Involvement of Mitogen-Activated Protein Kinase in Agonist-Induced Phosphorylation of the μ-Opioid Receptor in HEK 293 Cells

Authors


  • Abbreviations used: cAMP, cyclic AMP; DAMGO, [D-Ala2, MePhe4,Gly5-ol]enkephalin; ERK, extracellular signal-regulated kinase; GPCR, G protein-coupled receptor; GRK, G protein-coupled receptor kinase; HA, hemagglutinin; HEK, human embryonic kidney; Li, -like immunoreactivity; MAPK, mitogen-activated protein kinase; MBP, myelin basic protein; MEK, mitogen-activated protein kinase kinase; MOR1, rat μ-opioid receptor; PAGE, polyacrylamide gel electrophoresis; PD98059, 2-(2′-amino-3′-methoxyphenyl)oxanaphthalen-4-one; SDS, sodium dodecyl sulfate; U0126, 1,4-diamino-2,3-dicyano-1,4-bis(aminophenylthio)butadiene.

Address correspondence and reprint requests to Dr. V. Höllt at Department of Pharmacology and Toxicology, Otto-von-Guericke University, Leipziger Strasse 44, 39120 Magdeburg, Germany. E-mail: Volker.Hoellt@medizin.uni-magdeburg.de

Abstract

Abstract: Agonist exposure of many G protein-coupled receptors stimulates an activation of extracellular signal-regulated protein kinases (ERKs) 1 and 2, members of the mitogen-activated protein kinase (MAPK) family. Here, we show that treatment of human embryonic kidney (HEK) 293 cells stably transfected to express the rat μ-opioid receptor (MOR1) with [D-Ala2,MePhe4,Gly5-ol]enkephalin (DAMGO) stimulated a rapid and transient (3-5-min) activation and nuclear translocation of MAPK. Exposure of these cells to the MAPK kinase 1 inhibitor PD98059 not only prevented MAPK activation but also inhibited homologous desensitization of the μ-opioid receptor. We have therefore determined the effect of PD98059 on agonist-induced μ-receptor phosphorylation. DAMGO stimulated a threefold increase in MOR1 phosphorylation within 20 min that could be reversed by the antagonist naloxone. PD98059 produced a dose-dependent inhibition of agonist-promoted μ-receptor phosphorylation with an IC50 of 20 μM. DAMGO also induced MOR1 internalization that peaked at 30 min. Confocal microscopy revealed that DAMGO-induced MOR1 internalization was also largely inhibited in the presence of PD98059. U0126, another chemically unrelated inhibitor of the MAPK cascade, mimicked the effect of PD98059 on μ-receptor phosphorylation and desensitization. MOR1 itself, however, appears to be a poor substrate for MAPK because μ-receptors immunoprecipitated from stably transfected HEK 293 cells were not phosphorylated by exogenous ERK 2 in vitro. The fact that morphine also triggered MAPK activation but did not induce MOR1 internalization indicates that receptor internalization was not required for MOR1-mediated mitogenic signaling. We conclude that MOR1 stimulates a rapid and internalization-independent MAPK activation. Activation of the MAPK cascade in turn may not only relay mitogenic signals to the nucleus but also trigger initial events leading to phosphorylation and desensitization of the μ-opioid receptor.

There are at least three classes of opioid receptors (μ, δ, and κ), which structurally belong to the superfamily of G protein-coupled receptors (GPCRs). The rat μ-opioid receptor (MOR1) mediates the main analgesic effects of morphine and several other opioids. However, the clinical benefit of these drugs is limited by the development of tolerance and dependence during prolonged exposure. The molecular mechanisms underlying these effects are still poorly defined. On the cellular level the μ-opioid receptor undergoes a rapid homologous desensitization on repeated agonist exposure. It is now generally accepted that the agonist-induced desensitization involves phosphorylation of intracellular receptor domains (Lefkowitz, 1998). Phosphorylation facilitates interaction with β-arrestins and targeting of the receptor to the endocytotic machinery. The main route of μ-opioid receptor internalization is via clathrin-coated pits and vesicles into early endosomes. Within the acidic environment of the endosomes, the ligand is effectively separated from the receptor, which becomes dephosphorylated and thus resensitized. As a final step, the receptor recycles back to the cell surface (Koch et al., 1998; Lefkowitz, 1998; Wolf et al., 1999).

Mitogen-activated protein kinase (MAPK), also known as extracellular signal-regulated kinase (ERK), is regarded as a major pathway for growth factor signaling from the cell surface to the nucleus (Seger and Krebs, 1995). In addition, MAPK activation has increasingly been noted in response to agonist stimulation of many GPCRs, including opioid receptors (Burt et al., 1996; Fukuda et al., 1996; Li and Chang, 1996; Chuang et al., 1997; Belcheva et al., 1998; Hawes et al., 1998; Polakiewicz et al., 1998; Ignatova et al., 1999). Recently, much attention has been directed toward the elucidation of the molecular mechanisms of the GPCR-mediated MAPK recruitment. It has been shown that activation of the MAPK pathway by several GPCRs, including the serotonin 5-HT1A, β2-adrenergic, and δ-opioid receptors, requires at least functional β-arrestin and dynamin. Overexpression of dominant-negative inhibitory mutants of β-arrestin or dynamin attenuates both agonist-induced MAPK stimulation and receptor endocytosis (Daaka et al., 1998; Della Rocca et al., 1999; Luttrell et al., 1999). A specific function during β2-adrenergic receptor-mediated MAPK activation has been ascribed to β-arrestin 1 in that it permits recruitment of the tyrosine kinase c-Src to the receptor, which in turn initiates mitogenic signaling (Ahn et al., 1999; Luttrell et al., 1999). In contrast, other GPCRs, e.g., the κ-opioid receptor and the α2-adrenergic receptor, appear to require dynamin and arrestin only for endocytosis but not for MAPK activation (DeGraff et al., 1999; Li et al., 1999).

Although it is well established that activation of opioid receptors stimulates a rapid and transient increase in MAPK activity, little is known about the functional significance of this phenomenon. Recently, Polakiewicz et al. (1998) provided evidence for an involvement of the MAPK pathway in homologous desensitization of the μ-opioid receptor. This finding led us to examine the effects of specific inhibitors of the MAPK cascade on μ-opioid receptor signaling in human embryonic kidney (HEK) 293 cells. We find that PD98059 [2-(2′-amino-3′-methoxyphenyl)oxanaphthalen-4-one] and U0126 [1,4-diamino-2,3-dicyano-1,4-bis(aminophenylthio)butadiene] not only prevent the rapid recruitment of MAPK but also diminish the agonist-induced phosphorylation and internalization of the μ-opioid receptor in a dose-dependent manner. These findings strongly reinforce the hypothesis that the MAPK cascade may play a pivotal role in initiating μ-opioid receptor desensitization.

EXPERIMENTAL PROCEDURES

Materials

[D-Ala2,MePhe4,Gly5-ol]enkephalin (DAMGO) was from Bachem (Heidelberg, Germany), morphine from Synopharm (Barsbüttel, Germany), and naloxone from Sigma (Deisenhofen, Germany). PD98059, mouse monoclonal phosphospecific anti-MAPK antibody clone E10, and activated ERK 2 were purchased from New England Biolabs (Schwalbach, Germany). U0126 was obtained from Promega (Mannheim, Germany). Rat monoclonal anti-hemagglutinin (anti-HA) antibody was from Boehringer (Mannheim, Germany).

Cell culture and transfections

HEK 293 cells (American Type Culture Collection) were maintained in Dulbecco’s modified Eagle’s medium/Ham’s F-12 medium supplemented with 10% fetal calf serum in a humidified atmosphere containing 5% CO2. Cells were stably transfected with human influenza virus HA epitope-tagged rat μ-opioid receptor (kindly provided by Dr. P. Y. Law, Minneapolis, MN, U.S.A.) using the calcium phosphate precipitation method. Approximately 1.5 × 106 cells were transfected with 20 μg of plasmid DNA. Cells were selected in the presence of 500 μg/ml G418 (GibcoBRL, Eggenstein, Germany), and the whole pool of resistant cells was used without selection of individual clones. The maximal percent inhibition of adenylyl cyclase activity induced by 1 μM DAMGO was 38%, and the number of binding sites (Bmax) was 2,000 fmol/mg of protein as determined by radioligand binding using [3H]DAMGO (Koch et al., 1998; Wolf et al., 1999).

Determination of receptor desensitization by measurement of cyclic AMP (cAMP) accumulation

MOR1-transfected cells were seeded at a density of 1.5 × 105 per well in 22-mm diameter 12-well dishes. On the next day, cells were exposed to 1 μM DAMGO for 0 or 2 h. When indicated, cells were preincubated either with 20 μM PD98059 for 60 min or with 100 nM U0126 for 15 min and subsequently maintained under these conditions during agonist exposure. For the measurement of cAMP accumulation, medium was removed from individual wells and replaced with 0.5 ml of serum-free RPMI medium (Seromed, Berlin, Germany) containing 25 μM forskolin (Biotrend, Köln, Germany) or 25 μM forskolin plus 1 μM DAMGO. The cells were incubated at 37°C for 15 min. The reaction was terminated by removing the medium and sonicating the cells in 1 ml of ice-cold HCl/ethanol (1 volume of 1 M HCl/100 volumes of ethanol). After centrifugation, the supernatant was evaporated, the residue was dissolved in TE buffer, and the cAMP content was determined using commercially available radioimmunoassay kits (Amersham, Braunschweig, Germany). Statistical evaluation of results was performed using ANOVA followed by the Bonferroni test.

Direct observation of MAPK activation and receptor endocytosis using confocal microscopy

HEK 293 cells stably expressing HA-tagged MOR1 were grown on poly-L-lysine-treated coverslips overnight. Cells were then exposed to either 1 μM DAMGO or 10 μM morphine in the presence or absence of 0.4 M sucrose for 0, 3, 5, 7, 10, 20, 30, 45, or 60 min. When indicated cells were preincubated with either PD98059 in a concentration of 1, 5, 25, 50, or 100 μM for 60 min or U0126 in a concentration of 1, 10, 100, or 1,000 nM for 15 min. Cells were fixed with 4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer (pH 6.9) for 45 min at room temperature and subsequently washed several times in Tris/phosphate-buffered saline (10 mM Tris, 10 mM phosphate buffer, 137 mM NaCl, and 0.05% thimerosal, pH 7.4). Cells were then incubated for 3 min in 50% methanol and for 3 min in 100% methanol. After a 1-h preincubation in Tris/phosphate-buffered saline containing 3% normal goat serum, cells were incubated with either phosphospecific anti-MAPK antibodies at a dilution of 1:250 for detection of MAPK activation or anti-HA antibodies at a dilution of 1:5,000 for detection of receptor endocytosis in Tris/phosphate-buffered saline containing 1% normal goat serum overnight at room temperature. Bound primary antibody was detected with biotinylated secondary antibodies (1:100; Vector, Burlingame, CA, U.S.A.) followed by streptavidin-cyanine 3.18 (1:100; Amersham). Cells were then dehydrated, cleared in xylol, and permanently mounted in DPX (Fluka, Neu-Ulm, Germany). Specimens were examined using a Leica TCS-NT laser scanning confocal microscope. Cyanine 3.18 was imaged with 568 nm excitation and 570-630 nm bandpass emission filters. Confocal micrographs were taken by an individual blinded to the treatments who randomly selected one colony of four to 12 cells per coverslip.

Western blot analysis

MOR1-expressing and wild-type HEK 293 cells were plated onto poly-L-lysine-treated 150-mm-diameter dishes and grown to 80% confluence. Cells were then washed twice with phosphate-buffered saline and harvested into ice-cold lysis buffer [10 mM Tris-HCl (pH 7.6), 5 mM EDTA, 3 mM EGTA, 250 mM sucrose, 0.2 mM phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, 1 μg/ml pepstatin A, 1 μg/ml aprotinin, and 10 μg/ml bacitracin]. Cells were swollen for 15 min on ice and homogenized. The homogenate was spun at 500 g for 5 min at 4°C to remove unbroken cells and nuclei. Membranes were then pelleted at 20,000 g for 30 min at 4°C, and pelleted membranes were lysed in detergent buffer [20 mM HEPES (pH 7.4), 150 mM NaCl, 5 mM EDTA, 3 mM EGTA, 4 mg/ml β-dodecyl maltoside, 0.2 mM phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, 1 μg/ml pepstatin A, 1 μg/ml aprotinin, and 10 μg/ml bacitracin] for 1 h on ice. The lysate was centrifuged at 20,000 g for 30 min at 4°C. One milliliter of the supernatant was incubated with 100 μl of wheat germ agglutinin agarose beads (Pharmacia, Freiburg, Germany) for 90 min at 4°C with continuous agitation. Collected beads were washed five times with detergent buffer, and adsorbed glycoproteins were eluted into 200 μl of sodium dodecyl sulfate (SDS)-sample buffer [62.5 mM Tris-HCl (pH 6.8), 2% SDS, 20% glycerol, 200 mM DL-dithiothreitol, and 0.005% bromphenol blue] at 60°C for 20 min. Equal volumes of these fractions were subjected to 8% SDS-polyacrylamide gel electrophoresis (PAGE) and electroblotted onto nitrocellulose membranes. After blocking with 5% low-fat dried milk dissolved in phosphate-buffered saline containing 0.1% Tween-20, membranes were incubated for 12 h either with rabbit polyclonal anti-MOR1 antibody (9998) at a dilution of 1:5,000 or with a rat monoclonal anti-HA antibody at a concentration of 0.5 μg/ml. The anti-MOR1 antibody (9998) was generated to the peptide LENLEAETAPLP, corresponding to residues 387-398 of the carboxyl terminus of MOR1, and has been characterized extensively (Koch et al., 1998; Schulz et al., 1998). Bound primary antibodies were detected with peroxidase-labeled secondary antibodies, and immunoreactive proteins were visualized using an enhanced chemiluminescence detection system as described (Iwata et al., 1995).

Agonist-induced phosphorylation of μ-opioid receptor

HEK 293 cells were plated at a density of 5 × 106 cells per poly-L-lysine-treated 100-mm-diameter plate. After 48 h, cells were washed three times with serum-free, phosphate-free Dulbecco’s modified Eagle’s medium (GibcoBRL, Karlsruhe, Germany) and labeled with 300 μCi/ml carrier-free [32P]orthophosphate (285 Ci/mg of Pi; ICN, Eschwege, Germany) for 30 min. Either PD98059 in a concentration of 0, 1, 5, 25, or 100 μM or U0126 in a concentration of 0, 1, 10, 100, or 1,000 nM was then added, and incubation was continued for another 60 or 15 min, respectively. Labeled cells were exposed to 0, 1, or 10 μM DAMGO for 20 min in the presence or absence of 1 μM naloxone. After incubation, cells were placed on ice and washed three times with ice-cold phosphate-buffered saline. Proteins were extracted with 1.5 ml of RIPA buffer [50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 5 mM EDTA, 10 mM NaF, 10 mM disodium pyrophosphate, 0.1 mM sodium orthovana-date and 10 nM calyculin A (Alexis, Grünwald, Germany) as phosphatase inhibitors as well as proteinase inhibitors as described above] for 60 min on ice. All subsequent procedures were carried out at 4°C. After centrifugation at 20,000 g for 30 min, lysates were precleared with 50 μl of protein G Plusagarose beads (Calbiochem, Bad Soden, Germany) for 2 h. HA-tagged μ-receptors were then immunoprecipitated from 1 ml of the supernatant using 2.5 μg of rat monoclonal anti-HA antibody for 12 h. Immune complexes were collected using 100 μl of protein G Plus-agarose for 3 h. Beads were washed five times with RIPA buffer, and immunoprecipitates were eluted from the beads with 150 μl of SDS-sample buffer at 60°C for 20 min. Protein content of each sample was determined before preclearing using the BCA protein assay (Pierce, Rockford, IL, U.S.A.). These values were normalized and used to adjust gel loading volumes. Samples were then subjected to 8% SDS-PAGE, and gels were dried and autoradiographed with BioMax MR film (Kodak). Opioid receptor phosphorylation was quantitatively analyzed using a BAS 1000 PhosphorImager (Fuji). The mean ± SEM values are expressed for results obtained from three independent experiments. Statistical analysis of results was performed using ANOVA followed by the Bonferroni test.

Phosphorylation of the μ-opioid receptor by exogenous MAPK in vitro

Phosphorylation of MOR1 after immunoprecipitation from stably transfected HEK 293 cells by anti-HA antibody was carried out with the use of exogenous, activated ERK 2 by a protocol based on that described by Yang et al. (1997). In brief, MOR1-expressing and wild-type HEK 293 cells were grown in 150-mm-diameter dishes to 80% confluence. Cell membranes were prepared, and membrane proteins were extracted as described above. μ-Receptor proteins were immunoprecipitated from 1 mg of membrane proteins (∼2,000 fmol) using 2.5 μg of anti-HA antibody and 100 μl of protein G Plus-agarose in a total volume of 1 ml overnight at 4°C. Immunoprecipitates were washed three times with lysis buffer and twice with kinase buffer [50 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 1 mM EGTA, and 2 mM DL-dithiothreitol]. Beads were suspended in 30 μl of kinase buffer supplemented with 100 μM ATP and 100 μCi/μmol [γ-32P]ATP (3,000 Ci/mmol; NEN, Boston, MA, U.S.A.) with or without 1 U of activated ERK 2 and incubated for 30 min at 30°C. The reaction was stopped by addition of 100 μl of SDS-sample buffer and heated for 15 min at 60°C. Samples were subjected to 8% SDS-PAGE, and gels were dried and autoradiographed with BioMax MR film (Kodak). Reaction mixtures that contained 3 μg of myelin basic protein (MBP; Sigma) instead of μ-opioid receptor immuno-precipitates were used as the standard for phosphorylation assays. Samples containing phosphorylated MBP were electrophoresed separately on 15% SDS-polyacrylamide gels.

RESULTS

Rapid and internalization-independent MAPK activation by MOR1

To examine the time course of MAPK activation in response to μ-opioid receptor stimulation, stably transfected HEK 293 cells were exposed to either DAMGO or morphine in the presence or absence of sucrose, and the subcellular distributions of phosphorylated MAPK and receptor endocytosis were determined by immunocyto-chemistry. As shown in Fig. 1 (top row), DAMGO induced a sharp rise in phospho-MAPK-like immunore-activity (Li) within 3-5 min. Phospho-MAPK-Li was seen as punctuate staining predominantly localized to the nucleus. After 10 min, cellular levels of phosphorylated MAPK returned to control levels. DAMGO also promoted MOR1 endocytosis, which peaked at 30 min (Fig. 1, top row, rightmost panels; see also Koch et al., 1998). It is interesting that morphine stimulated a MAPK activation that was indistinguishable from that seen after DAMGO exposure. In contrast to DAMGO, however, morphine failed to promote receptor internalization (Fig. 1). Moreover, when cells were exposed to DAMGO in the presence of sucrose, an inhibitor of clathrin-mediated endocytosis (Koch et al., 1998; Wolf et al., 1999), an even more sustained MAPK activation was detectable, but receptor internalization was completely prevented (Fig. 1). These results demonstrate that agonist exposure of the μ-opioid receptor stimulated a rapid and transient MAPK activation and that activated MAPK predominantly localizes to the nucleus. Redistribution of the μ-receptor protein from plasma membrane into the endosomal compartment was not required for MAPK activation.

Figure 1.

Rapid and internalization-independent MAPK activation mediated by μ-opioid receptor stimulation. HEK 293 cells stably expressing MOR1 were exposed to 1μM DAMGO (top row), 10 μM morphine (middle row), or 1 μM DAMGO (bottom row) in the presence of 0.4 M sucrose for 0, 3, 5, 7, 10, or 30 min. Left panels: Cells were subsequently fixed, fluorescently labeled with anti-phospho-MAPK antibodies, and examined by confocal microscopy. Rightmost panels: Cells were subsequently fixed, fluorescently labeled with anti-MOR1 antibodies, and examined by confocal microscopy. Note that all three stimuli induced a rapid activation and nuclear translocation of MAPK; however, only DAMGO alone but not morphine or DAMGO plus sucrose promoted a concomitant endocytosis of the μ-opioid receptor. Representative results are shown from one of three independent experiments performed in duplicate. Bar = 20 μm.

FIG. 1.

We also examined the effect of specific inhibitors of the MAPK cascade on the opioid-mediated MAPK recruitment. When the cells were exposed to the MAPK kinase (MEK) 1/2 inhibitor PD98059 (Alessi et al., 1995; Dudley et al., 1995), at 20 μM before and during DAMGO exposure, MAPK activation was greatly inhibited (Fig. 2). A second, chemically unrelated inhibitor of the MAPK pathway, U0126 (DeSilva et al., 1998; Duncia et al., 1998; Favata et al., 1998), produced a similar effect when administered at a concentration of 100 nM (Fig. 2).

Figure 2.

Inhibition of MOR1-mediated MAPK recruitment by PD98059 and U0126. HEK 293 cells stably expressing MOR1 were exposed to 1 μM DAMGO for 0, 3, or 5 min in the absence (top row) or presence of 20 μM PD98059 (middle row) or 100 nM U0126 (bottom row). PD98059 and U0126 were added 60 and 15 min before DAMGO stimulation, respectively. Note that Figs. 1 and 2 show results from experiments conducted in parallel. Therefore, the same micrographs are shown in Fig. 1, top row, and in Fig. 2, top row. Representative results are shown from one of three independent experiments performed in duplicate. Bar = 20 μm.

FIG. 2.

Inhibition of the MAPK cascade diminishes agonist-induced desensitization of the μ-opioid receptor

Next, we determined the effect of PD98059 and U0126 on MOR1 desensitization. HEK 293 cells stably expressing MOR1 were preincubated with either 0 or 1 μM DAMGO, and the culture medium was either not supplemented or supplemented with 20 μM PD98059 or 100 nM U0126. Subsequently, the DAMGO-induced inhibition of forskolin-simulated adenylyl cyclase activity was determined by quantifying intracellular cAMP. As shown in Fig. 3, after 2 h of DAMGO preincubation, agonist responsiveness was reduced to 9% compared with cells that had not been preexposed to DAMGO. In contrast, agonist responsiveness was retained to 42 and 76% when cells were exposed to DAMGO in the presence of PD98059 and U0126, respectively. PD98059 and U0126 had no effect on basal cAMP or forskolin-stimulated cAMP levels in untransfected cells (data not shown). This result suggests that agonist-induced desensitization of the μ-opioid receptor was largely inhibited when MAPK recruitment was prevented by PD98059 and U0126.

Figure 3.

Inhibition of DAMGO-induced desensitization of MOR1 by PD98059 and U0126. HEK 293 cells stably expressing MOR1 were preincubated with either 0 or 1 μM DAMGO for 2 h in the presence or absence of 20 μM PD98059 or 100 nM U0126. After removal of agonist-containing medium, cells were treated with either 25 μM forskolin or 25 μM forskolin plus 1 μM DAMGO for 15 min, and cAMP levels were determined using a radioimmunoassay. Data are mean ± SE (bars) values of triplicate determinations from two independent experiments. DAMGO-induced inhibition of intracellular cAMP in cultures that had not been preexposed to DAMGO was defined as 100%. PD98059 and U0126 had no effect on basal or forskolin-stimulated intracellular cAMP levels, and DAMGO pretreatment also did not alter these responses. *p<0.05, significant difference in agonist responsiveness of DAMGO-preincubated cells that had been cultured in the presence or absence of PD98059 or U0126 as determined using ANOVA followed by the Bonferroni test.

FIG. 3.

Inhibition of the MAPK cascade attenuates agonist-induced phosphorylation of the μ-opioid receptor

As phosphorylation of serine and threonine residues within intracellular receptor domains is an early step during receptor desensitization, we explored the effects of inhibitors of MAPK activation on agonist-induced phosphorylation of the μ-opioid receptor. When membrane preparations from wild-type HEK 293 cells or HA-tagged MOR1-expressing HEK 293 cells were subjected to western blot analysis, the anti-MOR1 antibody directed against the carboxy-terminal tail of the μ-receptor detected a broad band migrating at a molecular mass corresponding to 65-80 kDa only in stably transfected but not in wild-type cells (Fig. 4A). A similar band was seen when the anti-HA antibody directed against the amino-terminal added tag was used (data not shown). In addition, a second band at ∼140-160 kDa was noted with both antibodies. This second band may represent a dimeric form of the receptor as has previously been shown for the δ and κ-opioid receptors (Cvejic and Devi, 1997; Jordan and Devi, 1999). To determine receptor phosphorylation, we immunoprecipitated epitope-tagged receptor proteins from agonist-stimulated cells that had been metabolically labeled with 32Pi. As shown in Fig. 4B, 1 μM DAMGO stimulated a phosphorylation of the μ-opioid receptor to a degree approximately threefold greater than that seen in the absence of agonist within 20 min. No further increase in phosphorylation was observed when 10 μM DAMGO was added or DAMGO exposure was extended as long as 60 min. The DAMGO-induced increase in phosphorylation was completely blocked by the opioid receptor antagonist naloxone. Only basal phosphorylation was observed in the presence of naloxone.

Figure 4.

Agonist-induced phosphorylation of the μ-opioid receptor. A: Membrane preparations from wild-type (WT; lane 1) and MOR1-expressing (lane 2) HEK 293 cells were separated on 8% SDS-PAGE and blotted onto nitrocellulose. Membranes were then incubated with anti-MOR1 antiserum (9998) at a dilution of 1:5,000, and immunoreactive proteins were detected using enhanced chemiluminescence. B: Stably transfected HEK 293 cells labeled with 32Pi were stimulated with 0, 1, or 10 μM DAMGO in the presence or absence of 1 μM naloxone for 20 min. Cells were lysed, epitope-tagged μ-receptors were immunoprecipitated with rat monoclonal anti-HA antibody, and immunoprecipitates were resolved on 8% SDS-PAGE. Gels were dried and subjected to autoradiography. Note that the μ-opioid receptor migrates as a broad band with a molecular mass of 65-80 kDa and that a band migrating at the identical molecular mass range becomes rapidly phosphorylated in response to DAMGO stimulation, an effect that is completely blocked in the presence of naloxone. The ordinate is migration of protein molecular mass markers. Depicted is a representative result of three independent experiments performed using the same method.

FIG. 4.

Next, we determined agonist-induced phosphorylation of the μ-receptor under conditions when activation of the MAPK cascade was prevented by the specific MEK1/2 inhibitor PD98059 (Alessi et al., 1995; Dudley et al., 1995). When cells were pretreated with various concentrations of PD98059 and DAMGO-induced phosphorylation was assessed, we observed a dose-dependent inhibition of μ-opioid receptor phosphorylation (Fig. 5). Whereas phosphorylation levels were not changed in the presence of 1 μM PD98059, the DAMGO-stimulated increase in phosphorylation was inhibited by 50% after treatment with 5 or 25 μM PD98059. Agonist-induced phosphorylation was completely blocked by 100 μM PD98059. Although PD98059 potently reduced the rate of agonist-specific phosphorylation, it did not change the level of basal phosphorylation as determined by incubation of the cells with various concentrations of PD98059 in the absence of DAMGO. Moreover, the novel MEK1/2 inhibitor U0126 (DeSilva et al., 1998; Duncia et al., 1998; Favata et al., 1998), which also inhibits activation of ERK1 and ERK2, induced a similar dose-dependent inhibition of DAMGO-stimulated phosphorylation of the μ-opioid receptor with an IC50 of ∼100 nM (data not shown). This concentration of U0126 corresponds well to its IC50 value for inhibition of MAPK activation observed in other cell systems (DeSilva et al., 1998; Scherle et al., 1998).

Figure 5.

Inhibition of agonist-promoted phosphorylation of the μ-opioid receptor by PD98059. A: Stably transfected HEK 293 cells were treated with 0, 1, 5, 25, or 100 μM PD98059 (lanes 2-6; see B) metabolically labeled with 32Pi, and either stimulated with 1 μM DAMGO or not stimulated (lane 1) for 20 min. Cells were lysed, epitope-tagged μ-receptors were immunoprecipitated with rat monoclonal anti-HA antibody, and immunoprecipitates were resolved on 8% SDS-PAGE. Gels were dried and subjected to autoradiography. The autoradiogram shown is representative of three independent experiments. The ordinate is migration of protein molecular mass markers. B: Receptor phosphorylation was quantitatively analyzed using a Phosphorlmager and expressed as fold increase over basal phosphorylation seen in the absence of agonist. Data are mean ± SE (bars) values of three separate experiments. *p < 0.05, significant inhibition of agonist-induced phosphorylation compared with control values as determined using ANOVA followed by the Bonferroni test.

FIG. 5.

Phosphorylation of intracellular receptor domains is believed to precede receptor internalization. To test the hypothesis that PD98059 would not only inhibit agonist-induced phosphorylation but also agonist-induced internalization, we treated MOR1-expressing HEK 293 cells with various concentrations of PD98059 and determined the extent of receptor internalization using confocal microscopy. As shown in Fig. 6 (top row, left panel), in the absence of DAMGO MOR1-Li was almost exclusively confined to the plasma membrane. After a 45-min agonist exposure, a dramatic loss of MOR1-Li from the plasma membrane with a concomitant accumulation in vesicle-like structures within the cytoplasm was observed (Fig. 6). When DAMGO-induced internalization was determined in PD98059-treated cells, a dose-dependent inhibition of μ-receptor sequestration was found (Fig. 6). Although 1 μM PD98059 did not influence internalization, the rate of intracellular trafficking of the μ-receptor progressively decreased with increasing concentrations of PD98059. After administration of 50 μM PD98059, the inhibition of μ-opioid receptor sequestration was nearly complete. U0126 also mimicked the effects of PD98059 on μ-opioid receptor endocytosis.

Figure 6.

Inhibition of agonist-induced internalization of the μ-opioid receptor by PD98059. Left panels:μ-Opioid receptor-expressing HEK 293 cells were either stimulated with 1 μM DAMGO for 45 min (bottom row) or not stimulated (top row). Right panels:μ-Opioid receptor-expressing HEK 293 cells were exposed to 1, 5, 25, or 50 μM PD98059 and stimulated with 1 μM DAMGO for 45 min. Cells were fluorescently labeled with anti-HA antibodies and examined by confocal microscopy. Representative results are shown from two independent experiments performed in duplicate. Bar = 20 μm.

FIG. 6.

The μ-opioid receptor is a poor substrate for MAPK

To investigate the possibility that MAPK could directly phosphorylate the μ-opioid receptor, we immunoprecipitated the receptor protein and incubated the immunoprecipitates derived from stably transfected and wild-type HEK cells with exogenous, activated MAPK in vitro. When the enzymatic reaction was stopped by adding SDS-sample buffer and the reaction mixtures were resolved using SDS-PAGE, it became apparent that no specific radioactivity was incorporated into μ-receptor protein. In contrast, MBP, which is known to be a good MAPK substrate, was readily phosphorylated under otherwise identical conditions (Fig. 7).

Figure 7.

In vitro phosphorylation of MOR1 by exogenous MAPK. MOR1 was isolated by immunoprecipitation from stably transfected HEK 293 cells as described in Experimental Procedures. A: Immunoprecipitates prepared from either MOR1-expressing or wild-type (wt) HEK 293 cells were incubated either with or without activated ERK 2 in the presence of [γ-32P] ATP for 30 min at 30°C. The reaction was stopped by adding SDS-sample buffer, and samples were subjected to 8% SDS-PAGE and visualized by autoradiography. B: MBP was used instead of immunoprecipitates under otherwise identical conditions. Samples were subjected to 15% SDS-PAGE and visualized by autoradiography. Representative autoradiograms are shown from two independent experiments performed using the same method. The ordinate is migration of protein molecular mass markers.

FIG. 7.

DISCUSSION

Activation of the MAPK pathway in response to agonist stimulation has been observed for many GPCRs in heterologous expression systems. In the present work, we show that DAMGO exposure of HEK 293 cells stably transfected to express MOR1 stimulates a rapid and transient activation of the MAPK pathway. Activated MAPK, in turn, translocates into the nucleus and promotes phosphorylation of transcription factors. Morphine, which is a major clinically used alkaloid analgesic drug, appears to differ fundamentally from opioid peptides in that it functionally coupled to the MAPK pathway but did not promote receptor endocytosis. In fact, redistribution of μ-receptor protein from the plasma membrane into the endosomal compartment was not required for opioid stimulation of MAPK activity, as evidenced by our finding that sucrose only prevented DAMGO-induced internalization but not MAPK recruitment.

For several GPCRs, including the δ-opioid receptor, MAPK activation has been found to require at least functional β-arrestin and dynamin. Overexpression of dominant-negative mutants of β-arrestin or dynamin results in an inhibition of both agonist-induced MAPK stimulation and receptor endocytosis (Daaka et al., 1998; Della Rocca et al., 1999; Ignatova et al., 1999). This finding has led to the conclusion that the internalized receptor may serve as the scaffold for the initiation of the mitogenic signal (Ahn et al., 1999; Della Rocca et al., 1999; Luttrell et al., 1999). However, the requirement of receptor endocytosis for MAPK activation seems not to be a general property of all GPCRs and all cellular systems, e.g., the κ-opioid receptor and the α2-adrenergic receptor require dynamin and arrestin only for endocytosis but not for MAPK activation (DeGraff et al., 1999; Li et al., 1999). Similarly, for the μ-opioid receptor the present data as well as several previous studies clearly show that the morphine-activated μ-receptor stimulates mitogenic signaling (Li and Chang, 1996; Chuang et al., 1997; Polakiewicz et al., 1998) but not receptor internalization (Arden et al., 1995; Keith et al., 1996; Sternini et al., 1996; Whistler and von Zastrow, 1998; Zhang et al., 1998).

The opioid-mediated MAPK recruitment not only represents a mitogenic signal but also a desensitizing signal for the μ-opioid receptor. Homologous desensitization involves phosphorylation of intracellular receptor domains and subsequent sequestration of phosphorylated receptors into the endosomal compartment. We show that both agonist-induced desensitization and phosphorylation were inhibited in the presence of PD98059 as well as the novel MEK1/2 inhibitor U0126. We also demonstrate that agonist-promoted internalization of the μ-opioid receptor is largely blocked when MAPK activation was prevented by administration of PD98059. We therefore conclude that this rapid agonist-promoted MAPK recruitment may be an initial event during μ-receptor desensitization preceding the phosphorylation and internalization of the receptor protein.

The simplest explanation for our findings would be that activated MAPK directly phosphorylates the μ-opioid receptor. Evidence for direct receptor phosphorylation by MAPK has recently been presented for the AT1 angiotensin II receptor, which contains a MAPK recognition sequence (Yang et al., 1997). Universal recognition motifs for MAPK are P-X-S/T-P and X-X-S/T-P [single-letter code; X, any amino acid (Jacobs et al., 1999)]. Because Thr180 in the second intracellular loop of the μ-opioid receptor, which is localized within the sequence motif 178FRTP181, represents a putative MAPK phosphorylation site, we performed in vitro phosphorylation assays using immunoprecipitated μ-opioid receptors and exogenous ERK2. However, both the lack of any detectable incorporation of 32P and the predominant perinuclear and nuclear localization of activated MAPK eliminate the possibility that the μ-opioid receptor may serve as substrate for MAPK in opioid-stimulated cells.

Putative phosphorylation sites within the μ-receptor have been identified for second messenger kinases, e.g., serine residues 261/266 in the third intracellular loop for calcium/calmodulin-dependent kinase II (Koch et al., 1997), as well as for GPCR kinases (GRKs), e.g., Thr394 at the carboxyl terminus for GRK2 (Pak et al., 1997). It is thus not unreasonable to speculate that MAPK may either directly or indirectly activate other second messenger kinases or GRKs that in turn phosphorylate the μ-opioid receptor. A direct inhibitory effect of PD98059 on calcium/calmodulin-dependent kinase II or GRKs can be ruled out because the IC50 values determined in the present study were <20 μM. At this concentration PD98059 has previously been shown to be highly specific in preventing MAPK activation, without affecting the activity of 18 other serine/threonine kinases tested (Alessi et al., 1995; Dudley et al., 1995). Similarly, U0126, which targets MEK1 and MEK2, inhibited agonist-induced phosphorylation with an IC50 of ∼100 nM. This compound appears to be highly selective for MAPK inhibition even at concentrations as high as 100 μM (DeSilva et al., 1998; Duncia et al., 1998; Favata et al., 1998).

Desensitization of the μ-opioid receptor has been mainly described as the attenuated reduction of forskolin-stimulated cAMP levels in response to the agonist. It is interesting that μ-receptor signaling via the MAPK cascade is also desensitized on prolonged exposure to agonist in transfected cells (Polakiewicz et al., 1998). In this context it should be noted that the MAPK cascade undergoes neuroadaptation during chronic opioid exposure in vivo. We have recently determined the activation state of MAPK using phosphospecific antibodies in morphine-dependent rats. Neuronal MAPK activity was potently repressed after repeated morphine administration in virtually all brain regions examined (Schulz and Höllt, 1998).

In summary, we have shown that exposure of the μ-opioid receptor to both peptidic and alkaloid agonists, including morphine, stimulates a rapid and transient MAPK activation. Unlike the δ-opioid receptor, the μ-opioid receptor does not require internalization for mitogenic signaling. We also show that an active MAPK pathway is essential for μ-receptor phosphorylation and internalization. Initial recruitment of MAPK in response to opioid stimulation could thus be a key signaling event not only for the transduction of mitogenic signals to the nucleus but also for the initiation of homologous desensitization of the μ-opioid receptor. Both mitogenic and desensitizing signals emanating from the MAPK cascade may contribute to long-term cellular effects of opioids, including the development of tolerance.

Acknowledgements

We thank Dana Wiborny, Dora Nüß, and Evelyn Raulf for excellent technical assistance. This work was supported by grant SCHU 924/4-1 (to S.S.) from the Deutsche Forschungsgemeinschaft, grant I/75 172 (to S.S.) from the Volkswagen-Stiftung, grant 1908A/0025 (to S.S.) from the Kultusministerium des Landes Sachsen/Anhalt, and a grant from the Fonds der Chemischen Industrie (to V.H.).

Ancillary