The mGlu5 Receptor Protomer-Mediated Dopamine D2 Receptor Trans-Inhibition Is Dependent on the Adenosine A2A Receptor Protomer: Implications for Parkinson’s Disease

The adenosine A2A receptor (A2AR), dopamine D2 receptor (D2R) and metabotropic glutamate receptor type 5 (mGluR5) form A2AR-D2R-mGluR5 heteroreceptor complexes in living cells and in rat striatal neurons. In the current study, we present experimental data supporting the view that the A2AR protomer plays a major role in the inhibitory modulation of the density and the allosteric receptor-receptor interaction within the D2R-mGluR5 heteromeric component of the A2AR-D2R-mGluR5 complex in vitro and in vivo. The A2AR and mGluR5 protomers interact and modulate D2R protomer recognition and signalling upon forming a trimeric complex from these receptors. Expression of A2AR in HEK293T cells co-expressing D2R and mGluR5 resulted in a significant and marked increase in the formation of the D2R-mGluR5 heteromeric component in both bioluminescence resonance energy transfer and proximity ligation assays. A highly significant increase of the the high-affinity component of D2R (D2RKi High) values was found upon cotreatment with the mGluR5 and A2AR agonists in the cells expressing A2AR, D2R and mGluR5 with a significant effect observed also with the mGluR5 agonist alone compared to cells expressing only D2R and mGluR5. In cells co-expressing A2AR, D2R and mGluR5, stimulation of the cells with an mGluR5 agonist like or D2R antagonist fully counteracted the D2R agonist-induced inhibition of the cAMP levels which was not true in cells only expressing mGluR5 and D2R. In agreement, the mGluR5-negative allosteric modulator raseglurant significantly reduced the haloperidol-induced catalepsy in mice, and in A2AR knockout mice, the haloperidol action had almost disappeared, supporting a functional role for mGluR5 and A2AR in enhancing D2R blockade resulting in catalepsy. The results represent a relevant example of integrative activity within higher-order heteroreceptor complexes. Supplementary Information The online version contains supplementary material available at 10.1007/s12035-022-02946-9.


Introduction
The first pieces of evidence for antagonistic glutamate receptor with dopamine D 2 receptor (D 2 R) interactions were found in 1983-1984 through the ability of glutamate to reduce the affinity of the high-affinity D 2 R agonist binding sites in striatal membrane preparations. Subsequently, it was observed that mGluR 5 agonists alone or combined with an A 2A R agonist (CGS-21680) can reduce the affinity of the high-affinity state of D 2 R for agonist binding sites in the rat striatum [1]. Co-immunoprecipitation experiments also indicated the existence of A 2A R-mGluR 5 heteroreceptor complexes in HEK293 cells and rat striatal membrane preparations [2]. The colocation of the receptors in striatal neurons was demonstrated [3,4] as well as their synergistic interactions as studied with in vivo microdialysis and intracellular signalling in striatal preparations [2,5,6].
In 1974, the discovery that the methylxanthines caffeine and theophylline could enhance the contralateral turning behaviour induced by levodopa and dopamine receptor agonists in the hemi-Parkinsonian rat model was one early finding leading to the hypothesis that antagonistic adenosine-dopamine interactions existed [7,8]. Today, a considerable amount of molecular and functional experimental data supports the view that A 2A R and D 2 R form heteroreceptor complexes with antagonistic receptor-receptor interactions on the plasma membrane [9][10][11][12][13][14][15][16].
The existence of A 2A R-D 2 R-mGluR 5 higher-order oligomers was postulated, and it was proposed that the receptor-receptor interactions within this high-order complex are important to modulate the dorsal and ventral striatal-pallidal GABA neurons [2,3,8]. Years later, it was proposed that combined treatment with A 2A R and mGluR 5 agonists targeting A 2A R-D 2 R-mGluR 5 heteroreceptor complexes in the ventral striatal-pallidal GABA pathway can represent a new strategy for the treatment of schizophrenia [17]. Also, the combine treatment with selective A 2A and mGluR 5 receptor antagonists represents an alternative therapeutic approach to Parkinson's disease [18][19][20].
A combination of bimolecular fluorescence complementation assays and bioluminescence resonance energy transfer assays as well as the sequential resonance energy transfer technique was used to show that A 2A R-D 2 R-mGluR 5 heteroreceptor complexes exist in living cells [21]. In addition, high-resolution immunoelectron microscopy was also used to further demonstrate their existence in striatal glutamate synapses [21]. An integrative role of these receptor complexes in adenosine, dopamine and glutamate transmission was also proposed [8,22,23]. Recently, A 2A R, D 2 R and mGluR 5 receptor-receptor interactions were also found to modulate the activity of the striatal-pallidal GABA neurons based on in vivo dual-probe microdialysis [24].
Herein, new findings that further expand the understanding of A 2A R-D 2 R-mGluR 5 heteroreceptor complexes are presented. Results in cellular models first demonstrated that A 2A R promotes the D 2 R and mGluR 5 receptor-receptor interactions, and its participation increases the density of the D 2 R-mGluR 5 heterocomplexes. Binding and functional experiments indicated that A 2A R and mGluR 5 upon agonist activation play a significant role in modulating the composition, density and signalling of A 2A R-D 2 R-mGluR 5 heteroreceptor complexes. This was also observed in A 2A R or D 2 R knockout mice when studying the effects of the mGluR 5 negative allosteric modulator raseglurant on locomotor activity.

Plasmid Constructs
The cDNA encoding the rat mGluR 5 was cloned (without stop codon) in pGFP 2 -N1 vector (PerkinElmer, Waltham, MA, USA) using standard molecular biology techniques.
The D 2 R Rluc construct used has been described previously in Borroto-Escuela et al. 2010 [25].

Cell Culture and Transfection
Human embryonic kidney 293T (HEK293T cells (American Type Culture Collection, Manassas, VA, USA) cells were grown in Dulbecco's Modified Eagle's Medium supplemented with 2 mM L-glutamine, 100 units/ml penicillin/ streptomycin and 10% (v/v) foetal bovine serum at 37 °C in an atmosphere of 5% CO 2 . Cells were plated in 6-well plates (1 × 10 6 cells/well), 96-well plates (1 × 10 4 cells/well) or in 75 cm 2 flasks and cultured overnight prior to transfection or experimental procedures. Cells were transiently transfected using linear polyethyleneimines (Polysciences Inc., Warrington, PA, USA) according to the manufacturer's instructions.

Animals
A 2A R −/− and D 2 R −/− mice generated on a CD-1 genetic background [30,32] and the corresponding wild-type littermates weighing 20-25 g were used. The animal protocol (no. 7085) was approved by the University of Barcelona Committee on Animal Use and Care. Animals were housed and tested in compliance with the guidelines provided by the Guide for the Care and Use of Laboratory Animals [33] and following the European Union directives (2010/63/EU), the ARRIVE guidelines [34]. Mice were housed in groups of five in standard cages with access to food and water ad libitum while maintained under a 12-h dark/light cycle (starting at 7:30 AM), 22 °C temperature and 66% humidity (standard conditions). All animal experimentation was carried out in a period comprehended between 9:00 AM and 6:00 PM by a researcher blind to drug treatments.

Locomotor Activity Tests
Mice spontaneous or drug-induced locomotor activity was assessed by the open field test. In brief, animals were administered intraperitoneal (i.p.) with raseglurant (1 mg/kg) or vehicle-saline with 5% DMSO and 5% Tween 20 30 min before the testing session. Non-habituated mice were placed in the centre of an activity field arena (30 × 30 cm, surrounded by four 50-cm-high black-painted walls) equipped with a camera above to record activity and connected to the light source. The total distance travelled was analysed using SPOT tracker function from ImageJ (NIH, Bethesda, MD, USA), as previously described [30].

Catalepsy Test
Mouse catalepsy was induced by the administration (i.p.) of haloperidol (1 mg/kg) [30]. After 1 h, haloperidol-induced catalepsy was measured as the duration in seconds of an abnormal upright posture in which the forepaws of the mouse were placed on a horizontal wooden bar (0.6 cm of diameter) that was located 4.5 cm above the floor. Subsequently, mice were administered (i.p.) with either vehicle (i.e. saline with 5% DMSO and 5% Tween) or raseglurant (1 mg/kg). After 20 min, a second haloperidol-induced catalepsy measurement was performed.
The rationale for the use of raseglurant (a mGluR5-negative allosteric modulator) instead of a full antagonist was based on the theoretical advantages that allosteric modulators offer compared with their competitive counterparts. mGluR5 allosteric modulators (negative allosteric modulators (NAM) and positive allosteric modulators (PAM)) have the potential for greater subtype selectivity when compared to orthosteric ligands. Also, mGluR5 NAM and PAM do not possess intrinsic activity and are assumed to be quiescent in the absence of an endogenous agonist and only modulate receptor function when the endogenous agonist is present. In this manner, NAM and PAM have the potential to retain spatial and temporal aspects of endogenous receptor signalling. This is of particular interest for CNS targets where optimal neurotransmission is likely to have an improved therapeutic outcome as opposed to sustained receptor blockade or activation.

Haloperidol-Induced Catalepsy
Mice (n = 10) were randomly assigned to treatment groups, and behavioural testing was performed blind to treatment. The dopamine D 2 receptor (D 2 R) antagonist, haloperidol (1 mg/kg, s.c.), was administered to induce catalepsy. Thirty minutes after the haloperidol administration, mice experienced a full cataleptic response. At this time point, for each mouse, the state of catalepsy was tested by gently placing their front limbs over an 8-cm-high horizontal bar. The intensity of catalepsy was assessed by measuring the time the mice remain in this position being completely immobile for a maximum of 120 s. Only mice that remained cataleptic for the entire 120 s were used for subsequent drug testing. After 30 min of the baseline measurement vehicle (0.5% methylcellulose and 2% DMSO), PBF509 was administered orally via gavage (3, 10 or 30 mg/kg, p.o.), and the catalepsy was then determined at 15, 30 and 60 min PBF509 administration. For each time point, the number of responding mice and the total cataleptic time for each animal were determined.

Membrane Preparation
HEK293T cells or mouse striata were homogenized in icecold 10 mM Tris HCl, pH 7.4, 1 mM EDTA and 300 mM KCl buffer containing a protease inhibitor cocktail (Roche, Penzberg, Germany) using a Polytron for three periods of 10 s each. The homogenate was centrifuged for 10 min at 1000 × g. The resulting supernatant was centrifuged for 30 min at 12,000 × g. The membranes were dispersed in 50 mM Tris HCl (pH 7.4) and 10 mM MgCl 2 , washed and resuspended in the same medium. Protein concentration was determined using the BCA protein assay kit (Thermo Fisher Scientific, Inc., Rockford, IL, USA).

Bioluminescence Resonance Energy Transfer Saturation Assay
BRET 2 saturation curves have been particularly used with the aim to establish the oligomeric order of receptor complexes, as well as the proportion of receptors engaged in dimers or oligomers (BRETmax). In the current work, bioluminescence resonance energy transfer (BRET 2 ) saturation assays were carried out using plasmids encoding for D 2 R Rluc and mGluR 5 GFP2 according to previously published methods [9,26,35,36]. The netBRET 2 ratio was defined as the BRET ratio for co-expressed Rluc and GFP 2 constructs normalized against the BRET ratio for the Rluc expression construct alone: netBRET 2 ratio = [(GFP 2 emission at 515 ± 30 nm)/ (Rluc emission 410 ± 80 nm)]-cf. The correction factor, cf, corresponds to (emission at 515 ± 30 nm)/(emission at 410 ± 80 nm) found with the receptor-Rluc construct expressed alone in the same experiment. The maximal value of BRET (netBRET 2 max) corresponds to the situation when all available donor molecules are paired up with acceptor molecules [8]. Also, saturation assay was used to compare the relative affinity of receptors for each other and their probability to form a complex, the so-called BRET50, which represents the acceptor/donor ratio giving 50% of the maximal signal. The ratio is calculated from fluorescence and bioluminescence values expressed as arbitrary units. BRET50 values should not be regarded as a common or classical value to expressed affinities as Molar units. Pairs with low BRET 50 value thought to form oligomers or an increased tendency to dimerize, while high BRET 50 values indicate weak interaction or the absence of interaction between the investigated receptors. The specificity of D 2 R Rluc -mGluR 5 GFP2 interactions was assessed by comparison with co-expression of A 1 R GFP2 and D 2 R Rluc .

In Situ PLA in Cultured Cells
In situ proximity ligation assay (PLA) in cultured cells was performed using the Duolink in situ PLA detection kit (Sigma-Aldrich, St. Louis, MO, USA), following the protocol described previously [11,37,38] using mouse monoclonal anti-D 2 R (2 μg/ml, MABN53; Millipore, Billerica, MA, USA) and rabbit polyclonal anti-mGluR 5 (2 μg/ml, AB5675; Millipore) primary antibodies. PLA control experiments employed only one primary antibody. The PLA signal was visualized and quantified by using a TCS-SL confocal microscope (Leica Lasertechnik GmbH, Heidelberg, Germany) and the Duolink ImageTool software. High magnifications of the microphotograph were taken and visualized using multiple z-scan projections.
The background signal was estimated from both PLA control experiments and from PLA experiments performed on non-transfected HEK293T cells (HEK293T cell line expresses endogenously small amount of D 2 R, A 2A R and mGluR 5 ). In general, the positive PLA values obtained in these experiments were residuals. The assay cut-off value was set to two standard deviations over the background signal. Therefore, samples with values below this cut-off were negative for the interaction of interest, while samples with values higher than the threshold were positive.
For in situ PLA in mouse brain, the Duolink in situ PLA detection kit (Sigma-Aldrich) was used as previously described [37,39,40]. Thus, the experimental procedure until the secondary antibody incubation step was the same as the IHF (see above). Subsequently, the following steps were performed according to the manufacturer's protocol. Images were acquired and analysed as previously described [39]. The background signal was estimated from PLA control experiments, and the assay cut-off value was performed as described above.

Radioligand Competition Binding Experiments
For the binding experiments, membrane preparations (60 μg protein/ml) were obtained from HEK293T cells expressing either D 2 R and mGluR 5 or A 2A R, D 2 R and mGluR 5 , and [ 3 H]-raclopride (Novandi Chemistry AB, Södertälje, Sweden) competition assays with minor modifications were performed according to previously published methods [26, 27,41]. [ 3 H]-raclopride (75 Ci/mmol), a D 2 -like receptor antagonist competing [42] with quinpirole for binding to D 2 -like receptors in HEK293T membrane preparations, was used to determine the D 2 R high-affinity (K i, High ) and D 2 R low-affinity (K i, Low ) values. (+)-Butaclamol, a selective D2R antagonist (100 μM, Sigma-Aldrich), was used to determine the non-specific binding. The amount of bound [ 3 H]-raclopride was determined by liquid scintillation spectrometry.

cAMP Functional Assay
Intracellular cAMP levels were determined using a cAMP-Glo™ assay detection kit (Promega, Madison, WI, USA). HEK293T cells expressing either D 2 R and mGluR 5 or A 2A R, D 2 R and mGluR 5 were plated at a density of 10,000 cells/ well in 96-well microtiter plates coated with poly-L-lysine (Sigma-Aldrich) and incubated overnight. Culture medium was then removed; cells were washed with 1 × PBS before the induction buffer (red phenol/serum-free DMEM containing 500 μM IBMX and 100 μM Ro 20-1724) was added. The cells were incubated for 1 h prior to drug incubation. To examine the G i protein-mediated inhibition of adenylyl cyclase, the levels of cAMP were first raised with 5 µM forskolin for 10 min. Drug dilutions were prepared in the induction buffer, and the temperature-and carbon dioxide-equilibrated drug dilutions (37 °C cell culture incubator for 30 min) were added as indicated, and cells were then incubated at 37 °C for 30 min. The assay was performed accordingly to the manufacturer's specifications (Promega, Sweden). Readings of luminescence intensity were performed using the POLARstar Optima plate reader (BMG Lab Technologies, Offenburg, Germany). cAMP levels in non-transfected, non-treated cells and non-transfected cells treated only with forskolin were defined as basal and control, respectively.
PVDF membranes were washed with PBS-T three times (5 min each) before incubation with either a HRP-conjugated rabbit anti-mouse IgG (1/10,000) or HRP-conjugated goat anti-rabbit IgG (1/30,000) in blocking solution at 20 °C during 2 h. After washing the PVDF membranes with PBS-T three times (5 min each), the immunoreactive bands were developed using a chemiluminescent detection kit (Thermo Fisher Scientific) and detected with an Amersham Imager 600 (GE Healthcare Europe, Barcelona, Spain).

Statistical Analysis
The number of independent experiments (n) in each group is indicated in figure legends. Data are represented as mean ± standard error of mean (SEM). Outliers were assessed by the ROUT method [43]; thus, subjects were excluded assuming a Q-value of 1% in GraphPad Prism 9 (San Diego, CA, USA). Data normality was assessed by the Shapiro-Wilk normality test (p < 0.05). When two groups were evaluated, unpaired Student's t-test or Mann-Whitney U-test was used. Comparisons among more than two experimental groups were performed by one-, two-or three-way factor analysis of variance (ANOVA) followed by either Dunnett's, Šídák's or Tukey post hoc test using GraphPad Prism 9, as indicated in the figure legends. A p-value ≤ 0.05 was considered significant.

BRET. 2 Experiments: Transient Co-expression of A 2A R with D 2 R and mGluR 5 Had a Significant Impact on D 2 R-mGluR 5 Heteroreceptor Complex Formation
HEK293T cells were transiently transfected with constant amounts of D 2 R Rluc and increasing amounts of plasmids encoding for mGluR 5 GFP2 with/without transient coexpression of A 2A R. The transient co-expression of A 2A R with D 2 R Rluc and mGluR 5 GFP2 had a significant impact on D 2 R Rluc -mGluR 5 GFP2 heteroreceptor complex formation (Fig. 1A). Transient co-expression of A 2A R promoted a significant increase of netBRET 2 max ratio value (0.084 ± 0.003 AU) compared to that found in cells without transient coexpression of A 2A R (0.043 ± 0.002 AU) (Fig. 1B). When the A 2A R was coexpressed with D 2 R Rluc and mGluR 5

GFP2
, these receptors hence showed an increased ability to heteromerize.
Also, saturation assay was used to compare the relative affinity of receptors for each other and their probability to form a complex, the so-called BRET50, which represents the acceptor/donor ratio giving 50% of the maximal signal. The netBRET 2 50 ratio value for D 2 R Rluc -mGluR 5 GFP2 heteromerization was significantly reduced by transient co-expression of A 2A R from (1.58 ± 0.09 AU) to (0.94 ± 0.11 AU) (Fig. 1C) indicating increased affinity of the two receptor protomers for each other. Pairs with low BRET 50 value thought to form oligomers or an increased tendency to dimerize, while high BRET 50 values indicate weak interaction or the absence of interaction between the investigated receptors.

Proximity Ligation Assay Experiments: Transient Co-expression of A 2A R Promoted the Formation D 2 R-mGluR 5 Heteroreceptor Complexes in HEK Cells
The role of A 2A R in the dynamics of the D 2 R-mGluR 5 heteromers was also evaluated by in situ proximity ligation assays (PLA) in transiently co-transfected HEK293T cells. The PLA results were in line with the results from the BRET 2 assays. The in situ PLA demonstrated the existence of D 2 R-mGluR 5 heteroreceptor complexes in cells to a low degree without transient co-expression of A 2A R (Fig. 2A). Furthermore, the transient co-expression of A 2A R highly significantly promoted the formation D 2 R-mGluR 5 heteroreceptor complexes as shown by the marked increase in the number of PLA-positive D 2 R-mGluR 5 complexes, while this was significantly reduced in HEK293T cells without coexpressing A 2A R (Fig. 2 B and D). Few and weak PLA clusters were detected in the PLA-negative controls (lack of D 2 R antibodies) representing background labelling (Fig. 2C).
The specificity of the PLA-positive D 2 R-mGluR 5 complexes, shown as red blobs in the mouse dorsal striatum (Fig. 3A), was demonstrated using D 2 R −/− mice (Fig. 3C). Results were expressed as mean ± SEM (n = 4 independent experiments). ****p < 0.0001 and **p < 0.01, Student's t-test In the sections from the mouse striatum, the appearance of the red PLA-positive D 2 R-GluR 5 complexes, shown as mean number of red blobs/Nucleus, was markedly and highly significantly reduced (Fig. 3D). Furthermore, the loss of the red D 2 R-mGluR 5 blobs to the same high degree in the A 2A R −/− mice (Fig. 3 B, D) likely reflects the requirement of D 2 R-mGluR 5 heterocomplexes to be part of an A 2A R-D 2 R-mGluR 5 to be expressed in the mouse striatum, probably by dorsal striatal-pallidal GABAergic neurons. In this way, it forms D 2 R-mGluR 5 complexes that are close enough to be visualized by PLA.

[. 3 H]-Raclopride/Quinpirole Competition Experiments: the A 2A R and mGluR 5 Protomers Interact and Modulate D 2 R Protomer Recognition
In HEK293T cells expressing D 2 R and mGluR 5 , the mGluR 5 agonist CHPG (500 nM) reduced the affinity of the highaffinity state (K i, High ) of the D 2 R for the agonist quinpirole with no effects on its low-affinity state (K i , Low ). Co-treatment with A 2A R agonist CGS-21680 (100 nM) did not significantly alter the D 2 R K i, High and K i, Low values obtained when the cells were treated only with CHPG (500 nM) ( Fig. 4A and Table 1). In HEK293T cells expressing A 2A R, D 2 R and mGluR 5 , mGluR 5 agonist stimulation also reduced the affinity of the high-affinity state (K i, High ) of the D 2 R for the agonist quinpirole with no statistically significant effects on its low-affinity state (Ki, Low ) ( Fig. 4A and Table 2). However, the transient co-expression of A 2A R by itself (without agonist stimulation) potentiates mGluR 5 agonist effects on the high-affinity D 2 R agonist binding sites (Fig. 4B, Tables 1  and 2). Finally, the co-stimulation of A 2A R and mGluR 5 synergistically increased in the K i, High values of the D 2 R protomer upon co-expression of the A 2A R (Table 2). Nevertheless, in cells expressing A 2A R, D 2 R and mGluR 5 , further analysis should be performed to test the effect of combine treatment of A 2A R (ZM-241385) and mGluR 5 (CHPG) to figure out if the expression of A 2A R, without agonist stimulation and its corresponding constitutive activity, is responsible for increased in the K i, High values of the D 2 R protomer upon co-expression of the A 2A R.
In both HEK293T cells expressing D 2 R and mGluR 5 or A 2A R, D 2 R and mGluR 5 , the incubation with A 2A R antagonist ZM-241385 (1 μM) and mGluR 5 antagonist MPEP (300 μM) alone or in combination resulted in an almost complete blockade of the mGluR 5 increase of the D 2 R K i, High values and A 2A R agonist-induced increase of mGluR 5 agonist effects on the high-affinity D 2 R agonist binding sites (Tables 1 and 2).

cAMP Functional Experiments: the A 2A R and mGluR 5 Protomers Interact and Modulate D 2 R Protomer Signalling
In cells expressing D 2 R and mGluR 5 forming D 2 R-mGluR 5 heterocomplexes (Fig. 2), the D 2 R agonist activation with quinpirole (100 nM) induced a G i protein-mediated inhibition of adenylyl cyclase that first was raised with 5 µM forskolin (Fig. 5A). This effect was highly significantly blocked by the D 2 R antagonist raclopride (1 μM). In these cells, the mGluR 5 agonist CHPG stimulation significantly counteracted the D 2 R agonist-induced reduction of cAMP accumulation (Fig. 5A). The significant effect of CHPG (500 nM) was significantly reduced by the mGluR 5 antagonist MPEP (300 μM). The cotreatment with the A 2A R agonist did not enhance the counteraction of the inhibitory D 2 R signalling by CHPG (Fig. 5A).
Likewise, quinpirole significantly reduced the cAMP level in cells expressing A 2A R, D 2 R and mGluR 5 (Fig. 5B). The mGluR 5 agonist CHPG had an improved ability to counteract the adenylyl cyclase inhibition produced by the D 2 R agonist in these cells, yielding cAMP levels similar to those obtained after blocking D 2 R signalling with raclopride (Fig. 5B). Upon A 2A R and mGluR 5 agonist co-activation, a larger counteraction of the D 2 R agonist action was found compared to that obtained with such a co-treatment performed in cells expressing only D 2 R It should be noted that CHPG agonist produces similar increases in cAMP levels as found after the A 2A R agonist GGS in HEK293T cells co-expressing D 2 R, A 2A R and mGluR 5 (Fig. 5C). Therefore, we should consider also that mGluR 5 might simply activate Gs, inducing cAMP accumulation, independently of D2-Gi-induced inhibition of adenylate cyclase (Fig. 5 A−C).  Table 1 Values for quinpirole binding site affinities to the D 2 -likeR by [. 3 H]-raclopride/quinpirole competition assays in HEK293T cells transiently expressing D 2 R and mGluR 5 incubated with agonist(s) or/ and antagonist(s) as indicated K i, High , D 2 R high-affinity value and K i, Low , D 2 R low-affinity value. Data are means ± SEM; n = 4, each determination performed at least in triplicate. Statistical analysis was performed by one-way ANOVA followed by the Tukey post hoc test. ***(p < 0.001); significant increased compared to vehicle. § § §(p < 0.001); significant reduced compared to cells incubated with CHPG. † † †(p < 0.001); significant reduced compared to cells incubated with CHPG and CGS-21680  Data are means ± SEM; n = 4, each determination performed at least in triplicate. Statistical analysis was performed by one-way ANOVA followed by the Tukey post hoc test. ***(p < 0.001); significant increased compared to vehicle. § § §(p < 0.001); significant differences compared to cells incubated with CHPG. † † †(p < 0.001); significant reduced compared to cells incubated with CHPG and CGS-21680

Experiments on Haloperidol-Induced Catalepsy
Catalepsy is a nervous condition characterized by loss of muscle control and fixity of posture. It is considered a symptom of certain nervous disorders such as Parkinson's diseases and epilepsy [44]. It is also a characteristic symptom of cocaine withdrawal, as well as one of the features of catatonia. The catalepsy is mainly produced by haloperidol induced blockade of D2R complexes in the dorsal striatal-pallidal GABA neurons within the dorsal striatum [44][45][46]. These GABA neurons mediate motor inhibition, counteracted by D 2 R agonist-induced activation of the D 2 R homo-and heterocomplexes like the D 2 R-A 2A R or the D 2 R-mGluR 5 heterocomplexes [24,[47][48][49][50]. The D 2 R activation of the dorsal striatal-pallidal GABA neurons is also essential for maintenance of normal locomotor activity. The catalepsy induced by the D 2 R antagonist haloperidol was evaluated in 10-min time intervals from 60 to 90 min after the injection of haloperidol (Fig. 6). In wild-type mice, the mGluR 5 -negative allosteric modulator raseglurant produced in this time period a significant reduction of the catalepsy time which was in the order of 25% (Fig. 6). In contrast, such a reduction of catalepsy was not observed by raseglurant treatment of A 2A R −/− mice. Furthermore, in vehicle-treated A 2A R −/− animals, Results are expressed as means ± SEM; n = 4 independent experiments, each determination performed in quadruplicates. ***p < 0.001, **p < 0.01 and *p < 0.05, one-way ANOVA followed by Tukey's post hoc test compared with cells treated only with forskolin (control). † † †p < 0.001, † †p < 0.01 and †p < 0.05, one-way ANOVA followed by Tukey's post hoc test when compared with cells treated only with quinpirole; § § §p < 0.001, § §p < 0.01 and §p < 0.05, when compared with cells treated only with quinpirole plus raclopride; δp < 0.05, when compared to cells treated with quinpirole and CHPG. C CHPG and CGS21680-induced cAMP levels. HEK293T cells transiently expressing D 2 R and mGluR 5  the haloperidol-induced catalepsy was markedly reduced compared to that obtained in vehicle treated wild-type mice (Fig. 6).

Discussion
The field of dopamine D 2 Rs changed markedly with the discovery of many types of D 2 R homo-and heteroreceptor complexes in subcortical limbic areas as well as the dorsal striatum [4,16,40]. The results indicate that the D 2 R is a hub receptor [51] which interacts not only with many other GPCRs including dopamine isoreceptors but also with ion-channel receptors, receptor tyrosine kinases, scaffolding proteins and dopamine transporters [24,52,53]. Disturbances in several of these D 2 R heteroreceptor complexes may contribute to the development of brain disorders through changes in the balance of diverse D 2 R homo-and heteroreceptor complexes mediating the dopamine signal, especially to the ventral striato-pallidal GABA pathway [37,52,54]. Of high relevance was the discovery of A 2A R-D 2 R and A 2A R-mGluR 5 heteroreceptor complexes in native tissue [4,16,40,55,56]. Furthermore, the existence of the D 2 R-mGluR 5 heterodimers in the biomembranes of living cells was demonstrated by bimolecular fluorescence complementation experiments in cellular models [21]. Although when tested by FRET microscopy in tsA 201 cells, D 2 R did not associate with mGluR 5 [57]. Nevertheless, by combination of bimolecular fluorescence complementation and bioluminescence resonance energy transfer techniques, as well as the sequential resonance energy transfer technique, the occurrence of an A 2A R-D 2 R-mGluR 5 heteroreceptor complexes was observed in living cells. Furthermore, by coimmunoprecipitation, experiments validated the existence of an association of mGluR 5 , D 2 R and A 2A R in rat striatum homogenates [21]. Herein, we present new findings that further expand the understanding of A 2A R-D 2 R-mGluR 5 heteroreceptor complexes. Also, strong evidences which support that the expression of the A 2A R is necessary to facilitate the association of D 2 R and mGluR 5 in a complex.
Our new findings are that transient co-expression of A 2A R in HEK293T cells together with D 2 R Rluc and mGluR 5 GFP2 resulted in a significant and marked increase in the formation of the D 2 R-mGluR 5 heterodimer, a component of the A 2A R-D 2 R-mGluR 5 heterocomplex, based on the increase in the BRET 2 max values. Such an increase could be related to the development of an increased affinity of the two D 2 R and mGluR 5 protomers for each other due to allosteric changes related to the formation of the A 2A R-D 2 R-mGluR 5 complex. In line with this hypothesis, the BRET 2 50 values were significantly reduced for the D 2 R-mGluR 5 heteromeric component of this trimeric heteroreceptor complex.
These results are also supported by the demonstration with PLA that an increased density of PLA-positive D 2 R-mGluR 5 clusters was observed when A 2A R expression had been added to the cells compared to cells only expressing D 2 R and mGluR 5 . In agreement, in the mouse dorsal striatum, the D 2 R-mGluR 5   on each other to improve or facilitate the formation of such complexes in the dorsal striatum. The different results obtained on haloperidol-induced catalepsy in wild-type mice vs A 2A R −/− mice are of substantial interest since they can indicate a functional role of the A 2A R-D 2 R-mGluR 5 heteroreceptor complexes in the dorsal striatum as previously discussed [8,17]. There was a marked reduction in the haloperidol-induced catalepsy in the A 2A R −/− mice compared to wild-type mice. Thus, in the absence of the A 2A R, the D 2 R antagonist haloperidol appears to have a substantially reduced potency to block the D 2 R which can be caused by the loss of the antagonistic A 2A R-D 2 R interaction [9,58]. According to the current findings in cell lines, the D 2 R-mGluR 5 heterocomplexes should be also formed to a much lower degree in the absence of A 2A R in view of their dependency of A 2A R according to the PLA experiments performed. The counteraction of the D 2 R-mediated inhibitory actions on cAMP signalling by CHPG, a mGluR 5 agonist, was in our cell line also more effective in cells co-expressing beside D 2 R and mGluR 5 , also A 2A R.
It seems likely that the for mation of the A 2A R-D 2 R-mGluR 5 complex enhances the affinity of the D 2 R and mGluR 5 protomers for each other in this complex. It is of high interest that the biochemical binding experiments reveal that the mGluR 5 CHPG agonist-induced increase in D 2 R K i, High values becomes significantly higher in the A 2A R-D 2 R-mGluR 5 complex compared to the D 2 R-mGluR 5 complex despite the absence of A 2A R agonist exposure. Thus, although agonist activation of the A 2A R seems necessary to exert negative allosteric modulation of the D 2 R protomer agonist binding via heteroreceptor complexes, an increased constitutive activity of the A 2A R protomer could explain the above results.
As expected, the combined incubation with CHPG and CGS-21680 led to an even stronger increase in the D 2 R K i, High values of the A 2A R-D 2 R-mGluR 5 complex, demonstrating the impact of the A 2A R protomer on the D 2 R-mGluR 5 allosteric interactions, which can involve both constitutive and A 2A R agonist-induced inhibition of D 2 R agonist binding. Our findings represent one of the first examples of integrative activity within a higher-order heteroreceptor complex and show how one receptor (A 2A R) can substantially modulate the structure and recognition of a participating receptor heterodimer (D 2 R-mGluR 5 ) in such a trimeric receptor complex.
The pharmacological analysis of the A 2A R-D 2 R-mGluR 5 complex and its impact on cAMP levels indicated that the A 2A R can modulate the effects of the D 2 R-mGluR 5 interactions on cAMP signalling. It was found that when the A 2A R-D 2 R-mGluR 5 complex was likely to be formed through the expression also of the A 2A R, the mGluR 5 agonist had an increased ability to counteract the D 2 agonist-induced G i/o -mediated inhibition of the cAMP levels in comparison with the counteraction observed in the absence of A 2A R expression. The same was also true for the combined treatment with the mGluR 5 agonist CHPG and the A 2A R agonist CGS-21680 when the A 2A R was coexpressed. A stronger counteraction of the D 2 R-induced inhibition of the cAMP levels was observed when A 2A R expression was present.
Taken together, our work on cell lines gives strong indications that, in the A 2A R-D 2 R-mGluR 5 complex, the A 2A R protomer enhances the formation of the D 2 R-mGluR 5 component of the complex with enhanced inhibition of D 2 R agonist binding recognition and its G i/o -mediated cAMP signalling. The inhibitory effects by A 2A R and mGluR 5 on D 2 R recognition and signalling reveal a significant molecular integration in A 2A R-D 2 R-mGluR 5 complexes, likely formed also in the dorsal striatum. The A 2A R and mGluR 5 antagonists targeting the A 2A R-D 2 R-mGluR 5 complexes in dorsal striatum may reduce the haloperidol-induced catalepsy by removal of the A 2A R and mGluR 5 protomer-mediated allosteric inhibition of the D 2 R protomer. Understanding of the trimeric complexes formed by these GPCRs could provide novel strategies for development of drugs against neuropsychiatric and neurodegenerative diseases by targeting their antagonistic receptor-receptor interactions.

Declarations
Ethics Approval This study was performed in line with the principles of the Declaration of Helsinki. The animal protocol (no. 7085) was approved by the University of Barcelona Committee on Animal Use and Care. Animals were housed and tested in compliance with the guidelines provided by the Guide for the Care and Use of Laboratory Animals [33] and following the European Union directives (2010/63/ EU), the ARRIVE guidelines [34].

Consent to Participate
Not applicable. The current research work does not involve human subjects.

Consent for Publication
Not applicable. This manuscript does not contain individual personal's data in any form.

Conflict of Interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.