Fiorentini C, Busi C, Gorruso E, Gotti C, Spano P, Missale CReciprocal regulation of dopamine D1 and D3 receptor function and trafficking by heterodimerization. Mol Pharmacol 74:59-69

Abstract

Colocalization of dopamine D1 (D1R) and D3 receptors (D3R) in specific neuronal populations suggests that their functional cross-talk might involve direct interactions. Here we report that the D1R coimmunoprecipitates with the D3R from striatal protein preparations, suggesting that they are clustered together in this region. Using bioluminescence resonance energy transfer (BRET(2)), we further suggest the existence of a physical interaction between D1R and D3R. Tagged D1R and D3R cotransfected in human embryonic kidney (HEK) 293 cells generated a significant BRET(2) signal that was insensitive to agonist stimulation, suggesting that they form a constitutive heterodimer. D1R and D3R regulate adenylyl cyclase (AC) in opposite ways. In HEK 293 cells coexpressing D1R and D3R, dopamine stimulated AC with higher potency and displaced [3H]R-(+)-7-chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine (SCH23390) binding with higher affinity than in cells expressing the D1R. In HEK 293 cells individually expressing D1R or D3R, agonist stimulation induces internalization of D1R but not of D3R. Heterodimerization with D3R abolishes agonist-induced D1R cytoplasmic sequestration induced by selective D1R agonists and enables internalization of the D1R/D3R complex in response to the paired stimulation of both D1R and D3R. This mechanism involves beta-arrestin binding because it was blocked by mutant beta-arrestinV53D. These data suggest that as a result of dimerization, the D3R is switched to the desensitization mechanisms typical of the D1R. These data give a novel insight into how D1R and D3R may function in an integrated way, providing a molecular mechanism by which to converge D1R- and D3R-related dysfunctions.

Full text

Reciprocal Regulation of Dopamine D1 and D3 Receptor
Function and Trafficking by Heterodimerization
Chiara Fiorentini, Chiara Busi, Emanuela Gorruso, Cecilia Gotti, PierFranco Spano,
and Cristina Missale
Section of Pharmacology, Department of Biomedical Sciences and Biotechnology (C.F., C.B., E.G., P.F.S., C.M.),
and Centre of Excellence on Diagnostic and Therapeutic Innovation (P.F.S., C.M.), University of Brescia, Brescia, Italy;
and Consiglio Nazionale delle Ricerche Institute of Neuroscience, Milano, Italy (C.G.)
Received November 27, 2007; accepted April 17, 2008
ABSTRACT
Colocalization of dopamine D1 (D1R) and D3 receptors (D3R) in
specific neuronal populations suggests that their functional
cross-talk might involve direct interactions. Here we report that
the D1R coimmunoprecipitates with the D3R from striatal pro-
tein preparations, suggesting that they are clustered together in
this region. Using bioluminescence resonance energy transfer
(BRET
2
), we further suggest the existence of a physical inter-
action between D1R and D3R. Tagged D1R and D3R cotrans-
fected in human embryonic kidney (HEK) 293 cells generated a
significant BRET
2
signal that was insensitive to agonist stimu-
lation, suggesting that they form a constitutive heterodimer.
D1R and D3R regulate adenylyl cyclase (AC) in opposite ways.
In HEK 293 cells coexpressing D1R and D3R, dopamine stim-
ulated AC with higher potency and displaced [
3
H]R-(⫹)-7-
chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-
benzazepine (SCH23390) binding with higher affinity than in
cells expressing the D1R. In HEK 293 cells individually express-
ing D1R or D3R, agonist stimulation induces internalization of
D1R but not of D3R. Heterodimerization with D3R abolishes
agonist-induced D1R cytoplasmic sequestration induced by
selective D1R agonists and enables internalization of the D1R/
D3R complex in response to the paired stimulation of both D1R
and D3R. This mechanism involves
␤
-arrestin binding because
it was blocked by mutant
␤
-arrestinV53D. These data suggest
that as a result of dimerization, the D3R is switched to the
desensitization mechanisms typical of the D1R. These data
give a novel insight into how D1R and D3R may function in an
integrated way, providing a molecular mechanism by which to
converge D1R- and D3R-related dysfunctions.
Dopamine (DA) controls various physiological functions,
including locomotor activity, learning and memory, and mo-
tivation and reward; dopaminergic dysfunctions have been
implicated in the development of Parkinson’s disease, schizo-
phrenia, and drug abuse. DA acts through five receptors,
belonging to the G protein-coupled receptor (GPCR) family,
that are divided into D1-like (D1 and D5) and D2-like (D2,
D3, and D4) subtypes. Each receptor displays unique prop-
erties, including affinity for DA and specificity for G protein
coupling and signaling and shows a peculiar neuronal distri-
bution (Missale et al., 1998). The D1 receptor (D1R) is the
most abundant and widespread DA receptor in the brain,
where it is found at high density in both motor and limbic
areas (Missale et al., 1998). The D3 receptor (D3R) is less
abundant and exhibits a more restricted pattern of distribu-
tion with high concentrations in the ventral striatum, par-
ticularly in the shell of the nucleus accumbens and islands of
Calleja (Sokoloff et al., 1990; Le´vesque et al., 1992) and lower
expression in other brain regions (Sokoloff et al., 1990;
Le´vesque et al., 1992; Schwartz et al., 1998). Both D1R and
D3R have been implicated in the regulation of rewarding
mechanisms and motivated behavior and in the modulation
of emotional and cognitive processes (Xu et al., 1997;
This study was supported by Ministero dell’Istruzione, dell’Universita`e
della Ricerca and University of Brescia (PRIN 2006054175) and partially by
Consiglio Nazionale delle Ricerche and Ministero dell’Istruzione,
dell’Universita` e della Ricerca – Fondo Integrativo Speciale per la Ricerca (to
C.M.) and partially by the Cariplo Foundation (grant 2006/0882/104878 to F.C.
and grant 2006/0779/109251 to C.G.).
C.F. and C.B. contributed equally to this work.
Article, publication date, and citation information can be found at
http://molpharm.aspetjournals.org.
doi:10.1124/mol.107.043885.
ABBREVIATIONS: DA, dopamine; GFP, green fluorescent protein; Rluc, Renilla reniformis luciferase; GPCR, G protein-coupled receptor; GRK,
G protein-coupled receptor kinase; PAGE, polyacrylamide gel electrophoresis; BRET
2
, bioluminescence resonance energy transfer; D1R, D1
receptor, D3R, D3 receptor; PBS, phosphate-buffered saline; HEK, human embryonic kidney; HRP, horseradish peroxidase; IP, immunoprecipi-
tation; WB, Western blot; AC, adenylyl cyclase; L-DOPA, 3,4-dihydroxy-L-phenylalanine; LID, 3,4-dihydroxy-L-phenylalanine-induced dyskinesia;
SCH23390, R-(⫹)-7-chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine; HA, hemoagglutinin; buffer A, NaCl, EDTA,
Na
2
HPO
4
, Nonidet P-40, and SDS; SKF 81297, (⫾)-6-chloro-7,8-dihydroxy-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine hydrobromide.
0026-895X/08/7401-59–69$20.00
MOLECULAR PHARMACOLOGY Vol. 74, No. 1
Copyright © 2008 The American Society for Pharmacology and Experimental Therapeutics 43885/3355430
Mol Pharmacol 74:59–69, 2008 Printed in U.S.A.
59
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

Schwartz et al., 1998; Karasinska et al., 2000, 2005). More-
over, both alterations of D1R function (Aubert et al., 2005)
and overexpression of D3R in the dorsal striatum have been
related to the development of motor dysfunctions (Bordet et
al., 2000; Guillin et al., 2001; Be´zard et al., 2003).
Biochemical and behavioral evidence suggests that D1R
and D3R may functionally interact. For example, D1R stim-
ulation induces D3R mRNA expression in rat striatum and
medulloblastoma cells (Levavi-Sivan et al., 1998; Bordet et
al., 2000), and coactivation of D1R and D3R in the shell of
nucleus accumbens synergistically enhances substance P
gene expression (Ridray et al., 1998; Schwartz et al., 1998).
Moreover, D3R-deficient mice exhibit increased behavioral
sensitivity to the stimulation of D1R and D2R (Xu et al.,
1997) and decreased D1R-induced c-fos expression (Jung and
Schmauss, 1999); furthermore, D1R and D3R interactions
are apparently involved in the rewarding properties of low
doses of cocaine and in cocaine-mediated inhibition of cAMP
response element-binding protein phosphorylation (Karasin-
ska et al., 2000, 2005). The cross-talk between D1R and D3R
could occur either at the level of neuronal networks or within
the same neuron. This latter type of interaction is supported
by the observation that D1R and D3R mRNAs are colocalized
in a large number of neurons within the shell of the nucleus
accumbens (Le Moine and Bloch, 1996; Ridray et al., 1998;
Schwartz et al., 1998) and the striatum (Surmeier et al.,
1996) and that L-DOPA administration to hemiparkinsonian
rats induces the overexpression of D3R in striatonigral neu-
rons that constitutively express the D1R (Bordet et al., 2000;
Guillin et al., 2001). Interaction between D1R and D3R in
single neurons might involve either the convergence of their
signaling pathways or the formation of heterodimeric com-
plexes. It has been shown, in fact, that a general property of
GPCR is to form heterodimeric receptor complexes with pe-
culiar pharmacological, signaling, and trafficking character-
istics (Angers et al., 2002), suggesting that receptor het-
erodimerization may represent a new integrative mechanism
at the synaptic level. On this line, it has been shown that the
D3R directly interacts with the D2R (Scarselli et al., 2001)
and with the adenosine A2AR (Torvinen et al., 2005), and
that the D1R interacts with the D2R (Rashid et al., 2007),
with the adenosine A1R (Gine´s et al., 2000), and with the
glutamate N-methyl-D-aspartate receptor (Lee et al., 2002;
Fiorentini et al., 2003; Scott et al., 2006), and that the for-
mation of these novel signaling units may represent the
molecular basis for the functional interactions between these
receptors.
The aim of this study was to investigate whether D1R and
D3R may form a heterodimeric receptor complex and to de-
fine the functional properties of this complex. The results
show that D1R and D3R directly interact in both striatal
membranes and cotransfected cells and that this interaction
influences D1R coupling to adenylyl cyclase and the adaptive
responses of both D1R and D3R to agonist stimulation.
Materials and Methods
Materials. Human embryonic kidney (HEK) 293 cells were pro-
vided by Deutsche Sammlung von Mikroorganismen und Zellcul-
turen GmbH (Braunschweg, Germany). Tissue culture media and
fetal bovine serum were purchased from Euroclone Celbio (Milano,
Italy). SKF 81297, quinpirole, and SCH23390 were purchased from
Tocris (Bristol, UK); dopamine, (⫺)-sulpiride, and the monoclonal
anti-D1R antibody (clone 1-1-F11-S.E6) were purchased from Sigma
(Milano, Italy). The anti-D3R antibody and the horseradish peroxi-
dase (HRP)-conjugated secondary antibodies were from Santa Cruz
(Santa Cruz Biotechnology Inc., Heidelberg, Germany). The anti-
hemoagglutinin (HA) antibody was from Sigma, and the anti-GFP
antibody was from Invitrogen (Carlsbad, CA). The Cy3-labeled sec-
ondary antibody was purchased from Jackson Immunoresearch Lab-
oratories, Inc. (West Grove, PA). [
3
H]Sulpiride (78.2 Ci/mmol),
[
3
H]SCH23390 (86 Ci/mmol), and [
3
H]raclopride (62.2 Ci/mmol) were
from PerkinElmer Life and Analytical Sciences (Waltham, MA).
Human D3R-GFP, human D3R, human D1R,
␤
-arrestin-1V53D, and
dynamin-IK44A were kindly provided by Dr. Marc Caron (Duke
University, Durham, NC); the ChemR23 chemokine receptor was
kindly provided by Dr. Silvano Sozzani (University of Brescia, Bres-
cia, Italy).
Generation of a Rabbit Anti-D1R Polyclonal Antibody. A
polyclonal antibody directed to the peptide GSSEDLKKEEAG-
GIAKPLEKLS, corresponding to the rat D1 receptor (D1R) amino
acids 396 to 417 (anti-D1R822), was produced in rabbits and was
affinity-purified as described previously (Vailati et al., 1999). The
sequence used does not match with the other DA receptor subtypes.
Protein Preparation, Immunoprecipitation, and Western
Blot. The rat striatum was homogenized with a glass-glass homog-
enizer in ice-cold 10 mM Tris-HCl containing 5 mM EDTA and a
complete set of protease inhibitors (Roche, Milano, Italy), pH 7.4, and
was centrifuged at 700gat 4°C for 10 min. The resulting supernatant
containing the total cell proteins was added with 1% SDS and stored
at ⫺80°C. To isolate the membrane fraction, the striatum was ho-
mogenized in 5 mM Tris-HCl containing 2 mM EDTA and a mixture
of protease inhibitors, pH 7.8, and was centrifuged at 80gfor 10 min
to pellet unbroken cells and nuclei. The supernatant was centrifuged
at 30,000gfor 20 min at 4°C to pellet the membrane fraction. Protein
concentration was determined by using the DC Protein Assay Re-
agent (Bio-Rad, Milano, Italy). To detect the D1R, 60
␮
g of protein
preparations was resolved by SDS-PAGE, transferred onto nitrocel-
lulose membranes, and blotted for1hatroom temperature in Tris-
buffered saline containing 0.1% Tween 20 and 5% nonfat powdered
milk. Membranes were incubated overnight at 4°C with the anti-
D1R822 antibody (1:700 dilution) or the anti-D3R antibody (1:200
dilution). Detection was performed by chemiluminescence (Chemi-
Lucent; Chemicon, Milano, Italy) with HRP-conjugated secondary
antibodies (1:3000 dilution). In the immunoprecipitation (IP) exper-
iments, 60
␮
g of striatal protein preparations were incubated over-
night at 4°C with either the anti-D1R822 antibody (1:50 dilution) or
the anti-D3R antibody (1:50 dilution) in 200 mM NaCl, 10 mM
EDTA, 10 mM Na
2
HPO
4
, 0.5% Nonidet P-40, and 0.1% SDS (buffer
A). Protein A-agarose beads were added, and incubation was contin-
uedfor2hatroom temperature. The beads were collected and
extensively washed with buffer A. The resulting proteins were re-
solved by SDS-PAGE, transferred onto nitrocellulose membranes,
and blotted for1hatroom temperature in Tris-buffered saline
containing 0.1% Tween 20 and 5% nonfat powdered milk. Mem-
branes were incubated overnight at 4°C with the anti-D3R antibody
(1:200 dilution) or the anti-D1R822 antibody (1:700 dilution). Detec-
tion was performed by chemiluminescence with HRP-conjugated sec-
ondary antibodies (1:3000 dilution). In another series of experi-
ments, HEK 293 cells were transfected with HA-tagged D1R and
GFP-tagged D3R. Total cell proteins were immunoprecipitated with
either the anti-HA (1:200 dilution) or the anti-GFP (1:200 dilution)
antibody, and the resulting proteins were immunoreacted with ei-
ther the anti-GFP (1:500 dilution) or the anti-HA (1:500 dilution)
antibody, respectively, and detected as described above.
Generation of Bioluminescence Resonance Energy Trans-
fer Fusion Constructs. The preparation of the D1R-luciferase con-
struct (D1R-Rluc) was described previously (Fiorentini et al., 2003).
The coding sequence of human D3R was amplified out of its original
vector using primers containing unique HindIII and BglII sites and
60 Fiorentini et al.
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

the native Pfu DNA polymerase (Stratagene, Milano, Italy) to gen-
erate a stop codon-free fragment. This fragment was cloned in-frame
into the pGFP
2
-N2(h) vector containing the green fluorescent protein
(GFP
2
) (PerkinElmer) to generate the plasmid D3R-GFP
2
. The cod-
ing sequence of ChemR23 receptor was amplified out of its original
vector using primers containing unique BamHI and EcoRI sites and
the native Pfu DNA polymerase (Stratagene) to generate a stop
codon-free fragment that was cloned into the pGFP
2
vector to gen-
erate the plasmid ChemR23-GFP
2
.
Cell Culture, Transfection, and Bioluminescence Reso-
nance Energy Transfer Assay. HEK 293 cells were cultured in
Dulbecco’s modified Eagle’s medium containing 10% fetal bovine
serum, 2 mM glutamine, 0,1 mM nonessential amino acids, 1 mM
sodium pyruvate, 100 U/ml penicillin, and 100
␮
g/ml streptomycin.
Semiconfluent cells were cotransfected with D1R-Rluc (0.2
␮
g) and
increasing concentrations of either D3R-GFP
2
(0.2–2
␮
g) or
ChemR23-GFP
2
(0.2–2
␮
g) using the LipofectAMINE 2000 reagent
(Invitrogen) according to the manufacturer’s instructions. The total
amount of DNA was kept at 2.2
␮
g. In competition experiments, cells
were transfected with D1R-Rluc (0.1
␮
g) and D3R-GFP
2
(0.5
␮
g) in
the absence or presence of different amounts of either untagged
pcDNA-D1R (0.1–1.5
␮
g) or untagged pcDNA-D3R (0.1–1.5
␮
g) or
pcDNA-D2R (0.1–1.5
␮
g) or pcDNA-ChemR23 (0.1–1.5
␮
g). Twenty-
four hours after transfection, cells were harvested, centrifuged, and
resuspended in PBS containing 0.1 mg/ml CaCl
2
, 0.1 mg/ml MgCl
2
,
and 1 mg/ml D-glucose. Approximately 15,000 cells/well were distrib-
uted in a 96-well microplate (white Optiplate; PerkinElmer). Deep-
BlueC coelenterazine (PerkinElmer) was added at the final concen-
tration of 5
␮
M, and bioluminescence resonance energy transfer
(BRET
2
) signals were determined using a Fusion universal micro-
plate analyzer (PerkinElmer), which allows sequential integration of
signals detected at 390/400 and 505/510 nm. To define the D1R-Rluc/
D3R-GFP
2
expression ratio in each sample, HEK 293 cells trans-
fected with increasing amounts of either D1R-Rluc or D3R-GFP
2
or
ChemR23-GFP
2
were evaluated for total luminescence or total fluo-
rescence and for D1R-Rluc or D3R-GFP
2
or ChemR23-GFP
2
protein
level expression. D1R-Rluc and D3R-GFP
2
levels were determined by
radioreceptor binding with [
3
H]SCH23390 and [
3
H]raclopide, respec-
tively. ChemR23-GFP
2
levels were determined by flow cytometry. In
brief, cells were labeled using an anti-ChemR23 monoclonal antibody
(IgG3; R&D System Inc., Minneapolis, MN) or an isotype control
(mouse IgG3; Biolegend, San Diego, CA) followed by a goat anti-
mouse-PE secondary antibody (Invitrogen). Samples were acquired
on a Pas II (Partec GmbH, Mu¨nster, Germany) and analyzed using
FlowJo version 7.2 (Tree Star, Ashland, OR). Luminescence was
plotted against D1R-RLuc expression levels, and fluorescence was
plotted against D3R-GFP
2
or ChemR23-GFP
2
expression levels. Be-
cause the relationship between measured luminescence or fluores-
cence and the corresponding receptor was linear, the acceptor/donor
ratio was expressed as the fluorescence/luminescence ratio. To test
the effects of agonists, cells cotransfected with D1R-Rluc and D3R-
GFP
2
at the 1:5 ratio were distributed in a 96-well microplate and
incubated in the absence or in the presence of 1
␮
M SKF 81297, 1
␮
M
quinpirole, or 10
␮
M dopamine for 10 min at 37°C. DeepBlueC
coelenterazine (5
␮
M) was added, and BRET
2
signals were deter-
mined as described previously. Untransfected cells and cells individ-
ually transfected with D1R-Rluc or D3R-GFP
2
were used to define
the nonspecific signals; cells transfected with the p-Rluc-GFP
2
con-
trol vector (PerkinElmer) were used as a positive controls. The
BRET
2
signal was calculated as [(emission at 505/510) ⫺(emission
at 390/400) ⫻Cf ]/(emission at 390/400), where Cf corresponds to
(emission at 505/510)/(emission at 390/400) for the D1R-Rluc ex-
pressed alone in the same experiment.
Generation of Cell Clones Stably Expressing the D1R, the
D3R, and both D1R and D3R. HEK 293 cells were transfected with
the D1R cDNA using the LipofectAMINE 2000 reagent according to
the manufacturer’s instructions (Invitrogen). Cell clones stably ex-
pressing D1R (HEK-D1R) were isolated by zeocin selection (100
␮
g/ml). HEK 293 cells were transfected with the D3R cDNA and
cultured in the presence of G418 (800
␮
g/ml) to select clones express-
ing the D3R (HEK-D3R). HEK-D1R cells, cultured in the standard
medium containing zeocin (100
␮
g/ml), were transfected with the
D3R cDNA, and cell clones stably expressing D1R and D3R (HEK-
D1R/D3R) were isolated by zeocin (100
␮
g/ml) and G418 (800
␮
g/ml)
selection. HEK-D1R cells were maintained in culture in the presence
of zeocin (100
␮
g/ml), HEK-D3R cells were cultured in the presence
of G418 (800
␮
g/ml), and HEK-D1R/D3R cells were cultured in the
presence of both zeocin (100
␮
g/ml) and G418 (800
␮
g/ml). Cell clones
expressing the D1R, the D3R, or both D1R and D3R were character-
ized for receptor levels in binding studies with [
3
H]SCH23390 and
[
3
H]sulpiride.
Receptor Sequestration and Recycling. HEK 293 and HEK-
D1R cells were transiently transfected with D3R-GFP (kindly pro-
vided by Dr. Marc Caron, Duke University, Durham, NC) in the
absence or in the presence of
␤
-arrestin-1V53D using the Lipo-
fectAMINE 2000 reagent. Cells expressing D1R or D3R-GFP or both
D1R and D3R-GFP were incubated for 5 to 60 min at 37°C with 1) the
D1R agonist SKF 81297 (10 nM to 10
␮
M); 2) the D3R agonist
quinpirole (0.5 nM to 1
␮
M); 3) a combination of 1
␮
M SKF 81297 and
1
␮
M quinpirole; 4) DA (100 nM to 10
␮
M); and 5) 1
␮
MDAinthe
presence of either the D1R antagonist SCH 23390 (1
␮
M) or the D3R
antagonist (⫺)sulpiride (1
␮
M). To study receptor recycling to the
plasma membrane, cells were exposed to a combination of 1
␮
M SKF
81297 and 1
␮
M quinpirole for 60 min at 37°C to promote seques-
tration. Agonists were removed by extensive washes with ice-cold
PBS, and cells were incubated in the standard medium at 37°C for 5
to 60 min. Receptor sequestration and recycling to the plasma mem-
brane were evaluated by both immunofluorescence and radiorecep-
tor binding.
Immunofluorescence. Cells expressing D1R and D3R-GFP were
fixed in 4% paraformaldehyde for 20 min at room temperature and
permeabilized with 0.1% Triton X-100 in PBS containing 5% bovine
serum albumin and 5% normal goat serum. Cells were incubated
with the monoclonal rat anti-D1R antibody (Sigma; 1:800 dilution in
PBS containing 1% normal goat serum) overnight at 4°C and then
with the Cy3-conjugated secondary antibody (1:1000 dilution) for 45
min at room temperature. The immunolabeled cells were recorded
with a fluorescence microscope (IX51; Olympus, Tokyo, Japan) at a
100⫻magnification. Nontransfected cells and omission of the pri-
mary antibody were used as negative controls.
[
3
H]Sulpiride Binding in Intact Cells. Sequestration of D3R
was measured according to Kim et al. (2001) exploiting the hydro-
philic properties of [
3
H]sulpiride. HEK-D3R and HEK-D1R/D3R cells
were plated at the density of 2 ⫻10
5
cells/well in 24-well plates,
allowed to recover for 24 h, and stimulated with agonists as de-
scribed previously. Incubation was blocked by cooling plates on ice
and extensively washing cells with ice-cold serum-free medium con-
taining 20 mM HEPES, pH 7.4. Intact cells were incubated at 4°C for
150 min with [
3
H]sulpiride at the final concentration of 2.2 nM. The
nonspecific binding was defined with either 10
␮
M(⫺)sulpiride or 10
␮
M haloperidol. The incubation was stopped by three washes with
the same medium, and 1% Triton X-100 was added. The amount of
radioactivity in each sample was determined on a liquid scintillation
analyzer.
Membrane Preparation and Radioreceptor Binding. Trans-
fected HEK 293 cells were rinsed, harvested, and centrifuged at 100g
for 10 min. Cells were homogenized with an Ultra Turrex homoge-
nizer in 5 mM Tris-HCl containing 2 mM EDTA and a mixture of
protease inhibitors, pH 7.8, and centrifuged at 80gfor 10 min. The
supernatant was centrifuged at 30,000gfor 20 min at 4°C, and the
resulting pellet, containing total cell membranes, was resuspended
in 50 mM Tris-HCl containing 5 mM MgCl
2
, 1 mM EGTA, and the
protease inhibitors, pH 7.8, layered on a 35% sucrose cushion and
centrifuged at 150,000gfor 90 min to separate the light vesicular and
heavy membrane fractions as described previously (Fiorentini et al.,
2003). The heavy fraction, at the bottom of the sucrose cushion, was
D1 and D3 Receptor Heterodimerization 61
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

resuspended in 50 mM Tris-HCl containing 5 mM EDTA, 1.5 mM
CaCl
2
, 5 mM MgCl
2
, 5 mM KCl, and 120 mM NaCl, pH 7.4, and used
for binding assay. Protein concentration was determined by using
the DC Protein Assay Reagent (Bio-Rad). Aliquots of membrane
suspension (50
␮
g of protein/sample) were incubated at room tem-
perature for 90 min with a saturating concentration (4 nM) of
[
3
H]SCH23390. The nonspecific binding was defined with 1
␮
M
d-butaclamol. To define the K
d
and B
max
of D1R and D3R and
D1R-RLuc and D3R-GFP
2
in HEK 293 cells, aliquots of total cell
membranes (50
␮
g protein/sample) were incubated with increasing
concentrations of [
3
H]SCH23390 (0.05–2.5 nM) or increasing concen-
trations of [
3
H]raclopride (0.5–7.5 nM) for 30 min at 37°C. The
nonspecific binding was defined with 1
␮
Md-butaclamol in the case
of [
3
H]SCH 23390 and with 1
␮
M(⫺)sulpiride in the case of [
3
H]-
raclopride. The reactions were stopped by rapid filtration under
reduced pressure through Whatman GF/C filters (Whatman, Clifton,
NJ).
Measurement of Adenylyl Cyclase Activity. Cells were ho-
mogenized in ice-cold 10 mM Tris-maleate, pH 7.4, containing 1.2
mM EGTA. Adenylyl cyclase (AC) activity was assayed in a 500-
␮
l
reaction mixture containing 80 mM Tris-maleate, 16 mM MgSO
4
, 0.5
mM 3–3-isobutyl-1-methylxanthine, 0.6 mM EGTA, 0.02% ascorbic
acid, pH 7.4, 2 mM ATP, 5 mM phosphocreatine, 50 U/ml creatine
phosphokinase, and various concentrations of DA (10 nM to 10
␮
M)
in the absence and presence of selective D1R (1
␮
M SCH23390) or
D3R [1
␮
M(⫺)sulpiride] antagonists. The reaction was started by
adding the cell homogenate (approximately 1.5
␮
g protein/sample),
and incubation was carried out at 30°C for 20 min and stopped by
placing samples in boiling water for 5 min. The cAMP in the super-
natant was measured by radioimmunoassay using the reagents sup-
plied by PerkinElmer.
Results
Development and Characterization of the Anti-D1R
Polyclonal Antibody. As shown in Fig. 1A (lanes 1 and 2),
the affinity-purified anti-D1R822 antibody (1.5
␮
g/ml) de-
tected a single band of ⬃70 kDa in Western blot (WB) exper-
iments with membrane preparations from rat striatum.
When the anti-D1R822 antibody was preabsorbed with an
excess of its specific immunizing peptide (80
␮
g/ml), the
signal corresponding to the ⬃70-kDa species was lost (Fig.
1A, lanes 3 and 4), suggesting that the immunoreactive band
is specific. It has been reported that the mature D1R in the
striatum is a glycosylated protein with a molecular size of
⬃72 kDa and that deglycosylation results in the appearance
of low molecular mass forms of ⬃60 and ⬃48 kDa (Amlaiky et
al., 1987; Jarvie et al., 1989). The ⬃70-kDa band recognized
by our antibody thus probably represents the fully glycosy-
lated form of the D1R. The anti-D1R822 antibody was also
tested by WB in different rat brain areas, characterized by
specific D1R expression. The immunoblot reported in Fig. 1B
shows that a major ⬃70-kDa band was present in mem-
branes from the striatum (lane 1), hippocampus (lane 2),
cerebellum (lane 3), and prefrontal cortex (lane 4). The in-
tensity of this signal was stronger in the striatum, hippocam-
pus, and prefrontal cortex, which express high levels of D1R,
than in the cerebellum, where the D1R is poorly expressed
(Missale et al., 1998). Moreover, no specific immunoreactivity
was detected by the anti-D1R822 antibody in the anterior
pituitary (lane 5), suggesting that this antibody does not
cross-react with the D2R, which is highly concentrated in
this region (Missale et al., 1998). To further evaluate the
specificity of this antibody, striatal proteins were immuno-
precipitated by a commercial anti-D1R antibody (anti-D1R
H-109; Santa Cruz Biotechnology), and the resulting mate-
rial was subjected to SDS-PAGE and blotted with the anti-
D1R822 antibody. As shown in Fig. 1C, the ⬃70-kDa band,
corresponding to the D1R, was detected by the anti-D1R822
antibody in both striatal proteins (lane 1) and in striatal
proteins immunoprecipitated by the anti-D1R H-109 anti-
body (lane 3). This band was undetectable when the precip-
itating antibody was omitted (lane 2). Moreover, as reported
in Fig. 1D, the anti-D1R822 antibody (6
␮
g/ml) immunopre-
cipitated a ⬃70-kDa species from striatal membranes that
was recognized by the anti-D1R H-109 antibody (lane 2),
further confirming the specificity of our antibody. Taken
together, these data support the selectivity of the anti-
D1R822 antibody for the D1R and suggest that it represents
a useful tool in both IP and WB assays. This antibody was
thus used in subsequent experiments.
D1R and D3R Coimmunoprecipitate from Striatal
Membranes and Transfected Cells. CoIP studies were
performed to determine whether D1R and D3R may directly
interact in rat striatum. As shown in Fig. 2A, incubation of
striatal proteins with the anti-D3R antibody immunoprecipi-
tated a ⬃70-kDa species that was recognized by the anti-D1R
antibody (lane 2) and was absent when the immunoprecipi-
tating antibody was omitted (lane 1). Moreover, according to
Nimchinsky et al. (1997), two major bands between ⬃60 and
⬃75 kDa, which were detected by the anti-D3R antibody,
were present in striatal proteins immunoprecipitated with
the anti-D1R antibody (Fig. 2B, lane 2). These species were
undetectable when the immunoprecipitating antibody was
omitted (lane 1). Taken together, these data indicate that a
significant proportion of striatal D1R and D3R might physi-
cally interact. To investigate whether D1R and D3R are
assembled into a complex also in transfected cell systems and
to exclude the possibility of artifacts generated by the recep-
Fig. 1. Characterization of the anti-D1R822 antibody. A, WB analysis in
striatal membranes. Increasing amounts of striatal proteins were immu-
noblotted with either the anti-D1R822 antibody (lanes 1 and 2) or the
anti-D1R822 antibody preabsorbed with an excess of its specific immu-
nizing peptide (lanes 3 and 4). B, WB analysis in different rat brain areas.
C and D, characterization of the anti-D1R822 antibody by IP. C, striatal
proteins were immunoprecipitated with a commercial anti-D1R antibody
(H-109; Santa Cruz Biotechnology), and the resulting proteins were im-
munoblotted with the anti-D1R822 antibody. A single specific band of
approximately 70 kDa was detected in the immunoprecipitated material.
D, striatal proteins were immunoprecipitated with the anti-D1R822 an-
tibody, and the resulting proteins were immunoblotted with the commer-
cial anti-D1R antibody. Each experiment was repeated four times. Rep-
resentative blots are shown.
62 Fiorentini et al.
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

tor-specific antibodies, HEK 293 cells were cotransfected
with HA-tagged D1R and GFP-tagged D3R, and proteins
were immunoprecipitated with anti-HA or anti-GFP antibod-
ies and revealed with anti-GFP or anti-HA antibodies, re-
spectively. As reported in Fig. 2C, a ⬃70-kDa species, corre-
sponding to HA-tagged D1R, was detectable in proteins
immunoprecipitated with the anti-GFP antibody and re-
vealed with the anti-HA antibody and two major bands be-
tween ⬃80 and ⬃100 kDa, corresponding to GFP-tagged
D3R, were detected in proteins immunoprecipitated with the
anti-HA antibody and revealed with the anti-GFP antibody.
D1R and D3R Constitutively Interact in Living Cells.
Although coimmunoprecipitation is a generally accepted
method to document protein-protein interactions, in the case
of membrane receptors, the interpretation of these experi-
ments may be complicated by detergent solubilization that
could promote artifactual aggregation. To asses whether
D1R/D3R heterodimers could be detected in living cells we
used the BRET
2
method, which detects energy transfer from
a luminescent donor to a fluorescent acceptor when they are
less than 50 to 80 Å apart. For this purpose, the D1R was
fused on its C terminus with the Renilla reniformis luciferase
(D1R-Rluc) and the D3R with the GFP
2
(D3R-GFP
2
). To
ensure that the expressed fusion proteins were properly
folded polypeptides capable of binding selective dopaminergic
ligands, we assessed their binding properties by saturation
binding assays with [
3
H]SCH23390 (D1R) or [
3
H]raclopride
(D3R). The obtained K
d
values were as follows: K
d
⫽1.2 ⫾0.1
and 1.0 ⫾0.06 nM for D1R and D1R-Rluc, respectively, and
K
d
⫽2.02 ⫾0.1 and 1.5 ⫾0.08 nM for D3R and D3R-GFP
2
,
respectively. BRET
2
signals were determined in HEK 293
cells simultaneously or individually expressing the D1R-Rluc
and D3R-GFP
2
constructs as described previously (Fiorentini
et al., 2003). Cells expressing a fusion construct covalently
linking Rluc to GFP
2
(pRluc-GFP
2
) were used as a positive
control. As shown in Fig. 3A, D1R-Rluc expressed in HEK
293 cells generated a small, nonspecific BRET
2
signal; like-
wise, no BRET
2
was observed in cells expressing the D3R-
GFP
2
construct. A significant BRET
2
signal was observed in
cells expressing the pRluc-GFP
2
construct (Fig. 3A), confirm-
ing the importance of molecular proximity between the
Fig. 2. Coimmunoprecipitation of D1R and D3R in rat striatum and
transfected HEK 293 cells. A, representative coIP of D1R from striatal
proteins by the anti-D3R antibody (lane 2) but not by omission of the
precipitating antibody (lane 1). B, representative coIP of D3R from stri-
atal proteins by the anti-D1R antibody (lane 2) but not by omission of the
precipitating antibody (lane 1). Sixty micrograms of striatal proteins was
used in each IP that was repeated four times. C and D, HEK 293 cells
were cotransfected with HA-tagged D1R and GFP-tagged D3R, and total
proteins were either IP with the anti-GFP antibody and revealed with the
anti-HA antibody (C) or IP with the anti-HA antibody and revealed with
the anti-GFP antibody (D). Data are representative of three experiments.
Fig. 3. Detection of D1R and D3R interaction by BRET
2
in transfected
HEK 293 cells. The D1R fused to R. reniformis luciferase (D1R-Rluc) and
the D3R fused to GFP
2
(D3R-GFP
2
) were transfected either individually
or simultaneously in HEK 293 cells. The DeepBlueC coelenterazine sub-
strate was added at a final concentration of 5
␮
M. A, quantification of
BRET
2
data from a series of control experiments with single receptor
constructs (D1R-Rluc, D3R-GFP
2
, and pRluc-GFP
2
) or with D1R-Rluc
and D3R-GFP
2
coexpressed in the same cells (D1R-Rluc/D3R-GFP
2
)or
with cells individually expressing D1R-Rluc and D3R-GFP
2
mixed to-
gether before the BRET
2
experiment (D1R-Rluc ⫹D3R-GFP
2
). Bars are
the means ⫾S.E. of five experiments. ⴱ,p⬍0.001 versus D1R-Rluc,
Student’s ttest. B, BRET
2
titration analysis. Cells were transfected with
D1R-Rluc in the presence of increasing concentrations of either D3R-
GFP
2
or ChemR23-GFP
2
. BRET
2
, total luminescence, and total fluores-
cence were determined. BRET
2
ratio values are plotted as a function of
the total fluorescence/total luminescence ratio. Data are representative of
three experiments. The curves were fitted using a nonlinear regression
equation assuming a single binding site (GraphPad Prism 4). F, specific
BRET
2
signal generated in the presence of D3R-GFP
2
;E, nonspecific
signals generated in the presence of ChemR23-GFP
2
. C, BRET
2
compe-
tition analysis. Experiments were carried out at the constant D1R-Rluc/
D3R-GFP
2
ratio of 3.4 ⫾0.5 in the absence or presence of increasing
concentrations of either untagged D1R (F), D3R (E), or D2R (f). The
expression of untagged receptors was determined by radioreceptor bind-
ing. Data are representative of three experiments. Inset, experiments
were carried out at the constant D1R-Rluc/D3R-GFP
2
ratio of 3.4 ⫾0.5 in
the absence or in the presence of increasing concentrations of untagged
ChemR23. The expression of ChemR23 in each sample was determined
by immunofluorescence and fluorescence-activated cell sorting analysis
and is expressed as Medial Fluorescence Intensity (MFI). Data are rep-
resentative of three experiments. D, cells transfected with D1R-Rluc and
D3R-GFP
2
were exposed to either 1
␮
M SKF 81297 or 1
␮
M quinpirole or
10
␮
M dopamine in the absence or in the presence of dynamin-IK44A.
Bars are the means ⫾S.E. of five independent experiments. ⴱ,p⬍0.005
versus untreated cells, Student’s ttest.
D1 and D3 Receptor Heterodimerization 63
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

BRET
2
partners for signal detection. Coexpression of D1R-
Rluc and D3R-GFP
2
yielded a significantly high BRET
2
sig-
nal that could be best explained with the formation of a
D1R/D3R complex. Energy transfer was in fact undetectable
when cells individually expressing D1R-Rluc and D3R-GFP
2
were mixed together before the BRET
2
analysis. The speci-
ficity of the BRET
2
signal was further confirmed in titration
and competition experiments. As shown in Fig. 3B, increas-
ing the amount of the D3R-GFP
2
acceptor in the presence of
a constant concentration of the D1R-Rluc donor resulted in a
hyperbolic increase of the BRET
2
signal as a function of
increasing D3R-GFP
2
/D1R-Rluc ratio, indicating specificity
of the interaction. The expression level of D1R-RLuc was
determined by measuring total luminescence, and the ex-
pression levels of D3R-GFP
2
were monitored by measuring
total fluorescence. BRET
2
signals were plotted as a function
of the fluorescence/luminescence ratio. The chemokine
ChemR23 receptor (Wittamer et al., 2003) was used as a
negative control. As reported in Fig. 3B, increasing the con-
centration of ChemR23-GFP
2
in cells expressing the D1R-
RLuc resulted in a nonspecific linear increase of the BRET
2
signal. Figure 3C shows the results of BRET
2
competition
experiments carried out at the constant D1R-Rluc/D3R-GFP
2
ratio of 3.4 ⫾0.5 in the presence of increasing concentrations
of either untagged D1R, or untagged D3R or untagged D2R,
or untagged ChemR23. BRET
2
signals were plotted as a
function of receptor expression. Both D1R and D3R competed
with their tagged counterparts in a concentration-dependent
way, as shown by the decrease of the BRET
2
signal as a
function of the concentration of the untagged competitor.
Moreover the D2R, that has been shown to interact with both
D1R (Rashid et al., 2007) and D3R (Scarselli et al., 2001)
decreased the BRET
2
signal generated by D1R-RLuc and
D3R-GFP
2
in a concentration-dependent manner. By con-
trast, the ChemR23 receptor does not interfere with the
BRET
2
signal generation (Fig. 3C, insert). Taken together,
these data suggest that D1R-Rluc and D3R-GFP
2
are true
interacting partners. As shown in Fig. 3D, the BRET
2
signal
recorded in cells cotransfected with D1R-Rluc and D3R-GFP
2
was insensitive to stimulation by the D1R agonist SKF 81297
(1
␮
M) and the D3R agonist quinpirole (1
␮
M). However, DA
induced a slight but significant increase of BRET
2
signal. To
evaluate whether this effect of DA might reflect the modula-
tion of D1R-D3R interaction at the plasma membrane or
internalization of the D1R/D3R complex, cells were cotrans-
fected with the dynamin-IK44A dominant-negative mutant,
which blocks agonist-induced GPCR internalization (Zhang
et al., 1997). As shown in Fig. 3D, DA-mediated increase of
BRET
2
signal but not the basal BRET
2
signal in cells coex-
pressing D1R-RLuc and D3R-GFP
2
was inhibited by dy-
namin-IK44A, suggesting that it could probably reflect D1R/
D3R complex internalization. Taken together, these data
demonstrate a physical proximity between D1R-Rluc and
D3R-GFP
2
that is consistent with the formation of a consti-
tutive heterodimeric complex and strongly support the re-
sults obtained by coimmunoprecipitation in the striatum.
D1R/D3R Heterodimerization Increases the Potency
of DA in Stimulating Adenylyl Cyclase through the
D1R. We investigated whether D1R/D3R heterodimeriza-
tion affects D1R coupling to the AC signaling pathway. It is
known in fact that D1R stimulates AC, whereas the D3R
inhibits this effector in different cell systems (Missale et al.,
1998). For this purpose, HEK 293 cell clones stably express-
ing the D1R, the D3R, and both D1R and D3R were gener-
ated. D1R and D3R expression levels in the different cell
clones were determined by [
3
H]SCH23390 and [
3
H]sulpiride
binding. D1R expression in the different clones was as fol-
lows: HEK-D1R, 525 ⫾56 fmol/mg protein; HEK-D1R/D3R,
512 ⫾46 fmol/mg protein. D3R expression was as follows:
HEK-D3R, 610 ⫾58 fmol/mg protein; HEK-D1R/D3R, 714 ⫾
67 fmol/mg protein. The HEK-D1R and HEK-D1R/D3R
clones thus had the same level of D1R expression. Moreover,
in HEK-D1R/D3R cells, the expression of D3R was in slight
excess, thus ensuring that a relevant amount of D1R present
was associated with D3R. The effects of DA on cAMP forma-
tion are reported in Fig. 4A. DA dose-dependently stimulated
AC activity with an EC
50
value of 560 ⫾52 nM and a
maximal stimulation of 75 ⫾5% at the dose of 50
␮
Min
HEK-D1R cells. On the other hand, according to previous
studies showing that the D3R only inhibits type V AC, which
is poorly expressed in HEK 293 cells (Robinson and Caron,
1997), DA weakly inhibited cAMP formation in HEK-D3R
cells (maximal inhibition, 20 ⫾2%). In HEK-D1R/D3R cells,
however, DA stimulated AC activity with higher potency
Fig. 4. Effects of DA on cAMP formation in cells expressing the D1R/D3R
complex. HEK-D1R cells stably expressing the D1R (525 ⫾56 fmol/mg
protein) and HEK-D1R/D3R expressing both D1R (512 ⫾46 fmol/mg
protein) and D3R (714 ⫾67 fmol/mg protein) were used. A, DA stimulated
cAMP formation with an EC
50
value of 560 ⫾52 nM in HEK-D1R cells (F)
and with an EC
50
of 39 ⫾1.5 nMⴱin HEK-D1R/D3R (E); ⴱ,p⬍0.001
versus EC
50
in HEK-D1R cells, Student’s ttest. Data are expressed as the
percentage increase of cAMP over basal and represent the mean ⫾S.E. of
five independent experiments. B, effects of D1R and D3R antagonists on
DA-induced cAMP formation. HEK-D1R/D3R cell homogenates were
treated with DA in the absence (E) or presence of either the D3R antag-
onist (⫺)sulpiride (䡺) or the D1R antagonist SCH23390 (f). The calcu-
lated EC
50
values for DA-stimulated cAMP formation were 39 ⫾1.5 nM
and 520 ⫾47 nMⴱin the absence or in the presence of (⫺)sulpiride,
respectively (ⴱ,p⬍0.001, Student’s ttest versus DA). In the presence of
SCH23390, DA lost the capability of stimulating AC and slightly inhib-
ited cAMP formation. Data are expressed as the percentage increase of
cAMP over basal and represent the mean ⫾S.E. of three independent
experiments.
64 Fiorentini et al.
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

than in HEK-D1R cells. The calculated EC
50
value for DA-
stimulated cAMP formation in cells expressing the D1R/D3R
complex was 39 ⫾1.5 nM. The maximal extent of DA stim-
ulation of AC in cells expressing the D1R/D3R complex was
similar to that found in cells expressing only the D1R (85 ⫾
7% increase over basal) but was detectable at a 10-fold lower
concentration (5
␮
M). To confirm the requirement of the
concurrent activation of both D1R and D3R to potentiate the
stimulatory effect of DA on cAMP formation, we used the
D1R antagonist SCH23390 and D3R antagonist (⫺)sulpiride.
As reported in Fig. 4B, the increased potency of DA in stim-
ulating AC in cells expressing the D1R/D3R complex was
counteracted by both antagonists. In particular, inhibition of
DA interaction with the D3R by (⫺)sulpiride (1
␮
M) shifted
the dose-response curve of DA to the right. The EC
50
value
for DA-stimulated AC was in fact shifted from 39 ⫾1.5 to
520 ⫾47 nM, a value consistent with the potency of DA at the
D1R expressed alone. The D1R antagonist SCH23390 (1
␮
M)
completely abolished cAMP formation induced by DA in cells
coexpressing the D1R and the D3R, suggesting that AC ac-
tivation was completely sustained by the D1R. In these con-
ditions, a weak D3R-dependent inhibition of AC activity was
detected. The increased potency of DA in stimulating AC in
HEK-D1R/D3R cells was correlated with the increased affin-
ity of DA for the D1R, as determined by radioreceptor binding
with [
3
H]SCH23390. In particular, the curves for DA dis-
placement of [
3
H]SCH23390 in cells expressing D1R alone or
in combination with the D3R, resolved by a two-site analysis
(GraphPad Prism 4; GraphPad Software Inc., San Diego,
CA), showed a high-affinity and a low-affinity site, probably
corresponding to G protein-coupled and -uncoupled recep-
tors, respectively. Interaction with the D3R increases the
affinity of DA for the high-affinity site. The calculated K
i
values (mean ⫾S.E. of three independent experiments) for
DA displacement of [
3
H]SCH23390 binding were 60 ⫾4nM
(high-affinity site) and 4.4 ⫾0.5
␮
M (low-affinity site) in
HEK-D1R cells and 1.3 ⫾0.3 nM (high-affinity site) and
4.0 ⫾0.6
␮
M (low-affinity site) in HEK-D1/D3.
Heterodimerization Influences Agonist-Mediated
D1R and D3R Sequestration. Because GPCR heterodimer-
ization may affect the trafficking of interacting receptors
(Angers et al., 2002), we investigated whether heteromeric
assembly modifies agonist-induced cytoplasmic sequestra-
tion of D1R and D3R. This issue is particularly relevant
because D1R and D3R are characterized by different adap-
tive properties (Oakley et al., 2000; Kim et al., 2001, 2005).
For this purpose, we used immunofluorescence microscopy
and receptor binding in transfected HEK 293 cells, which
endogenously express adequate amounts of both GRKs and
␤
-arrestin to allow DA receptor sequestration. As shown in
Fig. 5A, in unstimulated HEK 293 cells expressing the D1R,
the fluorescence distribution of D1R was confined to the
plasma membrane (a). Exposure to 1
␮
M SKF 81297 for 1 h
resulted in D1R sequestration into cytosolic compartments,
as shown by the D1R immunofluorescence that was detect-
able also in the cytoplasm with a punctate appearance (b).
The extent of D1R internalization was determined by
[
3
H]SCH23390 binding in the purified heavy membrane frac-
tion from HEK-D1R cells. As shown in Fig. 5B, a 1-h treat-
ment with 1
␮
M SKF 81297 promoted a 30 ⫾4% decrease of
cell surface D1R. This effect was dose-dependent over the
range of 10 nM to 10
␮
M with a maximum at 1
␮
M and was
detectable within 10 min of treatment, reaching the maxi-
mum after a 30-min exposure. DA (1
␮
M) also induced a
significant decrease of D1R at the cell surface (22 ⫾3%). The
effect of quinpirole (0.5 nM to 1
␮
M) or DA (1
␮
M) on D3R
trafficking was studied in HEK 293 cells transiently trans-
fected with D3R-GFP. Figure 5A shows the effect of a 1-h
treatment with 1
␮
M quinpirole on D3R cellular localization.
In line with previous data (Kim et al., 2001), this treatment
did not significantly modify the membrane localization of
D3R-GFP (c and d). Similar results were obtained by mea-
suring cell surface D3R in binding studies with [
3
H]sulpiride
in the HEK-D3R clone (Fig. 5B), suggesting that the D3R
does not internalize in response to agonist stimulation in this
heterologous system. The adaptive responses of the D1R/D3R
complex to agonist stimulation were studied by immunoflu-
orescence in the HEK-D1R cells transiently transfected with
D3R-GFP. As shown in Fig. 6, D1R and D3R expressed in
HEK 293 cells were mostly targeted to the plasma membrane
(a and b) where they were colocalized (c). A 1-h stimulation
with 1
␮
M quinpirole did not modify the membrane localiza-
tion of the D1R/D3R complex, as shown by the fluorescence of
both D1R (d) and D3R (e) that remained colocalized at the
plasma membrane (f). Likewise, exposure of cotransfected
cells to 1
␮
M SKF 81297, which induced internalization of
D1R in individually transfected cells (see Fig. 4), did not
modify the membrane retention of the D1R/D3R complex
(g–i), indicating that association with the D3R may affect
D1R function by impairing its desensitization. It is interest-
ing that the paired stimulation of the two receptor compo-
nents of the D1R/D3R heterodimer by a combination of SKF
81297 (1
␮
M) and quinpirole (1
␮
M) relieved the membrane
retention of the D1R/D3R complex, enabling its internaliza-
tion. In these conditions, D1R and D3R fluorescence was
colocalized at cytoplasmic sites with a punctate appearance
(l–n). Similar results were obtained with DA. As shown in o
to q, 1-h stimulation with 1
␮
M DA induced cytoplasmic
sequestration of both D1R and D3R. This effect, which was
dose-dependent over the range of 100 nM to 10
␮
M, was
Fig. 5. Effects of agonist stimulation on D1R and D3R localization in
transfected HEK 293 cells. HEK 293 cells were individually transfected
with D1R or D3R-GFP cDNAs and left untreated or exposed to either 1
␮
M SKF 81297 or 1
␮
M quinpirole or 1
␮
M dopamine for 60 min at 37°C.
A, immunofluorescence analysis of D1R localization in untreated cells (a)
and in SKF 81297-treated cells (b) and of D3R localization in untreated
cells (c) and quinpirole-treated cells (d). B, D1R and D3R receptor seques-
tration measured by radioreceptor binding in HEK 293 cell clones stably
expressing either the D1R (525 ⫾56 fmol/mg protein) or the D3R (610 ⫾
58 fmol/mg protein). D1R internalization was evaluated by [
3
H]-
SCH23390 binding in the heavy membrane fraction, and D3R internal-
ization was evaluated by measuring cell surface [
3
H]sulpiride binding in
intact cells. Bars are the means ⫾S.E. of three independent experiments;
ⴱ,p⬍0.001 versus untreated cells, Student’s ttest.
D1 and D3 Receptor Heterodimerization 65
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

detectable after a 5-min incubation and reached a maximum
at 30 min (data not shown). These observations were con-
firmed by radioreceptor binding with [
3
H]SCH23390 and
[
3
H]sulpiride in the HEK-D1R/D3R cell clone. As shown in
Fig. 7A, treatment with either SKF 81297 (1
␮
M) or quin-
pirole (1
␮
M) did not affect the abundance of cell surface
[
3
H]sulpiride and [
3
H]SCH23390 binding sites, confirming
that in these conditions, the D1R/D3R complex is retained at
the plasma membrane. By contrast, the coincident stimula-
tion of both D1R and D3R with a combination of SKF 81297
and quinpirole resulted in a 28 ⫾3% reduction of cell surface
[
3
H]sulpiride binding sites and 30 ⫾2% loss of membrane
[
3
H]SCH23390 binding sites. On the same line, DA treat-
ment (1
␮
M) induced a 30 ⫾4% decrease of membrane
[
3
H]sulpiride binding and a 25 ⫾3% decrease of membrane
[
3
H]SCH23390 binding sites (Fig. 7B). These effects were
prevented by either D1R- or D3R-selective antagonists. The
D1R antagonist SCH 23390 (1
␮
M) abolished the loss of both
cell surface [
3
H]sulpiride and [
3
H]SCH23390 binding sites in
HEK-D1R/D3R cells exposed to 1
␮
M DA (Fig. 7B). Likewise,
the D3R antagonist (⫺)sulpiride (1
␮
M) prevented DA-
induced decrease of membrane [
3
H]sulpiride and [
3
H]-
SCH23390 binding. The observation that stimulation of both
D1R and D3R induces the same extent of D1R and D3R
internalization suggests that in our experimental conditions,
a significantly high proportion of D1R and D3R is associated
into the heterodimeric complex with respect to the corre-
sponding homodimeric complexes.
These results point to the critical importance of the paired
stimulation of both receptor components to induce D1R/D3R
complex internalization. Moreover, as shown in Fig. 7B, DA-
induced cytoplasmic sequestration of both [
3
H]sulpiride and
[
3
H]SCH23390 binding sites was abolished by
␤
-arrestin-
1V53D, a dominant-negative
␤
-arrestin mutant that pre-
vents agonist-induced GPCR sequestration (Zhang et al.,
1997). Because internalization may target GPCR to either a
degradative pathway, leading to prolonged attenuation of cell
signaling, or to a cell surface recycling pathway, facilitating
receptor resensitization (Gainetdinov et al., 2004), we evalu-
ated the time course of D1R/D3R recycling to the plasma
membrane. Cells were treated with a combination of SKF
81297 (1
␮
M) and quinpirole (1
␮
M) for 60 min to promote
sequestration of the receptor complex. Agonists were then
removed, and the reappearance of D1R and D3R at the cell
surface was monitored over time. As shown in Fig. 8A, in
unstimulated cells, D1R and D3R-GFP were colocalized at
the plasma membrane (a–c). Exposure of transfected cells to
SKF 81297 (1
␮
M) and quinpirole (1
␮
M) for 60 min induced
the cointernalization of D1R and D3R-GFP (d–f). Fifteen
minutes after agonist removal, a significant proportion of
D1R and D3R was detected back at the plasma membrane,
where they were still colocalized (g–i). Figure 8B shows the
time course of D1R and D3R recycling in HEK-D1R/D3R cells
evaluated in binding studies with [
3
H]sulpiride and [
3
H]-
SCH23390. A significant amount of both [
3
H]sulpiride and
[
3
H]SCH23390 binding sites returned to the cell surface
within 15 min of treatment withdrawal. The density of
[
3
H]sulpiride and [
3
H]SCH23390 binding sites measured 30
and 60 min after treatment withdrawal was indistinguish-
able from that detected in untreated cells.
Discussion
In this study, we reveal heterodimerization between the
DA D1R and D3R in both the striatum and transfected cells.
The evidence for the physical interaction of these receptor
subtypes is derived from coimmunoprecipitation, BRET
2
,
Fig. 6. Effects of agonist stimulation on receptor localization in HEK 293
cells coexpressing D1R and D3R-GFP. HEK 293 cells stably expressing
the D1R were transfected with D3R-GFP and exposed to either 1
␮
M SKF
81297 or 1
␮
M quinpirole or a combination of 1
␮
M SKF 81297 and 1
␮
M
quinpirole or 1
␮
M DA for 60 min at 37°C. Cell localization of D1R was
evaluated by immunofluorescence with the rat monoclonal anti-D1R an-
tibody and the Cy3-conjugated secondary antibody as described under
Materials and Methods. a to c, colocalization of D1R and D3R-GFP at the
plasma membrane; d to f, quinpirole administration does not modify the
membrane colocalization of D1R and D3R-GFP; g to i, membrane colo-
calization of D1R and D3R-GFP in SKF 81297-treated cells; l to n,
intracellular colocalization of D1R and D3R-GFP in cells exposed to a
combination of SKF 81297 and quinpirole; o to q, intracellular colocaliza-
tion of D1R and D3R-GFP in cells exposed to DA. Data are representative
of five independent experiments.
Fig. 7. Quantitative analysis of agonist-induced D1R/D3R sequestration
by [
3
H]SCH23390 and [
3
H]sulpiride binding. A, HEK-D1R/D3R cells were
exposed to agonists for 60 min at 37°C. [
3
H]Sulpiride binding was carried
out in intact cells and [
3
H]SCH23390 binding in the heavy membrane
fraction. Data are expressed as the percentage of loss of cell surface
receptors. Bars are the means ⫾S.E. of five independent experiments; ⴱ,
p⬍0.001 versus either SKF 81297 or quinpirole. B, HEK-D1R/D3R cells
were exposed to 1
␮
M DA for 60 min at 37°C in the absence or presence
of either the D1R antagonist SCH23390 (1
␮
M) or the D3R antagonist
(⫺)sulpiride (1
␮
M); HEK-D1R/D3R cells were also transfected with
␤
-arrestin-1-V53D and exposed to DA. [
3
H]Sulpiride binding was carried
out in intact cells and [
3
H]SCH23390 binding in the heavy membrane
fraction. Data are expressed as the percentage loss of cell surface recep-
tors. Bars are the means ⫾S.E. of three independent experiments. ⴱ,p⬍
0.001 versus DA, Student’s ttest.
66 Fiorentini et al.
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

and cointernalization experiments. As a result of het-
erodimerization, these receptors display functional proper-
ties that are remarkably different from those of D1R and
D3R homo-oligomers. In particular, a unique characteristic
of D1R/D3R heterodimerization is that it increases the affin-
ity of DA for the D1R and the potency of DA in stimulating
AC through the D1R, abolishes agonist-induced D1R inter-
nalization, and enables the cytoplasmic sequestration of the
receptor complex in response to the paired stimulation of
D1R and D3R.
Using a conventional biochemical approach, we have
shown that the D3R was coimmunoprecipitated with the D1R
from striatal proteins, suggesting that these receptors may
be physically associated in this structure. The observation
that D1R and D3R are coexpressed in specific neuronal pop-
ulations of both limbic (Le Moine and Bloch, 1996; Ridray et
al., 1998; Schwartz et al., 1998) and motor areas (Surmeier et
al., 1996; Bordet et al., 2000; Guillin et al., 2001) supports
this finding and provides the anatomical basis for D1R-D3R
direct interactions. By using BRET
2
in transfected HEK 293
cells, we further demonstrated that D1R and D3R cocluster-
ing reflects the existence of a physical proximity between
these receptors that can be explained best by the formation of
protein heterodimers. Tagged D1R and D3R generated, in
fact, a significant and specific BRET
2
signal in cotransfected
HEK 293 cells that was insensitive to stimulation with either
D1R- or D3R-selective agonists. Costimulation of D1R and
D3R by DA, however, increased the BRET
2
signal, an effect
that could potentially reflect either the further clustering of
nonheteromeric D1R and D3R or the occurrence of conforma-
tional changes at preformed D1R/D3R complexes, increasing
the molecular proximity of BRET
2
partners or the clustering
of complexes into endocytotic vesicles, also resulting in in-
creased proximity of BRET
2
partners (Angers et al., 2002).
The observation that mutant dynamin I-K44A, which pre-
vents agonist-mediated GPCR internalization (Zhang et al.,
1997), antagonized DA-induced increase of BRET
2
signal
points to D1R/D3R complex internalization as the most likely
event to explain this finding. However, it cannot be excluded
that other mechanisms could contribute to the effect of DA in
the BRET
2
assay. The existence of a functional cross-talk
between D1R and D3R, involving the convergence of their
signaling pathways, has been reported previously (Ridray et
al., 1998; Schwartz et al., 1998). Our present data, showing
that D1R and D3R are constitutively assembled into a het-
erodimeric complex, extend these observations and provide
the molecular basis for the reported functional interactions
between these receptors.
In transfected cells, the interaction between D1R and D3R
finds an important functional implication in the modulation
of D1R-mediated stimulation of cAMP formation. D1R and
D3R primarily exert opposite effects on AC, being the D1R-
stimulatory and the D3R-inhibitory (Missale et al., 1998). In
HEK 293 cells, however, the D3R only marginally inhibits
cAMP formation, because these cells poorly express AC type
V, which is targeted by the D3R (Robinson and Caron, 1997).
Nevertheless, coexpression of D1R and D3R potentiated DA
stimulation of cAMP formation via the D1R. Whether this
effect is detectable also in cells expressing AC type V, which
is inhibited by the D3R, remains to be established. The in-
creased potency of DA in stimulating AC in HEK-D1R/D3R
cells was correlated with increased affinity of DA for the
high-affinity site of D1R. Whether the interaction between
D1R and D3R also modifies the affinity of selective com-
pounds for D1R or D3R is still matter of investigation. One
function of the D1R/D3R heteromeric complex may therefore
be to allow a stronger stimulatory coupling of the D1R to AC.
In animal models of L-DOPA-induced dyskinesias (LIDs)
D1R-related cAMP signaling is enhanced (Aubert et al.,
2005) and D3R expression is increased in striatal neurons
containing the D1R (Bordet et al., 2000; Guillin et al., 2001;
Be´zard et al., 2003). Both dysfunctions have been causally
linked to the development of LIDs. Our present data may
provide a mechanism by which to converge D1R- and D3R-
related alterations in the development of LIDs. It is possible,
in fact, that D1R/D3R interaction in striatal neurons is in-
creased in dyskinetic animals as a result of the increased
expression of the D3R, leading to supersensitivity of D1R-
mediated responses.
The interaction between D1R and D3R also influenced both
D1R and D3R trafficking from the plasma membrane to
intracellular compartments. Internalization, involving both
GRK-mediated phosphorylation and arrestin binding, is a
common adaptive response of GPCR to agonist stimulation
(Gainetdinov et al., 2004). This mechanism not only termi-
nates receptor signaling, but also promotes receptor resensi-
tization and recycling to the plasma membrane. In this
study, we demonstrated that D1R/D3R dimerization modifies
agonist-mediated internalization of both D1R and D3R, a
finding of relevance because D1R and D3R show different
adaptive properties. The D1R undergoes agonist-induced cy-
toplasmic sequestration and rapidly recycles back to the
plasma membrane fully resensitized (Oakley et al., 2000;
Gainetdinov et al., 2004), whereas D3R desensitization in-
volves GRK-mediated impairment of D3R binding to filamin
(Kim et al., 2005) resulting in decreased G protein coupling
with only marginal changes of membrane receptor density
(Kim et al., 2001, 2005). Our data show that heterodimeriza-
tion with the D3R abolished agonist-induced D1R cytoplas-
mic sequestration, suggesting that the adaptive responses of
Fig. 8. Recycling of internalized D1R/D3R receptors at the plasma mem-
brane. HEK 293 cells expressing D1R and D3R-GFP or HEK-D1R/D3R
cells were exposed to SKF 81297 (1
␮
M) and quinpirole (1
␮
M) for 60 min
at 37°C. Agonists were removed, and the reappearance of the D1R/D3R at
the cell surface was monitored over time. A, immunofluorescence analysis
of D1R/D3R recycling. a to c, colocalization of D1R and D3R-GFP at the
cell surface in untreated cells; d to f, cytoplasmic colocalization of D1R
and D3R-GFP in agonist-treated cells; g to i, reappearance and colocal-
ization of D1R and D3R-GFP at the cell surface 15 min after agonist
removal. Data are representative of three independent experiments. B,
time course of D1R/D3R recycling evaluated by [
3
H]sulpiride binding in
intact cells (F) and [
3
H]SCH23390 binding in the heavy membrane frac-
tion (E). Points are the means ⫾S.E. of four independent experiments. ⴱ,
p⬍0.001 versus time 0, Student’s ttest.
D1 and D3 Receptor Heterodimerization 67
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

D1R may differ from neuron to neuron or in different mi-
crodomains of the same neuron, depending on its interaction
with other membrane proteins. On the other hand, D1R/D3R
dimerization enabled cointernalization of both D1R and
D3R in response to the paired stimulation of both receptor
components within the heterodimer, suggesting that this
interaction could represent a novel mechanism of D1R-D3R
reciprocal regulation. Furthermore, our data point to an ad-
ditional mechanism of D3R desensitization, occurring when
this receptor is assembled with the D1R. Internalization of
D1R/D3R complex probably occurs via the clathrin-coated
vesicle-mediated endocytotic pathway involving
␤
-arrestin
binding because it was blocked by mutant
␤
-arrestin-1V53D,
which prevents GPCR internalization (Zhang et al., 1997).
These data thus suggest that as a result of dimerization, the
D3R is switched to the trafficking mechanisms typical of the
D1R. In line with our observations, changes in the trafficking
of a given receptor due to heterodimerization have been re-
ported previously. In some cases, agonist occupancy of only
one protomer within the complex is sufficient to induce in-
ternalization of the heterodimer (Angers et al., 2002). In
other cases, costimulation of both protomers within the
dimer is crucial to promote internalization. In particular,
internalization of the D1R/N-methyl-D-aspartate receptor
complex (Fiorentini et al., 2003) and recruitment of
␤
-arres-
tin-1 by M2/M3 muscarinic heterodimer and by adrenergic
␣
2/muscarinic M3 heterodimeric unit (Novi et al., 2005) have
been reported to require the paired activation of the single
receptors within the heterodimers. Different mechanisms
could explain the finding that oligomerization with D1R en-
ables D3R cytoplasmic sequestration. For example, dimeriza-
tion with the D1R might enable the recruitment of the endo-
cytotic machinery to the D3R itself or might enable the D3R
to access the endocytotic effectors linked to the D1R. How-
ever, it is also possible that the novel D1R/D3R unit has
different internalization characteristics compared with that
of D1R and D3R. This last possibility is supported by the
observation that SKF 81297 did not induce D1R cytoplasmic
sequestration in the presence of the D3R. It has been sug-
gested that in DA neurons, the function of D3 autoreceptors
might be regulated by DA through modulation of filamin
binding and G protein interaction to allow its fast desensiti-
zation and resensitization, a mechanism that may be crucial
to provide continuous control of synaptic DA concentrations
(Kim et al., 2005). On the other hand, our present data
suggest that in neurons coexpressing D3R and D1R at the
postsynaptic level, the D3R might undergo internalization in
response to DA as a result of heterodimerization with the
D1R, allowing a sustained adaptive cell response to the
strength of synaptic transmission. The internalized D1R and
D3R rapidly recycle back to the plasma membrane, where
they are still colocalized. Whether the intact heteromeric
complex recycles back to the cell surface or it is dissociated
after internalization and each receptor recycles indepen-
dently to form again the complex at the plasma membrane
cannot be established by our present data.
Both D1R and D3R have been implicated in several disor-
ders, including schizophrenia and motor dysfunctions. In
particular, both the symptoms of schizophrenia and the ab-
normal involuntary movements induced by L-DOPA in pa-
tients with Parkinson’s disease have been suggested to re-
flect imbalances in the relative abundance and function of
D1R and D3R (Schwartz et al., 1998; Bordet et al., 2000;
Be´zard et al., 2003; Aubert et al., 2005). Our present data
give a novel insight into how these receptors may function in
an integrated way, thus providing a molecular mechanism by
which to converge D1R- and D3R-related dysfunctions. The
D1R/D3R heterodimer could therefore represent a potential
and promising drug target for disorders related to the dopa-
minergic system.
Acknowledgments
We thank Dr. Marc Caron (Duke University, Durham, NC) for his
gift of the human D1 and D3 receptor cDNAs, the D3-GFP construct,
the dynamin-IK44A mutant, and the
␤
-arrestin-1V53D mutant, and
Drs. Silvano Sozzani and Daniela Bosisio, Department of Biomedical
Sciences and Biotechnology, University of Brescia (Brescia, Italy),
for their gift of the ChemR23 receptor and their help with fluores-
cence-activated cell sorting experiments.
References
Amlaiky N, Berger JG, Chang W, McQuade J, and Caron MG (1987) Identification of
the binding subunit of the D1-dopamine receptor by photoaffinity crosslinking.
Mol Pharmacol 31:129–134.
Angers S, Salahpour A, and Bouvier M (2002) Dimerization: an emerging concept for
G protein-coupled receptor ontogeny and function. Annu Rev Pharmacol Toxicol
42:409–435.
Aubert I, Guigoni C, Hakansson K, Li Q, Barche N, Bioulac BH, Gross CE, Fisone G,
Bloch B, and Be´zard E (2005) Increased D1 dopamine receptor signaling in levo-
dopa-induced dyskinesia. Ann Neurol 57:17–26.
Be´zard E, Ferry S, Mach U, Stark H, Leriche L, Boraud T, Gross C, and Sokoloff P
(2003) Attenuation of levodopa-induced dyskinesia by normalizing dopamine D3
receptor function. Nat Med 9:762–767.
Bordet R, Ridray S, Schwartz JC, and Sokoloff P (2000) Involvement of the direct
striatonigral pathway in levodopa-induced sensitization in 6-hydroxydopamine-
lesioned rats. Eur J Neurosci 12:2117–2123.
Fiorentini C, Gardoni F, Spano PF, Di Luca M, and Missale C (2003) Regulation of
dopamine D1 receptor trafficking and desensitization by oligomerization with
glutamate N-methyl-D-aspartate receptors. J Biol Chem 278:20196–20202.
Gainetdinov RR, Premont RT, Bohn LM, Lefkowitz RJ, and Caron MG (2004)
Desensitization of G protein-coupled receptors and neuronal functions. Annu Rev
Neurosci 27:107–144.
Gine´s S, Hillion J, Torvinen M, Le Crom S, Casado V, Canela EI, Rondin S, Lew JY,
Watson S, Zoli M, et al. (2000) Dopamine D
1
and adenosine A
1
receptors form
functionally interacting heteromeric complexes. Proc Natl Acad Sci U S A 97:
8606–8611.
Guillin O, Diaz J, Carroll P, Griffon N, Schwartz JC and Sokoloff P (2001) BDNF
controls dopamine D3 receptor expression and triggers behavioural sensitization.
Nature 411:86–89.
Jarvie KR, Booth G, Brown MB, and Niznik HB (1989) Glycoprotein nature of
dopamine D1 receptors in the brain and parathyroid gland. Mol Pharmacol 36:
566–574.
Jung MY and Schmauss C (1999) Decreased c-fos responses to dopamine D
1
receptor
agonist stimulation in mice deficient for D
3
receptors. J Biol Chem 274:29406–
29412.
Karasinska JM, George SR, Cheng R, and O’Dowd BF (2005) Deletion of dopamine
D1 and D3 receptors differentially affects spontaneous behaviour and cocaine-
induced locomotor activity, reward and CREB phosphorylation. Eur J Neurosci
22:1741–1750.
Karasinska JM, George SR, El-Ghundi M, Fletcher PJ, and O’Dowd BF (2000)
Modification of dopamine D
1
receptor knockout phenotype in mice lacking both
dopamine D
1
and D
3
receptors. Eur J Pharmacol 399:171–181.
Kim KM, Gainetdinov RR, Laporte SA, Caron MG, and Barak LS (2005) G protein-
coupled receptor kinase regulates dopamine D3 receptor signaling by modulating
the stability of a receptor-filamin-
␤
-arrestin complex. A case of autoreceptor reg-
ulation. J Biol Chem 280:12774–12780.
Kim KM, Valenzano KJ, Robinson SR, Yao WD, Barak LS, and Caron MG (2001)
Differential regulation of the dopamine D2 and D3 receptors by G protein-coupled
receptor kinases and
␤
-arrestins. J Biol Chem 276:37409–37414.
Le Moine C and Bloch B (1996) Expression of the D3 dopamine receptor in pepti-
dergic neurons of the nucleus accumbens: comparison with the D1 and D2 dopa-
mine receptors. Neuroscience 73:131–143.
Lee FJ, Xue S, Pei L, Vukusic B, Che´ry N, Wang Y, Wang YT, Niznik HB, Yu XM,
and Liu F (2002) Dual regulation of NMDA receptor functions by direct protein-
protein interactions with the dopamine D1 receptor. Cell 111:219–230.
Levavi-Sivan B, Park BH, Fuchs S, and Fishburn CS (1998) Human D3 dopamine
receptor in the medulloblastoma TE671 cell line: cross-talk between D1 and D3
receptors. FEBS Lett 439:138–142.
Le´vesque D, Diaz J, Pilon C, Martres MP, Giros B, Souil E, Schott D, Morgat JL,
Schwartz JC, and Sokoloff P (1992) Identification, characterization, and localiza-
tion of the dopamine D3 receptor in rat brain using 7-[
3
H]hydroxy-N,N-di-n-
propyl-2-aminotetralin. Proc Natl Acad SciUSA89:8155–8159.
Missale C, Nash SR, Robinson SW, Jaber M, and Caron MG (1998) Dopamine
receptors: from structure to function. Physiol Rev 78:189–225.
68 Fiorentini et al.
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from

Nimchinsky EA, Hof PR, Janssen WG, Morrison JH, and Schmauss C (1997) Ex-
pression of dopamine D3 receptor dimers and tetramers in brain and in transfected
cells. J Biol Chem 272:29229–29237.
Novi F, Stanasila L, Giorgi F, Corsini GU, Cotecchia S, and Maggio R (2005) Paired
activation of two components within muscarinic M3 receptor dimers is required for
recruitment of
␤
-arrestin-1 to the plasma membrane. J Biol Chem 280:19768–
19776.
Oakley RH, Laporte SA, Holt JA, Caron MG, and Barak LS (2000) Differential
affinities of visual arrestin,
␤
arrestin1, and
␤
arrestin2 for G protein-coupled
receptors delineate two major classes of receptors. J Biol Chem 275:17201–17210.
Rashid AJ, O’Dowd BF, Verma V, and Gorge SR (2007) Neuronal Gq/11-coupled
dopamine receptors: an uncharted role for dopamine. Trends Pharmacol Sci 28:
551–555.
Ridray S, Griffon N, Mignon V, Souil E, Carboni S, Diaz J, Schwartz JC, and Sokoloff
P (1998) Coexpression of dopamine D1 and D3 receptors in islands of Calleja and
shell of nucleus accumbens of the rat: opposite and synergistic functional interac-
tions. Eur J Neurosci 10:1676–1686.
Robinson SW and Caron MG (1997) Selective inhibition of adenylyl cyclase type V by
the dopamine D3 receptor. Mol Pharmacol 52:508–514.
Scarselli M, Novi F, Schallmach E, Lin R, Baragli A, Colzi A, Griffon N, Corsini GU,
Sokoloff P, Levenson R, et al. (2001) D2/D3 dopamine receptor heterodimers
exhibit unique functional properties. J Biol Chem 276:30308–30314.
Schwartz JC, Diaz J, Bordet R, Griffon N, Perachon S, Pilon C, Ridray S, and
Sokoloff P (1998) Functional implications of multiple dopamine receptor subtypes:
the D1/D3 receptor coexistence. Brain Res Rev 26:236–242.
Scott L, Zelenin S, Malmersjo S, Kowalewski JM, Markus EZ, Nairn AC, Greengard
P, Brismar H, and Aperia A (2006) Allosteric changes of the NMDA receptor trap
diffusible dopamine 1 receptors in spines. Proc Natl Acad SciUSA103:762–767.
Sokoloff P, Giros B, Martres MP, Bouthenet ML, and Schwartz JC (1990) Molecular
cloning and characterization of a novel dopamine receptor (D3) as a target for
neuroleptics. Nature 347:146–151.
Surmeier DJ, Song WJ, and Yan Z (1996) Coordinated expression of dopamine
receptors in neostriatal medium spiny neurons. J Neurosci 16:6579–6591.
Torvinen M, Marcellino D, Canals M, Agnati LF, Lluis C, Franco R, and Fuxe K
(2005) Adenosine A2A receptor and dopamine D3 receptor interactions: evidence of
functional A2A/D3 heteromeric complexes. Mol Pharmacol 67:400–407.
Vailati S, Hanke W, Bejan A, Barabino B, Longhi R, Balestra B, Moretti M, Clementi
F, and Gotti C (1999) Functional
␣
6-containing nicotinic receptors are present in
chick retina. Mol Pharmacol 56:11–19.
Wittamer V, Franssen JD, Vulcano M, Mirjolet JF, Le Poul E, Migeotte I, Bre´zillon
S, Tyldesley R, Blanpain C, Detheux M, et al. (2003) Specific recruitment of
antigen-presenting cells by chemerin, a novel processed ligand from human in-
flammatory fluids. J Exp Med 198:977–985.
Xu M, Koeltzow TE, Santiago GT, Moratalla R, Cooper DC, Hu XT, White NM,
Graybiel AM, White FJ, and Tonegawa S (1997) Dopamine D3 receptor mutant
mice exhibit increased behavioral sensitivity to concurrent stimulation of D1 and
D2 receptors. Neuron 19:837–848.
Zhang J, Barak LS, Winkler KE, Caron MG, and Ferguson SS (1997) A central role
for beta-arrestins and clathrin-coated vesicle-mediated endocytosis in
␤
2-
adrenergic receptor resensitization. Differential regulation of receptor resensiti-
zation in two distinct cell types. J Biol Chem 272:27005–27014.
Address correspondence to: Dr. Cristina Missale, Section of Pharmacology,
Department of Biomedical Sciences and Biotechnology, University of Brescia,
Viale Europa 11, 25124 Brescia, Italy. E-mail: cmissale@med.unibs.it
D1 and D3 Receptor Heterodimerization 69
at ASPET Journals on July 7, 2016molpharm.aspetjournals.orgDownloaded from