Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The clinical utility of opiates is hampered by the development of tolerance, which can be defined as a progressive decrease in drug potency with repeated administration. An improved understanding of the cellular and molecular mechanisms underlying opiate tolerance would have tremendous clinical implications.

Responses to opiates are mediated through three major classes of opioid receptors (ORs): µ, , and , all of which couple to the G_i family of G proteins. The µOR in particular is important for the analgesic and rewarding properties of opiate drugs (Matthes et al., 1996). The availability of cDNAs encoding ORs has opened many avenues of research, including studies of receptor localization, ligand-receptor interactions, receptor structure and function analyses, and receptor desensitization (Mansour et al., 1995). Although much has been learned regarding ORs and their signaling pathways, the mechanisms underlying receptor desensitization and development of drug tolerance remain relatively poorly understood (Nestler and Aghajanian, 1997).

Recently, a family of mammalian proteins with direct effects on the GTPase activities of G subunits has been discovered. To date, 19 mammalian regulators of G protein-signaling (RGS) proteins have been described (Dohlman and Thorner, 1997; Gold et al., 1997; Koelle, 1997; Berman and Gilman, 1998). These proteins mediate their physiological effects by facilitating the ability of G subunits to hydrolyze bound GTP and thereby increase the rate at which the G subunits reassociate with their corresponding G complexes. In this manner, RGS proteins hasten the return of G proteins to their inactive, GDP-bound state. However, more recent data indicate that RGS proteins can hasten the kinetics of receptor responses (Doupnik et al., 1997; Saitoh et al., 1997). As a result, the net effect of RGS proteins on the functioning of G protein-coupled receptors remains unclear.

The goal of the present study was to investigate a possible role for RGS proteins in µOR desensitization. We employed a bioassay developed to quantitate the functional effects of ligands on G protein-coupled receptors (Lerner, 1994). The bioassay relies on the ability of cultured dermal melanophores from Xenopus laevis to retain in vitro their ability to translocate pigment in response to changes in second messenger levels, including G_i-mediated responses, induced by activation of specific G protein-coupled receptors. Major advantages of this assay are that it: 1) provides measures of receptor-initiated functional responses over a wide time course (seconds to hours); 2) allows for multiple measurements of the same sample over time because cell sacrifice is unnecessary; 3) does not require ligand modification (e.g., radiolabeling) for response quantitation; and 4) enables the study of a wide range of receptors and intracellular pathways. For these reasons, the assay provides a rapid and high-throughput system to study functional responses of G protein-coupled receptors (Lerner, 1994) and is currently widely used in drug discovery efforts (Jayawickreme and Kost, 1997).

Using the melanophore-based bioassay system, the effects of morphine on the µOR were characterized. The responses of the D_2B dopamine receptor (D2R) and the cells' endogenous melatonin receptor, both of which are also G_i-coupled (Potenza et al., 1994), were studied for comparison. We investigated the effects of individual RGS proteins, as well as a mutant G_i subunit defective in its interaction with RGS proteins, on receptor responses. The results demonstrate differential abilities of specific RGS proteins to influence the functional responsiveness of the µOR and other receptors to agonist exposure.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture. Xenopus fibroblasts and melanophores were isolated and propagated as described previously (Potenza et al., 1994). Fibroblasts were grown at 27°C in room air in 70% L-15 Medium (Sigma, St. Louis, MO) supplemented with 20% fetal calf serum (Life Technologies, Inc., St. Louis, MO). The melanophores were grown at 27°C in room air in media conditioned for 3 to 4 days by the fibroblasts, collected by passage through a 0.22-µm sterile filter (Falcon, Franklin Lakes, NJ). Media and PBS were diluted with water to 70% strength for all work with the Xenopus lines. Pigment translocation assays were performed in serum-free 70% L-15 to minimize potential exposure of the cells to bioactive moieties from fetal calf serum during the assays. All melanophores used for experimentation stably expressed the human ₂-adrenergic receptor (2AR; Potenza et al., 1992), because the 2AR cell line propagates and electroporates more efficiently than wild-type melanophores (A. Roby-Shemkovitz and M. R. Lerner, personal communication). The stable cell line expressing the µOR was obtained as previously described (Huang, 1996) by cotransfection of the 2AR melanophore cell line with pJGµOR and pcDNAI/NEO (Invitrogen, Carlsbad, CA) and isolation of individual clonal colonies by the process of limiting dilution in the presence of G418.

Microtiter Plate Assays. Melanophores were seeded into standard 96-well tissue culture plates (Falcon) and grown to approximate confluence (about 7,000-10,000 cells/well). One day before assaying, the medium was removed and the cells were washed with either 70% PBS or L-15. Cells were then incubated overnight in 50 µl of 70% L-15/well. Fifty microliters of fresh 70% L-15 were added to each well and the cells were incubated in room light for 0.5 h (nontransfected µOR/2AR and 2AR cells) or in room light for 0.5 h and under a 40-W bulb (General Electric, Albany, NY) with light source located 20 inches from the cells to induce pigment dispersion (transiently transfected cells). Fifty microliters of 70% L-15 containing three times the final concentration of test ligand were added to each well at "time zero", and the cells were maintained in room light (nontransfected µOR/2AR and 2AR cells, transfected cells for naloxone experiments) or under 40-W bulb light source (transfected cells for ()quinpirole and RGS experiments) for the duration of the experiment, except during the 10- to 12-s periods during which measurements were being performed by the microtiter plate reader (SLT/Tecan, Hillsborough, NC). The additional photostimulation was generally employed with melanophores transiently transfected with G protein-coupled receptors because D2R-expressing cells are not fully pigment-dispersed under room light alone (Potenza et al., 1994). Morphine and melatonin were obtained from Sigma and ()quinpirole was obtained from Research Biochemicals International (Natick, MA). All drug preparations were made either on the day of use or obtained from aliquots stored at 20°C. Transmission at 620 nm was measured by a 340 ATTC microtiter plate reader (SLT/Tecan, Hillsborough, NC) using the agglutination mode of the SOFT 2000 program (SLT/Tecan). The agglutination mode acquires 20 separate transmission readings of each well for each time point. In each case, the final transmission reading was discarded (Potenza and Lerner, 1992), and the remaining 19 values were averaged and treated as an individual value. Data stored in ASCII files on a Zeos 386 computer were transferred to a MacIntosh using the Apple File Exchange program for reduction, using Microsoft Excel software. Using the curve-fit application of Kaleidagraph (v3.0), EC₅₀ values were calculated. When curve fitting data, the general equation employed was as follows:
where x = log[ligand]  log (EC₅₀ or IC₅₀), y = (A_F/A_I - 1), y_min = minimum plateau value of y, and y_max = maximum plateau value of y. (A_F/A_I  1) was employed to quantitate pigment aggregation as described previously (Potenza et al., 1994). Briefly, A_I represents the initial absorbance at 620 nm of light by the cells and A_F that at selected times following drug addition. A_F/A_I  1 yields a value of zero for no change in absorbance and decreasing values approaching  1 for decreasing absorbance secondary to pigment aggregation. For simplification of graphing and presenting data, the value (A_F/A_I  1), to be referred to as the "response ratio", has been employed in which the no drug value of A_F/A_I  1 has been subtracted from the A_F/A_I  1 value for the experimental sample (Potenza et al., 1994).

Plasmid DNA Constructs. The plasmid encoding the D2R, pJGD2R, contains the human D_2B dopamine receptor under the transcriptional regulation of a cytomeglavirus (CMV) promoter and was constructed as described previously (Potenza et al., 1994). The murine µOR was subcloned into the eukaryotic expression vector pJG3.6 (Potenza et al., 1994) under the transcriptional regulation of the CMV promoter as described (Huang, 1996) to create the plasmid pJGµOR. LacZ-encoding plasmids used in electroporations were either pON260 or pJGLacZ, as described previously (Potenza et al., 1994). RGS-encoding plasmids, containing human cDNA copies of RGS1 (prcCMV-RGS1), RGS2 (prcCMV-RGS2), RGS3 (prcCMV-RGS3), or RGS4 (pCR3-FLAG-RGS4), were generous gifts of John H. Kehrl (National Institutes of Health, Bethesda, MD) and used for electroporation without modification (Neill et al., 1997). Wild-type (H6pQE6Gia1, H6pQE6Go) and mutant (H6pQE6Gia1G184S, H6pQE6GoG184S) plasmid constructs of G_i1 and G_o, respectively, were generous gifts of Keng-Li Lan and Richard R. Neubig (University of Michigan, Ann Arbor, MI; Lan et al., 1998). Wild-type (_qEE2pcDNAIAmpa) and mutant (_qEE2GSpcDNAIAmpa) plasmid constructs of G_q were generous gifts of Donald Apanovitch and Henrik G. Dohlman (Yale University, New Haven, CT; DiBello et al., 1998). The coding sequences for the wild-type and mutant G_i1 and G_o subunits were excised from their respective plasmids and placed under the transcriptional regulation of a CMV promoter in JG3.6 using standard protocols. During the plasmid constructions, the sequences coding for the histidine tags were removed, generating wild-type and mutant G_i1 and G_o plasmids lacking sequences encoding tags. The identities of all constructs were verified by restriction enzyme digest analysis with enzymes purchased from Boehringer-Mannheim (Indianapolis, IN) or New England Biolabs (Beverly, MA). Plasmid DNA amplification was accomplished using mini-prep and endo-free maxi-prep protocols (Qiagen, Santa Clarita, CA).

Transfections. Plasmid DNA was transfected into the melanophores by the method of electroporation using an Electo Cell Manipulator 600 (BTX, San Diego, CA) at settings of 575 µF, 475 V, and R10 as described elsewhere (Potenza et al., 1994). Briefly, cells were washed with 70% PBS and detached by trypsinization (2.5% trypsin/3 mM EDTA; Life Technologies, Inc.) in 70% PBS. After inactivating the trypsin with serum (Life Technologies, Inc.) and collecting the cells by centrifugation at 2,000g at 4°C for 5 min, the supernatant was removed and the cell pellet resuspended in 5 ml of ice-cold 70% PBS and recentrifuged as above. After removing the supernatant, the cells were resuspended in ice-cold 70% PBS at a density of 5,000,000 cells/ml. Twenty micrograms of plasmid DNA were added to 400 µl of cells and the resulting mix was incubated on ice for 10 min in a prechilled electroporation cuvette (0.2 cm gap; BTX, San Diego, CA). The cuvette contents were triturated with a micropipetor just before electroporation and plated in media immediately after electroporation. Cells were plated onto tissue culture ware pretreated with Cell-Tak (Colloborative Research, Bedford, MA) to maximize survival of transfected cells following electroporation. Cells were assayed 3 days after electroporation, as described above.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Time-course and concentration-response analyses of µOR- or D2R-expressing melanophores in response to morphine, melatonin, or ()quinpirole. µOR-expressing and control cells were incubated in the continuing presence of morphine (A and B) or melatonin (C and D), or D2R-expressing and control cells were incubated in the continuing presence of ()quinpirole (E and F). Changes in absorbance through the cells were measured and used to quantitate receptor-mediated pigment translocation as described in the text. (AF/AI  1), a measure of pigment aggregation, is defined as the response ratio (see Materials and Methods). Error bars represent S.E. values of the mean for triplicate samples. For graphs of time-course analyses, concentrations of indicated drugs are displayed in the figure insets. For graphs of concentration-response curves, cell types used are displayed in the figure insets. Results shown are representative of at least duplicate experiments.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Morphine-Induced Pigment Aggregation in a Melanophore Cell Line Stably Expressing Murine µOR. As a first step in examining µOR function, the properties of a stable cell line of melanophores expressing the murine µOR were explored. After pretreating cells with light to induce pigment dispersion, cells were exposed to varying concentrations of morphine in the continued presence of light, and pigment aggregation was quantitated over time by determination of absorbance changes through the cells (Fig. 1). Increasing amounts of pigment aggregation, as is observed after stimulation of G_i-coupled receptors in the melanophores (Potenza et al., 1994), results in decreasing magnitudes of absorbance readings and negative values of the response ratio (A_F/A_I  1) of increasing magnitude. As expected, increasing concentrations of morphine produced successively increasing amounts of pigment aggregation in the cells (Fig. 1A). The response to morphine appeared within physiologically relevant concentrations, with maximal functional responses seen at 1 and 10 µM. Over the 7-h time course of the experiment, the responses to morphine diminished. The intermediate concentration of 100 nM produced a response peaking at 30 min and returning to baseline over the 7-h time course of the experiment. The higher concentrations of 1 and 10 µM produced responses peaking by 30 min, remaining maximal for an additional 90 min, and then, in a superimposable fashion, returning toward baseline thereafter. The EC₅₀ value for morphine on the cells was 91.4 ± 4.3 nM at 30 min (Fig. 1B).

Antagonism of Morphine-Induced Pigment Aggregation by Naloxone. To study further the characteristics of µOR responses in the cultured melanophores, the potential for naloxone, a specific µOR antagonist, to inhibit morphine-induced pigment aggregation was investigated (Fig. 2, A and B). Cells transiently transfected with µOR-encoding plasmid were pretreated with light to induce pigment dispersion and then exposed to 100 nM morphine and varying concentrations of naloxone. Naloxone blocked the morphine-induced pigment aggregation in a dose-dependent manner. These findings confirm that morphine's effects on the melanophores are mediated via the µOR.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Time-course and concentration-response analyses of naloxone-mediated antagonism of morphine-induced pigment aggregation in µOR-expressing melanophores. Pigment aggregation over time (A) and at 30 min (B) in response to 100 nM morphine and varying concentrations of the µOR antagonist naloxone was measured in cells transfected with µOR-encoding plasmid. LacZ-transfected cells did not show appreciable pigment aggregation in response to morphine. Changes in absorbance through the cells were measured and used to quantitate receptor-mediated pigment translocation as described in the text. (AF/AI  1), a measure of pigment aggregation, is defined as the response ratio (see Materials and Methods). Error bars represent S.E. values of the mean for triplicate samples. For the time-course graph, concentrations of naloxone are displayed in the figure inset. For the concentration-response graph, cell types used are displayed in the figure inset. Results shown are representative of at least duplicate experiments.

Rechallenge of µOR-Expressing Cells to Morphine and Melatonin. To investigate further possible desensitization in µOR responses, cells were exposed to morphine or melatonin and then rechallenged with the same drug (Fig. 3). Incubation proceeded overnight and rechallenge with the same concentration of fresh drug was made (second time zero in Fig. 3) as described in Materials and Methods. Over the 60-min period after the second application of morphine, the drug produced a less robust response for all of the concentrations tested (Fig. 3, A and B). It appears that the higher the concentration of morphine initially applied, the less robust a response was seen on restimulation. For example, re-exposure to 10 µM morphine produced virtually no response (Fig. 3B), whereas re-exposure to 100 nM morphine produced a response approximately two-thirds of that seen initially (Fig. 2A).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Rechallenge of morphine- and melatonin-treated µOR-expressing melanophores with additional drug. Cells were incubated overnight in the continued presence of morphine (A and B) or melatonin (C) and rechallenged 24 h later with the same concentration of the same drug to which they were originally exposed. Changes in absorbance through the cells were measured and used to quantitate receptor-mediated pigment translocation as described in the text. (AF/AI  1), a measure of pigment aggregation, is defined as the response ratio. The initial reading (original time zero) was used as the intrawell standard for calculations of AF/AI  1 for readings on both days 1 and 2 of the experiments, as described in greater detail in Materials and Methods. Error bars represent S.E. values of the mean for triplicate samples. Concentrations of indicated drugs are displayed in the figure inset for each graph. Results shown are representative of at least duplicate experiments.

Effect of RGS Proteins on Responses to Morphine in Transiently Transfected Melanophores Expressing µOR. To investigate the effect of specific RGS proteins on functional responses mediated by the µOR, melanophores were transiently transfected with a plasmid encoding the µOR alone or in combination with plasmids encoding RGS1, RGS2, RGS3, or RGS4. Transient transfection techniques were employed with the goal of targeting the µOR- and RGS-encoding plasmids to the same subset of cells. Cells were also cotransfected with a plasmid encoding LacZ to control for transfection efficiency by staining for -galactosidase activity. Cells were tested over time for their responses to various concentrations of morphine (Fig. 4). Time-course analysis over the 2-h time course of the experiment revealed that cells transfected with the RGS-encoding plasmids retained their abilities to translocate pigment in response to morphine. However, certain RGS proteins, particularly RGS2, were able to diminish the sensitivity of the cells to morphine, as is most evident from comparison of the cells' responses to intermediate concentrations (10 and 100 nM) of morphine (Fig. 4, A versus B). Determination of EC₅₀ values at multiple time points revealed the greatest changes with expression of RGS2, which resulted in approximately 2-fold shifts to the right in EC₅₀ values (Table 1). No consistent effect of the other RGS subtypes was observed. The differential effects of RGS2 and the other RGS proteins on µOR responses did not seem to be caused by differences in the levels of expression of the transfected RGS proteins: Northern blotting revealed comparable levels of expression of each of the RGS mRNAs (data not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Time-course analyses of morphine-induced pigment aggregation in melanophores cotransfected with plasmids encoding the µOR and RGS2. Pigment aggregation in response to morphine was measured in cells transfected with plasmids encoding the µOR alone (Fig. 4A) or in combination with one encoding RGS2 (B). LacZ-transfected cells did not show appreciable pigment aggregation in response to morphine (C). (AF/AI  1), a measure of pigment aggregation, is defined as the response ratio (see Materials and Methods). Error bars represent S.E. values of the mean for triplicate samples. Concentrations of morphine used are displayed in the figure inset for each graph. Results shown are representative of at least duplicate experiments.

View this table: [in this window] [in a new window]   TABLE 2 EC50 values for melatonin on µOR/RGS-expressing cells Data are expressed as mean ± S.D. (n = 3). There were no comparisons giving values of p  .0124, compared with the corresponding µOR/lacZ sample value. Data are representative of at least duplicate experiments.

Effect of RGS Proteins on Response to ()Quinpirole in Transiently Transfected Melanophores Expressing D2R. To investigate the effect of RGS proteins on another G_i-coupled receptor, the D2R was expressed transiently alone and in combination with various RGS proteins and tested with the D2R agonist ()quinpirole (Fig. 5). D2R-expressing cells, in the absence or presence of RGS proteins, showed the ability to aggregate pigment in response to (-)quinpirole, whereas LacZ-transfected cells did not. As seen with morphine on the µOR, RGS proteins influenced the D2R-mediated functional response to ()quinpirole, although different subtypes of RGS proteins were most effective for the two receptors. With the D2R, RGS3 produced the greatest decrement in ()quinpirole response, as is most evident from comparison of the cells' responses to intermediate concentrations (1, 10, and 100 nM) of the drug (Fig. 5, A versus B).

View larger version (31K): [in this window] [in a new window]   Fig. 5.   Time-course analyses of ()quinpirole-induced pigment aggregation in melanophores cotransfected with plasmids encoding the D2R and RGS3 or RGS2. Pigment aggregation in response to ()quinpirole was measured in cells transfected with plasmids encoding the D2R alone (A) or in combination with those encoding RGS3 (B) or RGS2 (C). LacZ-transfected cells did not show appreciable pigment aggregation in response to ()quinpirole (D). (AF/AI  1), a measure of pigment aggregation, is defined as the response ratio (see Materials and Methods). Error bars represent S.E. values of the mean for triplicate samples. Concentrations of ()quinpirole used are displayed in the figure inset for each graph. Results shown are representative of at least duplicate experiments.

View this table: [in this window] [in a new window]   TABLE 3 EC50 values for ()quinpirole on D2R/RGS-expressing cells Data are expressed as mean ± S.D. (n = 3). Data are representative of at least duplicate experiments. The D2R/RGS3/lacZ-expressing cells had the greatest shift in absolute magnitude from the D2R/lacZ-expressing ones.

Effect of Wild-Type and Mutant G_i1 Subunits on µOR Function. To investigate further the effect of RGS proteins on µOR-mediated responses, wild-type or mutant forms of G_i1 were transfected into the µOR-expressing melanophores. The mutant version contains a glycine to serine substitution at position 183 that markedly lowers its affinity for RGS proteins and makes it virtually unresponsive to stimulation by RGS proteins (DiBello et al., 1998; Lan et al., 1998). As shown in Fig. 6, overexpression of the mutant G_i1 potentiated responses of the µOR to morphine. Such increased and prolonged responses to morphine were most evident at intermediate concentrations of drug (e.g., 10 and 31.6 nM) in the cells expressing the mutant protein, compared with those expressing the wild-type version (Fig. 6, A versus B). EC₅₀ values of morphine on the melanophores expressing the mutated version of G_i1 were consistently one-half to two-thirds that observed in cells expressing the wild-type form over the time course of the experiment (Table 4). In contrast to the results obtained with the G_i1 constructs, similar experiments performed with the wild-type and corresponding glycine to serine mutant versions of G_o and G_q found minimal differences in morphine responses between cells transfected with the wild-type and mutant forms of the proteins (not shown).

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Time-course analyses of morphine-induced pigment aggregation in melanophores stably expressing the µOR and transiently transfected with plasmids encoding wild-type or mutant Gi1 proteins. Pigment aggregation in response to morphine was measured in µOR-expressing cells transiently transfected with plasmids encoding wild-type (A) or mutant (B) versions of Gi1. (AF/AI  1), a measure of pigment aggregation, is defined as the response ratio (see Materials and Methods). Error bars represent S.E.s of the mean for triplicate samples. Concentrations of morphine are listed in the figure insets. Results shown are representative of at least duplicate experiments.

View this table: [in this window] [in a new window]   TABLE 4 EC50 values for morphine on µOR/Gi1-expressing cells Data are expressed as mean ± S.D. (n = 3). Data are representative of at least duplicate experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this series of experiments, we investigated the regulation of agonist-induced responses over time for the µOR. Study of the µOR was facilitated by the creation of a stable melanophore cell line expressing this receptor. The inability of even high concentrations of morphine to sustain a maximal response over the course of hours in the µOR stable line suggested the possibility of homologous desensitization, consistent with findings in other cell systems (Zimprich et al., 1995). The finding of decreasing responsivity observed on rechallenge with morphine provides additional evidence consistent with such desensitization. Given that similar studies investigating the effects of melatonin on the endogenous G_i-coupled melatonin receptor showed a different pattern of response over hours and on drug rechallenge suggests receptor specificity in the responses. Moreover, given that the µOR is not under the transcriptional regulation of its endogenous promoter, it is likely that information necessary for the observed pattern of desensitization lies within the receptor protein itself.

We next sought to investigate the effects of specific RGS proteins in regulating functional responses of the µOR. RGS proteins have been shown to increase the rates of hydrolysis of GTP to GDP for many types of G subunits, including G_i subtypes, thereby facilitating their return to an inactive state. We investigated the abilities of four individual RGS proteins, RGS1-4, to influence responses of the µOR to morphine. Three of the four RGS proteins studied (RGS2-4) have been reported to be expressed in brain regions (e.g., striatum) of high µOR and D2R expression (Burchett et al., 1998). One of these proteins, RGS2, was found to have a greater effect on dampening morphine-induced µOR activation than the other RGS subtypes. In contrast, a different RGS protein, RGS3, was found to have relatively greater effects on D2R responses to ()quinpirole, with RGS2 displaying somewhat less of an effect. Responses of the melatonin receptor were not consistently affected by any of the RGS proteins examined. These findings raise the interesting possibility that some degree of RGS/receptor specificity might exist in the regulation of receptor responses.

The basis for this differential action of the RGS proteins on the µOR, D2R, and melatonin receptor remains unknown. All three receptors couple to the G_i family of G proteins, which raises the possibility that perhaps the receptors act through distinct subtypes, or at least distinct cellular pools, of these proteins. Support for selective actions of various RGS proteins on a given receptor comes from several recent studies. In one report, RGS3 displayed a greater ability, compared with RGS1, RGS2, and RGS4, to terminate physiological responses to gonadotropin-releasing hormone in an in vitro assay (Neill et al., 1997) and, in a separate study, RGS1, RGS3, and RGS4, but not RGS2, dramatically accelerated agonist activation of muscarinic m₂ or serotonin 1A receptors (Doupnik et al., 1997). More recently, the N-terminal domain of RGS4 was demonstrated to confer both high affinity and receptor selectivity to the protein (Zeng et al., 1998). These findings support the hypothesis that individual RGS proteins may preferentially interact directly with specific receptor-G protein complexes. A related question is whether specific RGS proteins exert differential effects on specific subtypes of G protein  subunits. However, despite the identification of cDNAs corresponding to 19 mammalian RGS proteins, it has been difficult to demonstrate high levels of specificity of any given RGS protein for a particular subtype of  subunit, with most proteins examined able to regulate both G_i and G_q subunits (see Berman and Gilman, 1998).

We also show in the present study that mutant forms of G_i1 can potentiate functional responses of the µOR. The mutant G_i1 is unresponsive to RGS proteins; that is, RGS proteins are unable to activate the GTPase activity of the  subunit (Lan et al., 1998). As a result, the mutant G_i1 functions in a sense as a dominant negative mutation for RGS proteins. Our findings, therefore, are consistent with the possibility that the desensitization observed in µOR responses in the cultured melanophores under normal conditions (e.g., Fig. 1) is dependent, in part, on RGS function. Thus, we show in this study not only that RGS proteins can promote diminution of the µOR response, but also that RGS proteins are required for the diminution normally seen in the µOR response in the melanophore system. Although the specific site of action of the RGS proteins on pigment translocation within the melanophores is not known with absolute certainty and could conceivably involve effects on downstream proteins, all available data support the notion that RGS proteins directly affect the function of specific G proteins and the immediate proteins (i.e., G protein-coupled receptors) with which they interact (Berman and Gilman, 1998; Zeng et al., 1998).

RGS proteins represent only one of several potential levels at which the functional responses of G protein-coupled receptors can be regulated. G protein-coupled receptor kinases (GRKs) can phosphorylate specific residues in the receptors that causes their association with arrestins and prevents subsequent receptor activation of G protein  subunits. Specific GRKs have been shown to phosphorylate and desensitize ORs in vitro (Pei et al., 1995; Kovoor et al., 1997; Zhang et al., 1998), and chronic morphine administration has been shown to up-regulate one subtype of GRK in specific brain regions in vivo (Terwilliger et al., 1994), consistent with a possible role for GRK-mediated phosphorylation in OR tolerance. ORs have also been shown to be phosphorylated by other protein kinases, which could result in altered receptor sensitivity as well (e.g., see Koch et al., 1997; Chakrabarti et al., 1998). Finally, it is known that agonists can induce internalization of ORs (which may be a phosphorylation-dependent process; Keith et al., 1998), as well as changes in levels of receptor expression (see Blake et al., 1997; Afify et al., 1998), both in vitro and in vivo. Results of the present study introduce another levelRGS proteinsat which agonist-initiated responses of ORs (and other G protein-coupled receptors) can be regulated. A major question for future research is to determine the relative contribution of each of these mechanisms to the OR desensitization and tolerance that occurs in specific neuronal cell types in vivo under various physiological and pharmacological conditions.

The ability of specific RGS proteins to modify responses to the µOR and D2R can begin to be understood within a functional context. Both receptor types are expressed at high levels in striatal regions of brain (Baik et al., 1995; Mansour et al., 1995), where they have been implicated in mediating the rewarding properties of opiates, psychostimulants (e.g., cocaine and amphetamine), and other drugs of abuse (Harris and Aston-Jones, 1994; Matthes et al., 1996; Nestler and Aghajanian, 1997). Recently, Burchett et al. (1998) reported the relative levels of specific RGS mRNAs in striatum. Interestingly, RGS2 mRNA was found to be most abundant of the several RGS subtypes examined, whereas levels of RGS3 mRNA were relatively low. The same study also reported that acute administration of amphetamine increased levels of RGS2 and RGS3 mRNA in this brain region. In another recent study, RGS2 mRNA was found to be highly regulated in striatum by electrical activity as well as by dopamine-acting agents (Ingi et al., 1998). Results of the present study indicate that such changes in levels of expression of RGS2 and RGS3 might be expected to result in alterations in the functional responses of µOR, D2R, and perhaps other G protein-coupled receptors in this brain region.

Additional work is now needed to directly demonstrate effects of RGS proteins on receptor function in vivo and to study the consequences of such regulation with respect to various behavioral phenomena. Nevertheless, results of the present study support a scheme wherein changes in the levels or activity of RGS proteins regulate the functional responsiveness of G protein-coupled receptors. The results also support the possibility (Gold et al., 1997) that specific RGS subtypes, because of their regional and functional selectivity, represent potential targets for the development of novel psychotherapeutic agents.