Cellular and Molecular Neurobiology, 11ol. 11, No. 5, 1991

Dopamine Receptors in the Human Retina: Cloning of cDNA and Localization of mRNA D2

Allen D e a r l y , 1'2'6 Pierre Falardeau, 1 Carol Shores, s and Marc G. Caron la'4 Received June 19, 1991; accepted November 27, 1991 KEY WORDS: dopamine; D 2 receptor; retina; cloning; in situ hybridization.

SUMMARY

1. We have obtained a cDNA clone encoding a human retinal D 2 dopamine receptor. 2. The longest open reading frame (1242 bp) of this clone encodes a protein of 414 amino acids having a predicted molecular weight of 47,000 and a transmembrane topology similar to that of other G protein-coupled receptors. 3. Transient transfection of COS-7 cells with an expression vector containing the clone resulted in expression of a protein possessing a pharmacological profile similar to that of the O 2 dopamine receptor found in striatum and retina. 4. Northern blot analysis indicated that, in rat brain and retina, the mRNA for this receptor was 2.9 kb in size. 5. In situ hybridization was performed to examine the distribution of the mRNA for this receptor in human retina. Specific hybridization was detected in both the inner and the outer nuclear layers. 6. These findings are consistent with prior physiological and autoradiographic studies describing the localization of D2 dopamine receptors in Department of Cell Biology, Duke University Medical Center, Durham, North Carolina 27710. 2 Department of Ophthalmology, Duke University Medical Center, Durham, North Carolina 27710. 3 Department of Medicine, Duke University Medical Center, Durham, North Carolina 27710. 4 Howard Hughes Medical Institute Laboratories, Duke University Medical Center, Durham, North Carolina 27710. 5 Department of Chemotherapy, Glaxo, Research Triangle Park, North Carolina 27709. 6To whom correspondence should be addressed at National Institutes of Health, Department of Health and Human Services, 5333 Westbard Ave., Rm. 319B, Bethesda, Maryland 20892. 437 0272-4340/91/1000-0437506.50/0(~) 1991PlenumPublishingCorporation

438

Dearry, Falardeau, Shores, and Caron

vertebrate retinas. Our observations suggest that photoreceptors as well as cells in the inner nuclear layer of human retinas may express the mRNA for this D2 dopamine receptor. INTRODUCTION Dopamine is the predominant catecholamine of vertebrate retinas (see Ehinger, 1983), and it has been postulated to mediate a variety of cellular responses to light onset or offset. I n several species, dopamine may participate in controlling electronic coupling between horizontal cells, the synthesis of melatonin, and retinomotor movements of photoreceptors and retinal pigment epithelium (RPE) cells (see Dowling, 1986; Besharse et al., 1988; Dearry and Burnside, 1988). In mammalian retinas, dopamine may regulate coupling between amacrine and cone bipolar cells (Vaney, 1985) and influence the responses of ganglion cells (Jensen and Daw, 1986; Robbins et al., 1988). In addition, dopamine may serve as an informational messenger between the retina and the RPE. For example, release of dopamine from the retina has been proposed to have a role in modulating electrical responses (Dawis and Niemeyer, 1986; Sato et al., 1987) and retinomotor movements (Dearry and Burnside, 1989; Dearry et al., 1990a) of RPE cells as well as phagocytosis of shed photoreceptor disks (Reme et al., 1986). These results are consistent with the idea that dopamine can act as a paracrine messenger by diffusing from its site of release in the inner retina to reach distant nonsynaptic receptors on RPE cells. Thus, dopamine may serve as both a synaptic transmitter and a diffusible neuromodulator within vertebrate retinas. The physiological effects of dopamine are mediated by its interaction with two basic types of receptor that have been classified as D1 and D2 (Kebabian and Calne, 1979). These receptors are coupled to stimulation (D1) and inhibition (D2) of adenylyl cyclase and possibly to other signaling mechanisms (see Andersen et al., 1990). Previous pharmacological and biochemical studies have demonstrated the presence of each of these receptor types in vertebrate retinas (Gredal et al., 1987; McGonigle et al., 1988; Qu et al., 1989). Furthermore, stimulation of each receptor type has been suggested to have specific functional effects. Activation of D1 receptors may modulate the coupling of horizontal cells in teleost fish (Teranishi et al., 1984; Lasater and Dowling, 1985), the activity of aromatic L-amino acid decarboxylase in rats (Rossetti et al., 1990), and the release of acetylcholine in rabbits (Hensler et al., 1987); D2 receptors, retinomotor movements in fish and X e n o p u s (Dearry and Burnside, 1985, 1986; Pierce and Besharse, 1985), melatonin synthesis in X e n o p u s and chick (Iuvone et al., 1987; Zawilska and Iuvone, 1989), and dopamine release in rabbits (Dubocovich, 1984). In some instances, the receptor type mediating these retinal responses may be species dependent. For example, both D1 and D2 receptors contribute to modulating the electrical response of horizontal cells to dopamine in X e n o p u s and turtle (Witkovsky et al., 1989; Piccolino et al., 1989), and D1 receptors modulate the retinomotor response of RPE cells to dopamine in bullfrogs (Dearry et aL, 1990a). Together, these results suggest that stimulation of retinal D~ or D2 dopamine receptors has a role in regulating a variety of physiological functions.

Retinal D2 DoparnineReceptors

439

As a first step in studying more closely the distribution, action, and regulation of these receptors in vertebrate retinas, we have cloned a cDNA or gene encoding each receptor type. In this paper, we describe the cloning and expression of a De dopamine receptor and localization of its mRNA in human retina. A separate report details the cloning of a D 1 dopamine receptor (Dearry et al., 1990b). These results will facilitate investigating the functional roles of these receptors at a molecular level.

MATERIALS A N D METHODS

Library Screening. A human retina cDNA library in ~,gtl0 (provided by Jeremy Nathans, Johns Hopkins University) was screened using a 72-base oligonucleotide derived from the sequence (nucleotides 199-270) of a previously cloned rat striatal D2 dopamine receptor (Bunzow et al., 1988). An oligonucleotide encoding most of the putative second transmembrane domain of the D2 receptor was selected for use as a probe because this region is highly conserved among catecholamine receptors. The probe was synthesized, purified, and labeled by phosphorylation of the 5'-OH group using 3~P-T-ATP and T4 polynucleotide kinase. The library was distributed on master agar plates, and 2 x 106 recombinants were screened by hybridizing duplicate nitrocellulose filters in 2x SSC (1 x = 0.15 M NaC1, 0.015 M Na3C6HsO7, pH 7), 10x Denhardt's solution (1 x = 0.02% polyvinylpyrrolidone, 0.02% Ficoll, 0.02% bovine serum albumin), 0.1% Na4P2OT, 0.1% sodium dodecyl sulfate (SDS), 50/~g/ml sheared salmon sperm DNA, and 3zp-labeled oligonucleotide (2-10 x 106cpm/ml) at 42°C for 18 hr. Filters were sequentially washed at 60°C in solutions containing decreasing SSC concentrations. Two positive clones, identified as D229 and D237, were identified by autoradiography after washing in 0.5x SSC and plaque purified. ,~DNA was prepared and restriction mapping was performed to compare the overall pattern of the clones. For sequencing, cDNA fragments were subcloned into plasmids (pTZ18R, Pharmacia, Piscataway, N.J.; or pSP64, Promega, Madison, Wis.). Nucleotide sequence analysis was done by the Sanger dideoxy nucleotide chain termination method on denatured double-stranded plasmid templates by the use of either the Klenow fragment of DNA polymerase I (Promega) or a modified bacteriophage T7 DNA polymerase (Sequenase; U.S. Biochemical Corp., Cleveland, Ohio). Polymerase Chain Reaction. Two partial cDNA clone fragments, corresponding to the 5' and 3' ends of a native cDNA encoding the D2 dopamine receptor, were amplified by use of the polymerase chain reaction (PCR) and ligated to reconstitute a full-length cDNA clone. The amplified segments were spliced together by constructing an XbaI site (TCTAGA) at bp 178-183 (see Fig. 1). The 5' segment of the composite was generated by amplifying clone D229 from bp - 9 9 to bp 183 (see Fig. 1) using the primers 5'ACGTAAGCTTGAAT-FCCGAGAGCCCCGGCGGGCA3' and 3'Cq-TCTCTCTAGACACAGCCATGCACAC5'. A HindlII site (AAGCq'T) was incorporated into the former primer for subcloning purposes. The 3' segment

440

Dearry, Falardeau, Shores, and Caron

GAATTccGAGAGCCCCGGC•GGCAGCAGCCGGCGCCGTCTC•GCCCCGGGGCGCC•TATGGCTTGAAGAGcCTGGCCACCCAGTGGcT•CAcCGCCCTG

-i

A T G GAT CCA CTG AAT CTG TCC TGG TAT GAT GAT GAT CTG GAG AGG CAG AAC TGG AGC CGG CCC TTC AAC GGG TCA MET A s p Pro Leu Ash Leu Set Trp Tyr Asp A s p A s p Leu Glu Arg Gln As___~nTrp Set A r g Pro Phe As_~n Gly Set

75 25

GAC GGG A A H GCG GAC AGA CCC CAC TAC AAC TAC TAT GCC ACA CTG CTC A C C CTG CTC ATC GCT GTC A T C GTC TTC A s p Gly Lys Ala Asp A r g Pro His Tvr A~m Tvr Tvr A)a Thr T.~u L~u Thr Leu Leu Tie AIR Va] If@ Val Phe

150 50

GHC A A C GTH CTG GTH TGC A T E GCT GTH TCT AGA GAG A A G GCG CTG CAG A C C A C C ACC AAC TAC CTG ATC GTC AGC Gly A s h Val L~n Val Cvs MET A]a Va] Set A r g Glu Lys Ala Leu Gln Thr Thr Thr Ash Tvr Leu Tle Val Set

225 75

CTC G C A GTG GCC GAC CTC CTC GTC GCC ACA CTG GTC A T G CCC TGG GTT HTC TAC CTG GAG GTG GTA GGT GAG TGG Leu A1a Val Ala A S p Leu Leu Val Ala Thr Len Val MET Pro T ro Val Val Tvr Leu Glu Val Val Gly G l u Trp

300 I00

A A A TTC AGC A G G A T T CAC TGT GAC A T C TTC GTC ACT CTG HAC GTC A T G A T G TGC A C G GCG A G C ATC C T G A A C TTG Lys Phe Set A r g Ile Hls Cvs A S p Ile Hhe Val Thr LeU Ash Val MET M E T Cvs Thr Ala Set Ile Leu A s h l~u

375 125

TGT G C C A T C A G C A T C HAC A G G TAC A C A GCT HTG GCC A T G CCC ATG C T G TAC A A T ACG CGC TAC A G C TCC A A G CGC cvs A]a Tle ~ r ~i@ A s m A r g Tyr Thr Ala Val Ala MET Pro MET Leu Tyr A s n T h r A r g Tyr Set Ser Lya Ara

450 150

CGG G T C ACC GTC A T G A T C TCC ATC GTC TGG GTC CTG TCC TTC A C C A T C TCC TGC C C A CTC C T C TTC G G A CTC AAT A r a Val Thr Val M E T T}a S@r []e Va) Trn Val l,~n Set Ph@ Thr T)e S@r Cvs Pro T,eu Lell P h m ~ ) v T ~ u A~n

525 175

A A C H C A GAC C A G A A C G A G THC ATC ATT GCC AAc CCG GCC TTC GTH GTC TAC TCC TCC ATC G T C TCC TTC TAC GTG A s n Ala A s p Gln A s h Hlu Cys ~le Ile Ala A S h Pro Ala Phe Va] Val Tvr Set Set Tle Val Set P h e Tvr Val

600 200

C C C TTC A T T GTC ACC C T G CTG GTC TAC A T C A A G A T C TAC ATT GTC CTC C G C A G A CGC CGC A A G C G A G T C A A C ACC Pro Phe Ile Val T h r Leu Leu Val Tvr Ile Lvs Ile Tvr Ile Val Leu A r g A r g A r g A r g Lys A r g Val A S h Thr

675 225

A A A CGC A G C AGC C G A GCT TTC AHG GCC C A C CTG A G G GCT CCA CTA A A G G A G GCT GCC CGG C G A G C C C A G GAG CTG Lys A r g Ser Set A r g A l a Phe Arg Ala His Leu A r g A l a Pro Leu Lys G l u A l a Ala A r g A r g A l a G l n Hlu Leu

750 250

GAG A T G GAG ATG CTC TCC A G C A C C AGC C C A CCC GAG A G G A C C CGH TAC A G C CCC A T C CCA CCC A G C C A C CAC CAG Hlu M E T Glu MET Leu Set Set Thr Set Pro Pro Glu A r g Thr A r g Tyr Set Pro Ile Pro Pro Set His His Gln

825 275

CTG A C T CTC CCC GAC CCG TCC CAC CAT GGT CTC CAC A H C ACT CCC HAC A G C CCC GCC AAA C C A G A G A A G A A T GGG Leo Thr Leu Pro A s p Pro Set His Hi~ Gly Leu His Set Thr Pro A s p Set Pro Ala Lys P r o G l u Lys A s h Gly

900 300

CAT GCC A A A GAC C A C CCC A A G ATT GCC AAG ATC TTT GAG ATC CAG A C C A T E C C C AAT GGC A A A A C C C G G ACC TCC His A l a Lys Asp His Pro Lys Ile Ala Lys Ile Phe Glu Ile Gln Thr MET P r o Ash Gly Lys T h r A r g Thr Set

975 325

CTC A A G ACC ATG A G C CGT A G G AAG CTC TCC CAG CAG A A G GAG A A G A A A HCC A C T CAG A T G C T T GCC A T T GTT CTC 1050 Leu Lys Thr MET Set A r g A r g Lys Leu Ser Gln Gln Lys Glu Lys Lys Ala Thr Gln MET T,en Ala Tie Va] Len 350 GGT HTG TTC ATC A T C TGC TGG CTG CCC TTC TTC A T C A C A CAC ATC CTG A A T A T A CAC TGT G A C T H C A A C ATC CCG 1125 ~Iv Val Hh~ Tle Ile Cvs Trn T,~n Pr~ Pbe Phe Yle Thr H~s I)e T.~II Ann ~le His Cvs A s p Cys A s n lle Pro 375 C C T G T C CTG TAC A G C GCC TTC ACG TGG CTG GGC TAT GTC AAC A G C GCC G T G A A C CCC A T C A T C TAC A C C A C C TTC 1200 P r o ~al Leu Tvr Set Ala Phe Thr Trn Leu Glv Tvr Val Ash S~r A l a Val A s h Pro Ile Ile Tvr Thr Thr Phe 400 A A C A T T GAG TTC C G C A A G GCC TTC CTH A A H A T C CTC CAC TGC TEA C T C T G C T G C C T G C C C G C A C A G C A G C C T G C T T C C C A C C T C A S h Zle Glu Phe A r g Lys A l a Phe Leu Lys lle Leu His Cys :

1284 414

•CTGCCCAGGCCGGCCAGCCGTCACCCTTGCGAACCGTGAGCAGGAAGGCCTGGGTGGATCGGCCTCCTCTTCACCCCGGCAGCCCTGCAGTGTTCGCT

1383 1482 1581 1680 1779 1878 1977 2076 2175 2274 2373 2375

TGGCTC~ATGCTCCTCACTGCCCGCACACC~TCACT~TGCCAGGGCAGTGCTAGTGAGCTGGGCATGGTACCAGCCCTGGGGCTGGCCC~CCAGCTcAG GGGCAGCTCATAGAGTCCCCCCTCCCACCTCCAGTCCCCCTATCCTTGGCACCAAAGATCGAGCCGCCTTCC TTGACCTTCCTCTGGGCTCTAGGGTTG CTGGAGCCTGAGTCAGGGCCCAGAGGCTGAGTTTTCTCTTTGTGGGGCTTGGCGTGGAGcAGGCHGTGGGGAGAGATGGACAGTTCA•ACCCTGCAAGG CCCACA~GAG~CAAGCAA~CTcTCTTGCCGA~AHCCAGGCAACTTCA~TCCTGG~AGACCCATHTAAATACCAGACTGCAG~TTGGAC~CCAGAGATT

CCCAAGCcAAAAACCTTAGCTCCCTC•C•CACCCCGAT•TGGACCTCTACTTTCCAGGCTAGTCCGGACCCAcCTcA•C•CGTTACAGCTCCCCAAGTG GTTTCCACAT~CTCTGA~AAGAGGAGCCCTCATCTTGAA~GGCCAGGAGGGTCTATGGGGAGAGGAACTCCTTGGCCTAGCCCACCCTGCTGCCTTCTG ACGGCCCTGCAATGTATCCCTTCTCACAGCACATGCTGGCCAHCCTGGGGCCTGGCAGGHAGGTCAGGCCCTGGAAC TCTATCTGGGCCTGGGCTAGGG

GA~ATCAGA~GTTCTTTGAGGGACTG~CT~T~C~ACACTCTGACGCAAAA~CACTTTCCTTTTCTATTCCTTCTG~CCTTTCCTCTCT~CTGTTTCCCT TCCCTTCCACT••CTCT•CCTTAGA••ACCCACG•cTAA•A•GCTGCT•AAAACCATCTGGCCT••CCTGGCCCTGCCCTGAGGAAG•AG•GGAAGCTG

•AGCTTG•GAGAG•CCCTGHGGCCTA•ACTCTGTAACATCACTATC•ATGCA•cAAACTAATAAAACTTTGACGA•TCAAAAAAAAAAAAAAACHGAAT TC

Fig. 1. Nucleotide and deduced amino acid sequence of a c D N A clone encoding a human retinal D 2 dopamine receptor, Nucleotide sequence is numbered in the 5' to 3' direction beginning with the first ATG of the open reading frame. Preceding bases are indicated by negative numbers. Nucleotides are numbered at the right-hand end of each line. The deduced amino acid sequence, shown below the nucleotide sequence, is numbered at the right-hand end of each line beginning with the initiator methionine (Met). Hydrophobic segments representing putative membrane domains are underlined. Putative sites of N-linked glycosylation are indicated by double underlines. The GAAT]'C on each end of the clone is an EcoRI insertion site into the ), arms. During construction of the PCR composite (see text), three nucleotide substitutions shown here were made: T for a C at bp 180 and A for C at bp 181 and bp183. No amino acids were changed. The PCR composite terminates at bp2191, and the remaining 3' untranslated sequence is from clones D229 and D237. A polyadenylation signal is found at bp 2334-2339, and a poly(dA) tail at bp 2353-2367.

Retinal D 2

DoparnineReceptors

441

of the composite was generated by amplifying clone D237 from bp 172 to bp 2191 using the primers 5'GCTGTGTCTAGAGAGAAGGCGCTGCAG3' and 3 ' A C G T A G T A C T C G A G G C A G T G G A A G G G A A G G G A A A C A G 5 ' . Restriction endonuclease sites (ScaI = AGTACT; XhoI = CTCGAG) were incorporated into the latter primer for subcloning purposes. Two PCR reactions were performed, each containing 100 ng of the appropriate cDNA template, 100 pmol each of the two primers, a dNTP mix containing 200/~M of each dNTP, a reaction buffer containing 1.5 mM MgCI2, and 2 U of Taq polymerase (PerkinElmer Cetus, Norwalk, Conn.). Components were subjected to 25 cycles of amplification (1-min denaturation at 94°C; 1-min annealing at 55°C; 2-min extension at 72°C) in a thermal cycler (Coy Model 60, Ann Arbor, Mich.). Each reaction resulted in a single product: a fragment approximately 300 bp in length using D229 as template and approximately 2 kb in length using D237 as template. Each reaction product was loaded in a separate lane of a 1% agarose gel, electrophoresed, and recovered from the gel using a D E A E membrane (Schleicher and Schuell NA-45). The PCR products were then combined and digested simultaneously with HindlII, XbaI, and XhoI. PCR products were subcloned into the plasmid pCMV5 for expression. The vector was digested with HindIII and SalI. Treatment with the latter endonuclease yields a terminus that is compatible with XhoI, although neither of the original sites is regenerated upon ligation. Digested PCR products and plasmid were combined in a three-way ligation. The resulting product was sequenced and found to be identical to the appropriate regions of D229 and D237. Expression. The cDNA putatively encoding a D2 dopamine receptor was subcloned into the expression vector pCMV5. The DEAE-dextran method was used for transient transfection of COS-7 cells (see Cullen, 1987). After 48 hr, cells were lysed and homogenized in ice-cold lysis buffer (10 mM Tris-HC1, pH 7.4, 5 mM EDTA). The homogenate was centrifuged at 43,000g for 30 min at 4°C. The resulting pellet was resuspended in an assay buffer containing 100 mM NaCI, 50mM Tris-HC1, pH7.4, and 2 m M MgC12. All binding experiments were initiated by the addition of 50 #1 of membrane suspension (approximately 20 ~g protein) to a total assay volume of 2 ml. Incubations were carried out at room temperature for 1 hr. They were terminated by the addition of 5 ml of ice-cold 50 rnM Tris-HCl, pH 7.4, 100 mM NaCI, followed by rapid vacuum filtration using a cell harvester (Brandel, Gaithersburg, Md.) through Whatman GF/C filters. Two subsequent 5-ml washes were performed. For saturation assays, increasing concentrations of 3H-spiperone (sp act, 27.5 Ci/mmol; NEN, Boston, Mass.) were added, and nonspecific binding was defined with 1/tM (+)butaclamol. Competition assays contained approximately 0.5 nM 3H-spiperone and varying concentrations of agonists or antagonists. Agonist incubations contained 1 mM Na2S205 to retard oxidation. Data were analyzed by nonlinear least-squares curve fitting on a VAX computer using a one-site model (DeLean et al., 1982). Northern Blot Analysis. Rat brain and retina poly(A) + RNA (Clontech, Palo Alto, Calif.) was denatured in 50% formamide and separated (1.7/~g/lane) on a 1.2% agarose gel containing 2.2M formaldehyde. RNA was transferred to

442

Dearry, Falardeau, Shores, and Caron

Nytran filters (Schleicher and Schuell, Keene, N.H.) in 20× SSC and UV cross-linked. Filters were prehybridized for 2hr at 65°C in 5× Denhardt's solution, 5× SSPE ( i x = 0.18M NaCI, 10 mM NaH2PO4, pH 7.7, 1 mM EDTA), 0.5% SDS, 200/~g/ml sheared salmon sperm DNA, 200/~g/ml total yeast RNA, and 50% formamide. Filters were then hybridized for 48hr at 65°C in 1× Denhardt's solution, 5× SSPE, 0.25% SDS, the same concentrations of salmon sperm DNA, total yeast RNA, and formamide as used in the prehybridization buffer, and 32p-labeled probe (2 × 106 cpm/ml). Filters were washed at 60°C in 0.1× SSC, 0.1% SDS. Riboprobes were generated from a 1516-bp SacI/BamHl fragment of a cDNA clone encoding a rat D2 dopamine receptor (provided by Olivier Civelli, Oregon Health Sciences University). This cDNA restriction fragment, encoding putative transmembrane domains VI and VII of the receptor as well as the 3' untranslated region, was subcloned into the same sites of pBluescript II S K + (Stratagene, La Jolla, Calif.). The vector was linearized with PvulI for antisense (complementary to mRNA) R N A synthesis from the T7 promoter or with BamHI for sense R N A synthesis from the T3 promoter. R N A probes were labeled with 32p-UTP during synthesis with the appropriate R N A polymerase. In Situ Hybridization Histochemistry. A human eye from a 40-year-old female donor was obtained within 1 hr of removal (provided by Dr. Gordon Klintworth, Duke University Eye Center). Sections of the posterior eyecup approximately 1 in. in diameter were immersed in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, for 6 hr at 4°C. Tissue was then cryoprotected in 20% sucrose in phosphate-buffered saline overnight at 4°C. All water was treated with 0.1% diethylpyrocarbonate to inhibit RNase activity. Tissue was frozen and 12-#m sections were cut in a cryostat. In situ hybridization was performed by a modification of the procedure described by Fremeau et al. (1986). Sections were mounted on gelatin-subbed slides, prehybridized for 2hr at 50°C in 600mMNaCI, 10mM Tris-HC1, pH7.5, 0.02% Ficoll, 0.02% polyvinylpyrrolidone, 0.1% bovine serum albumin, l m M E D T A , 500/~g/ml sheared salmon sperm DNA, 500#g/ml total yeast RNA, 50/~g/ml yeast tRNA, and 50% formamide. The labeled cRNA probe was then added to this solution at a concentration of 2-5 x 106cpm/ml along with 10mM dithiothreitol. Sections were hybridized overnight at 50°C, washed 30rain at room temperature in 2× SSC, digested with 30/~g/ml RNase A at 37°C for 60 min, washed 60 min at 50°C in 2× SSC, washed 3hr at 50°C in 0.1× SSC, 0 . 5 % N a pyrophosphate, 14mM~6-mercaptoethanol, and then washed overnight at room temperature in this same solution. They were dehydrated through 50, 70, and 90% ethanols containing 0.3 M ammonium acetate, and vacuum-dried overnight in a desiccator. Slides were dipped in emulsion (Kodak NTB2), exposed for 1-4 weeks, and counterstained with hematoxylin and eosin. Riboprobes were generated from a 297-bp BgllI fragment of the cDNA clone encoding the human retinal D2 dopamine receptor. This cDNA restriction fragment, encoding the putative third cytoplasmic loop of the receptor, was subcloned into the same site of pSP72 (Promega). The vector was linearized with HindlII for antisense R N A synthesis from the T7 promoter or with HpaI for

Retinal D2 Dopamine Receptors

443

sense R N A synthesis from the SP6 promoter. R N A probes were labeled with 35S-UTP during synthesis with the appropriate R N A polymerase.

RESULTS

Cloning. Two clones, designated D229 and D237, each approximately 2.5 kb in length, were isolated from the human retina cDNA library after filters were washed at 60°C, 0.5× SSC. Positive hybridization signals for these clones were no longer observed on the autoradiographs after filters were washed at 60°C, 0.2× SSC. Another clone, designated D233, was isolated at an earlier stage in the screening process, i.e., at lower stringency (see Dearry et al., 1990b). D229 and D237 were sequenced by the technique of primer extension. Each of these clones was found to encode an open reading frame that was highly homologous to the rat striatal D2 dopamine receptor. However, each of these clones exhibited a nucleotide rearrangement toward the 5' end. These rearrangements most likely represent a cloning artifact that occurred during construction of the cDNA library. Following self-priming to synthesize the second strand of cDNA, the double-stranded cDNA product is closed by a hairpin loop (see Sambrook et al., 1989). This loop is subsequently digested with nuclease $1. The latter treatment frequently results in loss or rearrangement of sequences corresponding to the 5' end of the mRNA (Land et al., 1981). A defective cleavage of the hairpin loop could account for the rearrangements observed in these clones. Sequencing demonstrated that D237 was homologous to the rat striatal D2 dopamine receptor from bp75 to the 3' end of the clone; the beginning 5' sequence was not homologous. D229 was homologous to the rat striatal D2 dopamine receptor from bp = - 9 9 (5' origin) to bp 300 and from bp 700 to the 3' end of the clone; the intervening 400 bp was not homologous. The two clones were identical in their regions of overlap that bore homology to the previously cloned D2 receptor. One such region, bp 75 to 300, represented an area in which the two clones could conceivably be spliced together to yield a composite clone (5' end from D229; 3' end from D237) entirely homologous to the rat striatal D2 receptor. However, no convenient unique restriction site existed within this region. We therefore used PCR to amplify the appropriate segment from each clone in order to produce a composite clone without changing the encoded amino acid sequence (see Materials and Methods). The nucleotide and deduced amino acid sequences of the composite clone are shown in Fig. 1. The longest open reading frame in this cDNA encodes a protein of 414 amino acids having a predicted molecular weight of approximately 47,000. The 3' untranslated region contains a polyadenylation signal ( A A T A A A at bp 2334-2339) and a poly(dA) tract beginning at bp 2353. A hydrophobicity plot (Fig. 2a) of the deduced protein sequence indicates the presence of seven hydrophobic stretches of 20-27 amino acids that putatively represent seven transmembrane domains. Based on this plot, the predicted

444

Dearry, Falardeau, Shores, and Caron

x "0 ,r-

2

.£

0

Q.

9

"0

-2

"r -4 50

(a)

1O0

150

200

250

300

350

400

Amino Acid Residue

Fig. 2. (a) Hydrophobicity profile of the amino acid sequence deduced from a e D N A clone encoding a human retinal D 2 dopamine receptor. Hydrophobic indices were calculated from the Kyte-Doolittle scale and averaged over a window of six amino acid residues. The seven highest peaks of hydrophobicity are numbered with Roman numerals in order of their appearance within the sequence. These stretches putatively represent seven transmembrane domains. (b) Predicted organization of a human retinal D 2 dopamine receptor in the plasma membrane. Amino acids are indicated by a single-letter code within circles. Three potential sites for N-linked glycosylation in the amino terminus are shown by shaded circles. The structural organization of the Dz receptor is similar to that of other G protein-coupled receptors.

organization of the protein in the plasma membrane is shown in Fig. 2b. The protein contains a long third cytoplasmic loop (135 residues) and a short carboxyl terminus (14 residues). These features are characteristic of other catecholamine receptors coupled to Gi, e.g., o~2-adrenergic receptors (see Lefkowitz and Caron, 1988). In addition, the protein possesses a carboxyl terminal cysteine (Cys 414) that may be palmitoylated as in the flz-adrenergic receptor (O'Dowd et al., 1989) and rhodopsin (Ovchinnikov et al., 1988). There are many Ser and Thr residues in the third cytoplasmic loop that could serve as potential sites of regulatory phosphorylation. For example, agonist-induced phosphorylation of the fladrenergic receptor plays a role in desensitization, i.e., in reducing the capacity of agonist-occupied receptor to stimulate its effector, adenylyl cyclase (see Benovic et al., 1988). The predicted amino acid sequence of the D2 dopamine receptor also contains three consensus sites for N-linked glycosylation in the amino terminus, and the predicted mass of this protein is similar to that reported for the deglycosylated D2 dopamine receptor from rat striatum (Grigoriadis et al., 1988). Together, these findings suggest that the structural organization of this protein is simialr to that of other G protein-coupled receptors. Expression. To verify that this eDNA clone encodes a D2 dopamine receptor, African green monkey kidney (COS-7) cells were transiently transfected with pCMV5 containing the composite PCR eDNA insert. Before transfection,

Retinal D 2 Dopamine Receptors

445

•:,'- • ~ ; ' : ~

×> "- .>