Distinct Roles for Gαi2, Gαi3, and Gβγ in Modulation of Forskolin- or Gs-mediated cAMP Accumulation and Calcium Mobilization by Dopamine D2S Receptors*

Previous studies have shown that a single G protein-coupled receptor can regulate different effector systems by signaling through multiple subtypes of heterotrimeric G proteins. In LD2S fibroblast cells, the dopamine D2S receptor couples to pertussis toxin (PTX)-sensitive Gi/Goproteins to inhibit forskolin- or prostaglandin E1-stimulated cAMP production and to stimulate calcium mobilization. To analyze the role of distinct Gαi/oprotein subtypes, LD2S cells were stably transfected with a series of PTX-insensitive Gαi/o protein Cys → Ser point mutants and assayed for D2S receptor signaling after PTX treatment. The level of expression of the transfected Gα mutant subunits was similar to the endogenous level of the most abundant Gαi/o proteins (Gαo, Gαi3). D2S receptor-mediated inhibition of forskolin-stimulated cAMP production was retained only in clones expressing mutant Gαi2. In contrast, the D2S receptor utilized Gαi3 to inhibit PGE1-induced (Gs-coupled) enhancement of cAMP production. Following stable or transient transfection, no single or pair set of mutant Gαi/o subtypes rescued the D2S-mediated calcium response following PTX pretreatment. On the other hand, in LD2S cells stably transfected with GRK-CT, a receptor kinase fragment that specifically antagonizes Gβγ subunit activity, D2S receptor-mediated calcium mobilization was blocked. The observed specificity of Gαi2 and Gαi3 for different states of adenylyl cyclase activation suggests a higher level of specificity for interaction of Gαi subunits with forskolin- versus Gs-activated states of adenylyl cyclase than has been previously appreciated.

A wide variety of physiological functions and pathological conditions are regulated by hormones and neurotransmitters which transduce intracellular signals by coupling to heterotrimeric guanine nucleotide-binding proteins (G proteins). 1 Upon receptor activation, G proteins dissociate into G␣ and G␤␥ subunits which in turn regulate the activity of effector molecules (1)(2)(3). The family of G␣ subunits is divided into structural and functional homologues, for example, G␣ s proteins couple positively to AC to increase intracellular production of cAMP; G␣ i/o proteins couple negatively to AC and are inactivated by PTX; and G␣ q proteins couple to PLC-␤ subtypes to increase [Ca 2ϩ ] i and are insensitive to PTX. The G␤␥ subunits of G proteins couple to a variety of cell-specific effectors including AC types II and IV, PLC-␤2 and PLC-␤3, inwardly rectifying potassium channels, and N-type calcium channels (4,5). In addition, G protein-coupled receptors appear to utilize particular combinations of subunits to initiate specific types of responses (6).
The dopamine D2S receptor couples to PTX-sensitive G proteins (G i/o ) to initiate multiple signaling pathways (7,8). In cells of neuroendocrine origin the D2S couples to "inhibitory" pathways, including inhibition of adenylyl cyclase, activation of potassium channels to hyperpolarize the cell membrane, and inhibition of L-type calcium channels (9 -12), which in concert mediate inhibition of hormone secretion and gene transcription, and inhibition of cell proliferation (13)(14)(15)(16)(17)(18). By contrast, when expressed in cells of mesenchymal lineage (e.g. LtkϪ fibroblast or Chinese hamster ovary cells), the same receptor mediates stimulation of phospholipase C activity to induce calcium mobilization, and activation of the mitogen-activated protein kinase cascade leading to enhanced gene transcription and cell proliferation (8, 14, 17, 19 -23). These findings suggest that the same receptor mediates different cellular responses depending on the repertoire of cell-specific effectors that are expressed. To address the pathways that underlie cell-specific signaling we have studied the G protein specificity of D2S receptor coupling, based on the hypothesis that different G protein subunits mediate receptor coupling to inhibitory versus stimulatory signaling events.
PTX acts to uncouple G␣ i/o proteins by ADP-ribosylating these subunits at a conserved carboxyl-terminal domain cysteine (Cys) residue (24). By mutating the conserved Cys residue in G␣ i 1, G␣ i 2, G␣ i 3, and G␣ o to a ribosylation-resistant serine (Ser) residue we have generated a series of PTX-insensitive mutants of G␣ i/o protein (G-PTX). Because the Cys 3 Ser mutation is a structurally conservative change, the mutant G proteins remain functional following PTX pretreatment (25)(26)(27)(28). We have assessed the contribution of individual or specific combinations of G protein subunits to D2S-mediated signaling. The D2S receptor utilizes distinct single G␣ subunits to inhibit cAMP accumulation depending on the method of AC activation. In contrast, calcium mobilization induced by the D2S receptor is not reconstituted with single or combinations of G␣ subunits,  h, the transfected cells were assayed for intracellular free calcium and ␤-galactosidase activity.
Western Blot Analysis-Cells (10 7 /10-cm plate) were harvested and resuspended in 200 l of RIPA-L buffer (10 mM Tris, pH 8, 1.5 mM MgCl 2 , 5 mM KCl, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 1% Nonidet P-40, 1% sodium lauryl sulfate, 0.5% sodium deoxycholate, 5 g/ml leupeptin) on ice. The cell lysate was then passed through a G-25 needle three times to shear genomic DNA and incubated on ice. After 30 min, the lysate was centrifuged (10,000 ϫ g, 10 min, 4°C) and the supernatant recovered and assayed for protein content by the bicinchonic acid protein assay kit (Pierce). Lysates (100 g/lane) were electrophoresed on sodium lauryl sulfate-containing 12% polyacrylamide gels at 100 V, 40 mA for 1 h, blotted on polyvinylpyrrolidone membranes for 1 h at 250 mA at 4°C. Blots were incubated overnight in 5% nonfat dry milk in TBS-T (10 mM Tris, 150 mM NaCl, pH 8.0, 0.05% Tween 20) at 4°C. The blots were then incubated for 1 h with primary antibody, followed by a 30-min incubation with horseradish peroxidase-conjugated secondary antibody at room temperature in TBS-T, and the peroxidase product was developed using the ECL for Western blot protocol.
cAMP Measurement-Equivalent numbers of cells were plated in 6-well plates and grown to 70 -80% confluence. After rinsing with HBBS buffer (118 mM NaCl, 4.6 mM KCl, 1.0 mM CaCl 2 , 10 mM Dglucose, and 20 mM Hepes, pH 7.2) the cells were incubated with or without experimental compounds in 1 ml/well of HBBS ϩ 100 M isobutylmethylxanthine at 37°C. After 20 min the media were recovered and stored at Ϫ20°C. Samples were analyzed by specific radioimmunoassay to detect cAMP (34). Percent inhibition was calculated using the following formula: % inhibition ϭ 100 Ϫ [100(D-C)/(S-C)], where D ϭ cAMP in dopamine-treated cells; C ϭ control or nontreated cells (basal cAMP); S ϭ stimulated cAMP in forskolin-or PGE 1 -treated cells.
Measurement of [Ca 2ϩ ] i -Cells were grown to 80% confluence, harvested with trypsin/EDTA, resuspended in 1 ml of HBBS with 2 M Fura-2 AM and incubated at 37°C for 45 min with shaking (100 rpm). The cells were washed twice with HBBS, resuspended in 2 ml of HBBS, and subjected to fluorometric measurement. The fluorescence ratio of Fura-2 was monitored in a Perkin-Elmer LS-50 spectrofluorometer at EX ϭ 340/380 nm and EM ϭ 510 nm. Calibration was done using 0.1% Triton X-100 and 20 mM Tris base to determine R max and 10 mM EGTA (pH Ͼ 8) to obtain R min (34) and the fluorescence ratio was converted to [Ca 2ϩ ] i based on a K d of 227 nM for the Fura 2-calcium complex. Experimental compounds were added directly to cuvette from 100-fold concentrated solutions at the times indicated in the figures.
Statistical Analysis-The data are presented as mean Ϯ S.E. of at least three independent experiments. The data were analyzed by repeated measure using ANOVA for each set of experiments. The percent inhibition data was analyzed with repeated measure using ANOVA and the data from G-PTX expressing clones were compared with LD2S cell (wild type) using Bonferroni multiple comparison post-test.

Expression of Mutant G␣ i/o Subtypes in LD2S
Cells-In order to investigate the importance of individual G␣ subtypes in dopamine-mediated responses, PTX-insensitive mutants of G␣ i/o were generated and stably transfected into LD2S cells (LtkϪ cells stably transfected with the rat D2S receptor cDNA). Transfected clones expressing the highest levels of individual mutant G␣ i/o RNA were identified by Northern blot analysis (data not shown) and were named RG o , RG i 1, RG i 2, and RG i 3 for clones expressing G␣ o -PTX, G␣ i 1-PTX, G␣ i 2-PTX, and G␣ i 3-PTX, respectively. Cell extracts from clones of interest were subjected to Western blot analysis to assess at the protein level the overexpression of G␣ proteins (Fig. 1). Wildtype LD2S cells expressed all four G␣ i/o subunits, although G␣ o and G␣ i 3 appeared to be the most abundant based on densitometric analysis. Comparison of G␣ o and G␣ i 3 expression in each transfectant to LD2S (wild type) indicates that transfectant cell lines expressed approximately 2-fold more than the corresponding endogenous G␣ subunit. This indicates that approximately equal amounts of mutant and wild-type protein were produced in the transfected cell lines.
G␣ i 2-PTX Mediates D2S-induced Inhibition of Forskolinstimulated cAMP Accumulation-In LD2S cells, dopamine did not alter the basal cAMP production (21). Upon addition of forskolin (10 M), cAMP levels were increased by 4.5-fold compared with basal (2.22 Ϯ 0.17 versus 0.50 Ϯ 0.04 pmol/ml) ( Fig.  2A). Dopamine (10 M) inhibited forskolin-stimulated cAMP accumulation in these cells by 84.5 Ϯ 12.2% (n ϭ 5), an action that was mimicked by apomorphine (1 M, not shown) and was largely reversed by pretreatment with PTX ( Fig. 2A), indicating the involvement of G i/o proteins. PGE 1 has been shown to induce a concentration-dependent increase in cAMP production indicating the presence of endogenous G s -coupled PGE 1 recep-tors (30). In LD2S cells, PGE 1 (1 M) increased cAMP accumulation by 7-8-fold basal cAMP (1.65 Ϯ 0.25 versus 0.196 Ϯ 0.002 pmol/ml) (Fig. 2B). The greater effect of PGE 1 may be related to the specific isoforms of adenylyl cyclase present in LD2S cells. Activation of D2S receptors with apomorphine (1 M) inhibited PGE 1 -induced cAMP production by 66.3 Ϯ 7.3% (Fig. 2B) and this action of apomorphine was largely reversed after PTX treatment, implicating G␣ i/o proteins.
The PTX sensitivity of dopamine-mediated inhibition of forskolin-induced cAMP production was examined in wild-type LD2S and stable clones expressing the mutant G␣ subunits. Dopamine inhibited forskolin-stimulated cAMP accumulation in all clones expressing mutant G␣ i/o proteins, as observed in LD2S cells (wild type). However, PTX blocked dopamine action in all clones except for those clones which express the mutant G i 2-PTX. In multiple experiments, the percent inhibition by dopamine of forskolin-stimulated cAMP accumulation was unaltered by PTX in only RG i 2-3 and RG i 2-4 clones, whereas in other clones a significant attenuation of dopamine action by PTX was observed (Fig. 3). These results indicate that the PTX-insensitive mutant of G␣ i 2 is functional and that G␣ i 2 is the only subunit required for D2S-mediated inhibition of forskolin-induced cAMP production in LD2S cells. G␣ i 3-PTX Mediates D2S Inhibition of PGE 1 -stimulated cAMP Production-The ability of D2S receptor activation to inhibit G s -coupled stimulation of cAMP accumulation was tested in the LD2S clones expressing PTX-insensitive G proteins. In these clones PGE 1 (1 M) induced a 7-8-fold increase in basal cAMP and apomorphine inhibited PGE 1 -stimulated cAMP production by 60 -70%, comparable to wild-type LD2S cells. Upon pretreatment with PTX, apomorphine-mediated inhibition was completely reversed in RG o , RG i 1, and RG i 2 clones. In contrast, PTX treatment of the RG i 3-2 clone did not block apomorphine-mediated inhibition of the PGE 1 response. In multiple experiments, clone RG i 3-2 retained significantly higher dopamine inhibitory activity following PTX treatment than any of the other clones (Fig. 4). Thus, the inhibitory action of the D2S receptor on G s -coupled enhancement of cAMP is mediated through the G␣ i 3 subunit, rather than the G␣ i 2 subunit as observed for forskolin-induced cAMP accumulation. These results show that the D2S receptor utilizes distinct G␣ i subtypes to inhibit forskolin-or G s -stimulated adenylyl cyclase activity.
Adenylyl Cyclase Expression in LD2S Cells-To further investigate the role of AC expression in LD2S cells, we performed semi-quantitative reverse transcriptase-PCR to determine the relative expression of AC subtypes I-VI, since their regulation has been well characterized compare with other subtypes (VII-X) (21, 35-37). The PCR was performed with different concentrations of cDNA (0.1, 0.5, and 1.0 g/reaction) and repeated at least twice for each concentration. Each PCR reaction amplified a single, specific product with the predicted sized for each AC subtypes, and the sequence was confirmed by sequencing the subcloned fragment. The specificity of the primers used was not altered with change in cDNA concentration, but the intensity of the product increased with concentration (Table I). Mouse brain RNA was used as positive control and was found to express all the subtypes of AC (data not shown). In LD2S cells, RNA for AC I and VI was expressed more abundantly than AC III, AC IV, and AC II (ACI ϭ ACVI Ͼ ACIII Ͼ ACIV Ͼ ACII) and AC V RNA was not detected in these cells (Table I). These results indicate that LD2S cells express AC subtypes at different levels, which may direct the specificity of signaling through AC in these cells.

G i/o Protein Subtypes Involved in Calcium Mobilization in
LD2S Cells-In LD2S cells, the D2S receptor couples to PI turnover to induce mobilization of calcium from ionomycinsensitive intracellular stores (8). In LD2S cells dopamine induced a 2-2.5-fold increase in [Ca 2ϩ ] i (Fig. 5) (Fig. 6). Following pretreatment with PTX, none of the mutant G protein transfectants exhibited a D2S-mediated calcium response (Fig. 6). In order to test whether more than one G protein could rescue the calcium response, LtkϪ cells were transfected with different pairs of the four G-PTX mutant constructs along with D2S receptor and assayed for [Ca 2ϩ ] i . In all sets dopamine increased [Ca 2ϩ ] i by 1.6-fold (Fig. 6), similar to that in LD2S cells (33). This indicates that the LtkϪ cells express sufficient levels of transiently transfected cDNAs to mediate a full functional response. However, after PTX treatment, none of the mutant combinations rescued the dopamine response (Fig. 6).
In these experiments, the increase of [Ca 2ϩ ] i induced by 100 M ATP served as a positive control for cellular responsiveness and was unchanged following PTX pretreatment, since the ATP response is mediated through a G q -coupled P2-purinergic receptor. Similarly, in LD2S cells transiently transfected with pairs of G-PTX plasmids, PTX pretreatment blocked completely the dopamine-mediated calcium responses for all combinations (data not shown). The dopamine response was also tested in another series of transfections in which a double dose (60 g) of single G-PTX plasmid was transiently transfected in LD2S cells. As for the stable LD2S clones (see above), the single G-PTX did not mediate the D2S calcium response in these The reverse transcriptase-PCR was performed for ACI-VI with 0.1, 0.5, and 1.0 g of cDNA as indicated. The data have been obtained from at least two independent experiments for each condition. Specific primers for each subtype of AC amplified only a single product with the predicted size corresponding to the subtype they have been targeted to. Positive (ϩ) and negative (Ϫ) signs indicate the presence and the absence of the specific product on ethidium bromide stained gels, respectively, and "Ϯ" represents a weakly detectable product. As a positive control, all primer pairs yielded same size products from mouse brain RNA (not shown).
transfections (data not shown). The protein level of the G␣ i 1 and G␣ i 2 was increased by more than 2-fold in LtkϪ cells transiently transfected by both of these proteins (Fig. 6, inset). Based on these results, none of the G␣ i/o subunits, alone or in combination, mediated calcium mobilization induced by dopamine in LD2S cells.
G␤␥ Subunits Mediate D2S-induced Calcium Mobilization in LD2S Cells-In order to investigate the role of G␤␥ subunit of G i /G o proteins in D2S-mediated increase in [Ca 2ϩ ] i , LD2S cells were stably transfected with the His 6 -tagged carboxyl-terminal of GRK-2 (GRK-CT), which contains a pleckstrin homology domain that is known to bind and inactivate free G␤␥ subunits (38). The relative level of His-GRK-CT protein in clones expressing His-GRK-CT was determined by Western blot using an antibody against the His epitope (Fig. 7, inset). As shown in Fig. 7, the dopamine-induced increase in [Ca 2ϩ ] i was reduced by 80% compared with LD2S cells in the clone expressing His-GRK-CT (RD-21). In contrast, dopamine mediated inhibition of forskolin-and PGE 1 -stimulated cAMP accumulation was not significantly different between LD2S and RD-21 cells (data not shown). In another clone expressing lower levels (20%) of His-GRK-CT than RD-21, the increase in [Ca 2ϩ ] i induced by dopamine was reduced by only 30% (data not shown). These results suggest that D2S-mediated stimulation of [Ca 2ϩ ] i is mediated by G␤␥ subunits and is more dependent on G␤␥ subunits than particular G␣ subunits.

State-dependent Modulation of Adenylyl Cyclase via Distinct G Proteins-
The dopamine-D2S receptor is coupled to inhibition of adenylyl cyclase in a wide variety of cell types (13, 14, 39 -41). Indeed, inhibition of adenylyl cyclase by receptors that couple to G␣ i/o appears to be a ubiquitous pathway (2,5,6). In intact cells, adenylyl cyclase can exist in at least three states: basal, forskolin-stimulated, or G␣ s -stimulated (36). In LtkϪ cells, the D2S receptor inhibits cAMP production stimulated by either forskolin or PGE 1 , but does not inhibit the basal level of cAMP level (Fig. 2, (21)). By contrast, in pituitary cells the D2S receptor inhibits all three states, i.e. basal, forskolin-stimulated, and vasoactive intestinal peptide-stimulated cAMP accumulation (13,14,39). The basal levels of cAMP are at least 5-fold lower in LtkϪ fibroblast cells as compared with GH4C1 pituitary cells (34), and perhaps it is already at a minimum level. Furthermore, our results show that LtkϪ cells have an undetectable level of ACV, and ACII and IV are weakly expressed (Table I). This could explain why D2S receptor activation induced no change in basal cAMP production in LD2S cells. For example, the dopamine-D3 receptor appears to couple exclusively to ACV (42).
To address the G protein specificity of D2S receptor signaling, LD2S cells were transfected stably with individual PTXinsensitive mutants of G␣ i/o subtypes and treated with PTX to inactivate endogenous G i/o proteins. When stimulated by forskolin, inhibition of cAMP accumulation by D2S receptor activation is mediated exclusively through the G␣ i 2 subtype (Fig.  3). This agrees with findings in pituitary cells. Using antibodies to G␣ i/o subunits, Izenwasser and Cote (43) have reported that inhibition of cAMP accumulation by D2 receptors in pituitary tumor cells utilizes G␣ i 1 and/or G␣ i 2, since their antibody detects both subtypes equally. Furthermore, using PTX-insensitive mutants, Senogles (39) has shown that in GH4C1 cells expressing D2S receptor, D2S inhibition of forskolin-induced adenylyl cyclase is routed through G␣ i 2. Our results indicate that in LtkϪ fibroblast cells, the D2S receptor couples preferentially to the G␣ i 2 subtype to inhibit activation of adenylyl cyclase by forskolin.
On the other hand, inhibition of PGE 1 -stimulated cAMP accumulation, a G␣ s -coupled AC pathway, by D2S receptor activation was mediated through G␣ i 3 in the LD2S cells, and not by G␣ i 2 as for forskolin (Fig. 4). This indicates that D2S receptor utilizes distinct G␣ i proteins to mediate inhibition of AC depending on the pathway of activation of AC. This is consistent with previous findings in GH4C1 pituitary cells, in which antisense depletion of G␣ i 2 only marginally reduced D2S-mediated inhibition of vasoactive intestinal peptide-stimulated cAMP levels (G s -coupled), but completely blocked D2Smediated inhibition of basal cAMP level (14). Interestingly, the coupling of other receptor subtypes (e.g. somatostatin or muscarinic-M4) to inhibition of vasoactive intestinal peptide-induced cAMP was completely blocked by depletion of G␣ i 2. Thus, G␣ i 2 plays an important role in D2S coupling to both basal and forskolin-induced AC, but not in coupling to G sstimulated AC in both transfected GH4C1 and LtkϪ cells. In contrast, G␣ i 3 mediates inhibition of G s -stimulated AC activity in LD2S cells. Taken together, these findings provide evidence for a precise regulation of AC activity that is dependent on specific interactions between different activation states of AC and distinct Ga i subtypes.
One explanation for the state-dependent specificity of inhibition of AC by G␣ i is that different subtypes of AC are recruited for forskolin-versus G s -mediated activation pathways. Sutkowski and co-workers (37) have shown that G␣ s activates ACII more efficaciously than ACI, ACV, or ACVI, whereas forskolin preferentially activates ACI over ACII, ACV, or ACVI. Furthermore, it has been shown that G␣ i 1 inhibits ACV more efficiently than ACI (44). By this interpretation, our results would be consistent with G␣ i 2 having greater activity to inhibit ACI, and G␣ i 3 inhibiting ACII. However, specific interactions between different G␣ i and AC subtypes have not been reported. Another possibility is the state-dependent modulation of individual AC subtypes by distinct G␣ i proteins. When expressed in Sf9 cells, G␣ i 1 did not inhibit G␣ s -mediated stimulation of ACI (44). However, when ACI was stimulated by calmodulin or forskolin, G␣ i 1 mediated inhibition of ACI. This indicates that the extent of inhibition of a particular AC subtype (ACI) by G␣ i 1 is dependent on the activation pathway. Following D2S receptor activation, G␣ i 2 may also preferentially inhibit the forskolin-activated state in AC subtypes that predominate in LtkϪ cells, whereas G␣ i 3 preferentially inhibits the G s -activated state. Distinct conformational changes in AC upon interaction with forskolin or G s could explain the selective inhibition of G␣ i subtypes. The crystal structure of the catalytic domain of ACII reveals that forskolin binds to two symetric sites to prevent hydration and enhance dimerization of the C1-C2 domains (45). The binding site for G␣ i/o has not been determined, but it may bind to the a2/a3 region of C1a, which is close to the catalytic domain, on the opposite site of the G␣ s site (46). In this case, G␣ i/o protein could alter the preferable alignment by blocking the "counterclockwise" rotation of C1 (47). In the forskolin-bound conformation, AC may preferentially recognize specific G␣ i/o protein subtypes distinct from those recognized by the G s -bound conformation of AC. Further structural studies may reveal the molecular basis for state-dependent G i protein selectivity in inhibition of AC.
G␤␥ and Calcium Mobilization-In LD2S cells transfected with single or pairs of PTX-insensitive G i/o proteins cDNAs to yield a greater than 2-fold protein expression, dopamine failed to induce calcium mobilization after PTX treatment, suggesting that G␣ subunits plays a minor or secondary role in this pathway. On the other hand, inhibition of G␤␥ signaling in LD2S cells expressing GRK-CT correlated with an inhibition of D2S-induced calcium mobilization, indicating that this process is mediated through G␤␥ subunits rather than the G␣ subunit. This result is consistent with the fact that G␤␥ subunits of G i/o proteins can activate PLC-␤2 and PLC-␤3 (5). Recent results indicate that the D2S receptor increases calcium mobilization in LD2S cells via activation of PLC-␤2 (33), implicating G␤␥signaling in the calcium mobilization pathway. The lack of activity of individual G␣ subunits to mediate calcium mobilization was partly unexpected. In NG108-15 neuroblastoma cells, PTX-insensitive G␣ o did mediate coupling to inhibition of calcium channel activation (25), a G␤␥-mediated response (5). It has been estimated that a 10-fold higher amount of G␤␥ is required to activate PLC-␤2 in vitro than is required for G␣ imediated activation of AC (48). It may be that multiple G i /G o subtypes, rather than a single subtype, must be activated to release sufficient G␤␥ subunits to induce calcium mobilization in LD2S cells.
Conclusion-The dopamine D2S receptor couples to G␣ i 2 to inhibit forskolin-induced cAMP production. On the other hand, when AC is activated by a G s -coupled receptor (PGE 1 receptor), dopamine-induced inhibition is mediated by G␣ i 3. Furthermore, D2S-induced increase in [Ca 2ϩ ] i in LD2S cells is not dependent on any particular G␣ i/o subtype, but is dependent on mobilization of G␤␥ subunit. Therefore, the dopamine D2S receptor utilizes different G i/o protein subunits to regulate a diversity of effector functions within the cell. Moreover, this study shows that the PTX-insensitive G i/o mutants provide useful tools for the dissection of G protein coupling to receptors.