C-terminal Splice Variants of the Mouse (cid:1) -Opioid Receptor Differ in Morphine-induced Internalization and Receptor Resensitization*

The main analgesic effects of the opioid alkaloid morphine are mediated by the (cid:1) -opioid receptor. In contrast to endogenous opioid peptides, morphine activates the (cid:1) -opioid receptor without causing its rapid endocytosis. Recently, three novel C-terminal splice variants (MOR1C, MOR1D, and MOR1E) of the mouse (cid:1) -opioid receptor (MOR1) have been identified. In the present study, we show that these receptors differ substantially in their agonist-selective membrane trafficking. MOR1 and MOR1C stably expressed in human embryonic kidney 293 cells exhibited phosphorylation, internalization, and down-regulation in the presence of the opioid peptide [ D -Ala 2 ,Me-Phe 4 ,Gly 5 -ol]enkephalin (DAMGO) but not in response to morphine. In contrast, MOR1D and MOR1E exhibited robust phosphorylation, internalization, and down-regulation in response to both DAMGO and morphine. DAMGO elicited a similar desensitization (during an 8-h exposure) and resensitization (during a 50-min drug-free interval) of all four (cid:1) -recep-tor splice variants. After morphine treatment, however, MOR1 and MOR1C showed a faster desensitization and no resensitization as compared with MOR1D and an N-terminal HA epitope tag, the forward primer 5 (cid:2) -CGT GAA AAG CTT ACC ATG TAC CCA TAC GAC GTC CCA GAC TAC GCT GAC AGC AGC-3 (cid:2) was synthesized and used for polymerase chain reaction amplification of all receptor splice variants. For construction of the mouse (cid:1) -opioid receptor splice variants MOR1D and MOR1E and in- troduction of an Xba I restriction site, polymerase chain reaction primers were designed based on the published DNA sequences of the vari- ants (15). The sequence for the reverse MOR1D and MOR1E primers were 5 (cid:2) -T (cid:2) and 5 (cid:2) , respectively. Generation of Cell Lines Expressing (cid:1) -Opioid Receptors— Transfec- tion of HEK 293 cells was performed by the calcium phosphate precipitation method (19). Approximately 1.5 (cid:3) 10 6 cells were transfected with 20 (cid:1) g of plasmid DNA. Cells were selected in the presence of 1 (cid:1) g/ml puromycin (Sigma), and resistant cells were grown in the pres- ence of 1 (cid:1) g/ml puromycin. Three to six clones were generated for each of the four (cid:1) -receptor isoforms. Receptor expression was monitored using saturation ligand binding assays and quantitative Western blot analysis as described below; clones expressing similar numbers of re- ceptors cells of (1 of of ethanol). After centrifugation, the supernatant was evapo-rated, the residue was dissolved in Tris-EDTA buffer, and the cAMP content was determined by using a commercially available radioimmu-noassay kit (Amersham Pharmacia Biotech). Statistical evaluation of results was carried out by using analysis of variance followed by the Bonferroni test. Radioligand Binding Assays— Binding studies were performed on membranes prepared from stably transfected cells. The dissociation constant ( K D ) and the number of [ 3 H]DAMGO binding sites ( B max ) were calculated by Scatchard using at least six concentrations

Opioid receptors couple via G proteins to a variety of downstream effectors including adenylate cyclase (1) and mitogenactivated protein kinases (2)(3)(4)(5). During repeated or continuous agonist stimulation these responses undergo rapid desensitization. An important mechanism of desensitization of G proteincoupled receptors is the phosphorylation of intracellular receptor domains by G protein-coupled receptor kinases or second messenger-regulated protein kinases such as Ca 2ϩ /calmodulindependent protein kinase II, cAMP-dependent protein kinase, or protein kinase C. After phosphorylation of the receptor, arrestins are frequently recruited to the plasma membrane, at which they facilitate endocytosis by serving as scaffolding proteins that bind to clathrin (6,7). For some time it has been assumed that the rapid removal of ligand-activated receptors from the cell surface plays a major role in the receptor degradation, thus enhancing functional desensitization. Recent studies have suggested that endocytozed receptors are predominantly recycled to the cell surface in a reactivated state (8,9).
We have shown previously that two C-terminal splice variants (rMOR1 and rMOR1B) 1 of the rat -opioid receptor differ in their DAMGO-mediated internalization and resensitization rates (10,11). The rapid internalizing variant rMOR1B revealed a faster resensitization and consequently a slower desensitization as compared with rMOR1. These results clearly show that receptor recycling after internalization affects the rate of agonist-induced desensitization of the -opioid receptor. Interestingly, morphine, which is a highly addictive opioid, can activate -opioid receptors but does not induce receptor internalization (12,13). Replacement of the C-tail of the -opioid receptor by the C-tail of the ␦-opioid receptor led to a receptor chimera that is internalized after morphine treatment, indicating that the C terminus of the -receptor can influence the agonist selectivity of endocytosis (14).
Recently, three novel splice variants (MOR1C, MOR1D, and MOR1E) were identified for the mouse -opioid receptor (15). The differences between MOR1 and the new variants are restricted to the terminal portion of the intracellular tail (MOR1, 387 LENLEAETAPLP 398 ; MOR1C, 387 PTLAVSVAQIFTGYP-SPTHVEKPCKSCMDRGMRNLLPDDGPRQESGEGQLGR 438 ; MOR1D, 387 RNEEPSS 393 ; and MOR1E, 387 KKKLDSQRGC-VQHPV 401 ). Immunohistochemical studies revealed marked differences in the regional distribution of the -opioid receptor splice variants in the central nervous system (16 -18). In the present study, we investigated whether these splice variants differ in their DAMGO-and morphine-induced phosphorylation, internalization, desensitization, and receptor resensitization. Our results provide strong evidence that DAMGO treatment resulted in a rapid internalization and similar desensitization and resensitization rates of all receptor splice variants, whereas after morphine treatment only the splice variants MOR1D and MOR1E showed receptor phosphorylation and internalization, a faster resensitization, and a slower desensitization as compared with MOR1 and MOR1C. EXPERIMENTAL restriction site and  an N-terminal HA epitope tag, the forward primer 5Ј-CGT GAA AAG  CTT ACC ATG TAC CCA TAC GAC GTC CCA GAC TAC GCT GAC  AGC AGC-3Ј was synthesized and used for polymerase chain reaction amplification of all receptor splice variants. For construction of the mouse -opioid receptor splice variants MOR1D and MOR1E and introduction of an XbaI restriction site, polymerase chain reaction primers were designed based on the published DNA sequences of the variants (15). The sequence for the reverse MOR1D and MOR1E primers were 5Ј-T TCC TGT CTA GAG CCA TCA TCA GGA AGA AGG TTC  CTC ATT CCT CTG GTG GTT AG-3Ј and 5Ј-AG ACA ATC TAG AGG  TCA CAC TGG ATG CTG TAC ACA CCC TCT CTG CGA GTC CAG  CTT TTT CTT CTG GTG GTT AG-3Ј, respectively. Generation of Cell Lines Expressing -Opioid Receptors-Transfection of HEK 293 cells was performed by the calcium phosphate precipitation method (19). Approximately 1.5 ϫ 10 6 cells were transfected with 20 g of plasmid DNA. Cells were selected in the presence of 1 g/ml puromycin (Sigma), and resistant cells were grown in the presence of 1 g/ml puromycin. Three to six clones were generated for each of the four -receptor isoforms. Receptor expression was monitored using saturation ligand binding assays and quantitative Western blot analysis as described below; clones expressing similar numbers of receptors were selected for further study. The B max and K D values are given in Table I.
Determination of Receptor Desensitization and Resensitization by Measurement of cAMP Accumulation-Approximately 1.5 ϫ 10 5 cells/ well were seeded in 22-mm 12-well dishes with Dulbecco's modified Eagle's medium nut F-12 medium containing 10% fetal calf serum. After 24 h, the cells were exposed to 1 M DAMGO (Bachem, Heidelberg, Germany) or 1 M morphine (Synopharm, Barsbü ttel, Germany) for 0, 0.5, 1, 2, 4, 6, or 8 h. For resensitization assays, cells were washed after 8 h of DAMGO or morphine exposure followed by an additional incubation period of either 0, 10, 20, 30, 40, or 50 min in the absence of agonist. For the measurement of cAMP accumulation, the medium was removed from individual wells and replaced by 0.5 ml of serum-free RPMI 1640 medium (Seromed, Berlin, Germany) containing 25 M forskolin (Biotrend, Köln, Germany) or 25 M forskolin plus 1 M DAMGO or 1 M morphine. 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 N HCl/100 volumes of ethanol). After centrifugation, the supernatant was evaporated, the residue was dissolved in Tris-EDTA buffer, and the cAMP content was determined by using a commercially available radioimmunoassay kit (Amersham Pharmacia Biotech). Statistical evaluation of results was carried out by using analysis of variance followed by the Bonferroni test.
Radioligand Binding Assays-Binding studies were performed on membranes prepared from stably transfected cells. The dissociation constant (K D ) and the number of [ 3 H]DAMGO binding sites (B max ) were calculated by Scatchard analysis using at least six concentrations of [ 3 H]DAMGO in a range from 0.3 to 9 nM as described previously (10). Nonspecific binding was determined as radioactivity bound in the presence of 1 M unlabeled DAMGO.
Confocal Microscopy-HEK 293 cells stably expressing the mouse -opioid receptor splice variants were grown on poly-L-lysine-treated coverslips overnight. After washing, the cells were incubated with 1 g/ml affinity-purified polyclonal rabbit anti-HA-tag antibody (Gramsch Laboratories, Schwabhausen, Germany) for 2 h at 4°C to label cell surface receptors. The cells were subsequently exposed to 1 M DAMGO or 1 M morphine for 0, 5, 10, 20, or 30 min at 37°C to induce receptor endocytosis. The cells were then fixed with 4% paraformaldehyde and 0.2% picric acid in phosphate buffer, pH 6.9, for 40 min at room temperature and washed several times in 10 mM Tris, 10 mM phosphate buffer, 137 mM NaCl, and 0.05% thimerosal, pH 7.4 (TPBS). Specimens were incubated for 3 min in 50% methanol and 3 min in 100% methanol, washed several times in TPBS, and preincubated with TPBS and 3% normal goat serum for 1 h at room temperature. Bound primary antibody was detected with biotinylated secondary antibodies (1:100; Vector, Burlingame, CA) followed by cyanine 3.18-conjugated streptavidin (Amersham Pharmacia Biotech). The cells were then dehydrated, cleared in xylol, and permanently mounted in DPX (Fluka, Neu-Ulm, Germany). Specimens were examined using a Leica TCS-NT laserscanning confocal microscope. Cyanine 3.18 was imaged with 568-nm excitation and 570 -630-nm bandpass emission filters. Confocal micrographs were taken by a person blinded to the treatments who was instructed to randomly select one colony of 4 -12 cells/coverslip.
Western Blot Analysis-Cells were plated onto poly-L-lysine-coated 150-mm dishes and grown to 80% confluence. When indicated, the cells were exposed to 1 M DAMGO, 1 M morphine, or 1 M naloxone for 0, 0.5, 2, 4, or 16 h. The cells were then washed twice with PBS and harvested into ice-cold lysis buffer (10 mM Tris-HCl, pH 7.6, 5 mM EDTA, 3 mM EGTA, 250 mM sucrose, 10 mM iodoacetamide, and the following proteinase inhibitors: 0.2 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 1 g/ml pepstatin A, 1 g/ml aprotinin, and10 g/ml bacitracin). Iodoacetamide was included in each buffer used for protein preparation to prevent nonspecific disulfide linkages. The cells were swollen for 15 min on ice and homogenized. The homogenate was spun at 500 ϫ g for 5 min at 4°C to remove unbroken cells and nuclei. Membranes were then pelleted at 20,000 ϫ 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 ␤-dodecylmaltoside, 10 mM iodoacetamide, and the proteinase inhibitors listed above) for 1 h on ice. The lysate was centrifuged at 20,000 ϫ g for 30 min at 4°C. The protein content of the resulting supernatant was determined using the BCA protein assay (Pierce). Samples containing equal amounts of protein (300 g) were then either subjected to immunoprecipitation or glycoproteins were purified using wheat germ lectin. For enrichment of glycoproteins, 1 ml of the supernatant was incubated with 100 l of wheat germ lectin-agarose beads (Amersham Pharmacia Biotech) for 90 min at 4°C with continuous agitation. Beads were washed five times with detergent buffer, and adsorbed glycoproteins were either subjected to enzymatic deglycosylation or directly eluted into 200 l of SDS-sample buffer (62.5 mM Tris-HCl, pH 6.8, 2% SDS, 20% glycerol, 100 mM DL-dithiotreitol, and 0.005% bromphenol blue) at 60°C for 20 min. Deglycosylation experiments were performed using peptide N-glycosidase F according to manufacturer protocol (New England BioLabs, Beverly, MA). After electroblotting, the membranes were incubated with 1 g/ml affinity-purified rabbit polyclonal anti-HA-tag antibody for 12 h at 4°C, followed by detection using an enhanced chemiluminescence detection system. Western blots exposed in the linear range of the x-ray film were densitometrically scanned, and the amount of immunoreactive material in each lane was quantified using NIH Image 1.57 software.
Whole-cell Phosphorylation Assays-Cells were plated onto 100-mm dishes and grown to 80% confluence. The cells were washed with serumand phosphate-free medium and then labeled with 200 Ci/ml carrierfree [ 32 P]orthophosphate (285 Ci/mg P i , ICN, Eschwege, Germany) for 60 min at 37°C. Labeled cells were exposed to either 1 M DAMGO or 1 M morphine for 20 min. After incubation, cells were placed on ice and washed with ice-cold PBS and then scraped into 1 ml of radioimmune precipitation buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 10 mM NaF, 10 mM disodium pyrophosphate, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.2 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 1 g/ml pepstatin A, 1 g/ml aprotinin, and 10 g/ml bacitracin). The cells were solubilized for 1 h at 4°C on a rotating platform. Supernatants were obtained by centrifugation at 13,000 ϫ g for 60 min at 4°C, after which aliquots were taken to determine total protein content in the supernatant of each sample to be immunoprecipitated. HA-tagged receptors were immunoprecipitated by adding 10 g of affinity-purified polyclonal rabbit anti-HA-tag antibody and 100 l of protein A-agarose for 2 h at 4°C with continuous agitation. The agarose beads were washed four times with radioimmune precipitation buffer, and receptors were eluted from the beads in 100 l of SDS sample buffer at 56°C for 20 min. In each experiment, each lane of the SDS-polyacrylamide gel was loaded with equivalent amounts of receptor protein. The amount of receptor in each sample was calculated as the function of receptor expression times the total protein content of the solubilized fraction of each sample subjected to immunoprecipitation. The receptor content of each sample was normalized to the sample with the least receptor content by dilution with sample buffer. Samples were then subjected to 8% SDS-polyacrylamide gel electrophoresis followed by autoradiography. The extent of receptor phosphorylation was quantitated using a Fuji PhosphorImager system and BAS 1000 software. The means and S.E. are expressed for values obtained from three independent experiments. Statistical evaluation of results was carried out by using analysis of variance followed by the Bonferroni test.

Functional Expression of -Opioid
Receptor Splice Variants-The -opioid receptor isoforms MOR1, MOR1C, MOR1D, and MOR1E were HA-epitope-tagged and stably expressed in HEK 293 cells. Receptor expression was monitored using saturation ligand binding assays. HEK 293 cells stably expressing the four mouse -opioid receptor splice variants revealed no markedly different numbers of binding sites (B max ) as assessed by [ 3 H]DAMGO binding. Furthermore, the affinities (K D ) for the peptide agonist DAMGO were similar for each of the receptor isoforms (Table I). In addition, the abilities of DAMGO or morphine to inhibit adenylate cyclase activity in HEK 293 cells were similar for all -receptor splice variants (Table I).
Western Blot Analysis of -Opioid Receptor Splice Variants-We then examined whether similar protein levels were expressed for all -receptor isoforms. When equal amounts of membrane proteins extracted from HEK 293 cells stably expressing the HA-tagged MOR1, MOR1C, MOR1D, or MOR1E receptors were subjected to Western blot analysis using an anti-HA antibody, a similar broad band migrating at 80 kDa was detected for each receptor isoform (Fig. 1A). Interestingly, we also observed an additional band with a higher molecular mass migrating at 160 kDa for MOR1 and MOR1C (Fig. 1A). In contrast, at this molecular mass, only very faint bands were detected for MOR1D and MOR1E (Fig. 1A). As shown in Fig.  1B, enzymatic deglycosylation of the MOR1 receptor protein reduced the size of the 80-kDa protein to 45 kDa and the size of the 160-kDa protein to 90 kDa, suggesting that this additional band may represent a -opioid receptor dimer. To further characterize the MOR1 homodimer, MOR1-expressing HEK 293 cells were exposed for extended periods (up to 16 h) to either DAMGO or the -receptor antagonist naloxone. Although DAMGO treatment resulted in a loss of MOR1 dimer and strong reduction of MOR1 monomer (46% of control), naloxone treatment resulted in an increase of both monomeric (153% of control) and dimeric (188% of control) forms of the receptor (Fig. 1B). The DAMGO-induced decrease in MOR1 receptor protein became apparent as soon as after 2 h of continuous exposure (data not shown). To exclude the possibility that this decrease in MOR1 receptor protein is not simply caused by a loss of cell surface receptors because of receptor internalization, glycoproteins were enriched not only from membrane extracts but also from the remaining cytosolic fraction and subjected to Western blot analysis. However, -receptor proteins were only detectable in membrane extracts and not in the cytosolic fractions, suggesting that the DAMGO-induced decrease of MOR1 dimers and monomers is caused by agonistinduced internalization, monomerization and lysosomal degradation leading to down-regulation of the receptor proteins. In contrast, the antagonist naloxone seems to stabilize MOR1 dimers at the plasma membrane and thus blocks receptor internalization and subsequent down-regulation of the receptor protein.
To test the possibility that the lack of detectable homodimers for MOR1D and MOR1E may be a result of constitutive internalization and subsequent monomerization of these receptors, cells stably expressing MOR1D and MOR1E were either not treated or treated with naloxone for 16 h. Again in untreated cells only monomeric forms of these receptors were detectable.
Agonist-selective Endocytosis of -Opioid Receptor Splice Variants-We asked whether the differences in the C-terminal tails of the -receptor splice variants may influence the agonist selectivity of receptor endocytosis. We therefore incubated HEK 293 cells stably expressing the N-terminal HA-tagged MOR1, MOR1C, MOR1D, or MOR1E receptors with anti-HAspecific antibodies at 4°C to label cell surface receptors. These cells were then treated with DAMGO or morphine for 30 min at 37°C. Control cells were incubated at 37°C for 30 min in the absence of agonist to examine whether these receptors may undergo constitutive or tonic internalization. The cells were subsequently fixed and permeabilized, and bound anti-HA antibody was immunofluorescently detected (Fig. 2). At low temperature (4°C), which prevents receptor endocytosis, all -opioid receptor splice variants were exclusively confined to the level of the plasma membrane as assessed by confocal microscopy (Fig. 2, Control 4°C). During a 30-min incubation at 37°C in the absence of agonist, only a proportion of MOR1D and MOR1E receptors showed constitutive internalization (Fig. 2,  Control 37°C). After 30 min of DAMGO treatment, all -opioid receptor isoforms exhibited robust endocytosis. However, only MOR1D and MOR1E appeared to be completely internalized, whereas a proportion of MOR1 and MOR1C receptors could still be detected at the plasma membrane (Fig. 2, DAMGO). After exposure to morphine, only MOR1D and MOR1E receptors revealed receptor endocytosis, whereas the MOR1 receptor was highly resistant to morphine-induced endocytosis. The MOR1C receptor revealed very little endocytosis in response to morphine (Fig. 2, Morphine).
Agonist-selective Desensitization and Resensitization of -Opioid Receptor Splice Variants-Next, we investigated whether the observed differences in receptor internalization would affect the rate of agonist-induced desensitization and resensitization of the -receptor splice variants. Receptor desensitization was measured as the decreased ability of the agonist to inhibit forskolin-stimulated adenylate cyclase activity after extended agonist pretreatment (up to 8 h). Thus, HEK 293 cells expressing MOR1, MOR1C, MOR1D, or MOR1E were pretreated with either 1 M DAMGO or 1 M morphine for various time periods followed by the determination of agonistinduced inhibition of forskolin-stimulated adenylate cyclase. Forskolin treatment resulted in a 5-fold increase of intracellular cAMP (up to 4 pmol) as compared with untreated HEK 293 cells. For each splice variant, the maximum agonist-induced inhibition of cAMP accumulation without agonist preincubation has been defined as 100%. Prolonged exposure (8 h) to DAMGO as well as to morphine led to complete desensitization of the four receptor variants (Fig. 3, A and B). Although the time courses of DAMGO-dependent loss of receptor activity were very similar for all -receptor variants (Fig. 3A), MOR1 and MOR1C showed a markedly facilitated desensitization in the presence of morphine compared with MOR1D and MOR1E (Fig. 3B). In contrast, the morphine-mediated desensitization rates of MOR1D and MOR1E were nearly identical to those obtained after DAMGO treatment. We next determined, whether the -opioid receptor splice variants differ in their resensitization rates after DAMGO-and morphine-induced receptor desensitization. Resensitization was measured as the increased ability of the previously desensitized receptor to inhibit forskolin-stimulated adenylate cyclase activity during a 50-min drug-free interval. Thus, HEK 293 cells expressing MOR1, MOR1C, MOR1D, or MOR1E were pretreated with either 1 M DAMGO or 1 M morphine for 8 h, the medium was removed, and after an additional drug-free interval of 0, 10, 20, 30, 40, or 50 min the agonist-induced inhibition of forskolin-stimulated adenylate cyclase was determined. As depicted in Fig. 3C, all splice variants completely resensitized during 50 min of agonist withdrawal after DAMGO-mediated desensitization. After morphine-mediated desensitization, the splice variants MOR1D and MOR1E showed only partial resensitization (ϳ25-30% of the maximum receptor activity) during 50 min of agonist withdrawal (Fig.  3D). Under otherwise identical conditions, however, the -opioid receptor isoforms MOR1 and MOR1C were not able to gain function after complete morphine-induced desensitization (Fig.  3D). These results suggest that the lack of morphine-induced internalization of MOR1 and MOR1C not only facilitates desensitization but also prevents resensitization of these receptors.
Agonist-selective Down-regulation of -Opioid Receptor Splice Variants-We then examined whether the observed agonist selectivity of receptor internalization would affect agonist-induced down-regulation of the -receptor splice variants. Thus, HEK 293 cells expressing MOR1, MOR1C, MOR1D, or MOR1E were pretreated with either 1 M DAMGO or 1 M morphine for 16 h, and the levels of the monomeric forms of these receptors were determined using quantitative Western blot analysis. As shown in Fig. 4, treatment with DAMGO for extended time periods resulted in a massive loss of receptor proteins of all -receptor variants. Specifically, quantitative analysis of three independent experiments revealed that the individual isoforms were down-regulated to 46% of control for MOR1, 58% for MOR1C, 10% for MOR1D, and 6% for MOR1E. In contrast, extended exposure to morphine resulted in a selective down-regulation of MOR1D (36%) and MOR1E (63%). No such down-regulation was observed after prolonged morphine treatment of MOR1 or MOR1C, suggesting that the lack of morphine-induced internalization of MOR1 and MOR1C also prevents morphine-mediated down-regulation of these receptors (Fig. 4).
Agonist-selective Phosphorylation of -Opioid Receptor Splice Variants-The -opioid receptor variants contain various lengths of sequence including serines and threonines in their cytoplasmic tails that may represent phosphoacceptor sites for G protein-coupled receptor kinases. To determine whether the observed differences in agonist-selective internalization and down-regulation may be associated with differences in agonist-induced phosphorylation of the four -receptor isoforms, we assessed whole-cell receptor phosphorylation in response to both DAMGO and morphine. DAMGO induced a rapid and robust phosphorylation of all -opioid receptor splice variants (Fig. 5A). Specifically, quantitative analysis of three independent experiments revealed that DAMGO promoted an increase in phosphorylation of ϳ4.4-fold for MOR1, ϳ2.0-fold for MOR1C, ϳ4.0-fold for MOR1D, and ϳ2.3 for MOR1E above basal levels. In contrast, the -opioid receptor isoforms markedly differed in their morphine-induced phosphorylation (Fig.  5A). Morphine promoted phosphorylation of MOR1D and MOR1E to similar levels of 111 Ϯ 4% and 87 Ϯ 9%, respectively, as compared with DAMGO-mediated receptor phosphorylation. However, the extent of the morphine-induced phosphorylation of MOR1 and MOR1C reached only 30 Ϯ 5% and 55 Ϯ 6%, respectively, of DAMGO-induced phosphorylation (Fig. 5B). DISCUSSION Alternative splicing of the cytoplasmic tail has been observed for a number of G protein-coupled receptors including the sst 2A somatostatin receptor (20), the D2 dopamine receptor (21), the EP3 prostaglandin receptor (22), and a number of serotonin receptor subtypes (23). C-terminal splicing is thought to modulate several aspects of G protein-coupled receptor physiology, i.e. cell-and tissue-specific expression, subcellular targeting, and coupling to specific G proteins. We have shown recently that the cytoplasmic tail of the rat -opioid receptor undergoes alternative splicing, giving rise to two isoforms, rMOR1 and rMOR1B (24). These isoforms exhibit similar pharmacological profiles; however, they differ in agonist-induced desensitization. Specifically, the shorter isoform rMOR1B desensitized at a slower rate than the longer isoform rMOR1 (10). In contrast, DAMGO-induced internalization and receptor resensitization of rMOR1B proceeded at a faster rate than that of rMOR1 (10). rMOR1B lacks only one phosphorylation site (Thr-394) compared with rMOR1. Changing this Thr-394 into alanine results in a reduction of DAMGO-induced phosphorylation and a slower DAMGO-mediated desensitization of the T394A receptor mutant (11,25,26). In addition, the lack of Thr-394 leads to facilitated receptor internalization, enhanced resensitization, and recycling of the T394A receptor mutant (10, 11). Thus,  Table I). The values represent means Ϯ S.E. from four separate measurements performed in triplicate. The asterisks indicate a significant difference (p Ͻ 0.05) between MOR1 and MOR1D or MOR1E as well as between MOR1C and MOR1D or MOR1E receptor isoforms (analysis of variance followed by the Bonferroni test). length and amino acid composition of the C terminus of the -opioid receptor seem to play a crucial role in agonist-induced endocytosis and receptor reactivation (10,11).
It has been reported that individual opioid agonists differ substantially in their ability to induce receptor endocytosis. Although endogenous opioid peptides promote a rapid endocytosis of the -opioid receptor, the highly addictive opioid alkaloid morphine fails to stimulate receptor phosphorylation and internalization (12,13). Interestingly, replacement of the entire -opioid receptor tail with the ␦-opioid receptor tail results in a receptor chimera that undergoes endocytosis after morphine treatment (27). It was demonstrated further that the /␦-tail swapping results in a receptor conformation that facilitates morphine-induced phosphorylation and subsequent ␤-arrestin binding (27).
Three novel splice variants, MOR1C, MOR1D, and MOR1E, were identified recently for the mouse -opioid receptor, MOR1, that differ in the number of potential S/T phosphorylation sites in their cytoplasmic tails (15). Instead of the Thr-394 in the very end of the MOR1 cytoplasmic tail, the C termini of MOR1C, MOR1D, and MOR1E contain seven (Thr-388, Ser-392, Thr-398, Ser-402, Thr-404, Ser-412, and Ser-431), two (Ser-392 and Ser-393), and one (Ser-392) new potential phosphorylation sites, respectively. We therefore addressed the question of whether C-terminal splicing of the -opioid receptor may influence agonist selectivity of receptor phosphorylation and internalization. The -receptor isoforms showed similar binding profiles and functional properties compared with the MOR1 receptor when expressed in HEK 293 cells (Table I). Our immunocytochemical analysis revealed that all splice variants underwent rapid endocytosis after 30 min of DAMGO treatment, whereas after 30 min of morphine treatment only MOR1D and MOR1E showed marked receptor endocytosis (Fig. 2). MOR1 and MOR1C revealed a faster receptor desensitization under morphine compared with DAMGO treatment, whereas MOR1D and MOR1E showed similar desensitization curves in response to either morphine or DAMGO (Fig. 3, A and  B). After complete desensitization, all splice variants resensitized during 50 min of DAMGO withdrawal (Fig. 3C). However, during 50 min of morphine withdrawal, only MOR1D and MOR1E but not MOR1 and MOR1C showed partial resensitization of receptor activity (Fig. 3D). These findings suggest a FIG. 5. Agonist-selective phosphorylation of -opioid receptor splice variants. HEK 293 cells expressing HA epitope-tagged MOR1, MOR1C, MOR1D, or MOR1E receptors were exposed to 1 M DAMGO or 1 M morphine for 20 min, and whole-cell receptor phosphorylation was determined as described under "Experimental Procedures." A, autoradiographs from a representative experiment are shown. B, mean Ϯ S.E. of three independent experiments quantified by PhosphorImager analysis. The data were normalized to the DAMGO-induced phosphorylation for each of the four -opioid receptor variants. Note that 1) morphine-induced MOR1 phosphorylation differed significantly from DAMGO-induced MOR1 phosphorylation (p Ͻ 0.001, double asterisks), and 2) morphine-induced MOR1C phosphorylation differed significantly from DAMGO-induced MOR1C phosphorylation (p Ͻ 0.05, asterisk, analysis of variance followed by the Bonferroni test). The positions of molecular mass markers are indicated on the left (in kDa). model in which the rapid morphine-induced endocytosis of MOR1D and MOR1E permits accelerated resensitization and recycling of these receptors, thus conferring resistance to agonist-induced desensitization. In contrast, the morphine-desensitized MOR1 and MOR1C receptors cannot gain entrance into the endocytotic-endosomal recycling pathway and are therefore not resensitized after morphine withdrawal. Several lines of evidence support the hypothesis that agonist-induced endocytosis is required for the reactivation of desensitized -opioid receptors and hence counteracts the development of physiological drug tolerance: 1) the -opioid receptor no longer resensitized when receptor recycling was blocked by monensin, an inhibitor of endosomal acidification (10), 2) the internalization rate of the -opioid receptor was increased in the presence of monensin (10), 3) the -opioid receptor desensitized faster when receptor recycling was blocked with monensin (10), and 4) the -opioid receptor desensitized faster after blocking receptor internalization by sucrose (28).
To delineate the mechanistic basis of these observations we carried out whole-cell phosphorylation studies for all splice variants. Previous studies have shown that morphine fails to induce intense -opioid receptor phosphorylation. It is believed that morphine restrains the MOR1 receptor in a conformation that is recalcitrant to phosphorylation by G protein-coupled receptor kinases (27,29). Our data clearly show that DAMGO mediates robust phosphorylation of all splice variants. In contrast, morphine promotes marked receptor phosphorylation only in MOR1D and MOR1E but fails to promote phosphorylation to a similar extent as DAMGO in the MOR1 and MOR1C receptor isoforms (Fig. 5). Thus, the sequence alterations of the cytoplasmic tail caused by C-terminal splicing result in a conformation that makes the morphine-activated MOR1D and MOR1E receptors accessible for G protein-coupled receptor kinase-mediated phosphorylation and permit subsequent receptor sequestration.
However, we also found that morphine activates and desensitizes all four -opioid receptor variants. Consequently, the question arises as to which mechanisms guide the morphinemediated desensitization of MOR1 and MOR1C. Our results show that in HEK 293 cells, morphine elicits phosphorylation of MOR1 and MOR1C to 30 and 55% of DAMGO-induced phosphorylation, respectively. It is therefore not unreasonable to speculate that partial phosphorylation of the MOR1 and MOR1C receptors elicited by morphine is sufficient to terminate signaling but not sufficient to promote receptor endocytosis (29).
After extended exposure (16 h) to DAMGO we observed a marked reduction of monomeric and dimeric forms of the MOR1 receptor (Fig. 1B). In contrast, treatment of MOR1 with naloxone led to an increase in the amount of receptor monomers and dimers (Fig. 1B). Moreover, among the four -opioid receptor splice variants, only MOR1 and MOR1C form homodimers that are stable under denaturing and reducing conditions, whereas only very faint bands were detected at the expected molecular masses for the MOR1D and MOR1E dimers (Fig. 1A). Interestingly, when constitutive internalization of MOR1D and MOR1E was blocked by prolonged exposure to naloxone, MOR1D and MOR1E homodimers were readily detectable (Fig. 1C). The simplest explanation for these findings is that the DAMGO-induced internalization and subsequent lysosomal degradation leads to down-regulation of the receptor proteins. On the other hand, the antagonist naloxone blocks receptor internalization and seems to stabilize opioid receptor dimers at the plasma membrane, thereby preventing monomerization and down-regulation of the receptor proteins. This conclusion is supported by the present findings that DAMGO promoted down-regulation of all four -opioid receptor splice variants, whereas morphine induced selectively the down-regulation of MOR1D and MOR1E but not of MOR1 or MOR1C (Fig. 4).
In conclusion, our results provide strong evidence that changes in the C-terminal sequence by alternative splicing determine agonist selectivity of phosphorylation and internalization of the -opioid receptor isoforms. We show that two splice variants, MOR1D and MOR1E, of the mouse -opioid receptor were rapidly phosphorylated and internalized after morphine treatment. The rapid morphine-induced endocytosis of MOR1D and MOR1E confers resistance to desensitization and permits receptor resensitization and down-regulation. In contrast, the lack of morphine-induced phosphorylation and internalization of MOR1 and MOR1C splice variants also prevents receptor resensitization and down-regulation. Together, our findings provide strong evidence that endocytosis is required for the -opioid receptor to enter either the endosomal recycling pathway, leading to receptor resensitization, or the lysosomal degradation pathway, leading to receptor down-regulation. Given the emerging evidence of cell-and tissue-specific expression of the four -opioid receptor isoforms, it is likely that C-terminal splicing may significantly modulate the development of tolerance to the various effects of morphine.