Heterodimerization of somatostatin and opioid receptors cross-modulates phosphorylation, internalization, and desensitization.

Heterodimerization has been shown to modulate the ligand binding, signaling, and trafficking properties of G protein-coupled receptors. However, to what extent heterodimerization may alter agonist-induced phosphorylation and desensitization of these receptors has not been documented. We have recently shown that heterodimerization of sst(2A) and sst(3) somatostatin receptors results in inactivation of sst(3) receptor function (Pfeiffer, M., Koch, T., Schröder, H., Klutzny, M., Kirscht, S., Kreienkamp, H. J., Höllt, V., and Schulz, S. (2001) J. Biol. Chem. 276, 14027-14036). Here we examine dimerization of the sst(2A) somatostatin receptor and the mu-opioid receptor, members of closely related G protein-coupled receptor families. In coimmunoprecipitation studies using differentially epitope-tagged receptors, we provide direct evidence for heterodimerization of sst(2A) and MOR1 in human embryonic kidney 293 cells. Unlike heteromeric assembly of sst(2A) and sst(3), sst(2A)-MOR1 heterodimerization did not substantially alter the ligand binding or coupling properties of these receptors. However, exposure of the sst(2A)-MOR1 heterodimer to the sst(2A)-selective ligand L-779,976 induced phosphorylation, internalization, and desensitization of sst(2A) as well as MOR1. Similarly, exposure of the sst(2A)-MOR1 heterodimer to the mu-selective ligand [d-Ala(2),Me-Phe(4),Gly(5)-ol]enkephalin induced phosphorylation and desensitization of both MOR1 and sst(2A) but not internalization of sst(2A). Cross-phosphorylation and cross-desensitization of the sst(2A)-MOR1 heterodimer were selective; they were neither observed with the sst(2A)-sst(3) heterodimer nor with the endogenously expressed lysophosphatidic acid receptor. Heterodimerization may thus represent a novel regulatory mechanism that could either restrict or enhance phosphorylation and desensitization of G protein-coupled receptors.

as homo-or heterodimeric complexes (1,2). Heterodimerization has been shown to alter both ligand binding affinity and signaling efficacy of GPCRs (1,2). ␦and -opioid receptors form stable heterodimers with ligand binding and signaling properties resembling that of the 2 receptor (3). Formation of heterodimers between the sst 1 and sst 5 somatostatin receptors has been found to modulate the pharmacology and signaling of both receptors (4). The ␥-aminobutyric acid receptor B is unique in that heterodimerization of the nonfunctional ␥-aminobutyric acid receptors B1 and B2 is required for native affinity for ligands and complete functional activity (5)(6)(7)(8)(9). Heteromeric assembly of fully functional AT 1 angiotensin II and B 2 bradykinin receptors results in increased efficacy of angiotensin II and decreased efficacy of bradykinin (10).
Heterodimerization has also been shown to alter endocytotic trafficking of GPCRs (3,4,10,11). The -␦ heterodimer exhibited a decrease in agonist-mediated receptor endocytosis (3). Oligomerization of ␦and -opioid receptors with the distantly related ␤ 2 -adrenergic receptor results in increased and decreased receptor endocytosis, respectively (11). AT 1 -B 2 heterodimerization induced a switch to a clathrin-and dynamindependent endocytotic pathway for both receptors (10). Signaling of GPCRs is often terminated by phosphorylation of intracellular serine and threonine residues. 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. Although changes in trafficking have been clearly documented, agonistinduced phosphorylation and desensitization of these GPCR heterodimers has not been examined.
We have recently shown that the sst 2A and sst 3 somatostatin receptors exist as constitutive homodimers when expressed alone and as constitutive heterodimers when coexpressed in human embryonic kidney (HEK) 293 cells (12). Whereas the sst 2A -sst 3 heterodimer behaved like the sst 2A homodimer, it did not reproduce the pharmacological characteristics of the sst 3 homodimer, suggesting that physical interaction of sst 3 with sst 2A induced functional inactivation of the sst 3 subtype (12). Here we report that the sst 2A receptor also forms stable heterodimers with the -opioid receptor (MOR1), a member of a closely related GPCR family. Unlike that observed for the sst 2A -sst 3 heterodimer, sst 2A -MOR1 heterodimerization did not significantly affect the ligand binding or coupling properties but promoted cross-modulation of phosphorylation, internalization, and desensitization of these receptors.
Cell Culture and Transfections-The wild-type rat sst 2A receptor was tagged at its amino terminus with the T7 epitope tag sequence MASMT-GGQQMG using polymerase chain reaction and subcloned into a pcDNA3.1 expression vector (Invitrogen) containing a neomycin resistance as described previously. The wild-type rat -opioid receptor MOR1 was tagged at its amino terminus with the HA epitope tag sequence YPYDVPDYA using polymerase chain reaction and subcloned into a pEAK10 expression vector (Edge Bio Systems, Gaithersburg, MD) containing a puromycin resistance as described previously (17). HEK 293 cells were obtained from ATCC and grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum in a humidified atmosphere containing 10% CO 2 . The cells were first transfected with plasmids containing the neomycin resistance using the calcium phosphate precipitation method. Stable transfectants were selected in the presence of 500 g/ml G418 (Invitrogen). To generate lines coexpressing two differentially epitope-tagged receptors, the cells were subjected to a second round of transfection using FuGENE 6 (Roche Diagnostics) and selected in the presence of 500 g/ml G418 and 1 g/ml puromycin (Sigma). Three clones expressing T7sst 2A alone, six clones expressing HAMOR1 alone, and four clones coexpressing T7sst 2A and HAMOR1 were generated. Receptor expression was monitored using saturation ligand binding assays as described below. In addition, quantitative Western blot analysis was carried out to ensure that clones coexpressing ϳ 1:1 ratio of sst 2A and MOR1 were selected. Finally, double immunofluorescent staining was performed to validate that sst 2A and MOR1 were coexpressed within the same cells. The B max and K D values of the cells that were used throughout this study are given in Table I.
Immunoprecipitation and Western Blot Analysis-Stably transfected HEK 293 cells were plated onto poly-L-lysine-coated 150-mm dishes and FIG. 1. Characterization of sst 2A -MOR1 heterodimers by coimmunoprecipitation. A, HEK 293 cells expressing either T7sst 2A or HAMOR1 alone or coexpressing T7sst 2A and HAMOR1 or a mixture of cells expressing either T7sst 2A or HAMOR1 alone were exposed to bis(sulfosuccinimidyl)suberate, lysed in detergent buffer, and subjected to immunoprecipitation using rat anti-HA antibody. The coimmunoprecipitates were immunoblotted using rabbit anti-sst 2A (6291) antibody. Coimmunoprecipitation of the T7sst 2A can be seen only when T7sst 2A and HAMOR1 are coexpressed in the same cell (third lane) but not when cells expressing T7sst 2A or HAMOR1 individually were mixed prior to immunoprecipitation (fourth lane). AЈ, the same blot was stripped and reprobed with rabbit anti-HA antibody. MOR1 dimers and monomers can be seen in cells expressing HAMOR1 alone, in cells coexpressing T7sst 2A and HAMOR1, and in a mixture of cells expressing T7sst 2A or HAMOR1 individually but not in cells expressing sst 2A alone. B, HEK 293 cells coexpressing T7sst 2A and HAMOR1 were lysed in radioimmune precipitation buffer and subjected to immunoprecipitation using rat anti-HA antibody and immunoblotted using rabbit anti-sst 2A (6291) antibody. BЈ, the same blot was stripped and reprobed with rabbit anti-HA antibody. Note that sst 2A -MOR1 heterodimers were nearly completely dissociated under these conditions. The positions of molecular mass markers are indicated on the left (in kDa). The arrows point to the dimeric and monomeric forms of the receptors. Three additional experiments gave similar results.  Radioligand binding studies and cAMP and ERK assays were carried out as described under "Experimental Procedures." The half-maximal inhibitory concentrations (IC 50 ) for competition binding and inhibition of forskolin-stimulated cAMP accumulation and the half-maximal effector concentration (EC 50 ) for ERK activation were analyzed by nonlinear regression curve fitting using the computer program GraphPad Prism 3.0. The data are presented as the means Ϯ S.E. of three or four independent experiments performed in triplicate. grown to 80% confluence. The cells were exposed to the cross-linking agents bis(sulfosuccinimidyl)suberate or dithiobis-(succinimidylpropionate) (both from Pierce) and subsequently lysed as described (12,17). The cell membranes were prepared and solubilized 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, 0.2 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 1 g/ml pepstatin A, 1 g/ml aprotinin, and 10 g/ml bacitracin) for 1 h on ice. Alternatively, the cells were lysed in radioimmune precipitation buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, and proteinase inhibitors) as described below. The receptor proteins were then immunoprecipitated with 100 l of protein A-agarose beads preloaded with 10 g of anti-HA, anti-T7, or anti-sst 2A (6291) antibodies. Immunocomplexes were eluted from the beads using SDS sample buffer for 20 min at 60°C and resolved by SDS-PAGE. After electroblotting, membranes were incubated with either mouse monoclonal anti-T7, rat monoclonal anti-HA, affinity-purified rabbit anti-sst 2A (6291), or anti-MOR1 (9998) antibodies at a concentration of 1 g/ml for 12 h at 4°C, followed by detection using an enhanced chemiluminescence detection system. When indicated, the membranes were placed in stripping buffer (100 mM 2-mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl, pH 6.7) for 30 min at 55°C and subsequently reprobed.
Whole Cell Phosphorylation Assays-The cells expressing T7sst 2A , T7sst 3 , or HAMOR1 alone as well as cells coexpressing T7sst 2A and Mycsst 3 or T7sst 2A and HAMOR1 were plated onto 100-mm dishes and grown to 80% confluence. The cells were washed with serum-and phosphate-free medium and then labeled with 200 Ci/ml carrier-free [ 32 P]orthophosphate (285 Ci/mg P i ; ICN, Eschwege, Germany) for 60 min at 37°C. The labeled cells were exposed to either 100 nM L-779,976, 1000 nM L-796,778, 100 nM SS-14, or 1000 nM DAMGO for 20 min. After incubation, the cells were placed on ice and washed with ice-cold phosphate-buffered saline 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. The supernatants were obtained by centrifugation at 13,000 ϫ g for 60 min at 4°C, after which aliquots were taken to determine the total protein content in the supernatant of each sample to be immunoprecipitated. Immunoprecipitations were carried out by adding 10 g of affinity-purified polyclonal rabbit anti-HA tag, anti-T7 tag, or anti-sst 3 (7986) antibodies as described above. Immunocomplexes were eluted from the beads using SDS sample buffer for 20 min at 60°C. 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. The samples were then subjected to 8% SDS-polyacrylamide gel electrophoresis followed by autoradiography. The extent of phosphorylation of receptor monomers was quantitated using a Fuji PhosphorImaging system and BAS 1000 software.
Immunocytochemistry-The cells were grown on poly-L-lysinetreated coverslips overnight and then exposed to agonists. The cells were fixed and permeabilized as described (12). For single immunofluorescence, the cells were then incubated with either mouse monoclonal anti-T7, rat monoclonal anti-HA, affinity-purified rabbit anti-HA, affinity-purified rabbit anti-sst 2A (6291), or affinity-purified rabbit anti-MOR1 (9998) antibody at a concentration of 1 g/ml in Tris/phosphatebuffered saline and 1% normal goat serum overnight. Bound primary antibody was detected with biotinylated secondary antibodies (1:100; Vector, Burlingame, CA) followed by cyanine 3.18-conjugated streptavidin (Amersham Biosciences). For double immunofluorescence, the cells were incubated either with a mixture of rat monoclonal anti-HA and affinity-purified rabbit anti-sst 2A (6291) or affinity-purified rabbit anti-HA and guinea pig anti-sst 2A (GP3) antibodies. Bound primary antibodies were detected with biotinylated anti-rabbit antibodies, followed by a mixture of cyanine 2.18-conjugated streptavidin and cyanine 5.18-conjugated anti-rat or anti-guinea pig antibodies (1:200, Jackson ImmunoResearch, West Grove, PA). The cells were then dehydrated, cleared in xylol, and permanently mounted in DPX (Fluka, Neu-Ulm, Germany).
Male Wistar rats (n ϭ 3, 200 -250 g; Tierzucht, Schönwalde, Germany) were deeply anesthetized with chloral hydrate and transcardially perfused with Tyrode's solution followed by Zamboni's fixative containing 4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer, pH 7.4. The brains were rapidly dissected and post-fixed in the same fixative for 2 h at room temperature. For all animal procedures ethical approval was sought prior to the experiments according to the requirements of the German National Act on the Use of Experimental Animals. The tissue was cryoprotected by immersion in 30% sucrose before sectioning using a freezing microtome. Free-floating sections (30 -40 m) were incubated with a mixture of guinea pig anti-sst 2A (GP3) and rabbit anti-MOR1 (9998) antibodies for 48 -72 h at room temperature. Bound primary antibodies were detected as above, and sections were permanently mounted in DPX. The specimens were examined using a Leica TCS-NT laser scanning confocal microscope (Heidelberg, Germany) equipped with a krypton/argon laser. Cyanine 2.18 was imaged with 488-nm excitation and 500 -560-nm band pass emission filters, cyanine 3.18 was imaged with 568-nm excitation and 570 -630-nm band pass emission filters, and cyanine 5.18 was imaged with 647-nm excitation and 665-nm long pass emission filters.
Internalization Assays-The cells were seeded at a density of 2 ϫ 10 5 /well onto poly-L-lysine-treated 24-well plates. The next day, the cells were preincubated with 1 g of affinity-purified rabbit anti-T7 or anti-HA antibody for 2 h in OPTIMEM 1 (Invitrogen) at 4°C. The cells were then treated with 100 nM L-779,976, 1000 nM DAMGO, or 100 nM PMA in OPTIMEM for 60 min. Subsequently, the cells were fixed and incubated with peroxidase-conjugated anti-rabbit antibody (1:1000; Amersham Biosciences) for 2 h at room temperature. After washing, the plates were developed with 250 l of ABTS solution (Roche). After 10 -30 min, 200 l of the substrate solution from each well was transferred to a 96-well plate and analyzed at 405 nm using a microplate reader (Bio-Rad).
Radioligand Binding Assays-Saturation binding assays were per-formed on membrane preparations from stably transfected cells as described previously. Measurements of cAMP Accumulation-Transfected cells were seeded at a density of 1.5 ϫ 10 5 /well onto poly-L-lysine-treated 22-mm 12-well dishes. On the next day, the cells were incubated in the presence or absence of 100 nM L-779,976 or 100 nM DAMGO for 0, 0.5, 1, 2, 4, or 6 h in OPTIMEM 1. The medium was then removed and replaced with 0.5 ml of serum-free RPMI medium containing 25 M forskolin or 25 M forskolin plus either L-779,976 or DAMGO in concentrations ranging from 10 Ϫ14 to 10 Ϫ6 M. The cells were incubated at 37°C for 15 min. The reaction was terminated by removal of the culture medium and subsequent addition of 1 ml of ice-cold HCl/ethanol (1 volume of 1 N HCl with 100 volumes of ethanol). After centrifugation, the supernatant was evaporated, the residue was dissolved in TE buffer (50 mM Tris-EDTA, pH 7.5), and the cAMP content was determined using a commercially available radioimmunoassay kit (Amersham Biosciences).
ERK Assays-The cells were seeded at a density of 1.5 ϫ 10 5 /well onto poly-L-lysine-treated 22-mm 12-well dishes, grown in Dulbecco's modified Eagle's medium containing 0.5% fetal calf serum overnight, and then pretreated with OPTIMEM 1 for 2 h. The cells were then incubated in the presence or absence of 100 nM L-779,976 or 100 nM DAMGO for 1, 2, 4, or 6 h in OPTIMEM 1 and then exposed to either L-779,976, DAMGO, or lysophosphatidic acid in concentrations ranging from 10 Ϫ14 to 10 Ϫ6 M for 5 min in RPMI medium at 37°C. Incubation was terminated by removal of the culture medium and subsequent addition of 250 l of boiling SDS sample buffer. Equal amounts of protein of each sample were separated on 10% SDS-polyacrylamide gels and electroblotted onto nitrocellulose membranes. The protein content was determined using the BCA method. After blocking, the membranes were incubated with mouse monoclonal phospho-specific anti-ERK1/2 antibody clone E10 (New England Biolabs, Beverly, MA) or phosphorylation-independent rabbit polyclonal anti-ERK2 antibody (Santa Cruz Biotechnology, Santa Cruz, CA). Blots were developed using peroxidase-conjugated secondary antibodies and enhanced chemiluminescence. Densitometric analysis of total ERK2 and phospho-ERK1/2 levels on Western blots exposed in the linear range of the x-ray film was performed using National Institutes of Health Image 1.57 software.
Phospho-ERK1/2 levels were normalized to total ERK1/2 per lane and expressed as the fold ERK1/2 phosphorylation over the basal value of untreated cells.
Data Analysis-Data from ligand binding, cAMP, and ERK assays were analyzed by nonlinear regression curve fitting using GraphPad Prism 3.0 software. Statistical analysis was carried out using the twotailed paired t test or two-way analysis of variance followed by the Bonferroni test. p values Ͻ 0.05 were considered to be statistically significant.

RESULTS
Heterodimerization of sst 2A and MOR1-To directly examine the association between the sst 2A and the MOR1 receptor, we stably coexpressed T7-tagged sst 2A receptors and HA-tagged MOR1 receptors in HEK 293 cells. Saturation binding experiments revealed that these cells coexpressed ϳ1:1 ratio of somatostatin binding sites (B max 2,458 Ϯ 418 fmol/mg membrane protein) and DAMGO binding sites (B max 2,986 Ϯ 135 fmol/mg membrane protein) ( Table I). When membrane extracts from these cells were prepared with detergent buffer and immunoprecipitated using the rat anti-HA antibody, the carboxyl-terminal anti-sst 2A antibody (6291) detected a single band migrating at 160 kDa, suggesting that this band represents a T7sst 2A -HAMOR1 heterodimer (Fig. 1A). Immunoprecipitation of sst 2A -MOR1 heterodimers was facilitated when cells were preincubated with the cross-linking agent bis(sulfosuccinimidyl) suberate (Fig. 1A). In contrast, no bands were detectable in immunoprecipitates prepared under identical conditions from cells expressing only T7sst 2A or HAMOR1 or from a mixture of T7sst 2A -and HAMOR1-expressing cells. These data strongly suggest that sst 2A -MOR1 heterodimers pre-existed in cells prior to cell lysis and were not artificially formed during sample preparation. When the blot shown in Fig. 1A was stripped and reprobed with rabbit anti-HA antibody, MOR1 monomers and dimers were revealed in immunoprecipitates from HAMOR1 cells, from T7sst 2A -HAMOR1 cells, and from a mixture of T7sst 2A and HAMOR1 cells (Fig. 1AЈ). To test the stability of sst 2A -MOR1 heterodimers under conditions used for whole cells phosphorylation assays, the cells were lysed in radioimmune precipitation buffer, immunoprecipitated using anti-HA antibody, and detected with anti-sst 2A antibody. As shown in Fig. 1B, sst 2A -MOR1 heterodimers were not detectable under these conditions. When the blot shown in Fig. 1B was stripped and reprobed with rabbit anti-HA antibody, it was apparent that cell lysis in SDS-containing radioimmune precipitation buffer in the absence of cross-linking agents leads to nearly complete dissociation of sst 2A -MOR1 heterodimers (Fig. 1BЈ).
Ligand  robust responses in cells coexpressing T7sst 2A and HAMOR1 compared with cells expressing T7sst 2A alone. Conversely, the MOR1-selective agonist DAMGO produced slightly more robust responses in cells coexpressing T7sst 2A and HAMOR1 as compared with cells expressing HAMOR1 alone. In contrast, L-779,976 neither inhibited forskolin-stimulated cAMP accumulation nor activated ERK1/2 in HAMOR1 cells. Similarly, DAMGO showed no significant functional responses in T7sst 2A cells. Although small changes in L-779,976 binding and signaling were detected, these findings suggest that two separate binding pockets were formed by the sst 2A -MOR1 heterodimer and that the ligand binding and coupling properties of the sst 2A and MOR1 receptors were not substantially altered after heterodimerization.
Endocytotic Trafficking of the sst 2A -MOR1 Heterodimer-We next examined the effect of sst 2A -HAMOR1 heterodimerization on receptor endocytosis using HEK 293 cells expressing either T7sst 2A or HAMOR1 or coexpressing both T7sst 2A and HAMOR1. The cells were exposed to either 100 nM L-779,976, 1000 nM DAMGO, or 100 nM PMA for 30, 60, 120, or 180 min at 37°C. The cells were subsequently fixed, permeabilized, and fluorescently labeled with T7sst 2A -and/or HAMOR1-specific antibodies. The subcellular distribution of receptor proteins was then analyzed by confocal microscopy. As depicted in Fig.  2, both sst 2A and MOR1 receptors were predominantly confined to the plasma membrane in the absence of agonist. Treatment with L-779,976 but not with DAMGO induced an accumulation of sst 2A receptors in vesicle-like structures within the cytoplasm in cells expressing T7sst 2A alone. Exposure to DAMGO but not to L-779,976 promoted internalization of MOR1 receptors in cells expressing HAMOR1 alone. Activation of protein kinase C (PKC) by phorbol esters is known to stimulate heterologous phosphorylation of both the sst 2A and the MOR1 receptor (18,19). However, PMA induced only the internalization of sst 2A but not of MOR1 in cells expressing these receptors alone (Fig. 2). In coexpressing cells both sst 2A and MOR1 were seen at the cell surface, revealing extensive colocalization (Fig. 3,  Control). Interestingly, after 30 min of L-779,976 exposure, both the sst 2A and the MOR1 receptor underwent robust internalization (Fig. 3, L-779,976). Similarly, after treatment with PMA, sst 2A and MOR1 were also cointernalized (Fig. 3, PMA). In contrast, DAMGO exposure induced only internalization of MOR1 but not of sst 2A (Fig. 3, DAMGO). Quantitative analysis of receptor internalization by enzyme-linked immunosorbent assay confirmed that MOR1 was internalized in response to L-779,976 as well as PMA only in cells coexpressing sst 2A and MOR1 but not in cells expressing MOR1 alone (Fig. 4). The fact that the MOR1 receptor was resistant to L-779,976-and PMAinduced endocytosis in cells expressing this receptor alone but not in coexpressing cells suggests that physical interaction of both receptor proteins was required to promote cointernalization of MOR1 together with sst 2A under these conditions. Desensitization of the sst 2A -MOR1 Heterodimer-We then examined agonist-induced desensitization of the sst 2A -MOR1 heterodimer. Cells coexpressing T7sst 2A and HAMOR1 were preincubated in the presence or absence of either 100 nM L-779,976 or 100 nM DAMGO for 0.5, 1, 2, 4, or 6 h. The medium was removed, and the ability of either L-779,976 or DAMGO to inhibit forskolin-stimulated cAMP accumulation was determined. Interestingly, the sst 2A -dependent responses of the sst 2A -MOR1 heterodimer underwent a rapid time-dependent loss of coupling to adenylate cyclase upon preincubation with either L-779,976 or DAMGO with a maximum desensitization at 6 h (Fig. 5). Conversely, the MOR1-dependent responses of the sst 2A -MOR1 heterodimer underwent a rapid time-dependent loss of coupling to adenylate cyclase upon preincubation with either DAMGO or L-779,976 with a maximum desensitization at 6 h (Fig. 5). The L-779,976-induced desensitization of the sst 2A -MOR1 heterodimer followed a similar time-course as that of the sst 2A homodimer (12). In contrast, DAMGO did not bind, activate, or desensitize the sst 2A homodimer (Table II). Similarly, the DAMGO-induced desensitization of the sst 2A -MOR1 heterodimer followed a time course similar to that of the MOR1 homodimer (16). In contrast, L-779,976 did not bind, activate, or desensitize the MOR1 homodimer (Table II).
We also examined the desensitization of mitogenic signaling of the sst 2A -MOR1 heterodimer. The cells coexpressing T7sst 2A and HAMOR1 were preincubated in the presence or absence of either 100 nM L-779,976 or 100 nM DAMGO for 0.5, 1, 2, 4, or 6 h. The medium was removed, and the ability of either L-779,976, DAMGO or lysophosphatidic acid to stimulate ERK1/2 activity was determined. As depicted in Fig. 6, preincubation with either L-779,976 or DAMGO for 4 h significantly attenuated both sst 2A -and MOR1-dependent responses. In contrast, mitogenic signaling of the lysophosphatidic acid receptor, a third receptor that is endogenously expressed in this system, was unchanged under these conditions, suggesting that the sst 2A -MOR1 heterodimer underwent homologous cross-desensitization under these conditions.
Phosphorylation of the sst 2A -MOR1 Heterodimer-To delineate a mechanistic basis for the observed cross-desensitization of the sst 2A -MOR1 heterodimer, we assessed whole cell receptor phosphorylation in response to both L-779,976 and DAMGO. As shown in Fig. 1, cell lysis in radioimmune precipitation buffer resulted in a nearly complete dissociation of receptor dimers, which enabled us to selectively analyze the phosphorylation level of the resulting receptor monomers. As depicted in Fig. 7 (A and B), L-779,976 induced a rapid and robust phosphorylation of the sst 2A receptor monomer (ϳ4.3fold over basal). DAMGO produced a rapid and robust phosphorylation of the MOR1 receptor monomer (ϳ5.9-fold over basal). Interestingly, L-779,976 also significantly increased phosphorylation of the MOR1 receptor monomer (ϳ2.5-fold over basal). Conversely, DAMGO also significantly increased phosphorylation of the sst 2A receptor monomer (ϳ2.0-fold over basal), indicating that activation of the sst 2A subunit of the sst 2A -MOR1 heterodimer resulted in cross-phosphorylation of the MOR1 subunit and vice versa. This cross-phosphorylation was not simply due to cross-reactivity of the agonists, because it was not observed in cells expressing either T7sst 2A or HAMOR1 alone (Fig. 7, C and D). To elucidate the selectivity of the observed sst 2A -MOR1 cross-phosphorylation, we examined agonist-induced phosphorylation of a sst 2A heterodimer with different functional properties, namely the sst 2A -sst 3 heterodimer. Like MOR1, the sst 3 receptor also forms stable heterodimers with the sst 2A receptor. Unlike MOR1, the sst 3 receptor is functionally inactivated upon heterodimerization with sst 2A (12). As shown in Fig. 8, whereas the sst 2 -selective agonist L-779,976 stimulated a robust phosphorylation of the sst 2A receptor monomer, it failed to increase phosphorylation of the sst 3 receptor monomer in cells coexpressing T7sst 2A and Mycsst 3 . Interestingly, the sst 3 -selective agonist L-796,778 promoted phosphorylation of the sst 3 receptor in cells expressing this receptor alone; however, it did not increase phosphorylation of the sst 3 or sst 2A receptor monomers above basal levels in coexpressing cells. These findings suggest that the specific pattern of agonist-induced phosphorylation of heterodimeric receptors may largely depend on the their functional properties. The loss of sst 3 -dependent binding and signaling of the sst 2A -sst 3 heterodimer is associated with diminished L-796,778-induced phosphorylation of this receptor. In addition, sst 2A -MOR1 cross-phosphorylation may provide a plausible explanation for homologous cross-desensitization of this heterodimer.
Colocalization of sst 2A and MOR1 in Rat Brain-A major prerequisite for the physiological assembly of sst 2A -MOR1 heterodimers is their coexpression in the same cells. We therefore examined the spatial relation between sst 2A and MOR1 in the central nervous system, and serial rat brain sections were The cells were washed and then exposed to either 100 nM L-779,976, 1000 nM DAMGO, or 1000 nM lysophosphatidic acid for 5 min. The cells were lysed, equal amounts of protein were resolved by SDS-PAGE, and the levels of total ERK1/2 and phosphorylated ERK1/2 were determined by immunoblotting. A, results were quantified by densitometric analysis. The data were normalized to total ERK1/2 and expressed as the fold ERK1/2 phosphorylation over the basal value in untreated cells. The values represent the means Ϯ S.E. of three independent experiments performed in duplicate. The asterisks indicate a significant difference (p Ͻ 0.05) between cells preincubated with either L-779,976 or DAMGO and cells that had not been preincubated (two-tailed paired t test). B and C, representative immunoblots for phospho-ERK1/2 and total ERK1/2, respectively. The positions of phospho-ERK1/2 (pERK1/2) and total ERK1/2 (ERK1/2) are indicated on the right. The positions of molecular mass markers are indicated on the left (in kDa). processed for dual immunofluorescence and examined under a confocal microscope (Fig. 9). Both sst 2A and MOR1 receptors were widely distributed throughout the central nervous system and mostly targeted to neuronal somata and dendrites. At low magnification it was apparent that immunoreactive sst 2A and MOR1 receptors exhibited closely overlapping distributions in many brain stem regions including the locus coeruleus, spinal trigeminal nucleus, and superficial layers of the spinal cord dorsal horn (Fig. 9, C, I, and O). At high power magnification a high degree of colocalization of the two receptors was observed only in the locus coeruleus (Fig. 9F). In contrast, immunoreactive sst 2A and MOR1 receptors were clearly confined to distinct neuronal somata and dendrites in the spinal trigeminal nucleus and the superficial layers of the spinal cord dorsal horn (Fig. 9, L and R). DISCUSSION The existence of homo-and heterodimers has been demonstrated for several GPCRs using coimmunoprecipitation, fluorescence and bioluminescence resonance energy transfer, and functional complementation techniques (3, 4, 10, 20 -25). Heterodimers can be formed between members of both closely and distantly related GPCR families (5-11, 22, 26). We have previously shown that members of the somatostatin receptor family exist as constitutive homodimers when expressed alone and as constitutive heterodimers when coexpressed (12). In the present study, we explored the functional consequences of a physical interaction between the sst 2A somatostatin receptor and the -opioid receptor a member of a closely related GPCR family. The sst 2A and the MOR1 receptor share 38% sequence homology. We find that the sst 2A receptor forms heterodimers with the -opioid receptor. The immunoprecipitation of HAtagged MOR1 receptors results in coprecipitation of T7-tagged sst 2A receptors only from coexpressing cells but not from a mixture of cells expressing these receptors separately, suggesting that sst 2A -MOR1 heterodimers preexisted in these cells prior to cell lysis and were not artificially formed during sample preparation.
The physical interaction between sst 2A and MOR1 has pro-FIG. 7. Agonist-induced cross-phosphorylation of the sst 2A -MOR1 heterodimer. HEK 293 cells coexpressing T7sst 2A and HAMOR1 or expressing either T7sst 2A or HAMOR1 alone were exposed to 100 nM L-779,976 or 1000 nM DAMGO for 20 min, and whole cell receptor phosphorylation was determined as described under "Experimental Procedures." T7sst 2A was immunoprecipitated (IP) with rabbit anti-T7 antibodies, and HAMOR1 was immunoprecipitated with rabbit anti-HA antibodies. A and C, autoradiographs from representative experiments are shown. B and D, means Ϯ S.E. of three independent experiments quantified by PhosphorImager analysis. The asterisks indicate significant agonist-induced phosphorylation compared with basal levels in the absence of agonist (p Ͻ 0.05; two-tailed paired t test). Note that phosphorylation of sst 2A subunit of the sst 2A -HAMOR1 heterodimer was significantly increased above basal levels in the presence of the MOR1-selective agonist DAMGO. Conversely, phosphorylation of the MOR1 subunit of the sst 2A -HAMOR1 heterodimer was significantly increased above basal levels in the presence of the sst 2 -selective agonist L-779,976. The data were normalized to basal phosphorylation in the absence of agonist for each receptor monomer. The positions of molecular mass markers are indicated on the left (in kDa).
found consequences on the trafficking properties of these receptors. Whereas MOR1 was resistant to L-779,976-and PMAinduced endocytosis in cells expressing this receptor alone, it was readily internalized together with the sst 2A receptor in response to both the sst 2 -selective agonist L-779,976 and the PKC activator PMA in coexpressing cells. This is in contrast to the trafficking properties of the sst 2A -sst 3 heterodimer (12). Whereas sst 3 was readily internalized in the presence of L-796,778 in cells expressing this receptor alone, it was resistant to endocytosis mediated by the sst 2 -selective agonist L-779,976, the sst 3 -selective agonist L-796,778, or the nonselective agonist SS-14 in cells coexpressing sst 2A and sst 3 (12). Thus it appears that heterodimerization differentially affects the properties of these closely related receptors, and this is unique for each heterodimeric complex.
Previous studies have reported similar effects of dimerization on the trafficking properties of GPCRs (3, 4, 10, 11). However, none of these studies has established a mechanistic basis for the observed differences. Phosphorylation of intracel-lular serine and threonine residues within the third intracellular loop and the carboxyl terminus is the initial step in the desensitization of opioid and somatostatin receptors (18,19,(27)(28)(29). After phosphorylation, ␤-arrestins are rapidly recruited to the plasma membrane where they facilitate endocytosis via clathrin-coated pits and vesicles. The present study shows that activation of the sst 2A subunit of the sst 2A -MOR1 heterodimer resulted in cross-phosphorylation of the MOR1 subunit and vice versa. In contrast, the binding-and signalingdeficient sst 3 subunit of the sst 2A -sst 3 heterodimer was resistant to phosphorylation induced by either sst 2A -or sst 3 -selective agonists. The simplest explanation for our findings is that these heterodimers exist in a physically restrained conformation as proposed in the three-dimensional dimer model by Gouldson et al. (30). Both domain-swapped and contact dimer models support the involvement of transmembrane helices five and six as dimerization interface. Interestingly, the two models predict that the third intracellular loop originating from each monomer would be parallel within the dimer. Activation of one binding pocket of the sst 2A -MOR1 heterodimer would then be expected to induce such a conformational change, which would facilitate phosphorylation of both the sst 2A and the MOR1 receptor monomers by G protein-coupled receptor kinases. In contrast, the binding-and signaling-deficient sst 3 subunit of the sst 2A -sst 3 heterodimer would be expected to be resistant to such an agonist-induced conformational change and may therefore represent a poor substrate for G protein-coupled receptor kinase-mediated phosphorylation.
Given that both sst 2A and MOR1 are phosphorylated upon activation of PKC by phorbol esters (18,19), an alternative explanation exists in which the sst 2A -MOR1 heterodimer may be subject to heterologous PKC-mediated phosphorylation. However, this hypothesis is unlikely, because sst 3 is also phosphorylated and internalized upon PMA-induced PKC activation in cells expressing sst 3 alone (not shown). The sst 3 subunit of the sst 2A -sst 3 heterodimer would therefore be expected to undergo heterologous PKC-mediated phosphorylation independent of its ability to acquire an active binding and signaling conformation in coexpressing cells as well. Thus the lack of phosphorylation of the sst 3 subunit of the sst 2A -sst 3 heterodimer in response to activation of the sst 2A subunit argues against a heterologous phosphorylation of the sst 2A -MOR1 heterodimer by PKC.
Although the sst 2A subunit of the sst 2A -MOR1 heterodimer underwent cross-phosphorylation and -desensitization in response to activation of the MOR1 subunit, it was not cointernalized with the MOR1 receptor. This suggests that the DAMGO-mediated cross-phosphorylation of the sst 2A subunit may involve sites distinct from those involved in L-779,976-or PMA-induced phosphorylation. The specific pattern of DAMGO-induced phosphorylation leads to separation of the sst 2A -MOR1 heterodimer at the plasma membrane and may thus facilitate desensitization of this receptor.
Cross-phosphorylation may provide a plausible explanation for homologous cross-desensitization of adenylyl cyclase and ERK1/2 signaling of the sst 2A -MOR1 heterodimer. Conversely, lack of cross-phosphorylation could explain increased resistance to agonist-induced desensitization of the sst 2A -sst 3 heterodimer (12). Interestingly, in a previous study cross-desensitization of somatostatin-and opioid-dependent signal transduction was noted upon expression of the -opioid but not the -opioid receptor in AtT-20 cells, which endogenously express the sst 2A receptor (31). These findings underscore the importance of physical interactions in the differential modulation of a diverse array of GPCR functions. sst 2A -MOR1 heterodimerization could have functional relevance in vivo. sst 2A and MOR1 receptors coexist and functionally interact in pain-processing pathways (14,15). Studies have shown extensive cross-talk between opioid-and somatostatinmediated analgesic responses (32,33). A major prerequisite for the physiological assembly of heterodimeric GPCRs is their coexpression in the same cells. We observed a particularly high degree of colocalization between the sst 2A and the -opioid receptor in the locus coeruleus a brain region known to be involved in the expression of the opioid withdrawal syndrome. It is possible that the attenuation of opioid withdrawal in humans by the sst 2 -preferring agonist octreotide may be due in part to the physical interactions of these two receptors in the locus coeruleus (34). However, future coimmunoprecipitation studies from rat brain tissue are necessary to elucidate whether physical interaction between sst 2A and MOR1 receptors may also occur in vivo.
In conclusion, we provide biochemical and functional evidence for heterodimerization of the sst 2A somatostatin and the -opioid receptor. We show that formation of heterodimers between somatostatin and opioid receptors selectively crossmodulates phosphorylation, internalization, and desensitization. Direct intramembrane protein-protein interactions may thus provide a novel regulatory mechanism that could either restrict or enhance the activation/deactivation cycle of G protein-coupled receptors.