Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Phosducin (Phd), a cytosolic phosphoprotein (Lee et al., 1992), has been shown to exist in the retina, in the peripheral organs, and in the brain (Lee et al., 1987; Danner and Lohse, 1996). The protein translocates upon receptor activation toward the cell membrane (Schulz et al., 1998a), where it is believed to bind with high affinity to subunits of G proteins (Lee et al., 1987; Gaudet et al., 1996). Neutralization of G is likely to interfere with distinct cellular mechanisms (Clapham and Neer, 1997), including the reassociation of G with G (Lee et al., 1992), the GTPase activities of G protein -subunits (Lee et al., 1992; Bauer et al., 1992), and the function of dynamin (Ferguson and Caron, 1998). Phd may prevent homologous desensitization of G protein-coupled receptors by their phosphorylation, as the responsible -adrenergic receptor kinases require freely available G to function (Hawes et al., 1994; Blüml et al., 1997; Schulz et al., 1998a; Pitcher et al., 1995). Considering these functions, it is of interest that the phosphoprotein has been demonstrated to increase the activity of adenylyl cyclase (AC) in olfactory cells (Boekhoff et al., 1997) and in neuroblastoma x glioma cells (NG 108-15), stably expressing Phd (Wehmeyer and Schulz, 1998).

It was proposed to add Phd to the family of endogenous compounds that function as regulators of G proteins (Bauer et al., 1992; Wehmeyer and Schulz, 1998). It was suggested that the phosphoprotein would exert a primarily inhibitory effect on G protein-mediated signaling by scavenging G (Bauer and Lohse, 1998). In this context, the process of agonist-stimulated receptor sequestration (Wong et al., 1994; Ferguson et al., 1996) becomes important, as endocytosis depends on the presence of freely available G (Lin et al., 1998). An application of these mechanisms to the function of opioid receptors would suggest that Phd does affect opioid-induced intracellular signaling. The present investigation uses NG 108-15 cells stably expressing the phosphoprotein (Wehmeyer and Schulz, 1998) to test whether the ability of Phd to scavenge G (Hawes et al., 1994; Gaudet et al., 1996) interacts with opioid-triggered signaling. The -opioid receptors of these cells couple with inhibitory G protein G_i (McKenzie and Milligan, 1990), and phosphorylation of the activated receptors strictly depends on the availability of G (Schulz et al., 1998a). Confocal microscopy supplements the study to investigate the internalization of µ-opioid receptors fused with enhanced green fluorescence protein (EGFP), which have been transiently expressed in control and in Phd overexpressing NG 108-15 cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Radio-labeled tracers were [¹²⁵I]cAMP (2000 Ci/mmol), [³⁵S]guanosine-5'-O-(-thio)-triphosphate (GTPS; >1000 Ci/mmol), [³H][D-Ala²,N-Me⁴,Gly⁵-ol]-enkephalin (DAMGO; 32 Ci/mmol) were obtained from Amersham Buchler (Braunschweig, Germany), and [³H][D-penicillamine²,D-penicillamine⁵]enkephalin (DPDPE; 36 Ci/mmol), [³H]forskolin (31 Ci/mmol), and [³²P]GTP (30 Ci/mmol) were obtained from NEN Research Products, Dreieich, Germany). DPDPE and DAMGO were purchased from Bachem (Bubendorf, Switzerland) and prostaglandin E₁ (PGE₁), forskolin, and all other reagents were purchased from Sigma (Deisenhofen, Germany). Vent DNA polymerase was purchased from New England Biolabs (Schwalbach, Germany), and the restriction enzymes were purchased from MBI Fermentas (St. Leon-Rot, Germany). The enzyme inhibitor COMPLETE came from Boehringer (Mannheim, Germany).

Cell Culture. Neuroblastoma x glioma hybrid (NG 108-15) cells were cultured as described (Ammer and Schulz, 1993). Dulbecco's modified Eagle's medium (DMEM) was supplemented with 10% fetal calf serum, 100 µM hypoxanthine, 1 µM aminopterine, 16 µM thymidine, L-glutamine (0.3 g/l), and penicillin (100 IU/ml)/streptomycin (100 µg/ml). Experiments were conducted at 60 percent confluency of cells. Human embroynic kidney (HEK) 293 cells were grown in DMEM supplemented with fetal calf serum, glutamine, and penicillin/streptomycin at above-mentioned concentrations.

DNA Transfection. The construction of the Phd expression vector and the generation of an NG 108-15 cell clone stably expressing the phosphoprotein were detailed by Wehmeyer and Schulz (1998). The cells expressing the empty vector were designated NGvec, and those expressing in addition Phd were termed NG8 cells. For transient transfection of HEK 293 cells, the calcium-phosphate precipitation method (Sambrook et al., 1989) was performed (40% confluency of cells), for NG 108-15 cells the N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulfate (DOTAP)-reagent was used according the manufacturer`s instructions (Boehringer, Mannheim, Germany). Unless otherwise stated cells were used for experiments 48 h after transient transfection.

Radioligand Binding. Opioid receptor binding was conducted with cell membranes, as detailed by Vachon et al. (1987), using freshly prepared membrane preparations. For estimation of ligand affinities (K_D) and binding capacities (B_max), saturation binding studies were conducted. Cell membranes were incubated for 30 min at 25°C with increasing concentrations of ligands (tracer [³H]DPDPE) in the absence and presence of 1 µM cold DPDPE to define nonspecific binding.

cAMP Assay. A slightly modified method by Ammer and Schulz (1997) was employed. Briefly, cells were seeded onto 96-well plates (1.8 × 10⁴ cells/well) and allowed to settle for 2 h at 37°C in supplemented DMEM (see above). The cAMP assay was conducted in supplemented DMEM (see above) in the presence of 0.25 mM 3-isobutyl-1-methylxanthine. cAMP accumulation was assayed over 20 min at 37°C. Assays were conducted in triplicate.

GTPase Assay. Agonist-stimulated GTPase activity was assayed according to Odagaki and Fuxe (1997). The studies included the use of 0.8 × 10⁶ dpm [³²P]GTP and 8 µg membrane protein per tube (30°C, 15 min, total volume 100 µl). Membranes were preincubated with Phd (1 h, 4°C) to test for drug-stimulated GTPase. Assays were conducted in triplicate.

[³⁵S]GTPS Binding Studies. The technique described by Thomas et al. (1995) was modified. Cell membranes (20 µg protein) were incubated for 30 min at 25°C in 500 µl reaction mixture (50 mM Tris, pH 7.4; 5 mM MgCl₂; 120 mM NaCl; 0.2 mM EGTA, 50 µM GDP; 0.1 nM [³⁵S]GTPS) in the presence and absence of an agonist (see Results section). The mixture also contained the enzyme inhibitor COMPLETE, according to the manufacturer`s instructions. Incubation was terminated by ice-cold buffer (50 mM Tris, pH 7.4; 5 mM MgCl₂) and filtration (Whatman GF/B filters). Filters were counted for radioactivity. Preincubation of membranes with Phd was for 1 h at 4°C. Assays were conducted in triplicate.

Construction of Expression Vectors pEGFP/µ-Opioid Receptor-C3 and µ-Opioid Receptor/EGFP-N3. The plasmids pEGFP-C3 and pEGFP-N3 (Cormack et al., 1996), encoding the red-shifted variant of wt GFP (enhanced GFP, EGFP Chalfie et al., 1994), were from Clontech (Palo Alto, CA). The µ-opioid receptor [pRc/cytomegalovirus-µ-opioid receptor (27)] was amplified by polymerase chain reaction (25 cycles of 1 min at 95°C, 1 min at 55°C, and 1 min at 72°C; Vent polymerase), using the following primers: µ-opioid receptor-F 5' GAT CTC GAG CTC ATG GAC AGC AGC ACC GGC 3'; µ-opioid receptor-R 5' GAT CCC GGG CCC GGCAAT GGA GCA GTT TCT GC 3' (stop codon eliminated). The amplified fragment was cleaved (underlined sequences) with SacI (µ-receptor F) and Bsp120 I (µ-receptor R), and cloned into the SacI/Bsp120 I multiple cloning site of pEGFP-N3 and pEGFP-C3, respectively. In-frame cloning and sequence of the µ-opioid receptor were verified by sequencing (MediGene, Martinsried, Germany).

Phd-Expressing NG 108-15 Cells. Plasmid construction and generation of cells stably expressing Phd (NG8 cells) and the vector only (NGvec), respectively, is reported by Wehmeyer and Schulz (1998). This paper also communicates the characteristics of NG8 cells, including the apparent absence of Phd in NG 108-15 cells.

Confocal Microscopy. Laser scanning confocal images were recorded by means of an inverted Zeiss LSM 410 microscope (Carl Zeiss, Inc., Oberkochen, Germany) using a 40 × 1.3 oil-immersion Plan-Neofluar objective. For excitation, the 488-nm argon-ion laser was used, and the emission was collected with a 520-nm longpass filter. These conditions allowed recording of cells without apparent damages, e.g., lysis or rounding up, during examination (up to 45 min). The digitized images were prepared as graphics by using Adobe Photoshop (version 4.0), and color prints were obtained with the aid of QuarkXPress software.

Flow Cytometric Analysis. Fluorescence analysis of cells (HEK 293, NG 108-15) transiently transfected to express the EGFP-tagged µ-opioid receptors followed essentially the method described by Schulz et al. (Schulz et al., 1998b).

Protein Assay. Protein was assayed by the method of Lowry et al. (1951), using BSA as standard.

Data Analysis. Receptor binding data were fitted with the aid of nonlinear regression analysis to obtain displacement curves and IC₅₀ values, using the Graphpad Prism (Version 2.0; San Diego, CA) computer program. For additional statistics we used Statview software (Abacus Concepts, Berkeley, CA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Opioid Induced Inhibition of AC Activity

Stimulation of AC. Stimulation of NGvec and NG8 cells, respectively, was conducted with forskolin (10^-9 to 10^-5 M) in complete DMEM (n = 4), causing cAMP concentrations that were not significantly different in NGvec or in NG8 cells. Maximal cAMP levels were observed at 10 µM forskolin, resulting in 408 ± 73 (NGvec) and 440 ± 64 (NG8) fmol per 4500 cells (p > .05, paired t test). Figure 1 displays the inhibitory effect of the -opioid receptor agonist DPDPE on cAMP accumulation in NGvec and NG8 cells. The effect of the opioid on forskolin-stimulated (5 µM) cAMP accumulation (Fig. 1A) revealed no significant differences to bringing about half-maximal inhibition (IC₅₀) in NGvec (6.2 nM) and NG8 cells (4.2 nM) (p > .05). It appears, however, that the opioid has a greater inhibitory activity in intact NG8 cells, indicating an elevated intrinsic activity of the opioid in the presence of Phd. Maximal inhibition was 75% in NG8 cells and 60% in NGvec cells. This difference proved statistically significant (p < .01). The use of deltorphine (-receptor agonist) parallels the experimental outcome with DPDPE (data not shown).

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Effect of DPDPE on adenylyl cyclase activities in NG 108-15 cells stably transfected to express the empty vector (NGvec, ) and Phd (NG8 cells, ), respectively. A, stimulation of NGvec cells with forskolin (5 µM) generated 21 ± 2.9 fmol cAMP/4500 cells × min1; NG8 cells displayed 23 ± 2.4 fmol/4500 cells × min1. The forskolin-stimulated cAMP levels were set to 100%, and the ordinate represents the percentage of inhibition of cAMP generation by DPDPE. Arrows indicate the DPDPE concentration required for half-maximal inhibition, which was 6.2 nM for NGvec and 4.2 for NG8 cells. The individual values represent the mean ± S.E.M. of 6 to 10 independent experiments. B, NGvec and NG cells, respectively, were stimulated with PGE1 (100 nM), and cAMP concentrations measured (NGvec 18 ± 0.7 fmol/4500 cells × min1; NG8 22 ± 1.6 fmol/4500 cells × min1) were set to 100%. The percent inhibition of cAMP accumulation by DPDPE is presented. Each value (mean ± S.E.M.) consisted of seven to nine independent experiments.

GTPase Activities

NG8 cells were examined for opioid-stimulated GTPase activity, as an inhibitory effect of Phd on the enzyme was proposed (Bauer et al., 1992; Bauer and Lohse, 1998). Indeed, a reduced GTPase activity of Gi could result in a prolonged inhibitory activity on AC. Experiments presented in Fig. 2 demonstrated a strong increase of GTPase activity in membranes from NGvec cells challenged with 20 nM DPDPE. This effect was significantly reduced in the presence of 0.1 nM exogenous Phd. Comparable studies with NG 8 membranes revealed basal GTPase activities similar to controls (NGvec), but exposure to DPDPE failed to liberate inorganic phosphate (P_i). The ³²P_i levels were even below basal concentrations (p < 01), indicating reduced incorporation of GTP into NG8 membranes (Fig. 3). This effect on GTPase was more pronounced when NG8 membranes were exposed in addition to exogenous Phd. Similar results were obtained with the opioid deltorphine (data not shown). The basal release of ³²P_i from NG8 cells (93 ± 7%, n = 4) as compared with NGvec cells (100%) failed to reach statistical significance (p > .05).

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Effect of Phd on DPDPE-stimulated GTPase activity of membranes (8 µg protein) from NGvec (open columns) and NG8 cells. Membranes were incubated with [32P]GTP, and release of 32Pi released from NGvec membranes was measured under basal conditions after stimulation with DPDPE (20 nM) and after exposure with DPDPE (20 nM) and PhD (0.1 nM). 32Pi released from NGvec membranes under basal conditions was 29.2 ± 1.3 pmol × min1 × mg1, and was set to 100%. The 32Pi release expressed in percent relates to the NGvec basal level. The individual values represent means ± S.E.M. of five to six independent experiments.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   [35S]GTPS binding to membranes of NGvec (open columns) and NG8 cells, respectively, in the presence of 50 µM GDP. Preparations were challenged with DPDPE (20 nM) in the absence and presence of Phd (0.1 nM). Basal [35S]GTPS binding of NGvec membranes (4.2 ± 0.6 fmol × mg protein × min1) was set to 100%. Incorporation of [35S]GTPS was expressed in relation to the basal binding of NGvec membranes. Columns represent mean ± S.E.M. values of six to eight experiments.

Opioid-Regulated [³⁵S]GTPS Binding. The GTPS assay informs about the direct activation of G subunits by receptors, e.g., -receptors in NG 108-15 cells (Breivogel et al., 1997). The studies were conducted with cell membranes separated from the cytosol and, thus, from Phd. Figure 3 demonstrates the opioid-stimulated binding of [³⁵S]GTPS conducted in the presence and absence of exogenous Phd. As expected, DPDPE concentration dependently stimulated the incorporation of [³⁵S]GTPS in NGvec cells up to a 333% over basal values (set to 100%). When Phd (0.1 nM) was added to unstimulated membranes [³⁵S]GTPS binding was not different from basal activity. However, simultaneous exposure of membranes to DPDPE (20 nM) and Phd (0.1 nM) resulted in a strong inhibition of [³⁵S]GTPS binding. Using membranes of NG8 cells, we found basal incorporations were moderately elevated (135%) as compared with NGvec membranes (set to 100%). DPDPE (20 nM) was found to weakly stimulate [³⁵S]GTPS incorporation (193%), reaching a level of statistical significance of p < .02. NG8 membranes resulted in basal levels not different from NGvec cells, and Phd in combination with DPDPE completely failed to stimulate [³⁵S]GTPS binding. Similar effects were observed using the opioid deltorphin instead of DPDPE. The actions of both opioid peptides were blocked by naloxone (1 µM), which itself failed to affect [³⁵S]GTPS incorporation.

Opioid Receptor Binding Studies

High-affinity binding capacities of -receptors of membranes from NGvec and NG8 cells, respectively, were assayed by means of [³H]DPDPE concentrations (0.5 to 12 nM). Scatchard analysis revealed a monophasic binding component both for NGvec (n = 4) and NG8 (n = 3) membranes, disclosing B_max values of 290 ± 39 fmol/mg membrane protein for NGvec membranes, and 328 ± 46 fmol/mg for NG8 membranes. The respective K_D values were 1.87 ± 0.27 nM for NGvec and 1.51 ± 0.32 nM for NG8 cell membranes. The differences between binding capacities and between ligand affinities were not significant (p > .05).

Opioid Receptor Internalization

Activation of opioid receptors triggers their phosphorylation in the presence of freely available G subunits (Schulz et al., 1998b), which may be followed by internalization. Because Phd functions as a scavenger of G, an interference of the phosphoprotein with the process of endocytosis is suggested (Lin et al., 1998). To examine this hypothesis the µ-opioid receptor was fused with EGFP, transiently expressed in both NGvec and NG8 cells, and studied by confocal microscopy in living cells.

Function of µ-Receptors Fused with EGFP. We examined whether fusion of the µ-receptor at the N- and the C-terminus, respectively, with the 27-kDa EGFP (Prasher et al., 1992) affected its function. The studies were conducted with HEK 293 cells transiently transfected to express the µ-opioid receptor or the receptor/EGFP fusion proteins, respectively, as the transfection rates in NG 108-15 cells were very low. HEK 293 cells were transfected by means of the calcium-phosphate precipitation technique, resulting in an expression efficiency for the fusion proteins of up to 40% of cells (determined by fluorescence-activated cell sorting analysis). In NG 108-15 cells the calcium-phosphate precipitation technique brought about transfection efficiency rates in the range of only 1% to 5% of total cells, while use of DOTAP resulted in an average transfection rate of 12%.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   µ-Opioid receptors fused at the C-terminus (µ/EGFP) and the N-terminus (EGFP/µ) with EGFP were expressed in HEK 293 cells. Cell preparations exhibiting a transfection rate of about 40% (determined by flow cytometry) were used to examine their function. A, membrane preparations of cells transfected to express µ-receptors fused at the C-() and the N-() terminus with EGFP were incubated with [3H]DAMGO, and the potency of DAMGO (0.1 nM-1 µM) to displace the tritiated ligand (expressed in percentage of displacement) was tested. When these experiments were conducted in the presence of GTPS (100 µM, open symbols) both displacement curves shifted to the right, indicating a loss of affinity for the opioid. Each value represents the mean ± S.E.M. of six to eight independent experiments. B, the potency of DAMGO to inhibit cAMP generation brought about by stimulation with forskolin (10 µM). Forskolin stimulation of HEK cells stably expressing the µ-opioid receptor (µ-wt) resulted in 56 ± 7 fmol cAMP/4500 cells × min1 (n = 6), and similar levels were established in cells transiently expressing the µ/EGFP (60 ± 9, n = 4) or the EGFP/µ construct (55 ± 6, n = 4). Arrows indicate the DAMGO concentrations required to cause half-maximal inhibition (percentage) of forskolin-stimulated cAMP accumulation. Each value represents the mean ± S.E.M. of six to eight independent experiments.

Internalization. The studies were conducted with living NGvec and NG8 cells transiently transfected to express µ-receptors fused with EGFP. These receptor constructs proved functionally active (see Fig. 4). The cellular distribution of the receptor construct (fluorescence) was monitored by laser-scanned confocal microscopy before and after etorphine exposure (1 µM, 25°C) (Fig. 5). The left side of the figure presents images from a single NGvec cell 2 days after transient transfection with cDNA coding for the µ-receptor/EGFP construct. The untreated cell exhibits fluorescence associated mainly with the cell membrane. The cytoplasm exhibits minor fluorescent material, and the cell nucleus is devoid of detectable labeling. This distribution pattern was largely unaffected up to 10 min after exposure to etorphine. After 12 min, disruption of membrane-located fluorescence was observed, followed by its translocation toward the cytoplasm. After 30 min, most of the fluorescence accumulated in the cytoplasm, most likely in endosomes (Sternini et al., 1996). Employing the same experimental set-up but using DAMGO (1 µM) instead of etorphine, receptor internalization proved of similar time course and efficiency. The same results were obtained with NG 108-15 cells transiently transfected to express the EGFP/µ-receptor construct (images not given). Regardless of the receptor construct expressed in NGvec cells, an etorphine concentration of 0.1 µM brought about a quite similar internalization pattern, which was completely blocked by naloxone (1 µM). Internalization of receptors caused by 1 µM etorphine was largely antagonized by 1 µM naloxone. In general, cells exhibiting detectable membrane-located fluorescence before opiate treatment disclose receptor internalization, that is, translocation of fluorescence towards the cytosol, when challenged with etorphine.

View larger version (88K): [in this window] [in a new window]   Fig. 5.   Confocal microscopic images of living NG 108-15 control cells (NGvec) and Phd overexpressing cells (NG8) 2 days after transient transfection to express the µ-receptor tagged at the C-terminus with EGFP (µ/EGFP construct). Frames represent time-lapse series (top to bottom) of a single confocal plane monitored from untreated cells ("untreated") and after challenge with etorphine (Eto, 1 µM, 25°C). Cells expressing the µ/EGFP construct (green fluorescence) are distinguished from nontransfected cells (red). Sections monitored from the NGvec cell disclose internalization of fluorescence within 30 min. In contrast, section images of the NG8 cell fail to demonstrate translocation of fluorescence. In these cells fluorescence remained membrane-located during the course of the experiment (30 min). The images given represent sections close through the center of the cells. Each cell shown is representative from at least 10 experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This investigation reports that NG 108-15 cells stably expressing Phd give distinctly different responses to opioid challenge as compared with wt control cells. Activation of -receptors results in an increased intrinsic activity in inhibiting forskolin-stimulated generation of cAMP, and the µ-receptor ligands lose activity to bring about receptor sequestration. It is suggested that the underlying biochemical mechanisms were related to the ability of Phd to sequester G subunits (1) and to inhibit opioid-stimulated GTPase (Bauer et al., 1992; Bauer and Lohse, 1998).

The increased efficacy of DPDPE in NG8 cells on cAMP generation may relate to the ability of Phd to inhibit opioid-stimulated GTPase, a concept deduced from the function of certain proteins regulating G protein signaling (RGS proteins; Berman and Gilman, 1998). In fact, Bauer et al. (1992) reported strongly reduced GTP-hydrolyzing activities of purified G-subunits in the presence of PhD. Thus, NG 108-15 cells release Gi2 upon activation of -receptors (McKenzie and Milligan, 1990), and Phd is thought to stabilize the G protein subunit in its GTP-bound state. Because the period when the G-GTP complex can activate its effector depends on the rate at which the -subunit hydrolyzes GTP to GDP, a prolonged functional activity of G_i2 to inhibit AC activity is likely in the case of a reduced GTPase activity. A change in -receptor affinity (IC₅₀) is not expected under these conditions and was not observed. A different mechanism likely to add to the increased DPDPE efficacy may relate to the capacity of Phd to sequester G (Lee et al., 1992), as G protein-coupled receptor kinases 2 and 3 require G to phosphorylate and thus desensitize receptors (Hawes et al., 1994, Blüml et al., 1997), including opioid receptors (Schulz et al., 1998a). These kinases are highly concentrated in NG 108-15 cells (Schulz et al., 1998a). These mechanisms have been stressed with respect to the observed supersensitivity of PGE1 in NG8 cells (Wehmeyer and Schulz, 1998), and the change of AC sensitivity in olfactory cells (Boekhoff et al., 1997).

If the elevated opioidergic efficacy in NG8 cells reflects the consequence of an altered ability of the opioid to activate its receptor, agonist-induced binding of [³⁵S]GTPS may be employed as parameter to examine G activation (Lorenzen et al., 1993, Selley et al., 1998). Using membranes of control cells (NGvec) stimulated by the -receptor agonist DPDPE,we confirmed reports demonstrating a concentration-dependent increase of [³⁵S]GTPS incorporation (Quock et al., 1997). However, in the presence of Phd the opioid lost considerable ability to stimulate incorporation of [³⁵S]GTPS. Basically the same outcome was obtained with membranes of NG8 cells. An analogous effect of Phd was reported by Bauer and Lohse showing in reconstitution experiments (Bauer and Lohse, 1998) that Phd hinders the binding of GTP by slowing the release of GDP. Indeed, experiments with membrane preparations suggest that agonist efficacy is determined by displacement of GDP from G proteins (Breivogel et al., 1997).

The inability of DPDPE to incorporate GTP in the presence of Phd conflicts with the outcome of the concentration response curve for DPDPE to inhibit forskolin-stimulated AC activity. The generation of cAMP was more strongly inhibited in NG8 cells, regardless of the DPDPE concentration (0.1-100 nM). If Phd acts to decrease the proportion of opioid-activated, GTP-bound G subunits (Bauer and Lohse, 1998), a reduced number of functional receptors may be available in the presence of Phd. Our experimental data do not support this notion as we did not observe any differences in the -receptor capacity of NGvec and NG8 cells. Furthermore, the efficacy of DPDPE in the presence of Phd to inhibit forskolin-stimulated AC was elevated, not decreased, as would be expected by a reduced generation of GTP-bound Gi. If Phd inhibited G protein-mediated signaling by reducing active GTP-bound G proteins (Bauer and Lohse, 1998), the opioid-induced inhibitory component on AC would be reduced. This effect is expected to result in an exaggerated function of Gs that would require an increased concentration of the opioid to overcome the stimulated AC activity. We never observed a shift to the right of the concentration-response curve for the opioids tested. Notably, receptor-binding studies, GTP incorporation, and GTPase tests were conducted with cell membranes separated from cytosol, which contains diverse kinases responsible for phosphorylation and desensitization of receptors (Schulz et al., 1998a). Thus, an important component interfering with the complex intracellular signal transmission is absent in these tests that may facilitate receptor-induced signal transmission in intact cells.

Phosphorylation of opioid receptors requires G (Schulz et al., 1998), which is believed to be a prerequisite for initiating internalization. We hypothesized that neutralization of G by Phd may in fact hinder endocytosis. We decided to test this notion by means of EGFP-tagged µ-receptors. Apparently, EGFP (27kDa) (Prasher et al., 1992) fused to either the C- or the N-terminal of the µ-receptor failed to affect their function as judged by their ability to shift the affinity to DAMGO in the presence of GTPS and to inhibit generation of cAMP. Similar data, that the functional capacity remains unaltered, has been reported for other biologically active proteins fused with GFP (Sloan-Lancaster et al., 1997; Tarasova et al., 1997; Fejes-Toth et al., 1998; Schulz et al., 1998b). Documentation of the function of the µ-receptor fused with EGFP is crucial, as G protein coupling has been shown to be essential for receptor internalization (Myburgh et al., 1998). Confocal microscopy revealed that NGvec and NG8 cells, expressing the µ-receptor/EGFP construct, concentrated fluorescent material in the plasma membrane. Etorphine, an opioid known to induce receptor internalization (Keith et al., 1996, 1998), caused fluorescence (receptor) translocation in NGvec control cells. Real time monitoring of living NGvec cells expressing the µ-receptor/EGFP construct revealed that 30 min after etorphine exposure only minor quantities of receptor-associated fluorescence were located in the membrane, confirming previous studies with opioid receptors (Sternini et al., 1996). An identical experimental approach conducted with NG8 cells documented an almost complete failure of etorphine with respect to the translocation of fluorescence (receptors). Even 30 min after challenge with the opioid no major effect was observed. This distinct difference between NGvec and NG8 cells is likely to reflect the potency of Phd to bind G, which consequently attenuates the activity of -adrenergic receptor kinases in these cells to phosphorylate receptors. This interpretation is strongly supported by most recent findings that sequestration of G inhibits receptor-mediated endocytosis (Lin et al., 1998). It was concluded by the authors that cells provide a G pool of uncertain capacity, which, beside other functions (Clapham and Neer, 1997; Ford et al., 1998), controls internalization of membrane receptors.