Address correspondence and reprint requests to M. T. Tebano, Section of Central Nervous System Pharmacology, Department of Therapeutic Research and Medicines Evaluation, Istituto Superiore di Sanità, Viale Regina Elena, 299, 00161 Rome, Italy. E-mail: firstname.lastname@example.org
Adenosine A2A, cannabinoid CB1 and metabotropic glutamate 5 (mGlu5) receptors are all highly expressed in the striatum. The aim of the present work was to investigate whether, and by which mechanisms, the above receptors interact in the regulation of striatal synaptic transmission. By extracellular field potentials (FPs) recordings in corticostriatal slices, we demonstrated that the ability of the selective type 1 cannabinoid receptor (CB1R) agonist WIN55,212-2 to depress synaptic transmission was prevented by the pharmacological blockade or the genetic inactivation of A2ARs. Such a permissive effect of A2ARs towards CB1Rs does not seem to occur pre-synaptically as the ability of WIN55,212-2 to increase the R2/R1 ratio under a protocol of paired-pulse stimulation was not modified by ZM241385. Furthermore, the effects of WIN55,212-2 were reduced in slices from mice lacking post-synaptic striatal A2ARs. The selective mGlu5R agonist (RS)-2-chloro-5-hydroxyphenylglycine (CHPG) potentiated the synaptic effects of WIN55,212-2, and such a potentiation was abolished by A2AR blockade. Unlike the synaptic effects, the ability of WIN55,212-2 to prevent NMDA-induced toxicity was not influenced by ZM241385. Altogether, these results show that the state of activation of A2ARs regulates the synaptic effects of CB1Rs and that A2ARs may control CB1 effects also indirectly, namely through mGlu5Rs.
Type 1 cannabinoid receptor (CB1R), is one of the most abundant G protein-coupled receptor in the brain (Herkenham et al. 1990, 1991). Endocannabinoids, the endogenous ligands of cannabinoid receptors [2-arachidonoylglycerol (2-AG) and anandamide], are released on demand from post-synaptic neurons and act as retrograde messengers on pre-synaptic CB1Rs (Alger 2002; Wilson and Nicoll 2002; Chevaleyre et al. 2006). By coupling to Gi proteins, CB1R activation inhibits neurotransmitter release (Freund et al. 2003; Piomelli 2003), thus influencing synaptic transmission in different brain areas. Although an exclusive pre-synaptic expression of striatal CB1Rs has been suggested by some studies (Matyas et al. 2006; Uchigashima et al. 2007), the application of a fractionation method has revealed that CB1Rs are also expressed in the post-synaptic density (Köfalvi et al. 2005).
Besides the direct A2A/CB1 interaction, there is at least another mechanism by which A2ARs may regulate CB1 effects, namely through the modulation of the metabotropic glutamate 5 receptors (mGlu5Rs). It is known, indeed, that the activation of striatal group-I mGlu5Rs induces endocannabinoid release. In striatal slices, this effect was proven to depend on the mGlu5R subtype, as the ability of dihydroxyphenylglycine (DHPG) to stimulate the biosynthesis of 2-AG was prevented by the selective mGlu5R antagonist 2-methyl-6-(phenylethynyl)pyridine hydrochloride (MPEP) (Jung et al. 2005).
It is also well known that striatal mGlu5Rs co-localize, constitute heteromeric complexes and functionally interact with adenosine A2ARs (Ferréet al. 2002). Worth of note, some of the effects exerted by mGlu5Rs, such as DARPP32 phosphorylation (Nishi et al. 2003) and potentiation of NMDA-mediated effects (Domenici et al. 2004), are tightly controlled by the state of activation of A2ARs.
The aim of the present work was to investigate whether, and by which mechanisms, the synaptic effects of CB1Rs could be influenced by the state of activation of striatal A2ARs.
Materials and methods
Male and pregnant female Wistar rats (2–3 months old, Harlan, Italy) were used.
A2AR KO (A2A−/−) and their wild-type (WT) littermate mice (A2A+/+) were generated on a CD1 background as previously described (Ledent et al. 1997). The Cre-loxP strategy was used to generate striatum-specific A2AR knockout (st-A2AR KO) mice as previously described (Bastia et al. 2005). Briefly, homozygous floxed (A2ARflox+/+) mice (F5 generation in mixed 129-Steel and C57BL/6 background) were crossbred with Dlx5/6-Cre transgenic mice expressing Cre recombinase under control of Dlx5/6 promoter, which is active exclusively in striatal neurons during development, to generate st-A2AR KO [Dlx5/6-Cre(+)A2ARflox+/−] mice. Genotyping was conducted by three primer PCR analysis of tail DNA.
The animals were kept under standardized temperature, humidity and lighting conditions and had free access to water and food. All animal procedures were carried out according to the European Community Guidelines for Animal Care, DL 116/92, application of the European Communities Council Directive (86/609/EEC).
Extracellular recordings from corticostriatal slices
Animals, rats and mice, were decapitated under ether anaesthesia, and coronal corticostriatal slices (300 μm) cut with a vibratome. Slices were maintained at room temperature (22–25°C) in artificial cerebrospinal fluid (ACSF) containing 126 mM NaCl, 3.5 mM KCl, 1.2 mM NaH2PO4, 1.2 mM MgCl2, 2 mM CaCl2, 25 mM NaHCO3 and 11 mM glucose (pH 7.3) saturated with 95% O2 and 5% CO2. After incubation in ACSF for at least 1 h, a single slice was transferred to a submerged recording chamber and continuously superfused at 32–33°C with ACSF at rate of 2.7–3 mL/min. The drugs were added to this superfusion solution.
Extracellular field potentials (FPs) were recorded in the dorsomedial striatum with a glass microelectrode filled with 2 M NaCl solution (pipette resistance 2–5 MΩ) and evoked at the frequency of 0.05 Hz by the stimulation of the white matter between the cortex and the striatum with a bipolar twisted NiCr-insulated electrodes (50 μm o.d. duration 100 us). One stimulus was delivered every 20 sec. Three consecutive responses were averaged. Signals were acquired with a DAM-80 AC differential amplifier (WPI Instruments, Woltham, MA, USA) and analysed with the LTP software (Anderson and Collingridge 2001).
At least 10 min of stable baseline recording preceded drug application. To allow comparisons between experiments, in each experiment the amplitude values were normalized, taking the average of the values obtained over the 10-min period immediately before applying the test compound as 100%. In each experiment, the mean basal FP amplitude (i.e. the mean from values obtained over the 10-min period immediately before drug application) was calculated and effects of the drugs were expressed as percentage variation with respect to basal values. The washout period lasted 30 min.
The influence of drugs on neurotransmitter release was studied using a Paired-Pulse Stimulation (PPS) protocol in which two consecutive pulses are applied 50-msec apart.
In control condition, this protocol elicits a Paired-Pulse Facilitation (PPF), in which the response elicited by the second stimulus (R2) is greater than that elicited by the first stimulus (R1). The degree of PPF is quantified by the R2/R1 ratio and a modification of this ratio is an indication of a pre-synaptic action on neurotransmitter release (Manabe et al. 1993; Schulz et al. 1994; Calabresi et al. 1997).
Data were expressed as mean ± SEM from N slices. Mann–Whitney U-test was used for statistical analysis of the data.
Western blot analysis
The expression of CB1Rs in the striatum of A2AR KO and their WT littermate mice was evaluated by western blotting experiments. Mice were decapitated under ether anaesthesia and the striata immediately dissected and stored at −80°C. Tissues were homogenated on ice in Ripa buffer [phosphate-buffered saline, 1% NP-40 (Sigma-Aldrich, Milan, Italy), 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulphate, 0.1 mM phenylmethylsulfonyl fluoride, 10 μg/mL aprotinin, 10 μg/mL pepstatin A, 10 μg/mL leupeptin and 10 μg/mL sodium orthovanadate] using a Micro Homogenizer (Omni International, International PBI, Milan, Italy). Tissue lysates were centrifuged at 12 000 g for 20 min at 4°C. Supernatants were removed; some aliquots of them were used for protein determination (Protein Assay Kit; BioRad, Hercules, CA, USA). Fifty micrograms of protein was diluted with Laemmli sample buffer 10x, boiled for 5′ at 100°C and separated by 8% sodium dodecyl sulphate–polyacrylamide gel electrophoresis gels. Proteins were transferred to polyvinylidene difluoride membranes (BioRad) by electroblotting overnight at 4°C. To avoid non-specific immunodetection, membranes were incubated for 1 h in T-TBS (50 mM Tris–HCl, 150 mM NaCl, 0.05% Tween-20, pH 7.4) containing 5% non-fat milk. Blots were then incubated overnight at 4°C with a rabbit anti-CB1 antibody (1 : 500 dilution, PA1-745; Affinity Bioreagents, Florence, Italy) and with a mouse anti-β-actin (1 : 3000 dilution; Chemicon International, Temecula, CA, USA) diluted in T-TBS containing 5% non-fat milk. After extensive washes in T-TBS, immunoreactive bands were detected by incubation with horseradish peroxidase-conjugated secondary antibodies (Chemicon International) and revealed by enhanced chemiluminescent substrate (Thermo scientific, Rockford, IL, USA) onto X-ray films (GE Healthcare, Buckinghamshire, UK). Densitometric analysis was conducted using the ImageJ64 program (http://rsbweb.nih.gov/) onto six different experiments. CB1-immunoreactive bands were normalized with respect to β-actin and statistical analysis was performed using an unpaired t-test.
Primary striatal cultures
Neurones were isolated from 17- to 18-day-old embryos. Pregnant rats were ether anaesthetised, decapitated and the foetuses collected and rapidly decapitated. After removal of the meninges, the striata were collected in Hank’s balanced salt solution and dissociated. Striatal cells were then washed in Hank’s balanced salt solution and resuspended in Neurobasal medium supplemented with 0.5 mM l-glutamine, 2% B-27 supplement and gentamicin (50 μg/ml) (referred as complete medium). Aliquots of 2–3 × 105 cells were placed in 24-well cultures plates coated with poly-l-lysine (5 μg/ml) and maintained at 37°C in humidified air with 5% CO2.
Every 4 days, 0.5 ml of medium was removed and replaced by the same volume of fresh complete medium. Assays were done on 12–14 days old cultures. At the time of the experiment, culture medium was removed and substituted by an appropriate volume of Neurobasal medium supplemented with 0.5 mM glutamine and gentamicin (50 μg/ml). Under these experimental conditions, the culture consists predominantly of neurons as previously described (Brewer et al. 1993).
After exposure of cultured cells to the drugs, culture medium was removed and replaced with fresh complete medium. Cultures were then returned to the incubator and cellular damage was evaluated, 24 h later, by measuring the amount of lactate dehydrogenase (LDH) released into the medium using a cytotoxicity detection kit (Roche Diagnostic, Indianapolis, IN, USA). The reaction was run at 22–23°C with light protection for 30 min.
Results are expressed as mean ± SEM of original values of 3–4 independent experiments, assayed in triplicate. anova one way analysis, followed by Neuman-Keuls multiple comparison test, was used for the statistical analysis.
Measurement of [Ca 2±]i levels by Fura-2AM
Striatal cells, obtained as above, were plated in complete medium with a density of 30 000–40 000/cm2. Experiments were started 13–15 days after plating. Optical fluorimetric recordings with Fura-2 acetoxymethyl ester (Fura-2AM) were used to evaluate the intracellular calcium concentration ([Ca2+]i]). Fura-2 AM stock solutions were obtained by adding 50 μg of Fura-2AM to 50 μl of 75% dimethylsulfoxide + 25% pluronic acid. Cells were bathed for 60 min at 22–23°C with 5 μl stock solution diluted in 1 ml of extracellular solution (composition in mM: NaCl 125; KCl 1; CaCl2 5; MgCl2 1; glucose 8; HEPES 20; pH 7.35). This solution was then removed, replaced with extracellular solution, and the dishes were placed quickly on the microscope stage. Drugs were directly dropped in the bath. To measure fluorescence changes, a Hamamatsu Argus 50 computerized analysis system was used (Hamamatsu, Shizouka, Japan), recording every 6 sec the ratio between the values of light intensity at 340- and 380-nm stimulation. For the statistical analysis, in each experiment, the mean 340/380 value obtained over 1 min immediately before and 10 min after the application of the different drugs was calculated.
WIN55,212-2, AM251, ZM241385, CHPG, MPEP were purchased from Tocris Cookson (Bristol, UK) and NMDA was purchased from Sigma Aldrich (Milan, Italy).
Influence of A2ARs on CB1-dependent synaptic effects
A2AR activation is required for CB1-dependent synaptic effects
Consistently with previous observations from our group (Pintor et al. 2006), corticostriatal slice perfusion with WIN55,212-2 (2 μM over 20 min) induced a long-lasting depression of FP amplitude (55.39 ± 2.8%; n = 7; p < 0.001 vs. basal) that reached 25.47 ± 1.27% of basal after 30 min of washout (p < 0.001 vs. control, Fig. 1a). This effect was completely blocked by AM251 (2 μM), the selective CB1 antagonist (99.15 ± 2.7%; n = 3 p < 0.01) that, by itself, did not affect synaptic transmission (Fig. 1b).
We then investigated the possible involvement of adenosine A2ARs in the modulation of WIN55,212-2-mediated effects. The co-application of the selective A2AR antagonist ZM241385 (100 nM, added 10 min before and then along with WIN55,212-2) significantly reduced WIN55,212-2-mediated effects (81.91 ± 4.65%; n = 8; p < 0.01 vs. WIN55,212-2 alone), suggesting that a basal level of A2AR is required for CB1R functioning (Fig. 1a and b). When applied alone, ZM241385 did not influence basal FP amplitude (data not shown).
To further verify the hypothesis that striatal A2ARs might play a permissive role in CB1R-mediated effects, we investigated the synaptic effects of WIN55,212-2 on corticostriatal slices from A2AR KO versus WT mice. As observed in rat corticostriatal slices, application of WIN55,212-2 2 μM (over 20 min) in WT mice induced a significant depression of FP amplitude that did not recover after washing (63.54 ± 1.94% and 38.93 ± 0.24% at the end of application and after 30 min of washout, respectively; n = 4; p < 0.05 vs. basal value, Fig. 2a). As shown in Fig. 2a and b, the ability of WIN55,212-2 to reduce synaptic transmission was significantly attenuated in A2AR KO mice (85.34 ± 2.64% and 57.03 ± 6.59% at the end of the application and of washout period, respectively, n = 5; p < 0.05 vs. WT mice).
The reduced effect of WIN55,212-2 in A2AR KO mice did not depend on a different density of CB1Rs in such a genotype, as no changes were revealed by western blot analysis (Fig. 2c).
The permissive effect of A2ARs towards CB1Rs occurs at post-synaptic level
To clarify whether the above reported interaction between A2A and CB1Rs occurred at pre- or post-synaptic level, we first performed experiments under a protocol of PPS in rat corticostriatal slices. In agreement with the known ability of WIN55,212-2 to inhibit pre-synaptic neurotransmitter release, WIN55,212-2-induced reduction of FP amplitude was accompanied by an increase in the magnitude of PPF ratio R2/R1 (1.21 ± 0.02 and 1.85 ± 0.16 in control conditions and after WIN55,212-2 application, respectively; n = 5; p < 0.01). The application of the selective CB1R antagonist AM251 (2 μM) completely prevented this effect (R2/R1 = 1.2 ± 0.02; p < 0.05 vs. WIN55,212-2 alone) whereas ZM241385 (100 nM) did not modify the ability of WIN55,212-2 to increase the R2/R1 ratio [1.54 ± 0.17; n = 8; NS (non significant) vs. WIN55,212-2 alone] (Fig. 3). Similarly, in A2A KO mice the application of WIN55,212-2 did not modify the PPS ratio (R2/R1 = 1.44 ± 0.16 in WT vs 1.35 ± 0.08 in KO mice, n = 5). These findings suggest that pre-synaptic A2ARs do not play a pivotal permissive role towards WIN55,212-2 effects.
To verify whether post-synaptic A2ARs could be involved, mice bearing a selective deletion of A2AR in post-synaptic striatal neurons (st-A2AR KO) were used. As shown in Fig. 4, the effect of WIN55,212-2 on the FP amplitude of R1 was attenuated in st-A2A KO (Fig. 4a). Although such a difference was very modest during the application of WIN55,212-2 and transient in nature, it reached a clear statistical significance during the washout period (% of FP amplitude at 40 min: WT: 50.5 ± 5.8, st-A2AKO: 82.2 ± 3.9, p = 0.002). The amplitude of R2 was almost unaffected (Fig. 4b). A significant reduction in the R2/R1 ratio was also observed (1.47 ± 0.08 in WT vs. 1.18 ± 0.06 in st-A2A KO mice: Fig. 4c). These results indicate a possible contribution of post-synaptic A2ARs receptors in WIN55,212-2-induced effects.
A2ARs regulate the mGlu5R-induced potentiation of WIN55,212-2 effects
To test the effects of the selective mGlu5R agonist CHPG, a concentration of WIN55,212-2 (1 μM) which induced a mild progressive decrease of FP amplitude was used in rat corticostriatal slices. Consistently with previous results (Domenici et al. 2004), we confirmed that CHPG 500 μM did not influence the FP amplitude in itself (95.96 ± 2.03% of basal, n = 3). However, the application, 10 min before and then along with WIN55,212-2, of CHPG 500 μM induced a marked potentiation of WIN55,212-2-induced reduction of synaptic transmission (% of basal FP amplitude at the end of WIN55,212-2 perfusion: WIN55,212-2 alone 90 ± 3.98%, n = 5; CHPG + WIN55,212-2: 65.59 ± 9%, n = 5; p < 0.05 vs. WIN55,212-2 alone, Fig. 5a and b), and this effect was reduced by the selective mGlu5R antagonist MPEP (30 μM) (80.94 ± 5.54%; n = 5; NS vs. WIN55,212-2 alone). This was an expected finding, as the activation of mGlu5Rs induces the formation of 2-AG which, in turn, contributes to the stimulation of CB1Rs. Given the strong functional interaction between striatal mGlu5 and A2ARs (see Introductory part), we next investigated whether A2ARs were also involved in CHPG-induced potentiation of WIN55,212-2 effects. ZM241385 (100 nM) completely prevented the potentiating influence of CHPG towards WIN55,212-2-induced effects (97.99 ± 3.7%; n = 4; p < 0.05 vs. WIN55,212-2 plus CHPG) (Fig. 5a and b). These data indicate that A2ARs may modulate CB1 effects also indirectly, namely by preventing the facilitatory influence of mGlu5Rs.
A2AR activation is not required for CB1-dependent protective effects
NMDA-induced neuronal injury in primary striatal cultures
Exposure of striatal neurons to NMDA 100 μM for 30 min caused a significant increase of LDH release with respect to basal levels (308 ± 14.58, Fig. 6a). WIN55,212-2 10 μM (applied 15 min before and then together with NMDA), significantly reduced NMDA toxicity (LDH release 248 ± 25.87; p < 0.001 vs. NMDA alone). This protective effect was abolished by the CB1R antagonist AM251 (10 μM) (LDH release 295 ± 15.98). Conversely, the A2AR antagonist ZM241385, 100 nM, applied 15 min before and then together with WIN55,212-2 plus NMDA, did not significantly reduce the neuroprotective effect of WIN55,212-2 (LDH release: 278 ± 19.18, p > 0.05 vs. WIN55,212-2 + NMDA) (Fig. 6a).
ZM241385 did not influence LDH release, either when applied alone or together.
NMDA-induced increase in intracellular calcium levels
To evaluate changes in [Ca2+]i, optical fluorimetric recordings with Fura-2AM were used and the ratio between the values of light intensity at 340 and 380 nm stimulation (F340/380) was recorded. The mean baseline value of the F340/380 ratio was 0.71 ± 0.01. Bath application of 30 μM NMDA induced a fast and persistent increase of [Ca2+]i (mean F340/380 measured 10 min after the application: 1.34 ± 0.05; n= 56 cells from five experiments). Pre-treatment with WIN55,212-2 (10 μM, applied 10 min before and then along with NMDA) reduced the increase of [Ca2+]i induced by NMDA (F340/380: 1.022 ± 0.07, n = 180 cells from 16 experiments; p < 0.05). The protective effect of WIN55,212-2 was prevented by AM251 10 μM (F340/380: 1.27 ± 0.05, n = 42 cells from four experiments). Finally, we tried to confirm the inability of A2AR antagonists on the protective effects of WIN55,212-2 towards NMDA. In agreement with the cytotoxicity findings, ZM241385 (100 nM) did not influence the effects of WIN55,212-2 on intracellular calcium concentration (F340/380 ratio 0.97 ± 0.1, n = 92 cells from five experiments, Fig. 6b).
The striatum is a brain area showing a strong expression of both A2ARs and CB1Rs. A functional cross-talk between these two receptors has been previously reported in different models. The involvement of A2ARs in the addictive properties of cannabinoids was demonstrated by the attenuation of Δ9-tetrahydrocannabinol-induced rewarding and aversive effects in A2AR KO mice (Soria et al. 2004), and by the finding that A2AR blockade prevents the synergy between μ-opiate and CB1 receptors (Yao et al. 2006). Furthermore, the A2AR antagonist ZM241385 reduced the effects of a CB1R agonist on forskolin-stimulated cAMP in human neuroblastoma cells and antagonized the motor depressant effect of CB1R activation (Andersson et al. 2005; Carriba et al. 2007). Worth of note, in the latter study, it has been demonstrated that A2ARs and CB1Rs are co-localized and constitute heteromeric complexes in the frame of which, as suggested by biochemical and behavioural experiments, the CB1R signalling depends on A2AR activation.
The present results strengthen the hypothesis that an endogenous activation of A2ARs is required to allow a proper CB1R functioning in the striatum. Specifically, we report that the blockade of A2ARs reduces WIN55,212-2-induced depression of synaptic transmission in corticostriatal slices and that the synaptic effects of WIN55,212-2 are reduced in slices from A2AR KO mice, thus confirming the occurrence of a permissive role of A2ARs towards CB1 effects. According to our results, post-synaptic A2ARs might contribute to the enabling of CB1-dependent synaptic effects. This is suggested by two orders of evidence: (i) the failure of ZM241385 to counteract the effect of WIN55,212-2 on PPF (see ‘Electrophysiological experiments in rats’) and (ii) the attenuation of WIN55,212-2-induced effects in slices from striatal A2AR KO mice. It should be considered, however, that the failure of ZM241385 towards WIN55,212-2-induced changes in R2/R1 ratio does not allow, per se, to definitely exclude a role of pre-synaptic A2ARs. In previous studies, in fact, we found that ZM241385 affects the R2/R1 ratio only in the presence of an increased neurotransmitter release (Tebano et al. 2005), a condition opposite to that induced by WIN55,212-2. Furthermore, also the data obtained in st-A2AR KO, although very intriguing, do not allow to definitely conclude on a major involvement of post-synaptic A2ARs. Indeed, although a clear difference was observed at some time points, the reduced ability of WIN55,212-2 to decrease FP amplitude was only transient in nature. Whatever the actual functional impact on CB1-dependent synaptic effects, however, according to our results, post-synaptic striatal A2ARs appear to play a role in WIN55,212-2 effects. Taking this concept in mind, it would be conceivable that the modulatory role of A2ARs may occur through post-synaptic A2A-CB1 heteromers, which activity depends on A2AR activation (Ferréet al. 2008). Otherwise, it is also possible that the A2A-dependent enabling of CB1 effects may involve other partners such as, for instance the D2 subtype of dopamine receptors. Indeed, CB1 and D2 receptors are co-localized in different striatal elements (Ferréet al. 2008) and heteromerize in co-transfected cells (Marcellino et al. 2008). Furthermore, striatal A2A and D2 receptors are strongly inter-independent on the functional ground (Ferréet al. 2007a,b,c) and some of the effects of CB1-A2A interactions seem to depend on D2R function (Andersson et al. 2005). Although the possible involvement of D2Rs was not explored here, it is thus conceivable that it does play a role. Indeed, CB1-D2-A2A receptor heteromers have been detected in HEK293T cells (Navarro et al. 2008), and they are likely to occur in the striatum (Ferréet al. 2008).
A further support to the critical role of A2ARs comes from the experiments on the interaction between mGlu5 and CB1 receptors. It is known that endocannabinoid release can be triggered by the activation of group I mGlu receptors/phospholipase C pathway (Wilson and Nicoll 2002; Doherty and Dingledine 2003) that yields to 1,2-diacylglycerol production and then, through diacylglycerol-lipase, to 2-AG. Accordingly, we found that the selective activation of mGlu5R by CHPG potentiated the synaptic effects of WIN55,212-2. The finding that the adenosine A2AR blockade prevents the facilitatory role of mGlu5R on CB1R-mediated effects confirms and extends the evidence of a permissive role played by A2ARs on striatal mGlu5Rs (Domenici et al. 2004; Rodrigues et al. 2005). These data are also in line with the general idea that A2ARs play the role of ‘fine tuners’ in the neuromodulation of synaptic transmission through functional and physical interactions with a variety of other receptors (Chen et al. 2007). Whether the above interaction reflects the existence of functional CB1-A2A-mGlu5 receptor heteromers (in analogy to what has been found for CB1-A2A-D2 receptors) is worthy of further investigations.
The permissive role played by A2ARs on WIN55,212-2-mediated synaptic effects does not appear to be also relevant in the modulation of NMDA-mediated toxicity. It is well established that the activation of CB1R represents an early step in the protective cascade against excitotoxicity, providing a protection on-demand (Marsicano et al. 2003). Here, we showed that the application of WIN55,212-2 significantly inhibited NMDA-induced increase in intracellular calcium and LDH release in striatal neurons, thus confirming the previously reported neuroprotective properties of cannabinoids in vitro (Hampson and Grimaldi 2001; Kim et al. 2006; Zhuang et al. 2005) and in vivo models (Marsicano et al. 2003; Pintor et al. 2006). Unlike its synaptic effects, however, the protective activity of WIN55,212-2 towards NMDA was not prevented by ZM241385. This finding indicates that not all the effects exerted by WIN55,212-2 are under the control of A2ARs. In agreement, although in the rat striatum most A2ARs are co-localized with CB1Rs, there is still a proportion of CB1Rs which does not co-localize with A2ARs (Carriba et al. 2007). Alternatively, the failure of ZM241385 to prevent the protective effects of WIN55,212-2 may simply indicate that the complex modulatory role exerted by A2ARs on CB1 effects requires a much higher level of integration than a cellular preparation. Another point of some relevance here is that the mechanisms by which CB1Rs modulate NMDA-induced calcium increase seem to be quite complex, as WIN55,212-2 itself may either increase (Lauckner et al. 2005) or decrease (Zhuang et al. 2005) the release of calcium from intracellular stores.
Altogether, the present results show that the state of activation of A2ARs regulates the synaptic (but, apparently, not the neuroprotective) effects exerted by CB1Rs, that post-synaptic A2ARs are at least partially involved in such permissive effects, and that A2ARs may control CB1 effects also indirectly, namely through mGlu5Rs.
Given the pivotal role played by A2A, CB1 and mGlu5Rs in the control of striatal functions, these results give new perspectives in the pathophysiology of striatal disorders.