Interactions between Calmodulin, Adenosine A2A, and Dopamine D2 Receptors*

The Ca2+-binding protein calmodulin (CaM) has been shown to bind directly to cytoplasmic domains of some G protein-coupled receptors, including the dopamine D2 receptor. CaM binds to the N-terminal portion of the long third intracellular loop of the D2 receptor, within an Arg-rich epitope that is also involved in the binding to Gi/o proteins and to the adenosine A2A receptor, with the formation of A2A-D2 receptor heteromers. In the present work, by using proteomics and bioluminescence resonance energy transfer (BRET) techniques, we provide evidence for the binding of CaM to the A2A receptor. By using BRET and sequential resonance energy transfer techniques, evidence was obtained for CaM-A2A-D2 receptor oligomerization. BRET competition experiments indicated that, in the A2A-D2 receptor heteromer, CaM binds preferentially to a proximal C terminus epitope of the A2A receptor. Furthermore, Ca2+ was found to induce conformational changes in the CaM-A2A-D2 receptor oligomer and to selectively modulate A2A and D2 receptor-mediated MAPK signaling in the A2A-D2 receptor heteromer. These results may have implications for basal ganglia disorders, since A2A-D2 receptor heteromers are being considered as a target for anti-parkinsonian agents.

G-protein-coupled receptors are able to form homo-and hetero-oligomers with unique biochemical and functional characteristics (1)(2)(3)(4)(5)(6)(7), and they are easily detected in vitro by using biophysical techniques (8 -10). Heteromers of adenosine A 2A and dopamine D 2 receptors were one of the first G-proteincoupled receptor heteromers to be described (11). A close physical interaction between both receptors was shown using coimmunoprecipitation and co-localization assays (11) and fluorescence and bioluminescence resonance energy transfer (FRET 2 or BRET) techniques (12)(13)(14). At the biochemical level, two types of antagonistic A 2A -D 2 receptor interactions have been discovered that may explain the A 2A -D 2 receptor interactions described both at the neuronal and behavioral level (11,(15)(16)(17)(18). First, by means of an allosteric interaction in the receptor heteromer, stimulation of A 2A receptor decreases the affinity of D 2 receptor for their agonists (12). Second, the stimulation of the G i/o -protein-coupled D 2 receptor inhibits the cAMP accumulation induced by the stimulation of the G s/olf -proteincoupled A 2A receptor (11,17,18). In view of the well known role of dopamine in Parkinson disease, schizophrenia, and drug addiction, it has been suggested that the A 2A -D 2 receptor interactions in the central nervous system may provide new therapeutic approaches to combat these disorders (16,19).
An epitope-epitope electrostatic interaction between an Arg-rich epitope of the N terminus of the third intracellular loop (3IL) of the D 2 receptor and an epitope containing a phosphorylated Ser localized in the distal part of the C terminus of the A 2A receptor is involved in A 2A -D 2 receptor heteromer interface (14,20,21). The same Arg-rich epitope of the D 2 receptor is able to interact with CaM (22)(23)(24)(25). In the absence of phosphorylated residues, adjacent aspartates or glutamates, which are abundant in CaM, may also form non-covalent complexes with Arg-rich epitopes (26). Therefore, CaM can potentially convey a Ca 2ϩ signal to the D 2 receptor through direct binding to the 3IL of the D 2 receptor (22). Mass spectrometry data have shown that bovine CaM can form multiple non-covalent complexes with an Arg-rich peptide corresponding to the N-terminal region of the 3IL of the D 2 receptor (VLR-RRRKRVN) (24) as well as a peptide from the proximal C terminus of the A 2A receptor (24). This epitope, whose sequence is 291 RIREFRQTFR 300 in the human A 2A receptor, also contains several Arg residues. Since the suspected interaction between the A 2A receptor and CaM was awaiting confirmation by assays using complete proteins, the present study was undertaken to demonstrate the existence of interactions between the A 2A * This work was supported, in whole or in part, by the National Institutes of Health, NIDA, Intramural Research Program (to S. F. and A. S. W.). This work was also supported by Spanish Ministry of Education and Science Grants SAF2008-00146 (to E. I. C.) and SAF2006-05481 (to R. F.) and Fundació La Marató de TV3 Grant 060110 (to E. I. C.). The Proteomics Laboratory at CIMA belongs to the network of the Spanish National Institute of Proteomics Facilities, ProteoRed. 1 To whom correspondence should be addressed: CIMA Neurociencias, Avda.
receptor and CaM both in a recombinant protein expression cell system and in the brain. A proteomics approach was used for the discovery of protein-protein interactions between the A 2A receptor and CaM in rat brain, whereas BRET in transfected cells demonstrated a direct interaction between CaM and this receptor. Furthermore, by using BRET and sequential resonance energy transfer (SRET) techniques and analyzing MAPK signaling in transfected cells, evidence was obtained for CaM-A 2A -D 2 receptor oligomerization and a selective Ca 2ϩmediated modulation of A 2A and D 2 receptor function in the A 2A -D 2 receptor heteromer.

EXPERIMENTAL PROCEDURES
Animals-Male Sprague-Dawley rats (250 g) were obtained from Harlan (Barcelona, Spain). Procedures involving animals were in accordance with the guidelines established by the normative of the European Council (86/609/EEC). The experimental design was approved by the Ethical Committee for Animal Testing of the University of Navarra (060/07).
Membrane Protein Purification-Animals were killed by decapitation, and the brains were rapidly removed. To obtain P2 membrane proteins, the striatum was dissected out and homogenized with a glass-Teflon homogenizer in ice-cold 0.32 M sucrose, 10 mM Hepes (pH 7,4), 2 mM EDTA, and complete protease inhibitors (Roche Applied Science). After spinning at 1,000 ϫ g for 15 min, the supernatant was removed and centrifuged at 200,000 ϫ g for 15 min to yield the crude membrane pellet (P2). The pellet was resuspended in homogenization buffer and spun again at 200,000 ϫ g to yield washed crude membrane pellet (P2Ј).
Immunoprecipitation-The pellet containing membrane proteins (P2Ј) was solubilized in immunoprecipitation buffer (phosphate-buffered saline (pH 7.4), complete protease inhibitor mixture, and 1% Igepal Ca-630 (Sigma)) for 1 h at 4°C. Protein concentration was determined by the BCA method (Pierce). A total amount of 0.5 mg of protein was used for immunoprecipitation. Samples were pretreated with 30 l of protein A/G-agarose beads (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) with gentle rocking for 1 h at 4°C and centrifuged at 10,000 ϫ g to remove the beads. Membrane proteins were incubated with 5 g of polyclonal anti A 2A receptor (Affinity Bioreagents, bioNova Científica, Madrid, Spain) or with 5 g of normal rabbit IgG (Upstate, Charlottesville, VA) for 90 min, and then 30 l of Protein A/G beads were added and left for 16 h at 4°C with gentle rocking. Immunoprecipitated proteins were washed three times with immunoprecipitation buffer.
Proteomic Analysis-Immunoprecipitated proteins were analyzed in the Proteomic Laboratory of CIMA. Briefly, proteins were eluted by incubation with 100 mM glycine-HCl (pH 2.5) for 10 min at room temperature with gentle rocking. The samples were centrifuged at 10,000 ϫ g for 1 min, and the supernatant content was precipitated with 25% trichloroacetic acid for 20 min at 4°C. After spinning at 10,000 ϫ g for 15 min, the pellet was washed twice with precooled acetone, completely dried in a SpeedVac, and resuspended in 30 l of 100 mM ammonium bicarbonate. Proteins were reduced with 10 mM dithiothreitol for 30 min at 56°C, alkylated with 55 mM iodoac-etamide for 20 min in darkness, and finally digested with trypsin 1:50 (v/v) for 16 h at 37°C. Peptide mass fingerprinting was obtained from tryptic digests by liquid chromatography-electrospray ionization-tandem mass spectrometry analysis (27). A peak list was generated by ProteinLynx Global Server 2.1 (Waters, Milford, MA). A data base search was done with ProteinLynx Global Server 2.1 (Waters) and Phenyx for raw data and mzXML data, respectively. The data base used in this study was SwissProt.
Cell Culture-HEK-293T cells were grown in Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine, 100 units/ml penicillin/streptomycin, and 5% (v/v) heatinactivated fetal bovine serum (all supplements were from Invitrogen). Cells were maintained at 37°C in an atmosphere of 5% CO 2 and were passaged when they were 80 -90% confluent (i.e. approximately twice per week . CaM was subsequently subcloned into the pGFP 2 -C1 (BioSignal Packard, Montreal, Canada) using EcoRI and BamHI, resulting in the GFP 2 -CaM plasmid. The cDNA encoding the 5HT 2B R-YFP fusion protein was kindly provided by Dr. Irma Nardi (University of Pisa, Italy). Expression of constructs was tested by confocal microscopy (see "Results"), and the receptor functionality was tested by second messengers, ERK1/2 phosphorylation, and cAMP production, as described previously (12,14) (data not shown).
Transient Transfection and Protein Determination-HEK-293T cells growing in 6-well dishes were transiently transfected with the corresponding fusion protein cDNA by the polyethyleneimine (Sigma) method. Cells were incubated (4 h) with the corresponding cDNA together with polyethyleneimine (5 l/g cDNA of 10 M polyethyleneimine) and 150 mM NaCl in a serum-free medium. After 4 h, the medium was changed to a fresh complete culture medium. Forty-eight hours after transfection, cells were washed twice in quick succession in HBSS with 10 mM glucose, detached by scraping, and resuspended in the same buffer unless otherwise indicated. To control the cell number, sample protein concentration was determined using a Bradford assay kit (Bio-Rad), using bovine serum albumin as the standard protein. Cell suspensions (20 g of protein) in HBSS buffer containing 1.26 mM CaCl 2 or, when indicated, in Ca 2ϩ -free HBSS buffer were distributed into 96-well micro-plates; black plates with a transparent bottom were used for fluorescence determinations, whereas white plates were used for BRET and SRET experiments.
BRET Assays-HEK-293T cells were transiently co-transfected with the indicated amounts of plasmid cDNAs corresponding to the indicated fusion proteins (see figure legends). To quantify fluorescence proteins, cells (20 g of protein) were distributed in 96-well microplates (black plates with a transparent bottom), and fluorescence was read at 400 nm in a Fluo Star Optima Fluorimeter (BMG Labtechnologies, Offenburg, Germany) equipped with a high energy xenon flash lamp, using a 10-nm bandwidth excitation filter. Receptor fluorescence expression was determined as fluorescence of the sample minus the fluorescence of cells expressing protein-Rluc alone. For BRET measurements, the equivalent of 20 g of cell protein was distributed in 96-well microplates (Corning 3600, white plates; Sigma), and 5 M coelenterazine H (Molecular Probes, Inc., Eugene, OR) was added. After 1 min of adding coelenterazine H, the readings were collected using a Mithras LB 940, which allows the integration of the signals detected in the 485-nm short (440 -500 nm) and the 530-nm long (510 -590 nm) wavelength filters. To quantify receptor-Rluc expression, luminescence readings were performed after 10 min of adding 5 M coelenterazine H. The same protocol was used to determine BRET with membranes obtained from cells (48 h post-transfection) disrupted using a Polytron homogenizer (PTA 20 TS rotor, setting 3; Kinematica, Basel, Switzerland). Disruption was performed for three 5-s periods in 10 volumes of free Ca 2ϩ HBSS, pH 7.4, containing a proteinase inhibitor mixture (Sigma). Cell debris were eliminated, membranes were obtained by centrifugation at 105,000 ϫ g (40 min, 4°C), and the pellet was resuspended and recentrifuged under the same conditions. The net BRET is defined as (long wavelength emission/short wavelength emission) Ϫ Cf, where Cf corresponds to long wavelength emission/short wavelength emission for the Rluc construct expressed alone in the same experiment.
SRET Experiments-HEK-293T cells were transiently cotransfected with the indicated amounts of plasmid cDNAs, corresponding to the indicated fusion proteins (see the legend for Fig. 6). Using aliquots of transfected cells (20 g of protein), three different determinations were performed in parallel. First, we performed quantification of protein-YFP expression by determination of the fluorescence due to protein-YFP. Cells distributed into 96-well microplates (black plates with a transparent bottom) were read in a Fluostar Optima fluorimeter (BMG Labtechnologies, Offenburg, Germany) equipped with a high energy xenon flash lamp, using an excitation filter at 485 nm and 10-nm bandwidth emission filters corresponding to 510 nm (506 -515 nm) (Ch1) and 530 nm (527-536 nm) (Ch2). The contribution of the GFP 2 and YFP proteins alone to the two detection channels (spectral signature) (28) was measured in experiments with cells expressing only one of these proteins and normalized to the sum of the signal obtained in the two detection channels. The spectral signatures of the different receptors fused to either GFP 2 or YFP did not vary significantly from the determined spectral signatures of the fluorescent proteins alone. For protein-YFP expres-sion quantification, linear unmixing was done taking into account the spectral signature as described by Zimmermann et al. (28,29) to separate the two emission spectra; the sample fluorescence is the fluorescence calculated as described minus the fluorescence of cells expressing only protein-Rluc and protein-GFP 2 . Second, we performed quantification of protein-Rluc expression by determination of the luminescence due to protein-Rluc. Cells were distributed in 96-well microplates (Corning 3600), and luminescence was determined 10 min after the addition of 5 M coelenterazine H in a Mithras LB 940 multimode reader (Berthold Technologies, DLReady, Germany). Third, we performed SRET 2 measurements. Cells were distributed in 96-well microplates (Corning 3600), and 5 M DeepBlueC (Molecular Probes) was added. The SRET 2 signal was collected using a Mithras LB 940 reader with detection filters for short wavelength (400 nm (370 -450 nm)) and long wavelength (530 nm (510 -590 nm)). By analogy with BRET, net SRET is defined as (long wavelength emission/short wavelength emission) Ϫ Cf, where Cf corresponds to long wavelength emission/short wavelength emission for cells expressing protein-Rluc, protein-GFP 2 , and the other protein partner not fused to a fluorescence protein (similar values were obtained measuring Cf in cells expressing protein-Rluc only and protein-GFP 2 ). Linear unmixing was done for SRET 2 quantification, taking into account the spectral signature (28) to separate the two fluorescence emission spectra.
Immunostaining-For immunocytochemistry, transiently transfected HEK-293T cells were fixed in 4% paraformaldehyde for 15 min and washed with phosphate-buffered saline containing 20 mM glycine (buffer A) to quench the aldehyde groups. Then, after permeabilization with buffer A containing 0.2% Triton X-100 for 5 min, cells were treated with phosphate-buffered saline containing 1% bovine serum albumin. After 1 h at room temperature, protein-Rluc was labeled with the primary mouse monoclonal anti-Rluc antibody (1:100; Chemicon, Billerica, MA) for 1 h, washed, and stained with the secondary antibody Cy3 donkey anti-mouse (1:100; Jackson Immunoresearch Laboratories, West Grove, PA). CaMYFP was detected by its fluorescence properties. Samples were rinsed and observed in a Leica SP2 confocal microscope (Leica Microsystems, Mannheim, Germany).
ERK Phosphorylation Assay-Cells were grown in 25-cm 2 flasks to 80% confluence and cultured in serum-free medium for 16 h before the addition of any agent. Cells resuspended in HBSS buffer containing 1.26 mM CaCl 2 were treated or not with 1 M ionomycin for 3 min before the addition of the following agonists: A 2A receptor agonist CGS2168 (200 nM), the D 2 receptor agonist quinpirole (1 M), or a mixture of both ligands for 5 min. Cells were rinsed with ice-cold phosphate-buffered saline and lysed by the addition of 500 l of ice-cold lysis buffer (50 mM Tris-HCl, pH 7.4, 50 mM NaF, 150 mM NaCl, 45 mM ␤-glycerophosphate, 1% Triton X-100, 20 M phenyl-arsine oxide, 0.4 mM NaVO 4 , and protease inhibitor mixture). The cellular debris was removed by centrifugation at 13,000 ϫ g for 5 min at 4°C, and the protein was quantified by the bicinchoninic acid method using bovine serum albumin dilutions as a standard. To determine the level of ERK1/2 phosphorylation, equivalent amounts of protein (10 g) were separated by electrophoresis on a denaturing 7.5% SDS-polyacrylamide gel and transferred onto polyvinylidene difluoride membranes. The membranes were then probed with a mouse anti-phospho-ERK1/2 antibody (1:2500; Sigma). In order to rule out the possibility that the differences observed were due to the application of unequal amounts of lysates, polyvinylidene difluoride blots were stripped and probed with a rabbit anti-ERK1/2 antibody that recognizes both phosphorylated and nonphosphorylated ERK1/2 (1:40,000; Sigma). Bands were visualized by the addition of anti-mouse horseradish peroxidase-conjugated (Dako, Glostrup, Denmark) or anti-rabbit horseradish peroxidaseconjugated (Sigma) secondary antibodies, respectively, and SuperSignal West Pico Chemiluminescent Substrate (Pierce). Band densities were quantified with a LAS-3000 imaging system (Fujifilm), and the level of phosphorylated ERK1/2 isoforms was normalized for differences in loading using the total ERK protein band intensities. Quantitative analysis of detected bands was performed by Image Gauge version 4.0 software. One-way analysis of variance and Student's t test for unpaired samples were used for statistical comparisons.

Identification of an Interaction between CaM and the A 2A
Receptor-According to the CaM target data base (available on the World Wide Web), the sequence RIREFRQTFR, which is present in the proximal part of the C terminus of the A 2A receptor, has a high probability to bind to CaM. In fact, we recently demonstrated that a synthetic peptide with the same sequence forms non-covalent complexes with bovine CaM (24). A proteomics approach was then used to find in situ evidence of direct interactions between the A 2A receptor and CaM in the brain. Samples from rat striatal membranes were immunoprecipitated using an antibody directed against a synthetic peptide of the A 2A receptor C terminus, SHGDMGLPDVELLSHELK, which does not overlap with the putative CaM-binding motif. Co-immunoprecipitates were digested with trypsin, and the resulting peptides were analyzed by mass spectrometry, as described under "Experimental Procedures." Among other peptides, two corresponding to the sequence of rat CaM (P62161) were detected after co-immunoprecipitation with the specific polyclonal antibody but not with rabbit IgG ( Table 1). The significance of the hits was confirmed by the low p values for both the 13-mer and the 9-mer peptides, which collectively covered 15% of the rat CaM sequence (Table 1). These results suggest that CaM and the A 2A receptor may interact, but the existence of a third bridging protein cannot be ruled out. For this purpose, BRET assays were performed in cells transfected with the cDNAs for the A 2A RRluc and CaMYFP fusion proteins. The hyperbola obtained upon increasing the YFP/Rluc ratio (Fig. 1A) indicates that a specific interaction between CaM and the A 2A receptor can occur in living cells. Similar experiments were performed using D 2 RRluc and CaMYFP to confirm that these two proteins may establish direct molecular interactions (Fig. 1B). The specificity of the CaM and A 2A receptor or CaM and D 2 receptor interactions in HEK cells was confirmed by the nonspecific (linear) BRET signal obtained when assaying A 1 RRluc and CaMYFP (Fig. 1) Fig. 2A). Accordingly, in cells expressing A 2A R AAA Rluc and CaMYFP, a linear and negligible BRET was detected in the absence or in the presence of ionomycin (Fig. 2B).   (Fig.  3). Co-expression of CaM and any of the two receptors did not modify the localization of the receptor. On the other hand, a significant membrane localization of CaM was only observed when the protein was co-expressed with either A 2A or D 2 receptors (Fig. 3). The translocation of CaM to the plasma membrane did not take place when co-expressed with the A 1 receptor (Fig.  3), which is not able to establish molecular interactions with the protein (see above). Treatment of cells with a Ca 2ϩ -free buffer and EDTA or treatment with ionomycin did not modify colo-calization of CaM and A 2A or D 2 receptors in cell membranes (Fig. 3).
Effect of Ca 2ϩ Levels on the Interaction between CaM and A 2A or D 2 Receptors-BRET experiments were performed to give some insight into the role of Ca 2ϩ in the interaction of CaM with either A 2A or D 2 receptors. In cells co-expressing D 2 RRluc and CaMYFP, resuspended in a medium containing 1.26 mM CaCl 2 , ionomycin treatment led to significant decreases in BRET measurements obtained after periods of incubation of 2, 5, and 10 min (Fig. 4B). In cells co-expressing A 2A RRluc and CaMYFP, ionomycin treatment also led to significant decreases in BRET measurements at 2 and 5 min of preincubation. In contrast, with a longer period of incubation (10 min), the BRET signal for A 2A RRluc and CaMYFP increased significantly above that found in untreated cells (Fig. 4A). On the one hand, these data indicate that Ca 2ϩ is not needed for CaM to be able to interact with A 2A or D 2 receptors. On the other hand, the ionophore-induced increase in intracellular Ca 2ϩ concentration does not disrupt the oligomerization of CaM with the receptors but leads to conformational changes in the CaM molecule and/or the receptors that alter the distance between Rluc (fused to the C terminus of the D 2 or A 2A receptors) and YFP (fused to CaM). The observed effects are not due to the Ca ϩ2 toxicity or changes in the membrane structure, since they were also observed in a cell-free system (Fig. 4D). Moreover, ionomycin did not induce significant changes in BRET measurements in cells co-expressing A 1 RRluc and CaMYFP (Fig. 4C).

Interactions of CaM with the A 2A -D 2 Receptor
Heteromer-By means of BRET competition experiments performed with cells expressing A 2 RRluc and CaMYFP (with expression levels adequate to give a suboptimal BRET value), the increase in D 2 receptor expression (tested by Western blot experiments) did not modify the BRET signal due to A 2 RRluc and CaMYFP (Fig.  5A). Analogously, increasing the expression of CaM did not modify the BRET signal due to A 2 RRluc and D 2 RYFP (Fig. 5B). As indicated in the Introduction, CaM seems to interact with the Arg-rich epitope of the 3IL of the D 2 receptor that is also involved in A 2A -D 2 receptor heteromerization (see the Introduction). In fact, competition assays demonstrated that increasing amounts of A 2A receptor led to significant reduction in the BRET signal due to the interaction between the D 2 receptor and CaM (Fig. 5C). However, a differential effect was obtained depending on the isoform of the D 2 receptor (D 2S or D 2L ). On the one hand, the BRET signal between D 2L RRluc and CaMYFP increased with low quantities of A 2A receptor but decreased dose-dependently with higher quantities of A 2A receptor. On the other hand, using the pair D 2S RRluc-CaMYFP, a dose-dependent decrease in BRET was found. These findings correlate with the potential ability of CaM and A 2A receptors to bind to any of the two Arg-rich epitopes of the 3IL of the D 2L receptor (one of which is not present in the D 2S structure; see "Discussion"). The specificity of these effects was demonstrated by the independence of the BRET signal with increasing expression of the A 1 receptor (Fig. 5C).
Oligomerization of CaM, A 2A , and D 2 Receptors-To test the possible existence of CaM-A 2A -D 2 receptor oligomerization, the recently described SRET technique was used (29). In SRET 2 , the oxidation of an Rluc substrate triggers GFP 2 acceptor excitation by BRET 2 and subsequent energy transfer to a YFP FRET acceptor. SRET 2 was attempted using A 2A RRluc, CaMGFP 2 , D 2 RYFP, and DeepBlueC as Rluc substrate. SRET 2 would only occur if the two acceptor/donor pairs, A 2A RRluc/CaMGFP 2 and CaMGFP 2 / D 2 RYFP, are well oriented at a distance of less than 10 nm. A SRET saturation curve obtained by augmenting D 2 RYFP expression while maintaining the same A 2A RRluc/ CaMGFP 2 ratio is shown in Fig. 6. To obtain optimal results (see Ref. 29), cells were transfected with an amount of cDNA (0.6 g) for A 2A RRluc giving readouts of about 100,000 luminescent units, an amount of cDNA (0.4 g) for CaMGFP 2 giving about 6,000 fluorescent units, and increasing amounts of cDNA (0.5-6 g) for D 2 RYFP (giving a range of 500 -8,000 fluorescence units). From the saturation curve (Fig. 6), apparent SRET max (0.29 Ϯ 0.03) and apparent SRET 50 (0.010 Ϯ 0.006) values were calculated. To prove whether SRET was possible when substituting A 2A receptors by A 1 receptors, cells were transfected with 0.7 g of cDNA for  A 1 RRluc (100,000 luminescent units), 0.4 g of cDNA for CaMGFP 2 (6,000 fluorescent units), and 0.5-6 g of cDNA for D 2 RYFP (500 -9,000 fluorescence units). The possible formation of a trimer in which the D 2 receptor was substituted by the serotonin 5-HT 2B receptor was tested using cells transfected with 0.6 g of cDNA for A 2A RRluc, 0.4 g of cDNA for CaMGFP 2 , and 0.5-6 g of cDNA for the 5-HT 2B receptor (500 -9,000 fluorescence units). Using these two combinations of fusion proteins, a very low and non-saturating (linear) SRET was observed, which indicates that oligomerization  6 g) and CaMYFP (0.4 g) and increasing amounts of D 2 R cDNA as competitor (0.3-2.5 g), whose expression was monitored by Western blotting (bottom) mBU, milli-BRET units. B, cells were cotransfected with the cDNA corresponding to A 2A RRluc (0.6 g) and D 2L RYFP (1.5 g) and increasing amounts of cDNA of CaMGFP 2 as competitor (0.2-1.5 g), whose expression was monitored by measuring the GFP 2 fluorescence (1,000 -20,000 fluorescence units), as described under "Experimental Procedures." To determine the YFP fluorescence in BRET experiments, the spectral signature was considered to control the GFP 2 contribution to the detection channel (28,29). Cells were cotransfected with the cDNA corresponding to D 2L RRluc (1 g; C) or D 2S RRluc (0.8 g; D) and CaMYFP (0.4 g) and increasing amounts of cDNA of A 2A R as competitor (black bars) or A 1 R as negative control (white bars), whose expression was monitored by dot blotting (results not shown). At the top of each panel, a scheme depicts the expressed proteins in BRET competition assays.
of CaM, A 1 , and D 2 receptors or CaM, A 2A , and 5-HT 2B receptors does not occur.

Effects of Intracellular Ca 2ϩ on CaM-A 2A -D 2 Receptor
Oligomerization-BRET was measured in the presence and in the absence of ionomycin in cells co-transfected with A 2A RRluc and D 2 RYFP. The calcium ionophore led to qualitative changes in the BRET curves (Fig. 7A) similar to those in the experiments with cells co-expressing A 2A RRluc and CaMYFP (see Fig. 4A). Thus, the BRET signal for A 2A RRluc and D 2 RYFP decreased at shorter times of incubation with ionomycin, whereas it increased at longer times (Fig. 7A). These results suggest that Ca 2ϩ binding to CaM induces conformational changes in the A 2A receptor such that the distance between the donor Rluc (fused to the C terminus of the A 2A ) and the acceptor YFP is modified. Taking all of the results into account, the most probable scenario is the existence of a CaM-A 2A -D 2 receptor oligomeric complex, Ca 2ϩ binding to CaM being unable to disrupt A 2A -D 2 receptor heteromers. To further test this possibility, SRET 2 for A 2A RRluc-CaMGFP 2 -D 2 RYFP was measured in cells treated with ionomycin. As shown in Fig. 6B, a positive SRET 2 was detected in the presence of ionomycin with an apparent SRET max of 0.24 Ϯ 0.08 and an apparent SRET 50 of 0.051 Ϯ 0.003. Therefore, calcium binding to CaM led to a decrease in the apparent SRET max and an increase in the apparent SRET 50 values. CaM-A 2A -D 2 receptor oligomers seem to be stable complexes that are not disrupted by calcium binding to CaM.  (Fig. 8B), respectively. The well known antagonistic interaction between A 2A and D 2 receptors was enhanced upon treatment with ionomycin. Results similar to those in Fig. 8, A and B, were obtained if cells were expressing higher amounts of CaM (1 g of cDNA; results not shown), indicating that moderate amounts of exogenous CaM are already able to modulate receptor operation. In contrast, the effects of ionomycin were not observed in cells expressing only endogenous CaM (Fig. 8C). Therefore, the effects of ionomycin were dependent on the intracellular levels of CaM. Results similar to those in Fig. 8, A and B, were also obtained using cells in suspension (not adherent cells) (i.e. in the same conditions as those used for BRET and SRET experiments) (results not shown). Overall, the results suggest a correlation between structural changes in the oligomer and the modulation of its functionality exerted by Ca ϩ2 binding.

DISCUSSION
Ca 2ϩ plays an important role in the physiology of higher order organisms and is involved in the regulation of many cellular events. Various stimuli, such as membrane depolarization or binding of ligands to plasma transmembrane receptors, trigger Ca 2ϩ channel opening, which results in a significant influx of Ca 2ϩ into the cytosol. Then Ca 2ϩ -binding proteins act as sensors and mediators of the initial Ca 2ϩ signal. CaM is a highly conserved, soluble, intracellular 17-kDa Ca 2ϩ -binding protein, regarded as a major transducer of Ca 2ϩ signals in mammalian cells (30). CaM has been shown to bind directly to cytoplasmic domains of plasma membrane proteins, including G-proteincoupled receptors, such as opioid , serotonin 5-HT 1A , acetylcholine muscarinic, and dopamine D 2 receptors (21,(31)(32)(33). In the present work, we also provide clear evidence for the binding of CaM to adenosine A 2A receptors. It was demonstrated that the sequence RIREFRQTFR, which is present in the proximal part of the C terminus of the A 2A receptor, is the CaM binding domain in these receptors. Furthermore, we provide evidence for CaM-A 2A -D 2 receptor oligomerization and for a specific Ca 2ϩ -dependent modulation of A 2A -D 2 receptor heteromer function. Previous studies have shown that an Arg-rich domain of the N-terminal portion of the 3IL of the D 2 receptor is involved in the binding to G i/o proteins, CaM, and the A 2A receptor (14,20,22,24,34). The binding of CaM to the D 2 receptor disrupts G i/o coupling and, therefore, its ability to inhibit the activity of adenylyl cyclase (22). We here report competition of CaM and A 2A receptors for the same domain in the D 2 receptor. This competition was particularly evident when the short isoform of the D 2 receptor was assayed. The two isoforms of the D 2 receptor (short and long, D 2S and D 2L , respectively) are generated by alternative splicing (35). D 2L , the isoform used in most of the experiments of the present study, contains 29 additional amino acid residues within the middle portion of the 3IL. The Arg-rich epitope of the D 2 receptor (VLRRRRKRVN) that is able to interact with the A 2A receptor is common to both D 2 receptor isoforms and is localized in the N-terminal portion of their long 3IL (14,20). However, there is an additional Arg-rich epitope in the 3IL of D 2L (VNRRRVEAA), which is not present in D 2S and which can potentially interact with CaM or with the A 2A receptor (21). The VNR-  RRVEAA peptide interacts in vitro with CaM and with a phosphorylated Ser-containing epitope of the C terminus of the A 2A receptor that is involved in A 2A -D 2 receptor heteromerization (data not shown). In BRET competition experiments using D 2S RRluc-CaMYFP, the increase of A 2A receptor expression produced a significant decrease of the BRET signal. On the other hand, when co-transfecting the A 2A receptor, the BRET signal between D 2L RRluc and CaMYFP first increased, whereas higher expression levels led to a reduced BRET signal. It is possible that, in the D 2L receptor, CaM binds preferentially to the VLRRRRKRVN epitope and that increasing quantities of transfected A 2A receptor induce, first, translocation of CaM to the VNRRRVEAA epitope, thus leading to an enhancement of the BRET signal. Upon further increase of receptor expression, A 2A R may also bind to the VNRRRVEAA epitope, with the consequent displacement of CaM from the D 2L receptor. The results from BRET competition experiments demonstrating that increasing expression of D 2 receptor did not modify the BRET signal due to A 2 RRluc and CaMYFP and that increasing expression of CaM did not modify the BRET signal due to A 2 RRluc and D 2 RYFP were also compatible with the possibility of CaM-A 2A -D 2 receptor oligomerization with A 2A receptors able to bind simultaneously to CaM and to the D 2 receptor. This was demonstrated with the use of the recently introduced sequential SRET techniques, which allow the demonstration of direct interaction of three protein molecules (29).
The results not only indicated that CaM may interact with A 2A receptors in cells that express A 2A -D 2 receptor heteromers but also that the conformation and function of A 2A -D 2 receptor heteromers are modulated by Ca 2ϩ . In fact ionomycin, an ionophore able to increase intracellular Ca 2ϩ , did affect the energy transfer between A 2A RRluc and D 2 RYFP. Agonist-induced conformational modifications in a receptor heteromer can be demonstrated by using resonance energy transfer techniques (36). In the case of the A 2A -D 2 receptor heteromer and by means of the same techniques used in the present study, no energy transfer variations were found after exposure to A 2A or D 2 receptor agonists (12). Therefore, the ability of BRET techniques to detect Ca 2ϩ -triggered conformational changes within the CaM-A 2A receptor, CaM-D 2 receptor, and CaM-A 2A -D 2 receptor oligomers suggests that Ca 2ϩ exerts a strong CaM-dependent control of the function of A 2A and D 2 receptors and A 2A -D 2 receptor heteromers. Interestingly, Ca 2ϩ exerted a selective modulation of A 2A -D 2 receptor heteromermediated activation of the MAPK pathway. On the other hand, Ca 2ϩ binding to CaM did not individually modulate the receptor-mediated signaling toward the MAPK pathway, and Ca 2ϩ significantly modified ERK1/2 phosphorylation induced by activation of A 2A or of D 2 receptors in cells co-expressing the two receptors. Furthermore, the previously described (18) antagonistic interaction between A 2A and D 2 receptors was regulated by ionomycin-mediated variation in the intracellular calcium concentration.
The present study demonstrates the existence of oligomeric intermolecular interactions between CaM, A 2A , and D 2 receptors. When not expressed together, A 2A or D 2 receptors may potentially bind CaM, but in cells expressing the A 2A -D 2 receptor heteromer, CaM binds preferentially to the A 2A receptor. Furthermore, CaM transduces Ca 2ϩ -dependent changes of MAPK signaling in the CaM-A 2A -D 2 receptor oligomer. This work adds new information about the function of A 2A -D 2 receptor heteromers, which are considered as a target for the development of anti-parkinsonian agents (37).