Neurochem. Int. Vol. 20, Suppl., pp. 17S-22S,1992 Printed in Great Britain
0197-0186/92$5.00+0.00 PergamonPressplc
MOLECULAR NEUROBIOLOGY OF DOPAMINE RECEPTOR SUBTYPES DAVID R. SIBLEY,l FREDERICKJ. MONSMA JR, l LORIS D. McVITTIE, ~ CHARLES R. GERFEN,2 RONALDM. BURCH3 and LAWRENCEC. MAHAN2 ~Experimental Therapeutics Branch, NINDS and 2Laboratory of Cell Biology, NIMH, National Institutes of Health, Bethesda, MD 20892 U.S.A. and 3NOVA Pharmaceutical Corporation, 6200 Freeport Centre, Baltimore, MD 21224, U.S.A.
Dopamine receptors belong to a large class of neurotransmitter and hormone receptors which are linked to their signal transduction pathways via guanine nucleotide binding regulatory (G) proteins. Pharmacological, biochemical and physiological criteria have been used to define two subcategories of dopamine receptors referred to as D1 and D2 (Creese and Fraser, 1987). Dt receptors are associated with the activation of adenylyl cyclase activity and are coupled with the G, regulatory protein. In contrast, activation of D2 receptors results in various responses including inhibition of adenylyl cyclase activity, inhibition of phosphatidylinositol turnover, increase in K + channel activity and inhibition of Ca 2+ mobilization (Vallar and Mendolesi, 1989). The G protein(s) linking the D2 receptors to these responses have not been identified, although D2 receptors have been shown to both copurify and functionally reconstitute with both "G{' and "Go" related proteins. One means of achieving the diversity of second messenger pathways associated with D2 receptor activation would be the existence of multiple D2 receptor subtypes, each being coupled with a different G protein-linked response. Efforts towards elucidating D2 receptor diversity were recently advanced by the cloning of a eDNA encoding a rat D2 receptor. This receptor exhibits considerable amino acid homology with other members of the G protein-coupled receptor super-family for which cDNAs and/or genes have been cloned. In the present investigation (Monsma et al., 1989a) we report the identification and cloning of a eDNA encoding an RNA splice variant of the rat D 2 receptor eDNA (Bunzow et al., 1988). This eDNA codes for a receptor isoform which is predominately expressed in the brain and contains an additional 29 amino acids in the 3rd cytoplasmic loop, a region believed to be involved with G protein coupling. This is the first example of a novel G-protein coupled receptor isoform generated by alternative R N A splicing.
As part of an effort to isolate cDNAs encoding dopamine receptor subtypes, we initially constructed a lambda ZAP II eDNA library using m R N A purified from rat striatum, the region of the brain known to contain the highest levels of both Dt and D2 dopamine receptors. This library was screened with a mix of two 36 mer synthetic oligonucleotides, the sequence of which was derived from amino acids 352-363 of the rat D 2 receptor eDNA (Bunzow et al., 1988). This region corresponds to the 6th transmemhrane spanning domain and is known to exhibit very high homology among previously cloned G protein-coupled receptors. Out of 1 x 106 recombinants screened, a total of 15 positive clones were isolated. Restriction analysis and partial sequence information indicated that 5 of these clones were related to the rat D 2receptor eDNA previously reported. One of the clones containing an insert of 2.5 kb was completely sequenced. The longest open reading frame in this eDNA codes for a 444 amino acid protein with a molecular weight of 50,887 Da. Fig. 1 depicts this D2 receptor protein as it is believed to be organized in the plasma membrane. The nucleotide and amino acid sequence within the coding region is identical to the rat D2 receptor eDNA previously reported (Bunzow et al., 1988) with the notable exception of an additional 87 bp sequence coding for a 29 amino acid insertion between residues 241 and 242. This is located within the predicted 3rd cytoplasmic loop approximately 30 amino acids away from the carboxy terminus of the 5th transmembrane spanning domain. In addition to this insertion sequence, and a slightly extended 5' untranslated sequence, we also noted 5 base differences within the 3' untranslated region in comparison with the previously published sequence (Bunzow et al., 1988). Subsequent sequence analysis- indicated that all 5 of the D2 receptor-related cDNAs isolated from this library contained the identical 87 bp insertion sequence. The nucleotide 17S
18S
I)op~llnine 90 O-IO
O--IO
EXTRACELLULAR
INTRACELLULAR
Fig. 1. The rat D 2dopamine receptor. The shaded amino acids indicate the splice variation insertion sequence. sequences delineating the boundaries of this insertion sequence correspond with consensus exon sequences for R N A splice junctions, suggesting that the cDNA resulted from alternative R N A splicing. In order to confirm the D2 subtype identity of this c D N A clone and to determine if the 29 amino acid insertion sequence results in a major alteration in the ligand binding properties of the D2 receptor, the c D N A was inserted into the SV40 promoter-driven vector, pEUK-C1, for expression in eukaryotic cells. The resulting plasmid (pEUK-D2L) was used to transiently transfect COS-7 cells. Three days after transfection, membranes were prepared and the binding of the D2 dopaminergic antagonist [3H]methylspiperone was examined. [3H]methylspiperone bound to the membranes in a saturable fashion with high specific activity (1 pmol/mg protein) and an affinity (62.1 + 2.1 pM) in good agreement with that found in the rat striatum. No specific binding activity was
detected in COS-7 cells that had not been transfected with pEUK-D2L or transfected with the pEUK-CI vector alone. A variety of dopaminergic ligands to compete for specific [3H]methylspiperone binding to transfected COS-7 cell membranes was examined. The high affinity D2-selective antagonist, spiperone c36 ~: 3.8 pM) is the most potent agent followed by the nonselective dopaminergic antagonist (+)butaclamol (0.52 + 0.01 nM) which ~s more than 4 orders of magnitude more potent than its inactive isomer. (-)butaclamol > 10 mML The D2-selective antagonist ( - ) s u l p i r i d e (7.9 ± 0.42 nM) also exhibits high affinity whereas the D -selective antagonist SCH23390 (0.41 + 0.047 mM) does not. This rank order of potency as well as the absolute affinities (K3 of the antagonists agree well with those previously demonstrated for D2 receptors (Creese and Fraser. 1987). Dopamine is also able to completely inhibit [3H]methylspiperone binding ( K , - 0.71 -4- 0.012 mM)
Dopamine 90 although the competition curve is homogeneous (Hill coefficient = I) and not significantly affected by guanine nucleotides indicating the absence of appropriate G protein coupling in the COS-7 cells. These experiments indicate that the insertion sequence does not appear to affect the basic properties of ligand recognition for the D2 receptor. In order to verify the expression of the D2 receptor variant containing the insertion sequence and determine the relative proportions of the two receptor isoforms, we subjected various rat tissues to Northern blot analysis using an oligonucleotide probe to a consensus region as well as an insert sequence-specific probe. The tissues expressing the highest levels of the 2.9 kb D2 receptor mRNA are the striatum and pituitary. The retina showed a moderate abundance of mRNA with low levels being observed in the mesencephalon and cortex and trace quantities detected in the olfactory bulb and hippocampus. Little to no mRNA was found in the cerebellum and kidney. This tissue distribution corresponds closely to that previously determined for De receptor expression. Of greatest interest, however, is the fact that in all of the tissues examined, the amount of mRNA detected with the two probes is very similar and in no instance did the consensus probe detect greater quantities of mRNA. To further investigate the relative distributions of the two mRNAs encoding the D2 receptor isoforms, we performed in situ hybridization analysis in the rat forebrain with the two oligonucleotide probes used for the Northern analysis. Identical patterns of labeling were obtained in a coronal section of rat brain which includes the striatum using both the consensus region probe and the insert sequence probe. The highest labeling occured in the striatal neurons where about 50% of the medium sized neurons were labeled. Larger sized neurons in the striatum also exhibited labeling. It is interesting that in the Northen blot and in situ hybridization analyses, there did not appear to be any difference in the levels of mRNA detected using the two oligonucleotide probes. If any tissue or brain area expressed mRNA containing the insertion sequence at a level equal to or less than the one lacking the insertion, then the consensus probe should detect mRNA levels that are at least 2-fold greater than those seen with the insert probe. These experiments thus indicate that not only is the longer De receptor variant (which we propose designating D~L)expressed in brain and other tissues, but in those areas which have been examined (especially the striatum), it appears to be the major if not exclusive isoform. Further experiments directed at determining the actual levels of the receptor
19S
proteins will be required to confirm this point. At present, the location of predominant expression of the shorter D2 receptor lacking the insertion sequence (now designated D2s) is unclear. With the exception of the visual opsins, the genes for the G protein-coupled receptor family have, in most instances, demonstrated a lack of introns within their coding sequences thus precluding the generation of receptor diversity through alternative RNA splicing. Recently, however, it has been determined that the serotonin 5HTlc (B.J. Hoffman, personal communication) and D2 dopamine (Bun~w et al., 1988) receptors are encoded by genes which contain introns. Our current data (Monsma et al., 1989a) on the rat De receptor now provides the first example of G-protein-coupled receptor isoforms which are generated through alternative RNA splicing. These isoforms are defined by the presence or absence of an internal 29 amino acid sequence within the receptor protein. This observation has recently been confirmed by several other groups (Giros et al., 1989; Selbie et al., 1989; Chio et al., 1990). This RNA splice variation could have arisen either through the existence of a "cassette exon" or through alternative internal acceptor or donor sites within the precursor mRNA. The isolation and sequencing of the D2 receptor gene has recently been achieved (Grandy et al., 1989; Dal Toso et al., 1989) and shown to contain an 87 bp exon encoding the splice variation sequence. The location of this optional amino acid sequence is particularly intriguing as it occurs within the predicted 3rd cytoplasmic loop of the receptor. Recent mutagenesis studies using the beta2-adrenergic catecholamine receptor have indicated that this region is highly involved in G protein-receptor coupling. It is thus tempting to speculate that the two D2 receptor isoforms are coupled to different G proteins thus resulting in the diversity of responses asssociated with D2 receptor activation. Further work involving the stable expression of the two D2 receptor isoforms in cells exhibiting appropriate G protein-linked effector systems will be required to test this hypothesis. Evidence has also accumulated suggesting heterogeneity in the D~ category of dopamine r e c e p t o r s . D t receptors have recently been described in renal tissue which stimulate phospholipase C activity independently from that of adenylyl cyclase (reviewed in Felder et al., 1989). We have also shown, using Xenopus ooeyte expression experiments, that rat striatal mRNA encodes D~ receptors which are coupled to phospholipase C and Ca 2+ mobilization in a cAMP-independent fashion (Mahan et aL, 1990). These data suggest that there may be multiple D t
20S
l)opamine 90 clio
~ CHO J ~
_
_
2
EXTRACELLULAR
INTRACELLULAR
Fig. 2. The rat D, dopamine receptor linked to adenyly!cyclase activation. receptors linked to different signal transduction pathways or that a single, muitifunctional D, receptor exists. In order to investigate these possibilities and to characterize the D, receptor(s) on a molecular level, we have cloned the D, receptor subtype expressed in mouse NS20Y neuroblastoma cells (Monsma et al., 1990). We have previously shown that these cells express functional D, receptors coupled to the stimulation of cAMP generation (Monsma et al., 1989b). Poly (A) + RNA was used to first synthesize cDNA by reverse transcription followed by polymerase chain reaction (PCR) amplification with a pair of highly degenerate primers derived from the third and sixth transmembrane regions of the previously cloned adrenergic, D 2dopaminergic, and serotonin receptors. This process resulted in the amplification of several cDNA fragments which were preliminarily characterized by DNA sequence analysis. One of these fragments was found to exhibit considerable sequence homology to previously cloned G protein-coupled
receptors and was used to screen a rat striatal cDNA library in order to isolate a full-length clone. One clone (pB73D,) with an insert of 3.6 kb was isolated and found to strongly hybridize with the 32P-labeled PCR probe on dot-blot analysis. The nucteotide sequence of this cDNA was found to exhibit 90% homology in the region of the PCR fragment, the divergence of which is probably attributable to species differences (mouse vs. rat). The longest open reading frame in this cDNA codes for a 487 residue protein with a theoretical molecular weight of 54,264 Da (Fig. 2). Although the neighboring sequence of the first ATG in this reading frame is similar to Kozak's consensus initiation sequence, the third Met codon at position 46 actually provides a better match (Fig. 2). Additional work will thus be necessary to definitively assign the translational start site for this mRNA. Hydrophobicity analysis of the translated protein reveals seven clusters of 24 hydrophobic residues, predicted to represent transmembrane-
Dopamine 90 spanning domains, connected by three extracellular and three intraceUular loops. This pattern is similar to that observed for other cloned G protein-coupled receptors where the NH, terminus is proposed to be extrac~llular and the COOH terminus projects into the cytoplasm. The NH2 terminus contains one consensus site for N-linked glycosylation while the predicted third cytoplasmic loop exhibits one consensus recognition site for phosphorylation by the cAMPdependent protein kinase (Fig. 2). In addition, the long COOH terminus contains several serine and threonin~ residues possibly representing additional sites for regulatory phosphorylation. Comparison of the deduced amino acid sequence for the pB73D~ eDNA clone with the sequences of various catecholamine receptors indicated that the regions of highest identity appear to occur within the predicted transmembrane spanning domains. Within these regions, the pB73D~ protein exhibits sequence homologies of 44% with the rat D2 dopaminergic receptor; 44%, 43% and 40% with the human betar, beta2-, beta3-adrenergic receptors, respectively; and 43% and 42% with the hamster alpha~a- and human alpha~-adrenergic receptors, respectively. When compared with various serotonin receptors, the transmembrane regions of the pB73D~ protein exhibited homologies of 40% for 5HTI^ and 37% for both 5HT~c and 5HT2 receptors. The NH 2and COOH termini and the extracellular and intracellular loops are significantly more divergent among these receptors. It is interesting to note that within the third putative transmembrane spanning domain ofpB73D~, there is a conserved aspartate residue which is common to all biogenic amine receptors that have been sequenced thus far (Strader et al., 1989). Moreover, the fifth transmembrane spanning domain of pB73D~ also contains two serine residues which are conserved among catecholamine receptors and are critical for the recognition of agonist ligands possessing a catechol group (Strader et al., 1989). These observations would tend to suggest the pB73D~ clone encodes a receptor for an endogenous cateeholamine ligand. In an initial attempt to establish the identity of pB73D[, we analyzed the tissue distribution of its corresponding mRNA by Northern blot and in situ hybridization analyses. Northern blot analysis in various neural tissues reveals a transcript size of 4.1 kb which is predominantly located in the striatum with lesser amounts in the cortex and retina. In contrast, no mRNA is observed in the cerebellum, hippocampus, olfactory bulb, mesencephalon, or pituitary. In situ hybridization analysis also indicates a high abundance
21S
of mRNA in the striatum as well as in the olfactory tubercle. Approximately half of the medium sized neurons in the striatum are identified using this technique. The tissue distribution ofpB73Dl mRNA is remarkably similar to that of the D~ dopamine receptor as demonstrated by receptor binding and autoradiography studies (Creese and Fraser, 1987). To definitively establish the identity of the receptor encoded by the pB73D~ clone, the eDNA insert was subcloned into the pCD-SRa vector for expression in eukaryotic cells. The resulting plasmid, pSRa-D~, was used to transiently transfect COS-7 cells. [3H]SCH23390, a Dcselective radiolabeiled antagonist, binds to transfected COS-7 membranes in a saturable fashion with high specific activity (400 fmol/mg protein) and an affinity (0.3 ± 0.03 nM) in good agreement with that found in the rat striatum. No specific binding was detected in COS-7 cells that had not been transfected with pSRa-Dm or transfected with the pCD-SRa vector alone. The ability of a variety of dopaminergic ligands to compete for specific [3H]SCH-23390 binding to transfected COS-7 cell membranes was examined. (+)-SCH-23390 is the most potent agent (0.2 + 0.01 riM) and is approximately 200-fold more potent than its enantiomer, (-)-SCH-23388 (41 -4- 1.2 nM). The non-selective dopaminergic antagonist (+)-butaclamol also exhibits high affinity (2.8 + 0.2 nM) and is more than 4 orders of magnitude more potent than its inactive isomer, (-)-butaclamol (31 4- 0.8 gM). The D2selective antagonist spiperone exhibits relatively low affinity (290 4- 7 nM) as do the serotonin antagonists, ketanserin (0.42 + 0.031 gM) and mianserin (0.18 ± 0.042/z M). The endogenous agonist, dopamine, is also able to completely inhibit [3H]SCH-23390 binding (0.64 4. 0.092 gM). This rank order of potency as well as the absolute affinities (Ki) of these compounds agree well with those previously demonstrated for striatal Dt receptors. It was also demonstrated that pSRa-D~-transfected COS-7 cells exhibit D~ receptor-mediated stimulation of cAMP production. Dopamine stimulates cAMP production by approximately 2-fold in these transfected cells. In contrast, no response to dopamine is observed in non-transfected cells. The D~ selective agonists (+)SKF-38393 and (+)SKF-82958 also stimulate cAMP accumulation to a similar extent as dopamine. Although SKF-38393 has been reported to be a partial agonist at DI receptors, its fuller efficacy observed here is probably due the presence of spare receptors resulting from over-expression of receptor protein. In addition, the stimulation by SKF-38393 exhibits appropriate stereoselectivity with the
22S
Dopamine 90
( - ) i s o m e r exhibiting a lower potency. Finally, the beta-adrenergic agonist, epinephrine, also exhibits a low potency relative to dopamine as expected for a D~ receptor. In these experiments, the beta-adrenergic antagonist propranolol was included in the assays to preclude stimulation of the endogenous C O S - 7 cell beta-adrenergic receptor. These expression data thus confirm that the c D N A which we have cloned encodes a functional D~ dopamine receptor protein. It is important to emphasize that the D~ receptor which we have presently cloned (Monsma et al., 1990) is one which is functionally coupled to the stimulation of adenylyl cyclase. Recently, unique D, receptors have been described in kidney which stimulate phospholipase C activity independently from the activation of adenylyl cyclase (Felder et al.. 1989). We have also shown, using Xenopus oocyte expression experiments, that rat striatum contains m R N A encoding D~ receptors which can couple to phospholipase C, inositol phosphate production and Ca 2+ mobilization in a cAMP-independent fashion (Mahan et al., t990). It is interesting that the m R N A which codes for this D~ receptor-stimulated phospholipase C response is 2.5 kb in size (Mahan et al., 1990) in comparison with the 4.1 kb D~ receptor m R N A observed here. Moreover, in preliminary experiments, we have found that when m R N A is transcribed from the pB73D~ D~ receptor c D N A clone and injected into Xenopus oocytes, dopamine will stimulate c A M P accumulation 2-fold but is incapable of producing a Ca 2+ mobilization response. These findings suggest that the striatum contains two separate D~ receptor proteins which are coupled to different signal transduction pathways. Consequently, we propose that the D~ receptor subtypes linked to the activation of adenylyl cyclase and phospholipase C be designated D,A and D~B, respectively. Further experimentation involving the cloning and expression of the D m receptor subtype will be required to confirm this hypothesis. REFERENCES
Chio, C.L, Hess, G.F., Graham, R.S., and Huff, R.M. (1990). Second Molecular Form of D2 Dopamine Receptor in Rat and Bovine Caudate Nucleus. Nature
(London) 343, 26(~269. Creese, I., and Fraser, C.M., eds. (1987). Receptor Biochemistry and Methodology: Dopamine Receptors, (Liss, New York) Vol. 8, pp. 1-245. Bunzow, J.R., Van Tol, H.H.M., Grandy, D.K., Albert, P., Salon, J., Christie, M., Machida, C.A., Neve, K.A., and Civelli, O. (1988). Cloning and Expression of a Rat D~ Dopamine Receptor cDNA~ Nature (London) 336, 783 787. Dal Toso, R., Sommer, B., Ewert, M., Herb, A., Pritchett, D.B., Bach, A., Shivers, B.D., and Seeburg, P.H. (1989). The Dopamine D 2 Receptor: Two Molecular Forms Generated by Alternative Splicing. EMBO J. 8. 4025 4034. Felder, R.A., Felder, C.C., Eisner, G.M., and Jose, P.A. (1989). The Dopamine Receptor in Adult and Maturing Kidney. Am. J. Physiol. 257, F315-F327. Giros, B., Sokotoff, P., Martres, M.P., Riou, J.F,, Emorine, L.J., and Schwartz, J.C. (1989). Alternative Splicing Directs the Expression of Two D 2 Dopamine Receptor Isoforms. Nature (London) 342,923-926. Grandy, D.K., Marchionni, M.A., Makam, H., Stofko, R.E., Alfano, M., Frothingham, L., Fischer, J,B., BurkeHowie, K.J., Bunzow, J.R., Server, A.C., and Civelli, O. (1989). Cloning of the cDNA and Gene for a Human D~ Dopamine Receptor. Proc. Natl. Acad. Sci. USA 86, 9762-9766. Mahan, L.C., Butch, R.M., Monsma, F.J., Jr., and Sibley, D.R. (1990) Expression of Striatal D~ Dopamine Receptors Coupled to Inositolphosphate Production and Ca:' Mobilization in Xenopus Oocytes. Proc. Natl. Acad. Sci. USA, 87, 2196-2200. Monsma, F.J., Jr., McVittie, L.D., Gerfen, C.R., Mahan, UC., and Sibley, D.R. (1989a). Multiple D2 Dopamine Receptors Produced by Alternative RNA Splicing. Nature, 342, 926-929. Monsma, F.J., Jr., Brassard, D.L., and Sibley, D.R. (1989b). Identification and Characterization of D~ and D~ Dopamine Receptors in Cultured Neuroblastoma and Retinoblastoma Clonal Cell Lines. Brain Research, 492, 314-324. Monsma, F.J., Jr., Mahan, L.C., McVittie, LD., Gerfen, C.R., and Sibley, D.R. (1990) Molecular Cloning and Expression of a D~ Dopamine Receptor Linked to Adenylyl Cyclase Activation. Proc. Natl. Acad Sci. USA, in press. Selbie, L.A., Hayes, G., and Shine, J. (1989). The major Dopamine D 2 Receptor: Molecular Analysis of the Human D2A Subtype. DNA 8, 683-689. Strader, C.D., Sigal, I.S., and Dixon, R.A.F. (1989). Structural Basis of Beta-Adrenergic Receptor Function. FASEB J. 3, 1825-1832. Vallar, L. and Meldolesi, J. (1989). Mechanisms of Signal Transduction at the Dopamine D, Receptor. Trends Pharmacol. Sci. 10, 74- 77.
0197-0186/92$5.00+0.00 PergamonPressplc
MOLECULAR NEUROBIOLOGY OF DOPAMINE RECEPTOR SUBTYPES DAVID R. SIBLEY,l FREDERICKJ. MONSMA JR, l LORIS D. McVITTIE, ~ CHARLES R. GERFEN,2 RONALDM. BURCH3 and LAWRENCEC. MAHAN2 ~Experimental Therapeutics Branch, NINDS and 2Laboratory of Cell Biology, NIMH, National Institutes of Health, Bethesda, MD 20892 U.S.A. and 3NOVA Pharmaceutical Corporation, 6200 Freeport Centre, Baltimore, MD 21224, U.S.A.
Dopamine receptors belong to a large class of neurotransmitter and hormone receptors which are linked to their signal transduction pathways via guanine nucleotide binding regulatory (G) proteins. Pharmacological, biochemical and physiological criteria have been used to define two subcategories of dopamine receptors referred to as D1 and D2 (Creese and Fraser, 1987). Dt receptors are associated with the activation of adenylyl cyclase activity and are coupled with the G, regulatory protein. In contrast, activation of D2 receptors results in various responses including inhibition of adenylyl cyclase activity, inhibition of phosphatidylinositol turnover, increase in K + channel activity and inhibition of Ca 2+ mobilization (Vallar and Mendolesi, 1989). The G protein(s) linking the D2 receptors to these responses have not been identified, although D2 receptors have been shown to both copurify and functionally reconstitute with both "G{' and "Go" related proteins. One means of achieving the diversity of second messenger pathways associated with D2 receptor activation would be the existence of multiple D2 receptor subtypes, each being coupled with a different G protein-linked response. Efforts towards elucidating D2 receptor diversity were recently advanced by the cloning of a eDNA encoding a rat D2 receptor. This receptor exhibits considerable amino acid homology with other members of the G protein-coupled receptor super-family for which cDNAs and/or genes have been cloned. In the present investigation (Monsma et al., 1989a) we report the identification and cloning of a eDNA encoding an RNA splice variant of the rat D 2 receptor eDNA (Bunzow et al., 1988). This eDNA codes for a receptor isoform which is predominately expressed in the brain and contains an additional 29 amino acids in the 3rd cytoplasmic loop, a region believed to be involved with G protein coupling. This is the first example of a novel G-protein coupled receptor isoform generated by alternative R N A splicing.
As part of an effort to isolate cDNAs encoding dopamine receptor subtypes, we initially constructed a lambda ZAP II eDNA library using m R N A purified from rat striatum, the region of the brain known to contain the highest levels of both Dt and D2 dopamine receptors. This library was screened with a mix of two 36 mer synthetic oligonucleotides, the sequence of which was derived from amino acids 352-363 of the rat D 2 receptor eDNA (Bunzow et al., 1988). This region corresponds to the 6th transmemhrane spanning domain and is known to exhibit very high homology among previously cloned G protein-coupled receptors. Out of 1 x 106 recombinants screened, a total of 15 positive clones were isolated. Restriction analysis and partial sequence information indicated that 5 of these clones were related to the rat D 2receptor eDNA previously reported. One of the clones containing an insert of 2.5 kb was completely sequenced. The longest open reading frame in this eDNA codes for a 444 amino acid protein with a molecular weight of 50,887 Da. Fig. 1 depicts this D2 receptor protein as it is believed to be organized in the plasma membrane. The nucleotide and amino acid sequence within the coding region is identical to the rat D2 receptor eDNA previously reported (Bunzow et al., 1988) with the notable exception of an additional 87 bp sequence coding for a 29 amino acid insertion between residues 241 and 242. This is located within the predicted 3rd cytoplasmic loop approximately 30 amino acids away from the carboxy terminus of the 5th transmembrane spanning domain. In addition to this insertion sequence, and a slightly extended 5' untranslated sequence, we also noted 5 base differences within the 3' untranslated region in comparison with the previously published sequence (Bunzow et al., 1988). Subsequent sequence analysis- indicated that all 5 of the D2 receptor-related cDNAs isolated from this library contained the identical 87 bp insertion sequence. The nucleotide 17S
18S
I)op~llnine 90 O-IO
O--IO
EXTRACELLULAR
INTRACELLULAR
Fig. 1. The rat D 2dopamine receptor. The shaded amino acids indicate the splice variation insertion sequence. sequences delineating the boundaries of this insertion sequence correspond with consensus exon sequences for R N A splice junctions, suggesting that the cDNA resulted from alternative R N A splicing. In order to confirm the D2 subtype identity of this c D N A clone and to determine if the 29 amino acid insertion sequence results in a major alteration in the ligand binding properties of the D2 receptor, the c D N A was inserted into the SV40 promoter-driven vector, pEUK-C1, for expression in eukaryotic cells. The resulting plasmid (pEUK-D2L) was used to transiently transfect COS-7 cells. Three days after transfection, membranes were prepared and the binding of the D2 dopaminergic antagonist [3H]methylspiperone was examined. [3H]methylspiperone bound to the membranes in a saturable fashion with high specific activity (1 pmol/mg protein) and an affinity (62.1 + 2.1 pM) in good agreement with that found in the rat striatum. No specific binding activity was
detected in COS-7 cells that had not been transfected with pEUK-D2L or transfected with the pEUK-CI vector alone. A variety of dopaminergic ligands to compete for specific [3H]methylspiperone binding to transfected COS-7 cell membranes was examined. The high affinity D2-selective antagonist, spiperone c36 ~: 3.8 pM) is the most potent agent followed by the nonselective dopaminergic antagonist (+)butaclamol (0.52 + 0.01 nM) which ~s more than 4 orders of magnitude more potent than its inactive isomer. (-)butaclamol > 10 mML The D2-selective antagonist ( - ) s u l p i r i d e (7.9 ± 0.42 nM) also exhibits high affinity whereas the D -selective antagonist SCH23390 (0.41 + 0.047 mM) does not. This rank order of potency as well as the absolute affinities (K3 of the antagonists agree well with those previously demonstrated for D2 receptors (Creese and Fraser. 1987). Dopamine is also able to completely inhibit [3H]methylspiperone binding ( K , - 0.71 -4- 0.012 mM)
Dopamine 90 although the competition curve is homogeneous (Hill coefficient = I) and not significantly affected by guanine nucleotides indicating the absence of appropriate G protein coupling in the COS-7 cells. These experiments indicate that the insertion sequence does not appear to affect the basic properties of ligand recognition for the D2 receptor. In order to verify the expression of the D2 receptor variant containing the insertion sequence and determine the relative proportions of the two receptor isoforms, we subjected various rat tissues to Northern blot analysis using an oligonucleotide probe to a consensus region as well as an insert sequence-specific probe. The tissues expressing the highest levels of the 2.9 kb D2 receptor mRNA are the striatum and pituitary. The retina showed a moderate abundance of mRNA with low levels being observed in the mesencephalon and cortex and trace quantities detected in the olfactory bulb and hippocampus. Little to no mRNA was found in the cerebellum and kidney. This tissue distribution corresponds closely to that previously determined for De receptor expression. Of greatest interest, however, is the fact that in all of the tissues examined, the amount of mRNA detected with the two probes is very similar and in no instance did the consensus probe detect greater quantities of mRNA. To further investigate the relative distributions of the two mRNAs encoding the D2 receptor isoforms, we performed in situ hybridization analysis in the rat forebrain with the two oligonucleotide probes used for the Northern analysis. Identical patterns of labeling were obtained in a coronal section of rat brain which includes the striatum using both the consensus region probe and the insert sequence probe. The highest labeling occured in the striatal neurons where about 50% of the medium sized neurons were labeled. Larger sized neurons in the striatum also exhibited labeling. It is interesting that in the Northen blot and in situ hybridization analyses, there did not appear to be any difference in the levels of mRNA detected using the two oligonucleotide probes. If any tissue or brain area expressed mRNA containing the insertion sequence at a level equal to or less than the one lacking the insertion, then the consensus probe should detect mRNA levels that are at least 2-fold greater than those seen with the insert probe. These experiments thus indicate that not only is the longer De receptor variant (which we propose designating D~L)expressed in brain and other tissues, but in those areas which have been examined (especially the striatum), it appears to be the major if not exclusive isoform. Further experiments directed at determining the actual levels of the receptor
19S
proteins will be required to confirm this point. At present, the location of predominant expression of the shorter D2 receptor lacking the insertion sequence (now designated D2s) is unclear. With the exception of the visual opsins, the genes for the G protein-coupled receptor family have, in most instances, demonstrated a lack of introns within their coding sequences thus precluding the generation of receptor diversity through alternative RNA splicing. Recently, however, it has been determined that the serotonin 5HTlc (B.J. Hoffman, personal communication) and D2 dopamine (Bun~w et al., 1988) receptors are encoded by genes which contain introns. Our current data (Monsma et al., 1989a) on the rat De receptor now provides the first example of G-protein-coupled receptor isoforms which are generated through alternative RNA splicing. These isoforms are defined by the presence or absence of an internal 29 amino acid sequence within the receptor protein. This observation has recently been confirmed by several other groups (Giros et al., 1989; Selbie et al., 1989; Chio et al., 1990). This RNA splice variation could have arisen either through the existence of a "cassette exon" or through alternative internal acceptor or donor sites within the precursor mRNA. The isolation and sequencing of the D2 receptor gene has recently been achieved (Grandy et al., 1989; Dal Toso et al., 1989) and shown to contain an 87 bp exon encoding the splice variation sequence. The location of this optional amino acid sequence is particularly intriguing as it occurs within the predicted 3rd cytoplasmic loop of the receptor. Recent mutagenesis studies using the beta2-adrenergic catecholamine receptor have indicated that this region is highly involved in G protein-receptor coupling. It is thus tempting to speculate that the two D2 receptor isoforms are coupled to different G proteins thus resulting in the diversity of responses asssociated with D2 receptor activation. Further work involving the stable expression of the two D2 receptor isoforms in cells exhibiting appropriate G protein-linked effector systems will be required to test this hypothesis. Evidence has also accumulated suggesting heterogeneity in the D~ category of dopamine r e c e p t o r s . D t receptors have recently been described in renal tissue which stimulate phospholipase C activity independently from that of adenylyl cyclase (reviewed in Felder et al., 1989). We have also shown, using Xenopus ooeyte expression experiments, that rat striatal mRNA encodes D~ receptors which are coupled to phospholipase C and Ca 2+ mobilization in a cAMP-independent fashion (Mahan et aL, 1990). These data suggest that there may be multiple D t
20S
l)opamine 90 clio
~ CHO J ~
_
_
2
EXTRACELLULAR
INTRACELLULAR
Fig. 2. The rat D, dopamine receptor linked to adenyly!cyclase activation. receptors linked to different signal transduction pathways or that a single, muitifunctional D, receptor exists. In order to investigate these possibilities and to characterize the D, receptor(s) on a molecular level, we have cloned the D, receptor subtype expressed in mouse NS20Y neuroblastoma cells (Monsma et al., 1990). We have previously shown that these cells express functional D, receptors coupled to the stimulation of cAMP generation (Monsma et al., 1989b). Poly (A) + RNA was used to first synthesize cDNA by reverse transcription followed by polymerase chain reaction (PCR) amplification with a pair of highly degenerate primers derived from the third and sixth transmembrane regions of the previously cloned adrenergic, D 2dopaminergic, and serotonin receptors. This process resulted in the amplification of several cDNA fragments which were preliminarily characterized by DNA sequence analysis. One of these fragments was found to exhibit considerable sequence homology to previously cloned G protein-coupled
receptors and was used to screen a rat striatal cDNA library in order to isolate a full-length clone. One clone (pB73D,) with an insert of 3.6 kb was isolated and found to strongly hybridize with the 32P-labeled PCR probe on dot-blot analysis. The nucteotide sequence of this cDNA was found to exhibit 90% homology in the region of the PCR fragment, the divergence of which is probably attributable to species differences (mouse vs. rat). The longest open reading frame in this cDNA codes for a 487 residue protein with a theoretical molecular weight of 54,264 Da (Fig. 2). Although the neighboring sequence of the first ATG in this reading frame is similar to Kozak's consensus initiation sequence, the third Met codon at position 46 actually provides a better match (Fig. 2). Additional work will thus be necessary to definitively assign the translational start site for this mRNA. Hydrophobicity analysis of the translated protein reveals seven clusters of 24 hydrophobic residues, predicted to represent transmembrane-
Dopamine 90 spanning domains, connected by three extracellular and three intraceUular loops. This pattern is similar to that observed for other cloned G protein-coupled receptors where the NH, terminus is proposed to be extrac~llular and the COOH terminus projects into the cytoplasm. The NH2 terminus contains one consensus site for N-linked glycosylation while the predicted third cytoplasmic loop exhibits one consensus recognition site for phosphorylation by the cAMPdependent protein kinase (Fig. 2). In addition, the long COOH terminus contains several serine and threonin~ residues possibly representing additional sites for regulatory phosphorylation. Comparison of the deduced amino acid sequence for the pB73D~ eDNA clone with the sequences of various catecholamine receptors indicated that the regions of highest identity appear to occur within the predicted transmembrane spanning domains. Within these regions, the pB73D~ protein exhibits sequence homologies of 44% with the rat D2 dopaminergic receptor; 44%, 43% and 40% with the human betar, beta2-, beta3-adrenergic receptors, respectively; and 43% and 42% with the hamster alpha~a- and human alpha~-adrenergic receptors, respectively. When compared with various serotonin receptors, the transmembrane regions of the pB73D~ protein exhibited homologies of 40% for 5HTI^ and 37% for both 5HT~c and 5HT2 receptors. The NH 2and COOH termini and the extracellular and intracellular loops are significantly more divergent among these receptors. It is interesting to note that within the third putative transmembrane spanning domain ofpB73D~, there is a conserved aspartate residue which is common to all biogenic amine receptors that have been sequenced thus far (Strader et al., 1989). Moreover, the fifth transmembrane spanning domain of pB73D~ also contains two serine residues which are conserved among catecholamine receptors and are critical for the recognition of agonist ligands possessing a catechol group (Strader et al., 1989). These observations would tend to suggest the pB73D~ clone encodes a receptor for an endogenous cateeholamine ligand. In an initial attempt to establish the identity of pB73D[, we analyzed the tissue distribution of its corresponding mRNA by Northern blot and in situ hybridization analyses. Northern blot analysis in various neural tissues reveals a transcript size of 4.1 kb which is predominantly located in the striatum with lesser amounts in the cortex and retina. In contrast, no mRNA is observed in the cerebellum, hippocampus, olfactory bulb, mesencephalon, or pituitary. In situ hybridization analysis also indicates a high abundance
21S
of mRNA in the striatum as well as in the olfactory tubercle. Approximately half of the medium sized neurons in the striatum are identified using this technique. The tissue distribution ofpB73Dl mRNA is remarkably similar to that of the D~ dopamine receptor as demonstrated by receptor binding and autoradiography studies (Creese and Fraser, 1987). To definitively establish the identity of the receptor encoded by the pB73D~ clone, the eDNA insert was subcloned into the pCD-SRa vector for expression in eukaryotic cells. The resulting plasmid, pSRa-D~, was used to transiently transfect COS-7 cells. [3H]SCH23390, a Dcselective radiolabeiled antagonist, binds to transfected COS-7 membranes in a saturable fashion with high specific activity (400 fmol/mg protein) and an affinity (0.3 ± 0.03 nM) in good agreement with that found in the rat striatum. No specific binding was detected in COS-7 cells that had not been transfected with pSRa-Dm or transfected with the pCD-SRa vector alone. The ability of a variety of dopaminergic ligands to compete for specific [3H]SCH-23390 binding to transfected COS-7 cell membranes was examined. (+)-SCH-23390 is the most potent agent (0.2 + 0.01 riM) and is approximately 200-fold more potent than its enantiomer, (-)-SCH-23388 (41 -4- 1.2 nM). The non-selective dopaminergic antagonist (+)-butaclamol also exhibits high affinity (2.8 + 0.2 nM) and is more than 4 orders of magnitude more potent than its inactive isomer, (-)-butaclamol (31 4- 0.8 gM). The D2selective antagonist spiperone exhibits relatively low affinity (290 4- 7 nM) as do the serotonin antagonists, ketanserin (0.42 + 0.031 gM) and mianserin (0.18 ± 0.042/z M). The endogenous agonist, dopamine, is also able to completely inhibit [3H]SCH-23390 binding (0.64 4. 0.092 gM). This rank order of potency as well as the absolute affinities (Ki) of these compounds agree well with those previously demonstrated for striatal Dt receptors. It was also demonstrated that pSRa-D~-transfected COS-7 cells exhibit D~ receptor-mediated stimulation of cAMP production. Dopamine stimulates cAMP production by approximately 2-fold in these transfected cells. In contrast, no response to dopamine is observed in non-transfected cells. The D~ selective agonists (+)SKF-38393 and (+)SKF-82958 also stimulate cAMP accumulation to a similar extent as dopamine. Although SKF-38393 has been reported to be a partial agonist at DI receptors, its fuller efficacy observed here is probably due the presence of spare receptors resulting from over-expression of receptor protein. In addition, the stimulation by SKF-38393 exhibits appropriate stereoselectivity with the
22S
Dopamine 90
( - ) i s o m e r exhibiting a lower potency. Finally, the beta-adrenergic agonist, epinephrine, also exhibits a low potency relative to dopamine as expected for a D~ receptor. In these experiments, the beta-adrenergic antagonist propranolol was included in the assays to preclude stimulation of the endogenous C O S - 7 cell beta-adrenergic receptor. These expression data thus confirm that the c D N A which we have cloned encodes a functional D~ dopamine receptor protein. It is important to emphasize that the D~ receptor which we have presently cloned (Monsma et al., 1990) is one which is functionally coupled to the stimulation of adenylyl cyclase. Recently, unique D, receptors have been described in kidney which stimulate phospholipase C activity independently from the activation of adenylyl cyclase (Felder et al.. 1989). We have also shown, using Xenopus oocyte expression experiments, that rat striatum contains m R N A encoding D~ receptors which can couple to phospholipase C, inositol phosphate production and Ca 2+ mobilization in a cAMP-independent fashion (Mahan et al., t990). It is interesting that the m R N A which codes for this D~ receptor-stimulated phospholipase C response is 2.5 kb in size (Mahan et al., 1990) in comparison with the 4.1 kb D~ receptor m R N A observed here. Moreover, in preliminary experiments, we have found that when m R N A is transcribed from the pB73D~ D~ receptor c D N A clone and injected into Xenopus oocytes, dopamine will stimulate c A M P accumulation 2-fold but is incapable of producing a Ca 2+ mobilization response. These findings suggest that the striatum contains two separate D~ receptor proteins which are coupled to different signal transduction pathways. Consequently, we propose that the D~ receptor subtypes linked to the activation of adenylyl cyclase and phospholipase C be designated D,A and D~B, respectively. Further experimentation involving the cloning and expression of the D m receptor subtype will be required to confirm this hypothesis. REFERENCES
Chio, C.L, Hess, G.F., Graham, R.S., and Huff, R.M. (1990). Second Molecular Form of D2 Dopamine Receptor in Rat and Bovine Caudate Nucleus. Nature
(London) 343, 26(~269. Creese, I., and Fraser, C.M., eds. (1987). Receptor Biochemistry and Methodology: Dopamine Receptors, (Liss, New York) Vol. 8, pp. 1-245. Bunzow, J.R., Van Tol, H.H.M., Grandy, D.K., Albert, P., Salon, J., Christie, M., Machida, C.A., Neve, K.A., and Civelli, O. (1988). Cloning and Expression of a Rat D~ Dopamine Receptor cDNA~ Nature (London) 336, 783 787. Dal Toso, R., Sommer, B., Ewert, M., Herb, A., Pritchett, D.B., Bach, A., Shivers, B.D., and Seeburg, P.H. (1989). The Dopamine D 2 Receptor: Two Molecular Forms Generated by Alternative Splicing. EMBO J. 8. 4025 4034. Felder, R.A., Felder, C.C., Eisner, G.M., and Jose, P.A. (1989). The Dopamine Receptor in Adult and Maturing Kidney. Am. J. Physiol. 257, F315-F327. Giros, B., Sokotoff, P., Martres, M.P., Riou, J.F,, Emorine, L.J., and Schwartz, J.C. (1989). Alternative Splicing Directs the Expression of Two D 2 Dopamine Receptor Isoforms. Nature (London) 342,923-926. Grandy, D.K., Marchionni, M.A., Makam, H., Stofko, R.E., Alfano, M., Frothingham, L., Fischer, J,B., BurkeHowie, K.J., Bunzow, J.R., Server, A.C., and Civelli, O. (1989). Cloning of the cDNA and Gene for a Human D~ Dopamine Receptor. Proc. Natl. Acad. Sci. USA 86, 9762-9766. Mahan, L.C., Butch, R.M., Monsma, F.J., Jr., and Sibley, D.R. (1990) Expression of Striatal D~ Dopamine Receptors Coupled to Inositolphosphate Production and Ca:' Mobilization in Xenopus Oocytes. Proc. Natl. Acad. Sci. USA, 87, 2196-2200. Monsma, F.J., Jr., McVittie, L.D., Gerfen, C.R., Mahan, UC., and Sibley, D.R. (1989a). Multiple D2 Dopamine Receptors Produced by Alternative RNA Splicing. Nature, 342, 926-929. Monsma, F.J., Jr., Brassard, D.L., and Sibley, D.R. (1989b). Identification and Characterization of D~ and D~ Dopamine Receptors in Cultured Neuroblastoma and Retinoblastoma Clonal Cell Lines. Brain Research, 492, 314-324. Monsma, F.J., Jr., Mahan, L.C., McVittie, LD., Gerfen, C.R., and Sibley, D.R. (1990) Molecular Cloning and Expression of a D~ Dopamine Receptor Linked to Adenylyl Cyclase Activation. Proc. Natl. Acad Sci. USA, in press. Selbie, L.A., Hayes, G., and Shine, J. (1989). The major Dopamine D 2 Receptor: Molecular Analysis of the Human D2A Subtype. DNA 8, 683-689. Strader, C.D., Sigal, I.S., and Dixon, R.A.F. (1989). Structural Basis of Beta-Adrenergic Receptor Function. FASEB J. 3, 1825-1832. Vallar, L. and Meldolesi, J. (1989). Mechanisms of Signal Transduction at the Dopamine D, Receptor. Trends Pharmacol. Sci. 10, 74- 77.
