Human A2B Adenosine Receptor Assay

The present invention concerns cloned human adenosine receptors of the A1, A2 and A3 class and their subtypes. The cloned dog A1 and A2a adenosine receptors have been reported. See F. Libert, et al., (1989) Science 244:569-572, C. Maennant, et al., Biochem. Biophys. Res. Comm., (1990) 173:1169-1178, and F. Libert, et al. (1991) EMBO J. 10:1677-1682. The cloned rat A1 adenosine receptor has been reported by L. C. Mahan, et al., (1991) Mol. Pharm. 40:1-7 and S. M. Reppert, et al., (1991) Mol. Endocrin. 5:1037-1048. We have now found that the human A1 adenosine receptor differs by 18 amino acids from the dog A1 sequence and 16 amino acids from the rat A1 sequence. The human A2a adenosine receptor differs by 28 amino acids from the dog A2a sequence.

Adenosine is a naturally occurring nucleoside which exhibits diverse and potent physiological actions in the cardiovascular, nervous, pulmonary, renal and immune systems. Adenosine has been demonstrated to terminate superventricular tachycardia through blockage of atrioventricular nodal conduction (J. P. DiMarco, et al., (1985) J. Am. Col. Cardiol. 6:417-425, A. Munoz, et al., (1984) Eur. Heart J. 5:735-738). Adenosine is a potent vasodilator except in the kidney and placenta (R. A. Olsson, (1981) Ann. Rev. Physiol. 43:385-395). Adenosine has been implicated as a preventative agent and in treatment of ventricular dysfunction following episodes of regional or global ischemia (M. B. Forman and C. E. Velasco (1991) Cardiovasc. Drugs and Therapy 5:901-908) and in cerebral ischemia(M. C. Evans, et al., (1987) Neurosci. Lett. 83:287, D. K. J. E., Von Lubitz, et al., (1988) Stroke 19:1133).

The instant invention also concerns an assay protocol which can be used for identifying and evaluating substances that bind to human adenosine receptors. The assay can be utilized to identify adenosine receptor agonists and antagonists and determine their binding affinity (R. F. Bruns, et al., (1983) Proc. Natl. Acad. Sci. USA 80:2077-2080; R. F. Bruns, et a1.,(1986) Mol. Pharmacol. 29:331-346; M. F. Jarvis, et al. (1989) J. Pharma. Exp. Therap. 251:888-893; K. A. Jacobson et al., (1989) J. Med. Chem. 32:1043-1051). Such adenosine receptor agonists, antagonists and binding enhancers have been identified and implicated for usage in the treatment of physiological complications resulting from cardiovascular, renal and neurological disorders. Adenosine receptor agonists have been identified for use as vasodilators ((1989) FASEB. J. 3(4) Abs 4770 and 4773, (19910 J. Med. Chem. (1988) 34:2570), antihypertensive agents (D. G. Taylor et al., FASEB J. (1988) 2:1799),and anti-psychotic agents (T. G. Heffner et al., (1989) Psychopharmacology 98:31-38). Adenosine receptor agonists have been identified for use in improving renal function (R. D. Murray and P. C. Churchill,(1985) J. Pharmacol. Exp. Therap. 232:189-193) Adenosine receptor allosteric or binding enhancers have shown utility in the treatment of ischemia, seizures or hypoxia of the brain (R. F. Bruns, et al. (1990) Mol. Pharmacol. 38:939-949; C. A. Janusz, et al., (1991) Brain Research 567:181-187). The cardioprotective agent, 5-amino-4imidazole carboxamide (AICA) ribose has utility in the treatment of ischemic heart conditions, including unstable angina and acute s myocardial infarction (H. E. Gruber, et al. (1989) Circulation 80: 1400-1414). Previous methods to date have proven inferior due to the presence of multiple subtypes present in the animal tissue utilized (R. F. Bruns et al., (1986) Mol. Pharm. 29:33 1-346) and the differences between species in the affinity for adenosine analogs and the physiological effects of adenosine (Ukera, et al., (1986) FEBS Lett, 209:122-128). Pure adenosine receptors make possible the identification and evaluation of compounds which have unique affinity for a single receptor subtype. Moreover, because of the variable effects of adenosine documented in other species, the utilization of human adenosine receptor subtypes is advantageous for the development of human therapeutic adenosine receptor agonists, antagonists or enhancers.

FIGS. 1A and 1B Full length amino acid sequence of human A1 adenosine receptor.

FIG. 2 Full length nucleotide sequence of the cloned human A1 adenosine receptor complementary DNA depicted from the 5' to 3' terminus.

FIGS. 3A and 3B Full length amino acid sequence of human A2a adenosine receptor.

FIG. 4 Full length nucleotide sequence of cloned human A2a adenosine receptor complementary DNA depicted from the 5' to 3' terminus.

FIGS. 5A and 5B Full length amino acid sequence of human A2b receptor.

FIG. 6 Full length nucleotide sequence of cloned human A2b adenosine receptor complementary DNA depicted from the 5' to 3' terminus.

FIGS. 7A and 7B Full length amino acid sequence of human A3 adenosine receptor.

FIG. 8 Full length nucleotide sequence of the cloned human A3 adenosine receptor complementary DNA depicted from the 5' to 3' terminus.

The invention concerns recombinant human adenosine receptors A1, A2a, A2b and A3, which have been prepared by cDNA cloning and polymerase chain reaction techniques. The invention also concerns expression systems for these receptors in COS cells, CHO cells and/or oocytes, as well as an assay using the expression system.

The recombinant adenosine receptors comprising the invention can be utilized in an assay to identify and evaluate entities that bind to or enhance binding to purified human adenosine receptors.

One embodiment of the invention concerns human A1 adenosine receptor, said receptor being free of other human receptor proteins.

Within this class, this embodiment concerns human A1 adenosine receptor from human tissue such as brain, kidney or heart, said receptor being free of other human proteins.

In a second class this embodiment concerns a protein comprising the amino acid sequence Leu-Thr-Gln-Ser-Ser (LTQSS) which is amino acids 90 to 94 in FIG. 1A. This sequence is also found in FIG. 3A, amino acids 87 to 91 and FIG. 5A, amino acids 88 to 92.

In a third class, this embodiment concerns a protein comprising the 326 amino acid sequence depicted in FIGS. 1A and 1B said protein being free of other human receptor proteins.

Within the third class this embodiment concerns a protein consisting of the 326 amino acid sequence as shown in FIGS. 1A and 1B.

This embodiment also concerns a cDNA sequence encoding human A1 adenosine receptor, said sequence being free of other human DNA sequences.

Within this aspect, the invention concerns the 981 nucleotide sequence of complementary DNA, as shown in FIG. 2 or a degenerate variation thereof.

A second embodiment concerns human A2a adenosine receptor, said receptor being free of other human receptor proteins.

Within this class, this embodiment concerns human A2a adenosine receptor from human tissue such as brain, kidney or heart, said receptor being free of other human proteins.

In a second class, this embodiment concerns a protein comprising the 412 amino acid sequence depicted in FIGS. 3A and 3B, said protein being free of other human receptor proteins.

Within the third class this embodiment concerns a protein consisting of the 412 amino acid sequence as shown in FIGS. 3A and 3B.

The second embodiment also concerns a complementary DNA sequence encoding human A2a adenosine receptor, said sequence being free of other human DNA sequences.

One class of the second embodiment the invention concerns the 1239 nucleotide sequence of complementary DNA, as shown in FIG. 4 or a degenerate variation thereof.

A third embodiment concerns human A2b adenosine receptor, said receptor being free of other human receptor proteins.

In one class this embodiment concerns human A2b adenosine receptor, said receptor being free of other human proteins.

Within this class, this embodiment concerns human A2b adenosine receptor from human tissue such as brain, said receptor being free of other human proteins.

In a second class, this embodiment concerns a protein comprising the 332 amino acid sequence depicted in FIGS. 5A and 5B, said protein being free of other human receptor proteins.

Within the third class this embodiment concerns a protein consisting of the 332 amino acid sequence as shown in FIGS. 5A and 5B.

The third embodiment also concerns a complementary DNA sequence encoding human A2b adenosine receptor, said sequence being free of other human DNA sequences.

One class of the third embodiment of the invention concerns the 999 nucleotide sequence of complementary DNA, as shown in FIG. 6 or a degenerate variation thereof.

A fourth embodiment concerns human A3 adenosine receptor, said receptor being free of other human receptor proteins.

In one class this embodiment concerns human A3 adenosine receptor, said receptor being free of other human proteins.

Within this class, this embodiment concerns human A3 adenosine receptor from human tissue such as brain, said receptor being free of other human proteins.

In a second class, this embodiment concerns a protein comprising the 3 18 amino acid sequence depicted in FIGS. 7A and 7B, said protein being free of other human receptor proteins.

Within the third class this embodiment concerns a protein consisting of the 318 amino acid sequence as shown in FIGS. 7A and 7B.

The fourth embodiment also concerns a complementary DNA sequence encoding human A3 adenosine receptor, said sequence being free of other human DNA sequences.

One class of the fourth embodiment of the invention concerns the 957 nucleotide sequence of complementary DNA, as shown in FIG. 8 or a degenerate variation thereof.

A fifth embodiment of this invention concerns systems for expressing human A1, A2a, A2b, or A3 adenosine receptor.

In one class this fifth embodiment of the invention comprises:

A plasmid which comprises:

(a) a mammalian expression vector, such as pSVL, and

(b) a nucleotide sequence encoding human A 1, A2a, A2b or A3 adenosine receptor protein.

A second class of this embodiment of the invention concerns a system for the expression of human A 1, A2a, A2b, or A3 adenosine receptor in a mammalian cell line (COS7 or CHO), or in oocytes.

A sixth embodiment of the invention concerns a method of using any of the above expression systems for determining the binding affinity of a test sample for human A 1, A2a, A2b, or A3 adenosine receptor respectively.

In one class this embodiment concerns an adenosine receptor binding assay for a test sample using a mammalian cell line, said mammalian cell line transfected with a plasmid, which plasmid comprises

(a) a mammalian expression vector, and

(b) the nucleotide sequence encoding a human adenosine receptor protein, the assay comprising the steps of:

(1) expressing said human adenosine receptor in a transfected cell line, selected from COS7 cells transfected with said plasmid or CHO cells transfected with said plasmid;

(2) in a solution, mixing a test sample with said transfected cell line and a radioactive ligand;

(3) incubating the mixture of Step 2;

(4) separating said radioactive ligand bound to said human adenosine receptor protein from unbound radioactive ligand; and

(5) measuring the radioactivity of said radioactive ligand bound to said human adenosine receptor;

said human adenosine receptor protein selected from the group comprising A 1, A2a, A2b or A3 human adenosine receptor protein.

There is a variety of ligands useful for characterizing the binding affinity of a given adenosine subtype. The following Table L is illustrative.

It will be a readily apparent to the skilled artisan that other ligands may be useful in adenosine receptor binding assays, including those listed in Table L but with a different radioactive label, and others. Testing of other ligands for high affinity binding is readily performed with the purified adenosine receptor subtypes described in the present invention.

As will be appreciated by those of skill in the art, there is a substantial amount of redundancy in the set of codons which translate specific amino acids. Accordingly, the invention also includes alternative base sequences wherein a codon (or codons) are replaced s with another codon, such that the amino acid sequence translated by the DNA sequence remains unchanged. For purposes of this specification, a sequence bearing one or more such replaced codons will be defined as a degenetate variation. Also included are mutations (exchange of individual amino acids) which one of skill in the art would expect to have no effect on functionality, such as valine for leucine, arginine for lysine and asparagine for glutamine. For purposes of this specification, amino acid sequences beating one or more of such mutations are used to substantially correspond to the sequence defined in FIGS. 1A and 1B, 3A and 3B, 4, 5A and 5B, 7A and 7B, respectively.

In overview, the present specification describes methods by which applicants have isolated human A1, A2a, A2b or A3 adenosine receptor complementary DNA (cDNA) without prior knowledge of its protein sequence or gene sequence. The human A1 and A2a cDNAs were obtained in part from a human heart cDNA library. In the approach, the regions of dog A1 and A2a adenosine receptors were selected by the applicants which they believed to be similar to the human A1 and A2a adenosine receptors. Oligonucleotide probes corresponding to these regions were designed and utilized to obtain part of the human A1 and A2a adenosine receptor cDNA sequence from a human heart cDNA library. The remaining parts of the human heart A1 and A2a adenosine receptor cDNAs were obtained by polymerase chain reaction (PCR) technique utilizing the above sequence information of human A1 and A2a adenosine receptor cDNA. The human A1 and A2a adenosine receptor cDNAs were also isolated from human brain and human kidney cDNA libraries utilizing the above sequence information of human A1 and A2a adenosine receptor cDNAs obtained from the human heart cDNA library. The A2b adenosine receptor subtype cDNA was obtained from a human cortex cDNA library utilizing the above sequence information of human adenosine receptor cDNA. The human A3 was isolated from human striata cDNA library, using a rat A3 fragment obtained by the polymerase chain reaction.

The complete sequence of the human A1, A2a, A2b and A3 adenosine receptor subtype was determined and their protein sequence deduced.

The cloned human adenosine receptor cDNAs were expressed in a heterologous expression system. Expression in COS cells (a monkey kidney cell line) was used to measure ligand binding properties of the human adenosine receptors. Such an expression system can be used for evaluation and identification of adenosine receptor agonists, antagonists or enhancers.

1) Isolation of human adenosine receptor cDNA

To isolate the human A1 and A2a adenosine receptor cDNAs in the absence of its sequence information, a human ventricle cDNA library was screened with oligonucleotide probes based on the published sequence of the dog A1 and A2a adenosine receptor cDNAs (F. Libert, et al., (1989) Science 244:569-572). Partial cDNA sequence was obtained for the human A1 and A2a adenosine receptor cDNAs. The remaining parts of the human A1 and A2a adenosine receptor cDNAs were obtained by PCR using a modification of the method described by Froham et al. ((1988), Proc. Natl. Acad. Sci., 85:8998-9002). PCR primers corresponding to human sequence obtained from the partial human ventricular A1 and A2a adenosine receptor cDNAs were designed and used in PCR amplification of human ventricular mRNA.

The human A1 and A2a adenosine receptors were also isolated from human brain and kidney cDNA libraries. Human oligonucleotide probes were designed based on the human ventricular A1 and A2a adenosine receptor cDNA sequence. The sequence of the brain and kidney A1 adenosine receptor cDNA was identical to the human ventricular A1 adenosine receptor cDNA obtained in part from the human ventricle cDNA library and by PCR methodology.

The human A2b adenosine receptor subtype was obtained from a human cortex cDNA library using human oligonueleotide primers corresponding to a region of the human adenosine receptor cDNA sequence.

The human A3 adenosine receptor subtype was obtained from a striata library using a rat cDNA probe generated by the polymerase chain reaction. Oligonucleotides were designed based on the rat A3 sequence (Zhou et al, Proc. Natl. Acad. Sci.89, 7432(1992)).

2) Expression of the cloned human adenosine receptors

The human A1, A2a, A2b and A3 receptor subtype cDNAs were subcloned into the expression vector pSVL (PHARMACIA). Transient expression in COS7 cells (monkey kidney cell line, ATCC CRL 1651, ATCC, Rockville, Md.) was accomplished by transfection of the cloned adenosine receptor cDNAs under the control of the SV40 promoter into mammalian cells (e.g., COS7). Membranes prepared from the transfected cells were utilized for the determination of binding affinity, selectivity and specificity of the human adenosine receptors for various ligands. Stable expression of the human adenosine receptors in mammalian cells (e.g., COS7 and CHO) was achieved after integration of the transfected cDNA into the chromosomes of the host cells. These stable cell lines constituently express the cloned human adenosine receptors and can be propagated infinitely. Stable cell lines expressing the human adenosine subtype cDNAs individually can be used in the binding assay to measure the affinity and selectivity of the receptors for adenosine agonists, antagonists and enhancers.

Membranes prepared from transfected COS7 cells were utilized in a binding assay to measure the affinity of the human adenosine receptors for the radiolabeled adenosine agonists, [.sup.3 H]-cyclohexyladenosine (CHA), [.sup.3 H]-CGS21680 (2-(p-(2-carboxyethyl)-phenylamino)-5'-N-ethyl-carboxamidoadenosine), [.sup.3 H]-5'-N-ethylcarboxamido adenosine ([.sup.3 H]-NECA), or [.sup.125 I]-N.sup.6 -aminobenzyl adenosine (.sup.125 I-ABA). Monolayer cell culture of transfected COS7 cells were dissociated with 1 mM EDTA in phosphate buffered saline and resuspended in 5 mM Tris, pH7.6/10 mM MgCl.sub.2. The cells were subjected to freeze-thaw lysis and the suspension was homogenized in a glass dounce homogenizer. The membranes were pelleted, resuspended in binding buffer, 50 mM Tris pH 7.6/10 mM MgCl.sub.2 and incubated with adenosine deaminase before the binding assay. The binding assay was performed by incubating 100 mg of membranes with increasing concentrations of radiolabeled adenosine agonists. Bound ligand was separated from free ligand by filtration on a SKATRON CELL HARVESTER equipped with a receptor binding filtermat. Bound radioactivity was measured by scintillation counting. Substances which bind to or enhance binding to expressed human adenosine receptors in COS and CHO cells can be identified in competition binding assays with radiolabeled adenosine analogs. For the competition binding assay, membranes were incubated with 5 nM [.sup.3 H]-CHA 5 nM [.sup.3 H]-CGS21680 or 10 Nm [.sup.3 H]-NECA and various concentrations of adenosine agonists or antagonists.

A transient expression system in Xenopus oocytes was established by microinjection of in vitro transcribed mRNA from the cloned adenosine receptor cDNAs. The expression system allows the measurement of the biological effects (i.e., changes in cAMP levels) upon activation of the expressed adenosine receptors with ligand binding. The cAMP levels are measured by a radioimmunoassay (RIANEN kit, DuPont/NEN) using the acetylation protocol. Activation of the expressed receptors by ligand binding are coupled to either increases or decreases in the intracellular cAMP levels dependent upon the subtype of adenosine receptor (Van Calker et al., I(1979) J. Neurochem. 33:999-1003; Londos et al. (1980) Proc. Natl. Acad. Sci. USA 77:2551-2554). The activity of any potential adenosine receptor agonist can be evaluated by measuring the changes in cAMP levels in oocytes injected with adenosine receptor mRNA but not in uninjected or negative control injected oocytes. The activity of any potential adenosine receptor antagonist can be evaluated by determining the inhibition of the cAMP response induced by adenosine in oocytes injected with adenosine receptor transcripts but not negative control or uninjected oocytes. The changes in cAMP accumulation can alternatively be measured in stably transfected CHO cells expressing the human adenosine receptor subtypes.

The cAMP accumulation assay has a number of advantages over the binding assay established in the mammalian cell expression system as a screen for adenosine receptor modulating agents. The assay allows the measurement of a biological effect (i.e., changes in cAMP levels) resulting from the activation of the expressed receptors by ligand binding. The native agonist adenosine is utilized in the assay to activate the expressed receptors. The functionality of additional adenosine receptor subtypes identified by molecular cloning which may not have defined ligands for binding analysis can be evaluated with the natural agonist and without prior identification of a selective, high affinity, radiolabeled ligand.

STEP A:

In the first step of obtaining the partial cDNAs encoding the human A1 and A2a adenosine receptors, total RNA was extracted by homogenizing 2.3 g human ventricle in 20 ml 5M guanidine isothiocyanate, 0.1M sodium citrate, pH 6.3, 1 mM EDTA, pH 7.0, 5% beta-mercaptoethanol, and 0.5% sodium lauryl sarcosinate. The homogenate was centrifuged for 10 min. at 10,000 rpm and the resulting supernatant was layered onto a cushion of 5.7M CsCl/0.1M EDTA, pH 7.0. After 20 hrs. of centrifugation at 24,000 rpm, the resulting pellet was precipitated one time and then passed over an oligo(dT)-cellulose (PHARMACIA, Piscataway, N.J.) column to isolate poly(A)+RNA.

An oligo(dT) primed library was synthesized from 5 .mu.g of the poly(A).sup.+ human ventricle RNA using the YOU-PRIME cDNA SYNTHESIS KIT (PHARMACIA, Piscataway, N.J.). See Gubler and Hoffman Gene 25:263 (1983). The resulting double-stranded cDNA was ligated into .lambda.gt10 EcoRI arms (PROMEGA, Madison, Wis.) and packaged according to the GIGAPACK II GOLD PACKAGING EXTRACT protocol (STRATAGENE, La Jolla, Calif.). See Huynh et al. (1985) DNA Cloning Techniques: A Practical Approach, IRL Press, Oxford, p.49 and Kretz et al. Res. 17:5409.

The E. coli strain C600Hfl (PROMEGA, Madison, Wis.) was infected with library phage, plated on agar plates, and incubated at 37.degree. C. The phage DNA was transferred to HYBOND-N nylon membranes (AMERSHAM, Arlington Heights, Ill.) according to the manufacturer's specifications.

Synthetic probes were constructed from overlapping oligonucleotides (A1 probe: 62+63, A2 probe: 52+53; see Table I for their sequences) based on the published dog A1 (RDC7) and A2a(RDC8) sequences (F Libert, et a1,(1989) Science 244:569-572). The oligonucleotides were annealed and filled-in with .alpha..sup.32 P-dCTP (NEN, Wilmington, Del.) and Klenow enzyme. The filters were hybridized with the appropriate probe in 5.times.SSC, 30% formamide, 5.times.Denhardt's solution, 0.1% SDS, and 0.1mg/ml sonicated salmon sperm DNA at 42.degree. C., overnight. Following hybridization the filters were washed to a final stringency of 6.times.SSC at 50.degree. C. and exposed to X-OMAT AR film (KODAK, Rochester, N.Y.) at -70.degree. C. The resulting positives were plaque purified by two additional rounds of plating and hybridization. Insert DNA was excised with NotI and ligated into NotI digested pBLUESCRIPT II KS+(STRATAGENE, La Jolla, Calif.). (Genebank #52327) DNA sequences were determined by the SEQUENASE protocol (USBC, Cleveland, Ohio). See Tabor and Richardsaon, J. Biol. Chem. 264 pp 6447-6458. Two clones were isolated in these screens. The human ventricle A1 cDNA (hval-3a) and human ventricle A2a cDNA (hva2-13) contain portions of coding sequences for proteins homologous to the reported dog A1 and A2a cDNAs, respectively. The coding region of the human A1 clone corresponds to nucleotides 482 through 981 (FIG. 2) and is 92% identical to the dog A1 sequence at the nucleotide level. The coding region of the human A2a clone corresponds to nucleotides 497 through 1239 (FIG. 4), and is 90% identical to the dog A2a sequence at the nucleotide level.

STEP B:

The human ventricle A1 adenosine receptor partial cDNA (hvA1-3a) is a 543 bp NotI fragment containing 23 bp 3' untranslated sequence and is 460 bp short of the initiation methionine based on sequence homology to the dog A1 cDNA. A modification of the 5' RACE (rapid amplification of cDNA ends) method (M. A. Frohman et al, (1988), Proc. Natl. Acad. Sci. USA, 85:8998-9002) was used to generate the 5' coding region of the cDNA. First strand cDNA was synthesized from 1 .mu.g of the human ventricle poly(A).sup.+ RNA in a total volume of 40 .mu.l containing 50 mM Tris, pH 8.0, 140 mM KCl, 10 mM MgCl.sub.2, 10 mM DTT, 15 mM each dNTP, 20 units RNasin (PROMEGA, Madison, Wis.), 20 pmol human primer 79 (see Table I), and 9.2 units AMV reverse transcriptase at 37.degree. C. for 2 hrs. The reaction was then diluted to 120 ml with 0.5 mM Tris, pH 7.6/0.05 mM EDTA and passed through a SEPHACRYL S-300 SPUN COLUMN (PHARMACIA, Piscataway, N.J.). The product in the column effluent was polyadenylated in 100 mM potassium cacodylate, pH 7.2, 2 mM CoCl.sub.2, 0.2 mM DTT, 0.15 mM dATP, and 14 units terminal deoxynucleotidyl transferase in a total volume of 31 .mu.l for 10 min. at 37.degree. C. The reaction was terminated by heating at 65.degree. C. for 15 min. and then diluted to 500 .mu.l with 10 mM Tris, pH 8.0/1 mM EDTA (TE).

Ten .mu.l of the poly(A)-tailed first strand cDNA was used as template in a primary PCR amplification reaction according to the GENEAMP protocol (PERKIN ELMER CETUS, Norwalk, Conn.; see Saiki et al. (1988) Science 239:487-491) containing 10 pmol primer 70, 25 pmol primer 71, and 25 pmol human primer 80 (see table I) in a total volume of 50 .mu.l. Primer 70 is 5'-gactcgagtcgacatcga(t).sub.17, primer 71 is 5'-gactcgagtcgacatcga, and both are based on M. A. Frohman, et al (1988), Proc. Natl. Acad. Sci. USA, 85:8998-9002. One cycle of PCR was performed of 1 min at 95.degree. C., 2 min at 50.degree. C., 40 min at 72.degree. C., followed by 40 cycles of 40 sec at 94.degree. C., 2 min at 56.degree. C., 3 min at 72.degree. C. The primary PCR amplification reaction product was electrophoresed through a 1.4% agarose gel and an area corresponding to approximately 600 bp was excised. The gel slice was melted and 1 .mu.l was used as template in a secondary PCR amplification reaction containing 100 pmol primer 71 and human primer 81 (see Table I) for 30 cycles of 1 min at 94.degree. C., 2 min at 56.degree. C., 3 min at 72.degree. C. The secondary PCR amplification product was digested with EcoRI and SalI and electrophoresed on a 1.4% agarose gel. An area corresponding to 500-600 bp was excised and ligated into EcoRI/SalI digested pBLUESCRIPT II KS+ (STRATAGENE, La Jolla, Calif.). The sequence of the 515 bp PCR product (5'HVA1-9) was determined by the SEQUENASE protocol (USBC, Cleveland, Ohio). The partial human ventricle A1 cDNA and the PCR product contain overlapping sequence and represent the complete coding region for the human A1 receptor, including 14 and 23 bp of 5' and 3' untranslated sequences, respectively. The sequence of the human A1 adenosine receptor cDNA so identified, is shown in FIG. 2.

STEP C:

A probe was generated by Klenow enzyme extension, including .alpha..sup.32 P-dCTP, of annealed oligonucleotides 62 and 63, and used to screen a human kidney cDNA library (CLONTECH, Palo Alto, Calif.). E. coli strain C600hfl (PROMEGA, Madison, Wis.) was infected with library phage and grown overnight on agar plates at 370C. Phage DNA was transferred to HYBOND-N nylon filters according to the manufacturer's protocol (AMERSHAM, Arlington Heights, Ill.). The probe was incubated with the filters in 750 mM NaCl, 75 mM sodium citrate, 30% formamide, 0.1% sodium dodecyl sulfate, 0.5 mg/mL polyvinylpyrrolidone, 0.5 mg/mL bovine serum albumin, 0.5 mg/mL Ficoll 400, and 0.1 mg/mL salmon sperm DNA, at 42.degree. C. overnight. The filters were washed in 0.9M NaCl and 90 mM sodium citrate at 50.degree. C. A positively hybridizing phage (hkA1-14), was identified and purified by replating and screening with the probe twice more. The final phage plaque was transferred to 0.5 mL 50 mM Tris, pH 7.5, 8 mM MgSO.sub.4, 85 mM NaCl, 1 mg/mL gelatin, and 1 .mu.L of a 1:50 dilution in water of the phage stock was used as template for PCR amplification. 50 pmol each of lamL and lamR (Table I) oligonucleotide primers were included, and subjected to 30 cycles of 40 sec at 94.degree. C., 1 min at 55.degree., 3 min at 72.degree., then a final 15 min at 72.degree., according to the GENEAMP protocol (PERKIN ELMER CETUS, Norwalk, Conn.). A 2.0 kb product was identified by agarose gel electrophoresis, and this was subcloned into the EcoRI site of pBLUESCRIPT II KS+ (STRATAGENE, La Jolla, Calif.). Sequence analysis by the SEQUENASE protocol (USBC, Cleveland, Ohio) demonstrated that this cDNA was homologous to the reported dog A1 clone. SmaI and EcoRI digestion released a DNA fragment containing coding sequence from base pair 76 through the translation STOP codon (FIG. 2) that is identical to the human ventricle A1 cDNA sequence (clones hval-3a and 5'hval-9). This fragment was used in construction of the full length coding sequence (see below). The human kidney cDNA also includes about 900 bp of 3' untranslated sequence.

STEP D:

The human ventricle A2a adenosine receptor partial cDNA (hvA2-13) is a 1.6 kb NotI fragment containing approximately 900 bp 3' untranslated sequence and is 496 bp short of the initiation methionine based on sequence homology to the dog A2a cDNA clone. Two consecutive rounds of 5' RACE were utilized to generate the 5' coding region of the cDNA. First strand cDNA was synthesized from 1 .mu.g of the human ventricle poly(A).sup.+ RNA in a total volume of 40 .mu.l containing 50 mM Tris, pH 8.0, 140 mM KCl, 10 mM MgCl.sub.2, 10 mM DTT, 15 mM each dNTP, 20 units RNasin, 20 pmol human primer 68 or 74 (for 1st or 2nd round RACE respectively), and 9.2 units AMV reverse transcriptase at 37.degree. C. for 2 hrs. The reaction was then diluted to 120 .mu.l with 0.5 mM Tris, pH 7.6/0.05 mM EDTA and passed through a SEPHACRYL S-300 SPUN COLUMN. The products in the column effluents were polyadenylated in 100 mM potassium cacodylate, pH 7.2, 2 mM CoCl.sub.2, 0.2 mM DTT, 0.15 mM dATP, and 14 units terminal deoxynucleotidyl transferase in a total volume of 31 .mu.l for 10 min. at 37.degree. C. The poly(A) tailing reaction was terminated by heating at 65.degree. C. for 15 min. and then diluted to 500 .mu.l with TE.

Five or 10 .mu.l (for 1st or 2nd round RACE respectively) of the poly(A) tailed first strand cDNA was used as template in the PCR amplification reaction according to the GENEAMP protocol containing 10 pmol primer 70, 25 pmol primer 71 (primer 70 and 71 sequences are given above), and 25 pmol human primer 69 or 75 (1st or 2nd round RACE respectively; see Table I) in a total volume of 50 .mu.l. One cycle of PCR was performed of 1 min at 95.degree. C., 2 min at 50.degree. C., 40 min at 72.degree. C., followed by 40 cycles of 40 sec at 94.degree. C., 2 min at 56.degree. C., 3 min at 72.degree. C. The PCR amplification products were digested with EcoRI and SalI and electrophoresed on a 1.4% agarose gel. Areas corresponding to 200-400 bp were excised and ligated into EcoRI/SalI digested pBLUESCRIPT II KS+ (STRATAGENE, La Jolla, Calif.). The sequences of the two A2a PCR products, the 332 bp 1st round RACE product (5'hvA2-14) and the 275 bp 2nd round RACE product (5'hvA2-29) were determined by the SEQUENASE (USBC, Cleveland, Ohio) protocol. By sequence homology comparisons with the dog A2a adenosine receptor cDNA sequence, the 1st round RACE product (5'hvA2-14) was 258 bp short of the initiation methionine and the second round RACE product (5'HVA2-29) was determined to extend 1 bp upstream of the initiation methionine. The human ventricle A2a partial cDNA clone (hvA2-13) and the human A2a PCR products (5'hvA2-14 and 5'hva2-29) contain overlapping sequence and together represent the complete coding sequence for the human adenosine A2a receptor, and include 1 bp and 0.8 kb of 5' and 3' untranslated sequence, respectively. The sequence of the human A2a adenosine receptor is shown in FIG. 4.

STEP E:

A double-stranded DNA probe was generated by Klenow enzyme extension, including .alpha..sup.32 P-dCTP, of annealed oligonucleotides 66 and 67, and used to screen a human striata cDNA library (STRATAGENE, La Jolla, Calif.). The oligonucleotide sequence was based on a region of the human ventricle A2a cDNA sequence. E. coli strain XL 1-blue (STRATAGENE, La Jolla, Calif.) cells were infected with library phage and grown overnight on agar plates at 37.degree. C. Phage DNA was transferred to HYBOND-N nylon filters according to the manufacturer's protocol (AMERSHAM, Arlington Heights, Ill.). The probe was incubated with the filters in 750 mM NaCl, 75 mM sodium citrate, 10% formamide, 0.5% sodium dodecyl sulfate, 0.5 mg/mL polyvinylpyrrolidone, 0.5 mg/mL bovine serum albumin, 0.5 mg/mL Ficoll 400, and 0.02 mg/mL salmon sperm DNA, at 42.degree. C. overnight. The filters were washed in 0.9M NaCl and 90 mM sodium citrate at 50.degree. C. A positively hybridizing phage (hbA2-22A) was identified and purified by replating and screening with the probe twice more, and subcloned into the plasmid pBLUESCRIPT SK-- by the manufacturer's protocol (STRATAGENE, La Jolla, Calif.). See Short et al. (1988) Nucl. Acids Res. 16:7583-7600; Sorge (1988) Stratagies 1:3-7. The human brain A2a adenosine receptor cDNA (hbA2-22A) spans bp 43 of the A2 coding sequence (FIG. 4) through the translation STOP codon, and includes about 900 bp of 3' untranslated sequence. The sequence of this human brain A2a cDNA is identical to the human ventricle A2a adenosine receptor cDNA (hvA2-13, 5'hvA2-14 and 5'hvA2-29).

STEP F:

A double-stranded DNA probe was generated by Klenow enzyme extension of annealed oligonucleotides 129 and 130, including .alpha..sup.32 P-dCTP, and used to screen a human frontal cortex cDNA library (STRATAGENE, La Jolla, Calif.). The oligonucleotide sequence was based on a region of the human A2a and A1 cDNA sequence. E. coli strain XL-1 blue (STRATAGENE, La Jolla, Calif.) cells were infected with library phage and grown overnight at 37.degree. C. Phage DNA was transferred to HYBOND-N nylon filters according to the manufacturer's protocol (AMERSHAM, Arlington Heights, Ill.). The probe was incubated with the filters in 750 mM NaCl, 75 mM sodium citrate, 10% formamide, 0.5% sodium dodecyl sulfate, 0.5 mg/mL polyvinyl-pyrrolidone, 0.5 mg/mL bovine serum albumin, 0.5 mg/mL Ficoll 400, and 0.02 mg/mL salmon sperm DNA, at 42.degree. C. overnight. The filters were washed in 0.9M NaCl and 90 mM sodium citrate at 50.degree. C. A positively hybridizing phage (hb-32c), was identified and purified by replating and screening with the probe twice more. The insert was subcloned to the plasmid pBLUESCRIPT SK--according to the manufacturer's protocol (STRATAGENE, La Jolla, Calif.). Sequence analysis by the SEQUENASE protocol (USBC, Cleveland, Ohio) demonstrated a complete open reading frame coding for amino acid sequence homologous to both of the previously isolated human A1 and A2a clones. This homologous adenosine receptor subtype cDNA is the A2b subtype having the sequences in FIGS. 5A, 5B and 6. A 1.3 kb SmaI-XmnI fragment was ligated into the SmaI site of pSVL (PHARMACIA, Piscataway, N.J.), giving the full length coding sequence of this homologous adenosine receptor in a plasmid suitable for its expression in COS cells. See Sprague et al. (1983) J. Virology 4.5:773; Templeton and Eckhart (1984) Mol. Cell Biol. 4:817.

Table I:

Sequences and directions of the primers used in the isolation of cDNA's and construction of expression plasmids, along with the positions in the clones upon which the sequences are based. Dog A1 and A2a cDNA clones are from F. Libert, et al, (1989) Science 244:569-572. Primers LamL and LamR are based on the sequence of .lambda.gt10 (T. V. Hyunh, et al. (1985) DNA Cloning: A Practical Approach, Vol 1, D. Glover, ed, IRL Press, Oxford). The homologous adenosine receptor subtype encoded by the clone hb32C was determined to be the A2b adenosine receptor subtype on the basis of the binding profile of the adenosine receptor agonist NECA and affinities for adenosine receptor antagonists measured on membranes prepared from pSVLhb32C transfected COS7 cells.

EXAMPLE 2

Human A1 adenosine receptor expression construct:

To express the human adenosine receptor cDNA in COS cells, the 118bp SalI-SmaI fragment of the human ventricle A1 PCR product (5'HVA1-9) was ligated together with the 1.8 SmaI-EcoRI fragment of the human kidney A1 adenosine receptor cDNA (hkA1-14) and the 3.0 kb SalI-EcoRI fragment of pBLUESCRIPT II KS+, resulting in a plasmid containing the contiguous full length coding sequence for the human A1 adenosine receptor cDNA and some 5' and 3' untranslated sequence. This plasmid was digested first with EcoRI, the resulting ends were filled in by Klenow enzyme extension and then the plasmid was digested with XhoI to release a fragment of 1.9 kb containing the full length human A1 adenosine receptor cDNA. The fragment was subcloned into the expression vector pSVL (PHARMACIA) which had been digested with XhoI-SmaI.

Human A2a adenosine receptor expression construct:

To express the human A2a adenosine receptor cDNA in COS cells, a contiguous A2a cDNA sequence was constructed before subcloning into the expression vector, pSVL. Primer 131, containing an XbaI recognition site, 14 bp of 5' untranslated sequence of human A1 adenosine receptor cDNA, and the first 18 bp of human A2a adenosine receptor cDNA coding sequence was used with primer 75 in PCR with 1 ng of the plasmid containing the human ventricle A2a 2nd round RACE product (5'hvA2-29) as template. Twenty-five cycles of 40 sec at 94.degree. C., 1 min at 55.degree. C., and 3 min at 72.degree. C., then a final incubation of 15 min at 72.degree. C., with 1 ng of plasmid template and 50 pmol of each primer in a volume of 50 .mu.L according to the GENEAMP protocol (PERKIN ELMER CETUS, Norwalk, Conn.), resulted in the expected 302 bp product determined by agarose gel electrophoresis. The 172 bp XbaI-EagI digestion product of this DNA fragment was ligated together with 1125 bp EagI-BglII digestion product of the human striata A2a adenosine receptor cDNA (hbA2-22A) and XbaI-SmaI digested pSVL (PHARMACIA), generating the full length human A2a adenosine receptor cDNA coding sequence in a plasmid suitable for its expression in COS cells.

Mammalian cell expression:

COS7 cells (ATCC #1651-CRL) were grown in complete medium, Dulbecco's modified Eagles's medium, DMEM (GIBCO, Grand Island, N.Y.) containing 10% fetal bovine serum, 100 U/mL penicillin-streptomycin and 2 mM glutamine, in 5% CO.sub.2 at 37.degree. C. Transient transfection of COS7 cells was performed by the CaPO.sub.4 method (Graham, F. L. and Van Der Erb, A. J. (1973) Virology 52:456-567) using the Mammalian Transfection Kit (STRATAGENE). See Chen and Okayama Mol. Cell Biol. 7:2745-2752. Plasmid DNA (15 .mu.g) was precipitated with 125 mM CaCl.sub.2 in BBS (N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid buffered saline) at room temperature for 30 minutes. The DNA precipitate was added to the COS7 cells and incubated for 18 h in 5% CO.sub.2 at 37.degree. C. The precipitate was removed and the cells were washed twice with serum free DMEM. Cells were incubated in complete medium in 5% CO.sub.2 at 37.degree. C. for 48 h prior to the binding assay.

Stable expression in COS 7 or CHO cells:

To establish stable cell lines, COS7 cells or CHO cells were co-transfected with 20 .mu.g of pSVL containing the adenosine receptor cDNA and 1 .mu.g of pWLneo (STRATAGENE) containing the neomycin gene. See Southern and Berg (1982) J. Mol. App. Gen. 1:327-341. Transfection was performed by the CaPO.sub.4 method. DNA was precipitated at room temperature for 30 minutes, added to the COS7 cells and incubated 18 h in 5% CO.sub.2 at 37.degree. C. The precipitate was removed and the cells were washed twice with serum free DMEM. Cells were incubated for 24 h in 5% CO.sub.2 at 37.degree. C., replated in 24-well dishes at a dilution of 1:10, and incubated an additional 24 h before adding selection medium, DMEM containing 10% fetal bovine serum, 100 U/mL penicillin-streptomycin, 2 mM glutamine and 0.5 mg/mL G418 (GIBCO). Transfected cells were incubated at 5% CO.sub.2, 37.degree. C. until viable colonies were visible, approximately 14-21 days. Colonies were selected and propagated. The cell clone with the highest number of human adenosine receptors was selected for subsequent application in the binding assay.

Binding studies:

Membranes were prepared from transiently transfected COS7 cells 48 h after transfection or from G418-selected stably transfected COS7 or CHO cells. Cells were harvested in 1 mM EDTA in phosphate buffered saline and centrifuged at 2000.times. g for 10 minutes. The cell pellet was washed once with phosphate buffered saline. The cell pellet was resuspended in 2 mL of 5 mM Tris, pH 7.6/5 mM MgCl.sub.2. Membranes were prepared from the cells by freeze-thaw lysis in which the suspension was frozen in a dry ice/ethanol bath and thawed at 25.degree. C. twice. The suspension was homogenized after adding an additional 2 mL of 5 mM Tris, pH 7.6/5 mM MgCl.sub.2, in a glass dounce homogenizer with 20 strokes. The membranes were pelleted at 40,000.times. g at 4.degree. C. for 20 minutes. The membrane pellet was resuspended at a protein concentration of 1-2 mg/mL in binding assay buffer, 50 mM Tris, pH 7.6/10 mM MgCl.sub.2. Protein concentration was determined by the method of Bradford ((1976) Anal. Biochem. 72: 248-250). Before the binding assay was performed, the membranes were incubated with adenosine deaminase (BOEHRINGER MANNHEIM), 2 U/mL for 30 minutes at 37.degree. C. Saturation binding of [.sup.3 H]-cyclohexyladenosine (CHA) was performed on membranes prepared from pSVLA1 transfected COS7 or CHO cells.

Membranes (100 .mu.g) were incubated in the presence of 0.2 U/mL adenosine deaminase with increasing concentrations of CHA (NEN, 32 Ci/mmol) in the range of 0.62-30 Nm for 120 minutes at 25.degree. C. in a total volume of 500 .mu.L. The binding assay was terminated by rapid filtration and three washes with ice-cold 50 mM Tris,pH 7.6/10 mM MgCl.sub.2 on a SKATRON CELL HARVESTER equipped with a receptor binding filtermat (SKATRON INSTRUMENTS, INC.). Non-specific binding was determined in the presence of 100 .mu.M N.sup.6 -cyclopentyladenosine (CPA). Bound radioactivity was measured by scintillation counting in READY SAFE SCINTILLATION COCKTAIL. (BECKMAN). For competition binding experiments, membranes were incubated with 5 Nm [.sup.3 H]-CHA and various concentrations of A1 adenosine receptor agonists. Saturation binding of [.sup.3 H]CGS-21680 was performed on membranes prepared from pSVLA2a transfected COS7 cells. Membranes (100 .mu.g) were incubated in the presence of 0.2 U/mL adenosine deaminase with increasing concentrations of CGS21680 (NEN, 48.6 Ci/mmol) in the range of 0.62-80 Nm for 90 minutes at 25.degree. C. in a total volume of 500 .mu.L. The binding assay was terminated by rapid filtration with three washes with ice-cold 50 mM Tris, pH 7.6/10 mM MgCl.sub.2 on a Skatron cell harvester equipped with a receptor binding filtermat (SKATRON INSTRUMENTS, INC.). Non-specific binding was determined in the presence of 100 .mu.M CPA. Bound radioactivity was measured by scintillation counting in READY SAFE LIQUID SCINTILLATION COCKTAIL (BECKMAN). For competition binding experiments, membranes were incubated with 5 Nm [.sup.3 H]-CGS21680 and various concentrations of A2 adenosine receptor agonists.

Saturation binding of [.sup.3 H]5'-N-ethylcarboxamidoadenosine (NECA) was performed on membranes (100 .mu.g) prepared from pSVLhb32C transfected COS7 cells in the presence of adenosine deaminase with increasing concentrations of NECA (NEN, 15.1 Ci/mmol) in the range of 1.3-106 Nm for 90 minutes at 25.degree. C. in a total volume of 500 .mu.L. The assay was terminated by rapid filtration and three washes with ice-cold binding buffer on a cell harvester equipped with a receptor binding filtermat (SKATRON INSTRUMENTS, INC.). Bound radioactivity was measured by scintillation counting. Non-specific binding was measured on membranes prepared from non-transfected COS7 cells. For competition binding experiments, membranes from transfected cells were incubated with 10 Nm [.sup.3 H]NECA and varying concentrations of adenosine receptor antagonists.

The human A3 adenosine receptor was cloned from a human striata cDNA library. Oligonucleotide probes were designed based on the rat A3 sequence of Zhou et al., Proc. Natl. Acad. Sci. 89, 7432 (1992). The complete sequence of the human A3 adenosine receptor was determined and the protein sequence deduced. The cloned human A3 adenosine receptor is expressed in a heterologous expression system in COS and CHO cells. Radiolabeled adenosine receptor agonists and antagonists are used to measure the binding properties of the expressed receptor. Stable cell lines can be used to evaluate and identify adenosine receptor agonists, antagonists and enhancers.

STEP A:

A synthetic probe homologous to the rat A3 adenosine receptor was generated using the polymerase chain reaction (PCR). Three .mu.l of rat brain cDNA was used as template in a PCR amplification reaction according to the GENEAMP protocol (PERKIN ELMER CETUS, Norwalk, Conn.) containing 50 pmol of primers 207 (5'-cccaagcttatgaaagccaacaatacc) (SEQ. ID NO: 27) and 208 (5'-tgctctagactctggtatcttcacatt) (SEQ. ID NO: 28) in a total volume of 50 .mu.l. Primers 207 and 208 are based on the published rat A3 adenosine receptor sequence (Zhou, et al, (1992), Proc. Natl. Acad. Sci. USA, 89:7432-7406). Forty cycles of 40 sec at 94.degree. C., 1 min at 55.degree. C., 3 min at 72.degree. C. were performed and the resulting 788 bp fragment was subcloned into HindIII-XbaI digested pBLUESCRIPT II KS+ (STRATAGENE, La Jolla, Calif.). The sequence was verified by the SEQUENASE protocol (USBC, Cleveland, Ohio).

STEP B:

The 788 bp PCR fragment was labeled with .alpha..sup.32 P-dCTP using the MULTIPRIME DNA LABELLING SYSTEM (AMERSHAM, Arlington Heights, Ill.) and used to screen a human striata cDNA library (STRATAGENE, La Jolla, Calif.). E. coli strain XL-1 Blue (STRATAGENE, La Jolla, Calif.) cells were infected with library phage and grown overnight at 37.degree. C. Phage DNA was transferred to HYBOND-N nylon filters according to the manufacturer's protocol (AMERSHAM, Arlington Heights, Ill.). The probe was incubated with the filters in 5.times.SSC, 30% formamide, 5 .times.Denhardt's solution, 0.5% sodium dodecyl sulfate, and 50 .mu.g/ml sonicated salmon testis DNA. The filters were washed in 2.times.SSC at 55.degree. C. A positively hybridizing phage (HS-21 a) was identified and plaque purified by two additional rounds of plating and hybridization. The insert was subcloned to the plasmid pBLUESCRIPT II SK- according to the manufacturer's protocol (STRATAGENE, La Jolla, Calif.). Upon sequence analysis using the SEQUENASE protocol (USBC, Cleveland, Ohio) it was determined that clone HS-21 a contained the complete open reading frame corresponding to the human homolog of the rat A3 adenosine receptor. The coding region of the human A3 adenosine receptor cDNA is 78% identical to the rat sequence at the nucleotide level and contains 265 bp and 517 bp of 5' and 3' untranslated sequence, respectively. The 1.7 kb fragment was excised using sites present in the multiple cloning site of pBLUESCRIPT II SK- (STRATAGENE, La Jolla, Calif.) and subcloned into Xhol/SacI digested pSVL (PHARMACIA, Piscataway, N.J.) for its expression in COS and CHO cells.

Mammalian cell expression:

COS7 cells (ATCC #1651-CRL) were grown in complete medium, Dulbecco's modified Eagles's medium, DMEM (GIBCO, Grand Island, N.Y.) containing 10% fetal bovine serum, 100 U/mL penicillin-streptomycin and 2 mM glutamine, in 5% CO.sub.2 at 37.degree. C. Transient transfection of COS7 cells was performed by the CaPO.sub.4 method (Graham, F. L. and Van Der Erb, A. J. (1973) Virology 52:456-567) using the Mammalian Transfection Kit (STRATAGENE). Plasmid DNA (15 .mu.g) was precipitated with 125 mM CaCl.sub.2 in BBS (N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid buffered saline) at room temperature for 30 minutes. The DNA precipitate was added to the COS7 cells and incubated for 18 h in 5% CO.sub.2 at 37.degree. C. The precipitate was removed and the cells were washed twice with serum free DMEM. Cells were incubated in complete medium in 5% CO.sub.2 at 37.degree. C. for 48 h prior to the binding assay.

Stable expression in CHO cells:

To establish stable cell lines, CHO cells were cotransfected with 20 .mu.g of pSVL containing the adenosine receptor cDNA and 1 .mu.g of pWLneo (STRATAGENE) containing the neomycin gene. Transfection was performed by the CaPO.sub.4 method. DNA was precipitated at room temperature for 30 minutes, added to the COS7 cells and incubated 18 h in 5% CO.sub.2 at 37.degree. C. The precipitate was removed and the cells were washed twice with serum free DMEM. Cells were incubated for 24 h in 5% CO.sub.2 at 37.degree. C., replated in 24-well dishes at a dilution of 1:10, and incubated an additional 24 h before adding selection medium, DMEM containing 10% fetal bovine serum, 100 U/mL penicillin-streptomycin, 2 mM glutamine and 1.0 mg/mL G418 (GIBCO). Transfected cells were incubated at 5% CO.sub.2, 37.degree. C. until viable colonies were visible, approximately 14-21 days. Colonies were selected and propagated. The cell clone with the highest number of human adenosine receptors was selected for subsequent application in the binding assay.

Binding assay:

Membranes were prepared from transiently transfected COS7 cells 48 h after transfection or from G418-selected stably transfected CHO cells. Cells were harvested in 1 mM EDTA in phosphate buffered saline and centrifuged at 2000.times.g for 10 minutes. The cell pellet was washed once with phosphate buffered saline. The cell pellet was resuspended in 2 mL of 5 mM Tris, pH 7.6/5 mM MgCl.sub.2. Membranes were prepared from the cells by freeze-thaw lysis in which the suspension was frozen in a dry ice/ethanol bath and thawed at 25.degree. C. twice. The suspension was homogenized after adding an additional 2 mL of 5 mM Tris, pH 7.6/5 mM MgCl.sub.2, in a glass dounce homogenizer with 20 strokes. The membranes were pelleted at 40,000.times.g at 4.degree. C. for 20 minutes. The membrane pellet was resuspended at a protein concentration of 1-2 mg/mL in binding assay buffer, 50 mM Tris, pH 7.6/10 mM MgCl.sub.2. Protein concentration was determined by the method of Bradford ((1976) Anal. Biochem. 72: 248-250). Before the binding assay was performed, the membranes were incubated with adenosine deaminase (BOEHRINGER MANNHEIM), 2 U/mL for 30 minutes at 37.degree. C. Saturation binding of [.sup.125 I]-N.sup.6 -aminobenzyladenosine (.sup.125 I-ABA) or [.sup.125 I]-N.sup.6 -2-(4-amino-3.iodophenyl)ethyladenosine adenosine (APNEA) was performed on membranes prepared from pSVLA3 transfected COS7 cells. Membranes (100 .mu.g) were incubated in the presence of 0.2 U/mL adenosine deaminase with increasing concentrations of .sup.125 I-ABA in the range of 0.1-30 Nm for 120 minutes at 25.degree. C. in a total volume of 500 .mu.L. The binding assay was terminated by rapid filtration and three washes with ice-cold 50 mM Tris, pH 7.6/10 mM MgCl.sub.2 on a Skatron cell harvester equipped with a receptor binding filtermat (SKATRON INSTRUMENTS, INC.). Non-specific binding was determined on non-transfected cells. Bound radioactivity was measured by scintillation counting in Ready Safe Scintillation Cocktail (BECKMAN).

In vitro transcription and oocyte expression:

The 1.3 kb XhoI-BamHI fragment of the pSVL expression construct (described in Example 2) containing the full length human A2a adenosine receptor coding sequence was ligated into SaII-SpeI digested pGEMA (Swanson, et al, (1990) Neuron 4:929-939). The resulting plasmid, pGEMA2, was linearized with NotI, forming a template for in vitro transcription with T7 RNA polymerase. The homologous adenosine receptor subtype cDNA in pBluescript SK- was used as a template for in vitro transcription by T3 polymerase after removal of most of the 5' untranslated region, with the exception of 20 bp, as a 0.3 kb SmaI fragment. The K.sup.+ channel cDNA, Kv3.2b was employed as a negative control in the cAMP accumulation assay. The generation of Kv3.2b RNA was described by Luneau, et al, ((1991) FEBS Letters 1:163-167). Linearized plasmid templates were used with the STRATAGENE mCAP kit according to the manufacturer's protocol, except that the SP6 RNA polymerase reaction was performed at 40.degree. C. Oocytes were harvested from mature female Xenopus laevis, treated with collagenase, and maintained at 18.degree. C. in ND96 medium (GIBCO) supplemented with 1 mM sodium pyruvate and 100 .mu.g/mL gentamycin. Fifty nanoliters (10 ng) of RNA diluted in H.sub.2 O was injected and oocytes were incubated at 18.degree. C. for 48 hours.

cAMP accumulation assay in oocytes:

Oocytes injected with either human adenosine receptor transcript or the Kv3.2b transcript were transferred to fresh medium supplemented with 1 mM of the phosphodiesterase inhibitor, Ro 20-1724 (RBI, Natick, Mass.) and 1 mg/mL bovine serum albumin incubated for 30 minutes and transferred to an identical medium with or without the agonist adenosine (10 mM) for an additional 30 minutes at room temperature. Groups of 5-10 oocytes were lysed by transfer to ND96/100 mM HCl/1 mM Ro 20-1724 in microfuge tubes, shaken, incubated at 95.degree. C. for 3 min, and centrifuged at 12000 g for 5 min. Supernatants were stored at -70.degree. C. before cAMP measurements. Cyclic AMP levels were determined by radioimmunoassay (RIANEN kit, DuPont/NEN) using the acetylation protocol. The adenosine receptor antagonist, 8-(p-sulfophenyl)theophylline (100 .mu.M) was utilized to inhibit the cAMP response induced by adenosine in oocytes expressing the adenosine receptors.

cAMP accumulation in stable CHO cell lines:

The changes in cAMP accumulation can alternatively be measured in stably transfected CHO cells expressing the human adenosine receptor subtypes. CHO cells are washed twice in phosphate buffered saline (PBS) and detached in 0.2% EDTA in PBS. The cells are pelleted at 800 rpm for 10 min and resuspended in KRH buffer (140 mM NaCl/5 mM KCl/2 mM CaCl.sub.2 /1.2 mM MgSO.sub.4 /1.2 mM KH.sub.2 PO.sub.4 /6 mM glucose/25 mM Hepes buffer, pH 7.4). The cells are washed once in KRH buffer and resuspended at 10.sup.7 cells/mL. The cell suspension (100 .mu.L) is mixed with 100 .mu.L of KRH buffer containing 200 .mu.M Ro 20-1724 and incubated at 37.degree. C. for 10 minutes. Adenosine (10 .mu.M) was added in 200 .mu.L KRH buffer containing 200 .mu.M Ro 20-1724 and incubated at 37.degree. C. for 20 minutes. After the incubation, 400 .mu.L of 0.5 mM NaOAc (pH 6.2) was added and the sample was boiled for 20 minutes. The supernatant was recovered by centrifugation for 15 minutes and stored at -70.degree. C. cAMP levels were determined by radioimmunoassay (RIANEN kit, DuPont/NEN) using the acetylation protocol. The effect of antagonists on cAMP accumulation are measured by preincubation for 20 minutes before adding adenosine. Results show highly sensitive measurement of cAMP accumulation in response to adenosine (ADO) in a stable CHO cell line expressing the human A2b receptor and the effect of the antagonist 1,3-diethyl-8-phenylxanthine (DPX).

While the foregoing specification teaches the principles of the present invention, with examples provided for the purpose of illustration, it will be understood that the practice of the invention encompasses all of the usual variations, adaptations, modifications, as come within the scope of the following claims and its equivalents.