Proc. Natl. Acad. Sci. USAVol. 90, pp. 5113-5117, June 1993Pharmacology

Expression cloning of an ATP receptor from mouseneuroblastoma cellsKEVIN D. LUSTIG, ANDREW K. SHIAU, ANTHONY J. BRAKE, AND DAVID JULIUS*Department of Pharmacology, Programs in Cell Biology and Neuroscience, University of California, San Francisco, CA 94143

Communicated by Stephen Heinemann, March 1, 1993

ABSTRACT Extracellular ATP activates cell-surface me-tabotropic and ionotropic nucleotide (P2) receptors in vascular,neural, connective, and immune tissues. These P2 receptorsmediate a wealth of physiological processes, including nitricoxide-dependent vasodilation of vascular smooth muscle andfast excitatory neurotransmission in sensory afferents. Al-though ATP is now recognized as a signaling molecule, thecellular and molecular mechanisms underlying its actions havebeen difficult to study due to the absence of selective P2 receptorantagonists and cloned receptor genes. Nonetheless, five mam-malian P2 receptor subtypes have been tentatively assignedbased solely on agonist specificity and signaling properties.Here we report the cloning of a mouse cDNA encoding a P2receptor that shares striking homology with several G protein-coupled peptide receptors. When expressed in Xenopus laevisoocytes, the cloned receptor resembles a metabotropic P2Ureceptor; activation by either ATP or UTP elicits the mobili-zation of intracellular calcium. mRNA encoding the P2U puri-nergic receptor is found in neural and nonneural tissues.

There is considerable evidence suggesting that ATP functionsas an extracellular signaling molecule in neural and nonneuralmammalian tissues (1, 2). In central and peripheral synapses,ATP mediates fast excitatory neurotransmission (3, 4). In theautonomic nervous system, ATP is a major purinergiccotransmitter that is often colocalized in secretory vesicleswith norepinephrine or acetylcholine (5, 6). In the vascularsystem, aggregating platelets secrete ATP and ADP, whichstimulate the release of nitric oxide and other vasodilatorsfrom the endothelium (7). In the immune system, ATPmodulates macrophage phagocytosis (8) and mast cell de-granulation (9). In the human airway epithelium, ATP stim-ulates transepithelial ion transport (10), an effect that mayunderlie the therapeutic effect of ATP and UTP in thetreatment of cystic fibrosis-related lung disease (11).

It has been postulated that these responses to extracellularATP are mediated by specific plasma membrane receptors,called P2 purinergic receptors (12, 13). Based on agonistselectivity and signaling properties, five subclasses of P2receptor have been tentatively defined: three subclasses ofreceptors (P2T, P2U, and P2y) that are believed to signalthrough G proteins, one subclass (P2X) that is believed to bea ligand-gated cation channel, and one subclass (P2z) that ispresent on mast cells, macrophages, and fibroblasts, butwhose signaling mechanism is less well understood (1, 14-16). G protein-coupled P2 receptors are found in numerouscultured cell lines, where they have been shown to activatesignal transduction systems that involve the breakdown ofmembrane phospholipids and the elevation of cytoplasmicfree Ca2+ (17, 18). Ionotropic P2X receptors carry Na+, K+,and Ca2+ currents and appear to be predominantly expressedin neural and neuromuscular tissues (16).

A more complete characterization of this putative family ofP2 purinergic receptors has been hampered by the completelack of specific P2 receptor antagonists, radioligands, andcloned receptor cDNAs. To circumvent this difficulty, wehave used a Xenopus laevis oocyte expression cloning strat-egy (19, 20) to isolate a cDNA encoding a functional P2Ureceptor from NG108-15 neuroblastoma x glioma hybridcells.t We have examined the agonist selectivity, signalingproperties, species origin, and tissue distribution of thecloned P2U receptor.

EXPERIMENTAL PROCEDURESExpression Cloning. NG108-15 cells were grown in mono-

layer culture (21), total cellular RNA was isolated by theguanidine thiocyanate method (22), and poly(A)+ RNA wasselected on oligo(dT)-cellulose. A directional cDNA library(2 x 106 recombinants) was constructed in pCCM6XL asdescribed (23). pCCM6XL is a derivative of pCDM6XL inwhich a 778-bp Taq I fragment containing a gene encodingchloramphenicol acetyltransferase was inserted into theBstBI site of the supF gene. Initially, the NG108-15 cDNAlibrary was subdivided into 10 pools of 2 x 105 clones.Templates for in vitro transcription were prepared by linear-izing plasmid DNA isolated from these pools with Not I.Complementary RNA (cRNA) transcripts were synthesizedusing SP6 RNA polymerase as described (20).X. laevis oocytes were surgically isolated (24) and enzy-

matically defolliculated by incubation with 2 mg of collagen-ase per ml for 2 hr at room temperature. Defolliculatedoocytes were washed five times with modified Barth's solu-tion [MBS1: 7.5 mM Tris, pH 7.6/88 mM NaCl/1 mMKCI/2.4 mM NaHCO3/8.2 mM MgSO4/0.33 mM Ca(NO3)2/0.4 mM CaCl2/100 units of penicillin per ml/100 tg ofstreptomycin per ml/2% Ficoll-400] and maintained in MBS1at 18°C. On the following day, oocytes were injected with 50nl of NG108 cRNA transcripts (=1 ug/,ul) and incubated at18°C for 2 days prior to analysis. Voltage-clamp recordingwas performed using a single electrode (Axoclamp 2A) indSEVC mode. cRNA from 1 of the 10 pools rendered theoocytes responsive to ATP and UTP (not shown). A sibselection procedure was used to progressively subdivide thispositive pool into smaller pools of 20,000, 2000, 200, and 10clones, finally yielding a single clone, called pP2R. Bothstrands of the cDNA insert were sequenced using the dide-oxynucleotide chain-termination method (25).

45Ca2+ Release Assay. Defolliculated oocytes were micro-injected with 0.5 ng ofpP2R cRNA transcripts and incubatedat 18°C. After 48 hr, the oocytes were washed four times with5 ml of MBS2 (Ca2+-free MBS1 containing 0.1% bovine

Abbreviations: cRNA, complementary RNA; ATP[,yS], adenosine5'-[y-thio]triphosphate; 2-MeSATP, 2-methylthioadenosine 5'-triphosphate; AMP-PCP, adenosine 5'-[f3,'y.methylene]triphosphate;AMP-CPP, adenosine 5'-[a,4-methylene]triphosphate.*To whom reprint requests should be addressed.tThe sequence reported in this paper has been deposited in theGenBank data base (accession no. L14751).

5113

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 90 (1993)

serum albumin) and incubated for 3 hr at room temperaturein 0.5 ml of MBS2 containing 30 ,uCi of 45CaC12 (1 Ci = 37GBq). Labeled oocytes were washed 10 times with 5 ml ofMBS1 and incubated in 3 ml ofMBS1 for 90 min. Each oocytewas then separately placed into 100 ,ul of MBS1 in a well ofa 96-well polystyrene plate. After 20 min at room tempera-ture, individual oocytes were transferred to 100 ,ul of freshMBS1 containing the indicated concentration ofnucleotide ornucleoside. The medium was sampled 20 min later and theradioactivity in the sample was quantitated by liquid scintil-lation counting. EC50 values (concentrations of agonists thatresult in half-maximal stimulation) were determined by non-linear regression analysis.Northern Analysis. Poly(A)+ RNA was isolated from cell

lines or mouse tissues and electrophoresed in 0.8% agarose/formaldehyde gels (26). The mRNA was transferred to Hy-bond-N nylon membranes (Amersham), prehybridized for 6hr, and then hybridized overnight at 42°C in solution con-taining 50% formamide with 6x SSC (lx SSC = 0.15 MNaCl/15 mM sodium citrate). The probe was prepared byrandom prime-labeling (Boehringer Mannheim) of a 2.2-kbXba I DNA fragment containing the entire coding region ofP2R. Filters were washed in O.lx SSC containing 0.1% SDSat 65°C and exposed to KodakXAR film with two intensifyingscreens at -80°C for 2 days for cell lines and 4 days fortissues.

RESULTSIdentification of an ATP Receptor cDNA. Using a X. Iaevis

oocyte expression cloning strategy (19, 20), we isolated acDNA clone encoding a P2 receptor (P2R) from NG108-15cells, a mouse N18TG2 neuroblastoma x rat C6 glioma cellhybrid cell line. In Xenopus oocytes injected with poly(A)+

ADP 2-MeSATP ATP

AMP-PCP AMP-CPP UTP

ADO ATPyS

Vf < 200 nA

lOs

FIG. 1. Nucleotides elicit inward currents in Xenopus oocytesinjected with 0.5 ng of cRNA transcribed in vitro from pP2R. Theoocytes were single-electrode voltage clamped 48 hr after cRNAinjection and held at -60 mV while superfused with MBS1. Threeoriginal current traces are shown, and the thick lines above the tracesindicate the length of time the oocytes were incubated with thefollowing compounds (50 uM): ATP, UTP, adenosine 5'-[y-thio]triphosphate (ATP[yS]), 2-methylthioadenosine 5'-triphosphate(2-MeSATP), adenosine 5'-[,8,-methylene]triphosphate (AMP-PCP), ADP, adenosine 5'-[a,4-methylene]triphosphate (AMP-CPP),and adenosine (ADO).

RNA from NG108-15 cells, bath-applied ATP orUTP (1 mM)evoked inward currents (between 20 and 200 nA) that aretypical of the activation of phospholipase C-coupled recep-tors (not shown). To identify a cDNA encoding the P2R,pools of 2 x 105 individual clones from a NG108-15 cDNAlibrary were transcribed in vitro and the resultant cRNA wasinjected into oocytes. The pool that rendered the oocytesresponsive to ATP or UTP was progressively subdivided toobtain a single plasmid clone called pP2R (Fig. 1).

Predicted Structure of P2R. Sequence analysis revealedthat the cloned P2R cDNA has a 1119-bp open reading frame(Fig. 2), 269 bp of 5' untranslated sequence, and -1 kb of 3'untranslated sequence. The largest open reading frame en-codes a putative protein of 373 amino acids with a predictedrelative molecular mass of 42,172. Analysis of the deducedamino acid sequence suggests that the protein contains asmall N-terminal extracellular domain, seven hydrophobictransmembrane domains, and a large C-terminal intracellulardomain. There are two consensus N-linked glycosylationsites near the N terminus and several potential phosphory-lation sites near the C terminus (Fig. 2). The only site thatresembles a consensus nucleotide-binding motif [G-(X)4-G-K, ref. 27] is comprised of residues 18-24 (GDELGYK)located in the putative N-terminal extracellular domain.Sequence comparisons (Fig. 3) indicate that the cloned P2R

is a member of the G protein-coupled receptor superfamily.The predicted amino acid sequence of the cloned P2R con-tains a number of highly conserved amino acids (Asn-51,Asp-79, Leu-81, Arg-131, Pro-167, and Pro-303) that areATGGCAGCAGACCTGGAACCCTGGAATAGCACCATCAATGGCACCTGGGAGGGGGACGAA 329M A A D L E P W * S T I g G T W E G D E 20

CTGGGATACAAGTGTCGCTTCAACGAGGACTTCAAGTACGTGCTGTTGCCCGTGTCCTAT 389L G Y K C R F N E D F K Y V L L P V S Y 40

GGCGTGGTGTGCGTGCTCGGGTTGTGCCTGAACGTCGTGGCTCTCTATATCTTCCTATGC 449G V V C V L G L C L N V V A L Y I F L C 60

CGCCTCAAAACCTGGAACGCCTCCACCACCTACATGTTTCACCTGGCAGTTTCGGACTCTR L K T W N A S T T Y M F H L A V S D S

CTCTACGCAGCGTCCCTGCCGCTGTTGGTTTATTACTACGCCCGGGGTGACCACTGGCCAL Y A AS L P L L V Y Y Y A R G D H W P

TTTAGCACGGTGCTCTGCAAGCTGGTGCGTTTCCTCTTCTACACCAACCTCTACTGCAGCF S T V L C K L V R F L F Y T N L Y C S

ATCCTCTTCCTCACCTGCATCAGCGTGCACCGGTGCCTGGGAGTCCTGCGCCCTCTGCACI L F L T C T S V H R C L G V L R P L H

TCCCTGCGTTGGGGCCGGGCCCGTTATGCCCGCCGGGTGGCTGCGGTTGTGTGGGTGCTGS L R W G R A R Y A R R V A A V V W V L

GTGCTGGCCTGCCAGGCACCCGTGCTCTACTTCGTCACCACCAGCGTGCGGGGAACCCGGV L A C O A P V L Y F V T T S V R G T R

ATCACTTGCCATGACACCTCGGCCCGAGAGCTCTTTAGCCATTTTGTGGCTTACAGCTCCI T C H D T S A R E L F S H F V A Y S S

GTCATGCTGGGTCTGCTTTTTGCTGTGCCCTTTTCCGTAATCCTGGTCTGTTACGTGCTTV M L G L L F A V P F S V I L V C Y V L

50980

569100

629120

689140

749160

809180

869200

929220

ATGGCCAGGCGGCTGCTCAAACCGGCTTATGGGACCACAGGAGGTCTGCCTCGGGCCAAG 989M A R R L L K P A Y G T T G G L P R A K 240

CGCAAGTCTGTGCGCACCATTGCCTTGGTACTGGCCGTCTTCGCCCTCTGCTTTCTGCCT 1049R K S V R T I A L V L A V F A L C F L P 260

TTCCACGTCACGCGCACCCTCTACTACTCCTTCCGATCACTTGACCTCAGCTGCCACACC 1109F H V T R T L Y Y S F R S L D L S C H T 280

1169L N A I N M A Y K I T R P L A S A N S C 300

CTTGACCCGGTACTCTACTTCCTGGCAGGGCAGAGACTTGTCCGCTTTGCCCGAGATGCC 1229L D P V L Y F L A G Q R L V R F A R D A 320

AAGCCACCCACGGAGCCTACCCCCAGCCCACAGGCTCGTCGCAAGCTGGGCCTGCACAGG 1289K P P T E P T P S P Q A R R K L G L H R 340

CCTAACAGAACTGTGAGGAAAGATTTGTCAGTCAGCAGTGACGACTCAAGACGGACAGAG 1349P N R T V R K D L S V S S D D S R R T E 360

TCCACACCAGCTGGAAGTGAGACTAAGGACATTCGGCTATAG 1391S T P A G S E T K D I R L - 373

FIG. 2. The nucleotide sequence of the largest open readingframe of the cloned P2R is shown above the deduced amino acidsequence. A putative nucleotide-binding region in the N-terminaldomain is indicated by a black bar, two potential N-linked glycosy-lation sites are indicated by asterisks, and the seven putativetransmembrane domains are underlined.

CTCAACGCCATCAACATGGCATATAAGATCACCCGGCCGCTGGCCAGCGCCAACAGTTGT

5114 Pharmacology: Lustig et al.

Pharmacology: Lustig et al. Proc. Natl. Acad. Sci. USA 90 (1993) 5115

P2 M Ar D. E W - - - - - - -r STHROMBIN MGPRRLLLVAACFSLCGPLLSARTRARRPESKATN TN D RS F L L R N PUND

P2 TINGT|i GD LGYKC ---------------- - - - - - - -0FR ETHROMBIN KY E P F DEW KNESGLTEYRLVSINKSSPLQKOLPAFISEDA S GY L T S SWPAF MELNSSS - - - - - - - - - - - - - - - - EV D S EHANGII M I L N SST - - - EDGIKRIQDDCPKAGRHNYGPRN1 MDLHLFDYAEPGNFSD SWPCNSSD-----C VVDTVMCPNMPNKSVIL-8 MEVNVWNMTOLWTWFEDEFANATGMPPVEKDYSPCLVVTQTLjNjKYV

P2 K tL V S GVfOVt V lIP--ORLf8VAASTT TiFHLAVSTHROMBIN L T L FV SV 1TVTF VS P I4I VVj L K V K K PA VVA} -LLATADPAF RA0T8 FkII V S F ANGYWVW: AR YPt K LE E K IF V AANGII F M T L S F VV IF NS L V VI V I- YFY K L KTVASV F L LN :L ALA DGPRNI L L YTdS'FII:Y I F I FLILM4] L[VMVWV - - IQAK GY D THOC'- .LN,jJAI A DIL-8 VVQ SA LL A L L S L S L V M L V -L NY S R LMNRSDVTDVLL LN -A MA 0

II IIIP2 SRY9ASIFfLEI YYARDHWPFS'TVLK 17T1~~~THROMBIN V F V S V PK SY FSSD GS E.A T AF Y LLPAF L F L TILPTJW VY-SNGGNN P KFLL GCAOFI S ITY

IL-B L..JL...I N4dIWAVS- ~~~~~~~~K E K~~~ ~ AFLFTNLV0$.......L.....VANG II LOF L LTWP2JWAV\ M

TAMEYAWPWMGN4 -A S A SVYSFINLAiTISIFALTIVGPRN1 LWVVLVL T PIW9VSLVQHNW MGELTCKVT H L IVKEVs NFSLL:LAOIL-8 L[gF[RL T 1WAVS KE K G ILSffP> }<|ES L VK E V S L A

IV

P2 H CL GL PLHL G(faA R 9RAA.LVL. Q AP-VYF VTTSRTHROMBIN DFY:A:: V P|MQ$WS-R T L G ASF T C ALA IGAVV PL E VPPAF N|F OALK Y|PIIKTAQATT R K R G A L S L0V I: V EAA S Y F L S T N V V S NANGII D YLAIVHMK[gR MLTML VKV T C II L- A GL A S LFT I H RN V F F EGPRN1 LYIWS T YFTNT P S S RKK MV VE|C L L-H - AFJl V S LlPID TYL KQ|V T S AIL8 D:YLSITYFTNTPSSTKKMHVAVFICLGI LAFSLVSLPIDTS0RVFS

v

P2 GTRI-T .......C.H:D.B|TSAAERF-dHFOAYSSVMLGWL(TT......................................... *.-V---P-T-SVT3-VL A AL....

THROMBIN L NTT CHDV L N E TLEGYYA F S A V F L S T CYVS CLPAF K A G S G NI T R C F E H Y E K G00K PL II H C VIF V FL LI LFCL LV H TIL

ANGII NONET VA F H Y E S N S T -L PUG L G L T K N I4FLf JLIITSfT W K TLGPRN1 S N N E T R S FYPEHS KEWL GMEL92SVVIFAVQ IHELJYIF LLAIIL-8 NNSSPVCYEDLGHNTAK-WRMVLRI LPHT F GTGFTLbTg

VI

P2 E3PlGTT GG LP R A K S V[qTB | .FAL FFX t T R Tld - - Y ::F R STHROMBIN S S SWA N - - - - R S KSR A L F LSWAIVF. IlI L L ITA-L - HL:!j-PAF M0 P V K a - Q R N A E V R R A L W M V C T|V L A V Fl V OFV HjPV QL P W T L A E L G M

ANGII K E.E - O K N K P UD D F K ITLA V L FFFSW VPHQ I FTFMDVL IQ L G LGPRN1 S A - S SD 0O- E K HSSR- - - - - K IF S Y V V F L V: A V LQ|D F S L H YIL-8 FQ-HMG - QKHRAM- [-VI.!|FA4V L I LYYN LV L L A DT L MRT HV

VII

P2THROMBINPAFANG IIGPRN1IL-8

W.SHTSTLI -EAAVFAY.LLCVCYVSSS 1. .........-.Y ASSECYVYSILWP--SSN ADHL K L L$TN D IFLTKKFRHLSEK L N I

I~~~~~~~~~~~~I N-k NLMPTICX^ F Y~| F L G K K F K K YflL 0 L L5YIPFTARL E H A L F TA L H V TOO CL S L V HO V NBIP---t-I YSF IN RN YIRL I Y VF F

GET GANDR1§IC .IDAA8 L DA -T- E I LGFLHSOLNI1FVF1!A TtI GQN F R NKGHL KML A A

1 050

311 001 3264246

791 4863749090

1 2 919811 31241401 38

1 772471 631 731 891 85

22529721 3222239234

273339262270282277

32238731031 9332327

FIG. 3. Sequence comparisons of P2R and G protein-coupled receptors. The deduced amino acid sequence of P2R (P2) was aligned to thatof the human thrombin receptor precursor (425 amino acids, ref. 28), guinea pig platelet-activating factor receptor (PAF; 342 amino acids, ref.29), bovine angiotensin II type I receptor (ANG II; 359 amino acids, ref. 30), GPRN1, a putative human vasoactive intestinal peptide receptor(362 amino acids, ref. 31), and rabbit interleukin 8 receptor (IL-8; 355 amino acids, ref. 32). The boxed and shaded amino acids indicate identicalresidues. Roman numerals and brackets denote the seven putative transmembrane domains.

believed to be important for the function ofG protein-coupledreceptors (33). The cloned receptor is most similar to recep-tors for thrombin (25% identity), platelet-activating factor(25% identity), angiotensin II (22% identity), interleukin 8(23% identity), and GPRN1, a putative vasoactive intestinalpeptide receptor (21% identity). The cloned receptor issubstantially less similar (<12% identity) to G protein-coupled receptors for adenosine and cAMP.

Functional Characterization of the Cloned P2R. Functionalcharacterization of the cloned receptor using transfectedmammalian cells was hindered by the presence of endoge-nous P2 receptors and ubiquitous nucleotide-binding proteins

and the absence of selective high-affinity radioligands. Wetherefore examined the pharmacology of the cloned receptorin Xenopus oocytes injected with pP2R cRNA transcripts. Involtage-clamped oocytes expressing the cloned P2R, inwardcurrents (,s500 nA) were elicited by bath application of50 AuMATP, UTP, or ATP[yS] but not 50 AM 2-MeSATP, AMP-PCP, ADP, AMP-CPP, or adenosine (Fig. 1). Agonist stim-ulation of phospholipase C-coupled receptors in oocytesactivates endogenous calcium-dependent chloride channelsthat carry inward currents (34). Thus, ATP also elicited anincrease in the rate of 45Ca2+ efflux (Fig. 4), indicative of anincrease in intracellular free Ca2+ (35). The order of agonist

Proc. Natl. Acad. Sci. USA 90 (1993)

a)U)

C:0cna)01)

.,,

x

0\0

11

-8 -7 -6 -5 -4 -3[Nucleotide] (log M)

FIG. 4. Characterization of the agonist specificity of the clonedP2R using a 45Ca2+ release assay. Xenopus oocytes expressing theP2R were labeled with 45Ca2+ and then incubated for 20 min with theindicated concentration of ATP (o), UTP (o), ATP[yS] (m), 2-Me-SATP (n), AMP-PCP (A), ADP (O), AMP-CPP (v), or adenosine (*).The % maximal response was determined for each experiment bydividing the cpm in each sample by the cpm in samples taken fromoocytes treated with a maximally effective concentration of ATP.The data are the mean ±SEM of three separate experiments.

potency for 45Ca2+ efflux was ATP (EC5o = 0.7 ,uM) UTP(EC50 = 1.1 ,JM) > ATP[yS] (EC50 = 7.9 ,uM) >> 2-MeSATPAMP-PCP ADP AMP-CPP> adenosine (Fig. 4). In

uninjected or water-injected oocytes, none ofthe compoundsreproducibly evoked inward currents or enhanced the rate of45Ca2+ efflux (not shown). In these control oocytes, ATP andUTP occasionally induced inward currents (typically <20nA), which may be due to the activation of a P2 receptorendogenous to the oocytes. Of the subtypes of P2 receptorsthat have been tentatively assigned, this agonist specificity ismost similar to that of the P2U (nucleotide) receptor subtype,for which ATP and UTP are more potent agonists than2-MeSATP (15). In NG108-15 cells, ATP and UTP activate anendogenous P2u receptor that mobilizes intracellular Ca2+(36, 37).ATP hydrolysis does not appear to be necessary for

receptor activation since the cloned P2R (Fig. 1) and theendogenous P2 receptor in NG108-15 cells (37) are activatedby ATP[yS], a slowly-hydrolyzed ATP analog. A search forATP-binding motifs in the P2R revealed one site in theputative N-terminal extracellular domain that resembles aconsensus ATP-binding site. To determine whether this siteis essential for receptor activation by ATP, we constructed amutant receptor in which residues 18-24 (GDELGYK) havebeen deleted. Voltage-clamp analysis showed that oocytesexpressing this mutant receptor exhibited similar responsesto ATP and UTP as did oocytes expressing the wild-typereceptor (not shown).

Species Type and Tissue Distribution of the Cloned P2R.Polymerase chain reaction (PCR) and Northern blot analysesrevealed that the receptor is derived from the mouse N18TG2neuroblastoma and not the rat C6 glioma parent cell line. Twooligonucleotide primers corresponding to positions 277-306and 646-675 of the P2R cDNA (Fig. 2) were used to amplifytotal mouse or rat genomic DNA by standard PCR. A PCRproduct corresponding to an =0.4-kb region of the P2 recep-tor could be amplified from pP2R, rat genomic or mousegenomic DNA templates. Amplification at higher stringency,however, yielded an =0.4-kb PCR product only with pP2Rand mouse genomic DNA (not shown). When PCR productsof the lower stringency reactions were digested with Ava I,BstEII, Eae I, Pst I, or Sma I and resolved on a 1.5% agarosegel, the restriction pattern of the mouse PCR product was

identical to that of the pP2R PCR product, whereas therestriction map of the rat PCR product differed at one Ava Iand one Sma I site (not shown). In addition, on Northernblots, an -2.4-kb RNA species was detected in NG108-15cells and mouse N18TG2 cells, whereas an -3.0-kb RNAspecies was detected in rat C6 glioma cells (Fig. 5). Takentogether, these observations demonstrate that the cloned P2RcDNA is of murine origin.Northern blot analysis revealed that mRNA encoding the

P2R is widely distributed in mouse tissues. An %2.4-kb RNAspecies was detected in spleen, testes, kidney, liver, lung,heart, and brain (Fig. 5). An -2.4-kb RNA species was alsodetected in mouse NlE-115 neuroblastoma cells.

DISCUSSIONWe have isolated a cDNA clone encoding the first member ofwhat, to our knowledge, is likely to be a new gene family ofnucleotide receptors. Our results now provide moleculargenetic evidence for the existence of plasma membranereceptors for ATP. The cloned P2R has predicted structuralfeatures that are characteristic of most known G protein-coupled receptors: an N-terminal extracellular domain withtwo consensus sites for N-linked glycosylation, seven hy-drophobic putative transmembrane a-helices, and a largeC-terminal intracellular domain containing several consensusphosphorylation sites. A surprising finding is that the clonedP2R is considerably more homologous to G protein-coupledpeptide receptors than to receptors for the structurally re-lated ligands adenosine and cAMP.

Five mammalian P2 receptor subtypes have been tenta-tively assigned based on agonist selectivity and signalingproperties (2, 15). The characteristics of the cloned P2R aremost similar to those of the P2u receptor subtype (15). Thecloned P2R and the endogenous P2U receptor expressed inNG108-15 cells (37) are activated by ATP or UTP and bothcouple to a signal transduction system involving increases incytoplasmic free Ca2+. Our findings demonstrate that thepharmacological profile of the P2U site can be attributed to asingle gene product. This direct relationship has provendifficult to establish due to the lack of subtype-selectivepharmacological or molecular probes.

P2 receptors have been implicated in the regulation ofneurotransmission, vascular tone, wound healing, inflamma-tion, muscle contraction, immune responses, pulmonaryfunction, and cell growth (1, 2, 7, 16). Ca2+-linked P2 recep-tors that respond to ATP or UTP are expressed in cell lines

1 2 3 4 5 6 7 8 9 10k b

7.5-

4.4-

2.4 Jl* :.

1.4-

1I

k b-7.5

-4.4

-2.4

-1.4

FIG. 5. Determination of the tissue distribution ofP2R mRNA byNorthern analysis. Shown is an autoradiograph of a blot probed withrandom primer labeled DNA synthesized from an Xba I fragment ofP2R. Northern blots were prepared using poly(A)+ mRNA isolatedfrom the following cells and mouse tissues: mouse NlE-115 neuro-blastoma cells (5 ,g, lane 1), mouse N18TG2 neuronal cells (5 Ag,lane 2), rat C6 glioma cells (5 ,ug, lane 3), mouse neuroblastoma x ratglioma NG108-15 cells (5 ,g, lane 4), spleen (0.7 ,ug, lane 5), testes(2 .g, lane 6), kidney (1 ,ug, lane 7), liver (2 j.g, lane 8), lung (2 ,g,lane 9), heart (2 ,ug, lane 10), and brain (10 ,g, lane 11).

5116 Pharmacology: Lustig et al.

Proc. Natl. Acad. Sci. USA 90 (1993) 5117

derived from many different sources, including neural, mus-cular, vascular, connective, and epithelial tissues. Consistentwith the idea that we have cloned a ubiquitous P2 receptor,P2R mRNA is expressed in all mouse tissues that we haveanalyzed. Presumably, P2R-mediated changes in intracellu-lar free Ca2+ or other cytoplasmic second messenger systemspromote tissue- or cell-type-specific physiological responses.ATP has been demonstrated to act as a rapid excitatory

neurotransmitter in the nervous system and as a hormone inmany nonneural tissues (1, 2). Cloning of this P2 receptorprovides a crucial reagent for examining the structure, func-tion, and expression of members of the ATP receptor family.

We thank Nila Patil, Andrew Peterson, and Svetlana Shtrom foradvice about cDNA library construction and RNA isolation, LaurieErb and Gary Weisman for discussion about P2 receptors, and HenryBourne for comments on the manuscript. This work was supportedby the National Institutes of Health, by a National Science Foun-dation Presidential Young Investigator Award (D.J.), and by apredoctoral fellowship from the Howard Hughes Medical Institute(A.K.S.). D.J. is a Fellow of the Pew Memorial Trust and theMcKnight Foundation for Neuroscience.

1. Gordon, J. L. (1986) Biochem. J. 233, 309-319.2. Dubyak, G. R. & Fedan, J. S., eds. (1991) Ann. N.Y. Acad.

Sci. 603.3. Edwards, F. A., Gibb, A. J. & Colquhoun, D. (1992) Nature

(London) 359, 144-147.4. Evans, R. J., Derkach, V. & Surprenant, A. (1992) Nature

(London) 357, 503-505.5. Von Kugelgen, I. & Starke, K. (1991) Trends Pharmacol. Sci.

12, 319-324.6. Westfall, D. P., Sedaa, K. O., Shinozuka, K., Bjur, R. A. &

Buxton, I. (1990) Ann. N. Y. Acad. Sci. 603, 300-310.7. Boeynaems, J. M. & Pearson, J. D. (1990) Trends Pharmacol.

Sci. 11, 34-37.8. Steinberg, T. H., Buisman, H. P., Greenberg, S., Di Virgilio,

F. & Silverstein, S. C. (1990) Ann. N.Y. Acad. Sci. 603,120-129.

9. Osipchuk, Y. & Cahalan, M. (1992) Nature (London) 359,241-244.

10. Mason, S. J., Paradiso, A. M. & Boucher, R. C. (1991) Br. J.Pharmacol. 103, 1649-1656.

11. Knowles, M. R., Clarke, L. L. & Boucher, R. C. (1991) N.Engl. J. Med. 325, 533-538.

12. Burnstock, G. (1978) in Cell Membranes Receptors for Drugsand Hormones, eds. Straub, R. W. & Bolis, L. (Raven, NewYork), pp. 107-118.

13. Burnstock, G. & Kennedy, C. (1985) Gen. Pharmacol. 167,433-440.

14. Gordon, J. L. (1990) Ann. N. Y. Acad. Sci. 603, 46-52.15. O'Connor, S. E., Dainty, I. A. & Leff, P. (1991) Trends Phar-

macol. Sci. 12, 137-141.16. Bean, B. (1992) Trends Pharmacol. Sci. 13, 87-90.17. Dubyak, G. R. (1991) Am. J. Respir. Cell Mol. Biol. 4,295-300.18. Irving, H. & Exton, J. H. (1987) J. Biol. Chem. 262, 3440-3443.19. Masu, Y., Nakayama, K., Tamaki, H., Harada, Y., Kuno, M.

& Nakanishi, S. (1987) Nature (London) 329, 836-838.20. Julius, D., MacDermott, A. D., Axel, R. & Jesseli, T. M. (1988)

Science 241, 558-564.21. Klee, W. A. & Nirenberg, M. (1974) Proc. Natl. Acad. Sci.

USA 71, 3474-3477.22. Cathala, G., Savouret, J. F., Mendez, B., West, B. L., Karin,

M., Martial, J. A. & Baxter, J. D. (1983) DNA 2, 329-335.23. Maricq, A. V., Peterson, A. S., Brake, A. J., Myers, R. M. &

Julius, D. (1991) Science 254, 432-437.24. Marcus-Sekura, C. J. & Hitchcock, M. J. M. (1987) Methods

Enzymol. 152, 284-288.25. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl.

Acad. Sci. USA 74, 5463-5467.26. Sambrook, J., Fritsch, E. F. & Maniatis, T., eds. (1989)

Molecular Cloning:A Laboratory Manual (Cold Spring HarborLab., Plainview, NY), 2nd Ed.

27. Saraste, M., Sibbald, P. R. & Wittinghofer, A. (1990) TrendsBiochem. Sci. 15, 430-434.

28. Vu, T. K., Hung, D. T., Wheaton, V. I. & Coughlin, S. R.(1991) Cell 64, 1057-1068.

29. Honda, Z., Nakamura, M., Miki, I., Minami, M., Watanabe,T., Seyama, Y., Okado, H., Toh, H., Ito, K., Miyamoto, T. &Shimizu, T. (1991) Nature (London) 349, 342-346.

30. Sasaki, K., Yamano, Y., Bardhan, S., Iwai, N., Murray, J. J.,Hasegawa, M., Matsuda, Y. & Inagami, T. (1991) Nature(London) 351, 230-233.

31. Sreedharan, S. P., Robichon, A., Peterson, K. E. & Goetzl,E. J. (1991) Proc. Natl. Acad. Sci. USA 88, 4986-4990.

32. Lee, J., Kuang, W. J., Rice, G. C. & Wood, W. I. (1992) J.Immunol. 148, 1261-1264.

33. Savarese, T. M. & Fraser, C. M. (1992) Biochem. J. 283, 1-19.34. Snutch, T. (1988) Trends Neurosci. 11, 250-256.35. Williams, J. A., McChesney, D. J., Calayag, M. C., Lingappa,

V. R. & Logsdon, C. D. (1988) Proc. Natl. Acad. Sci. USA 85,4939-4943.

36. Ehrlich, Y. H., Snider, R. M., Kornecki, E., Garfield, M. G. &Lenox, R. H. (1988) J. Neurochem. 50, 295-301.

37. Lin, T. A., Lustig, K. D., Sportiello, M. G., Weisman, G. A.& Sun, G. Y. (1993) J. Neurochem. 60, 1115-1125.

Pharmacology: Lustig et al.