Abstract
To explore the feasibility of developing inhibitors of signaling by opioid receptors and other G protein-coupled receptors (GPCRs) that use the same G protein pool, we investigated the capacity of a minigene encoding the third intracellular loop of the δ-opioid receptor (δ-i3L) to act as competitive antagonist of the receptor-G protein interface interaction. In δ-i3L-expressing cells, the peptide blocked high-affinity agonist binding to both the δ- and the μ-opioid (δ-OR and μ-OR) and attenuated opioid and α2-adrenergic receptor (α2AR)-dependent [35S]guanosine-5′-O-(3-thio)triphosphate binding. Furthermore, δ-i3L expression resulted in inhibition of δ-, μ-OR-, and α2AR-receptor-mediated cAMP accumulation, whereas the cAMP response produced by activation of the β2-adrenergic receptor was unaffected, suggesting that the inhibitory effects of δ-i3L expression were selective for Gi/Go proteins. Moreover, although δ-i3L expression also attenuated drastically phospholipase C accumulation and Ca2+ release following μ- and δ-OR stimulation, it failed to inhibit carbachol-mediated stimulation of inositol phosphate accumulation in M1-muscarinic receptor-expressing human embryonic kidney 293 cells. Finally, we also examined the effects of δ-i3L expression on the regulation of the extracellular signal-regulated kinase (ERK) mitogen-activated protein kinase pathway. Our results demonstrate that, although ERK activation by μ- and δ-ORs is attenuated by the presence of δ-i3L, ERK activation mediated by α2AR remained unaffected. Collectively, our data demonstrate that the δ-i3L can be used as potent inhibitor of G protein signaling for various GPCRs that use a common pool of G proteins.
Opiate drugs mediate their analgesic, euphoriant, and rewarding effects by activating opioid receptors (Waldhoer et al., 2004). Pharmacological and molecular studies have demonstrated the existence of three opioid receptor subtypes, μ, δ, and κ (μ-OR, δ-OR, and κ-OR), that couple to Gi/Go types of G proteins to inhibit adenylyl cyclase (Law et al., 1999, 2000). Opioid receptors regulate a number of second messenger systems, such as phospholipase C (PLC), mitogen-activated kinase (MAPK), and various ion channels and other signaling intermediates (Standifer and Pasternak, 1997; Law et al., 2000; Lo and Wong, 2004; Mazarakou and Georgoussi, 2005). Such diverse signaling events are mediated by various G protein subtypes in a PTX-sensitive or -insensitive manner depending on the system studied and the opioid agonists used (Georgoussi et al., 1993; Chan et al., 1995; Garzon et al., 1997; Sanchez-Blazquez et al., 2001; Lo and Wong, 2004).
Opioid receptors are activated by both endogenously produced opioid peptides and exogenously administered opiate compounds, some of which are not only among the most effective analgesics known but also highly addictive drugs of abuse. Various agents that act as agonists or antagonists of GPCRs represent the most common type of drug in clinical use today. Irrespective of chemical composition, these agents share a common feature of acting extracellularly and either mimicking or precluding agonist binding to the receptor. Thus, targeting the external ligand binding sites of GPCRs is the generally accepted strategy for designing antagonists.
An alternative approach for achieving functional inhibition of GPCRs is the targeting of the receptor-G protein interface with agents that block the intracellular coupling of the receptors to the G protein. Such an approach differs fundamentally from classical GPCR pharmacology, because the blockage of receptor-G protein coupling may result in the achievement of G protein rather than receptor-specific antagonism. Several successful applications of this strategy, using polypeptides derived from the putative contact surfaces of the receptor or the Gα subunits of G proteins, have been reported (Luttrell et al., 1993; Hawes et al., 1994; Gilchrist et al., 1999; Vanhauwe et al., 2002). For example, expression of peptides derived from the third intracellular domains of the Gq/11-coupled α1B-adrenergic and M1-muscarinic acetylcholine receptors, the Gi-coupled α2A-adrenergic and M2-acetylcholine receptors, and the Gs-coupled D1A dopamine receptor have been shown to inhibit G protein mediated-receptor signaling (Luttrell et al., 1993; Hawes et al., 1994). On the other hand, minigene plasmids encoding oligopeptides representing the carboxyl termini of various G proteins, such as Giα, Gqα, Gsα, and Gα12,13, have also been used to determine the contribution of different G protein pools to the signaling of various GPCRs (Gilchrist et al., 1999, 2001; Feldman et al., 2002).
Studies related to opioid receptor signaling mechanisms have demonstrated that the cytoplasmic face of these receptors, particularly the third intracellular loop and the COOH tail, are critical in mediating signal transduction by G proteins. Indeed, previous observations using receptor-derived peptides from specific regions of the δ- and the μ-ORs have shown that the third intracellular loop (i3L) and the juxtamembranous region of the C-terminal tail of the μ- and δ-OR are critical for functional receptor-G protein interaction (Merkouris et al., 1996; Georgoussi et al., 1997). These observations have confirmed that there are different determinants for receptor-G protein coupling and G protein activation.
Given these observations and the importance of the i3 loop for receptor-G protein coupling and G protein-mediated signal transduction upon receptor activation, we have investigated the feasibility of developing an analog that could inhibit G protein-mediated downstream effects for activated GPCRs that couple to a common G protein pool. Accordingly, we developed a minigene construct capable of directing the expression of the i3 loop of the murine δ-OR in various cells lines and measured its ability to inhibit the downstream signaling of a number of receptors that couple to Gi/Go proteins. Our results suggest that the presence of the δ-i3L minigene construct in the various cell lines not only prevents G protein coupling of δ-OR and μ-OR but also affects the functional coupling of other GPCRs that selectively interact with the same Gi/Go protein population. In contrast, δ-i3L expression has no effect on Gs- or Gq-coupled receptors. These results provide insight into the importance of the interplay between GPCRs and selective G protein pools in cellular regulation, suggest that opioid receptor-derived peptides can be used as selective inhibitors for such interactions, and point to novel strategies related to drug development for physiological perturbations caused by excessive GPCR activity.
Materials and Methods
Materials. [3H]DADLE (57 Ci/mmol), [3H]DAMGO (60 Ci/mmol), [3H]diprenorphine (50 Ci/mmol), and [3H]adenine (23 Ci/mmol) were from GE Healthcare (Little Chalfont, Buckinghamshire, UK). All receptor ligands were from Sigma-Aldrich (St. Louis, MO). The rat myc-tagged μ-OR (in the pcDNA3 vector) was generously provided by Dr. S. George (University of Toronto, Toronto, ON, Canada). Lipofectamine was purchased from Invitrogen (Carlsbad, CA). Protein A and protein AG-Sepharose beads were from GE Healthcare. PTX, GppNHp, protease, phosphatase inhibitor cocktails, and all other reagents were purchased from Sigma-Aldrich.
Construction of the δ-i3L and GST-δ-i3L Minigene Encoding the Third Intracellular Loop of the δ-OR. Two partially complementary oligonucleotides encoding the 23 amino acids of the third intracellular loop of murine δ-OR (amino acids 239–261) were engineered into 5′ and 3′ ends. The 5′ end contained a BamHI site followed by the ribosome binding consensus sequence (5′-GCC GCC ACC-3′), a methionine (ATG) for translation initiation, and a glycine (GGA) to protect the ribosome-binding site during translation and the nascent peptide from proteolytic degradation. An EcoRI site was designed at the 3′ end immediately following the translation stop codon (TAA). Partially complementary oligonucleotides were allowed to anneal, and full-length (completely complementary) double-stranded DNA was synthesized using T4 DNA polymerase in the presence of dNTPs. After digestion with the restriction enzymes BamHI and EcoRI, the cDNA was ligated to the eukaryotic plasmid vector pcDNA3. The presence of the insert was verified by automated dideoxynucleotide sequencing (ABI Prism 377 DNA sequencer; Applied Biosystems, Foster City, CA). The GST fusion construct encompassing the i3L (amino acids 239–261) of the δ-OR was generated using the cDNA clone of the rat δ-OR, as template for PCR and the TOPO cloning procedure (Invitrogen). Transformants were transferred from entry clones to the destination vector pDest-27 (for N-terminal GST fusion protein expression) using the GATEWAY cloning technology (Invitrogen).
Cell Culture and Transfection. HEK293 cells, stably expressing the EYMPME (EE)-tagged μ-opioid, δ-opioid, and β2-adrenergic receptors, were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 2 mM glutamine, 100 units/ml penicillin, and 0.1 mg/ml streptomycin under 5% CO2 at 37°C. Rat-1 fibroblasts stably expressing the α2-adrenergic receptor (generously provided by Prof. G. Milligan, University of Glasgow, Glasgow, Scotland) were grown in the same medium. Transfections were performed on ∼80% confluent monolayers in 100-mm dishes using Lipofectamine according to the manufacturer's instructions (Invitrogen). All assays on transiently transfected cells were performed after 48 to 72 h.
Expression of δ-i3L Peptide. To determine the presence of the δ-i3L minigene in the transiently transfected HEK293 cells, total RNA was extracted 68 h post-transfection and subjected to RT-PCR (Access RT-PCR system; Promega, Madison, WI). The PCR analysis was performed using cDNA as template with primers specific for the 3′ and 5′ ends of the insert (5′-GCCTGCGCAGCGTGCG-3′, 5′-TGATGCGCCGCAGGCT-3′). The separation of the PCR products was performed on 4% MetaPhor agarose gel (FMC Bioproducts, Rockland, ME). Control reactions were performed using RNA of cells transfected with the expression plasmid for δ-OR under the same experimental conditions. Minigene expression following transient transfection of HEK293 cells with the δ-i3L peptide was detected by Western blotting using an antibody raised against the amino acid sequence of the third intracellular loop of the δ-OR that recognizes both μ- and δ-ORs. Immunoreactive bands were detected by ECL detection kit (GE Healthcare) according to the manufacturer's instructions. The presence of the i3 loop peptide encoded by the δ-i3L minigene was also verified in the case of a GST fusion protein encompassing the third intracellular loop of the δ-OR (GST-δ-i3L construct). Briefly, HEK293 cells were transiently transfected with the GST-δ-i3L plasmid, and cell lysates were used for Western blotting. Immunoreactivity was detected using a GST antibody (1:5000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and horseradish peroxidase-conjugated anti-mouse IgG using enhanced chemiluminescence.
Generation of IgG Fractions from the i3L of the δ-OR. The antiserum Δi3L was produced in New Zealand White rabbits using a conjugate of a synthetic peptide corresponding to the i3L of the mouse δ-OR biochemically coupled to keyhole limpet hemocyanin (Calbiochem, San Diego, CA) as antigen. Antiserum Δi3L raised against this peptide was affinity-purified on protein A-Sepharose 4B (Sigma-Aldrich) as described by Georgoussi et al. (1993).
Coimmunoprecipitation Assays. HEK293 cells transiently transfected with myc-tagged μ-OR and GST-δ-i3L plasmid 72 h post-transfection were rinsed in PBS buffer containing 100 nM phenylmethylsulfonyl fluoride and 100 nM sodium orthovanadate. Cells were lysed in lysis buffer A containing 1% Triton X-100, 10 mM Tris, pH 7.6, 5 mM Na2EDTA, 50 mM NaCl, 50 mM NaF, and 30 mM Na4P2O7 supplemented with 2 μg/ml antipain, 2 μg/ml leupeptin, 2 μg/ml benzamidine, 1 mM phenylmethylsulfonyl fluoride, and 100 μM Na3VO4. Approximately 1 mg of the clarified cell lysate was incubated overnight at 4°C with either an anti-myc monoclonal antibody or with antibodies recognizing Goα and Gsα. Immune complexes were recovered on protein AG-Sepharose beads when monoclonal antibodies were used or on protein A-Sepharose beads in case of polyclonal antibodies. The samples were washed extensively and separated on SDS-PAGE. Precipitated samples were immunoblotted using an anti-GST antibody (Santa Cruz Biotechnology, Inc.).
Cell Membrane Preparations. Seventy-two hours post-transfection confluent monolayers of HEK293 cells and Rat-1 fibroblasts transiently transfected with vector alone or with the δ-i3L minigene were harvested, collected by centrifugation at 2000g for 5 min, and washed once with PBS at pH 7.5. The pellets were resuspended in ice-cold buffer B (10 mM Tris, pH 7.5, and 0.1 mM EDTA) and lysed with 30 strokes of the pestle of a Dounce homogenizer. Nuclei and unbroken cells were pelleted by centrifugation at 1000g for 2 min, and the supernatants were further centrifuged at 150,000g for 20 min at 4°C. The membrane pellet was resuspended in ice-cold buffer B at a protein concentration of approximately 1 mg/ml and stored in aliquots at –70°C. Protein concentration was measured according to the method of Bradford (Bradford, 1976).
Receptor Binding Experiments. In experiments designed to define ligand specificity, membranes (15–50 μg) were incubated at 30°C for 45 min in buffer containing 50 mM Tris-HCl, pH 7.5, 100 mM NaCl, and 10 mM MgCl2 when the opioid antagonist [3H]diprenorphine was used and in buffer containing 20 mM Tris-HCl, pH 7.5, 50 mM sucrose, and 20 mM MgCl2 when the opioid peptide agonists [3H]DAMGO and [3H]DSLET were used. In saturation experiments, the concentration of radioligand varied between 0.1 and 6 nM for the opioid agonists. Nonspecific binding was assessed in the presence of 10 μM naloxone and increased in a linear manner depending on the 3H-ligand concentration. The reaction was stopped by rapid filtration and three washes in ice-cold 50 mM Tris-HCl, pH 7.4, through GF/C filters (Whatman, Maidstone, UK) using an automated cell harvester (Brandel Inc., Gaithersburg, MD). The radioactivity was measured by liquid scintillation counting. In experiments where uncoupling of opioid receptors from G proteins was determined, membranes were preincubated for 45 min at 30°C with 100 μM GppNHp before ligand binding. Analysis of the binding data was performed using the RadLig 4.0 software (Biosoft, Cambridge, UK).
[35S]GTPγS Binding Studies. Specific [35S]GTPγS binding was measured as described by Merkouris et al. (1996). Briefly, 15 to 20 μg of membrane proteins was incubated in 20 mM HEPES, pH 7.4, 3 mM MgCl2, 100 mM NaCl, 10 μM GDP, 0.2 mM ascorbate, 0.3 to 0.5 nM [35S]GTPγS, and the appropriate ligand (1 nM–10 μM) in a final volume of 100 μl for 1 h at 4°C. Incubation was terminated by rapid filtration on presoaked GF/B filters followed by three washes with ice-cold 20 mM HEPES, pH 7.4, and 3 mM MgCl2 using a Brandel cell harvester. Bound radioactivity was determined by scintillation counting. Nonspecific binding was determined in the presence of 10 μM GTPγS.
Measurements of cAMP Accumulation. Measurements of adenylyl cyclase activity were performed as described by Merkouris et al. (1997). Briefly, HEK293 cells expressing the μ- and δ-opioid and the β2AR receptor and Rat-1 fibroblasts expressing the α2AR were transiently transfected with empty pcDNA3 vector or the δ-i3L minigene construct and cultured in 12-well plates. Forty-eight hours post-transfection, cells were incubated in medium containing [3H]adenine (1.5 μCi/ml) for 24 h. The generation of [3H]cAMP was assessed in response to treatment of the cells with the appropriate selective agonists (1 nM–10 μM) depending on the receptor expressed using 50 μM forskolin for 30 min at 37°C. Results are calculated as agonist-mediated percentage of inhibition of cAMP accumulation.
Measurement of Ins(1,4,5)P3 Formation. Forty-eight hours post-transfection, HEK293 cells expressing the μ- or δ-ORs were transiently transfected with empty pcDNA3 vector or δ-i3L minigene construct and seeded onto 35-mm culture dishes at a density of 2 × 105 cells/plate. [3H]Myoinositol (2 μCi/ml) was added to each plate, and cells were further incubated for 24 h. Cells were challenged with variable amounts of agonist in Hanks' balanced salt solution buffer consisting of 5.4 mM KCl, 0.3 mM Na2HPO4, 0.4 mM KH2PO4, 4.2 mM NaHCO3, 1.3 mM CaCl2, 0.5 mM MgCl2, 0.6 mM MgSO4, 137 mM NaCl, 5.6 mM d-glucose, and 21 mM HEPES, pH 7.4, supplemented with 10 mM LiCl at 37°C for 30 min. The reaction was stopped by removal of the drug and addition (1 ml) of stop solution [96% (v/v) methanol and 0.23% (v/v) HCl]. Following addition of chloroform (0.7 ml) and spinning at 1000g, for 5 min, the aqueous phase was removed, mixed with 2 ml of 5 μM myoinositol, and loaded onto ion exchange AG-501-X8 resin (Bio-Rad, Hercules, CA). The columns were further washed 3 × 4 ml with 5 μM myoinositol. The tritiated labeled product was eluted with 2 ml of 1 M ammonium formate. The levels of Ins(1,4,5)P3 production were determined by the ratio of the radioactivity bound in the aqueous and lipidic phases. Basal rates of Ins(1,4,5)P3 were determined by incubating transfected cells in the absence of agonist.
Detection of MAPK Phosphorylation. Cells stably expressing the δ-OR, μ-OR, α2AR, or EGF receptors were transiently transfected with empty pcDNA3 vector or the δ-i3L minigene expression construct and cultured in six-well plates for 48 h. Six hours before the addition of drugs, the culture medium was removed and replaced by fresh serum-free medium. The agonists were added to the cells and allowed to incubate in a time-dependent manner at 37°C. Cell monolayers were rinsed with PBS, and whole lysates were prepared by the addition of 120 μl of lysis buffer A supplemented with protease (P8340; Sigma-Aldrich) and phosphatase inhibitor cocktails (P5726; Sigma-Aldrich) for 30 min. Soluble proteins were separated by centrifugation at 15,000g for 10 min at 4°C. The cell lysates were analyzed for protein concentration using the Bradford method. The proteins were prepared for SDS-polyacrylamide gel electrophoresis (10% acrylamide running gel and 4% acrylamide stacking gel) and electroblotted onto polyvinylidene difluoride membranes. Membranes were exposed overnight at 4°C to mouse monoclonal antibodies that are specific to phosphorylated MAPK (1:1000 dilution in blocking solution, phospho-ERK; Santa Cruz Biotechnology, Inc.). Immunoreactive proteins were visualized using a horseradish peroxidase-sensitive ECL chemiluminescent Western blotting kit (GE Healthcare).
Intracellular Ca2+ Measurements. μ-HEK293 cells were incubated for 30 to 45 min in 10 mM HEPES, pH 7.4, 140 mM NaCl, 10 mM glucose, 5 mM KCl, 1 mM MgSO4, and 1.8 mM CaCl2 containing 2.5 mM probenecid (MP Biomedicals, Irvine, CA) (buffer C) and 5 mM Fluo-3 (Invitrogen). The cells were then washed three times with buffer C and incubated for 20 to 30 min in the same buffer. Fluorometric measurements after DAMGO (Sigma-Aldrich) stimulation at various concentrations were performed in a stirring cuvette using a LS50 fluorometer (PerkinElmer Life and Analytical Sciences, Boston, MA) with filter sets for excitation at 488 nm and emission at 525 nm (Swevers et al., 2005).
Results
Construction and Expression of a Minigene Encoding the Third Intracellular Loop of the δ-OR. The i3Ls of the δ- and μ-ORs are comprise some 23 amino acids and share a high degree of identity, with the exception of amino acids Arg241, Leu245, and Ser255 (in δ-OR), which are replaced with Lys260, Met264, and Asn274 in μ-OR (Fig. 1A). The i3L domain has been shown to be critical for G protein coupling and activation (Merkouris et al., 1996; Georgoussi et al., 1997) and to participate in receptor desensitization (Cen et al., 2001) and other aspects of signaling (Megaritis et al., 2000; Chaipatikul et al., 2003).
To determine whether we could selectively antagonize G protein signaling events of homologous or heterologous GPCRs by expressing peptides that interact with the receptor-G protein interface, we generated a minigene that encodes the third intracellular loop of the mouse δ-opioid receptor (δ-i3L) (Fig. 1A) and introduced it into HEK293 cells expressing the opioid receptors and Rat-1 fibroblasts expressing the α2AR by transfection. To confirm the expression of the δ-i3L minigene in transfected cells, total RNA was isolated 48 h post-transfection and analyzed by RT-PCR. The analysis confirmed the presence of a single 70-base pair band corresponding to the RT-PCR product of the δ-i3L minigene in the transfected cells (Fig. 1B). To verify the expression of the δ-i3L peptide in the transfected HEK293 cells, 48 h post-transfection the cells were harvested and subjected to SDS-PAGE. As shown in Fig. 1C, a band of approximately 3-kDa molecular mass was identified in cells that have been transfected with the minigene construct, which comigrated with the purified synthetic i3L peptide. On the other hand, upon transient transfection into HEK293 cells with another construct in which a GST-epitope was added to the NH2-terminal region of the i3L, a 30-kDa polypeptide was detectable in the transfected cell lysates after immunoblotting with a GST antibody (Fig. 1D).
δ-i3L Interacts with μ-OR and Goα but Not Gsα. To detect whether expression of the δ-i3L peptide targets the opioid receptor and/or the G protein(s), HEK293 cells were transiently transfected with expression constructs for a myc-tagged μ-OR and GST-δ-i3L. Lysates from these cells obtained 48 h post-transfection were immunoprecipitated with antibodies recognizing the GST, Goα, Gsα, or the myc-tagged receptor and immunoblotted with an anti-GST antibody. As demonstrated in Fig. 2, both the anti-myc and Goα antibodies coimmunoprecipitated the GST-δ-i3L, suggesting that once the δ-i3L peptide is expressed in HEK293 cells, it is capable of binding to the plasma membrane receptor and/or Goα. The coprecipitation of Goα with the i3L was independent of DAMGO stimulation status, confirming that the association of Goα with the i3L does not require activation of the receptor (Fig. 2, compare lanes 3 and 4). No bands were detected in similarly treated samples of nontransfected HEK293 cells (data not shown) or in lysates of transfected cells immunoprecipitated with normal mouse serum (Fig. 2A, lane 2). Similarly, no band was detected in cells that were transfected with a construct expressing GST alone (data not shown).
When the same cell lysates were immunoprecipitated with anti-Gsα, no coimmunoprecipitated δ-i3L was detectable with the GST-antibody, suggesting that Gsα does not interact with the expressed peptide. The absence of δ-i3L immunoreactivity in the immunoprecipitates cannot be attributed to lack of Gsα expression in HEK293 cells, because the latter was clearly detectable in the cell lysates (Fig. 2C). Collectively, these results suggest that the peptide not only recognizes and targets the expressed receptor but also that it directly interacts with Goα but not Gsα in HEK293 cells. In turn, this supports the notion that the peptide associates specifically with a certain G protein pool in HEK293 cells.
Effect of δ-i3L Expression on High-Affinity Agonist Binding to the μ- and δ-ORs. As shown in Table 1, specific [3H]DADLE and [3H]DAMGO binding to membrane protein preparations of δ- or μ-HEK293 cells expressing or not the δ-i3L minigene amounted to approximately 400 and 900 fmol/mg membrane protein, respectively. Expression of the δi3L peptide did not alter the Kd values for both ligands. On the other hand, the binding parameters of the specific [3H]diprenorphine demonstrated comparable Kd values and the existence of a higher number of receptors for both μ- and δ-opioid receptors. This is due to the higher potency of the opioid antagonist diprenorphine to bind to opioid receptors.
Given that the expressed δ-i3L peptide targets the receptor and the Goα protein, we examined whether its presence blocks the receptor-G protein interface, causing irreversible dissociation of the G protein(s). Accordingly, we measured the effectiveness of DSLET to displace specific [3H]diprenorphine binding from δ-HEK293 membranes obtained from cells that express stably the δ-OR under conditions that interfere with receptor-G protein interaction; i.e., in the presence of the poorly hydrolyzable GTP analog GppNHp. Figure 3A compares the competition binding curves generated for the displacement of [3H]diprenorphine by DSLET in the presence of either GppNHp or the δ-i3L peptide. GppNHp produced a marked rightward shift in the displacement of [3H]diprenorphine binding by DSLET (IC50 of 2.6 and 13.8 nM in the absence and presence of GppNHp, respectively). In an analogous manner, a rightward shift was observed for the ability of DSLET to displace [3H]diprenorphine binding in δ-HEK293 cells (IC50 of 2.6 and 9.9 nM in the absence and presence of the δ-i3L minigene, respectively. The effect of the δ-i3L minigene on high-affinity binding in cell membrane preparations is clearly seen in the competition experiments where the concentration range of the agonists extends to higher levels, unlike the saturation experiments where the tritiated agonists are used in lower concentrations.
Similar results were also obtained for the μ-OR-expressing HEK293 cells. Analysis of the displacement curves indicated that approximately 40% of the specific [3H]diprenorphine binding was displaced at 100 nM DAMGO in the absence of GppNHp, whereas in its presence, 100 nM DAMGO was unable to displace [3H]diprenorphine binding (Fig. 3B). Expression of δ-i3L, on the other hand, abolished the ability of DAMGO to displace [3H]diprenorphine binding. No differences were detected in the IC50 values between GppNHp-treated and δ-i3L-containing μ-HEK293 membranes. These results suggest that the presence of the peptide interferes in the receptor-G protein interaction of both δ- and μ-ORs and thus results in the conversion of these receptors to a low-affinity state, supporting the notion that δ-i3L binding to the G protein may prevent receptor-G protein coupling.
δ-i3L Expression Inhibits G Protein Activation. Because δ-i3L expression seems to target the receptor-G protein interface, we sought to determine whether the presence of the peptide in the cells affects G protein activation by measuring agonist-induced stimulation of [35S]GTPγS binding to membranes from δ- or μ-HEK293 cells after transfection with the δ-i3L or vector alone in the presence of 10 μM GDP. As shown in Fig. 4A, in untransfected cells and in cells transfected with the empty vector, DSLET produced a consistent stimulation of [35S]GTPγS binding in a dose-dependent manner. This stimulation was abolished when membranes from cells that coexpressed the δ-receptor and the δ-i3L minigene were used (Fig. 4A). When similar experiments were performed with membranes of cells that express the μ-OR, the expression of δ-i3L minigene also abolished the DAMGO-mediated stimulation of [35S]GTPγS binding that normally occurs upon receptor stimulation (Fig. 4B). These results suggest that the presence of the δ-i3L minigene blocks the association of G proteins with both the δ- and μ-ORs.
To assess whether the presence of δ-i3L peptide could block G protein-mediated activation of other receptors that interact with the same population of G proteins, we also expressed transiently the δ-i3L minigene in Rat-1 fibroblasts that stably express the α2-adrenergic receptor. Membrane protein preparations from these cells were measured for their ability to stimulate [35S]GTPγS binding by the α2AR agonist UK-14,304. Figure 4C clearly demonstrates that, although UK-14,304 stimulated [35S]GTPγS binding in a dose-dependent manner to a maximum of 90 ± 5% over basal values, coexpression of the i3L-peptide resulted in complete inhibition of UK-14,304-stimulated binding of [35S]GTPγS. Thus, the δ-i3L peptide inhibits agonist-mediated stimulation of [35S]GTPγS binding to Giα/Goα proteins interacting with opioid as well as other GPCRs interacting with them.
Effect of δ-i3L Peptide on Adenylyl Cyclase Inhibition by Opioid or Other GPCRs. Opioid receptors typically signal through coupling to Gi/Go proteins to inhibit adenylyl cyclase. To examine whether expression of the δ-i3L peptide selectively uncouples GPCRs from Giα/Goα proteins, we examined the ability of δ-i3L minigene to affect agonist-mediated adenylyl cyclase inhibition in δ-HEK293, μ-HEK293, and Rat-1 fibroblasts expressing the α2AR and in HEK293 cells, which also express endogenously the β2AR. In δ-HEK293 cells transfected with vector alone, 1 μM DSLET resulted in 40 ± 1.5% inhibition of cAMP accumulation, whereas this inhibition was abolished in cells expressing the minigene (Fig. 5A). Similarly, the expression of δ-i3L reduced dramatically DAMGO-mediated inhibition of forskolin-stimulated adenylyl cyclase activity (Fig. 5B).
To test whether the expression of the δ-i3L minigene could also interfere with other receptors interacting with the same population of G proteins as the opioid receptors, we measured the ability of this peptide to modify α2AR-mediated adenylyl cyclase inhibition by UK-14,304. Our results revealed that, although UK-14,304 inhibited forskolin-stimulated adenylyl cyclase by more than 80% at 1 μM UK-14,304, coexpression of the effective peptide δ-i3L essentially abolished this inhibition (Fig. 5C). In contrast, no alteration in adenylyl cyclase activity by δ-i3L was observed upon stimulation with isoproterenol in HEK293 cells that express endogenously the β2AR, which couples to Gsα (Fig. 5D). These results suggest that the expression of the third loop of the δ-OR can influence G protein signaling by blocking intracellularly a specific G protein pool (Gi/Go) that interacts with other receptors besides opioid receptors.
Effect of δ-i3L Expression on PLC Activation and Intracellular Ca2+ Release. Besides coupling to Gi/Go proteins, opioid receptors have been shown to couple to other G proteins, to stimulate Ins(1,4,5)P3 formation as well as Ca2+ release (Smart et al., 1994, 1997; Smart and Lambert, 1996; Connor et al., 1997; Law et al., 2000; Swevers et al., 2005). To examine the possible effects of δ-i3L expression on the interaction of various GPCRs with other G proteins, we examined its influence on PLC activation in HEK293 cells expressing the δ- or μ-ORs following agonist stimulation. As shown from the concentration-response curves for DSLET and DAMGO increases of 120 ± 5.2 and 100 ± 4.8% in Ins(1,4,5)P3 formation were obtained upon agonist activation of the δ- and μ-ORs, respectively (Fig. 6, A and B), suggesting that the presence of endogenous G proteins in HEK293 cells can effectively activate PLC. In contrast, expression of δ-i3L in the δ- and μ-HEK293 cells abolished opioid-agonist Ins(1,4,5)P3 accumulation essentially at all agonist concentrations. Moreover, expression of the i3L peptide in the same cells, which also express the M1-muscarinic receptor, had no inhibitory effect on carbachol-dependent stimulation of Ins(1,4,5)P3 formation (Fig. 6C), suggesting that the peptide does not interfere with the function of Gq-coupled receptors.
Previous work on neural SHSY-5Y, neuroblastoma NG108-15, and other cell lines has shown that OR activation leads to increases of Ca2+ mobilization from intracellular stores (Smart and Lambert, 1996; Connor et al., 1997; Swevers et al., 2005). To capitalize on these observations toward the development of a quick and reliable assay for signaling inhibitors, which abolishes downstream G protein signaling via the PLC/Ca2+ pathway upon GPCR activation, we measured the ability of δ-i3L to influence intracellular Ca2+ mobilization in δ- and μ-HEK293 cells following activation of the opioid receptors by the agonists DSLET and DAMGO, respectively. As shown in Fig. 7, A and B, significant dose-dependent increases in Ca2+ release were observed in both δ- and μ-HEK293 cells 5 s after the addition of agonists. In clear contrast, expression of δ-i3L abolished the relative increase in Ca2+ release mediated by both opioid agonist-activated receptors.
Effects of δ-i3L Expression on ERK Activation by the δ- and μ-Opioid and α2-Adrenergic Receptors. It has been previously demonstrated that δ- or μ-ORs expressed in HEK293 cells or Rat-1 fibroblasts stimulate ERK1/2 activity via PTX-sensitive or -insensitive G protein signaling mechanisms (Burt et al., 1996; Tso et al., 2000; Belcheva et al., 2002). To examine whether the δ-i3L peptide can be also used as potential inhibitor of this type of GPCR signaling, ERK activation was measured in response to δ- and μ-opioid agonist stimulation in serum-starved HEK293 cells expressing these receptors. As shown in Fig. 8, A and B, Western blotting with a specific phospho-ERK1/2 antibody revealed an increase in ERK1/2 phosphorylation by δ-OR and μ-OR within 2 min of the respective agonist stimulation. In contrast, ERK phosphorylation was abolished in the δ-HEK293 and μ-HEK293 cells that express the δ-i3L peptide, suggesting that the latter could block the δ-OR and μ-OR signaling to ERK1/2. Consistent with these findings were the results obtained upon PTX treatment of the cells, which abolished ERK phosphorylation (Fig. 8B, right). These data suggest that δ- and μ-ORs in HEK293 cells activate ERK1/2 via PTX-sensitive G proteins, which can be blocked by the presence of the minigene.
In view of these findings, we also administered the tyrosine kinase receptor ligand EGF (10 nM) in HEK293 cells, a treatment known to cause rapid activation of ERK kinases (Ho et al., 2005). EGF-mediated phosphorylation of ERK was indeed readily observed; in contrast to the situation with the opioid agonists, however, ERK phosphorylation was not affected by the presence of δ-i3L, suggesting that δ-i3L expression cannot influence ERK1/2 activation by a receptor that signals through mediators other than G proteins (Fig. 8C).
Last, because the α2AR couples to Gi/Go family proteins and mediates ERK activation through a PTX-sensitive pathway (Anderson and Milligan, 1994), we examined whether δ-i3L expression could also block downstream signaling of ERK phosphorylation by this receptor upon agonist stimulation of Rat-1 fibroblasts. As shown, in Fig. 8D, 2 min of exposure of Rat-1 cells to UK-14,304 were sufficient to induce robust phosphorylation of ERK1/2 kinases. The induction of ERK phosphorylation was unaffected by the δ-i3L expression, suggesting that the mechanism of ERK phosphorylation in Rat-1 fibroblasts by the α2-AR in HEK293 cells involves a different pathway than that of the δ- and μ-ORs.
Discussion
The precise structural determinants underlying the activation of heterotrimeric G proteins by heptahelical receptors are not yet fully understood and could be explained by the notion that G protein selectivity is defined by the overall conformation of the intracellular regions of the receptors. It is thought that there are multiple sites of contact between an activated receptor and G protein(s), but the structural determinants of the opioid receptors involved in controlling the fidelity of interactions with specific G proteins and/or other signaling molecules is not yet fully defined. Previous observations from our laboratory using peptides representing specific intracellular domains of the δ- and μ-ORs have shown that the i3L and part of the C-terminal tail are responsible for G protein coupling and play a distinct role in mediating protein-protein interactions by recruiting cytoplasmic proteins at specific modular domains located in them (Merkouris et al., 1996; Georgoussi et al., 1997, 2005). Although both the C-terminal tail and the i3 loop have been implicated in receptor coupling with G proteins and other signaling molecules, either one alone could be sufficient for recognition and productive G protein coupling (Chan et al., 1995; Georgoussi et al., 1997; Mazarakou and Georgoussi, 2005).
As part of an effort to define potential activators or inhibitors of G protein signaling, the i3L of the δ-OR was engineered in an expression vector and its ability to modify G proteins or other signaling components for the opioid or other GPCR signaling was investigated. Our attempts were specifically focused in the third loop of δ-OR, because this region was implicated in direct G protein binding and activation (Chan et al., 1995; Georgoussi et al., 1997), arrestin binding (DeGraff et al., 2002), and ligand-selective activation (Chaipatikul et al., 2003). We found that cellular expression of a 23-amino acid polypeptide derived from the i3L of δ-OR was able to target the opioid receptor as assessed by coprecipitation studies. Another finding was the demonstration of the direct interaction between the i3L portion of δ-OR with Goα but not Gsα after immunoprecipitation using selective G protein antisera in intact cells. This observation is in accordance with previous findings from our laboratory using specific G protein antibodies demonstrating that both δ- and μ-opioid receptors form stable complexes with G proteins in the absence of agonists (Georgoussi et al., 1993). The fact that the i3L peptide recognizes and binds the receptor and at the same time interacts directly with Goα suggests that the expressed intracellular portion of the receptor forms an assembly that interferes with opioid receptor signaling independent of receptor stimulation. Recruitment of the δ-i3L to G proteins blocks receptor-Gi/Go protein coupling as demonstrated by the ability of the expressed peptide to modify high-affinity opioid agonist binding in competition experiments using the opioid antagonist [3H]diprenorphine. When [3H]DADLE and [3H]DAMGO were used to characterize binding to δ- and μ-receptors in saturation experiments in membranes from HEK293 cells stably transfected to express these receptors, the limited range of labeled agonist concentrations that could be used for the estimation of specific binding prohibited detection or characterization of sites that had low affinity for these peptides in the presence of minigene. Both δ- and μ-opioid receptors demonstrated agonist-stimulated [35S]GTPγS in membranes, indicating that they are coupled to functional G proteins present in HEK293 cells. This agonist-mediated [35S]GTPγS binding to the G proteins was abolished in membranes expressing this peptide not only for the δ- and μ-opioid but also for the α2AR receptor, suggesting that the δ-i3L peptide interferes with a specific Gi/Go protein pool that couples with opioid receptors as well as with other GPCRs.
These findings were further supported by studies measuring alterations in the generation of second messengers following expression of δ-i3L peptide in intact cells. Indeed, expression of the peptide attenuates adenylyl cyclase inhibition of activated δ- and μ-opioid and also α2AR receptors, suggesting that the δ-i3L minigene specifically blocks receptor-Gi/Go protein coupling for different GPCRs. In addition, opioid-mediated activation of Ins(1,4,5)P3 accumulation and Ca2+ release for both δ- and μ-ORs was blocked in the presence of the expressible peptide. These data suggest that discrete amino acid sequences within a single i3 intracellular portion of the receptor might play a major role in determining the selectivity of receptor Gi/Go protein interaction. The finding that coexpression of the δ-i3L did not affect the signal transduction pathway mediated by the activated β2-adrenergic or M1-muscarinic receptors, respectively, suggests that the expressible peptide targets to a specific Gi/Go protein pool rather than to Gs or Gq. These observations can be interpreted as evidence for the specificity in the inhibitory effect of the δ-i3L domain for receptor coupling to a certain G protein pool. The inhibition of a signaling pathway mediated by a nonhomologous receptor domain may also reflect competition between the intracellular domains of different receptors and the δ-i3L for the same region of the Gi/Go α subunit. Cross-reactivity between i3L of the δ for the μ- and α2AR receptors occurs at the G protein level, because sequence homology within these receptors cannot account for the effect of the peptide to impair receptor-G protein coupling and contribute to the observed blockage of adenylyl cyclase activity. In the present study, we also demonstrate that the i3 loop peptide plays a critical role in MAPK phosphorylation of the activated opioid receptors. Recent observations encompassing fusion peptides corresponding to the third intracellular loop and the C-terminal tails of the δ- and μ-ORs have shown the direct binding of Gβγ complexes within these domains of the opioid receptors (Georgoussi et al., 2005). Therefore, a plausible scenario could be that the minigene can also disrupt Gβγ subunit association with the receptor a hypothesis that merits further investigation. However, although the presence of the minigene δ-i3L disrupts G protein signaling of the activated α2AR, it does not block UK-14,304-mediated ERK1/2 phosphorylation. This could be attributed to the fact that mitogen-activated protein phosphorylation mediated by the activated α2AR in the various cell lines tested is regulated via a different mechanism involving other signaling components and not merely by using the same mechanism as that of the opioid receptors (Anderson and Milligan 1994; Belcheva et al., 2002; Feldman et al., 2002).
In accordance with our observations, minigene constructs encoding various intracellular domains of heptahelical receptors have been used to examine the coupling mechanisms of GPCRs toward the development of inhibitors of G protein activation (Luttrell et al., 1993; Hawes et al., 1994; Thompson et al., 1998; Ulloa-Aguirre et al., 1998). It was shown that expression of the i3L of the α1BAR with its homologous receptor resulted in specific antagonist of receptor-mediated signaling (Luttrell et al., 1993), whereas expression of the i2 or i3 loops of the AT1-angiotensin receptor inhibited angiotensin-dependent activation of PLC in HEK293 cells by 40 and 27%, respectively (Thompson et al., 1998). Moreover, heterologous expression of the i3 loops of the α1AR and the M1-muscarinic receptors resulted in (25–55%) inhibition of gonadotropin releasing hormone-evoked inositol phosphate turnover (Ulloa-Aguirre et al., 1998). In addition, although inositol phosphate production mediated by the α2AR and M2-muscarinic receptors was sensitive to coexpression of their homologous i3 loops, the heterologous expression of these domains has little effect (0–20%). In most of these cases, the extent of inhibition observed was lower than that produced in our studies. This can be attributed to the relatively small length and the particular amino acid composition of our peptide. Indeed, the inner i3L of δ-OR contains two consecutive arginine residues part of the BBXXB motif (where B is any basic residue and X is a nonbasic residue) close to the carboxyl-terminal region and a BXXB motif in its N-terminal region, which is known to be an interactive domain of G proteins (Law et al., 2000; Waldhoer et al., 2004). These critical domains may favor a structure and adapt such a conformation that block G protein activation of Gi/Go-coupled receptors.
Ample evidence also exists describing the development and the mechanism of action for a number of other constructs encoding the carboxy-terminal domains of various G proteins, such as the Giα, Gsα, Gqα, and the G13,14α. These studies have provided evidence for the specificity and the ability of these G protein-derived peptides to block competitively receptor-mediated activation of signaling pathways in an analogous manner to that described for receptor expressible peptides (Gilchrist et al., 1999; Feldman et al., 2002; Vanhauwe et al., 2002).
Our data suggest that the δ-i3L domain peptide blocks receptor-mediated signal transduction by binding to the G protein in a manner analogous to that of the receptor, thereby disrupting receptor-Gi/Go interaction. Indeed, we found that the δ-i3L peptide 1) directly interacts with the G protein(s) but does not have the ability to directly activate the G protein population; 2) competes with the homologous third loop of the intact receptor for G protein binding; and 3) does not behave as partial agonist since it has no effect on the basal levels of adenylyl cyclase or PLC. Collectively, our observations suggest that inhibition resulting from expression of an intracellular domain of GPCR can be used to interfere in receptor-mediated signaling events at the level of the receptor-G protein interface rather than at the level of ligand-receptor binding and can provide a useful tool for determining the contribution of a given G protein population to signaling by a receptor that couples to multiple G proteins. Moreover, the use of this peptide might represent a new approach for studying potential coupling specificity of Gi/Go with the family of regulator of G protein signaling proteins that have been found to interact with G proteins and the receptors (Roy et al., 2003; Georgoussi et al., 2005). In addition, taken into account that both μ- and δ-opioid receptors oligomerize with themselves or with other closely related GPCRs, including those of the β2AR (Ramsay et al., 2002) and the α2AR (Jordan et al., 2003) this δi3L minigene can also be a valuable probe for identifying specific G protein interactions mediated by these hetero-oligomeric receptors in living cells. Finally, since expression of the minigene blocks intracellular Ca2+ mobilization, it can be used to silence responses of protein kinase C-mediated receptor desensitization.
Acknowledgments
We thank Prof. K. Iatrou (Institute of Biology, National Center for Scientific Research “Demokritos”) for critically reading this manuscript and valuable comments. We also thank Prof. G. Milligan (University of Glasgow) for providing the antibody for the Goα and Prof. R. Schulz (University of Munich, Munich, Germany) for providing the δ-HEK293 cell line that stably expresses δ-OR.
Footnotes
-
This work was supported by grants from the Hellenic General Secretariat of Science and Technology, Greek Ministry of Development (97EKBAN2-112, EPETII) and the pharmaceutical company ELPEN Pharmaceutical Co. Inc. (Attica, Greece).
-
doi:10.1124/jpet.105.089946.
-
ABBREVIATIONS: δ-OR, δ-opioid receptor; μ-OR, μ-opioid receptor; PLC, phospholipase C; MAPK, mitogen-activated kinase; PTX, Bordetella pertussis toxin; GPCR, G protein-coupled receptor; i3L, third intracellular loop; α2AR, α2-adrenergic receptor; β2AR, β2-adrenergic receptor; DADLE, [d-Ala2,d-Leu5]-enkephalin; DAMGO, [d-Ala2,N-MePhe4,Gly-ol5]-enkephalin; GppNHp, guanosine 5′-[β,γ-imido]triphosphate; GST, glutathione S-transferase; PCR, polymerase chain reaction; HEK, human embryonic kidney; RT-PCR, reverse transcription-polymerase chain reaction; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; DSLET, [d-Ser2,Leu5]-enkephalin-Thr; GTPγS, [35S]guanosine 5′-O-(3-thio)triphosphate; Ins(1,4,5)P3, inositol 1,4,5-trisphosphate; EGF, epidermal growth factor; ERK, extracellular signal-regulated kinase; UK-14,304, 5-bromo-N-(4,5-dihydro-1H-imidazol-2-yl)-6-quinoxalinamine.
- Received May 23, 2005.
- Accepted September 8, 2005.
- The American Society for Pharmacology and Experimental Therapeutics