Biochemical Journal

Research article

A2A adenosine receptor ligand binding and signalling is allosterically modulated by adenosine deaminase

Eduard Gracia, Kamil Pérez-Capote, Estefanía Moreno, Jana Barkešová, Josefa Mallol, Carme Lluís, Rafael Franco, Antoni Cortés, Vicent Casadó, Enric I. Canela

Abstract

A2ARs (adenosine A2A receptors) are highly enriched in the striatum, which is the main motor control CNS (central nervous system) area. BRET (bioluminescence resonance energy transfer) assays showed that A2AR homomers may act as cell-surface ADA (adenosine deaminase; EC 3.5.4.4)-binding proteins. ADA binding affected the quaternary structure of A2ARs present on the cell surface. ADA binding to adenosine A2ARs increased both agonist and antagonist affinity on ligand binding to striatal membranes where these proteins are co-expressed. ADA also increased receptor-mediated ERK1/2 (extracellular-signal-regulated kinase 1/2) phosphorylation. Collectively, the results of the present study show that ADA, apart from regulating the concentration of extracellular adenosine, may behave as an allosteric modulator that markedly enhances ligand affinity and receptor function. This powerful regulation may have implications for the physiology and pharmacology of neuronal A2ARs.

  • adenosine deaminase
  • adenosine receptor
  • allosteric interaction
  • G-protein-coupled receptor
  • protein–protein interaction
  • receptor binding parameter

INTRODUCTION

Self-association of proteins to form dimers and higher-order oligomers and/or interaction with other proteins are key factors in cell signalling [13]. A paradigmatic example are adenosine receptors. The nucleoside adenosine exerts a modulatory action in many areas of the CNS (central nervous system) via its four GPCR (G-protein-coupled receptor) subtypes: A1Rs (adenosine A1 receptors) and A3Rs (adenosine A3 receptors) that are negatively coupled to the adenylate cyclase, and A2ARs (adenosine A2A receptors) and A2BRs (adenosine A2B receptors) that mediate the stimulation of adenylate cyclase activity [4]. Along the plasma membrane (horizontal plane), A1Rs and A2ARs may form homo-oligomers [57] and heteromers with other receptors [811], and the oligomerization generates new and unique biochemical and functional characteristics by modulating the binding properties, G-protein coupling and receptor trafficking [3,12,13]. Across the membrane (vertical to the plane of the membrane), A1Rs interact with intracellular proteins that are not directly involved in the signalling cascade, such as the Hsc73 (heat-shock cognate 73 stress protein), and this direct interaction is relevant for receptor function [14]. Also across the membrane, both A1Rs and A2BRs interact with a protein that has an extracellular topology, ADA (adenosine deaminase) [1518].

ADA is an enzyme involved in purine metabolism which catalyses the hydrolytic deamination of adenosine and 2′-deoxyadenosine to inosine or 2′-deoxyinosine and ammonia. Congenital defects of ADA lead to SCID (severe combined immunodeficiency), which is characterized by the absence of functional T- and B-lymphocytes in affected individuals [19,20]. Neurological abnormalities, which are less life threatening than immunological abnormalities, have also been described in a portion of patients [21]. Neurological alterations may be secondary to infections, or may be due to the accumulation of adenosine and derivatives in brain. Although the location of ADA is mainly cytosolic, it has been found on the cell surface of many cell types, including neurons [22]; therefore it can be considered as an ecto-enzyme [19]. Since ADA is a peripheral membrane protein it needs integral membrane proteins to be anchored to the membrane. Apart from A1Rs and A2BRs, another class of ecto-ADA-binding protein is CD26, a multifunctional transmembrane glycoprotein, acting as a receptor and a proteolytic enzyme [23]. It has been shown that ADA anchored to the dendritic cell surface, probably by the A2BR, binds to CD26 expressed on the surface of T-cells, triggering co-stimulation and enabling an enhanced immune response [2426].

We have also demonstrated that binding of enzymatically active or inactive ADA to A2BR increases its affinity and signalling by a protein–protein interaction [17]. In the case of A1Rs, the ADA–A1R interaction is very relevant since the enzyme potentiates signal transduction and modulates the desensitization of A1Rs [15,18,27]. Despite the well-established positive modulation exerted by ADA on A1Rs and A2BRs, it is not known whether the enzyme is able to modulate the A2AR subtype. There is currently a major interest in the ability of central A2ARs to control synaptic plasticity at glutamatergic synapses due to a combined ability of these receptors to facilitate the release of glutamate and the activation of NMDA; furthermore, A2ARs also control glial function and brain metabolic adaptation, and are important in controlling the demise of neurodegeneration [28]. In the present paper we report the molecular interaction between ADA and A2AR that results in ADA-induced conformational changes in the quaternary structure of A2ARs homodimers and in the pharmacological and functional characteristics of brain striatal A2ARs. A fine-tune regulation exerted by ADA probably has important implications for the physiology and pharmacology of neuronal A2ARs.

EXPERIMENTAL

Fusion proteins and expression vectors

The human cDNA for the A2ARs or GABAB2 (γ-aminobutyric acid B2) receptors cloned into pcDNA3.1 were amplified (removing stop codons) using sense and antisense primers harbouring either unique EcoRI or KpnI sites. The fragments were then subcloned to be in-frame with Rluc (Renilla luciferase) into the EcoRI and KpnI restriction site of an Rluc-expressing vector (pRluc-N1; PerkinElmer), or into the EcoRI and KpnI or BamHI restriction site of the variant of GFP (green fluorescent protein) (EYFP-N3; enhanced yellow variant of GFP; Clontech), to give the plasmids that express A2ARs or GABAB2 receptors fused to Rluc or YFP (yellow fluorescent protein) on the C-terminal end of the receptor (A2AR–Rluc, A2AR–YFP or GABAB2R–Rluc). As previously reported [9,11], when analysed by confocal microscopy, it was observed that all fusion proteins showed a similar membrane distribution as naïve receptors, and fusion of Rluc and YFP to A2ARs did not modify receptor function, as determined by cAMP assays.

Transient transfection

HEK-293T [HEK-293 cells expressing the large T-antigen of SV40 (simian virus 40)] cells were grown in DMEM (Dulbecco's modified Eagle's medium; Gibco) supplemented with 2 mM L-glutamine, 100 units/ml penicillin/streptomycin and 5% (v/v) heat-inactivated FBS (fetal bovine serum) (all supplements were from Invitrogen). HEK-293T cells growing in six-well dishes were transiently transfected with the corresponding fusion protein cDNA using the PEI (polyethylenimine; Sigma) method. Cells were incubated (for 4 h) with the corresponding cDNA together with PEI (5.47 mM nitrogen residues) and 150 mM NaCl in a serum-starved medium. After 4 h, the medium was changed to a fresh complete culture medium. At 48 h after transfection, cells were washed twice in quick succession in HBSS (Hanks balanced salt solution) with 10 mM glucose, detached and resuspended in the same buffer containing 1 mM EDTA. To control the cell number, the protein concentration of the sample was determined using the BCA (bicinchoninic acid) method (Pierce) using BSA dilutions as standards.

Generation of a CHO (Chinese-hamster ovary) cell clone expressing A2ARs

CHO cells were maintained at 37°C in an atmosphere of 5% CO2 in αMEM (α-minimal essential medium) without nucleosides (Invitrogen), containing 10% FBS, 50 μg/ml penicillin, 50 μg/ml streptomycin and 2 mM L-glutamine (300 μg/ml). CHO cells were transfected with the cDNA corresponding to human A2AR and cloned into a pcDNA3.1/Hygro vector with a hygromycin-resistance gene using the Lipofectamine™ (Invitrogen) method following the manufacturer's instructions. At 1 day after transfection, the selection antibiotic was added at a concentration that was previously determined using a selection antibiotic test. The antibiotic-resistant clones were isolated and cultured in six-well plates in the presence of the selection antibiotic. After an appropriate number of days/passages, a stable line expressing 6±1 pmol/mg of protein, with an affinity constant for the A2AR antagonist ZM 241385 of 1±0.3 nM, was selected and cultured in the presence of hygromycin (300 μg/ml).

BRET (bioluminescence resonance energy tranfer)

HEK-293T cells were co-transfected with 0.15 μg of cDNA corresponding to A2AR–Rluc acting as a BRET donor, and increasing amounts of cDNA corresponding to A2AR–YFP (0.8–3 μg of cDNA) acting as a BRET acceptor. As a negative control, HEK-293T cells were co-transfected with 0.15 μg of A2AR–Rluc and increasing amounts of cDNAs corresponding to the GABAB2–YFP receptor (0.3–3 μg of cDNA). After 48 h of transfection, the cell suspension (20 μg of protein) was dispensed in duplicate into 96-well black microplates with a transparent bottom (Porvair), and the fluorescence was measured using a Mithras LB940 fluorescence-luminescence detector (Berthold) with an excitation filter of 485 nm and an emission filter of 535 nm. For BRET measurement, 20 μg of cell suspension was distributed in duplicate into 96-well white opaque microplates (Porvair), and coelenterazine H (Molecular Probes) was added at a final concentration of 5 μmol/l. After 1 min the readings were collected in a Mithras LB 940 instrument which allows the integration of the signals detected in the short-wavelength filter at 485 nm (440–500 nm) and the long-wavelength filter at 530 nm (510–590 nm). The same samples were incubated for 10 min, and the luminescence was measured to quantify the donor. The BRET ratio is defined as: Embedded Image where Cf corresponds to (emission at 510–590)/(emission 440–500) for the A2A–Rluc construct expressed alone in the same experiment. Curves were fitted to a non-linear regression equation, assuming a single phase with GraphPad Prism software (San Diego, CA, U.S.A.).

Immunostaining

Wild-type CHO cells and A2AR-expressing CHO cells, grown on glass coverslips, were washed with PBS and fixed with 2% paraformaldehyde and 60 mM sucrose (pH 7.4) for 15 min at room temperature (25°C). Cells were washed twice with PBS containing 15 mM glycine, and treated with 1% BSA, 20 mM glycine and 0.05% sodium azide for 20 min before the addition of the antibodies. Then, cells were labelled for 45 min either with 100 μg/ml of the anti-A2AR antibody [14,29] or 50 μg/ml of the anti-ADA antibody [30], both conjugated with FITC as described previously [14]. Cells were washed with PBS containing 1% BSA, 20 mM glycine and 0.05% sodium azide, and placed on coverslips for the subsequent fluorescence microscopy analysis in a Leica TCS 4D confocal laser-scanning microscope (Leica Lasertechnik).

Brain striatal membrane preparation and protein determination

Sheep brains were obtained from the local slaughterhouse. Membrane suspensions from sheep brain striatum were prepared as described previously [31]. Tissue was disrupted with a Polytron homogenizer (PTA 20 TS rotor, setting 3; Kinematica) for three 5 s periods in 10 vol. of 50 mM Tris/HCl buffer (pH 7.4), containing a protease inhibitor cocktail (Sigma, 1:1000). After eliminating cell debris by centrifugation at 1000 g for 10 min, membranes were obtained by centrifugation at 35000 rev./min (40 min at 4°C; rotor type 90 Ti, Beckman) and the pellet was resuspended and recentrifuged under the same conditions. The pellet was stored at −80°C and was washed once more as described above and resuspended in 50 mM Tris/HCl buffer for immediate use. Protein was quantified using the BCA method (Pierce) using BSA dilutions as the standard.

Enzyme activity of ADA and ADA inhibition by Hg2+

Bovine ADA (Roche) enzyme activity was determined at 25°C with 0.1 mM adenosine as the substrate in 50 mM Tris/HCl buffer (pH 7.4). The decrease in the absorbance at 265 nm (Δϵ=7800 M−1·cm−1) was monitored in an Ultrospec 3300 pro spectrophotometer (Biochrom); 1 ml cuvettes with a 1 cm light pathlength were used. Hg2+-inactivation of bovine ADA was performed by a pre-incubation (2 h), of 15 units/ml desalted ADA with 100 μM HgCl2, and removal of free Hg2+ by gel filtration as described previously [16]. No residual activity was found after a 4 h incubation with 0.1 mM adenosine and a high excess (10 μg/ml) of inhibited enzyme in the conditions described above.

Radioligand-binding experiments

ADA dose-dependent curves were obtained by incubating (2 h) sheep brain striatal membrane suspensions (0.3 mg of protein/ml) with the indicated concentration of A2AR agonist [3H]CGS 21680 (42.7 Ci/mmol; PerkinElmer) or A2AR antagonist [3H]ZM 241385 (27 Ci/mmol; American Radiolabelled Chemicals) in the presence or the absence of the indicated amounts of desalted bovine ADA at 25°C in 50 mM Tris/HCl buffer (pH 7.4), containing 10 mM MgCl2.

Saturation experiments were performed by incubating striatal membrane suspensions (0.3 mg of protein/ml) with increasing concentrations of the A2AR antagonist [3H]ZM 241385 (triplicates of ten different concentrations, from 0.1 to 27 nM), at 25°C in 50 mM Tris/HCl buffer (pH 7.4), containing 10 mM MgCl2, in the absence or the presence of 0.2 i.u./ml (1 μg/ml) ADA.

Competition experiments were performed by incubating striatal membrane suspensions (0.3 mg of protein/ml) with a constant amount of [3H]CGS 21680 or [3H]ZM 241385 and it was increasing concentrations of CGS 21680 (triplicates of ten different concentrations from 1 nM to 10 μM; Tocris) or ZM 241385 (triplicates of 11 different concentrations, from 0.01 nM to 10 μM; Tocris) in the absence or presence of 0.2 i.u./ml (1 μg/ml) desalted ADA at 25°C in 50 mM Tris/HCl buffer (pH 7.4), containing 10 mM MgCl2, providing sufficient time to achieve equilibrium for the lowest radioligand concentration (5 h). In all experiments, non-specific binding was determined in the presence of 10 μM CGS 21680 or 10 μM ZM 241385 and it was confirmed that the value was the same as calculated by extrapolation of the competition curves. Free and membrane-bound ligand were separated by rapid filtration of 500 μl aliquots in a cell harvester (Brandel) through Whatman GF/C filters embedded in 0.3% PEI, which were subsequently washed for 5 s with 5 ml of ice-cold Tris/HCl buffer (pH 7.4). The filters were incubated with 10 ml of Ecoscint H scintillation cocktail (National Diagnostics) overnight at room temperature, and radioactivity counts were determined using a Tri-Carb 1600 scintillation counter (PerkinElmer) with an efficiency of 62% [14].

Binding-data analysis

Since A2ARs are expressed as dimers or higher-order oligomers [6,13], radioligand competition curves were analysed by non-linear regression using the commercial Grafit curve-fitting software (Erithacus Software), by fitting the specific binding data to the mechanistic two-state dimer receptor model [32,33]. This model considers a homodimer as the minimal structural unit of the receptor. To calculate the macroscopic equilibrium dissociation constants from saturation binding experiments the following equation previously deduced [34] was considered (eqn 1): Embedded Image(1) where A represents the free radioligand (the A2AR antagonist [3H]ZM 241385) concentration, RT is the total amount of receptor dimers, and KDA1 and KDA2 are the macroscopic dissociation constants describing the binding of the first and the second radioligand molecule to the dimeric receptor.

When binding of A to the dimer is non-co-operative, KDA2/KDA1=4 (see [32,33] for details) and, therefore, KDA1 is enough to characterize the binding. In this case, the above equation can be reduced to (eqn 2): Embedded Image(2) To calculate the macroscopic equilibrium dissociation constants from competition binding experiments the following equation previously deduced [34,35] was considered (eqn 3): Embedded Image(3) Here A represents free radioligand (the A2AR agonist [3H]CGS 21680 or the A2AR antagonist [3H]ZM 241385) concentration, B represents the assayed competing compound (CGS 21680 or ZM 241385) concentration, and KDB1 and KDB2 are, respectively, the macroscopic equilibrium dissociation constants of the first and second binding of B; KDAB is the hybrid equilibrium radioligand/competitor dissociation constant, which is the dissociation constant of B binding to a receptor dimer semi-occupied by A.

Binding to GPCRs can display negative co-operativity and in these circumstances KD2/KD1>4. On the other hand, for positive co-operativity, KD2/KD1<4 [34]. To measure the degree of co-operativity, the two-state dimer receptor model also introduces a co-operativity index (DC). The dimer co-operativity index for the radioligand A ([3H]ZM 241385) or the competing ligand B (CGS 21680 or ZM 241385) was calculated as [13,34,35] (eqn 4): Embedded Image(4) DC measures the affinity modifications occurring when a protomer senses the binding of the same ligand molecule to the partner protomer in a dimer. The way the index is defined is such that its value is ‘0’ for non-co-operative binding, positive values of DC indicate positive co-operativity, whereas negative values imply negative co-operativity [13,34,35].

In the experimental conditions when both the radioligand A ([3H]CGS 21680 or [3H]ZM 241385) and the competitor B (CGS 21680 or ZM 241385) show non-co-operativity (DC=0), it results that KDA2=4KDA1 and KDB2=4KDB1, and eqn (3) was simplified to (eqn 5): Embedded Image(5) When both the radioligand A ([3H]CGS 21680 or [3H]ZM 241385) and the competitor B are the same compound and the binding is non-co-operative, eqn (5) simplifies to (eqn 6): Embedded Image(6) Goodness of fit was tested according to a reduced χ2 value given by the non-linear regression program. The test of significance for two different population variances was based upon the F-distribution (see [32] for details). Using this F test, a probability greater than 95% (P<0.05) was considered the criterion to select a more complex equation to fit binding data over the simplest one. In all cases, a probability of less than 70% (P>0.30) resulted when one equation to fit binding data was not significantly better than the other. Results are given as parameter values±S.E.M. of three to four independent experiments.

ERK (extracellular-signal-regulated kinase) phosphorylation assay

A2AR-expressing CHO cells were cultured in serum-free medium for 16 h before the addition of any agent. Cells were treated (for 1 h at 37°C) with medium or the indicated concentration of ADA before the addition of the A2AR agonist CGS 21680 for a further incubation of 5 min. Cell were washed with ice-cold PBS and lysed by the addition of 500 μl of ice-cold lysis buffer [50 mM Tris/HCl (pH 7.4), 50 mM NaF, 150 mM NaCl, 45 mM 2-glycerophosphate, 1% Triton X-100, 20 μM phenyl-arsine oxide, 0.4 mM sodium orthovanadate and protease inhibitor cocktail]. Cell debris was removed by centrifugation at 13000 g for 5 min at 4°C and the protein was quantified using the BCA method using BSA dilutions as standards. To determine the level of ERK1/2 phosphorylation, equivalent amounts of protein (15 μg) were separated by electrophoresis on denaturing SDS/PAGE (10% gels) and transferred on to PVDF-FL membranes. Odyssey blocking buffer (LI-COR Biosciences) was then added, and membranes were rocked for 90 min. Membranes were then probed with a mixture of a mouse anti-(phospho-ERK 1/2) antibody (1:2500 dilution; Sigma) and rabbit anti-ERK 1/2 antibody (1:40000 dilution; Sigma) for 2–3 h. Bands were visualized by the addition of a mixture of IRDye 800 (anti-mouse) antibody (1:10000 dilution; Sigma) and IRDye 680 (anti-rabbit) antibody (1:10000 dilution; Sigma) for 1 h and scanned by the Odyssey IR scanner (LI-COR Biosciences). Bands densities were quantified using the scanner software and exported to Excel (Microsoft). The level of phosphorylated ERK1/2 isoforms was normalized for differences in loading using the total ERK protein band intensities.

RESULTS

ADA was anchored to the cell surface of A2AR-expressing cells

To investigate a potential direct interaction of ADA and A2ARs, wild-type CHO cells and a CHO–A2AR clone were selected, since CHO cells do not constitutively express adenosine receptors and since rodent CD26 endogenously expressed in CHO cells does not interact with ADA [36]. Parental CHO cells did not express A2ARs since they could not be labelled using a specific anti-A2AR antibody (Figure 1b). The CHO–A2AR clone showed a marked staining for A2AR (Figure 1a). ADA, which was detected in the cytoplasm using permeabilized CHO cells (results not shown), did not appear at the cell surface of parental CHO cells (Figure 1d). However, cell-surface ADA was detected in CHO–A2AR cells (Figure 1c), indicating that the ADA released to the cell culture may bind to the cell surface only in cells expressing A2ARs. These results indicate that the cell-surface A2AR behaved as an ADA-anchoring protein.

Figure 1 Expression of ADA on the cell surface of wild-type and A2AR-expressing CHO cells

Non-permeabilized wild-type CHO cells (b and d) or CHO–A2AR cell clone (a and c) were labelled with FITC-conjugated anti-A2AR antibody (a and b) or with FITC-conjugated anti-ADA antibody (c and d). Cells were processed for confocal microscopy analysis as described in the Experimental section.

ADA binding affected the quaternary structure of A2ARs

To investigate the consequences of the ADA–A2AR interaction, and taking into consideration that A2ARs are expressed as dimers or higher-order oligomers [6], the effect of ADA on the quaternary structure of A2AR–A2AR homomers was analysed by BRET experiments. Cells were co-transfected with 0.15 μg of the cDNA encoding A2AR–Rluc and increasing amounts of the cDNA corresponding to A2AR–YFP. At 48 h post-transfection, cells were treated (20 min at 37°C) with medium or with 1 μg/ml ADA in medium, and BRET was measured. In the absence of ADA, the hyperbola obtained upon increasing the acceptor expression indicated a specific interaction between the two fusion proteins (Figure 2). The BRETmax was 43±3 mBU and the BRET50 was 9±2. The specificity of the A2AR homomerization was confirmed by the unspecific (linear) BRET signal obtained in cells co-transfected with the cDNA corresponding to A2AR–Rluc and increasing amounts of the cDNA corresponding to GABAB2–YFP receptor (Figure 2). Interestingly, in the presence of ADA, a significant (P<0.01) increase in the BRETmax was observed (60±2 mBU) without significant alterations in BRET50 (9±1). These results can be interpreted in two ways. In one, ADA led to conformational changes in A2AR homomers that reduces the distance between Rluc and YFP fused to the C-terminal domain of the two A2AR-containing fusion proteins. In the other, ADA increases the receptor homomerization by increasing the affinity between protomers. In this last case, a decrease in the BRET50 values could be expected as there is binding between monomers to give homomers; BRET50 might represent the affinity between protomers. Since the BRET50 values were not changed in the presence of ADA we favour the first interpretation, that of ADA causing conformational changes.

Figure 2 Effect of ADA on A2AR homomerization detected by BRET experiments

BRET saturation experiments were performed as described in the Experimental section using cells transfected with 0.15 μg of cDNA corresponding to A2AR–Rluc and increasing amounts of cDNA corresponding to A2AR–YFP (0.8–3 μg of cDNA) (■ and ●) or to GABAB2–YFP receptor (0.3–3 μg of cDNA) as a negative control (▲). After 48 h of transfection, cells were treated for 20 min with medium (● and ▲) or with 1 μg/ml ADA (■) before BRET determination. Both fluorescence and luminescence for each sample were measured before every experiment to confirm similar donor expressions (approximately 120000 bioluminescence units) while monitoring the increase in acceptor expression (10000–50000 fluorescence units). The relative amount of BRET is given as the ratio between the fluorescence of the acceptor (YFP) and the luciferase activity of the donor (Rluc). BRET data are expressed as means±S.E.M. of three to four different experiments grouped as a function of the amount of BRET acceptor.

ADA modulated the agonist and antagonist binding to A2ARs

The effect of ADA on ligand binding to A2ARs was first determined using A2ARs expressed in a more physiological context. For this purpose striatal membranes, which express a high amount of A2AR, were selected. Isolated membranes were incubated with increasing concentrations of ADA and 17 nM of the radiolabelled A2AR agonist ([3H]CGS 21680, see the Experimental section). ADA enhanced in a dose-dependent manner the agonist binding to A2ARs (Figure 3a) with an EC50 value of 0.26±0.03 ng/ml, which approximately corresponds to 6 pM. To test whether the effect of ADA was independent of its enzymatic activity, a preparation containing an irreversible-inhibited enzyme was used. ADA was inactivated using a preparation containing 100 μM Hg2+; non-bound Hg2+ was removed by gel filtration prior to the assays (see the Experimental section). Membrane suspensions were incubated with 17 nM [3H]CGS 21680 in the absence or in the presence of 1 μg/ml of active or Hg2+-inactivated ADA. Both, active or Hg2+-inactivated ADA enhanced to a similar extent agonist binding to striatal A2ARs (Figure 3a, inset), thus demonstrating that the effect was independent of the enzyme activity and suggesting that, in our exhaustively washed membrane preparation, there is not enough endogenous adenosine to interfere with the ligand binding to receptors. ADA also enhanced the A2AR antagonist [3H]ZM 241385 binding to striatal membranes in a dose-dependent manner (Figure 3b) with an EC50 value of 0.13±0.06 ng/ml, which is approximately equivalent to 3 pM ADA. Purified BSA (1–10 nM) did not modify agonist or antagonist binding to striatal A2ARs, showing that the ADA effect was specific (results not shown). All of these results suggest that ADA is an allosteric modulator of A2ARs.

Figure 3. Effect of ADA on A2AR agonist and antagonist binding to brain striatal membranes

Binding of 17 nM [3H]CGS 21680 (a) or 1.6 nM [3H]ZM 241385 (b) to striatal membranes (0.3 mg of protein/ml) was performed as described in the Experimental section, in the presence of increasing concentrations of ADA. Data points on the y axis correspond to the binding in the absence of ADA. Inset in (a): 17 nM [3H]CGS 21680 binding in the absence (white bar) or in the presence of 1 μg/ml of active (grey bar) or Hg2+-inactivated (black bar) ADA was performed as described above. Data are means±S.E.M. (n=3). Significant differences with respect to the samples in the absence of ADA were calculated by an unpaired Student's t test (*P<0.05).

To further investigate the modulating effect of ADA on agonist and antagonist binding, the pharmacological parameters for ligand binding to A2ARs were calculated by means of saturation and competition experiments. To investigate the modulating effect of ADA on the A2AR antagonist equilibrium dissociation constants, brain striatal membranes were incubated with increasing concentrations of [3H]ZM 241385 in the absence or in the presence of 1 μg/ml ADA, and saturation experiments were performed as indicated in the Experimental section. Since A2ARs are expressed as dimers or higher-order oligomers [6], radioligand saturation curves were analysed by fitting the specific binding data to the mechanistic two-state dimer receptor model [32,33], which considers a homodimer as the minimal structural unit of the receptor. In the absence or in the presence of ADA, the saturation curves (Figure 4a) were monophasic (DC=0) according to the non-co-operative behaviour of ZM 241385 binding to A2ARs [35]. The resulting equilibrium constants from fitting data to eqn (2) were 4.6±0.8 nM and 1.9±0.4 nM in the absence or in the presence of ADA respectively (mean±S.E.M. of three different assays). This effect of ADA on antagonist affinity was also analysed by competition-binding experiments with 1.6 nM [3H]ZM 241385 and increasing concentrations of ZM 241385 in the absence or in the presence of 1 μg/ml ADA. In the absence or in the presence of ADA, the competition curves (Figure 4b) were also monophasic (DC=0). The resulting equilibrium constants from fitting data to eqn (6) were 5.1±0.7 nM and 3.3±0.8 nM in the absence or in the presence of ADA respectively (mean±S.E.M. of three different assays), not significantly different from saturation parameters. Thus ADA significantly (P<0.05) increased the affinity of A2ARs for the antagonist.

Figure 4 Effect of ADA on A2AR antagonist affinity constants

(a) Saturation binding experiments of increasing concentrations of the radiolabelled antagonist [3H]ZM 241385 (0.1–27 nM) or (b) competition experiments of the antagonist [3H]ZM 241385 (1.6 nM) binding against increasing concentrations of ZM 241385, in the absence (●) or in the presence (○) of 1 μg/ml ADA. Data are means±S.E.M. from a representative experiment (n=3) performed in triplicate.

To determine the modulating effect of ADA on the A2AR agonist CGS 21680 equilibrium dissociation constants, we only carried out competition-binding experiments since saturation experiments with a low-affinity ligand are not reliable. Radioligand binding was therefore determined in brain striatal membranes incubated with a constant amount of [3H]CGS 21680 (17 nM) and increasing concentrations of CGS 21680, in the absence or presence of 1 μg/ml ADA. As shown in Figure 5, competition curves of [3H]CGS 21680 against CGS 21680 were monophasic (DC=0) according to the non-co-operative behaviour expected for CGS 21680 binding [37]. The resulting equilibrium constant from fitting data to eqn (6) were 90±20 nM and 41±4 nM in the absence or in the presence of ADA respectively (mean±S.E.M. of three different assays). Thus ADA also significantly (P<0.05) increased the affinity of A2ARs for the agonist.

Figure 5 Effect of ADA on A2AR agonist affinity constants

Competition experiments of the agonist [3H]CGS 21680 (17 nM) binding against increasing concentrations of CGS 21680, in the absence (●) or in the presence (○) of 1 μg/ml ADA. Data are means±S.E.M. from a representative experiment (n=3) performed in triplicate.

Signalling consequences of the ADA–A2AR interaction

To investigate the functional consequences of the interaction of ADA with A2ARs, the A2AR-mediated signal transduction was determined in cells expressing the receptors. Accordingly, CHO–A2AR cells were treated for 5 min at 37°C with increasing amounts of the A2AR agonist CGS 21680 in the absence or presence of 1 μg/ml ADA, and ERK1/2 phosphorylation was determined as indicated in the Experimental section. In the absence of ADA, CGS 21680 up to 200 nM dose-dependently increased ERK1/2 phosphorylation followed by a decrease of signalling at high CGS 21680 concentrations (Figure 6). The phenomenon in which previous or continued exposure of receptor to agonist results in a diminished functional response of the receptor upon subsequent or sustained agonist treatment has been defined as desensitization [38]. It has been described that A2AR-mediated adenylate cyclase stimulation desensitizes rapidly in cultured cells (see [38] for a review). The results of the present study suggest that in A2AR-expressing CHO cells there is also a CGS 21680-promoted desensitization of ERK1/2 phosphorylation. In the presence of ADA, a significant increase in the CGS 21680-induced ERK1/2 phosphorylation was observed, resulting in a bell-shaped concentration–response curve (Figure 6). According to an ADA-induced increase in ligand affinity for A2ARs, ADA also increased the A2AR signalling, determined as ERK1/2 phosphorylation. These results show that ADA not only increased ligand affinity for A2ARs, but also was able to modulate, in a positive manner, signal transduction. ADA may then be considered an enhancer of ligand binding and of A2AR-mediated signalling events.

Figure 6 Effect of ADA on A2AR-mediated ERK1/2 phosphorylation

A2AR-expressing CHO cells were stimulated with increasing concentrations of the A2AR agonist CGS 21680 in the presence or in the absence of 1 μg/ml ADA. In (a) a representative Western blot is shown. In (b) values are means±S.E.M. of three independent experiments. Grey columns are in the presence of 1 μg/ml ADA, white columns are in the absence of 1 μg/ml ADA. Significant differences with respect to the samples in the absence of ADA were calculated by an unpaired Student's t test (*P<0.05 and **P<0.01).

DISCUSSION

Cell-surface ADA needs to be anchored to the plasma membrane by means of specific receptors. In the present paper we describe that ADA may bind to A2ARs on the surface of living cells. By FRET or BRET it has previously been demonstrated that A2ARs form homomers and that homomers, but not monomers, appear to be the functional species at the cell surface of transfected cells [6]. Thus the quaternary structure of A2ARs is constituted by, at least, two protomers that form a dimer. Probably resulting from a decrease in the distance between the C-termini of the A2AR protomers fused to Rluc and YFP, ADA binding led to modifications in the quaternary structure of A2AR homomers that could be detected by BRET experiments. Using a similar set up Canals et al. [6] showed that A2AR agonists are not able to modify the BRET signal. Therefore the ability of BRET to detect ADA-triggered conformational changes within the A2AR homomers suggests that ADA exerts a control of the function of A2AR homomers by a strong modification of their quaternary structure. In fact, the ADA-induced structural changes in the A2AR molecule correlated with marked affinity modifications in the binding of both agonist and antagonist. Irrespective of its enzymatic activity, ADA was able to significantly decrease agonist and antagonist equilibrium dissociation constants. The ADA-induced increase in the ligand affinities indicates that ADA behaved as a positive modulator of A2ARs.

In addition to orthosteric sites, many GPCRs have been found to possess structurally distinct allosteric domains. One characteristic feature of the allosteric interaction is that the receptor is able to simultaneously bind an orthosteric and an allosteric ligand, introducing complexity into pharmacological responses by modifying the affinity or the signal imparted by the orthosteric ligand [39]. An allosteric effect results in a positive modulation if the modulator facilitates the interaction, or in a negative modulation if it inhibits the interaction of the ligand with the orthosteric-binding site [39,40]. According to these concepts, ADA is an allosteric ligand of A2ARs that positively modulates the agonist and antagonist binding to the orthosteric site of the receptor. Kreth et al. [41] have shown that an endogenous allosteric modulator leads to a reduced ligand affinity and to an impaired function of the A2AR of human granulocytes in sepsis. Furthermore, some compounds have been synthesized and evaluated as positive enhancers of agonist and antagonist radioligands for the neuronal A2AR [42,43]. A2ARs are allosterically modulated by sodium ions binding to an allosteric site linked to Glu13 in TM1 (TM is transmembrane domain) and His278 in TM7, and by the potassium-sparing diuretic amiloride [4345]. The ability of allosteric modulators to fine-tune pharmacological responses has sparked interest in their potential applications in both clinical and basic science settings [40]. This interest is more relevant in the case of neurotransmitter receptor targets due to the fact that synaptic neurotransmission occurs in extremely complex circuits implicated in many neurological functions. Owing to the implication of A2ARs in many neurodegenerative diseases, such as Parkinson's and Huntington's disease, obsessive-compulsive disorders and drug addiction [46], different approaches have been tested to find allosteric modulators, i.e. a structure-based ligand-discovery methodology provided new routes for modulation of this neuronal key target [4749]. Conceptually the allosteric interaction described in the present study is different from the one exerted by small molecules since it comes from the interaction across the membrane with a protein that has an extracellular topology. By means of the interaction with an extracellular domain of A2ARs, ADA exerts a fine-tune modulation of adenosine neuroregulation that may have important implications for the function of neuronal A2ARs, which are enriched in and play a key role in the brain striatum. The presence of ADA bound to the cell surface of neurons has been demonstrated [22], reinforcing the concept that this allosteric effect of ADA is likely to occur in vivo. With this in mind one may hypothesize that ADA SCID patients with ADA mutations affecting the binding of ADA to A2AR may manifest neurological alterations that are predicted to be different from those resulting from mutations not affecting the ADA–A2AR interface. Probably, mutations affecting the interaction would be less deleterious for striatal function since it would attenuate overactivation of A2AR exerted by the elevated adenosine levels. Irrespective of this, the results described in the present study show that ADA, apart from reducing the adenosine concentration, binds to A2AR behaving as an allosteric effector that markedly enhances agonist-induced signalling thought to be the MAPK (mitogen-activated protein kinase) pathway, increasing ERK1/2 phosphorylation. Thus the physiological role of the ADA–adenosine receptor interaction is to make those receptors more functional.

AUTHOR CONTRIBUTION

Eduard Gracia, Carme Lluís, Antoni Cortés, Vicent Casadó, Rafael Franco and Enric Canela conceived and designed the experiments; Eduard Gracia, Kamil Perez-Capote, Estefanía Moreno, Jana Barkesová, Josefa Mallol, Antoni Cortés and Vicent Casadó performed the experiments; Eduard Gracia, Kamil Pérez-Capote, Estefanía Moreno, Jana Barkešová, Josefa Mallol, Carme Lluís, Enric Canela, Antoni Cortés and Vicent Casadó discussed and analysed data; Carme Lluís, Rafael Franco, Antoni Cortés, Vicent Casadó and Enric Canela wrote the paper.

FUNDING

This work was supported by the Spanish Ministerio de Ciencia y Tecnología [grant numbers SAF2008-00146, SAF2008-03229-E, SAF2009-07276]; and the Fundació La Marató de TV3 [grant number 060110].

Acknowledgments

We thank Jasmina Jiménez for technical help (Molecular Neurobiology Laboratory, Barcelona University, Barcelona, Spain).

Abbreviations: A2AR, adenosine A2A receptor; A2BR, adenosine A2B receptor; ADA, adenosine deaminase; BCA, bicinchoninic acid; BRET, bioluminescence resonance energy transfer; CHO, Chinese-hamster ovary; ERK, extracellular-signal-regulated kinase; FBS, fetal bovine serum; GABA, γ-aminobutyric acid; GFP, green fluorescent protein; GPCR, G-protein-coupled receptor; HEK-293T, HEK-293 cells expressing the large T-antigen of SV40 (simian virus 40); PEI, polyethylenimine; Rluc, Renilla luciferase; SCID, severe combined immunodeficiency; TM, transmembrane domain; YFP, yellow fluorescent protein

References

View Abstract