Differential Dependence of the D1 and D5Dopamine Receptors on the G Protein γ7 Subunit for Activation of Adenylylcyclase*

The D1 dopamine receptor, G protein γ7 subunit, and adenylylcyclase are selectively expressed in the striatum, suggesting their potential interaction in a common signaling pathway. To evaluate this possibility, a ribozyme strategy was used to suppress the expression of the G protein γ7 subunit in HEK 293 cells stably expressing the human D1 dopamine receptor. Prior in vitro analysis revealed that the γ7 ribozyme possessed cleavage activity directed exclusively toward the γ7 RNA transcript (Wang, Q., Mullah, B., Hansen, C., Asundi, J., and Robishaw, J. D. (1997)J. Biol. Chem. 272, 26040–26048). In vivoanalysis of cells transfected with the γ7 ribozyme showed a specific reduction in the expression of the γ7 protein. Coincident with the loss of the γ7 protein, there was a noticeable reduction in the expression of the β1 protein, confirming their interaction in these cells. Finally, functional analysis of ribozyme-mediated suppression of the β1 and γ7 proteins revealed a significant attenuation of SKF81297-stimulated adenylylcyclase activity in D1 dopamine receptor-expressing cells. By contrast, ribozyme-mediated suppression of the β1 and γ7 proteins showed no reduction of SKF81297-stimulated adenylylcyclase activity in D5 dopamine receptor-expressing cells. Taken together, these data indicate that the structurally related D1 and D5 dopamine receptor subtypes utilize G proteins composed of distinct βγ subunits to stimulate adenylylcyclase in HEK 293 cells. Underscoring the physiological relevance of these findings, single cell reverse transcriptase-polymerase chain reaction analysis revealed that the D1 dopamine receptor and the G protein γ7 subunit are coordinately expressed in substance P containing neurons in rat striatum, suggesting that the G protein γ7 subunit may be a new target for drugs to selectively alter dopaminergic signaling within the brain.

Dopaminergic signaling in brain is mediated by five receptor subtypes that can be grouped into D 1 -like (D 1 and D 5 ) and D 2 -like (D 2 , D 3 , and D 4 ) classes based on pharmacological and physiological criteria (1). Studies suggest that imbalances between these two opposing classes lead to deficiencies in movement and cognitive performance (2,3). In particular, alterations in the D 1 -like dopamine receptors are implicated in a variety of neurologic and psychiatric disorders, such as Parkinson's disease, Tourette's syndrome, and schizophrenia. Thus, achieving a better understanding of the D 1 -like dopamine receptors and the signaling pathways they activate may suggest more selective therapeutic targets in these diseases.
The D 1 and D 5 dopamine receptors stimulate adenylylcyclase activity through their coupling to heterotrimeric G proteins (1, 4 -6). Because the function of these heterotrimeric G proteins was originally ascribed to the ␣ subunit, most research has focused on determining its identity. Of the several ␣ subunits identified to date, reconstitution studies have shown that coupling of D 1 dopamine receptors to adenylylcyclase can be mediated only by the ␣ s and ␣ olf subunits of the G s subclass (4,5,(7)(8)(9). Although sharing 88% amino acid homology, the ␣s and ␣ olf subunits show very divergent expression patterns, ranging from the ubiquitous expression of the ␣ s subunit to the olfactory and neuron-specific expression of the ␣ olf subunit (10). Immunoprecipitation studies (11) have confirmed the interaction between the ␣ s subunit and the D 1 dopamine receptors in cells, whereas in situ hybridization (12,13) and gene targeting (14) studies have suggested a possible interaction between the ␣ olf subunit and the D 1 dopamine receptor in striatum.
By contrast, little effort has focused on determining the identity of the ␤␥ subunits of the G protein despite mounting evidence for their importance in receptor recognition (15,16). Of particular interest, reconstitution studies (17)(18)(19)(20)(21) and reverse genetic approaches (22,23) have shown that the nature of the ␥ subunit is an important determinant of its interaction with receptor. Consistent with such a role, 12 ␥ subunit genes have been identified that show extensive structural diversity (24). Recently, we used a ribozyme approach to begin to elucidate their functions (23,25,26). This approach identified the ␥ 7 subunit as a specific component of the G protein that couples the ␤-adrenergic receptor, but not the prostaglandin E 1 receptor, to stimulation of adenylylcyclase in HEK 293 cells (23). Although expressed in a variety of tissues and cell types (27,28), the ␥ 7 subunit expression is most abundant in medium spiny neurons in striatum (13). This pattern of expression is shared by the D 1 dopamine receptor and adenylylcyclase (4, 12, 29 -31), raising the possibility that all of these components may be involved in the same signaling pathway. In the present study, we used the ribozyme approach to identify a role for the ␥ 7 subunit as a specific component of the G protein that couples the D 1 dopamine receptor to activation of adenylylcyclase.

MATERIALS AND METHODS
Single Cell Reverse Transcriptase-Polymerase Chain Reaction Analysis-Following dissociation and plating at low density, single neostriatal neurons from 4-week-old rats, cells were aspirated into electrodes by applying negative pressure. The electrode contents (ϳ5 l) were ejected into a thin walled PCR 1 tube (MJ Research, Watertown, MA) containing diethylpyrocarbonate-treated water, RNAsin (Promega, Madison, WI), dithiothreitol, and oligo(dT), as described previously (31). First strand cDNA was generated from mRNA using SuperScript II reverse transcriptase (Life Technologies, Inc.). After reverse transcription, mRNA was eliminated by the addition of 1 l of RNase H (2 units/l) and by heating the PCR tube to 37°C for 20 min. Reverse transcriptase-PCR analyses on the contents of single cells were performed, as described previously (31,32). For detection of enkephalin, substance P, and G protein ␥ 7 subunit, PCR amplifications were performed with a thermal cycler (MJ Research), using 2 l of first strand cDNA template and Taq polymerase. For detection of the D 1 dopamine receptor, two rounds of PCR amplification were necessary. The first round consisted of a 35cycle amplification using 2 l of first strand cDNA template, with one-tenth of the product generated serving as template in the second round of a 35-cycle amplification (31). The primers for enkephalin, substance P, and D 1 dopamine receptor amplify 476-, 513-, and 609base pair products, as described previously (31). The primers for the G protein ␥ 7 subunit (GenBank TM accession number L23219) were made to nucleotides 248 -269 (5Ј-ATGTCAGGTACTAACAACGTCG-3Ј) and nucleotides 429 -450 (5Ј-CTAGAGAATTATGCAAGGCTT-3Ј) to amplify a 202-base pair product. The resulting PCR products were separated by electrophoresis on 5% polyacrylamide gels and visualized by ethidium bromide staining. The gels were digitally imaged using a Bio-Rad gel documentation system.
Generation of HEK 293 Cells Stably Expressing Either the D 1 or D 5 Dopamine Receptor-HEK 293 cells were grown in Dulbecco's modified Eagle's medium (Sigma) containing 2 mM glutamine and 10% fetal calf serum. Clonally derived cell lines stably expressing the D 1 or D 5 dopamine receptor subtype were established by calcium phosphate transfection of the pLXSN plasmid (CLONTECH) containing the corresponding human (33,34) cDNAs followed by selection in medium containing 700 g/ml G418 for 4 weeks and continued maintenance in medium containing 200 (g/ml G418 (Life Technologies, Inc.). Receptor expression was confirmed by radioligand binding of intact cells (35). To determine the level of receptor present on the cell surface, radioligand binding experiments were performed on intact cells stably expressing the D 1 or D 5 dopamine receptor subtype. The cells were transfected with either wild type or mutant ribozyme using Effectene (Qiagen, Valencia, CA) or mock transfected. At 24 h after transfection, the cells were removed from the dish by mild trypsinization, and the amount of radioligand bound to the cells was then determined by a rapid filtration technique. In our experience, mild trypsinization does not affect cell viability or the ability of these receptor subtypes to bind the radioligand (35). This is consistent with the observation that biogenic amine receptors (e.g. adrenergic and dopaminergic receptors) that bind ligands within their transmembrane domains are not as susceptible to trypsinization as are peptide receptors (e.g. glucagon receptor) that utilize extracellular domains (36). In any event, loss of receptors because of mild trypsinization would be expected to be the same in both the D 1 and D 5 dopamine receptor-expressing cell lines because they are handled in the same way and they share high homology within their transmembrane domains that govern ligand binding. For each replicate, 5 ϫ 10 5 cells were incubated at room temperature for 90 min with 0.25-16 nm [ 3 H]SCH23390. Incubations were terminated by the addition of 2.5 ml of ice-cold washing buffer (50 mM Tris-HCl, pH 7.4) and rapid filtration through Whatman GF/B filters. The filters were washed three times with 2.5 ml of ice-cold washing buffer, and the bound radioactivity was measured with a Beckman LS 5000TD liquid scintillation counter. Specific and nonspecific binding was determined at each concentration of antagonist in duplicate. Similar results were obtained in at least two independent sets of transfection experiments. The data were analyzed using the nonlinear curve-fitting software in Origin 6.0 (Microcal Software, Northampton, MA).
Synthesis of G Protein ␥ 7 Ribozyme-A chimeric DNA-RNA hammerhead ribozyme targeted against the G protein ␥ 7 subunit mRNA was chemically synthesized and modified by the addition of two phosphorothioate linkages at the 3Ј-end (23). Two mutant ribozymes were chemically synthesized in the same way with the following exceptions. The first mutant ribozyme (RZm1) substituted two nucleotides in the catalytic core to destroy cleavage activity (26) and scrambled the flanking arms to prevent any possible antisense effect. The second mutant ribozyme (RZm2) retained catalytic activity but scrambled the flanking arms. The cleavage activities of the wild type and mutant ␥ 7 ribozymes were confirmed in a cell-free system, as described previously (23). Briefly, the in vitro synthesized ␥ 7 RNA transcript was labeled by inclusion of 50 Ci of [␣-32 P]CTP (10 mCi/ml, 3000 Ci/mmol; PerkinElmer Life Sciences) in the reaction. To perform the cleavage reaction, the ␥ 7 RNA transcript and either the wild type or mutant ␥ 7 ribozyme (RZm1) were denatured in separate tubes in 50 mM Tris-HCl, pH 8.5, at 85°C for 1 min. The reaction was initiated by mixing the contents of the two tubes together in the presence of 20 mM MgCl 2 . In the reaction, the ␥ 7 RNA transcript was present at 50 nM, whereas the ␥ 7 ribozyme was present at a 20 -100-fold higher concentration. Following incubation at 42°C for 1 h, the reaction was stopped by the addition of 1 volume of stop solution consisting of 95% formamide and 60 mM EDTA, pH 8.0. The resulting cleavage products were analyzed on a 5% polyacrylamide gel containing 7 M urea.
Cellular Analysis of G protein ␥ 7 Ribozyme-For immunoblot analysis, HEK cells stably expressing the D 1 or D 5 dopamine receptor were transfected at ϳ60 -80% confluence with serum-free medium containing 0 or 2 M ␥ 7 ribozyme (wild type or mutant) premixed with 20 g/ml LipofectAMINE (Life Technologies, Inc.). At 5 h post-transfection, the medium was supplemented with 6% fetal bovine serum. Between 5 and 48 h post-transfection, 0 or 0.5 M ␥ 7 ribozyme was added incrementally to a final concentration of 0 or 4 M. At 72 h post-transfection, the cells were harvested. The membrane fractions were prepared from control and ribozyme-treated cells and then extracted overnight in the cold room with 1% sodium cholate (23,25). The concentrations of the cholate-solubilized proteins were determined by Amido Black assay (37). Equal amounts of cholate-solubilized proteins were resolved on 15% SDS-polyacrylamide gels and transferred to Nitro-plus nitrocellulose (0.45-m pore size; Micron Separations Inc.), using a high temperature procedure (38). Following transfer, the nitrocellulose blots were probed with G protein-specific, primary antibodies (23,25,38) and 125 I-labeled goat anti-rabbit F(abЈ) 2 fragment, secondary antibody (1 ϫ 10 5 dpm/ml; PerkinElmer Life Sciences) in high detergent blotto (38). After washing, the blots were subjected to autoradiography by exposure to BioMax MS film (Eastman Kodak Co.), and the intensities of the immunodetectable bands were quantified using a PhosphorImager SI (Molecular Dynamics Co.).
cAMP Analysis-Cells stably expressing the D 1 or D 5 dopamine receptor were plated into 6-well plates and transfected with ␥ 7 ribozyme (wild type or mutant). At 72 h post-transfection, the cells were stimulated with the indicated agonist for 5 min and then harvested for measurement of cAMP accumulation (39). The full D 1 dopamine receptor agonist (Ϯ)-SKF-81297 was used (40).

Characterization of HEK 293 Cell Lines Stably Expressing the D 1 or D 5 Dopamine
Receptor-HEK 293 cells are a useful model for studying D 1 or D 5 dopamine receptors that are linked in a stimulatory manner to adenylylcyclase (11). Accordingly, we generated clonal cell lines stably expressing one or the other receptor subtype. Expression was confirmed by radioligand binding and functional coupling to adenylylcyclase. As shown in Fig. 1A, saturation binding studies performed on intact cells stably expressing the D 1 dopamine receptor revealed that the antagonist [ 3 H]SCH23390 bound with high affinity to a homogenous population of receptor, with a K d value of 2.1 nM and a B max value of 150 Ϯ 8 fmol/5 ϫ 10 5 cells (n ϭ 2). The K d value of the expressed receptor was consistent with that of the native D 1 dopamine receptor. As shown in Fig. 1B, similar studies performed on intact cells stably expressing the D 5 dopamine receptor revealed that the antagonist [ 3 H]SCH23390 also bound with high affinity to a homogenous population of receptor, with a K d value of 3.7 nM and a B max value of 224 Ϯ 77 fmol/5 ϫ 10 5 cells (n ϭ 2). Thus, both receptor subtypes were similarly expressed in the plasma membrane.
To demonstrate functional coupling with adenylylcyclase, control and receptor-expressing cell lines were treated with the selective D 1 dopamine receptor agonist (SKF81297). As shown in Fig. 1C, control cells showed no elevation of cAMP accumulation in response to SKF81297. By contrast, the D 1 and D 5 dopamine receptor-expressing cells displayed comparably large increases in cAMP accumulation following exposure to 1 M SKF81297 for 5 min. Taken together, these results show that both receptor subtypes were similarly expressed in the plasma Effect of ␥ 7 Ribozyme on Cleavage of ␥ 7 RNA in Vitro-To compare the involvement of the G protein ␥ 7 subunit in the dopaminergic receptor signaling pathways, a ribozyme was designed to target the GUC sequence at positions ϩ3 to ϩ5 in relation to the translational start site of the ␥ 7 mRNA. The ␥ 7 ribozyme was chemically synthesized as a DNA-RNA chimera ( Fig. 2A). The flanking sequences, which confer the specific binding to the ␥ 7 mRNA, were composed of deoxyribonucleotides (in capital letters), whereas the catalytic core region, which confers the cleavage activity, was composed of ribonucleotides (in lowercase letters). In addition, two phosphorothioate linkages were added to the 3Ј-end of the ribozyme to enhance the intracellular stability of the ribozyme, as described previously (23). To rule out nonspecific effects, two mutant ␥ 7 ribozymes were also constructed ( Fig. 2A). For the first mutant ␥ 7 ribozyme (RZm1), the flanking sequences were scrambled to prevent hybridization to the ␥ 7 mRNA, thereby eliminating any possible antisense effect, and the catalytic core region included two substitutions (underlined) that abolish ribozyme cleavage activity (26). For the second mutant ␥ 7 ribozyme (RZm2), only the flanking sequences were scrambled to prevent hybridization to the ␥ 7 mRNA.
To assess the cleavage activity in vitro, the wild type or mutant ␥ 7 ribozyme (RZ or RZm1, respectively) was incubated with ␥ 7 RNA transcript at varying ribozyme to template ratios for 1 h. Addition of the mutant ␥ 7 ribozyme did not result in cleavage of the ␥ 7 RNA transcript even at a 100-fold excess of ribozyme to template (Fig. 2B). However, addition of the wild type ␥ 7 ribozyme readily induced cleavage of the ␥ 7 RNA transcript, yielding products of the expected sizes (indicated by arrows). Taken together, these results confirmed that the wild type ␥ 7 ribozyme showed the expected cleavage activity toward the ␥ 7 RNA transcript, whereas the mutant ␥ 7 ribozyme (RZm1) does not.
Effect of ␥ 7 Ribozyme on the G Protein ␥ 7 Protein Level-Having previously established the ability of the wild type ␥ 7 ribozyme to specifically suppress the level of the ␥ 7 mRNA in intact cells (23), the present study confirmed that mRNA suppression was paralleled by a reduction in the amount of ␥ 7 protein. For this purpose, D 1 and D 5 dopamine receptor-expressing cells were transfected with either no ribozyme (CON), wild type ␥ 7 ribozyme (RZ), or mutant ␥ 7 ribozymes (RZm1/2). At 72 h after transfection, membrane fractions were prepared from these cells and evaluated by immunoblotting with various antibodies. Fig. 3A shows a representative immunoblot in which the bottom half of the gel was blotted with antibody (A-67) specific for the G protein ␥ 7 subunit (23,27), and the top half of the gel was blotted with antibodies (B-69 and 584) specific for the G protein ␤ 1 and ␣ s subunits, respectively (41). Of particular interest, the wild type ␥ 7 ribozyme produced a dramatic reduction in the level of the ␥ 7 protein in both the D 1 and D 5 dopamine receptor-expressing cell lines. Along with loss of the ␥ 7 subunit, there was also a significant reduction in the amount of the ␤ 1 protein, in agreement with results of our previous papers showing a functional association between these two proteins in HEK 293 cells (23,25). Fig. 3B shows the quantitation of these results in which the relative amounts of the ␥ 7 , ␤ 1 , and ␣ s proteins were expressed as percentages of their control levels. Based on this analysis, the level of the ␥ 7 protein was markedly suppressed in both the D 1 and D 5 dopamine receptor-expressing cells transfected with the wild type ␥ 7 ribozyme (for D 1 dopamine receptor-expressing cells, 20 Ϯ 6%, n ϭ 7; for D 5 dopamine receptor-expressing cells, 18 Ϯ 6%, n ϭ 4) compared with control cells. By comparison, the level of the ␥ 7 protein was not reduced in cells transfected with the two mutant ␥ 7 ribozymes (for D 1 dopamine receptor-expressing cells, 111 Ϯ 18%, n ϭ 7 for RZm1 and 125 Ϯ 23%, n ϭ 4 for RZm2; for D 5 dopamine receptor-expressing cells, 101 Ϯ 19%, n ϭ 4 for RZm1 and 93 Ϯ 4%, n ϭ 4 for RZm2). These data demonstrated that suppression of the ␥ 7 protein was a specific consequence of the wild type ␥ 7 ribozyme and that the effect was similar in both the D 1 and D 5 dopamine receptor-expressing cells. Moreover, the wild type ␥ 7 ribozyme had no impact on the levels of other G protein subunits, such as the ␣ i and the ␥ 7 -like subunits (data not shown), attesting to the specificity of this effect.
Next, we asked whether loss of the ␥ 7 subunit had any effect on the expression of the associated ␤ and ␣ s subunits that comprise the G s heterotrimer (Fig. 3B). Previously, we showed that expression of the ␤ 1 and ␥ 7 proteins are tightly linked in HEK 293 cells (23,25), consistent with their functional interaction to form a ␤ 1 ␥ 7 dimer. Confirming this finding, we show that the level of the ␤ 1 protein was substantially reduced in both D 1 and D 5 dopamine receptor-expressing cells transfected with the wild type ␥ 7 ribozyme (for D 1 dopamine receptorexpressing cells, 50 Ϯ 5%, n ϭ 4; for D 5 dopamine receptorexpressing cells, 57 Ϯ 8%, n ϭ 4) compared with control cells. Moreover, the level of the ␤ 1 protein was not significantly altered in cells transfected with the two mutant ␥ 7 ribozymes (for D 1 dopamine receptor-expressing cells, 93 Ϯ 11%, n ϭ 4 for RZm1 and 90 Ϯ 10%, n ϭ 4 for RZm2; for D 5 dopamine receptor-expressing cells, 117 Ϯ 12%, n ϭ 4 for RZm1 and 98 Ϯ 17%, n ϭ 4 for RZm2). By comparison, the total level of the ␣ s proteins (sum of the 45-and 52-kDa bands) was only slightly reduced in cells transfected with the wild type ␥ 7 ribozyme (for D 1 dopamine receptor-expressing cells, 82 Ϯ 5%, n ϭ 4; for D 5 dopamine receptor-expressing cells, 65 Ϯ 14%, n ϭ 4) compared with control cells. Taken together, these results indicated that ribozyme-mediated loss of the ␥ 7 protein occurs in concert with suppression of the ␤ 1 protein and, to a lesser extent, the ␣ s protein.
Effect of Ribozyme-mediated Loss of the G Protein ␥ 7 Subunit on the D 1  At 72 h post-transfection, the membrane proteins were extracted with 1% cholate, and equal amounts of membrane proteins (180 g/lane) were resolved on 15% polyacrylamide-SDS gels, transferred to nitrocellulose, and immunoblotted. Following transfer, the nitrocellulose blots were cut along the 30-kDa marker; the bottom halves were probed with the ␥ 7 -specific antibody (A-67), and the top halves were probed with either the ␤ 1 antibody (B-69) or the G protein ␣ s -specific antibody (584). The ␣ s -specific antibody recognizes two alternatively spliced proteins of 45 and 52 kDa. B, quantitation of ribozyme suppression. The intensities of the bands were determined by PhosphorImager analysis. The relative amounts of proteins were expressed as percentages of the control levels. In the case of the ␣ s protein, the sum of the two alternatively spliced species is plotted. The data shown are the means Ϯ S.E. obtained from at least three separate experiments (n ϭ 4 -12).
protein ␣, ␤, and ␥ subunits (19,21). Therefore, ribozymemediated loss of the ␥ 7 subunit would be expected to compromise functional coupling to its associated receptor. Accordingly, we compared the relative dependence of the D 1 and D 5 dopamine receptor signaling pathways on the G protein ␥ 7 subunit. For this purpose, cells were transfected with either no ribozyme (CON), wild type ␥ 7 ribozyme (RZ), or mutant ␥ 7 ribozymes (RZm1/2). At 72 h after transfection, the cells were incubated with various agonists for 5 min, and then agoniststimulated cAMP accumulation was quantitated and expressed as a percentage of the control value. As shown in Fig. 4A, the D 1 dopamine receptor-expressing cells transfected with the wild type ␥ 7 ribozyme showed a significant reduction in SKF81297induced cAMP accumulation (68 Ϯ 4%, n ϭ 22) compared with control cells. Attesting to the specificity of ribozyme action, cells transfected with the two mutant ␥ 7 ribozymes showed no significant attenuation in SKF81297-stimulated cAMP accumulation (108 Ϯ 14%, n ϭ 6 for RZm1 and 95 Ϯ 4%, n ϭ 6 for RZm2). This contrasts with results obtained for the D 5 dopamine receptor-expressing cells. As shown in Fig. 4B, the D 5 dopamine receptor-expressing cells transfected with wild type ␥ 7 ribozyme showed no significant reduction in SKF81297induced cAMP accumulation (107 Ϯ 8%, n ϭ 16). Taken together, these data demonstrated that the D 1 dopamine receptor has a greater dependence on the G protein ␥ 7 subunit for activation of adenylylcyclase activity than does the D 5 dopamine receptor. That this is a real difference and reflects ribozyme-mediated loss of the G protein ␥ 7 subunit and its associated subunits was shown by the response of these two cell lines to a ␤-adrenergic receptor agonist. Consistent with our previous results showing that endogenous ␤-adrenergic receptor activation of adenylylcyclase activity is dependent on the G protein ␤ 1 and ␥ 7 subunits (23), both the D 1 and D 5 dopamine receptor-expressing cells transfected with the wild type ␥ 7 ribozyme showed similarly dramatic reductions in isoproterenolinduced cAMP accumulation (75 Ϯ 2%, n ϭ 6 and 59 Ϯ 9%, n ϭ 6, respectively) compared with control cells. Likewise, the finding that the D 1 dopamine receptor-expressing cells transfected with the wild type ␥ 7 ribozyme showed no attenuation in prostaglandin E 1 -stimulated cAMP accumulation compared with control cells (108 Ϯ 8%, n ϭ 9) indicates that there is no inherent defect in adenylylcyclase activity. Taken together these results identified the ␤ 1 and ␥ 7 subunits as components of the G s protein that couples the D 1 dopamine receptor, but not the D 5 dopamine receptor, to stimulation of adenylylcyclase activity.
Co-expression of the D 1 Dopamine Receptor and G Protein ␥ 7 Subunit in Rat Striatum-To extend these results to a more physiologic setting, we next performed single-cell reverse transcriptase-PCR analysis to localize these signaling components within neurochemically distinct neurons of the rat striatum (31). In this regard, studies have shown that neurons containing substance P (SP) and projecting to the substantia nigra express the D 1 dopamine receptor (12,29), whereas neurons containing enkephalin (ENK) and projecting to the globus pal- lidus express the D 2 dopamine receptor (42) . Fig. 5 shows the results of this analysis for single neurons representative of each type. Of the 10 neurons analyzed, three neurons resulted in amplification of PCR product for SP but not for ENK from first strand cDNA (for example, Neuron 1). Those neurons expressing SP also generated PCR products for the D 1 dopamine receptor and the G protein ␥ 7 subunit. By contrast, two neurons resulted in detection of PCR products for ENK but not for SP (for example, Neuron 10). Those neurons expressing ENK did not generate PCR products for the D 1 dopamine receptor or the G protein ␥ 7 subunit. One neuron expressing both SP and ENK resulted in amplification of PCR product for the G protein ␥ 7 subunit but not for the D 1 dopamine receptor, whereas the remaining four neurons did not result in amplification of any of these PCR products (data not shown). Although only a small number of neurons were sampled, this analysis indicated that a subset of neurons co-express the D 1 dopamine receptor and the G protein ␥ 7 subunit, consistent with their involvement in a common signaling pathway. DISCUSSION A growing body of evidence suggests that the G protein ␤␥ dimer composition is an important determinant for receptor recognition (15)(16)(17)(18)(19)(20)(21)(22)(23). In the present study, a ribozyme approach is used to provide further support for this hypothesis by showing that the closely related D 1 and D 5 dopamine receptors utilize distinct G s proteins that vary in their ␤␥ subunit composition.
Ribozyme Suppression of the G Protein ␥ 7 Subunit-Ribozymes are powerful tools for suppressing gene expression and studying the functional consequences thereof. In the present study, ribozymes directed against the ␥ 7 mRNA (23) were introduced into HEK 293 cell lines stably expressing either the D 1 or the D 5 dopamine receptor. After 72 h, the level of the ␥ 7 protein was reduced to a similar extent in both cell lines transfected with the wild type ␥ 7 ribozyme (Fig. 3). Attesting to the specificity of ribozyme action, the amount of the ␥ 7 protein was not altered in both cell lines transfected with the mutant ␥ 7 ribozymes. Associated with the ribozyme-mediated loss of the ␥ 7 protein, the level of the ␤ 1 protein was also reduced in both cell lines (Fig. 3). As shown previously, this occurs when there is not a sufficient amount of the ␥ 7 protein to dimerize with the ␤ 1 protein to prevent its degradation (23,25). The expression of the ␣ s proteins was only slightly reduced (Fig. 3). Presumably, this reflects a lesser dependence of the ␣ s proteins on the presence of the ␤ 1 ␥ 7 subunit complex for stabilization. Although not directly examined, this is consistent with studies showing the ␣ subunits contain their own membrane targeting signals (43). Taken together, these results demonstrate the ability of the wild type ␥ 7 ribozyme to suppress the G protein ␥ 7 , ␤ 1 , and ␣ s proteins in that order.
Dependence of the D 1 Dopamine Receptor Subtype on the G Protein ␥ 7 Subunit for Activation of Adenylylcyclase Activity-The D 1 dopamine receptor, the G protein ␥ 7 subunit, and adenylylcyclase are highly expressed in striatum (4,12,13,29,30,31), suggesting that these components may be involved in the same signaling pathway. This possibility is supported by the finding that the D 1 dopamine receptor-expressing cells transfected with the wild type ␥ 7 ribozyme showed a 32% reduction in SKF21897-induced cAMP accumulation (Fig. 4A). Because the reduced cAMP response was not observed in cells transfected with the two mutant ␥ 7 ribozymes, this strongly argues that the attenuated response was due to the specific loss of the ␥ 7 subunit and its binding partners rather than a nonspecific effect of the ␥ 7 ribozyme. Other possible nonspecific effects of the wild type ␥ 7 ribozyme were also ruled out. Thus, the wild type ␥ 7 ribozyme did not reduce the D 1 dopamine receptor content (150 Ϯ 8 fmol/5 ϫ 10 5 cells for control cells, and 205 Ϯ 77 fmol/5 ϫ 10 5 for cells transfected with wild type ␥ 7 ribozyme), indicating that there was no defect in the receptor upstream of the G protein. Moreover, the ribozyme did not alter the cAMP response to the prostaglandin receptor agonist, prostaglandin E 1 (108 Ϯ 8% of control), confirming that there was no defect in the effector downstream of the G protein (Fig. 4A). Taken together, these results indicate that the G protein ␥ 7 subunit participates in signal transduction between the D 1 dopamine receptor and adenylylcyclase in this model system. Moreover, because loss of the ␥ 7 subunit was associated with a similar reduction in ␤ 1 subunit, a participatory role of the ␤ 1 subunit in this same pathway is suggested.
Although the present study does not directly address the role of the G protein ␤ 1 and ␥ 7 subunits, two possible mechanisms are suggested. First, it is known from previous studies that the ␤␥ dimer itself can regulate certain types of adenylylcyclases (44). In this scenario, the ribozyme-mediated loss of the ␤ 1 ␥ 7 dimer might compromise the regulation of adenylylcyclase activity, thereby accounting for the attenuated cAMP response. However, the finding that HEK 293 cells do not express ␤␥regulated form of adenylylcyclase appears to argue against this scenario (44). 2 Second, it is known from previous studies that a receptor requires the combined interaction of the G protein ␣␤␥ heterotrimer (19,21). In this scenario, the ribozyme-mediated loss of the ␤ 1 ␥ 7 dimer would lead to reduced assembly of the required G ␣ s ␤ 1 ␥ 7 heterotrimer. At this point, we favor this scenario. Because the G ␣ s ␤ 1 ␥ 7 heterotrimer is required for activation by the D 1 dopamine receptor, the net effect would be to disrupt activation of adenylylcyclase activity. The question then arises as to why D 1 dopamine receptor activation of adenylylcyclase was only partially attenuated. An obvious explanation relates to the transient transfection nature of these studies. Because only 70% of the cells were transfected with the ␥ 7 ribozyme (23), the remaining 30% of the cells that were not transfected would be expected to show a normal cAMP response to SKF81297.
Lack of Dependence of the D 5 Dopamine Receptor Subtype on the G Protein ␥ 7 Subunit for Activation of Adenylylcyclase Activity-Both D 1 and D 5 dopamine receptor-expressing cell lines contained a similar number of receptors in the plasma membrane and showed comparable levels of SKF81297-stimulated cAMP accumulation (compare Fig. 1). Nevertheless, the D 5 dopamine receptor-expressing cell line transfected with the wild type ␥ 7 ribozyme showed no attenuation of SKF21897induced cAMP accumulation compared with the 32% reduction observed in the D 1 dopamine receptor-expressing cell line (Fig.  4). That this difference is real is shown by the similar reduction in the levels of the ␥ 7 , ␤ 1 , and ␣ s proteins in both cell lines (Fig.  3) and by the the comparable attenuation in the levels of isoproterenol-stimulated cAMP accumulation (Fig. 4). Taken together, these results are most consistent with the notion that the D 1 and D 5 dopamine receptor subtypes recruit different G s heterotrimers to mediate this response. Furthermore, these findings suggest that the D 1 dopamine receptor interacts with a form of G s containing the ␤ 1 and ␥ 7 subunits, whereas the D 5 dopamine receptor interacts with a form of G s containing a different combination of ␤ and ␥ subunits.
Implications for Specificity of the D 1 and D 5 Dopamine Receptor Signaling Pathways in Vivo-The results obtained from this heterologous expression system are intriguing because they point to intrinsic differences between the two receptor subtypes in terms of their G protein coupling properties. Future structure-function mapping of the two receptor subtypes 2 J. D. Robishaw, unpublished observations. should allow the molecular basis for this to be identified. In this regard, fluorescence energy transfer (15) and cross-linking (16) studies have identified direct interactions between receptor and the G protein ␤␥ dimer. Although cross-linking studies have yet to identify contact site(s) between the receptor and the ␥ subunit, numerous reconstitution studies have underscored the importance of the carboxyl-terminal region of the ␥ subunit in receptor recognition, with both primary structure and type of prenyl group contributing to the specificity (17,18,21). Thus, these results lay the groundwork for more precisely identifying the molecular determinants specifying the differential interaction of the D 1 and D 5 dopamine receptors with the ␥ 7 subunit and, ultimately, in designing tools to selectively disrupt their individual signaling pathways.
On the other hand, we recognize that results obtained from a heterologous expression system do not constitute proof that the D 1 dopamine receptor subtype interacts with a form of G s containing the ␥ 7 subunit in a native system. However, in support of this possibility, we used reverse transcriptase-PCR analysis of individual, striatal neurons (31,32) to show that the D 1 dopamine receptor and G protein ␥ 7 subunit are predominantly and coordinately expressed in SP containing neurons that project to the substantia nigra par reticulata. Intriguingly, the Gng7 locus encoding the G protein ␥ 7 subunit has been localized to mouse chromosome 10 in the vicinity of several neurological mutations, such as jittery, Ames Waltzer, and mocha (13). Thus, on the basis of these findings, the G protein ␥ 7 subunit should be considered as a candidate for these and other genetic disorders that map near this region in the future.