Investigation of novel ropinirole analogues: Synthesis, pharmacological evaluation and computational analysis of dopamine D2 receptor functionalized congeners and homobivalent ligands

Article abstract of DOI:10.1039/c4md00066h

Herein, we report the development of novel functionalized congeners of ropinirole toward the design of pharmacological tools to probe structural requirements at the dopamine D₂ receptor. Subsequently, we have used the functionalized amine congener 11 and synthesized and pharmacologically evaluated a series of homobivalent ligands of ropinirole with designated spacer lengths ranging from 14 to 30 atoms. The most potent homobivalent ligands (22-, 26- and 30-atom spacers) showed approximately 20- to 80-fold greater potency (EC₅₀ = 3.9, 6.2 and 14 nM, respectively) than ropinirole (304 nM) in a [³⁵S]GTPγS functional assay. Molecular modeling studies suggest that the observed increase in potency of the homobivalent ligands is possibly due to a bitopic binding mode involving the orthosteric site and an allosteric interaction at the dopamine D₂ receptor protomer rather than bridging interactions at two orthosteric sites across a dopamine D ₂ receptor dimer. This research has the potential to advance the development of structurally related bitopic ligands, biomarkers such as radioligands and fluorescently labeled probes, and furnish new homo- and heterobivalent ligands towards a better understanding of the dopamine D ₂ receptor and potential novel treatment for Parkinson's disease.

Full text of DOI:10.1039/c4md00066h

MedChemComm
View Article Online
View Journal | View Issue
CONCISE ARTICLE
Investigation of novel ropinirole analogues:
synthesis, pharmacological evaluation and
computational analysis of dopamine D2 receptor
functionalized congeners and homobivalent
ligands†
Cite this: Med. Chem. Commun., 2014,
5, 891
Manuela Jorg,a Agnieszka A. Kaczor,bc Frankie S. Mak,d Kiew Ching K. Lee,d
¨
c
d
a
a
*
*
Antti Poso, Neil D. Miller, Peter J. Scammells and Ben Capuano
Herein, we report the development of novel functionalized congeners of ropinirole toward the design of
pharmacological tools to probe structural requirements at the dopamine D2 receptor. Subsequently, we
have used the functionalized amine congener 11 and synthesized and pharmacologically evaluated a
series of homobivalent ligands of ropinirole with designated spacer lengths ranging from 14 to 30 atoms.
The most potent homobivalent ligands (22-, 26- and 30-atom spacers) showed approximately 20- to
80-fold greater potency (EC50 ¼ 3.9, 6.2 and 14 nM, respectively) than ropinirole (304 nM) in a [35S]
GTPgS functional assay. Molecular modeling studies suggest that the observed increase in potency of
the homobivalent ligands is possibly due to a bitopic binding mode involving the orthosteric site and an
allosteric interaction at the dopamine D2 receptor protomer rather than bridging interactions at two
orthosteric sites across a dopamine D2 receptor dimer. This research has the potential to advance the
development of structurally related bitopic ligands, biomarkers such as radioligands and fluorescently
labeled probes, and furnish new homo- and heterobivalent ligands towards a better understanding of the
dopamine D2 receptor and potential novel treatment for Parkinson's disease.
Received 19th February 2014
Accepted 3rd May 2014
DOI: 10.1039/c4md00066h
www.rsc.org/medchemcomm
broad range of diseases including Parkinson's disease (PD),
restless legs syndrome, sexual dysfunction, dementia, depres-
Introduction
sion and schizophrenia.
The dopamine receptors, are members of the G protein-coupled
receptor (GPCR) super family, and implicated in many neuro-
logical processes, such as motivation, pleasure, cognition,
memory, learning, and ne motor control.1 Dopamine initiates
its effects through the activation of ve GPCR subtypes; namely
dopamine receptors D1 to D5.1 Commercially available dopa-
minergic drugs (pro-drugs, agonists, antagonists and enzyme
inhibitors) demonstrate clinical utility in the treatment of a
More recent studies provide evidence that some of the tar-
geted dopaminergic receptors exist not only in monomeric
form, but also as homo- and heterodimers and/or higher
ordered oligomers.2–4 Homobivalent ligands are described as
molecules comprised of two identical pharmacophoric enti-
ties5,6 covalently tethered by an appropriate spacer and have
been used to investigate the properties and function of dopa-
mine homodimers and higher ordered oligomers. To date, there
are numerous published examples of dopamine D2 receptor
homobivalent ligands based on 1,4-disubstituted aromatic
piperidine/piperazines,7–10 clozapine,11 5-hydroxy-2-(dipropyla-
mino)tetralin (5-OH-DPAT)12 and apomorphine13 pharmaco-
phores. The most remarkable gain in binding affinity or
functional potency compared to the original pharmacophore
has been reported for homobivalent ligands of the dopamine D2
antagonist clozapine11 (ꢀ80-fold) and the dopamine D2 agonist
5-OH-DPAT12 (ꢀ95-fold).
aMedicinal Chemistry, Monash Institute of Pharmaceutical Sciences, 381 Royal
Parade, Parkville, Victoria 3052, Australia. E-mail: ben.capuano@monash.edu;
peter.scammells@monash.edu
bDepartment of Synthesis and Chemical Technology of Pharmaceutical Substances with
Computer Modeling Lab, Faculty of Pharmacy with Division of Medical Analytics,
´
Medical University of Lublin, 4A Chodzki St., PL-20093, Lublin, Poland
cSchool of Pharmacy, University of Eastern Finland, Yliopistonranta 1, P.O. Box 1627,
FI-70211, Kuopio, Finland
dGSK R&D, Neural Pathways DPU, Neurosciences TAU, 11 Biopolis Way, Helios Bldg
#03-01/02, Singapore 138667, Singapore
This study is focused on agonism of the dopamine D2
receptor which is linked to the treatment of motor dysfunction
in PD. Currently, the dopamine pro-drug L-dopa as well as
dopamine D2 receptor agonists are used for the treatment of
† Electronic supplementary information (ESI) available: Details of chemical
synthesis and characterization of all the reported compounds. Details of the
[
35S]GTPgS assay including example curves. Details of molecular modeling
study. See DOI: 10.1039/c4md00066h
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 891–898 | 891

View Article Online
MedChemComm
Concise Article
PD.14,15 These drugs give some relief to patients by compen- molecular weight associated with some of these concepts,
sating for the progressive loss of dopaminergic neurons of which may limit their usefulness as potential drugs, they are
substantia nigra in the basal ganglia. Nevertheless, clinically valuable pharmacological tools to further explore the structural
used L-dopa and dopamine D2 receptor agonists are not able to features and functional properties of GPCRs.
cure patients with PD and can cause side effects such as
Structure–activity relationship data from the literature were
induced dyskinesia, ‘on-off’ oscillation and ‘wearing off’ used to guide the design of our ropinirole-based functional
effects.15 Consequently, a better understanding of the dopamine congeners. The design of our congeners and homobivalent
D2 receptor in relation to its intrinsic structural and functional ligands was based on the ndings published by Namil et al.;
properties is desperately needed to develop novel treatment who successfully replaced one n-propyl chain with either ethyl-
methods for patients with PD.
or methyl 4-butanoate, while maintaining binding affinity
(Ki) values for the dopamine D2 receptor similar to that of
ropinirole (1).24 This general approach was adopted based
upon literature precedent as well as the synthetic accessibility
of the tertiary amine (Scheme 1).
Ropinirole
In 1996, ropinirole (1) entered the market as a non-ergoline
dopamine D2 receptor agonist for the treatment of PD.16,17 This
molecule acts as a full agonist at the D2, D3, and D4 receptors
and has little affinity towards the D1 and D5 receptors
(Fig. 1).18–20 Ropinirole (1) has been chosen for this study due to
its low molecular weight, its simple and accessible chemical
structure and the absence of any stereocentres. To our knowl-
edge this is the rst example of a series of homobivalent ligands
investigating dopamine D2 agonism using a pharmacophore of
clinical relevance.
The commercially available hydrochloride salt of ropinirole
(2) was transformed to the free base 1, followed by dealkylation
of a single propyl chain with 1-chloroethyl chloroformate. The
hydrochloride salt of the dealkylated ropinirole 4 was formed in
43% yield as well as the doubly dealkylated compound 3 as the
minor byproduct (12% yield, product/byproduct ratio 4 : 1).
Alternatively compound 4 was synthesized from the alcohol 4-
(2-hydroxyethyl)indolin-2-one (5), which was subsequently
activated to the stable, crystalline tosylate derivative 6, then
further reacted with n-propylamine at reux to give 4 in 74%
yield. The second pathway was preferred due to robustness,
simple purication (no column chromatography required) and
higher yields. Compound 4 was converted to the free base and
then further reacted with either ethyl 4-bromobutyrate or tert-
butyl(3-bromopropyl)carbamate to afford intermediates 7 and 9
in moderate yields of 28% and 38%, respectively. It is note-
worthy that for the scenario involving alkylation with the amine
linker (tert-butyl(3-bromopropyl)carbamate), the use of a stoi-
chiometric amount of the reagents was critical to avoid the
dialkylated byproduct 10. Alkaline hydrolysis of 7, followed by
acidic workup, potentially leads to the functionalized oxindole
carboxylic acid congener (OCAC) (8) which can then be used to
synthesize derivatives that further extend towards the extracel-
lular space. On the other hand, removal of the Boc-group of 9
Fig. 1 Chemical structure of the dopamine D2 receptor agonist
ropinirole (1) and Ki values for the different dopamine receptor
subtypes from experiments in human cell-lines.18
Design, synthesis and pharmacological furnished the functionalized oxindole amine congener (OAC)
(11) which could be utilized for the same purpose. It is impor-
evaluation of ropinirole congeners
tant to note that compound 9 was stable in the solid state at 5 ꢁC
To rationally design and synthesize a functionalized congener – but slowly degraded in solution. Compound 11 rapidly
a molecule that comprises a chemical functional group (e.g. degraded in both states therefore immediate use is recom-
amine or carboxylic acid) that is suitably linked to a pharma- mended. Consequently, the free amine linker 11 was reacted
cophore21,22 – it is crucial to locate the optimal attachment point with decanoyl chloride to afford the monovalent variant 12 in
for a linker on the pharmacophore. Not only must the activity of 17% yield (Scheme 2). This compound was utilized to evaluate if
the functionalized pharmacophore be retained but also its extensions in this position are well tolerated in view of
chemical accessibility is essential. The functional congener designing more elaborate pharmacological tools.
approach has been successfully used to investigate GPCRs such
The monovalent ligands were tested in a [35S]GTPgS assay
as the adenosine, muscarinic and adrenergic receptor as well as (results Table 1) using Chinese hamster ovary cells. This assay
the purinergic P2Y nucleotide receptors.23 Adding further measures the agonist-stimulated activation of GPCRs close to
functionality to the correct position of an existing pharmaco- the receptor in the signaling cascade. As a result little ampli-
phore potentially enhances ligand affinity and selectivity of cation of the signal is perceived compared to other assays
drug candidates. These functionalized congeners also provide a downstream in the signaling cascade.25 The results were
suitable starting point for the design of bitopic ligands, compared to that of quinelorane, which behaved as a full
biomarkers such as radioligands and uorophores as well as agonist at the dopamine D2 receptor with a pEC50 of 7.35 ꢂ 0.14
hetero- and homobivalent ligands. Despite the handicap of high (EC50 ¼ 45 nM).
892 | Med. Chem. Commun., 2014, 5, 891–898
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
Scheme 1 Reaction pathway to synthesize the functionalized oxindole based congeners 8 and 11. Reagents and conditions: (a) 1 M NaOH, DCM,
rt, 30 min, 94% (1); (b) 1-chloroethyl chloroformate, Na2CO3, DME, 85 ꢁC, 17 h followed by MeOH, reflux, 18 h, 12% (3), 43% (4); (c) p-tolue-
nesulfonyl chloride, DCM, pyridine, 5–10 ꢁC, 4 h, 80%; (d) n-propylamine, reflux, 1.5 h, 74%; (e) 1 M NaOH, DCM, rt, 20 min, 92%; (f) ethyl 4-
bromobutyrate, K2CO3, acetone, reflux, 27 h, 28%; (g) tert-butyl(3-bromopropyl)carbamate, K2CO3, acetonitrile, reflux, 19 h, 38%; (h) 4 M HCl in
dioxane, MeOH, rt, 1 h, 84%.
compound 12 as the standout. These results indicate that
variations to one of the alkyl chains are very well tolerated,
maintaining functional potency at the dopamine D2 receptor;
neither the removal of one propyl chain (compound 4) nor the
attachment of a long spacer (compound 12) diminished the
potency or efficacy of the original pharmacophore. Linker
attachment to the lactam nitrogen (compound 10) on the other
hand not only displayed a 5-fold decrease in potency relative to
ropinirole (1) but also a substantial drop in the maximum
response. In summary, the tertiary amine of ropinirole (1) has
been shown to be a suitable position to extend towards the
extracellular space of the dopamine D2 receptor. As a conse-
quence, the functionalized congeners 8 and 11 proved to be
promising starting points for the synthesis of pharmacological
tools to target the dopamine D2 receptor.
Scheme 2 Synthesis of monovalent ligand 12 as the decanamide.
Reagents and conditions: (a) decanoyl chloride, DIPEA, DCM, 45 min, 17%.
All the synthesized monovalent compounds with the excep-
tion of 10 showed higher functional potencies (EC50 values of
42–219 nM) compared to ropinirole (1, EC50 ¼ 304 nM) with
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 891–898 | 893

View Article Online
MedChemComm
Concise Article
Table 1 Functional potencya and efficacya of synthesized dopamine Table 2 Functional potencya and efficacya of synthesized dopamine
D2 monovalent ligands using a [35S]GTPgS assay. Emax values were D2 homobivalent ligands 15a–e and monovalent ligand 16 using a [35S]
referenced to quinelorane at 100 mM
GTPgS assay. Emax values were referenced to quinelorane at 100 mM
Compound #
pEC50 ꢂ SEM
EC50 (nM)
Emax (%)
Compound #
n
Spacer pEC50 ꢂ SEM EC50 (nM) Emax (%)
Quineloraneb
7.35 ꢂ 0.14
6.52 ꢂ 0.02
7.07 ꢂ 0.09
7.26 ꢂ 0.06
6.66 ꢂ 0.02
5.84 ꢂ 0.16
7.38 ꢂ 0.15
45
304
85
100 ꢂ 2
74 ꢂ 1
92 ꢂ 4
90 ꢂ 2
69 ꢂ 7
Quinelorane
7.35 ꢂ 0.14
6.52 ꢂ 0.02
7.55 ꢂ 0.17
7.65 ꢂ 0.13
8.41 ꢂ 0.06
8.21 ꢂ 0.11
7.85 ꢂ 0.12
5.33 ꢂ 0.06
45
304
28
100 ꢂ 2
74 ꢂ 1
94 ꢂ 3
99 ꢂ 2
85 ꢂ 4
85 ꢂ 3
58 ꢂ 4
72 ꢂ 4
Ropinirole (1)c
Ropinirole (1)b
4
7
15a
15b
15c
2
6
14
18
55
22
9d
10
12b
219
1413
42
10 22
14 26
18 30
3.9
6.2
14
4677
45 ꢂ 11 15d
78 ꢂ 4
15e
16c
2
9
a
Data represent the mean ꢂ SEM of four separate experiments performed
b
a
in duplicate. Data represent the mean
ꢂ
SEM of six separate
Data represent the mean
ꢂ
SEM of six separate experiments
c
experiments performed in duplicate. Literature value (Ghosh et al.);27
Ropinirole was not tested in this assay but the potency of quinelorane
is consistent with the literature26 thereby permitting subsequent
b
performed in duplicate. Literature value (Ghosh et al.);27 Ropinirole
was not tested in this assay but the potency of quinelorane is
consistent with the literature26 thereby permitting subsequent
d
c
comparisons of test compounds to ropinirole (1). Data represent the
comparisons of test compounds to ropinirole (1). Data represent the
mean ꢂ SEM of two separate experiments performed in duplicate due
mean ꢂ SEM of four separate experiments performed in duplicate
to instability of the stock solution.
due to insufficient amount of sample.
and degree of agonism (results Table 2). The synthesized
homobivalent ligands 15a–e exhibited higher potency than the
original pharmacophore 1 (Fig. 2, top), whereas the monovalent
Synthesis and pharmacological
evaluation of homobivalent ligands of
ropinirole
compound 16 showed a signicant decrease in potency (EC50
>
4 mM). The analogues with a spacer length of 14 to 18 atoms
(compound 15a and b) showed approximately 11- to 14-fold
greater functional potencies relative to ropinirole (1), which are
comparable to the most active monovalent ligands represented
in Table 1. An additional substantial gain in potency, ranging
from 20- up to 80-fold compared to ropinirole (1) and 3- to 11-
fold compared to quinelorane, was observed for the homo-
bivalent ligands with a spacer length of 22 to 30 atoms
(compounds 15c–e). Interestingly, compounds 15a–d main-
tained full agonism whereas the largest homobivalent ligand
15e showed a substantial drop in the maximum response
(Fig. 2, bottom). On the other hand, the monovalent analogue
16 exhibited a similar maximum response to ropinirole (1) but
with signicantly reduced functional potency.
On the basis of the results from our functionalized congeners,
linking through the tertiary amine was used for the synthesis of
a series of homobivalent ligands with spacer lengths (counting
the atoms between the ionizable nitrogen atoms) from 14 to 30
atoms. The homobivalent ligands of ropinirole ranging from 2
to 16 methylene units in length were synthesized (Scheme 3) by
adding the functionalized OAC (11) to the appropriate diacid
chloride 14a–e, that was formed in situ from the corresponding
dicarboxylic acid, oxalyl chloride and catalytic dimethylforma-
mide (Vilsmeier reaction). The reaction was generally complete
within one hour at room temperature affording product yields
varying from 4–31% and was strongly dependent on the degree
of degradation of amine 11. In one case, where the degradation
of 11 had progressed extensively, only the monovalent carbox-
ylic acid with the four carbon linker (n ¼ 2) was isolated
(compound 16).
Docking
The homobivalent ligands were tested in the previously
described [35S]GTPgS assay to ascertain their functional potency
A dopamine D2 receptor dimer model was built to correlate the
pharmacological ndings with a possible binding mode. In this
Scheme 3 Synthesis of homobivalent ligands 15a–e. Reagents and conditions: (a) oxalyl chloride, DCM, DMF, rt, 1 h; (b) 11, DCM, DIPEA, rt, 1 h,
4% (15a), 16% (15b), 17% (15c), 22% (15d), 31% (15e) and 4% (16).
894 | Med. Chem. Commun., 2014, 5, 891–898
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
then re-docked for pose renement. The main interactions
observed for ropinirole (1) originate from the carbonyl group of
the oxindole scaffold and the ionizable nitrogen of the side
chain with Ser197(5.46) and Asp114(3.32), respectively
(numbers in parenthesis indicate residue numbers according to
Ballesteros–Weinstein system).32 Linker attachment to the
tertiary amine (compounds 7, 9 and 12) lead to a slight increase
in functional potency (up to 7-fold compared to ropinirole)
which is possibly due to an additional interaction between the
introduced side chain carbonyl group of the analogues and
Thr412(7.39) (Fig. 3). Compound 10, which represents a mole-
cule with a linker attached to the lactam group and the tertiary
amine not surprisingly was unable to dock in the same pose as
ropinirole (1), potentially explaining the poor biological results.
The low potency ascribed to compound 16 could not be
explained with the docking results and may result from unfav-
ourable receptor interactions due to the presence of the
carboxylic acid group (as the carboxylate anion) and/or the
zwitterionic state of the molecule at physiological pH.
Docking of homobivalent ligands 15a–e was performed
methodologically to ensure the interaction of the ionizale
nitrogen of the ligand with Asp114(3.32) in at least one proto-
mer. Docking of the homobivalent ligands 15a and 15b showed
similar interactions observed with the monovalent ligands (7, 9
and 12). The docking results showed that 15b is unable to reach
the second protomer whereas the homobivalent ligands 15c and
15d are sufficiently extended to exhibit signicant interactions
at both protomers across the homodimer (Fig. 4). The docking
poses of homobivalent ligands 15c to 15e direct towards the
allosteric sites in protomer B involving the extracellular end of
helices TM4 and TM5 as well as the extracellular loop ecl2. Fig. 5
illustrates a binding mode where the homobivalent ligand 15e
interacts with the extracellular region of the second protomer
(A) or crosses the membrane region allowing interactions with
both orthosteric sites simultaneously (B). The second binding
mode, although theoretically possible, is unlikely to occur in a
biological system due to steric considerations. It is noteworthy
that the aforementioned unfavourable binding mode for 15e
was the only docking position conserved that ensured ligand
interaction with the critical Asp114(3.32) residues in both pro-
tomers. The docking results of the homobivalent ligands 15a–e
are in accordance with preliminary performed measurements of
the size of the molecules (Table 4) and the comparison with the
distances between the different binding sites (Table 3). The
results suggest that the additional increase in potency of
homobivalent ligands 15c–e is possibly due to a bitopic mode of
binding involving the orthosteric site of protomer A with an
allosteric site on protomer B. It should be noted that this pre-
dicted binding mode is one of many possible binding modes;
the same increase in potency could be explained by a bitopic
binding mode involving the orthosteric and allosteric sites
present within the same protomer. Similar conclusions were
previously made by Gmeiner's group for another class of D2
receptor bivalent ligands (whereby bivalent ligands displace just
one equivalent of orthosteric ligand).8
Fig. 2 Graphical representation of the functional potency (top) and
efficacy (Emax, % maximum stimulation quinelorane) (bottom) of
ropinirole (1), homobivalent ligands (15a–e) with increasing length and
monovalent ligand 16.
study, a symmetric homodimer with a TM3TM4TM5 interface
from both individual protomers was built. This particular
interface was chosen since it is the commonly used approach to
build GPCR dimers28 and the same interface has been used for
the dopamine D2 homodimer29,30 and dopamine D2- adenosine
A
2A heterodimer models.31 The dimensions of the dopamine D2
homodimer model were measured out and the approximate
distances between the different binding sites are documented
in Table 3.
Ropinirole (1), the monovalent ligands 4, 7, 9, 10, 12 and 16
as well as the homobivalent ligands 15a were docked into the
model of the receptor monomer, whereas the compounds 15b–e
were manually placed in the dopamine D2 receptor dimer and
Table 3 Distances between different binding sites in the dopamine D2
receptor dimer model (in angstroms)
Orthosteric–orthosteric in different protomers
through the membrane region
ꢀ34
Orthosteric–orthosteric in different protomers
without crossing a membrane region
ꢀ50–60
ꢀ18
Orthosteric–allosteric in one protomer
Orthosteric–allosteric in different protomers
Although the absolute binding mode of the synthesized
ropinirole-based homobivalent ligands remains elusive, this
across the homodimer
ꢀ30–40
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 891–898 | 895

View Article Online
MedChemComm
Concise Article
Fig. 3 Schematic representation of the binding site and the most important interactions of ligand 7 with the dopamine D2 receptor. (A) Overview
of the ligand–receptor complex; (B) details of the binding site; (C) schematic representation of the binding site. The crystal structure of the b2
adrenergic receptor in a complex with the Gs protein (PDB ID: 3SN6)33 was used as a template for the homology model of the dopamine D2
receptor in active conformation. There are two disulfide bridges present in the model: one linking TM3 and e2 (Cys107-Cys182) and the other in
e3 loop (Cys399-Cys401).
study represents the rst time a dopamine D2 receptor dimer additional interaction of the introduced carbonyl group with
model has been used to demonstrate that the spacer length of Thr412 within the same protomer. However, the designed
30 carbon atoms is probably insufficient to allow interaction of monovalent ligands (7, 9 and 12) have a deciency in hydrogen-
the ligand with the orthosteric sites of both protomers in a bond acceptor and donor atoms whilst the homobivalent
favorable manner, i.e. with the linker in the extracellular region, ligands 15a and 15b lack molecular size in order to interact with
rather than across the membrane region.
the key residues in the proximity of the allosteric region of the
D2 receptor, therefore limiting their gains in potency. The
further extended homobivalent ligands with a spacer length of
22 up to 30 carbon atoms (compounds 15c–e) showed a 20- to
80-fold increase in functional potency compared to the parent
molecule and a 3- to 11-fold increase compared to the reference
compound quinelorane. Interestingly, compound 15e was the
only homobivalent ligand which showed a substantial drop in
efficacy. Molecular modeling studies of these molecules
proposed that the substantial increase in potency is unlikely
due to interactions at both orthosteric sites within a dopamine
D2 receptor homodimer. The lengths of the homobivalent
ligands do not allow interactions at both orthosteric sites
simultaneously due to steric considerations, however we predict
that the second pharmacophore possibly interacts with an
allosteric site present at the second protomer of the homo-
dimer. The bitopic binding mode across the dimer is just one of
many possible binding modes. The increased ligand potency
resulting from additional interactions at an allosteric site
Discussion and conclusions
In summary, we have developed a functionalized oxindole
carboxylic acid congener (OCAC) (8) and oxindole amine
congener (OAC) (11) which offer excellent starting points for the
design and synthesis of biological tools to further investigate
the dopamine D2 receptor topology and its role in biological
processes. On the basis of our functionalized congener 11, a
series of homobivalent ligands with spacer lengths from 14 to
30 atoms were synthesized and pharmacologically evaluated in
a [35S]GTPgS assay to ascertain functional potency and ago-
nism. Ropinirole analogues with the linker attached to the
tertiary amine and the homobivalent ligands with a spacer
length of 14 and 18 carbons atoms showed an approximate 1.5-
to 14-fold increase in functional potency compared to ropinirole
as well as full agonism; molecular modeling studies indicate
that the gain in ligand functional potency is possibly due to an
896 | Med. Chem. Commun., 2014, 5, 891–898
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
Table 4 The calculated overall size and the distance between the
ionizable nitrogen atoms of the homobivalent ligands 15a–e (in
angstroms)
Length of
compounds
Distance between ionizable
nitrogen atoms (A)
˚
Compound #
15a
15b
15c
15d
15e
ꢀ29
ꢀ31
ꢀ36
ꢀ37
ꢀ48
14
19
24
28
33
simultaneously. Consequently, it is feasible that published
homobivalent ligands with similar spacer spans between the
two pharmacophores (1,4-disubstituted aromatic piperidine/
piperazines,7–10 clozapine,11 5-hydroxy-2-(dipropylamino)tetra-
lin (5-OH-DPAT)12 and apomorphine13 based molecules) may
also act in a bitopic mode rather than targeting two orthosteric
sites across a homodimer.
The nding that the most potent ropinirole-based homo-
bivalent ligands exhibited 20- to 80-fold greater potency than
ropinirole is intriguing, even more so considering the reduced
likelihood of these molecules interacting at two orthosteric sites
across a homodimer. Consequently, further investigations
should consider: (a) if dopamine D2 receptor homobivalent
Fig. 4 Representation of dopamine D2 receptor dimer model with
homobivalent ligands docked to the binding cavity. (A) Compound ligands published in the literature are generally too short to
15b; (B) compound 15c; (C) compound 15d.
interact at two orthosteric sites of a homodimer, would
exchanging one of the orthosteric pharmacophores with an
allosteric dopamine D2 receptor fragment further improve the
within a single protomer cannot be excluded. Similar changes
in potency have been reported with homobivalent ligands based
on the dopamine D2 receptor antagonist clozapine and the
dopamine D2 receptor agonist 5-hydroxy-2-(dipropylamino)tet-
ralin (5-OH-DPAT).12 The molecular modeling study presented
gives supporting evidence that the ropinirole-based homo-
bivalent ligands described herein are unlikely to interact at two
orthosteric sites of a dopamine D2 receptor homodimer
potency, selectivity and ligand efficiency of such pharmacolog-
ical tools? (b) would the incorporation of a longer linker that
facilitates the targeting of the two orthosteric sites simulta-
neously further increase the potency of homobivalent ligands?
and (c) if, as suggested in the literature, only one protomer of a
homodimer becomes activated and signals to a G protein, then
would the design and synthesis of a molecule that incorporates
both a dopamine D2 receptor agonist and a dopamine D2
Fig. 5 Alternative poses for homobivalent ligands 15e. Pose B is unrealistic due to steric clashes and was generated to depict how this ligand can
reach orthosteric sites in two protomers.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 891–898 | 897

View Article Online
MedChemComm
Concise Article
receptor antagonist be of greater benet? In addition, comple- 11 F. M. McRobb, I. T. Crosby, E. Yuriev, J. R. Lane and
mentary mutagenesis studies should be considered to deter- B. Capuano, J. Med. Chem., 2012, 55, 1622–1634.
mine the amino acid residues that are involved in the 12 S. Gogoi, S. Biswas, G. Modi, T. Antonio, M. E. A. Reith and
interactions of the aforementioned pharmacological tools and
the dopamine D2 receptor.
A. K. Dutta, ACS Med. Chem. Lett., 2012, 3, 991–996.
13 J. Shonberg, J. R. Lane, P. J. Scammells and B. Capuano,
MedChemComm, 2013, 4, 1290–1296.
The study presented provides further insight that will assist
in the understanding of the complex mechanism of receptor 14 R. Tintner and J. Jankovic, Expert Opin. Invest. Drugs, 2003,
dimers and signalling pathways associated with them.
12, 1803–1820.
¨
15 T. Muller, Transl. Neurodegener., 2012, 1, 10.
16 G. Gallagher Jr., US4452808A, 1984.
Acknowledgements
17 M. C. De and H. N. Sree, WO9200735A1, 1992.
18 P. M. Luthra and J. B. S. Kumar, Mini-Rev. Med. Chem., 2012,
12, 1556–1564.
19 R. J. Eden, B. Costall, A. M. Domeney, P. A. Gerrard,
C. A. Harvey, M. E. Kelly, R. J. Naylor, D. A. A. Owen
and A. Wright, Pharmacol., Biochem. Behav., 1991, 38,
147–154.
20 C. Reavill, I. Boyeld, M. Coldwell and P. Nelson, J. Pharm.
Pharmacol., 2000, 52, 1129–1135.
21 K. A. Jacobson, D. Ukena, W. Padgett, K. L. Kirk and
J. W. Daly, Biochem. Pharmacol., 1987, 36, 1697–1707.
M. J. is a recipient of the Australian Postgraduate Award Industry
(APAI) and Monash University Postgraduate Publication Award.
We thank GSK R&D Singapore for nancial support. Thank you to
Francisco Javie Colilla-Rodriguez (GSK R&D Tres Cantos, Spain)
for the dopamine D2 receptor testing of our compounds. Thanks
to the members of the Medicinal Chemistry department in GSK
Singapore where some of this work was performed. A part of
research was carried out during the postdoctoral fellowship of A.
A. K. under Marie Curie IEF grant. Calculations were performed
under a computational grant by Interdisciplinary Center for
Mathematical and Computational Modelling (ICM), Warsaw,
Poland, grant number G30-18 and under resources of CSC, Fin-
land. Thanks to Dr Stephen Headey (Monash University) for his
valuable support with the NMR experiments.
¨
22 M. Jorg, J. Shonberg, F. S. Mak, N. D. Miller, E. Yuriev,
P. J. Scammells and B. Capuano, Bioorg. Med. Chem. Lett.,
2013, 23, 3427–3433.
23 K. A. Jacobson, Bioconjugate Chem., 2009, 20, 1816–1835.
24 A. Namil, M. Benoit-Guyod and G. Leclerc, Eur. J. Med.
Chem., 1995, 30, 973–981.
25 P. G. Strange, Br. J. Pharmacol., 2010, 161, 1238–1249.
26 A. Gobert, A. Dekeyne and M. J. Millan, Neuropharmacology,
2000, 39, 1608–1616.
Notes and references
1 J.-M. Beaulieu and R. R. Gainetdinov, Pharmacol. Rev., 2011,
63, 182–217.
27 B. Ghosh, T. Antonio, M. E. A. Reith and A. K. Dutta, J. Med.
Chem., 2010, 53, 2114–2125.
2 R. K. Somvanshi and U. Kumar, Pharmaceuticals, 2012, 5,
417–446.
28 A. A. Kaczor and J. Selent, Curr. Med. Chem., 2011, 18, 4606–
4634.
3 S. Gines, J. Hillion, M. Torvinen, C. S. Le, V. Casado,
E. I. Canela, S. Rondin, J. Y. Lew, S. Watson, M. Zoli,
L. F. Agnati, P. Vernier, C. Lluis, S. Ferre, K. Fuxe and
R. Franco, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 8606–8611.
4 M. Rocheville, D. C. Lange, U. Kumar, S. C. Patel, R. C. Patel
and Y. C. Patel, Science, 2000, 288, 154–157.
29 S. P. Lee, B. F. O'Dowd, R. D. Rajaram, T. Nguyen and
S. R. George, Biochemistry, 2003, 42, 11023–11031.
30 Y. Liang, D. Fotiadis, S. Filipek, D. A. Saperstein,
K. Palczewski and A. Engel, J. Biol. Chem., 2003, 278,
21655–21662.
5 J. Shonberg, P. J. Scammells and B. Capuano,
ChemMedChem, 2011, 6, 963–974.
´
31 A. Soriano, R. Ventura, A. Molero, R. Hoen, V. Casado,
´
´
A. Cortes, F. Fanelli, F. Albericio, C. Lluiıs, R. Franco and
6 I. Berque-Bestel, F. Lezoulac'h and R. Jockers, Curr. Drug
Discovery Technol., 2008, 5, 312.
M. Royo, J. Med. Chem., 2009, 52, 5590–5602.
32 J. A. Ballesteros and H. Weinstein, Methods Neurosci., 1995,
25, 366–428.
7 D. Huber, H. Hubner and P. Gmeiner, J. Med. Chem., 2009,
52, 6860–6870.
33 S. G. F. Rasmussen, B. T. DeVree, Y. Zou, A. C. Kruse,
K. Y. Chung, T. S. Kobilka, F. S. Thian, P. S. Chae,
E. Pardon, D. Calinski, J. M. Mathiesen, S. T. A. Shah,
J. A. Lyons, M. Caffrey, S. H. Gellman, J. Steyaert,
G. Skiniotis, W. I. Weis, R. K. Sunahara and B. K. Kobilka,
Nature, 2011, 477, 549–555.
8 D. Huber, S. Loeber, H. Huebner and P. Gmeiner, Bioorg.
Med. Chem., 2012, 20, 455–466.
9 J. Kuehhorn, A. Goetz, H. Huebner, D. Thompson, J. Whistler
and P. Gmeiner, J. Med. Chem., 2011, 54, 7911–7919.
¨
¨
10 J. Kuhhorn, H. Hubner and P. Gmeiner, J. Med. Chem., 2011,
54, 4896–4903.
898 | Med. Chem. Commun., 2014, 5, 891–898
This journal is © The Royal Society of Chemistry 2014