Background of the invention The nucleoside adenosine can influence most mammalian cell types via stimulation of four G protein coupled receptors subtypes Ai, A2A, A2B and As (Fredholm et al., 1994). The nature of the responses to adenosine and other agonists depends on the selective coupling of the activated receptor to distinct G proteins. The A1 (Freissmuth et al., 1991; Jockers et al., 1994; Munshi et al., 1991) and As (Palmer et al, 1995; Zhou et al, 1992) receptors are coupled with Gi proteins inhibiting adenylyl cyclase, whereas the A2A and A2B receptors (Londos et al., 1980; Pierce et al., 1992; van Calker et al., 1979) couple to Gs-like proteins and activate adenylyl cyclase. Three stimulatory G proteins have been biochemically characterized ; Gs short, Gs long and Golf (Jones et al., 1990). They show 88% identity in their amino acid sequences (Jones and Reed, 1989). The Gs subunits have a widespread distribution, whereas the Golf subunit is distributed in a more restricted manner (Herve et al., 1993), and are particularly high in striatum (Drinnan et al., 1991 ; Herve et al., 1993). The striatal expression may be functionally important since recently developed transgenic mice deficient in Golf are not only anosmic but are also hyperactive (Belluscio et al., 1998).

Thus, Golf present in striatum probably plays a key role in the signal transduction mediated by G protein coupled receptors in this area. Previous work has shown that Golf is the main stimulatory G protein coupling dopamine D1 receptors to adenylyl cyclase in the striatum (Herve et al., 1993). D1 receptors are abundantly expressed in approximately 50% of the striatal neurons, mostly in those projecting to the substantia nigra and containing GABA, substance P and dynorphin (Gerfen et al., 1990; Le Moine et al., 1991). Adenosine A2A receptors are also highly expressed in the striatum (Jarvis and Williams, 1989 ; Parkinson and Fredholm, 1990).

However, adenosine A2A receptors are segregated from dopamine D1 receptors and are selectively expressed in the striatopallidal neurons, that also contain enkephalin and dopamine D2 receptors (Fink et al., 1992; Schiffmann et al., 1991; Svenningsson et al., 1997). This is the other major subpopulation of projection neuron within the striatum. It is also known that A2A receptors stimulate adenylyl cyclase and cAMP-dependent signal transduction in striatum (Fredholm, 1977; Svenningsson et al., 1998).

The A2A receptor is particularly abundant in the striatopallidal GABAergic neurons that are also characterized by having a high density of dopamine D2 receptors. Adenosine A2A and dopamine D2 receptors are functionally antagonistic. Under normal physiological conditions, endogenous adenosine exerts a tonic influence on the striatopallidal neurons, causing an increase in neuronal activity. This tonic influence can be counteracted either by dopamine released from the dopaminergic neurons acting on dopamine D2 receptors or by drugs that inhibit adenosine A2A receptors. In Parkinson's disease the dopaminergic neurotransmission is impaired which leads less dopamine in striatum and consequently to overactive A2A receptors.

When the adenosine receptor is activated it counteracts D2 receptor activation and endogenous dopamine will be less effective (Ferre et al. 1997 ; Svenningsson et al. 1999). Conversely, if the adenosine A2A receptor is blocked, endogenous dopamine will be more effective, a desired biological mechanism in the treatment of Parkinson's disease. Indeed, there is excellent evidence that blockade of A2A receptors prevents Parkinson-like symptoms in animals and even the progression of the underlying neurodegeneration (Chen et al.). Furthermore, several human studires with the non-selective antagonist caffeine show that this works also in man (See Fredholm et al. 1999).

Summary of the invention The hitherto identified problem with targeting A2A receptors, is due to that A2A receptors is not only expressed in striatopallidal neurons, but also abundant in endothelial cells, leukocytes, blood plates and blood vessels. For these target tissues and cells, it is not desirable to use drugs affecting the adenosine A2A receptors function since that could lead to unwanted side effects. Abundant reports in the literature convincingly show that endogenous adenosine can activate AsA receptors of importance in blood platelet aggregation and in acute inflammatory reactions.

It has been for a long time suggested and demonstrated that adenosine A2A receptors are coupled to the stimulatory G protein subunit Gs in several tissue types. This is true in several tissues and cells where A2A receptors are expressed. However, the present inventors have found that adenosine A2A receptors are co-localized with the stimulatory G protein Golf in the same neurons in the striatum and have also shown that Golf is the natural partner to A2A receptors rather than Gs in striatum. Thus, targeting the A2A-G,, if complex rather than using antagonists developed for A2A-Gs could potentially avoid unwanted side effects. If such specific antagonists are discovered, they hold the potential to specifically act on A2A receptors expressed in striatum and not at the A2A receptors expressed in other tissues.

A drug that selectively interacts with the combination of adenosine A2A receptors and Golf would have several potential applications: disorders linked to an altered dopaminergic neurotransmission such as Parkinson's disease, Hurltington's chorea and other CNS diseases involving the dopaminergic or GABA transmittor systems which are modulated by adenosine A2A receptors.

Present therapy shows some degree of efficacy, however it is well-known that present therapy is not uniformly successful over the longer term and that additional treatment alternatives are necessary (see e. g. Jankovivic 2000).

One additional and intriguing possibility is that such a drug could be used to decrease the number of side effects of current therapy in Schizophrenia.

The vast majority of these drugs are more or less selective for dopamine D2 receptors. It is known that the most potent of these drugs when given over the course of several weeks can lead to not only transient movement disorders, but also to so called tardive dyskinesia.

Thus, in a first aspect the invention relates to a screening method comprising use of an A2A receptor coupled to Golf to screen for agents selectively interacting with A2A receptor coupled to Golf. The interaction may be directly or indirectly with A2A-Gotf. Thus, the invention relates to use of A2A-Golf as a drug target.

The drug target may be the A2A receptor either at its ligand binding or G protein-binding site. Alternatively, the drug target may be the Golf protein.

The screening method can, for example, be a high throughput screening method.

In high throughput screening for ligands (agonists or antagonists) the inventors will use a cell system to study adenosine A2A receptors coupled to either Gs or Golf. The cyc-variant of S49 mouse lymphoma cells, which is deficient in endogenous stimulatory G protein, provides an ideal experimental system in which to measure the efficacy of adenylyl cyclase activation by exogenous introduced G proteins. A reporter gene might be used to measure receptor activation or inactivation.

To express adenosine A2A receptors and G proteins in S49 cells the inventors will use a retroviral expression (RevTet) system which allows a fast and efficient establishment of regulated gene expression in the cells.

Five different stable cell lines are produced; A2A, A2A-Golf, A2A-Gs, Golf and Gs Cells are selected with G418 and hygromycin after virus infections and inducible clones isolated.

Alternatively, a yeast two-hybrid system could be used for screening of Axa Golf selective antagonists or agonists. A further alternative is GTP binding to purified Golf and Gs proteins as indicator for screening of Golf selective agonists.

In a second aspect, the invention relates to antagonists or agonists for adenosine A2A receptor coupled to Golf obtained by any of the above screening-methods of the invention. Antagonists are intended for the primary indication (Parkinson) and agonists for secondary indication (Schizophrenia).

In a third aspect, the invention relates to a pharmaceutical preparation comprising an antagonist or agonist for adenosine A2A receptor coupled to Golf. The pharmaceutical preparation is formulated according to conventional practice as concerns additives etc.

The pharmaceutical preparation may be used for treatment of disorders linked to dopaminergic neurotransmission, such as Parkinson's disease and Schizophrenia.

A further alternative is to use the pharmaceutical preparation according to invention for treatment of CNS diseases involving dopaminergic or GABA transmittor systems.

Detailed description of the invention Abbreviations CHO-K1, Chinese hamster ovary cells, strain Kl ; SDS-PAGE, sodium dodecylsulfate-polyacryl-amide gel electrophoresis; EDC, N-ethyl-N-(3- dimethylamino-propyl) carbodi-imide hydrochloride; MES, 2- [N- morpholino] ethanesulfonic acid ; Mops, 4-morpholinepropanesulfonic acid; m-AcAGTP, m-acetylanilido-GTP; DTT, dithiothreitol; EDTA, ethylendiamine- tetra-acetate; PCR, polymerase chain reaction; PFA, paraformaldehyde ; SSC, saline sodium citrate buffer; PMSF, phenymethylsulfonylfluoride; cAMP, adenosine 3', 5'-cyclic monophosphate; a-MEM, a-minimum essential medium.

Materials and Methods Chemicals Aprotinin, m-aminoacetophenone, 1, 4-dioxane, Triton X-100, sodium deoxycholate, phenymethylsulfonylfluoride (PMSF), protein A-sepharose, tergitol NP-40,2- [N-morpholino] ethanesulfonic acid (MES), N-ethyl-N'- (3- dimethylamino-propyl) carbodi-imide hydrochloride (EDC), and adenosine 3', 5'-cyclic monophosphate (cAMP) were purchased from Sigma (St. Louis, MO, USA). 2-[p-(2-carbonyl-ethyl)-phenylethylamino]-5'-N- ethylcarboxamidoadenosine (CGS 21680) was obtained from RBI (Research Biochemicals International, Natick, MA, USA). [3H] cAMP was obtained from Du Pont-NEN (Boston, MA, USA). 5-Amino-7- (2-phenylethyl)-2- (2-furyl)- pyrazolo [4,3-e]-1,2,4-triazolo [1, 5-c] pyrimidine (SCH 58261) was a gift from Dr. E. Ongini, Schering-Plough Research Institute, Milan, Italy. Digoxgenin- 11-UTP and FuGENETM 6 transfection reagent was purchased from Boehringer Mannheim, Germany. [35S] UTP, [a32P] GTP and 4-morpholine- propanesulfonic acid (Mops) were purchased from Amersham, (Little Chalfont, England).

Tissue preparation Adult male Sprague-Dawley rats (200-250 grams) were briefly anesthetized with C02 and killed immediately by decapitation. For the in situ hybridization experiments, the brains were dissected out and frozen on dry ice. Consecutive coronal sections (14 um) were made through the rostral part of striatum. For the photolabeling experiments striatum was dissected out and homogenized on ice in 50 mM Tris-HCl (pH 7.4) with an ultra turrax (3 x 10 seconds). The homogenate was centrifuged for 10 min (1000 g) after which the pellet was discarded. The supernatant was centrifuged for 50 min (30000 g) and washed once with 50 mM Tris-HCl (pH 7.4). After the second centrifugation the pellet, containing the membranes, was resuspended in an incubation buffer (20 mM MgCl2, 400 mM NaCI, 120 mM HEPES and 0.4 mM EDTA) and stored at-80°C in aliquots until used.

Subcloning of cDNA fragments for in vitro transcription templates cDNA fragments corresponding to the rat adenosine A2A receptor, the rat alpha subunit Gs and the rat alpha subunit Golf were amplified with polymerase chain reaction (PCR). The PCR products were subcloned into pBluescript II (Stratagene, La Jolla, CA, USA. In order to verify the incorporation of the cDNA into pBluescript II, the subcloned cDNA fragments were sequenced using an automated DNA sequencer, ABI 373A (Applied Biosystems Inc. Foster City, CA, USA).

Probe synthesis and labeling for in situ hybridization 35S-or digoxigenin-labeled antisense and sense cRNA probes were prepared by in vitro transcription from cDNA clones corresponding to fragments of rat adenosine A2A receptor (Fink et al., 1992), rat G protein alpha subunit Gs (Jones and Reed, 1987) and rat G protein alpha subunit Golf (Jones and Reed, 1989). The transcription was performed using MAXI-scriptTM in vitro transcription kit according to the manufacturer's protocol (Ambion Inc., Austin, TX, USA).

Single in situ hybridization The in situ hybridization experiments were performed as previously described by Le Moine and Bloch (1995). Cryostat sections were postfixed in 4% paraformaldehyde (PFA), dehydrated in graded alcohol, and were the hybridized overnight at 55°C with 106 c. p. m. of 35S-labeled probe in 50 ul of hybridization solution. After washing, the slides were dipped into Ilford K5 emulsion (diluted 1/3 in SSC) and exposed for eight to twelve weeks, developed, and stained with cresyl violet.

Double in situ hybridization In double in situ hybridization experiments, the adenosine A2A receptor cRNA probe was labeled with (35S) UTP, whereas the probes against Gs and Golf mRNAs were labeled with digoxigenin-11-UTP. Cryostat sections were pretreated as described above. The sections were hybridized overnight at 55° C with a combination of 35S-and digoxigenin-labeled probes (106 c. p. m. of 35S-labeled probe and approximately 20 ng of digoxigenin-labeled probe in 50 ul of hybridization solution). The slides were washed in RNAse A and various concentrations of SSC, but without DTT. At the end of the washes, the slides were put in 0. 1 xSSC at room temperature. The sections were rinsed twice for 5 min in buffer A (1 M NaCl/0. 1 M Tris/2 mM MgCl2, pH 7.5), and then for 30 min in buffer A containing 3% normal goat serum and 0. 3% Triton X-100. After 5 h of incubation at room temperature with alkaline phosphatase-conjugated anti-digoxigenin antiserum (Boehringer Mannheim, 1: 1000 in buffer A/3% normal goat serum/0.3% Triton X-100), the sections were rinsed in buffer A (5 min, twice), then for 10 min twice in buffer B (1 M NaCl/0. 1 M Tris/5 mM MgCl2, pH 9.5), and 10 min twice in 0.1 M buffer B (containing 0.1 M NaCl). The sections were then incubated overnight in the dark at room temperature in 0.1 M buffer B (pH 9.5) containing 0.34 mg/ml nitroblue tetrazolium and 0.18 mg/ml bromochloro-indolylphosphate. The sections were thereafter rinsed in 0.1 M buffer B (pH 9.5), then in 1 x SSC, dried and dipped into Ilford K5 emulsion. After being exposed for seven to eleven weeks the sections were developed and mounted without counterstaining.

Preparation of radioactive m-AcAGTP The preparation of radioactive m-AcAGTP was performed as described by Zor et al. (1995). Briefly, [a32P] GTP (3000 Ci/mmol) was freeze-dried and redissolved in 50 ptl of 0.125 M MES buffer (pH 6.5). EDC (10 pmol) was then dissolved in this solution and m-aminoacetophenone (20 limol) in 20 KLl of 1,4-dioxane was added. After 5 h at room temperature the mixture was freeze dried and redissolved in 10 mM Mops buffer (pH 7.0). Insoluble m- aminoacetophenone was removed by centrifugation. The product was stored at-18°C until used.

After the synthesis of [a32p] m-AcAGTP the photoreactivity of the product was tested by u. v. irradiation of the compound spotted onto polyethyleneimine- cellulose plates followed by chromatography with 1 M LiCl as mobile phase.

Approximately 30-40% of the irradiated m-AcAGTP did not migrate from the origin (data not shown). Non-irradiated m-AcAGTP had an Rf value twice that of [a32P] GTP which agrees with that shown previously (Zor et al., 1995).

Photoaffinity labeling of striatal membranes with [a32P] m-AcAGTP First we optimized the photoaffinity reaction with regard to Mg2+, NaCl and GDP concentrations. The reaction mixture that gave the largest ratio of incorporated [a32P] m-AcAGTP between stimulated and unstimulated membranes was selected for further use. The optimal reaction mixture contained, in a total volume of 60 ; nl, 50 jug of membranes, 5 mM MgCl2, 100 mM NaCl, 0. 1 mM EDTA, 0. 6 mM ATP, 100 uM GDP and the adenosine A2A agonist CGS 21680 when indicated. After 3 min incubation at 37°C, 1.5 ACi of [a32P] m-AcAGTP was added and the samples were incubated for additional 5 min at 37° C. The reaction was terminated by dilution with 400 PI of ice- cold incubation buffer (5 mM MgCl2, 100 mM NaCI, 30 mM HEPES, 0.1 mM EDTA and 2 mM DTT). Excess photoaffinity label was removed by centrifugation at 4°C (12000 g for 5 min) and the pellet was resuspended in 60 nl of incubation buffer. The membranes were pipetted into dimples made in an aluminium foil pressed against an empty eppendorf tube rack. The aluminum foil was wrapped over a glass dish filled with ice. UV irradiation (k =320 nm) was applied for 20 min from a distance of 2 cm from the samples.

Photolabeled membranes were pelleted and solubilized in 2% SDS for 10 min at room temperature prior to SDS/PAGE (12%) and autoradiography or immunoprecipitation.

Immunoprecipitation of alpha Golf protein The solubilized photolabeled membranes were diluted 1 : 1 with immunoprecipitation buffer (10 mM Tris-HCl (pH 7.4), 1% Triton X-100, 1% sodium deoxycholate, 0.5% SDS, 150 mM NaCl, 1 mM DTT, 1 mM EDTA, 10 , ug/ml aprotinin and 0.2 mM PMSF) and centrifuged at 12000 g for 10 min at 4°C. The pellet from this centrifugation was discarded and the supernatant was mixed with 3 ul of non-diluted subtype-specific G protein a subunit Golf antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA.), which does not cross-react with the G protein a subunit Gs, and was incubated for 1 h at 4° C under constant rotation. Then 60 pl of protein A-Sepharose (4 mg) in immunoprecipitation buffer was added to each sample and incubated overnight at 4°C under constant rotation. Thereafter the Sepharose beads were pelleted (1 min 12000 g, 4°C) and washed twice with 1 ml of washing buffer A (50 mM Tris/HCl, 600 mM NaCl, 0.5% SDS and 1% Tergitol NP-40, pH 7.4) and twice with washing buffer B (100 mM Tris/HCl, 300 mM NaCI and 10 mM EDTA, pH 7.4). The washed Protein A-Sepharose was then resuspended in 100 nl Laemmli buffer and heated at 80° C for 5 min and then centrifuged as above. Fifty microliters of the supernatant was then subjected to SDS/PAGE (12%), as described by Laemmli (1970). The gel was dried in a gel drier and exposed to X-ray film.

Subcloning of the rat alpha Go f cDNA The rat G alpha subunit Golf cDNA was amplified with polymerase chain reaction (PCR) using a plasmid (Bluescript, Stratagene) containing the rat Golf cDNA sequence (a gift from Dr. A. G. Gilman). The PCR product was subcloned into the vector pCI-neo (Promega, Scandinavian Diagnostic Services, Falkenberg, Sweden). In order to verify the incorporation of the cDNA into pCI-neo, the subcloned cDNA fragments were sequenced using an automated DNA sequencer, ABI 373A (Applied Biosystems Inc. Foster City, Ca, USA).

Transient transfection of Go f cDNA into CHO cells expressing human adenosine A2A receptors Chinese hamster ovary cells (CHO-K1 cells; CCL61, American Type Culture Collection) stably expressing human adenosine A2A receptors (Kull et al., 1999) were grown adherent and maintained in a-minimum essential medium (a-MEM) without nucleosides, containing 10% fetal calf serum, penicillin (50 U/ml), streptomycin (50 llg/ml), L-glutamine (2 mM) and geneticin (Gibco; 500 pg/ml) at 37°C in 5% C02/95% air.

Transient transfection of Golf cDNA or control plasmid was performed using FuGENETM 6 transfection reagent (Boehringer Mannheim, Germany) according to the instruction manual.

Measurement of cAMP accumulation Thirty-six hours after transfection, both Golf cDNA-transfected cells and control cells were sown out in 12-well plates (200,000 cells per well) and were allowed to grow for 36 h. Cells were then washed 2 times with HEPES- buffered (20 mM) a-MEM (pH 7.4). The cells were incubated at 37° for 10 min in 0.9 ml of HEPES buffered medium. The adenosine A2A agonist CGS 21680 was added in 0.1 ml of medium and the cells were incubated for another ten min. The reactions were terminated by the addition of perchloric acid to a final concentration of 0.4 M. After 1 h at 4°C the acidified cell suspensions were transferred to tubes and neutralized with 4 M KOH/1 M Tris-HCl. The cAMP content in the samples was determined using a competitive radioligand-binding assay (Nordstedt and Fredholm, 1990).

Radioactivity was measured in an LKB/Pharmacia scintillation counter with 3 ml of ReadySafe (LKB/Pharmacia) scintillation fluid.

Data analysis In the double in situ hybridization experiments with probes against adenosine A2A receptor mRNA and Golf or Gs mRNAs in the striatum, three categories of neurons were counted: those that only exhibited a radioactive signal (i. e. at least two times the background), those that showed only a non- radioactive signal, and those that showed both signals. Quantification were made in the lateral and medial parts of striatum and in the core and shell regions of nucleus accumbens. All neurons within the examined areas were counted.

Autoradiographic data from photolabeling experiments were quantified using densitometry (MCID system; Imaging Research, St. Catharines, Canada).

Analysis of dose response curves from cAMP measurements, non-linear regression analysis and statistical analysis (t test) were performed using GraphPad Prism version 3.00 for Windows (GraphPad Software, San Diego, CA, USA).

Results Regional distribution of Go f, Gs and A2A receptors in rat forebrain In order to study the abundance and distribution of mRNAs for Golf, Gs and A2A receptors we performed in situ hybridization on coronal sections from the rat brain (Fig. 1). Adenosine A2A receptor mRNA was, as shown previously (Svenningsson et al., 1997), abundant in the caudate putamen, the nucleus accumbens and the olfactory tubercle (Fig. la). Golf mRNA was also found in the caudate putamen, the nucleus accumbens and the olfactory tubercle. In addition, Golf mRNA, but not adenosine A2A receptor mRNA, was detected in e. g. the pyramidal cell layer of piriform cortex and in the Islands of Calleja (Fig. lb). These neurons in the Islands of Calleja are known to be interconnected with the olfactory tubercle (Fallon, 1983) and to express dopamine D1 and D3 receptors (Le Moine and Bloch, 1996; Svenningsson et al., 1997). Gs mRNA was most abundant in areas not expressing adenosine A2A receptor mRNA, such as pyramidal cells of the piriform cortex, cerebral cortex, claustrum, endopiriform nucleus and the diagonal band of Broca. In addition Gs mRNA was found throughout septum, whereas mRNA encoding A2A receptors were found only in occasional cells in lateral septum. Moderate levels of Gs mRNA were detected in nucleus accumbens, whereas caudate putamen only exhibited very low expression (Fig. lc).

Cellular localization of Go f, Gs and A2A receptor mRNA within the striatum There are three major subcategorize of striatal neurons; medium-sized (app.

20 im in diameter) neurons that are further subdivided into spiny and aspiny neurons and large-sized (app. 40 llm in diameter) neurons. 95 % of the striatal neurons are medium-sized spiny neurons and the remaining 5 % are medium-sized aspiny neurons and large-sized neurons. Since no staining for detecting spines was used in the present study, we could only distinguish between neurons based on their size.

In caudate putamen and nucleus accumbens 49% of the medium-sized neuron-like cells exhibited a strong signal for adenosine A2A receptor mRNA, which is in agreement of our previous report (Svenningsson et al., 1997). Golf mRNA was found in the majority (98%) of the medium-sized neuron within the striatum. No expression of A2A receptor or Golf mRNAs was found in large-sized neurons. The labeling of Gs mRNA in striatum was weak and most of the label appeared to be located in glial rather than neuronal cells.

Gs mRNA was occasionally found in large-sized neurons. Of the medium- sized neurons, only 12% in caudate putamen and 27% in nucleus accumbens showed detectable Gs mRNA.

In order to determine the co-expression of adenosine A2A receptor mRNA with that of Golf or Gs in caudate putamen and nucleus accumbens we performed double in situ hybridization. Quantifications were made in the lateral and medial parts of striatum and in the core and shell regions of nucleus accumbens. As expected from the above-mentioned data, the large majority of the adenosine A2A receptor mRNA-containing neurons also expressed Golf mRNA in the lateral and medial parts of striatum as well as in the core and shell regions of nucleus accumbens (95,91,91 and 88 %, respectively) (Fig.

2a, 3). By contrast, only a minority (3-5 %) of the adenosine A2A receptor mRNA-containing neurons co-expressed Gs mRNA in caudate-putamen (Fig.

2b, 3). In nucleus accumbens, the extent of co-expression of A2A receptor mRNA and Gs mRNA was somewhat higher (13-21 %). When Golf mRNA positive neurons were considered, 47-56 % of them were also positive for A2A receptor mRNA labeling in the examined areas of caudate-putamen and nucleus accumbens (Fig. 2a and 3). Out of the few neurons that expressed Gs mRNA, 17-36 % in caudate putamen and 33-42% in nucleus accumbens, also expressed A2A receptor mRNA (Fig. 3). We could also detect a high level of Os mRNA expression in some large-sized neurons which did not express A2A or Golf mRNAs (not shown).

Photolabeling of G proteins in striatal membranes To investigate if Golf is activated upon stimulation with an A2A receptor agonist, we used a photolabeling technique (Zor et al., 1995). We first established that G proteins present in the membranes can be photolabeled by [a32P] m-AcAGTP and that activation of G proteins leads to an increase in this photolabeling. We incubated striatal membranes with [a32PIm-AcAGTP with and without a high concentration (10-, uM) of the adenosine A2A agonist CGS 21680 and with different concentrations (3-100 uM) of GDP. The labeling was then visualized by SDS/PAGE followed by X-ray autoradiography. The crude photolabeled membranes gave a broad band with an apparent molecular mass of 40-50 kDa (Fig. 4a). In a parallel experiment the Golf subunits were immunoprecipitated with a specific antibody directed against Golf, and separated on SDS-PAGE. The incorporated [a32P] m-AcAGTP label was determined by autoradiography (Fig.

4b). The immunoprecipitate gave a single band with an apparent molecular mass of 44-45 kDa, a size identical to the molecular mass of the protein visualized in immunoblot experiments (data not shown). By comparing the intensity of [a32P] m-AcAGTP labeling in the crude membranes it was not possible to distinguish between control conditions and in the presence of agonist at any of the GDP concentrations. However, by comparing the intensity of labeling after immunoprecipitation, it was clear that the labeling was enhanced in the stimulated samples. At low (3-10 AM) GDP concentrations the basal labeling was strongest, but the ratio of agonist- stimulated to basal photolabeling of the Golf protein was largest at high (30- 100 pu) GDP concentrations (Fig. 4b and 4c).

To determine if the agonist induced enhancement of [a32P] m-AcAGTP labeling of Golf was dose-dependent, photolabeling experiments were performed using different concentrations of the adenosine A2A receptor agonist CGS 21680 in the presence of a high concentration of GDP (100 uM).

The immunoprecipitated samples were visualized by SDS/PAGE followed by X-ray autoradiography. As demonstrated by the representative autoradiogram in Fig. 5a incorporation of la32P] m-AcAGTP into Golf protein increased in a concentration dependent manner. The intensity of the labeling was used to construct a dose-response curve (Fig. 5b). The potency of CGS 21680 in this assay was estimated to 16 nM (4.7-54; 95% confidence interval). The agonist induced increase of labeling was blocked by the specific adenosine A2A receptor antagonist SCH 58261 (30 nM) (data not shown). cAMP accumulation experiments The adenosine A2A agonist CGS 21680 induced a concentration dependent increase in cAMP accumulation both in the control cells and in the Golf cDNA transfected CHO cells (Fig. 6). There was no significant change in the potency of CGS 21680 : 19 nM (13-27) for Golf transfected cells and 29 nM (16-52) for control cells. However, there was a significant increase (t test P=0.0082) in the efficacy of the agonist. The plateau was 2.8 (2.6-3.0) pmol/50 jul for Golf transfected cells and 2.2 (1.9-2.4) pmol/50 PI for control cells (95% confidence intervals within parentheses).

Discussion It has been taken for granted that adenosine A2A receptors are coupled to the stimulatory G protein subunit Gs. The present data suggest that this may not be the complete truth, at least not in the dopamine-rich areas of the brain.

There are some previous reports that suggest that Gs plays a minor role in striatal neurons, which contain the highest levels of A2A receptors in the brain. Whereas the highest levels of activated adenylyl cyclase, measured using [3H] forskolin binding, are found in the cautate putamen (Worley et al., 1986), this region possesses the lowest levels of Gs mRNA in brain (Largent et al., 1988) (see also Fig. lc, 2b and 3). It can also be mentioned that Alright and IvIcCune-Albright syndromes, which are associated with mutations of the Gs gene, do not exhibit any extrapyramidal neurological symptoms (Schwindinger et al., 1992; Weinstein et al., 1990).

Herve and co-workers (1993) have shown that selective lesioning of striatonigral neurons, using a retrograde neurotoxin, markedly decreases the levels of dopamine D1 receptors and Golf in striatum. Moreover, 6-OHDA lesioning of the dopaminergic axons in neonatal rats induced hypersensitivity to dopamine receptor agonists in adulthood without any change in Di receptor binding-but there is an increase in Golf expression.

There is also an increase in Gs protein, but mainly in glial cells (Penit-Soria et al., 1997). Furthermore, mice deficient in Golf are hyperactive (Belluscio et al., 1998). These results could indicate that Golf may, in striatum, play some of the role (s) otherwise attributed to Gs.

Here we confirm that mRNAs for the adenosine A2A receptor (Svenningsson et al., 1997) and for Golf (Drinnan et al., 1991 ; Herve et al., 1993) were most abundant in the caudate putamen, the nucleus accumbens and the olfactory tubercle, whereas these areas express little Gs (Largent et al., 1988). We also show directly that adenosine A2A receptors and Golf are co-localized. In situ hybridization experiments showed that approximately 50% of the medium- sized neuron-like cells exhibited a strong signal for adenosine A2A receptor mRNA and a large majority (95%) of these neurons co-expressed Golf mRNA.

Golf mRNA was also abundant in A2A receptor mRNA negative neurons. Thus, the relative enrichment of A2A receptors and Golf mRNA in the striatum is likely to reflect the preferential localization in the striatopallidal neurons (A2A receptors) and in striatonigral and striatopallidal neurons (Golf). The labeling for Gs mRNA in striatum was weak and seemed to be preferentially located in glial cells as previously described (Feinstein et al., 1992). In the present study, only3-5% of the neurons in caudate putamen and 13-21% of the neurons in nucleus accumbens expressed both A2A receptor mRNA and Gs mRNA.

Using a photoaffinity labeling method we show that the adenosine A2A receptor is not only colocalized with Golf in striatum, but also functionally coupled to it. Photoaffinity labeling is a method used for identification of distinct molecular components in biochemical processes. The simplest procedure for labeling G proteins is by u. v. irradiation of [a32P] GTP (Basu and Modak, 1987). However, this method has low sensitivity. The most widely used photoaffinity label for G proteins is azidoanilido-GTP and was developed by Pfeuffer (Pfeuffer, 1977). However, this compound has some disadvantages: it is instable and its synthesis, purification and application must be performed in complete darkness. The photoaffinity label we have used, m-acetylanilido-GTP, developed by Zor and co-workers (1995), has several advantages: (1) No need for purification; (2) Quantitative conversion of GTP into m-AcAGTP; (3) Excellent stability in solution ; (4) Resistant to hydrolysis and remains bound to the G protein during centrifugal washing.

Using this photolabeling method we could demonstrate that the adenosine A2A agonist CGS 21680 induced an increased incorporation of [a32P] m- AcAGTP in immunoprecipitated Golf subunits. The magnitude of the increase was GDP-concentration-dependent, with the largest increase seen at a high GDP concentration (100 jus). CGS 21680 increased labeling in a dose- dependent manner with a calculated ECso value of 16 nM (4.7-54). This increase was receptor mediated since it was blocked by the A2A receptor antagonist SCH 58261 (data not shown). The CGS 21680 potency is in agreement with the potency in the cAMP assay.

In an attempt to, determine whether A2A receptor-activated Golf subunits could activate adenylyl cyclase we transiently transfected Golf into CHO cells stably expressing A2A receptors (Kull et al., 1999). In the Golf transfected cells there was a significant increase in the maximal agonist stimulated adenylyl cyclase activity without any change in the potency of CGS 21680. A change in the maximal effect was expected from the previous demonstration of increased constitutive accumulation of cAMP and the observation that agonist-stimulated cAMP levels were proportional to the amount of Gs expression (Yang et al., 1997).

Conclusion The present study showed that adenosine A2A receptor mRNA was co- expressed with Golf mRNA in striatal medium-sized neurons to a much higher extent than with Gs mRNA. Using a photolabeling technique, we showed that activation of adenosine A2A receptors in striatal membranes led to activation of Golf. Moreover, transfection of Golf cDNA into cells that express human adenosine A2A receptors led to an increase of the maximal agonist stimulated cAMP level. When taken together, these results provide strong anatomical and biochemical evidence that adenosine A2A receptors stimulate adenylyl cyclase and cAMP-dependent signal transduction in striatum by activating Golf rather than Gs. These findings have the more general implication that a given receptor may couple to different G proteins in different locations.

References Basu A and Modak MJ (1987) An affinity labeling of ras p21 protein and its use in the identification of ras p21 in cellular and tissue extracts. J Biol Chem 262: 2369-2373.

Belluscio L, Gold GH, Nemes A and Axel R (1998) Mice deficient in G (oii) are anosmic. Neuron 20 : 69-81.

Chen JF, Xu K, Petzer JP, Staal R, Xu YH, Beilstein M, Sonsalla PK, Castagnoli K, Castagnoli N Jr, Schwarzschild MA. Neuroprotection by caffeine and A (2A) adenosine receptor inactivation in a model of Parkinson's disease. J Neurosci 2001 May 15; 21 (10): RC143.

Drinnan SL, Hope BT, Snutch TP and Vincent SR (1991) Golf in the basal ganglia. Mol Cell Neurosci 2: 66-70.

Fallon JH (1983) The islands of Calleja complex of rat basal forebrain II: connections of medium and large sized cells. Brain Res Bull 10: 775-793.

Feinsten DL, Mumby SM and Milner RJ (1992) Characterization of Gs alpha mRNA transcripts in primary cultures of rat brain astrocytes. Glia 5: 139- 145.

Ferre S, Fredholm BB, Morelli M, Popoli P, Fuxe K. Adenosine-dopamine receptor-receptor interactions as an integrative mechanism in the basal ganglia. Trends Neurosci 1997 Oct; 20 (10): 482-7 Fink JS, Weaver DR, Rivkees SA, Peterfreund R, Pollak R, Adler E and Reppert SM (1992) Molecular cloning of a rat A2 adenosine receptor : Selective co-expression with D2 dopamine receptors in rat striatum. Mol Brain Res 14: 186-195.

Fredholm BB (1977) Activation of adenylate cyclase from rat striatum and tuberculum olfactorium by adenosine. Med Biol 55 : 262-267.

Fredholm BB, Abbracchio MP, Burnstock G, Daly JW, Harden TK, Jacobson KA, Leff P and Williams M (1994) Nomenclature and classification of purinoceptors. Pharmacol Rev 46: 143-156.

Fredholm B. B., Battit K., Holmen J., Nehlig A. and Zvartau E. (1999) Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev 51,83-153.

Freissmuth M, Schutz W and Linder ME (1991) Interactions of the bovine brain Al-adenosine receptor with recombinant G protein alpha-subunits.

Selectivity for rGi alpha-3. JBiol Chem 266: 17778-17783.

Gerfen CR, Engber TM, Mahan LC, Susel Z, Chase TN, Monsma Jr FJ and Sibley DR (1990) Di and D2 dopamine receptor-regulated gene expression of striatonigral and striatopallidal neurons. Science 250: 1429-1432.

Herve D, Levi-Strauss M, Marey-Semper I, Verney C, Tassin JP, Glowinski J and Girault JA (1993) Golf and Gs in rat basal ganglia: possible involvement of Golf in the coupling of dopamine Dl receptor with adenylyl cyclase. J Neurosci 13 : 2237-2248.

Jankovic J. Complications and limitations of drug therapy for Parkinson's disease. Neurology. 2000 ; 55 (12 Suppl 6): S2-6.

Jarvis MF and Williams M (1989) Direct autoradiographic localization of adenosine A2 receptors in the rat brain using the A2-selective agonist, [3H] CGS 21680. EurJPharmacol 168: 243-246.

Jockers R, Linder ME, Hoehnegger M, Nanoff C, Bertin B, Strosberg AD, Marullo S and Freissmuth M (1994) Species difference in the G protein selectivity of the human and bovine Al-adenosine receptor. J Biol Chem 269: 32077-32084.

Jones DT, Masters SB, Bourne HR and Reed RR (1990) Biochemical characteriztion of three stimulatory GTP-binding proteins. The large and small forms of Gs and the olfactory-specific G-protein, Golf. JBiol Chem 265 : 2671-2676.

Jones DT and Reed RR (1987) Molecular cloning of five GTP-binding protein cDNA species from rat olfactory neuroepithelium. JBiol Chem 262: 14241- 14249.

Jones DT and Reed RR (1989) Golf : an olfactory neuron-specific G-protein involved in odorant signal transduction. Science 244: 790-795.

Kull B, Arslan G, Nilsson C, Owman C, Lorenzen A, Schwabe U and Fredholm BB (1999) Differences in the order of potency for agonists, but not antagonists, at human and rat adenosine A2A receptors. Biochem Pharmacol 57:65-75.

Laemmli UK (1970) Cleavage of structural proteins during the assembly of the heat of bacteriophage T4. Nature 227: 680-685.

Largent BL, Jones DT, Reed RR, Pearson RC and Snyder SH (1988) G protein mRNA mapped in rat brain by in situ hybridization. Proc Natl Acad Sci U S A 85: 2864-2868.

Le Moine C and Bloch B (1995) Di and D2 dopamine receptor gene expression in the rat striatum: sensitive cRNA probes demonstrate prominent segregation of Dl and D2 mRNAs in distinct neuronal populations of the dorsal and ventral striatum. J Comp Neurol 355: 418-426.

Le Moine C and Bloch B (1996) Expression of the D3 dopamine receptor in peptidergic neurons of the nucleus accumbens: comparison with the Di and D2 dopamine receptors. Neuroscience 73 : 131-143.

Le Moine C, Normand E and Bloch B (1991) Phenotypical characterization of the rat striatal neurons expressing the Di dopamine receptor gene. Proc Natl Acad Sci USA 88: 4205-4209.

Londos C, Cooper DM and Wolff J (1980) Subclasses of external adenosine receptors. Proc Natl Acad Sci USA ? ? : 2551-2554.

Munshi R, Pang 1-H, Sternweis PC and Linden J (1991) A1 adenosine receptors of bovine brain couple to guanine nucleotide-binding proteins Gii, Gi2 and Go. JBiol Chem 266: 22285-22289.

Nordstedt C and Fredholm BB (1990) A modification of a protein-binding method for rapid quantification of cAMP in cell-culture supernatants and body fluid. Anal Biochem 189 : 231-234.

Palmer TM, Benovic JL and Stiles GL (1995) Agonist-dependent phosphorylation and desensitization of the rat As adenosine receptor.

Evidence for a G-protein-coupled receptor kinase-mediated mechanism. J Biol Chem 270 : 29607-29613.

Parkinson FE and Fredholm BB (1990) Autoradiographic evidence for G- protein coupled A2-receptors in rat neostriatum using [3H]-CGS 21680 as a ligand. Naunyn Schmiedebergs Arch Pharmacol 342: 85-89.

Penit-Soria J, Durand C, Besson MJ and Herve D (1997) Levels of stimulatory G protein are increased in the rat striatum after neonatal lesion of dopamine neurons. Neuroreport 8: 829-833.

Pfeuffer T (1977) GTP-binding proteins in membranes and the control of adenylate cyclase activity. JBiol Chem 252: 7224-7234.

Pierce KD, Furlong TJ, Selbie LA and Shine J (1992) Molecular cloning and expression of an adenosine A2B receptors from human brain. Biochem Biophys Res Commun 187 : 86-93.

Schiffmann SN, Jacobs O and Vanderhaeghen JJ (1991) Striatal restricted adenosine A2 receptor (RDC8) is expressed by enkephalin but not by substance P neurons: an in situ hybridization histochemistry study. J Neurochem 57: 1062-1067.

Schwindinger WF, Francomano CA and Levine MA (1992) Identification of a mutation in the gene encoding the alpha subunit of the stimulatory G protein of adenylyl cyclase in McCune-Albright syndrome. Proc Natl Acad Sci USA 89: 5152-5156.

Svenningsson P, Le Moine C, Kull B, Sunahara R, Bloch B and Fredholm BB (1997) Cellular expression of adenosine A2A receptor messenger RNA in the rat central nervous system with special reference to dopamine innervated areas. Neuroscience 80 : 1171-1185.

Svenningsson P, Lindskog M, Rognoni F, Fredholm BB, Greengard P and Fisone G (1998) Activation of adenosine A2A and dopamine Di receptors stimulates cyclic AMP-dependent phosphorylation of DARPP-32 in distinct populations of striatal projection neurons. Neuroscience 84: 223-228.

Svenningsson P, Le Moine C, Fisone G, Fredholm BB. Distribution, biochemistry and function of striatal adenosine A2A receptors. Progress in Neurobiology 1999 ; 59: 355-396. van Calker D, Muller M and Hamprecht B (1979) Adenosine regulates via two different types of receptors, the accumulation of cyclic AMP in cultured brain cells. JNeurochem 33 : 999-1005.

Weinstein LS, Gejman PV, Friedman E, Kadowaki T, Collins RM, Gershon ES and Spiegel AM (1990) Mutations of the Gs alpha-subunit gene in Albright hereditary osteodystrophy detected by denaturing gradient gel electrophoresis. Proc Natl Acad Sci U S A 87: 8287-8290.

Worley PF, Baraban JM, De Souza EB and Snyder SH (1986) Mapping second messenger systems in the brain: differential localizations of adenylate cyclase and protein kinase C. Proc Natl Acad Sci U S A 83: 4053-4057.

Yang X, Lee Sr FY and Wand GS (1997) Increased expression of Gs (alpha) enhances activation of the adenylyl cyclase signal transduction cascade. Mol Endocrinol 11: 1053-1061.

Zhou QY, Li C, Olah ME, Johnson RA, Stiles GL and Civelli O (1992) Molecular cloning and characterization of an adenosine receptor: the As adenosine receptor. Proc Natl Acad Sci U S A 89: 7432-7436.

Zor T, Halifa I, Kleinhaus S, Chorev M and Selinger Z (1995) m-Acetylanilido- GTP, a novel photoaffinity label for GTP-binding proteins: synthesis and application. Biochem J 306 : 253-258.

Legends for figures : Figure 1 Distnbution of adenosine A2A, Golf and Gs in rat brain. Darkfield autoradiograms showing the expression of (a) adenosine A2A receptor mRNA, (b) Golf mRNA and (c) Gs mRNA in coronal sections of the rat forebrain.

Regions expressing the genes appear white. Scale bars = 1 mm.

Figure 2 Co-localization of adenosine A2A receptors with Go f and Gs. Emulsion autoradiograms from double in situ hybridization experiments showing (a) the co-localization of adenosine A2A receptor mRNA (silver grains) with Golf mRNA (dark cells) in caudate putamen (arrows indicate co-localization) and (b) the absence of co-localization of adenosine A2A receptor mRNA (silver grains) with Gs mRNA (dark cell) in caudate putamen (arrow indicates a single labeled neuron for Gs mRNA). Scale bars = 10, um.

Figure 3 Histograms showing the numbers of neurons in the lateral and medial caudate-putamen and in the core and shell of nucleus accumbens that are labeled (+) or not labeled (-) with probes for adenosine A2A receptor mRNA, Golf mRNA and C-s mRNA. All neurons within the examined areas were counted.

Figure 4 Agonist-stimulated photoaffinity labeling of Go f is influenced by GDP.

Membranes from striatum were photolabeled with [a32P] m-AcAGTP at various GDP concentrations in the absence (-) or presence (+) of CGS 21680 (10 1M). (A) autoradiogram showing solubilized labeled membranes directly subjected to SDS-PAGE or (B) one representative autoradiogram out of three independent experiments showing immunoprecipitated membranes with the subtype-specific G protein a subunit Golf antibody subjected to SDS-PAGE.

(C) Histograms showing the quantification of 10 uM CGS 21680 induced increased labeling of immunoprecipitated Golf subunits in percent of control at various GDP concentrations (n=3). (error bars = SEM).

Figure 5 Concentration dependent increase of Go f photolabeling by CGS 21680.

Membranes from striatum were stimulated in the presence of 100 uM GDP with increasing concentrations of CGS 21680 (1 nM to 10 uM). Goif was immunoprecipitated with the subtype-specific G protein a subunit Golf antibody and subjected to SDS-PAGE. The autoradiogram (A) showing the 45-kDa region is representative of three independent experiments. (B) Dose- response curve of CGS 21680 induced incorporation of [a32P] m-AcAGTP as percent of maximal incorporation (n=3) (error bars = SEM).

Figure 6 Dose-response curves of CGS 21680 induced cAMP accumulation in CHO cells stably expressing (A) adenosine A2A receptors and (N) Golf cDNA transiently transfected CHO cells expressing adenosine A2A receptors.

Results are mean of 2 experiments with duplicates and are presented as pmol cAMP/50 ul (error bars = SEM). The plateau is significant increased in Golf transfected cells compared with non transfected cells (t test P=0.0082).