(P<0.05
), comparable to those of dOTK (3) (4.0 ± 1.6 nM (SE), Hill coefficient = ꢀ0.51 and OT (1)
(2.1 ± 1.3 nM (SE), Hill coefficient = ꢀ0.49). We can therefore exclude a significant contribution to
the biphasic activation process of the spacer alone.
The binding features of dOTK (2) and dOTK
2
ꢀC8 (4b) to the OTR were then investigated in
ꢀC8 (4b) data were best fitted with a biphasic
curve characterized by a high affinity (KiHi = 0.08 ± 0.05 pM (SE)) and low affinity component
competition binding experiments (Figure 3). dOTK
2
(
KiLo = 1.24 ± 0.6 nM (SE)), in agreement with BRET curves. dOTK (2) could be fitted by a twoꢀ
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 55
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
site binding model (P<0.01) with KiHi of 21 ± 6.93 pM (SE) and a KiLo of 3.35 ± 2.07 nM (SE)
(Figure 3). KiHi and KiLo values obtained in binding assays resulted leftꢀshifted by a factor 10 with
respect to EC50Hi and EC50Lo values obtained in activation experiments, a phenomenon that can
result from the different nature of the assays performed (binding is performed in competitive
conditions using as a radioligand a monovalent antagonist while activation is a non competitive
assay) and/or by the experimental conditions in which the two assays were performed (temperature,
buffers, incubation times).
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
Binding and activation data suggest the coexistence of different receptor subpopulations, a
35, 36
concept observed before in single molecule studies looking at other receptors.
Although
alternative explanations could in principle arise from cooperativity, no available model based on
such a mechanism is able to account at the same time for biphasic binding and activation.
37
Moreover, only negative cooperativity has been reported for the OTR. Since the most plausible
explanation of our observations is thus the presence of different receptor subpopulations, we
analyzed the pharmacological features of the activation curves by dissecting their possible
components (Figure 4).
dOTK (2) activation, although best fitted by a monotonic curve, is characterized by a Hill
coefficient = ꢀ0.51 in the BRET assay. As shown in Figure 4B (dashed black lines, whose sum is
the continuous black line), this same activation curve can be broken up as a sum of two independent
decays with Hill coefficient = ꢀ1, resulting in two EC50 (EC50Hi and EC50Lo) differing by up to three
orders of magnitude. Direct comparison of the EC50Hi and EC50Lo between dOTK (2) and dOTK
2
ꢀC8
(
4b) shows that while the EC50Lo values (dOTK (2): 74.8 nM; dOTK ꢀC8 (4b): 67.3 nM) are
2
statistically not different, the EC50Hi values (dOTK (2): 265 pM; dOTK ꢀC8 (4b): 0.8 pM) are
2
significantly different (Table 1). Consistently, our binding studies show low affinity sites with
identical Ki values for dOTK (2) (1.24 nM) and dOTK
2
ꢀC8 (4b) (3.35 nM) and a high affinity site
ꢀC8 (4b) (0.08
with three orders of magnitude different K values for dOTK (2) (21 pM) and dOTK
i
2
pM). These data indicate that the OTR exists in i) a high affinity state whose affinity is markedly
8
ACS Paragon Plus Environment

Page 9 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
boosted only by bivalent ligands and ii) a low affinity state that does not discriminate between
monovalent and bivalent analogs. Based on this evidence, we hypothesize that the first component
of the bivalent activation curve is sustained by high affinity sites, corresponding to dimeric
ꢀ13
ꢀ11
receptors, that are activated in a doseꢀdependent way at concentrations between 10 and 10 M;
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
11
ꢀ11
ꢀ9
the maximal activation of these sites is achieved at 10 M and, from 10 to 10 M, the curve
ꢀ9
stays flat, as expected for a saturating phenomenon. At 10 M, a population of low affinity sites,
corresponding to monomeric receptors, starts to be activated, leading to the second component of
ꢀ9
the biphasic curve; these sites are activated in a doseꢀdependent way at concentrations between 10
ꢀ7
and 10 M.
Because these data strongly suggest that the high affinity sites correspond to dimeric receptors
while the low affinity sites corresponds to monomeric receptors, we engaged in receptor modeling
and docking analysis to identify the monomeric and dimeric receptor conformations sustaining
monovalent and bivalent binding and activation.
Molecular modeling and docking of monovalent and bivalent dOTK ligands into monomeric
and dimeric OTR
The molecular models for monomeric OTR complexed with dOTK (2) (Figure S1) and
protomers of dimeric OTRs complexed with bivalent dOTK
2
ꢀC8 were developed (Figure 5) (for
details see Supporting Material).
Our docking studies suggest that the monomeric OTR binds OT (1) and dOTK (2) in a region
between the extracellular loops (ECLs) and extracellular ends of the transmembrane helices
30, 38
(TMHs).
The model (supporting material, structural model file_1) implicates the following
ligand/receptor aminoacid pairs involved in the binding of dOTK to the OTR (in the binding
pocket): Tyr2/Tyr200ꢀPhe291; Ile3/Met123ꢀLeu201ꢀIle201; Gln4/Ser298; Asn5/Gln119ꢀGln295;
Lys8/(Glu42)ꢀAsp100. This docking also supports an orientation of the ligand's Cꢀterminal tail
towards the ECL1/extracellular part of TMH1 and the cyclic dOTK ring in the transmembrane
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 55
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
region close to the ECL2 (see Figure S1 for details and Supporting Experimental Section for the
list of amino acids constituting the ligandꢀbinding pocket).
Stimulated by experimental evidence that suggests different protomerꢀprotomer interfaces in
39ꢀ43
different GPCR dimers,
we designed two models of OTRs protomerꢀprotomer interfaces, either
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
constituted by amino acid sideꢀchains between TMH1ꢀTMH2ꢀhelix8 (Figure 5A) or between
TMH5ꢀTMH6 (Figure 5B). Remarkably, only the OTR dimers with a protomer interface at TMH1ꢀ
TMH2ꢀhelix8 allow for the computational binding of both moieties of our homobivalent C8 (4b)
and C10 (4c) ligands into the two orthosteric binding sites (Figure 5A and 5B). Of note, the
docking of dOTKꢀC8 (4b) inside dimers with a TMH1ꢀTMH2ꢀhelix8 protomerꢀprotomer interface
is possible thanks to the presence of a channelꢀlike passage between the two interacting helices
(TMH1ꢀTMH2) (Figure 5C). This channelꢀlike passage allows the docking of the C8 (Figure 5A)
and C10 spacers maintaining the pharmacophore binding mode of the monomeric OTR/dOTK
complex. In this model, the channel opening towards the membrane is mainly constituted at the
protomer surface by Ala39, Val43, Leu97, Leu98 and is accompanied by further residues of ECL1,
such as Ile101. Indeed, the presence of a shortcutting channel is crucial to combine the limited
length of C8 and C10 spacers with the intrinsic size of the OTR and an optimal docking of dOTK
(
2) into the orthosteric binding sites. This fact can be immediately recognized by analyzing Figure
D, which shows that the conformation adopted by the spacers to connect the two dOTK
pharmacophores is almost linear. No alternative arrangements of dOTK ꢀC8 (4b) and dOTK ꢀC10
4c) into a dimerized OTR appear possible: any connection between the binding sites external to
5
2
2
(
TMHs would have required a much longer spacer. The OTR/dOTK arrangement in the TMH1ꢀ
TMH2ꢀhelix8 dimer is also supported by the fact that the C6 spacer is too short, leading to an
extraction of the ligandꢀmoieties from the orthosteric binding sites in the protomers. Likewise, by
constraining ligand moieties with the C12 and C14 spacers inside the orthosteric binding sites, the
conformation of the spacers within the channel would become kinked (Figure 5D), suggesting that
their excessive length could prevent appropriate binding at the dimer by steric hindrance. A model
10
ACS Paragon Plus Environment

Page 11 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
based on TMH5ꢀTMH6 (Figure 5B) will also require a much longer spacer, as the estimated
straight linear distance between the two orthosteric binding sites is, in this case, approximately 50
Ǻ; however, as the spacer would lay above the ECLs regions, its length will need to be further
increased, with a high probability that a very long alkane chain will negatively impact on docking
and binding for its higher hydrophobicity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Our modeling predicted that receptor mutations of the dimeric TMH1ꢀTMH2ꢀhelix8 interface
and/or synthetic peptides mimicking TMH1 would interfere with a channel constitution, altering the
critical relationship between the channel length and the length of the C8 and C10 spacers. If either
were altered then we predicted this would abolish the reduced entropic cost linked to the favorable
docking, resulting in the loss of superactivity. Experimentally, if unable to dock into the two
orthosteric sites of a receptor dimer, we expect the C8 and C10 bivalent ligands to lose the first,
high potency phase of the activation curve. To specifically test this hypothesis, we used a series of
point mutations in the TMH1ꢀTMH2ꢀhelix8 interface of OTR dimers and synthetic peptides
interfering with the TMH1 and TMH5 dimer interfaces.
Point mutations at the TMH1 dimer interface and a synthetic peptide mimicking TMH1
modifies the activation induced by bivalent dOTK ligands
Receptors bearing single or combined mutations of all residues exposed at the TMH1ꢀTMH2ꢀ
helix8 interface (C47, V43, L50, L51, L97, L98, V341, V345) were preliminary evaluated for their
expression at the cell surface by fluorescence microscopy assays (Figure S2) and FACS analysis
(Figure S3). We found that receptors bearing multiple mutations at the TMH1ꢀTMH2ꢀhelix8
interface were retained inside the intracellular compartments. Only, the OTR C47A and OTR
V43A/C47A receptor mutants were found to be correctly localized and expressed at the plasma
membrane where they formed dimers (Figure S4), and were further characterized.
The OTR C47A and OTR V43A/C47A (Figure 6A) were probed with the full C6ꢀC14 bivalent
series (4a-e) (Figure 6B, 6C). In the OTR C47A, the C6, C8 and C10 bivalent ligands presented
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 55
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
monophasic activation curves (Hill coefficient different from ꢀ1, P<0.05), all rightward shifted with
respect to that of dOTK (2) (Table 2). Intriguingly, the bivalent ligands with longer spacers, C12
(
4d) and C14 (4e), displayed a biphasic trend at this mutant, with a calculated EC50Hi < 10 pM and
EC50Lo > 100 nM (Figure 6B and Table 2). In the OTRꢀmutant V43A/C47A, only the bivalent
ligand with the longest spacer, dOTK ꢀC14 (4e), activated Gq with a biphasic curve with calculated
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
2
EC50Hi of 11 ± 14 pM (SE) and EC50Lo of 867 ± 250 nM (SE) (Figure 6C and Table 2). These data
suggest that by perturbing the TMH1ꢀTMH2ꢀhelix 8 interface and the spacer channelꢀopening
between the helices at residues C47 and V43, “relaxes" or "stretches” the dimers allowing enhanced
binding and functional activity of bivalent analogs with increased spacer lengths, but excludes the
biphasic effects of C8 (4b) and C10 (4c) observed in the OTR wild type (WT). We conclude that
these two point mutations most likely lead to a weakening in protomerꢀprotomer interactions with a
consequent rearrangement between the protomers which, in turn, causes a shift in the spacerꢀlength
required to enable the binding of two dOTK pharmacophores into both orthosteric sites.
To ensure that these results are indeed due to proteinꢀprotein interactions, we used synthetic
peptides mimicking single transmembrane helices, an approach previously shown to destabilize in
44ꢀ48
vitro and in vivo the quaternary structure and the interaction interfaces of heterodimeric GPCR
49
and rhodopsin . Cells were incubated with TMH1 and TMH5 peptides fused at their Cꢀterminal
end with a portion of the HIV transactivator of transcription (TAT) sequences to promote their
50
integration and correct orientation in the plasma membrane .
First, we tested the interference of TMH1 and TMH5 (Figure 7A) with OTR dimer
2
formation/stabilization in cells transfected with OTRꢀGFP and OTRꢀRluc (Figure 7B). As shown
in Figure 7B, incubation with TMH1ꢀTAT and TMH5ꢀTAT resulted in a reduction of the maximal
BRET of 36 % and 30 %, indicating that both peptides induced a destabilization of the OTR
dimers/oligomers. The residual BRET signal observed in the presence of both TMH1ꢀTAT and
TMH5ꢀTAT is consistent with the coꢀexistence of two dimeric subpopulations characterized by a
TMH1 or a TMH5 interface; in addition, it is likely that the peptides disrupt the quaternary structure
12
ACS Paragon Plus Environment

Page 13 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
between the receptors but do not necessarily block dimerization completely. Thus, a component of
this residual maximal BRET could arise from dimeric/oligomeric receptors that cannot be
modulated by exogenous peptides, such as those residing in intracellular compartments.
Importantly, there was no measurable decrease in BRET signal in the presence of a control
transmembrane peptide TMH1CꢀTAT, in which all residues exposed to the putative interacting
interface of TMH1 have been changed into alanine. These results confirm the specificity of the
functional interference exerted by the TMH1ꢀTAT and TMH5ꢀTAT peptides on OTR
dimers/oligomers.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Next, we tested the effects of the bivalent ligands in the presence of TMH1ꢀTAT or TMH5ꢀ
TAT. If the superagonist effects were indeed due to binding to two separate receptors in a dimer
then disruption of the quaternary structure with the disrupting peptides should lead to a monophasic
response curve. If however, the bivalent ligands are binding to two sites on the same receptor, or
binding a single receptor in a novel orthosteric site (Figure 4D panel V), then the peptides should
have no effect on the response curve. As shown in Figure 7C, the addition of TMH1ꢀTAT
completely disrupts the biphasic response of bivalent ligands which now present monophasic
(P<0.05) activation curves (Hill coefficient in the range 0.41 ꢀ 0.59; Table 3). Moreover, the EC50
values were all rightward shifted with respect to that of dOTK (2) (Table 3). On the contrary, no
effect of TMH5ꢀTAT was observed on the biphasic response (P<0.05) of the activation curves
(Figure 7D) and on activation parameters (Table 3) of bivalent ligands. Finally, neither TMH1 nor
TMH5 affected the activation properties of the monovalent dOTK (2) ligand.
These data solidify the model presented in Figure 4 and demonstrate that the high affinity state
of the OTR targeted by the C8 (4b) and C10 (4c) bivalent analogs are the dimers having a
TMH1/TMH2 interface.
2
Potency of dOTK -C8 in vivo on social behavior
51
OTR agonists are currently in clinical trial for autism related disorders. To test whether a
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 55
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
bivalent OTR ligand may be clinically relevant, we tested dOTK ꢀC8 (4b) in vivo in two animal
2
models in which the pharmacological effects of OT agonists on social behavior have been
previously extensively characterized.
The first model is represented by genetically modified mice in which one allele coding for the
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
+
/ꢀ
mouse OTR gene (Oxtr) has been deleted; as previously reported, Oxtr mice display a deficit in
5
2
social behavior. In particular, when tested for sociability in the three chamber test (Figure 8A),
these animals spent an equal amount of time exploring the compartment occupied by a stranger
mouse and that occupied by an empty cage, with a resulting difference score, calculated as the
difference between the time spent in the chamber with the stranger and the time spent in the
5
2
chamber with the empty cage, close to zero (Figure 8B and 8C). As previously reported, a single
dose of OT (1) of 0.5 ng/mouse administered i.c.v. immediately prior to the test, corrected this
abnormal behavior by significantly increasing the time spent exploring the stranger mouse over the
time spent exploring the empty cage, as indicated by a significant increase in the difference score
[
2
OT (1): F(3,36)=20.53, P < 0.001] (Figure 8B). To assess the effects of dOTK ꢀC8 (4b) on social
+
/ꢀ
behavior, we administered this compound i.c.v. to male Oxtr mice. As shown in Figure 8C,
dOTK ꢀC8 (4b) significantly restored sociability, i.e. significantly increased the time spent
2
exploring the unfamiliar mouse over the time spent exploring the empty cage, already at a dose of
0
.005 ng/mouse [dOTK
2
ꢀC8 (4b): F(3,36)=14.16, P < 0.001]. The dose of 0.05 ng/mouse resulted to
be statistically different from 0.005 ng/mouse (P<0.05) but not from 0.5 ng/mouse. These data
indicate that dOTK
2
ꢀC8 is active at a 100ꢀfold lower dose in comparison to the endogenous peptide
+/ꢀ
OT in restoring sociability in Oxtr mice.
The second model is represented by the shoaling behavior of zebrafish, a social behavior that
53
we have previously shown to be modulated, in this vertebrate, by several OTꢀrelated peptides
(Figure 8D). We demonstrated that the total time spent by WT zebrafish swimming parallel to
conspecifics is greater than that spent shoaling in proximity of Nacre, mutant zebrafish with a
defect in pigmentation. When this parameter was expressed as a difference in the time spent by the
14
ACS Paragon Plus Environment

Page 15 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
test animal with each phenotype (Nacre–WT), it resulted negative in vehicle treated animals. As
53
shown in Figure 8E and as previously reported, treatment with Isotocin (IT), the analog of OT in
fish, injected in the caudal musculature significantly increased the time spent in proximity of Nacre
at a concentration of 0.75 ng/kg, with a consequent significant decrease in the difference in the time
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
NacreꢀWT) that becomes positive, indicating that the fish prefer to shoal in proximity of Nacre
than of WT [IT: F(4,45)=34.00, P < 0.001].
As shown in Figure 8F, dOTK ꢀC8 (4b) significantly increased the shoaling preference to the
2
Nacre, i.e. significantly increased the time spent swimming parallel to the Nacre over the time spent
shoaling close to the conspecific already at a dose of 0.02 ng/kg [F(6,63)=17.78, P < 0.0001]. The
dose of 0.2 ng/kg resulted to be statistically different from the lower doses 0.002 and 0.02 ng/kg but
not from 0.5, 0.75 and 1 ng/kg. These data indicate dOTK ꢀC8 (4b) is active at approximately a 40ꢀ
2
fold lower dose as compared to the endogenous peptide IT on zebrafish sociability in vivo. The
slightly lower potency of dOTK
2
ꢀC8 (4b) in zebrafish compared to mice can be ascribed either to
the differences between the involved zebrafish receptor (native ligand IT) and the mouse receptor
(native ligand OT (1)) across species and/or to the administration route employed (dOTK
2
ꢀC8 (4b)
was administered i.c.v. in mice and i.m. in zebrafish).
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 55
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
DISCUSSION
Despite the promise of GPCR dimers as drug targets, clinically relevant molecules specific to
homoꢀoligomers have not yet been reported. In this study, we report the synthesis and
characterization of a series of homobivalent agonistic ligands in which two identical oxytocinꢀ
derived peptides were linked by aliphatic spacers of increasing length. Using this series of
compounds we are able to show that: i) the OTR exists in both a low affinity and high affinity state
that most likely reflect changes in oligomeric assembly; ii) the low affinity site is indifferent to the
spacer length of the bivalent ligands; iii) the high affinity population achieves superagonistic
responses upon binding of bivalent ligands having a specific spacer length (C8ꢀC10; ~25 Ǻ),
indicating the presence of a receptor entity with a specific structure; iv) mutagenesis and TMH
interference assays experimentally show that this structure is represented by homodimes interacting
via a TMH1ꢀTMH2 interface; v) dimeric OTR can be exploited in vivo as drug target of
homobivalent based therapeutic strategies.
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 development of bivalent ligands such as shown here requires matching the spacer
lengths of bivalent ligands with the distance between the receptor protomers. In the OTR, this effect
is dependent on the proposed “channelꢀlike passage”, which determines and constrains the length of
the bivalent spacer and prevents biphasic activation for bivalent analogs with shorter or longer
spacers. This channelꢀlike passage is a feature that emerges in the model only when a specific
dimeric arrangement based on a TMH1ꢀTMH2 interface is adopted; the relevance of the channel
and involvement of the TMH1 interface in an active dimeric conformation targeted by bivalent
ligands was predicted by our modeling data indicating that modification of single residues located
in TMH1 prevent bivalent docking into the OTR dimer. This model has been validated by
interference experiments with synthetic peptides mimicking transmembrane helices and
mutagenesis data of key residues located at the TMH1ꢀTMH2 interface in which we observed a loss
of the superagonist effects of dOTK ꢀC8 (4b) and dOTK ꢀC10 (4c) but a gain in the superagonist
2
2
16
ACS Paragon Plus Environment

Page 17 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
effects of dOTK ꢀC12 (4d) and dOTK ꢀC14 (4e). This behavior is coherent with a weakening in the
2
2
TMH1 protomerꢀprotomer interactions induced by these mutations, which may include mutual
azimuthal rotation. The channelꢀpassage described here could be of general interest: the search and
characterization of channelꢀlike opening(s) might be an important step in the design of superactive
bivalent ligands for other members of the GPCR superfamily, with broad applications for drug
design and development.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A key question concerns the mechanism responsible for the boost in EC50Hi observed for the
superactive bivalent ligands docking into the TMH1ꢀTMH2 dimer. A reduction in the entropy
cost for bivalent ligands (and an increase in binding affinity) is expected when their two
pharmacophores, connected by a spacer of length L, are both able to bind to the docking sites of a
5
4, 55
dimeric receptor.
within a volume (V) that could be roughly approximated as a hemisphere of radius L. Here, L ≈
5 Ǻ, is the distance between the Cꢀalpha atoms of the two lysine residues present in dOTK ꢀC8
4b), and thus includes the C8 spacer plus the lateral chains of the two lysine residues (Figure 1
Indeed, the docking of the first pharmacophore constrains the second one
2
2
(
and Figure 4D). Inside V, the effective local concentration (CLOC) of the unbound pharmacophore
is roughly CLOC ≈ 50 mM. Since, at such high concentrations, the second pharmacophore becomes
fully bound to the receptor, the condition of single pharmacophore binding of a bivalent ligand to
a dimeric receptor is unstable. This effect can be formally expressed in terms of a reduced entropy
cost for the binding of bivalent ligands, leading to a markedly enhanced binding affinity. Its
analysis also enables the formulation, under a few simplifying assumption, of a simple expression
2
for the expected activation efficiency of bivalent ligand: EC50Hi ∼ (EC50Lo) / CLOC (Supporting
note 2). This simple estimate is verified by our data, supporting the notion that the huge increase
in potency intrinsically brought about by the double docking is the dominant factor in our
observations, over shadowing other effects of less quantitative relevance.
Emerging data indicate that, in GPCRs, different interfaces can be found in different oligomers
3, 35
and even in different conformations (activity states) of the same oligomer.
For example,
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 55
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
molecular simulation of β1ꢀ and β2ꢀadrenergic receptor homodimers stability in a lipid bilayer is
consistent with a model of oligomerization in which more stable TMH1 homodimers diffuse
through the membrane, transiently interacting with other protomers at interfaces involving other
56
TMH helices (e.g. TMH4) . The data obtained in this work suggest that the OTR also exists in at
least two dimeric conformations, one involving a TMH1/TMH2 interface and the other a TMH5
interface, and that only the TMH1/TMH2 dimers can be activated by the bivalent ligands with
increased affinity and potency. Importantly, the analysis of the amplitude of the first phase of the
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
dOTK
fraction of the TMH1/TMH2 dimeric OTR specifically targeted by the bivalent compounds. The
amplitude of the first phase of the dOTK ꢀC8 (4b) and dOTK ꢀC10 (4c) biphasic curves, expressed
2 2
ꢀC8 (4b) and dOTK ꢀC10 (4c) biphasic activation curves allowed for the computation of the
2
2
as the percentage of the maximal amplitude of the curve (Emax values reported in Table 2), resulted
of 32 ± 0.5 % (SE) and 44 ± 0.5 % (SE), suggesting that, in our experimental conditions, the
dimeric OTR population targeted by the superactive bivalent ligand ranges between 32ꢀ44% of all
activable OTR present in the cells, a percentage in line with previous single molecule attempts to
5
7, 58
measure oligomerization.
To quantitatively disclose dimeric entities in living cells is currently a
very critical issue in the GPCR field as the techniques employed to determine the composition of
monomeric/oligomeric GPCRs, including resonance energy transfer (RET), fluorescence lifetime
imaging (FLIM), total internal reflected fluorescence (TIRF), fluorescence recovery after
photobleaching (FRAP) and fluorescence correlation spectroscopy (FCS), have led to a variety of
different results, with no apparent consensus on what equilibrium amongst these forms should be
3
expected for each GPCR.
Finally, to demonstrate the translational value of the superactive OT analogs we used two wellꢀ
established in vivo models (social behavior in zebrafish and mice) showing that the homobivalent
dOTK
2
ꢀC8 (4b) ligand displays a significant gain (40ꢀ and 100ꢀfolds) over the endogenous ligands,
ꢀC8 (4b) in vivo is consistent with the presence of
IT and OT (1). The increased potency of dOTK
2
OTR dimers or higherꢀorder oligomers in native tissues, thus extending the finding of OTR
18
ACS Paragon Plus Environment

Page 19 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
9
oligomers previously reported in the rat mammary gland to the central nervous system.
Remarkably, the only slight decrease in potency observed in vivo in mice and zebrafish as compared
to in vitro studies on human receptors, represent a proof of principle of the high value of these
analogs for translational studies in neurodevelopmental and psychiatric disorders. As dOTK (2)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6
0
2
binds with good affinity also to mouse V1a and V1b vasopressin receptors, dOTK ꢀC8 (4b) could
in principle also be hitting V1a and/or V1b homodimers as well as OTR/V1a, OTR/V1b, or
V1a/V1b heterodimers. However, because we have previously shown that the selectivity of dOTK
(
2) towards the OTR increases by a factor 100 when the amino group of its lysine in position 8 is no
30
2
longer available, we believe that the selectivity profile of the bivalent dOTK ꢀC8 (4b) analog
towards the OTR could be improved as well. In vivo experiments in vasopressin receptor null
models and/or with selective antagonists should be performed to fully address this point.
In conclusion, this new bivalent ligand class shows great potential in paving the way for the
development of advanced molecular tools and therapeutics targeting the OTR in important diseases
such as autism and schizophrenia. The findings and methodology reported and developed here
should also be highly applicable to other GPCR signaling systems.
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 55
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
EXPERIMENTAL SECTION
Peptide synthesis
Peptidic monovalent ligands were synthesized on a CS Bio CS336X peptide synthesizer using
61
Fmoc solidꢀphase peptide synthesis (SPPS). Following cleavage from resin, peptides were folded
by air oxidation in 0.1 mol ammonium bicarbonate (pH 8.2) for 24 h. Peptides were analyzed and
purified by analytical RPꢀHPLC (Phenomenex Jupiter C18, 4u, 150 mm x 4.6 mm) using a linear
gradient of 0ꢀ60% B (solvent A, H O/0.05% TFA; solvent B, 90% CH CN/10% H O/0.043% TFA)
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
2
3
2
in 60 min at 1 mL/min while monitoring UV absorbance at 214 nm. MS analysis was carried out on
an API 2000 from AB Sciex (Framingham, MA, USA). The purity of all compounds was >95%.
6
2
The bivalent ligands were synthesized via an adapted method of Gera et al.
that utilized a
disuccinimidyl (DSS) spacer to conjugate two monomers to form the bivalent ligand. Briefly, two
equivalents of dOTK and one equivalent of the DSS spacer were dissolved in anhydrous N,Nꢀ
dimethylformamide (DMF), followed by addition of three equivalents of N,Nꢀ
diisopropylethylamine (DIEA) in anhydrous DMF to the solution. The reaction mixture was left for
6 h and monitored by analytical RPꢀHPLC and LCMS. For the compounds, where the DSS ester of
the dicarboxylic acid was not commercially available (C6 and C14), the DSS esters were created in
situ using Nꢀhydroxysuccinimide (NHS) in the presence of N,N'ꢀdiisopropylcarbodiimide (DIC)
8
and DIEA, as described in . Additional details on the synthesis can be found in Supporting
Experimental Section.
OTR transmembrane (TMH) peptides were designed and synthesized together with a portion of
the highly polar HIV transactivator of transcription, HIV TAT (GRKKRRQRRR), added to the Cꢀ
terminus of TMH peptides to aid in plasma membrane insertion and correct orientation and to
5
0
improve the peptide’s solubility. Their sequences are shown in Figure 7A and more details are
reported in Supporting Experimental Section.
Cell culture and transfection
20
ACS Paragon Plus Environment

Page 21 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Human Embryonic Kidney 293 (HEK293) cells were purchased from the American Type Culture
Collection (Manassas, VA) and were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM)
(Sigma Aldrich, St. Louis, MO, USA) supplemented with 5% (vol/vol) FBS (Sigma Aldrich), 100
Units of penicillin (Sigma Aldrich), 100 g/mL streptomycin (Sigma Aldrich) and 2 mM Lꢀ
glutamine (Sigma Aldrich). The day before transfection, cells were seeded in 100ꢀmm plates or 35ꢀ
mm plates. Transient transfections were performed using linear polyethyleimine 25 KDa (PEI MW
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
5,000, Polysciences, Inc., Warrington, PA, USA) as transfecting agent, at a ratio of 3 to 1 PEI /
plasmid DNA. 24 h after transfection, the supplemented DMEM was renewed, and the cells were
cultured for a further 24 h before the experiments.
BRET assays
In Gq BRET activation assay DNA encoding for Renilla Luciferase (Rluc8)ꢀtagged Gα (the
q
1
0
BRET energy donor) (4 µg), GFP ꢀGγ
2
1
(BRET energy acceptor) (5 µg), Gβ (5 µg) G proteins
subunits and OTR constructs (7 µg) were used. In OTR dimerization experiments fixed amount of
2
OTRꢀRluc (3µg) and saturating amount of OTRꢀGFP (9µg) were used. HEK293 cells were
33, 34
transfected as described in
. 48 h following transfection, cells were washed twice with PBS,
detached and resuspended in PBS plus 0.1% (w/v) glucose at room temperature. Protein
concentration, was determined using a colorimetric protein assay (DC protein assay, Biorad, Milan,
Italy). For TMH peptide treatment transfected cells were incubated with 1µM TMH peptides for 2h
at room temperature before BRET assays. Cells were then distributed (80 µg of proteins per well) in
a 96ꢀwell microplate (Optiplate, Perkin Elmer, Milan, Italy) and incubated in the presence of
2
various ligands’ concentration or vehicle (PBS) for 2 min prior to the addition of the BRET
substrate, coelenterazine 400A, also called Deep Blue C (Biotium, Hayward, CA). Immediately
1
0
2
after the addition of coelenterazine 400A (5 µM), Rluc and GFP /GFP emissions were recorded
using a multidetector plate reader Infinite F500 (Tecan, Milan, Italy) that allows the sequential
integration of light signals detected with two filter settings (Rluc filter, 370ꢀ450 nm; and
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 55
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
1
0
2
2
GFP /GFP filter, 510ꢀ540 nm). Data were collected and the BRET signal determined as the ratio
1
0
2
of the light emitted by acceptors (GFP /GFP ) over donors (Rluc8). The “ligand promoted BRET”
value was calculated by subtracting the BRET signal measured with ligand from the one obtained in
34
the vehicle condition as in . In BRET saturation experiments, constant amounts of wild type or
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
mutant C47A or V43A/C47A OTRꢀRluc plasmid (0.5 µg) were coꢀtransfected in HEK293 cells
2
with increasing amounts of the corresponding OTRꢀGFP plasmids (0ꢀ1.5 µg in 35 mm/well). CD4ꢀ
2
GFP was used as a negative control of dimerization. The expression level of each tagged receptor
was determined by direct measurement of total fluorescence and luminescence in an aliquot of the
transfected cells. Total fluorescence was measured using a TECAN Infinite F500 microplate reader
2
10
with an excitation filter at 340 ± 25 nm and an emission filter at 535 ± 25 nm for GFP /GFP . After
fluorescence measurement, the same cell sample was incubated for 8 min with 5 µM coelenterazine
h (Biotium, Hayward, CA, USA) and the total luminescence was measured (from 370 to 450 nm)
using a TECAN Infinite F500 microplate reader. BRET signals were calculated immediately after
2
the adjunction of coelenterazine 400A (5 µM, Biotium) using the Rluc filter, 370ꢀ450 nm and GFP
2
filter, 510ꢀ540 nm, and were calculated as the ratio of the emission of GFP over the emission of
Rluc. BRET saturation data were analyzed using the Graph Pad Prism 5.0 software by nonlinear
regression assuming a oneꢀphase exponential association.
Ligand binding assays
Binding assays were performed at 30 °C on membranes prepared from COS7 cells as
38
125
2
4
9
125
previously described. The [ I]d(CH ) [Tyr(Me) , Thr , TyrꢀNH ]OVT ([ I]ꢀOVTA) was
2
5
2
purchased from Perkin Elmer (Walthman, MA, USA). Compound affinities were determined by
ꢀ
means of competition experiments in which the unlabeled compound concentrations varied from 10
1
5
ꢀ6
125
ꢀ11
to 10 M, and the concentration of the radioligand [ I]ꢀOVTA was 10 M. Nonspecific binding
ꢀ5
was determined in the presence of unlabeled OT (10 M). All assays were performed in triplicate
and repeated at least thrice.
22
ACS Paragon Plus Environment

Page 23 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Homology modeling and ligand docking
Details are provided in the Supporting Experimental Section. Briefly, the monomeric OTR
active conformation was built mainly on the basis of the β2ꢀadrenergic crystal structure (PDB
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6
3
3
SN6). Homodimeric OTR constellations were designed by usage of Kꢀopioid receptor (KOR)
(PDB 4DJH) and chemokine receptor type 4 (CXCR4) (PDB 3ODU) crystal structures as templates
for protomer arrangements and to simulate dimer interfaces TMH1ꢀTMH2ꢀH8 and TMH5ꢀTMH6ꢀ
intracellular loop 2 (ICL2), respectively. Monomeric OTR/dOTK (2) complexes were superimposed
with the protomers of KOR and CXCR4 dimers in an inactive conformation (the only dimerꢀ
conformations available at present in the literature).
For binding of the bivalent ligands into the dimeric OTR we assumed identical binding modes
of dOTK (2) in the putative orthosteric binding sites of OTR protomers forming the dimer, that was
comparable to the monovalent ligand in a monomeric OTR. Therefore, in a first step, monovalent
dOTKs were constraint in their binding mode at both protomers of the two hypothesized main OTR
dimer constellations (see above) and connected by spacers with different lengths. These complexes
were optimized concerning the spacers’ bondꢀlength and conformation. Initial minimization was
performed by constraining the backbone of the TMHs and the ligandꢀbackbone atoms. As none of
the bivalent spacerꢀlengths were suitable to bridge the distance between the dOTK moieties in the
OTR dimer with a TMH5ꢀTMH6ꢀICL2 interface by maintenance of the two orthosteric ligand
binding sites, further molecular dynamic simulations were conducted only to the dimeric models
with the TMH1ꢀTMH2ꢀhelix 8 interface, followed by energy minimization to justify and
energetically optimize the bivalent ligand orientation inside the OTR dimers.
Site-directed mutagenesis
Enzymes for molecular cloning were purchased from New England Biolabs (Ipswich, MA, USA)
and Promega (Madison, WI, USA). Oligonucleotides were synthesized by Eurofins MWG Operon
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 55
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
(Ebersberg, Germany). DNA sequencing was performed by Eurofins MWG Operon (Ebersberg,
Germany). DNA purification kits were obtained from Promega. PCR amplification was performed
using the GCꢀrich PCR system (Roche, Monza, Italy). For cloning purposes, E.coli DH5α bacteria
were used (Invitrogen, Life Technologies, NY, USA).
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
6
4
All mutations were introduced into the hOTRꢀpRK5 using standard PCR mutagenesis
65
technique . The oligonucleotidic primers (reported in the Supporting Experimental Section)
were designed on the basis of the sequence deposited at the NCBI (accession number NM 000196).
Sociability and social preference
+
/-
66
Oxtr mice
were obtained from L. Young (Emory University, Atlanta, GA, USA) and
housed as reported in the Supporting Experimental Section. Adult WT zebrafish were obtained
by local Aquarium supply store (Aquarium Center, Milan, Italy). All of the experimental procedures
followed the guidelines established by the Italian Council on Animal Care and were approved by
Italian Government Decree No. 17/2013 for mice and 18/2013 for zebrafish. Every effort was made
to minimize the number of animals used and their suffering.
For the sociability test in mice, a rectangular, threeꢀchamber transparent polycarbonate box was
52
employed and the test was conducted as previously described; procedural details are reported in
Supporting Experimental Section.
Social preference in zebrafish was carried out in a large glass aquarium according to a previous
53
study. Briefly, a WT zebrafish test fish was allowed to recognize two shoal stimuli, a Nacre
mutant and a WT counterpart, located in two opposite areas of the aquarium for up to 15 minutes
(
further details are described in Supporting Experimental Section). Shoaling preference was
quantified by recording the total time spent by the WT test fish in proximity of each stimulus shoal.
OT (1) and dOTK ꢀC8 (4b) were dissolved in cerebrospinal fluid and administered intraꢀ
2
cerebroventricularly (i.c.v) to mice via an implanted cannula (Supporting Experimental Section)
in a total volume of 2 ꢀL/mouse with a dose ranging from 0.005 to 0.5 ng/mouse, 10 min prior to
24
ACS Paragon Plus Environment

Page 25 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the sociability test. For zebrafish, isotocin (IT) (Bachem, Bubendorf, Switzerland) and dOTK ꢀC8
2
were dissolved in saline and administered intramuscularly (i.m.) (Supporting Experimental
Section) immediately prior the start of the test, with doses ranging between 0.002 and 1 ng/kg. All
solutions were freshly prepared daily from stocks stored at ꢀ20 °C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Data analysis
Data analysis, statistical significance and curve fitting, were calculated using GraphꢀPad Prism
software, version 5 (GraphPad Software, Inc., San Diego, CA). To determine the best fitting,
monophasic vs biphasic models were compared for each data set. Data points are reported as mean
i
± SEM. Estimated parameter values (EC50, Hill coefficient, Emax, K ) represent the mean ± SE of at
least three independent experiments each performed at least in triplicate.
BRET data of Gq activation were analyzed with a nonꢀlinear regression curve fitting procedure.
Monophasic curves where analyzed using the fourꢀparameter logistic model:
Y=Bottom + (Top-Bottom)/(1+10^((LogEC50-X)*HillSlope))
where Top and Bottom represent the lower and upper plateaus.
6
7
Biphasic curves were analyzed using a seven parameters model for nonꢀmonotonic curves
as
68
modified in
Where: δ = response; X = Log concentration of the agonist; Bottom = response when X = 0; EC50ꢀ1
and EC50ꢀ2 = concentrations of the agonist that produce half of the response of the first and second
component, respectively; b
Plateau = maximal response of the first component; Emax = response for an infinite concentration of
X, with constrained b and b equal to ꢀ1. Data from radioligand binding were evaluated by oneꢀsite
1 2
and b = slopes of the first and second component, respectively;
1
1
2
and twoꢀsites competitive binding curveꢀfitting procedure using GraphPad Prism version 5.
Synthetic transmembrane peptides effect on dimerization were analyzed by means of oneꢀway
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 55
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
ANOVA followed by Dunnett’s postꢀhoc test comparing all columns to the control (NT) column.
Mice and zebrafish behavioral data were analyzed by means of oneꢀway ANOVA for multiple
comparisons followed by Tukey’s or Bonferroni’s postꢀhoc test. A statistical level of significance of
P < 0.05 was accepted.
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
Corresponding author
Dr. Bice Chini, CNR, Institute of Neuroscience, Via Vanvitelli 32, 20129 Milano, Italy. Phone
+39ꢀ02ꢀ50316958; Fax +39ꢀ02ꢀ50317132; Eꢀmail: b.chini@in.cnr.it
Author Contribution
M.B. performed radioligand binding, functional assays, BRET experiments, FACS analysis and
fluorescence microscopy, generated mutant receptors, performed data analysis, and wrote the
manuscript. G.K. conducted docking and homology modeling, suggested receptor mutations and
wrote the manuscript. M.Mu. synthesized dOTK, dOTK
2 2 2 2
ꢀC6, dOTK ꢀC8, dOTK ꢀC10, dOTK ꢀ
C12, dOTK ꢀC14 and wrote the m. S.S. synthesized dOTK, dOTK
2
2
ꢀC8, dOTKꢀC8 and wrote the
corresponding experimental section. LB, LAH, and PJM designed and synthesized the disrupting
peptides and edited the manuscript.
2
M.Ma. planned and supervised dOTK, dOTK ꢀC8, dOTKꢀC8 chemical synthesis and wrote the
corresponding experimental section. D.B. and M.S. conceived and performed animal behavior
assays, performed data analysis and wrote the corresponding experimental section. S.D. generated
mutant receptors. GE.R. performed data analysis. T.B. performed data analysis and wrote the
manuscript. B.C. conceived and supervised the project, performed radioligand binding and
functional assays, performed data analysis and wrote the manuscript.
26
ACS Paragon Plus Environment

Page 27 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Competing Financial Interest
The authors declare that this study was conducted in the absence of any commercial or
financial relationships that could represent a potential conflict of interest.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Acknowledgements
This work was supported by the Telethon Foundation (grant GGP12207), CNR Research
Project on Aging and Regione Lombardia (Project MbMMꢀconvenzione n°18099/RCC) to B.C.;
European Union Seventh Framework Program (Marie Curie grant 254897 and 2013ꢀBPꢀBꢀ00109),
the Secretary of Universities and Research of the Economy and Knowledge Department of the
Government of Catalonia and ARC DECRA (DE150100784) to M.Mu.; DFG (grant KL2334/2ꢀ2)
and EKFS project 2014_A114 to GK; the National Institute of General Medical Sciences (grant
GMꢀ25280) toꢀ M.Ma. and BBSRC DTP funding to PJM and LAH (MCCORMICK_U15DTP1).
M.B. is the recipient of a Foundation Umberto Veronesi postꢀdoctoral fellowship. We thank Dr. C.
Galés, INSERM, France who provided the BRET biosensor cDNAs, and Fondazione Monzino for
gifting the LSM510 META confocal laserꢀscanning microscope (Zeiss, Jena, Germany).
Abbreviations Used
BRET, bioluminescence resonance energy transfer; cAMP, cyclic adenosine 3’,5’ꢀ
monophosphate; CXCR4, chemokine receptor type 4; CSF, cerebrospinal fluid; DOTA, 1,4,7,10
tetraazacyclododecaneꢀ1,4,7,10ꢀtetraacetic acid; ECL, extracellular loop; ICL, intracellular loop;
2
10
FACS, fluorescence activated cell sorting; GFP and GFP , blue shift variants of the green
fluorescent protein; GPCR, Gꢀprotein coupled receptors; Gx: Gꢀprotein subunit; i.c.v.,
intracerebroventricular; i.m. intramuscular; IT, isotocin; KOR, Kꢀopioid receptor; L, length; OT,
2
4
9
2 5 2
oxytocin; OTR, oxytocin receptor; OVTA, d(CH ) [Tyr(Me) , Thr , TyrꢀNH ]ornithine vasotocin;
Rluc, Renilla reniformis Luciferase; TAT, portion of the HIV transactivator of transcription; TMH,
transmembrane helix; V, volume; WT, wildꢀtype.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 55
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
ANCILLARY INFORMATION
Authors will release the atomic coordinates of presented 3Dꢀmodels and experimental data
upon article publication.
(1) Oxytocin; (2) dOTK; (3) dOTKꢀC8; (4a) dOTK
2 2 2
ꢀC6; (4b) dOTK ꢀC8; (4c) dOTK ꢀ
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
C10; (4d) dOTK ꢀC12; (4e) dOTK ꢀC14;
2
2
Supporting Information
Supporting information are available and include supporting figures, supporting experimental
section, supporting notes. PDB coordinates for docking pose of 2 in the monomeric OTR, PDB
coordinates for docking of 4b in the dimeric OTR and PDB coordinates for docking of 4e in the
dimeric OTR have been included in the supporting information as pdb files
28
ACS Paragon Plus Environment

Page 29 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
REFERENCES
Bouvier, M. Oligomerization of Gꢀproteinꢀcoupled transmitter receptors. Nat. Rev. Neurosci.
001, 2, 274ꢀ286.
Pin, J. P.; Neubig, R.; Bouvier, M.; Devi, L.; Filizola, M.; Javitch, J. A.; Lohse, M. J.;
1
.
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.
Milligan, G.; Palczewski, K.; Parmentier, M.; Spedding, M. International Union of Basic and
Clinical Pharmacology. LXVII. Recommendations for the recognition and nomenclature of G
proteinꢀcoupled receptor heteromultimers. Pharmacol. Rev. 2007, 59, 5ꢀ13.
3
.
Ferre, S.; Casado, V.; Devi, L. A.; Filizola, M.; Jockers, R.; Lohse, M. J.; Milligan, G.; Pin,
J. P.; Guitart, X. G proteinꢀcoupled receptor oligomerization revisited: functional and
pharmacological perspectives. Pharmacol. Rev. 2014, 66, 413ꢀ434.
4
.
Hiller, C.; Kuhhorn, J.; Gmeiner, P. Class A Gꢀproteinꢀcoupled receptor (GPCR) dimers and
bivalent ligands. J. Med. Chem. 2013, 56, 6542ꢀ6559.
Shonberg, J.; Scammells, P. J.; Capuano, B. Design strategies for bivalent ligands targeting
GPCRs. Chem. Med. Chem. 2011, 6, 963ꢀ974.
Valant, C.; Robert Lane, J.; Sexton, P. M.; Christopoulos, A. The best of both worlds?
5
.
6
.
Bitopic orthosteric/allosteric ligands of G proteinꢀcoupled receptors. Annu. Rev. Pharmacol.
Toxicol. 2012, 52, 153ꢀ178.
7
.
Lane, J. R.; Sexton, P. M.; Christopoulos, A. Bridging the gap: bitopic ligands of Gꢀproteinꢀ
coupled receptors. Trends Pharmacol. Sci. 2013, 34, 59ꢀ66.
Daniels, D. J.; Lenard, N. R.; Etienne, C. L.; Law, P. Y.; Roerig, S. C.; Portoghese, P. S.
8
.
Opioidꢀinduced tolerance and dependence in mice is modulated by the distance between
pharmacophores in a bivalent ligand series. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 19208ꢀ19213.
29
ACS Paragon Plus Environment