Differential coupling of dopaminergic D2 receptors expressed in different cell types. Stimulation of phosphatidylinositol 4,5-bisphosphate hydrolysis in LtK- fibroblasts, hyperpolarization, and cytosolic-free Ca2+ concentration decrease in GH4C1 cells.

Dopaminergic D2 receptors are widely regarded as typical inhibitory receptors, as they both inhibit adenylyl cyclase and decrease the cytosolic free Ca2+ concentration ([Ca2+]i) by activating K+ channels. A D2 receptor has recently been cloned (Bunzow, J. R., Van Tol, H. H. M., Grandy, D. K., Albert, P., Salon, J., Christie, M. D., Machida, C. A., Neve, K. A., and Civelli, O. (1988) Nature 336, 783-787) and expressed in two different cell lines, pituitary GH4C1 cells and Ltk- fibroblasts, where it has been shown to induce inhibition of adenylyl cyclase. We have investigated the additional effector systems coupled to this receptor. The responses observed in the two cells lines, which express similar levels of receptors (0.5-1 x 10(5)/cell), were surprisingly different. In GH4C1 cells D2 receptors failed to affect phosphoinositide hydrolysis and induced a decrease of [Ca2+]i. This latter effect appears to be mediated by hyperpolarization, most likely due to the activation of K+ channels. In striking contrast, in Ltk- fibroblasts the D2 receptor induced a rapid stimulation of inositol(1,4,5)-trisphosphate (+73% at 15 s) followed by the other inositol phosphates, and an immediate increase of [Ca2+]i due to both Ca2+ mobilization from internal stores and influx from the extracellular medium. In both GH4C1 and Ltk- cells, the D2 receptor response was mediated by G protein(s) sensitive to pertussis toxin. The increases of inositol trisphosphate and [Ca2+]i observed in Ltk- cells required dopamine concentrations only slightly higher than those inhibiting adenylyl cyclase (EG50 = 25, 29, and 11 nM, respectively) and were comparable in magnitude to the responses induced by the endogenous stimulatory receptor agonists, thrombin and ATP. The results demonstrate that in certain cells D2 receptors are efficiently coupled to the stimulation of phosphoinositide hydrolysis. The nature of receptor responses appears therefore to depend on the specific properties not only of the receptor molecule but also of the cell type in which it is expressed.


Dopaminergic
Dz receptors are widely regarded as typical inhibitory receptors, as they both inhibit adenylyl cyclase and decrease the cytosolic free Ca2+ concentration ([Ca"'] The receptors for many hormones and neurotransmitters transduce their signals by coupling to GTP binding (G) proteins, which in turn regulate the activity of effector molecules, such as adenylyl cyclase, phospholipase C, and K' and Ca2+ channels (l-3). Each of these receptors was initially believed to selectively activate a single effector pathway. However, results accumulated during the last few years clearly indicate that individual receptor molecules can generate multiple signals by coupling to more than one effector system. Strong evidence in favor of this possibility comes from the study of a group of receptors, which were first demonstrated to inhibit adenylyl cyclase via G,' (4). In some systems agonists to these inhibitory receptors also induce, by means of PTx-sensitive G protein(s), opening of K' channels and/or inhibition of Ca2+ channels, both resulting in the reduction of [Ca2+], (l-3, 5-9). In the case of the heart muscarinic M2 receptor, the current evidence unambiguously indicates that the inhibition of adenylyl cyclase and the activation of K' channels are mediated by the same receptor molecule rather than by different coexisting muscarinic receptor subtypes with specialized functions (5,8,(10)(11)(12). In this and in other systems, the multiple events triggered by inhibitory receptors appear to cooperate in inducing a negative effect on cell function via the decrease of second messenger levels. It was therefore rather unexpected that the cloned muscarinic MP and MS receptors stimulate phosphoinositide hydrolysis in various transfected cells by coupling to G proteins sensitive to . This response, however, required much higher receptor densities and/or agonist concentrations in comparison to the inhibition of adenylyl cyclase and, even under optimal conditions, was rather weak (11-13). It would therefore appear that, although inhibitory receptors can mediate stimulation of phosphoinositide hydrolysis, such a coupling is inefficient and of uncertain biological significance.
Further insight into the multiple effector coupling mechanisms of inhibitory receptors and their relative roles in mediating the final response in the target cells can be provided by the extension of these studies to other members of the same receptor family expressed in various cell types. Dopaminergic Dz receptors are known to inhibit adenylyl cyclase as well as  (1,4,5)P3 and Ins (1,3,4)P3, inositol 1,4,5-and 1,3,4-trisphosphate. to activate K+ channels (14-17). However, the characterization of the transduction pathways used by this inhibitory receptor has been limited by the lack of convenient experimental cell models, and our present knowledge results from the study of very few cell systems (17). A Dz receptor has recently been cloned from a rat brain cDNA library and expressed in two cell lines, pituitary GH4C1 cells and Ltkmouse fibroblasts (18)(19)(20). In both cell types this receptor induces inhibition of adenylyl cyclase, and in the GHICl line it also blocks prolactin secretion (19,20). We report here the effects of receptor activation on phosphoinositide hydrolysis, [Ca2+li, and membrane potential. The results demonstrate that in GH4C1 cells the DP receptor does not affect phosphoinositide hydrolysis and decreases [Ca2+]i most likely via a K' channel-dependent hyperpolarization, whereas in Ltk-cells it efficiently stimulates PtdInsP, hydrolysis with consequent increase of [Ca2+11. The properties of this latter response indicate that the stimulation of phosphoinositide hydrolysis is a coupling of primary relevance of the DO receptor in certain cells. Thus, an individual receptor molecule can induce profoundly different responses depending on the cell type in which it is expressed.

Expression
of D2 Receptors-The cell lines employed in this study were wild-type GHICl and Ltk-cells, which are devoid of endogenous D2 receptors, and the previously described GH4ZR7 and LZR, (formerly referred as L-RGB2Zem-1) clones transfected with D2 receptor cDNA (18)(19)(20). In GH,ZR7 cells transcription of the D2 receptor cDNA is regulated by a zinc-sensitive metallothionein promoter (20). The experiments were therefore carried out after treatment of the cells with ZnSOl as described under "Experimental Procedures." Under these conditions GH,ZR7 cells express approximately 1 x lo5 D2 receptors/cell. The LZRl clone stably expresses about 0.5 X lo5 D2 receptors/cell.

Effects of D2 Receptor Activation in GHJR,
Cells-We first examined the ability of the D2 receptor to affect phosphoinositide hydrolysis. As shown in Fig. lA, in GH4ZR7 cells D2 receptor activation failed to induce a detectable change in inositol phosphate production, even in the presence of very high dopamine concentrations (1 mM Fig. 2A). This dopamine effect was abolished by the selective DP antagonists 1-sulpiride and butaclamol and was not detected in wild-type GH&i cells ( Fig. 2A and results not shown).
Endogenous Dz receptors in pituitary lactotrophs, and somatostatin receptors in cells of the GH lines, reduce [Ca2+]i by inhibiting Ca*+ entry through voltage-dependent Ca*+ channels (25)(26)(27). The latter effect is in turn mediated, at least in part, by hyperpolarization due to the opening of K+ channels (15,16,(25)(26)(27)(28). The following results indicate that the action on [Ca2+]i of the D2 receptor transfected into GH4ZR7 cells can be accounted for by the same mechanism. (a) The effect of DA on [Ca'+]i was similar and not additive to those induced by withdrawal of external Ca*+ and inhibition of voltage-gated Ca*+ channels by verapamil (Fig. 2,B and C). (b) When the fluorescent probe bis-oxonol was used to evaluate changes in membrane potential dopamine was found to cause hyperpolarization of GH,ZRT cells (Fig. 3A). This effect was antagonized by l-sulpiride and butaclamol ( As shown in Fig. 1, the dopamine concentrations required for both [Ca*+], decrease and membrane hyperpolarization (EC!,, = 9 + 2 and 12 f 2 nM, respectively, n = 3) were similar to those inducing inhibition of CAMP production (EC& = 14 f 4 nM, n = 3) (see also Ref. 20). The first two effects, however, were independent of the action of DA on adenylyl cyclase, as they were unaffected by the addition of the membrane permeant CAMP analogue,  Fig. 4A shows the time course of the stimulation of inositol phosphate production observed in the presence of 1 PM dopamine. Total InsPa, measured by conventional anion-exchange chromatography on Dowex columns, showed a relatively slow increase, which was clearly detectable only at 1 min and maximal at 10 min (+55%, n = 30). However, analysis of the individual isomers by HPLC (Fig. 4A, inset) revealed that Ins (1,4,5)P3, i.e. the inositol phosphate which originates directly from PtdInsP2 hydrolysis, was promptly stimulated by dopamine (at 15 s: +73%, n = 6). This rapid increase of Ins (1,4,5)Pz was accompanied by a delayed and gradual rise of Ins(1,3,4)P,. It should be noted that in resting LZR, cells Ins (1,3,4)P3 is by far the predominant InsPB isomer (see legend to Fig. 4). This easily explains why the early increase of Ins(l,4,5)P3 is not reflected by a parallel change of the total InsPs fraction separated by Dowex chromatography. Dopamine also caused accumulation of InsP, (maximal increase +75%, n = 30, at 10 min) and InsP, (at 10 min +24, n = 30; at 20 min +70%, n = 6) (Fig. 4A) 5). In the presence of extracellular Ca2+ an immediate rise, going on the average from 191 f 15 to 664 f 78 nM (n = 15), was observed. After this initial peak [Ca"+], gradually decreased to reach a lower plateau that was maintained for several minutes (truce A). The ability of dopamine to increase [Ca'+], was retained in Ca*+-free medium, although under these conditions the rise was short-lived, with return to resting levels within l-2 min (trace B). Thus, the Dz receptor expressed in LZR, cells induce the [Ca*+]' response described for many classical receptors coupled to PtdInsPp hydrolysis: an initial rapid release of Ca2+ from intracellular stores, which is believed to be triggered by Ins (1,4,5)P3, accompanied by sustained influx of external Ca*+ (29). Many Ca*+ mobilizing receptors induce a concomitant hyperpolarization due to the opening of Ca'+-activated K' channels (30). Hyperpolarization was observed also in LZR, cells when the effect of dopamine on membrane potential was investigated (Fig. 6A). Consistent with the involvement of Ca*+-dependent K+ channels the dopamine-induced hyperpolarization 1) was shortened in the absence of external Ca*+ and nearly abolished when cells incubated in Ca'+-free medium were pretreated with another PtdInsP, hydrolysis stimulating agonist, i.e. thrombin (see below), to deplete intracellular stores (Fig. 6 (not shown). The effects of dopamine on [Ca"+], and membrane potential were antagonized by 1-sulpiride and butaclamol and were not observed in non-transfected Ltk-cells (Fig.  5, C and D, and results not shown). Fig. 7 shows the dose-response analysis of the stimulation of inositol phosphate production and [Ca"], in LZR, cells. The effect of dopamine on inositol phosphate generation was clearly detectable at 10 nM and maximal between 100 nM and 1 FM, with EC& values of 25 f 4, 29 + 2, and 18 f 3 nM for InsPB, InsPz, and InsP,, respectively (n = 3). Similar dopamine concentrations were required for raising [Ca*']i (EC& = 29 + 3 nM, n = 3). These values are only slightly higher than those estimated for the dopamine inhibition of CAMP production in LZR, cells (ECso = 11 f 3, n = 3) (Fig. 7) as well as in GH,ZR7 cells (see above and Fig. 1).
The response of LZR, cells to dopamine was compared with those induced in the same cell type by endogenous receptors coupled to PtdInsP, hydrolysis. Screening with a number of agonists revealed that these cells are endowed with both thrombin and ATP receptors. Fig. 4 (panels C and 0) shows the effect of maximal concentration of thrombin and ATP on inositol phosphate production.
With thrombin the maximal stimulations observed over lo-min time course experiments were 36% for InsPB (at 15 s), 39% for InsPz (at 1 min), and 16% for InsPI (at 10 min), (n = 12); with ATP 34, 53 and 26% for InsPs, InsP1, and InsPI, respectively (at 10 min, n = 12). Although the responses induced by the three agonists are difficult to compare because they exhibit different time courses, it is clear from the data that the effects of endogenous thrombin and ATP receptors do not differ substantially in terms of magnitude from that mediated by the transfected D2 receptor (see Fig. L4  and dopamine (see above and Fig. 5A) were comparable. All the effects so far attributed to endogenous DP receptors, as well as those now observed with the cloned D2 receptor expressed in GH,ZRT cells, are mediated by PTx-sensitive G protein(s).
Whether this is the case also for the novel Dg response observed in LZRl cells it was interesting to establish because recent results in various systems have revealed that receptors can employ different G proteins, either sensitive or insensitive to PTx, to stimulate PtdInsP* hydrolysis (13, 31). As shown in Fig. 8, dopamine lost its ability to stimulate inositol phosphate generation in cells pretreated with 100 ng/ ml PTx for 4.5 h. Fig. 8  Cells were labeled and incubated as described in the legend to Fig. 4. The results (n = 6-9) are presented as percentage of the increase of InsPB obtained in the absence of PTx after 15-s incubations with thrombin (7 units/ml) and lo-min incubations with dopamine (1 PM) and ATP (100 PM). Cells were pretreated with the indicated PTx concentrations for 4.5 h. These last results were confirmed also when using a higher toxin concentration (1 pg/ml). The above experiments indicate that stimulation of Ptd-InsPz hydrolysis and inhibition of CAMP production by the DP receptor expressed in LZR, cells are both PTx-sensitive responses and require similar agonist concentrations.
To rule out the possibility of a major interference of the CAMP decrease in the effect of dopamine on phosphoinositide hydrolysis, we carried out experiments in which either the intracellular CAMP levels were increased by forskolin, or the CAMP analogue 8-BrcAMP was administered to the cells (Table I). Both agents slightly decreased basal inositol phosphates (about -20%) in LZR1 cells. Under these conditions the net increase of inositol phosphates induced by dopamine was proportionally reduced, so that the fold stimulation was the same as in control cells. Similar results were obtained with thrombin and ATP (not shown).

DISCUSSION
The present knowledge of the effector systems coupled to dopaminergic Dz receptors has been inferred from studies carried out exclusively on pituitary and brain tissue and primary culture preparations (17). Even in the few cases in which a homogeneous cell system, such as purified pituitary lactotrophs, was employed the interpretation of the results remains problematical due to the possible existence of multiple D2 receptor subtypes that might have differentially contributed to the overall response. Indeed, although pharmacological studies failed so far to provide conclusive evidence for receptor subtypes, the molecular biology approach has revealed that at least two forms of D2 receptor exist (X3,32,33). These molecules, which appear to be generated from the same gene by alternative splicing, differ from each other only for the insertion of a stretch of 29 amino acids in the third cytoplasmic loop (18,32,33).
Whether or not the "short" and the "long" receptor represent functionally distinct subtypes remains to be elucidated. In this report we characterize the effects of the short receptor and demonstrate that this molecule can transduce with similar efficiency at least three distinct responses: the inhibition of adenylyl cyclase, a K' channel-dependent hyperpolarization, and the stimulation of PtdInsP* hydrolysis.
The results obtained in the GH,ZR7 clone clearly indicate that in these cells dopamine concentrations similar to those required for the inhibition of adenylyl cyclase induce hyperpolarization.
The dependence of this effect on K+ channel activation is demonstrated by the blockade observed with high extracellular K' and the K' channel blocker, quinidine. Hyperpolarization was independent of the effect of Dz receptors on adenylyl cyclase, as shown by the experiments with &BrcAMP, and was prevented by pretreatment of the cells with PTx. Taken as a whole, these results appear remarkably similar to those previously obtained in pituitary lactotrophs (15,16,25) and confirm the inference (17) that a single Dz receptor can mediate, by coupling to PTx-sensitive G protein(s), both the inhibition of adenylyl cyclase and the activation of K' channels. The latter effect can explain the ability of DZ receptors to decrease the resting [Ca2+lr in GH,ZR7 cells. Indeed, hyperpolarization is expected to prevent the firing of spontaneous Ca2+ action potentials, as previously described in both individual lactotrophs and cells of the GH lines treated with dopamine and somatostatin, respectively (25, 34). Whether the hyperpolarization mechanism accounts entirely for the [Ca"], decrease observed in GH,ZR7 cells or whether, in addition, the D:! receptor is also negatively coupled to voltage-gated Ca2+ channels, as suggested based on indirect evidence in lactotrophs and striatal neurons (17), cannot be established from our present data. The ability of Dz receptors to stimulate PtdInsPs hydrolysis is demonstrated by the experiments in LZRl fibroblasts where dopamine induced 1) a rapid increase of Ins (1,4,5) The significance of these small differences is doubtful. 2) The stimulation of PtdInsP* hydrolysis was observed in cells expressing relatively low receptor levels (0.5 X lo5 receptors/cell).
3) The D2 response was comparable in magnitude to those induced by the endogenous Ltk-cell receptors coupled to PtdInsPz hydrolysis, i.e. thrombin and ATP receptors. Taken all together these findings indicate that the stimulation of PtdInsP* hydrolysis cannot be regarded as a coupling of secondary relevance of the DP receptor. Further work is needed to establish whether the difference between the DP receptor investigated in this study and other inhibitory receptors is due to intrinsic functional properties of the different receptor molecules or whether similar results can be obtained with all these receptors when expressed in Ltk-cells (and possibly in other cell types).
The response induced by Dz receptors in LZR, cells is shown here to be mediated by G protein(s) sensitive to PTx. This appears to be a general feature when receptors normally considered to inhibit cell activation are coupled to the stimulation of PtdInsPp hydrolysis (11-13, 35, 36 The results obtained in LZRl cells suggest that a similar response may occur also in some of the cell types which endogenously express the same D, receptor. However, such a possibility needs to be substantiated. So far D, receptors have been shown to stimulate [Ca2+]' only in a subpopulation of pituitary lactotrophs cells in which, however, PtdInsPs hydrolysis was not investigated (37). In other cell systems DZ receptors have been reported either to have no effect on inositol phosphate generation or even to inhibit the response induced by stimulatory agonists, such as TRH (21,(38)(39)(40). In pituitary lactotrophs the latter effect is, at least in part, indirect depending on the [Ca'+]i decrease induced by DA (21). While we report here that dopamine does not modify basal inositol phosphate production in GH,ZR7 cells, the possibility (and the mechanisms) of a Dz-mediated inhibition of the TRH response has not been investigated yet. A major finding of the present work is that, with the exception of the inhibition of adenylyl cyclase, the DP receptor appears to selectively activate different effector pathways in the two cell types investigated.
In fact, the coupling to K' channels observed in GH,ZRT cells most probably does not operate in LZR, cells. On the other hand, the D2 receptor stimulates PtdInsPp hydrolysis in the latter cell type but not in GH,ZR7 cells. A similar heterogeneity of responses most likely occurs also with other inhibitory receptors. Indeed, the transfected Mt and Ms receptors, which stimulate PtdInsPa hydrolysis in various cell types (see above), fail to induce such an effect in neuroblastoma x glioma cells (41). As the GH,ZR7 and LZRl cells employed in the present study express similar numbers of Dz receptors, the simplest interpretation of our results is a different expression in the two cell types of the post-receptor molecules required for the response, either the G proteins or the effecters themselves. In particular, in the case of GH& cells the PTx-sensitive G protein mediating PtdInsPz hydrolysis may be not (or not sufficiently) expressed. Alternatively, it can be imagined that despite the availability of the relevant G protein these cells lack its effector, i.e. a specific phospholipase C different from the enzyme activated by TRH via a PTx-insensitive G protein. In this respect it is worth to emphasize that (i) none of the stimulatory receptors identified so far in GHICl cells operates through a PTx-sensitive pathway, and (ii) as many as five different phospholipases C are known to exist, which might be selectively activated by specific G proteins (42). Analogous hypothesis could be proposed about the G protein and/or its K' channel target to explain the dopamine effect observed in GH,ZR7 but not in LZR, cells. Regardless of their actual explanation, these results indicate that the responses induced by an individual D2 receptor molecule differs not simply in degree but even in nature depending on the cell type. Whereas in GH,C, cells this receptor appears to function in complete agreement with its definition of "inhibitory" receptor, in Ltkcells it clearly activates a "stimulatory" pathway. We conclude therefore that the role of a given receptor in transmembrane signaling is determined not only by the functional properties of the receptor molecule itself but also by the specific features of the cell in which it does operate.