Oligomerization of µ and δ Opioid Receptors: Generation of Novel Functional Properties

Transcription

1 JBC Papers in Press. Published on June 6, 2 as Manuscript M3452 Oligomerization of µ and δ Opioid Receptors: Generation of Novel Functional Properties Susan R. George 1,2,3, Theresa Fan 3, Zhidong Xie 1, Roderick Tse 1, Vincent Tam 1, George Varghese 1 and Brian F. O Dowd 1,3 Departments of 1 Pharmacology and 2 Medicine, University of Toronto, Toronto and the 3 Centre for Addiction and Mental Health, Toronto, Ontario, Canada Address correspondence to: Dr. Susan R. George Department of Pharmacology University of Toronto Medical Sciences Building, Room King s College Circle Toronto, Ontario M5S 1A8 Canada Tel: (416) Fax: (416) s.george@utoronto.ca Running Title: Oligomerization of mu and delta opioid receptors Copyright 2 by The American Society for Biochemistry and Molecular Biology, Inc.

2 SUMMARY The existence of dimers and oligomers for many G protein coupled receptors (GPCRs) have been described by us and others. Since many GPCR subtypes are highly homologous to each other, we examined whether closely related receptors may interact with each other directly, and thus have the potential to create novel signaling units. Using µ and δ opioid receptors, we show that each receptor expressed individually was pharmacologically distinct and could be visualized following electrophoresis as monomers, homodimers, homotetramers and higher MW oligomers. When µ and δ opioid receptors were coexpressed, the highly selective synthetic agonists for each had reduced potency and altered rank order, whereas endomorphin-1 and Leu-enkephalin had enhanced affinity, suggesting the formation of a novel binding pocket. No heterodimers were visualized in the membranes coexpressing µ and δ receptors by the methods available. However, heterooligomers were identified by the ability to coimmuno-precipitate µ receptors with δ receptors, and vice versa, using differentially epitope tagged receptors. In contrast to the individually expressed µ and δ receptors, the coexpressed receptors showed insensitivity to pertussis toxin and continued signal transduction, likely due to interaction with a different subtype of G protein. In the present study, we provide for the first time, evidence for the direct interaction of µ and δ opioid receptors to form oligomers, with the generation of novel pharmacology and G protein coupling properties. 2

3 INTRODUCTION Opioid receptors have distinct pharmacological profiles and discrete but overlapping distributions in brain. The relatively recent cloning of opioid receptors has established that the products of three genes form the known subtypes, the µ, δ and κ opioid receptors, that interact with the complex family of endogenous opioid peptides (reviewed,1). The endogenous opioid peptide-receptor systems mediate important physiologic functions related to pain perception, locomotion, motivation, reward, autonomic function, immunomodulation and hormone secretion. The analysis of the contribution of each receptor type to the various opioid functions documented has been limited by the selectivity and cross-reactivity of the available opioid ligands and the postulation that multiple receptor subtypes are present. Since the cloning of the opioid receptors, the individual pharmacological and biochemical profiles of the µ, δ and κ opioid receptors have been better defined, however there are many aspects of opioid receptor biology that still remain poorly understood. A major problem that still remains is that the pharmacology of opioid receptors in brain tissue predicts a greater number of receptor subsites than revealed by opioid receptor cloning (1, 2). One possible explanation may be that the precise receptor-effector interactions present in endogenous brain regions may have not been adequately replicated in the heterologous expression systems in which the cloned receptors have been studied, or alternatively, that the relative proportion of the three receptor subtypes, variably expressed endogenously, may predicate the pharmacological profile, possibly by 3

4 receptor-receptor or distal interactions. The existence of such interactions or receptor cross-talk has long been postulated in the field of opioid receptor pharmacology (2,3), with functional (4,5) and biochemical (3,5) evidence for interactions between the µ and δ opioid receptors. Moreover, such interactions may be region-specific, as indicated by their occurrence in brain but not in peripheral locations (6,7). Intriguing observations in µ receptor knockout mice show that δ opioid receptor mediated analgesia was reduced and δ mediated respiratory depression was abolished, whereas κ opioid receptor function was preserved (8). These findings suggest that significant functional interactions may take place between µ and δ opioid receptors in specific neuronal pathways in brain. We and many other groups have described dimers and oligomers for many GPCRs mediating the actions of neurotransmitters, such as dopamine (9-12), serotonin (13), and for neuropeptide receptors such as for vasopressin (14) and opioid peptides (15). Dimerization of the δ opioid receptor has been suggested to have an important role in activation and internalization of the receptor (15). Since homodimerization appears to be an universal occurrence for GPCRs, we examined whether closely related receptors may heterodimerize and thus have the potential to create novel signaling units. For the Family A GPCRs, we showed that the related serotonin 5HT-1B and 5HT-1D receptors (68% identity overall, 77% in transmembrane domains) form homodimers and heterodimers (16). Among the Family B GPCRs, it has been shown that the GABA-B1 and GABA-B2 receptors heterodimerize and were dependent on this process for some aspects of signal transduction and cell surface localization (reviewed, 17). Since the µ and δ opioid receptors are highly homologous (share 65% amino acid 4

5 identity overall), we examined whether these closely related receptors could heterodimerize and whether they exhibited any evidence for direct modification of the distinct pharmacological profile of one when coexpressed with the other. While this work was in progress, a study showing heterodimerization of κ and δ opioid receptors was reported (18). In the present study, we provide for the first time, evidence for the direct interaction of µ and δ opioid receptors to form oligomers, and the generation of novel pharmacology and functional characteristics when µ and δ opioid receptors are expressed together, indicating that such interactions between these receptors may lead to novel signaling properties. 5

6 EXPERIMENTAL PROCEDURES Construction of µ and δ opioid receptor expression vectors cdnas encoding the rat µ and δ opioid receptors were inserted separately into the mammalian expression vector pcdna3. For the immunoprecipitation studies, the receptors were tagged with the c-myc or FLAG epitopes. The receptor cdnas were modified using the Transformer Site Directed Mutagenesis kit (Clontech) to insert after the NH 2 -terminal start methionine, an 11 residue c-myc epitope (EQKLISEEDL) for the µ opioid receptor and an 8 residue FLAG epitope (DYKDDDDK) for the δ opioid receptor. The absence of sequence errors and the correct orientation of the PCR products into the expression vectors were verified by sequencing on both strands. Expression in mammalian cells COS-7 monkey kidney cells and Chinese hamster ovary (CHO-K1) cells (American Type Culture Collection) were maintained as monolayer cultures at 37 C in minimal essential medium supplemented with 1% fetal bovine serum and antibiotics. COS-7 or CHO-K1 cells were transfected with the pcdna3 vectors using Lipofectamine (BRL). An equal amount of pcdna3 vector was co-transfected with each receptor construct so that the total amount of DNA used was consistent with studies involving transfections with two constructs. The receptor expression levels ranged between 7-12 fmol/mg protein. 6

7 Membrane Preparation Cells were washed extensively with PBS. Cell lysate was prepared by polytron disruption in ice-cold 5 mm Tris-HCl, 2 mm EDTA buffer, containing 5 µg/ml leupeptin, 1 µg/ml benzamidine and 5 µg/ml soybean trypsin inhibitor as described previously (13). The lysate was subjected to centrifugation at 1 x g to pellet unbroken cells and nuclei and to recover the supernatant which was centrifuged at high speed (4 x g for 2 min at 4 C) to prepare the crude membrane fraction. Membranes were washed with 5 mm Tris-HCl, 2 mm EDTA buffer containing protease inhibitors and centrifuged at high speed to prepare P2 membranes. Membrane protein was determined by the Bradford assay (BioRad) according to the manufacturer s instructions. Radioligand binding assays Saturation binding experiments were performed with ~2 µg of P2 membrane protein from Cos cells incubated with increasing concentrations of [3H]naloxone or [3H]diprenorphine and used to determine receptor densities (Bmax) and ligand affinities (KD) as previously described (19). Each concentration was examined in duplicate and incubated for 2 hours at 22 C in a total volume of 1 ml binding buffer (5 mm Tris-HCl, 5 mm EDTA, 1.5 mm CaCl2, 5 mm MgCl2, 5 mm KCl, 12 mm NaCl) with protease inhibitors. Nonspecific binding was defined as that not displaced by 1 µm naltrexone. The whole cell binding assays were conducted at 4 C and the incubation period was 4 hours. Competition experiments were performed in triplicate with increasing 7

8 concentrations of competing ligand ( M). The concentration of radioligand used in the competition assays was approximately equivalent to its KD. Bound ligand was isolated by rapid filtration through a 48-well cell harvester (Brandel) using GF/C filters (Whatman). Filters were washed with 1 ml of cold 5 mm Tris-HCl buffer (ph 7.4) and placed in glass vials with scintillation fluid (Universol) and counted for tritium. Data were analysed by nonlinear least-squares regression, using the computer-fitting program Prism (GraphPad). A two-site fit was designated only when a statistically significant improvement of the fit over a one-site model was obtained (by comparison of the coefficients of the goodness-of-fit by an F test). Adenylyl Cyclase Activity Adenylyl cyclase assays were conducted essentially as described (13). The assay mix contained.2 ml of membrane suspension from CHO cells (1-25 µg of protein),.12 mm ATP,.1 mm camp,.53 mm GTP, 2.7 mm phosphoenolpyruvate,.2 U pyruvate kinase, 1 U myokinase, 1 µm forskolin and.13 µci of [ 32 P]ATP in a final volume of.5 ml. The mixture was incubated with M agonist at 22 C for 2 min and enzyme activities were determined. Reactions were stopped by the addition of 1 ml of an ice-cold solution containing.4 mm ATP,.3 mm camp and [3H]-cAMP (25, cpm). Cyclic AMP was isolated by sequential column chromatography using Dowex cation exchange resin and aluminum oxide. Data were analyzed by computer fitted nonlinear least-squares regression. 8

9 Gel electrophoresis and immunoblotting The membrane samples were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) under reducing conditions using 1% or 12% precast acrylamide gels (Novex) and transferred to nitrocellulose as described previously (16). Immunoreactivity was revealed with antibodies raised against the c-myc epitope (Santa Cruz) and FLAG epitope (Sigma) and HRP-conjugated goat antimouse IgG (BioRad) and the Enhanced Chemiluminescence Detection kit (Amersham). Immunoprecipitation Studies The P2 membrane pellet was resuspended and stirred at 4 C overnight in 5 ml of freshly prepared solubilization buffer, consisting of 1 mm NaCl, 1 mm Tris-HCl ph 7.4, 2% digitonin and 5 mm EDTA with protease inhibitors. The homogenate was centrifuged at 27 g for 2 min, and the solubilized fraction was washed and concentrated in Centriprep-3 four times with 2 ml cold Buffer A: 1 mm NaCl, 1 mm Tris-HCl ph 7.4, with protease inhibitors. The washed fraction was precleared at 4 C with a 1/4 of agarose-fixed goat anti-mouse IgG overnight. The solubilized receptors were immunoprecipitated with the mouse monoclonal anti-myc or anti-flag antibody at a 1/5 dilution in buffer A for 6 h and agitated gently overnight with a 1/4 dilution of agarose-fixed goat anti-mouse IgG. The immunoprecipitate was washed with cold buffer A, solubilized in SDS sample buffer, and electrophoresed by SDS-PAGE. 9

10 RESULTS Interactions between µ and δ receptors Membranes expressing epitope-tagged µ or δ opioid receptors subjected to electrophoresis revealed receptor species detected as monomers, dimers, tetramers and higher order oligomers (Fig. 1A). The approximate molecular weights of the specific bands detected were as follows: for the µ opioid receptor, 9 kda (monomer), 18 kda (dimer), >3 kda (tetramer), and for the δ opioid receptor, 6 kda (monomer), 12 kda (dimer) and ~24 kda (tetramer). The addition of an epitope to the NH 2 -terminus had no effect on the electrophoretic mobility of the receptor species visualized on SDS- PAGE (data not shown). Also, the c-myc-µ receptor could not be immunoprecipitated by the FLAG antibody or vice versa for the FLAG-δ receptor from membranes expressing only one receptor subtype. No distinct heterodimer or heterooligomer bands could be discerned in membranes from cells coexpressing both receptors, whereas the monomers, homodimers and tetramers of each were easily discerned. When the receptors were coexpressed, both µ and δ receptors were detectable by Western blotting in the same membranes (Fig. 1B). In membranes coexpressing both receptors, using differential immunoprecipitation, c-myc-µ receptors were co-precipitated by the FLAG antibody, and FLAG-δ receptors by the c-myc antibody (Fig.1C), indicating an interaction between the µ and δ receptors in the cell, probably at an oligomeric level. The co- 1

11 immunoprecipitation was specific and did not occur for other GPCRs epitope-tagged similarly and expressed with the c-myc-µ opioid receptor or when membranes expressing the receptors individually were mixed together (data not shown). FIGURE 1 Pharmacological profile of coexpressed µ and δ receptors Membranes from Cos cells transfected with cdna encoding µ, δ or both receptors bound the non-selective [ 3 H]-naloxone and [ 3 H]-diprenorphine with high affinity, which was not altered under the coexpression conditions. The ability of highly selective µ and δ opioid ligands to interact with the receptors expressed separately or coexpressed was examined. The µ- and δ-selective synthetic opioid agonists had high affinity for the receptors expressed alone, but had altered affinity and a different rank order of affinity for the coexpressed µ-δ receptors. The µ-selective agonist DAMGO exhibited >16 fold greater affinity for the high affinity state of the µ receptor (indicated by the lower K H ) compared to the δ receptor high affinity state, and the δ-selective agonist DPDPE had >1, fold greater affinity for the δ receptor over µ (Table I). TABLE I When µ and δ receptors were coexpressed, there was a 1 fold reduction in affinity for DAMGO, DPDPE and morphine, as indicated by the higher affinity constant KH detected (Fig. 2A, B, C). To confirm that these findings did not result from the simple admixture of discrete µ and δ binding sites, these studies were repeated in membranes pooled together from cells separately expressing either µ or δ receptors. As shown in Fig. 3, recognition of µ opioid receptors by the µselective ligand DAMGO was not affected by 11

12 the mere presence of δ receptors in the membrane mixture (Fig. 3A), and similarly for DPDPE recognition of δ opioid receptors (Fig. 3B), as the respective KH values and receptor fractions in the high and low affinity states remained unchanged. FIGURE 2, FIGURE 3 The rank order of the affinities of a series of compounds at the agonist-detected high and low affinity states of the receptor was examined as shown in Table I. For the coexpressed µ and δ opioid receptors, a comparison of the binding profiles of several of the endogenous opioid peptides revealed ~2-3 fold enhanced potencies of one of the endomorphin peptides and Leu-enkephalin at the high affinity site, with a small reduction in the affinity for Met-enkephalin. For the coexpressed µ-δ receptors the affinities for DAMGO and DPDPE were equivalent. In order to confirm that the observed alterations in pharmacology resulted entirely from µ and δ receptor interactions at the cell surface, radioligand binding studies were repeated in whole cells. The effects of receptor coexpression to result in reduced affinity for DAMGO and DPDPE were replicated in a manner identical to that observed in the membrane assays, as shown in Fig. 4. FIGURE 4 Sensitivity to guanine nucleotides Incubation of membranes expressing µ opioid receptors with GTPγS 5-8 µm resulted in a shift of DAMGO-detected affinities to 1 fold lower values, with no change in the proportion of receptors in the high (RH 64% in control vs. 7% with GTPγS 8 µm) and low affinity states. We have previously reported partial desensitization of 12

13 µ opioid receptors occurring with guanine nucleotide analogs, with a shift of receptor affinity to the right but without receptor uncoupling from G protein (19). In contrast to membranes exclusively expressing µ (Fig. 5A) and δ (Fig. 5B) opioid receptors, GTPγS had no effect on the two affinity states detected by DAMGO when the µ and δ receptors were coexpressed (Fig. 5C), suggesting insensitivity to guanine nucleotides (RH 6 % in control vs. 72% with GTPγS 8 µm). FIGURE 5 Sensitivity to pertussis toxin To determine the involvement of G proteins in the generation of the agonistdetected high affinity states, cells expressing the µ or δ receptor or coexpressing µ and δ opioid receptors were treated with pertussis toxin 1 µg/ml for 24 hrs and the membranes harvested. Pertussis toxin treatment resulted in a complete loss of the DAMGO-detected high affinity state of the µ opioid receptor (RH 5% in control and % with PTX) as reported previously (2), with 1% of the receptors existing in a single low affinity state, indicating uncoupling from G protein (Fig. 6A). Similarly in cells expressing the δ receptor, pertussis toxin treatment resulted in complete loss of the DPDPE-detected high affinity sites (RH 48% in control and % with PTX) (Fig. 6B). In the µ-δ expressing membranes, pertussis toxin treatment had no effect on the two affinity states detected by either DAMGO or DPDPE (Fig. 6C, D). These results suggest that the affinity states detected when the receptors are coexpressed, are either not indicative of subpopulations of receptors coupled to or uncoupled from G protein, or alternatively, may result from 13

14 coupling of the receptors to a pertussis toxin-insensitive G protein. FIGURE 6 Desensitization and internalization The effect of agonist activation on receptor desensitization and internalization were assessed for µ and δ opioid receptors expressed separately and together. In cells expressing µ opioid receptors, 74% of receptors were in the agonist-detected high affinity state and 26% in the low affinity state, and following exposure to DAMGO for 1 hr, receptor density on the cell surface was reduced by 3%, but the affinities for agonist (KH and KL) and the relative proportion of receptors in the two states was unchanged (Fig. 7A). For the δ opioid receptor, exposure to DPDPE for 1 hr resulted in a reduction of agonist potency and a 23% reduction in cell surface receptor density (Fig. 7B). In contrast, in cells coexpressing µ and δ opioid receptors no desensitization was evident, however, agonist-induced internalization was preserved and even enhanced with DAMGO but abolished with DPDPE with no changes in affinity for either agonist detected (Fig. 7C and D). FIGURE 7 Effect of µ and δ receptor blockade In order to determine whether the pharmacological profile of the coexpressed µ-δ opioid receptors could be affected by blockade of the component µ and δ opioid receptors, the agonist competition studies were performed in the presence of selective antagonists of each receptor subtype. FIGURE 8 14

15 In the presence of naltrindole, a selective antagonist of the δ opioid receptor, the overall density of the coexpressed receptors detected was reduced by 4-5 %, however the K H for DAMGO and the proportion of receptors in the two affinity states were unaffected, although the K L was shifted to the left (Fig. 8A). Treatment of the coexpressed receptors with the irreversible µ opioid receptor antagonist β FNA also reduced apparent receptor density by 5-7%, and revealed no changes in K H for DPDPE (Fig. 8C). In separate experiments, the effects of the naltrindole concentration on DAMGO competition of µ opioid receptor binding (Fig. 8B), and of β FNA concentration on DPDPE competition of δ opioid receptor binding (Fig. 8D) were assessed and found to have no significant effect on the agonist-detected sites detected. Adenylyl Cyclase Activity Adenylyl cyclase activity was assessed in membranes from cells expressing the µ and δ receptors separately or coexpressing them. Membranes expressing µ opioid receptors revealed inhibition of forskolin-stimulated cyclic AMP production by DAMGO in a dose-responsive manner, with EC nm; this activity was completely abolished by treatment of the cells with pertussis toxin (Fig. 9A). Similarly, in membranes expressing δ opioid receptors, DPDPE inhibited forskolin-stimulated adenylyl cyclase activity with an EC nm, also abolished by pertussis toxin treatment (Fig. 9B). In membranes coexpressing µ and δ receptors, following pertussis 15

16 toxin treatment, there was continued ability of DAMGO (Fig. 9C) and DPDPE (data not shown) to inhibit forskolin stimulated camp production at concentrations ~1 µm (n=3), suggesting the involvement of a pertussis toxin-insensitive G protein. FIGURE 9 DISCUSSION In this study we report that µ opioid receptors heterooligomerize with δ opioid receptors to form a novel signaling entity with distinct radioligand binding and functional properties, different from either the µ or δ opioid receptors. When µ opioid receptors were coexpressed with δ opioid receptors, the resulting µ-δ oligomeric complex had 1 fold lower affinity for the µ-selective and δ-selective agonists DAMGO and DPDPE respectively. The rank order of agonist affinities for a series of selected agonists were different for the µ-δ receptors compared to the profiles for µ or δ receptors, suggesting the formation of a binding pocket different from that present when µ or δ receptors were expressed singly. In addition, there was a small increase in the affinities of some of the endogenous opioid peptides for the high affinity site of the µ-δ receptor complex when compared to the µ opioid receptor by a factor of 2 to 3, suggesting a possible preferential 16

17 interaction with this entity over µ receptors. The novel binding site was dependent on the coexpression of the two receptors and could not be reproduced by simple admixture of membranes expressing each receptor separately. The preservation of the binding characteristics of DAMGO-detected µ opioid sites and of DPDPE-detected δ opioid sites in the mixed membrane preparation adds credence to the notion that a novel µ-δ binding site has been created in the coexpressing membranes. The agonist affinities for the µ-δ receptor were not affected by addition of guanine nucleotide, or pertussis toxin treatment of the cells, suggesting that the agonistdetected binding sites may not result from coupling to a G protein, or alternatively, may result from coupling to a pertussis toxin-insensitive G protein. Precedence for the latter possibility has been established previously (21, 22), with evidence that opioid receptors can couple to multiple G proteins, which in turn may couple to one or more effectors in different experimental models. More specifically, opioid peptide activation of pertussis toxin-insensitive G proteins such as Gz and possibly others, has been shown in physiological settings, such as mouse brain (21) and in a neuroblastoma cell line (22). The ability of the µ-δ receptor complex to inhibit adenylyl cyclase activity following pertussis toxin treatment suggests coupling to a G protein which is pertussis toxin insensitive, such as Gz, which has been shown to link to inhibition of adenylyl cyclase activity (23). The involvement of other effector pathways by the µ-δ complex remains to be established. Agonist treatment of the coexpressed µ-δ receptors revealed significant differences compared to µ or δ receptors expressed alone. In the combined presence of µ 17

18 and δ opioid receptors, there was resistance to desensitization and internalization upon exposure to DPDPE, although exposure to DAMGO resulted in an accentuated loss of receptors from the membrane. The ability to immunoprecipitate the µ opioid receptor using the epitope on the δ opioid receptor, and vice versa, suggests a very close interaction between these receptors in the cell membrane. The relative SDS-resistance of the homodimers of µ and δ receptors seen when each was expressed separately, suggests the presence of covalent and/or strong hydrophobic interactions between receptor monomers to form homodimers, however no heterodimers were visualized. If indeed heterodimers of µ and δ opioid receptors were present in the cell, the inability of the heterodimers to survive SDS-PAGE may indicate the presence of a less stable interaction, or alternatively, may indicate that the homodimers of each may participate in the formation of heterotetramers and heterooligomers. The precise nature of the structural determinants resulting in the interactions between µ and δ receptors remains to be elucidated. That the interaction between µ and δ opioid receptors resulted in the creation of a unique binding site is evident from the pharmacological profile of the coexpressed receptors. This interaction is present at the cell surface, as indicated by the identical radioligand binding parameters in whole cell binding as in membranes. This would suggest that distinct conformational changes occurred, altering the original binding pockets of the µ and δ receptors and even altering the conformation of the G protein interacting intracellular domains. The finding that blockade of one receptor with a selective antagonist did not restore binding of the other suggests that the binding site is 18

19 indeed novel, rather than occurring as a result of altered cooperativity between ligand binding sites on adjacent µ and δ receptors. Several studies have attempted to delineate the sites of interaction of receptors with G proteins, which although not identified, have been shown to depend on agonist mediated conformational changes. For the opioid receptor, different agonists have been shown to promote the exposure of distinct intracellular regions that activate specific G proteins, such as Gi, Go or Gz (22). In an analogous manner, conformational changes resulting from receptor:receptor interactions may also present a different aspect of the intracellular domains to interacting proteins and thus may alter the specificity of these interactions. In summary, we have provided evidence for the heterooligomerization of µ and δ opioid receptors, which results in the generation of a novel binding site with a significantly altered pharmacological profile compared to the receptors expressed individually. This novel binding entity may interact with a different complement of signal transducing G proteins and appears to have distinct patterns of desensitization and internalization upon exposure to µ and δ agonists. These data indicate that such interactions between the µ and δ receptors, and between the κ and δ receptors described recently (18), will enable the generation of a greater diversity of opioid signaling receptor units than predicted by the three cloned opioid receptor genes, and provides greater physiological significance for the overlapping distributions of the closely related opioid receptor subtypes present in brain. 19

20 Acknowledgements: This work was supported by grants from the Medical Research Council of Canada, the National Institutes of Health NIDA, and the Smokeless Tobacco Research Council, Inc. REFERENCES 1. Kieffer, B. L. Opioids: first lessons from knockout mice. Trends Pharmacol. Sci. (1999) 2, Traynor, J. R. and Elliott, J. delta-opioid receptor subtypes and cross-talk with mu-receptors. Trends Pharmacol Sci. (1993) 14, Rothman, R. B., Holaday, J. W. and Porreca, F. Allosteric coupling among opioid receptors: evidence for an opioid receptor complex. (1992) In: Herz A., Ed. Handbook of experimental pharmacology, Vol. 14 pp Springer Verlag, Berlin 2

21 4. Porreca, F., Takemori, A. E., Sultana, M., Portoghese, P. S., Bowen, W. D and, Mosberg, H. I. Modulation of mu-mediated antinociception in the mouse involves opioid delta-2 receptors. J. Pharmacol. Exp. Ther. (1992) 263, Rothman. R. B., Long, J. B., Bykov, V., Jacobson, A. E., Rice, K. C. and Holaday, J. W. Beta-FNA binds irreversibly to the opiate receptor complex: in vivo and in vitro evidence. J Pharmacol Exp Ther (1988) 247, Franklin, T. G. and Traynor, J. R. Alkylation with beta-funaltrexamine suggests differences between mu-opioid receptor systems in guinea-pig brain and myentericplexus. Br J Pharmacol. (1991) 12, Elliott, J. and Traynor, J. R. Evidence for lack of modulation of mu-opioid agonist action by delta-opioid agonists in the mouse vas deferens and guinea-pig ileum. Br J Pharmacol. (1995) 114, Matthes, H. W., Smadja, C., Valverde, O., Vonesch, J. L., Foutz, A. S., Boudinot, E., Denavit-Saubie, M., Severini, C., Negri, L., Roques, B. P., Maldonado, R. and Kieffer, B. L. Activity of the delta-opioid receptor is partially reduced, whereas activity of the kappa-receptor is maintained in mice lacking the mu-receptor. J Neurosci. (1998) 18, Zawarynski, P., Tallerico, T., Seeman, P., Lee, S. P., O Dowd, B. F. and George, S. R. Dopamine D2 receptor dimers in human and rat brain. FEBS Lett. (1998) 441, Ng, G. Y., Mouillac, B., George, S. R., Caron, M., Dennis, M., Bouvier, M. and O Dowd B. F. Desensitization, phosphorylation and palmitoylation of the human 21

22 dopamine D1 receptor. Eur J Pharmacol. (1994) 267, George, S. R., Lee, S., Varghese, G., Zeman, P., Seeman, P., Ng, G. Y. K. and O Dowd, B. F. A transmembrane domain-derived peptide inhibits D1 dopamine receptor function without affecting receptor oligomerization. J. Biol. Chem. (1998) 273, Nimchinsky, E. A., Hof, P. R., Janssen, W. G. M., Morrison, J. H. and Schmauss, C. Expression of dopamine D3 receptor dimers and tetramers in brain and in transfected cells. J Biol Chem. (1997) 272, Ng, G. Y., George, S. R., Zastawny, R. L., Caron, M., Bouvier, M., Dennis, M., O Dowd, B. F. Human serotonin1b receptor expression in Sf9 cells: phosphorylation, palmitoylation, and adenylyl cyclase inhibition. Biochem. (1993) 32, Zhu, X. and Wess, J. Truncated V2 vasopressin receptors as negative regulators of wild-type V2 receptor function. Biochem.. (1998) 37, Cvejic, S., Devi, L. A. Dimerization of the delta opioid receptor: implication for a role in receptor internalization. J Biol. Chem. (1997) 272, Xie, Z., Lee, S. P., O Dowd, B. F. and George, S. R. Serotonin 5HT-1B and 5HT-1D receptors form homodimers when expressed alone and heterodimers when coexpressed. (1999) FEBS Lett. 456, Marshall, F. H., Jones, K. A., Kaupmann, K. and Bettler, B. GABAB receptors - the first 7TM heterodimers. (1999) Trends Pharmacol Sci. 2, , 18. Jordan, B. A., Devi, L. A. G-protein-coupled receptor heterodimerization modulates receptor function. (1999) Nature. 399,

23 19. Pak, Y. S., Kouvelas, A., Schiedeler, M., Rassmussen, J., O Dowd, B. F. and George, S. R. Functional desensitization of the mu opioid receptor is mediated by loss of receptors from the membrane rather than uncoupling from G protein. (1996) Mol. Pharm. 5, Pak, Y., O Dowd, B.F. and George, S.R. Agonist-induced, G protein-dependent and -independent internalization of the mu opioid receptor. (1999) J. Biol. Chem. 274, Garzon, J., Castro, M. and Sanchez-Blazquez, P. Influence of Gz and Gi2 transducer proteins in the affinity of opioid agonists to µ receptors. (1998) Eur. J. Neurosci. 1, Allouche, S., Polastron, J., Hasbi, A., Homberger, V. and Jauzac, P. Differential G-protein activation by alkaloid and peptide opioid agonists in the human neuroblastoma cell line SK-N-BE. (1999) Biochem. J. 342, Obadiah, J., Avidor-Reiss, T., Fishburn, C.S., Carmon, S., Bayewitch, M., Vogel, Z., Fuchs, S. and Levevi-Sivan, B. Adenylyl cyclase interaction with the D2 dopamine receptor family; differential coupling to Gi, Gz and Gs. (1999) Cell. Mol. Neurobiol. 19,

24 Footnote: Abbreviations: K H = affinity constant of the agonist-detected high affinity site; DADLE = [D-Ala 2, D-Leu 5 ]-enkephalin; DAMGO = [D-Ala 2, N-Me-Phe 4, Gly 5 -ol]- enkephalin; DPDPE = [D-Pen 2,5 ]-enkephalin. 24

25 TABLE I AGONIST µ δ µ-δ K H / K L (nm) K H / K L (nm) K H / K L (nm) DADLE.26/536.7/125.5/167 DAMGO / / / / / / DPDPE Morphine.22/ / /262 Endomorphin-2 (YPFF) 4.45/ /1969 Endomorphin-1 (YPWF) Met-enkephalin 6.3/ / / / /54 Leu-enkephalin 6.3/ / /653 β-endorphin Dynorphin A 17./47 22./ / / / /122 Values shown are the mean and S.E.M. of n=7-8 determinations, or the average of n=2 determinations. Abbreviations: K H/L = affinity constant of the agonist-detected high (H) or low (L) affinity sites or the single site detected; DADLE = [D-Ala 2, D-Leu 5 ]-enkephalin; DAMGO = [D-Ala 2, N-Me-Phe 4, Gly 5 -ol]-enkephalin; DPDPE = [D-Pen 2,5 ]- enkephalin. FIGURE LEGENDS Figure 1: Western blots of membranes from Cos cells (A) expressing either µ or δ 25

26 opioid receptors, (B) coexpressing µ and δ receptors and (C) immunoprecipitated from membranes coexpressing both receptors. Epitope-tagged c-myc-µ and FLAG-δ were immunoblotted (i.b.) with the c-myc and FLAG antibodies in membranes from cells expressing each receptor singly (Lanes 2, 4), from untransfected cells (Lanes 1, 3) and from cells coexpressing both receptors (Lanes 5, 6). Coexpressed receptors were immunoprecipitated (i.p.) with FLAG antibody and immunoblotted with c-myc antibody (Lane 7) and immunoprecipitated with c-myc antibody and immunoblotted with FLAG antibody (Lane 8). Figure 2: Competition of 3 H-naloxone binding to membranes from Cos cells expressing µ or δ opioid receptors singly (clear symbols) or coexpressing them (solid symbols), by (A) DAMGO, (B) DPDPE and (C) morphine. Each assay was performed in triplicate, and the data shown are representative of n=8 experiments for DAMGO and DPDPE, and of n=2 for morphine. Figure 3: Competition by (A) DAMGO and (B) DPDPE of 3 H-naloxone binding to membranes from Cos cells expressing µ opioid receptors or δ opioid receptors singly (clear symbols) or mixed together (solid symbols). Each assay was performed in triplicate, and the data shown are representative of n=3 experiments. Figure 4: Competition by (A) DAMGO and (B) DPDPE of 3 H-naloxone binding to 26

27 whole cells (Cos) expressing µ opioid or δ opioid receptors singly or coexpressing them. Each assay was performed in triplicate and the data shown are representative of n=2 experiments. Figure 5: Effect of GTPγS (8 µm) on DAMGO competition of 3 H-naloxone binding to membranes from Cos cells (A) expressing µ opioid receptors and (B) expressing δ opioid receptors, and (C) coexpressing µ and δ opioid receptors. Figure 6: Effect of pertussis toxin (PTX) treatment of Cos cells expressing µ or δ opioid receptors separately, or coexpressing µ and δ opioid receptors. Competition of 3H-naloxone binding by (A, C) DAMGO and (B, D) DPDPE are shown. Figure 7: Effect of selective agonist treatment for 1 hour on competition of 3 H- naloxone binding by (A) DAMGO in membranes from cells expressing µ receptors, (B) DPDPE in membranes from cells expressing δ receptors, (C) DAMGO and (D) DPDPE in membranes from cells coexpressing µ and δ receptors. Figure 8: Effect of selective antagonist treatment on competition of 3 H-naloxone binding to membranes from Cos cells expressing µ opioid or δ opioid receptors singly or together. Shown is DAMGO competition with and without naltrindole 5nM in membranes from cells (A) coexpressing µ and δ receptors and in (B) expressing µ 27

28 receptors, and DPDPE competition with and without βfna 5nM in membranes from cells (C) coexpressing µ and δ receptors and in (D) expressing δ receptors. Figure 9: Effect of DAMGO and DPDPE on forskolin-stimulated adenylyl cyclase activity in membranes from CHO cells expressing (A) µ opioid receptors, (B) δ opioid receptors, and (C) coexpressing µ and δ receptors following vehicle treatment (control) or pertussis toxin treatment (PTX). Values shown are representative of 3 separate experiments. 28

29 29

30 B. myc-µ/ flag-δ myc-µ/ flag-δ - myc - flag kda C. myc-µ/ flag-δ myc-µ/ flag-δ flag myc myc flag kda A. myc-µ flag-δ i.p. i.b. - myc - flag kda myc UT UT - flag kda -25

31 % bound A. B. C µ µ-δ [D AM GO] M % bound δ µ-δ [DPD PE]M % bound µ µ-δ [MOR PHINE]M Fig.2 George et al.

32 A. B. % bound MIXED µ µ+δ [DAM GO ]M % bound δ MIXED µ+δ [DPD PE] M Fig.3 George etal.

33 A. B. % bound µ µ-δ [D AM GO] M δ % bound δ µ-δ [DPD PE]M µ Fig.4 G eorge etal.

34 % bound A µ +GTPγS [DA M GO] M % bound B. C δ [DPD PE] M +GTPγS % bound µ-δ +GTPγS [DA M GO] M Fig.5 Ge orge etal.

35 A. B. % bound µ +PTX % bound δ +PTX 2 2 % bound [DAM G O ] M C. D. µ-δ +PTX [D AM GO] M % bound [DPDPE] M µ-δ +PTX [DPD PE]M Fig.6 George et al.

36 A. B. receptor density fmol/mg receptor density fmol/mg DA MGO treated µ control [DAM GO] M 8 µ δ DA M G O treated µ δ control receptor density fmol/mg C. D. receptor density fmol/mg 1 DP DPE treated δ control [DPD PE] M µ δ control DP DP E treated [DAM GO] M [DPD PE] M Fig.7 Ge orge etal.

37 A. B. % bound % bound NTD µ-δ [D AM GO] M µ-δ +βfna [DPD PE] M % bound 1 C. D. % bound µ +NTD [D AM GO] M δ +βfna [DPD PE]M Fig.8 George et al.

38 Fig.9 George etal. % forskolin-s timulated cam P accum ulation 1 9 control µ [DAM G O ] M PTX %forskolin-stimulated cam P accum ulation 1 9 control δ PTX [DPDPE]M %forskolin-stimulated cam P accum ulation 1 9 control µ-δ PTX [DAM GO] M

39 Oligomerization of mu and delta opioid receptors: generation of novel functional properties Susan R. George, Theresa Fan, Zhidong Xie, Roderick Tse, Vincent Tam, George Varghese and Brian F. O'Dowd J. Biol. Chem. published online June 6, 2 Access the most updated version of this article at doi: 1.174/jbc.M3452 Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's alerts