Adenosine A 1 Receptor-mediated Modulation of Dopamine D 1 Receptors in Stably Cotransfected Fibroblast Cells*

, The antagonistic interactions between adenosine A 1 and dopamine D 1 receptors were studied in a mouse Ltk 2 cell line stably cotransfected with human adenosine A 1 receptor and dopamine D 1 receptor cDNAs. In membrane preparations, both the adenosine A 1 receptor agonist N 6 -cyclopentyladenosine and the GTP analogue guanyl-5 * -yl imidodiphospate induced a decrease in the proportion of dopamine D 1 receptors in a high affinity state. In the cotransfected cells, the adenosine A 1 agonist induced a concentration-dependent inhibition of dopamine-induced cAMP accumulation. Blockade of adenosine A 1 receptor signal transduction with the adenosine A 1 receptor antagonist 1,3-dipropyl-8-cyclo- pentylxanthine or with pertussis toxin pretreatment increased both basal and dopamine-stimulated cAMP levels, indicating the existence of tonic adenosine A 1 receptor activation. Pretreatment with pertussis toxin also counteracted the effects of low concentrations of the A 1 agonist on D 1 receptor-agonist binding. The results suggest that adenosine A 1 receptors antagonistically modulate dopamine D 1 receptors at the level of receptor binding and the generation of second messengers. characteristics of units/ml). The homogenate was centrifuged at 3000 rpm for 10 min at 4 °C; the precipitated nucleic fraction was discarded; and the supernatant was incubated for 30 min at 37 °C (to activate adenosine deaminase and to remove endogenous adenosine) and centrifuged at 20,000 rpm for 40 min at 4 °C. The membrane pellet was then resuspended by sonication in the incubation buffer without adenosine deami- nase (final protein concentration of ; 0.2 mg/ml). In the experiments with [ 3 H]SCH 23390 (NEN Life Science Products), the incubation buffer was 50 m M Tris-HCl (pH 7.4) containing 120 m M NaCl, 5 m M KCl, 2 m M CaCl 2 , and 1 m M MgCl 2 . In the experiments with 1,3-[ 3 H]dipropyl-8- cyclopentylxanthine ([ 3 H]DPCPX; NEN Life Science Products), the incubation buffer was 50 m M Tris-HCl (pH 7.4) containing 2 m M MgCl 2 . Pertussis Toxin Pretreatment and [ 32 P]ADP-ribosylation of Mem-branes— Experiments with pertussis toxin (PTX) were performed with A 1 D 1 cells exposed to PTX (200 ng/ml) for 4 h before membrane prep- aration for radioligand binding experiments or cAMP accumulation experiments. The effectiveness of PTX pretreatment and the conse- quent inactivation of G i proteins were evaluated in membrane preparations by subsequent [ 32 P]ADP-ribosylation in the presence of [ 32 P]NAD and PTX (12). PTX was preactivated in 50 m M 1,4-dithiothre-itol for 1 h at 25 °C. Thereaction mixture (100 m l) contained 0.5–1 mg/ml protein, 1 m M ATP, 10 m M thymidine, 250 m M GTP, 1 m M [ 32 P]NAD (0.5–1 Ci/mmol), and 20 m g/ml PTX a 50 m M Tris-HCl (pH At end of a 45-min incubation period at 30 °C, the reaction was stopped by adding 10 m l of 1% sodium deoxycholate followed by 12 m l of 4 M perchloric acid. The samples were kept on ice for 20 min and centrifuged for 5 min at 20,000 rpm. The pellet was neu- tralized with 1 M NaOH; 30 m l of Laemmli buffer was added; and the proteins were separated by SDS-polyacrylamide gel electrophoresis with 10% acrylamide. Autoradiography was performed by exposure of dried gels to Fuji RXNIF film for 2–4 days. Radioligand Binding Experiments— Saturation experiments with the D 1 receptor antagonist [ 3 H]SCH 23390 were carried out with 10 concentrations (0.2–6.0 n M ) of [ 3 H]SCH 23390 (70.3 Ci/mmol) in the presence or absence of the A 1 receptor agonist N 6 -cyclopentyladenosine (CPA; 10 n M ) by incubation for 15 min at 37 °C. Nonspecific binding is defined as the binding in the presence of 100 m M dopamine. Saturation experiments with the A 1 receptor antagonist [ 3 H]DPCPX were carried out with 10 concentrations (0.6–27.7 n M ) of [ 3 H]DPCPX (120.0 Ci/mmol) by incubation for 2 h at room temperature. Nonspecific binding is defined as that occurring in the presence of the adenosine A 1 receptor agonist N 6 -cyclohexyladenosine (40 m M ). Competition experiments of dopamine (1 n M to 10 m M ) versus the dopamine D 1 antagonist [ 3 H]SCH 23390 ( ; 2 n M ) were performed by incubation for 15 min at 37 °C in the presence or absence of CPA or the nonhydrolyzable GTP analogue Gpp(NH)p. Competition experiments of CPA (10 p M to 10 m M ) versus [ 3 H]DPCPX ( ; 1 n M ) were by terminated with 50 m M perchloric acid to a final concentration of 0.1 M after a 10-min incubation at 37 °C. Samples were neutralized with 60 m l of KOH, and the cAMP content in the supernatants was determined with a protein binding assay (13). The following experiments were performed: first, the effect of different concentrations of CPA (0.3–300 n M ) on the cAMP accumulation induced by forskolin (30 m M ) and dopamine (0.1 m M ); second, the effect of CPA (30 n M ) and DPCPX (30 n M ) on the cAMP accumulation induced by different concentrations of dopamine (10 n M to 30 m M ); and third, the effect of different concentrations of CPA (0.01–10 m M ) on the cAMP accumulation induced by dopamine (10 m M ) after PTX pretreatment (200 ng/ml for 4 h). Student’s t test and one-way and bifactorial ANOVA (followed by post hoc protected least square difference method) were used for statistical analysis.

The antagonistic interactions between adenosine A 1 and dopamine D 1 receptors were studied in a mouse Ltk ؊ cell line stably cotransfected with human adenosine A 1 receptor and dopamine D 1 receptor cDNAs. In membrane preparations, both the adenosine A 1 receptor agonist N 6 -cyclopentyladenosine and the GTP analogue guanyl-5-yl imidodiphospate induced a decrease in the proportion of dopamine D 1

receptors in a high affinity state. In the cotransfected cells, the adenosine A 1 agonist induced a concentration-dependent inhibition of dopamine-induced cAMP accumulation. Blockade of adenosine A 1 receptor signal transduction with the adenosine A 1 receptor antagonist 1,3-dipropyl-8-cyclopentylxanthine or with pertussis toxin pretreatment increased both basal and dopamine-stimulated cAMP levels, indicating the existence of tonic adenosine A 1 receptor activation. Pretreatment with pertussis toxin also counteracted the effects of low concentrations of the A 1 agonist on D 1 receptor-agonist binding. The results suggest that adenosine A 1 receptors antagonistically modulate dopamine D 1 receptors at the level of receptor binding and the generation of second messengers.
It has been shown that the binding characteristics of one type of G protein-coupled receptor can be altered by the stimulation of another type of G protein-coupled receptor in crude membrane preparations (1). Such intramembrane interactions have been postulated to represent direct interactions between the receptor molecules and/or to involve G proteins or other mobile molecules associated with the membrane (1). There is increasing evidence suggesting that antagonistic intramembrane interactions between specific subtypes of adenosine and dopamine receptors constitute an important integrative mechanism in the basal ganglia (2,3). Adenosine A 1 and A 2A receptors antagonistically and specifically modulate the binding characteristics of dopamine D 1 and D 2 receptors, respectively (2,3). In membrane preparations from rat striatum, the stimulation of A 2A receptors decreases the affinity of D 2 receptors for agonists (4). On the other hand, the stimulation of A 1 receptors was shown to decrease the proportion of D 1 receptors in the high affinity state, without modifying the dissociation constants of high and low affinity D 1 agonist-binding sites (5).
Thus, the A 1 receptor agonist had the same effect as that induced by the GTP analogue Gpp(NH)p. 1 It was hypothesized that A 1 receptor stimulation might uncouple the striatal D 1 receptor from the G protein (5). There is evidence that the antagonistic A 2A -D 2 and A 1 -D 1 intramembrane interactions are involved in the motor depressant effects of adenosine receptor agonists and the motor stimulant effects of adenosine receptor antagonists, such as caffeine (2)(3)(4)(5).
The same changes in the binding characteristics of striatal D 2 receptors after A 2A receptor stimulation have been obtained in membrane preparations from a mouse fibroblast cell line (Ltk Ϫ ) stably cotransfected with the dog A 2A receptor and human D 2 (long-form) receptor cDNAs (6). In these transfection studies, it was also found that activation of adenylyl cyclase was not involved in the intramembrane A 2A -D 2 interaction (6). Altogether, these results showed that stably cotransfected cell lines constitute a valuable model to study the mechanistic aspects involved in the intramembrane receptor-receptor interactions. In the present work, this methodology has been applied to study the antagonistic interaction between A 1 and D 1 receptors. The first aim of the study was to demonstrate the existence of an antagonistic A 1 -D 1 intramembrane interaction in mammalian cells stably cotransfected with A 1 receptor and D 1 receptor cDNAs. The second aim was to demonstrate the existence of a functional antagonistic interaction between A 1 and D 1 receptors in the cotransfected cells by means of cAMP accumulation experiments. Finally, the third aim of the study was to find a functional significance of the antagonistic A 1 -D 1 intramembrane interaction.

EXPERIMENTAL PROCEDURES
Transfection and Maintenance of Fibroblast Ltk Ϫ Cells-Cells from the mouse fibroblast Ltk Ϫ cell line previously transfected with the human D 1 receptor cDNA (7) were used. The expression vector pZEM-3 (8) containing the full coding sequence of the human D 1 receptor in front of mouse metallothionein promoter I had been cotransfected with the plasmid pRSV-neo, which confers resistance to neomycin and Geneticin (G418). Metallothionein promoter I allows transcriptional induction by including zinc sulfate in the cell culture (8). Nevertheless, a clone expressing a relatively high level of D 1 receptor mRNA and protein was obtained (D 1 cells) without zinc-mediated induction (7). The D 1 cells were cotransfected with the human adenosine A 1 receptor cDNA (A 1 D 1 cells). The expression vector pcDNA3 (Invitrogen) containing the full coding sequence of the human A 1 receptor (gift from M. Lohse) (9, 10) in front of enhancer-promoter sequences from the immediate-early gene of the human cytomegalovirus was cotransfected with a hygromycin resistance plasmid (pHyg; gift from G. Vassart) with the calcium phosphate precipitation method as described in detail (6). The expression of the A 1 receptor was verified by Northern blot and radioligand binding techniques (see below), and a clone expressing similar levels of A 1 and D 1 antagonist-binding sites (A 1 D 1 cells) was chosen for further experiments. A 1 D 1 cells were cultured routinely at 37°C with 5% CO 2 in Dulbecco's minimal essential medium with 4.5 mg/ml glucose and 0.11 mg/ml sodium pyruvate supplemented with 10% fetal calf serum, 2 mM glutamine, 100 units/ml penicillin, 100 units/ml streptomycin, 200 g/ml G418, and 300 g/ml hygromycin in plastic Petri dishes. D 1 cells were cultured as described for A 1 D 1 cells, but without hygromycin. The splitting of cell cultures was performed by replacing the medium with a modified Puck's saline containing trypsin (0.5 mg/ml) and EDTA (0.2 mg/ml). The experiments were performed at a cell confluence of ϳ80%.
Analysis of RNA-Isolation of RNA from the Ltk Ϫ cells was carried out according to the method of Chomcyznski and Sacchi (11). For preparation of the Northern blots, 20 g of total RNA/lane was denatured in a 2.1 M formaldehyde and 50% formamide solution by heating for 2 min at 95°C, separated by electrophoresis on a 2.2 M formaldehyde and 1.0% agarose gel, and transferred to a nitrocellulose membrane. Blots were hybridized with 32 P-labeled adenosine A 1 receptor cDNA by nick translation. Following hybridization, the membrane was washed and exposed to Kodak XAR-5 film with an intensifying screen at Ϫ70°C. The optical density of the bands was measured by computerassisted densitometric analysis (IBAS image analyzer).
Membrane Preparation-The D 1 and A 1 D 1 cells were lifted from Petri dishes with a cell scraper. Harvested cells were washed twice with ice-cold phosphate-buffered saline and centrifuged at 2000 rpm for 5 min at 4°C. The cell pellet was sonicated (30 s) and resuspended in the incubation buffer containing adenosine deaminase (Boehringer Mannheim; 10 units/ml). The homogenate was centrifuged at 3000 rpm for 10 min at 4°C; the precipitated nucleic fraction was discarded; and the supernatant was incubated for 30 min at 37°C (to activate adenosine deaminase and to remove endogenous adenosine) and centrifuged at 20,000 rpm for 40 min at 4°C. The membrane pellet was then resuspended by sonication in the incubation buffer without adenosine deaminase (final protein concentration of ϳ0.2 mg/ml Pertussis Toxin Pretreatment and [ 32 P]ADP-ribosylation of Membranes-Experiments with pertussis toxin (PTX) were performed with A 1 D 1 cells exposed to PTX (200 ng/ml) for 4 h before membrane preparation for radioligand binding experiments or cAMP accumulation experiments. The effectiveness of PTX pretreatment and the consequent inactivation of G i proteins were evaluated in membrane preparations by subsequent [ 32 P]ADP-ribosylation in the presence of [ 32 P]NAD and PTX (12). PTX was preactivated in 50 mM 1,4-dithiothreitol for 1 h at 25°C. The reaction mixture (100 l) contained 0.5-1 mg/ml protein, 1 mM ATP, 10 mM thymidine, 250 M GTP, 1 M [ 32 P]NAD (0.5-1 Ci/mmol), and 20 g/ml PTX in a buffer containing 50 mM Tris-HCl (pH 7.4). At the end of a 45-min incubation period at 30°C, the reaction was stopped by adding 10 l of 1% sodium deoxycholate followed by 12 l of 4 M perchloric acid. The samples were kept on ice for 20 min and centrifuged for 5 min at 20,000 rpm. The pellet was neutralized with 1 M NaOH; 30 l of Laemmli buffer was added; and the proteins were separated by SDS-polyacrylamide gel electrophoresis with 10% acrylamide. Autoradiography was performed by exposure of dried gels to Fuji RXNIF film for 2-4 days.
Radioligand . The radioactivity content of the filters was detected by liquid scintillation spectrometry. To avoid the variability of the binding pa-rameters associated with the assay conditions, such as cell confluency (ϳ80%) and number of passages (up to 10), the same membrane preparation was used to study the effect of different drugs on the binding characteristics of dopamine D 1 receptors, and each experiment was independently analyzed. Data from saturation experiments were analyzed by nonlinear regression analysis (GraphPad) for the determination of dissociation constants (K D ) and the number of receptors (B max ). Data from competition experiments were also analyzed by nonlinear regression analysis, and the fitting for either one or two binding sites was statistically compared (F test). For a two-binding site fit, the dissociation constants for the high (K H ) and low affinity (K L ) binding sites and for the proportion of binding sites in the high affinity state (R H ) were determined. For a one-binding site fit, the concentration of agonist that displaced 50% of the labeled antagonist (IC 50 ) was determined. The amount of nonspecific binding was calculated by extrapolation of the displacement curve. Protein determinations were performed using bovine serum albumin as a standard. The Kruskal-Wallis test and Mann-Whitney's U test were used to analyze differences in K D , B max , K H , and K L values.
cAMP Accumulation Experiments-After scraping the cells off the culture plates, they were washed twice with phosphate-buffered saline and resuspended in serum-free medium at a concentration of 0.5-1.

Northern Blotting and Saturation Experiments with the Adenosine A 1 Antagonist [ 3 H]DPCPX-Different levels of A 1
receptor mRNA were obtained with the Northern blot analysis of the different hygromycin-resistant clones. A 1 receptor mRNA was not detected in cells not transfected with A 1 receptor cDNA (D 1 cells) (Fig. 1). A significant correlation (linear regression analysis, r 2 ϭ 0.93 and p ϭ 0.0001) was found between the relative A 1 receptor mRNA content (optical density) obtained from the Northern blot analysis and the A 1 receptor density (log B max values) obtained from the saturation experiments with [ 3 H]DPCPX in membrane preparations from the different clones (Fig. 1). The nonspecific binding was Ͻ5% of the total binding. The D 1 cells did not show any significant  (Fig. 2). The determined K D value is in close agreement with the values previously reported for membrane preparations from Chinese hamster ovary cells and Escherichia coli cells expressing human A 1 receptors (9,10,14). For comparison, the B max and K D values for [ 3 H]DPCPX binding in membrane preparations from rat striatum have been reported to be ϳ1 pmol/mg of protein and 1 nM, respectively (15).  (Fig. 3).

Competition Experiments of Dopamine Versus the Dopamine D 1 Antagonist [ 3 H]SCH 23390 -Competition experiments of dopamine versus the dopamine D 1 antagonist [ 3 H]SCH 23390
in membrane preparations from both D 1 and A 1 D 1 cells showed a significantly better fit for two binding sites than for one binding site (F test, p Ͻ 0.05). Similar K H and K L values were obtained in membrane preparations from D 1 and A 1 D 1 cells, and the proportion of D 1 receptors in the high affinity state (R H values) was ϳ10% in both cases ( Fig. 4 and Table I). In the presence of Gpp(NH)p (100 M), a significantly better fit for one binding site (R H ϭ 0) was obtained in most of the membrane preparations from either D 1 or A 1 D 1 cells, with IC 50 values very similar to the K L values obtained in the absence of Gpp(NH)p. The same effect as that induced by Gpp(NH)p was obtained in the presence of the A 1 agonist CPA (1-100 nM) in membrane preparations from A 1 D 1 cells. On the other hand, CPA (0.1 and 10 M) was ineffective in membrane preparations from D 1 cells ( Fig. 4 and Table I). Pretreatment of the A 1 D 1 cells with PTX counteracted the effect of a low concentration of CPA (10 nM), but it did not counteract the effect of 10 M CPA or 100 M Gpp(NH)p ( Fig. 5 and Table I). The degree of PTX-induced [ 32 P]ADP-ribosylation was markedly reduced in membrane preparations from PTX-pretreated A 1 D 1 cells compared with nonpretreated cells (Fig. 6).
cAMP Accumulation Experiments-In A 1 D 1 cells, but not in D 1 cells (data not shown), CPA induced a significant concentration-dependent inhibition of cAMP accumulation induced by 30 M forskolin and 0.1 M dopamine (one-way ANOVA, p Ͻ 0.001 in both cases), with IC 50 values (95% confidence intervals in parentheses) of 0.9 (0.2-2.8) nM and 0.8 (0.3-1.8) nM, respectively (Fig. 7). In A 1 D 1 cells, dopamine induced a significant concentration-dependent increase in cAMP accumulation (Fig.  8). The effect of dopamine was significantly antagonized by CPA (30 nM) (bifactorial ANOVA, p Ͻ 0.001 for the factors dopamine and CPA) (Fig. 8). The EC 50 values (95% confidence pmol/50 l, respectively. Since they were independently analyzed, the basal levels of cAMP (dopamine concentration ϭ 0) were not included in the ANOVA. cAMP basal levels were significantly increased by DPCPX (Student's t test, p Ͻ 0.05), and they were not modified by CPA (Fig. 8). In PTX-pretreated A 1 D 1 cells, the basal levels of cAMP were significantly higher than in the control experiment, without PTX pretreatment (0.41 Ϯ 0.01 and 0.23 Ϯ 0.01 pmol/50 l (means Ϯ S.E.), respectively). CPA also induced a significant concentration-dependent inhibition of cAMP accumulation induced by a high concentration of dopamine (10 M), and the cAMP accumulation induced by dopamine (10 M) was significantly higher in PTX-pretreated cells (bifactorial ANOVA, p Ͻ 0.001 for the factors CPA and PTX) (Fig. 9). However, CPA was more effective in control cells. In PTX-pretreated cells, only a high con-centration of CPA (10 M) significantly antagonized cAMP accumulation induced by dopamine (10 M) (post hoc one-way ANOVA, p Ͻ 0.05) (Fig. 9). DISCUSSION A mouse fibroblast Ltk Ϫ cell line stably cotransfected with human A 1 and D 1 receptor cDNAs was obtained, and a clone containing similar amounts of both receptors (ϳ4 pmol/mg of protein) was chosen (A 1 D 1 cells). Both receptors were shown to be functional in cAMP accumulation experiments. It has been previously shown that dopamine induces a concentration-dependent increase in cAMP accumulation in Ltk Ϫ cells transfected with the human D 1 receptor cDNA (D 1 cells), but not in nontransfected cells (7). It was also shown that the effect of dopamine is mediated by D 1 receptors since it was selectively counteracted by a D 1 but not a D 2 receptor antagonist (7). In the present experiments, dopamine-induced cAMP accumulation was also demonstrated in the A 1 D 1 cells, with an EC 50 value very similar to the K H value obtained in the [ 3 H]SCH 23390-dopamine competition experiments (ϳ0.3 M). Furthermore, in A 1 D 1 cells, the A 1 agonist CPA counteracted the cAMP accumulation induced by dopamine or forskolin, with an IC 50 value similar to the K H value shown in [ 3 H]DPCPX-CPA competition experiments (1 nM range). In addition, the A 1 antagonist DPCPX was found to significantly increase the basal levels of cAMP and to potentiate dopamine-induced cAMP accumulation. Altogether, these results show the existence of a functional antagonistic interaction between A 1 and D 1 receptors in A 1 D 1 cells. Nonlinear regression analysis indicated that the CPA-and DPCPX-mediated effects on dopamine-induced cAMP accumulation were mainly due to changes in the maximal stimulation without changes in EC 50 . This suggests that, in agreement with the results obtained from radioligand binding experiments, the A 1 receptor-mediated modulation of D 1 receptors does not involve changes in the affinity of D 1 receptors for agonists. This is in contrast to the A 2A -D 2 interaction (see the Introduction), where A 2A receptor stimulation induces a decrease in the affinity of D 2 receptors for agonists (4,6). The effects of the A 1 antagonist also suggest that adenosine released by these cells exerts a tonic inhibition of D 1 receptormediated function through the A 1 -D 1 interaction.
The radioligand binding experiments carried out with membrane preparations from the cotransfected A 1 D 1 cells showed results very similar to those obtained with rat striatal membrane preparations (5). The only difference was the lower proportion of D 1 receptors in the high affinity state (D 1H ) in the A 1 D 1 cells (ϳ10%) as compared with the rat striatal membranes (ϳ30%) (5). Since it has been previously shown that the density of D 1H correlates with the G protein content and with the endogenous dopamine levels (16), this difference might reflect either a low content of G proteins or the absence of a previous exposure of the D 1 receptors to dopamine. Both the GTP analogue Gpp(NH)p and CPA induced a significant reduction in the proportion of D 1H . Gpp(NH)p, but not CPA, was also effective in membrane preparations from control cells containing D 1 but not A 1 receptors, which demonstrates that the A 1 receptors are required for CPA to have an effect. Since D 1H represents the D 1 receptors coupled to the G protein, these results can be interpreted as an uncoupling of the D 1 receptor from its G protein induced by A 1 receptor stimulation.
PTX induces an uncoupling of the A 1 receptor from its G protein by inducing an ADP-ribosylation of the G ␣ subunit of the G i (and G o ) protein family. This results in a reduction of the number of A 1 receptors in the high affinity state and in a blockade of A 1 receptor signal transduction (17)(18)(19). A 1 D 1 cells were exposed to PTX to study the possible involvement of G i proteins in the A 1 receptor-mediated uncoupling of the D 1 receptor from the G s protein. As with the blockade of A 1 receptors with the A 1 antagonist DPCPX, PTX induced a significant increase in the basal levels of cAMP and potentiated dopamineinduced cAMP accumulation. This gives functional support for the blockade of A 1 receptor signal transduction by PTX in these experiments. It was found that PTX counteracted the effect of CPA (10 nM), but not of Gpp(NH)p, on D 1 receptor binding characteristics, suggesting that the G i protein was, in fact, necessary for the intramembrane A 1 -D 1 interaction. However, a higher concentration of CPA (10 M), which is sufficient to  bind to the A 1 receptor in the low affinity state in the A 1 D 1 cells (see competitive inhibition curves of CPA versus [ 3 H]DPCPX), could still uncouple D 1 receptors from the G s protein after PTX pretreatment. This effect of CPA was not reproduced in D 1 cells, which shows that it is not a nonspecific effect, but is A 1 receptor-mediated. Furthermore, in agreement with the radioligand binding experiments, a high concentration of CPA (10 M) was still able to significantly decrease dopamine-induced cAMP accumulation after PTX pretreatment. The PTX-induced ribosylation, although very distinct, was not complete (see the SDS-polyacrylamide gel in Fig. 6). Therefore, it is still possible that stimulation of the low amount of A 1 receptors in the high affinity state left after PTX pretreatment with the high concentration of CPA would be able to reproduce the same effect as the low concentration of CPA in PTX-nonpretreated cells. Nevertheless, the clear correlation between the results obtained with the radioligand binding and cAMP accumulation experiments suggests that the intramembrane A 1 -D 1 interaction involved in the binding experiments is related to the A 1 -D 1 interaction found at the adenylyl cyclase level.
In summary, three main findings have been obtained in this work. The first finding is that in membrane preparations from stably cotransfected A 1 D 1 cells, the stimulation of A 1 receptors induces an uncoupling of the D 1 receptor from its G protein.
This intramembrane A 1 -D 1 interaction has the same characteristics as that previously found in rat striatum (5). The dem-onstration of this interaction in an artificial and very different cellular type and cellular environment strongly suggests that these kind of intramembrane receptor-receptor interactions (1) represent a generalized functionally important mechanism in mammalian cells. The second finding is a functional antagonistic A 1 -D 1 interaction at the adenylyl cyclase level. Although previously shown in homogenates of rat striatum (20), this is FIG. 6. Effects of PTX on the [ 32 P]ADP-ribosylation of G proteins in membrane preparations from A 1 D 1 cells. PTX ADP-ribosylates G proteins with molecular masses of of 39 -41 kDa. [ 32 P]ADPribosylation is markedly reduced in A 1 D 1 cells previously exposed to PTX compared with controls. The positive control (ϩ) contained whole rat brain; the negative control (Ϫ) contained buffer. PTX and C (control) represent A 1 D 1 cells previously pretreated or not with PTX, respectively. the first time that such an interaction has been demonstrated at the cellular level. Finally, the third finding is the correlation between the results obtained with the radioligand binding and cAMP accumulation experiments, suggesting that the intramembrane A 1 -D 1 interaction involved in the binding experiments is related to the A 1 -D 1 interaction found at the adenylyl cyclase level. Similar interactions are likely to occur in nerve cells that express both A 1 and D 1 receptors, such as the ␥-aminobutyric acidergic strionigral-strioentopeducular neurons (21).