- Email: [email protected]
Homodimerization and internalization of galanin type 1 receptor in living CHO cells
Neuropeptides Neuropeptides 39 (2005) 535–546 www.elsevier.com/locate/npep
Homodimerization and internalization of galanin type 1 receptor in living CHO cells Sebastian A. Wirz a, Christopher N. Davis a,1, Xiaoying Lu a, Tomasz Zal b,2, Tamas Bartfai a,* a
Department of Neuropharmacology, Harold L. Dorris Neurological Research Center, The Scripps Research Institute, 10550 North Torrey Pines Road, SR 307, La Jolla, CA 92037, United States b Department of Immunology, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, United States Received 6 July 2005; accepted 2 September 2005 Available online 20 October 2005
Abstract Galanin is a 29- to 30-aa-long neuropeptide affecting feeding, cognitive, and sexual behavior. It exerts its effects through galanin receptors 1, 2 and 3, which are all seven transmembrane domain G-protein coupled receptors (GPCRs). The GPCRs have been shown to function as monomers, homodimers, heterodimers and oligomers. In this study, we examined the extent of galanin receptor 1 (GalR1) dimerization and internalization in living CHO cells using fluorescence resonance energy transfer (FRET) and time lapse confocal imaging. Ratio imaging analysis and emission spectral analysis revealed substantial homodimerization of GalR1. In addition, internalization of GalR1 after 1 h of agonist stimulation with the GalR1 agonist galanin (1–29) was observed with time lapse fluorescence imaging, whereas stimulation with the GalR2 specific agonist galanin (2–11) did not lead to internalization. Treatment of GalR1 transfected cells with the non-selective adenylyl cyclase activator forskolin influenced the rate of internalization when administered together with galanin (1–29). These results indicate that GalR1 can act as a dimer on the cell surface and that receptor desensitization and internalization was observed after stimulation with the agonist galanin (1–29). Western blots further confirm the FRET data that GalR1-XFP dimerizes and can be detected in the cell as a monomer or dimer using antibodies to XFP. Internalization and dimerization of GalR1 is shown, contributing to the regulation of galanergic signaling. 2005 Elsevier Ltd. All rights reserved. Keywords: Galanin; GPCR; Homodimerization; FRET
1. Introduction Abbreviations: GalR1, galanin receptor 1; GalR2, galanin receptor 2; FRET, fluorescence resonance energy transfer; CFP, cyan fluorescent protein; YFP, yellow fluorescent protein; GPCR, G-protein coupled receptor; CHO cells, Chinese hamster ovarian cells; PBS, phosphate-buffered saline. * Corresponding author. Tel.: +1 858 784 8404; fax: +1 858 784 9099. E-mail addresses: [email protected] (C.N. Davis), tbartfai@ scripps.edu (T. Bartfai). 1 Tel.: +1 858 784 9442. 2 Currently at M.D. Anderson Cancer Center, Department of Immunology, South Campus Bldg Unit 901, 7455 Fannin, Houston, TX 77030, United States. 0143-4179/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.npep.2005.09.001
Galanin has been shown to be important in endocrine, pain, mood and seizure regulation (Crawley, 1995; Vrontakis, 2002). Galanin exerts its effects at three galanin receptors (GalR1-3), each belonging to the superfamily of GPCRs. Galanin type 1 receptor (GalR1) mediates the effects of galanin in regulation of insulin release (Ahren et al., 1986; Kwok et al., 1988), growth hormone release (Bauer et al., 1986), pain threshold (Liu et al., 2001; Holmes et al., 2003), cognitive performance (Wrenn et al., 2004) and seizure threshold (Mazarati et al., 1998). The local concentrations of the endogenous
536
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
agonist galanin can vary broadly, e.g., it is highly inducible upon nerve injury (up to 120-fold upregulation) (Villar et al., 1989) or it is rapidly depleted by seizure activity (Mazarati et al., 1998). Oligomerization and internalization of the galanin receptors may influence receptor–agonist binding and subsequent downstream signaling and may be involved in coping with the large variation in ligand concentrations. Several seven transmembrane domain receptors that couple to the trimeric G proteins have been shown to undergo oligomerization. The past five years have provided multiple examples of GPCR homodimerization (Bouvier, 2001; Dean et al., 2001; Javitch, 2004; Milligan, 2004) among monoamine receptors such as the 5HT2C receptor (Herrick-Davis et al., 2004), a1 adrenoreceptors (Uberti et al., 2003), D2 dopamine receptor (Lee et al., 2000) and metabotropic glutamate receptors mGluR1 (Robbins et al., 1999). Neuropeptide receptor dimerization has been shown with the d opioid receptor (Cvejic and Devi, 1997), neuropeptide Y1, Y2, and Y5 receptors (Dinger et al., 2003), melanocortin receptors MC1R and MC3R (Mandrika et al., 2005). Furthermore, dimerization of CB1 cannabinoid receptor (Wager-Miller et al., 2002), the growth factor receptor Frizzled Xfz3 (Carron et al., 2003) and the cytokine receptor CXCR2 (Trettel et al., 2003) was also shown. The existence of heterodimers of GPCRs has also been demonstrated between the d and j opioid receptors (Jordan and Devi, 1999), d and l opioid receptors (Gomes et al., 2000), type A and B cholecystokinin receptors (Cheng et al., 2003), a 1D and 1B adrenergic receptors (Hague et al., 2004), b 2 adrenergic with d and j (Jordan et al., 2001) and MCR1 and MCR3 melanocortin receptors (Mandrika et al., 2005). The existence of GPCR dimers has been shown by fluorescence energy transfer (FRET) of tagged receptors (Dinger et al., 2003), by bioluminescence energy transfer and acceptor photobleaching (Herrick-Davis et al., 2004), by immunoprecipitation and Western blot analysis of the receptors (Cvejic and Devi, 1997). The structural basis of the dimerization of GPCRs is slowly emerging. For example, the long intracellular loop of the muscarinic M3 receptor has been implicated in dimerization (Maggio et al., 1993) and the large extracellular N-terminal domain of mGluR1 is involved in dimerization, respectively (Kniazeff et al., 2004). Interactions between transmembrane domains of GPCRs including the first, third and fourth transmembrane domains have been suggested to participate in the formation of dimeric GPCRs (Carrillo et al., 2004). While most reported cases of dimerization involve noncovalent binding of monomers, others such as the mGluR1 dimer is formed with a disulfide bond between the cysteine residues of each monomer (Ray and Hauschild, 2000), or in the case of the frizzled receptor Xfz3, dimer disulfide bonds are formed between the intracellular domains of the monomers (Carron et al., 2003).
GPCR homo- and heterodimerization results in altered agonist binding, rate of internalization, receptor ligand pharmacology and G protein mediated signaling (Ferguson, 2001). The most striking example so far is in the case of Interleukin-8 (IL-8) binding to CXCR1 that requires the CXCR1 dimer to dissociate, thus suggesting that the dimerization equilibrium alters agonist binding, leading to changes in receptor signaling (Fernando et al., 2004). In the case of the d opiate receptor dimer, dissociation to monomers precedes agonist binding and internalization of the receptor/agonist complex (Cvejic and Devi, 1997). Frizzled Xfz3 dimerization is sufficient to start the Wnt/b-catenin signaling (Carron et al., 2003), while agonist binding or G-a protein binding had no effect on the state of oligomerization of the NPY 1 receptors (Dinger et al., 2003). The discovery of homo- and heterodimerization of GPCRs provides an interesting explanation of differences in GPCR signaling upon ligand binding. Different agonists to the same GPCR can differ greatly in their ability to stabilize oligomers of the receptor: morphine binding to the d opiate receptor promotes the dissociation of dimers leading to reduced internalization (Cvejic and Devi, 1997), however, other d agonists promote dimerization resulting in accelerated internalization of the receptor dimer. Original studies indicate that two spectral variants of GFP, the cyan fluorescent protein (CFP) and the yellow fluorescent protein (YFP), show the most promising spectral properties for FRET (Tsien, 1998). Here, we report data from combined FRET and Western blot studies on the dimerization and internalization of the GalR1 receptors tagged with either CFP or YFP and heterologously expressed in CHO cells.
2. Materials and methods 2.1. Vector construction C-terminally tagged fusion protein constructs were generated by cloning full length murine coding sequence of galanin receptor 1 into pECFP-N1 and pEYFP-N1 vectors (Clontech Laboratories, Palo Alto, CA, USA). GalR1 cDNA was amplified using Platinum Taq High Fidelity Polymerase (Invitrogen, Carlsbad, CA, USA) with the following primers: GalR1 forward primer: 5 0 GAGCTCAAGCTTATGGAACTGGCTATGGTGAACCT – 3 0 which contained a HindIII site; GalR1 reverse primer: 5 0 – ACCGTCGACTTCACGTGGGTGCAGTTGGT – 3 0 which contained a SalI site and replaced the stop codon. PCR products were digested and cloned into pECFP-N1 and pEYFP-N1 vectors. The constructs were sequenced (The Scripps Research Sequencing Core Facility, La Jolla, CA, USA) to ensure that no mutations had occurred during the PCR and cloning process. As a positive control for FRET between CFP
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
537
and YFP, a CFP–YFP fusion protein linked via 18 amino acids was used (kindly provided by Dr. Peter Vanderklish, The Scripps Research Institute, La Jolla, CA, USA).
using Prism software (GraphPad, San Diego, CA, USA).
2.2. Cell culture and transfection
Membranes were isolated 48 h after transfection from transiently transfected CHO cells expressing CFP or YFP tagged GalR1 as previously described (McIlhinney, 2004). Membranes were resuspended in 10 mM NaHCO3 containing protease inhibitors 16 lg/ml benzamidine HCl, 10 lg/ml phenanthroline, 10 lg/ml aprotinin, 10 lg/ml leupeptin, 10 lg/ml pepstatin A, 1 mM PMSF (protease inhibitor cocktail, BD Biosciences, Palo Alto, CA, USA) and incubated at 60 C for 15 min in the absence or presence of 20 mM DTT before electrophoresis. To determine the extent of monomer and dimer formation in the cell, equal amounts of mock transfected CHO cell and chimeric receptor protein (20 lg) were separated by electrophoresis in a 7% SDS–PAGE gel in the presence of sample buffer, electrophoretically transferred to nitrocellulose membrane and probed with anti-XFP Living Colors antibody (BD Biosciences, Palo Alto, CA, USA).
Chinese hamster ovary (CHO) cells were cultured in DulbeccoÕs modified EagleÕs medium supplemented with 10% (v/v) fetal bovine serum (FBS), 4 mM L-glutamine, 2 mg/ml L-proline, 100 U/ml penicillin/streptomycin (Gibco, Invitrogen Corporation, Carlsbad, CA, USA) at 37 C and in an atmosphere of 5% CO2. Cells were grown in 35 · 10 mm culture dishes and transiently transfected with 4 lL lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) at 80% confluency with 1 lg of either GalR1-CFP, GalR1-YFP or 1 lg of CFP–YFP construct (positive control). For cotransfection experiments, 500 ng GalR1-CFP and 500 ng GalR1-YFP were used. 2.3. Membrane preparation and radioligand binding assay Forty eight hours after transfection, cells were washed three times with cold phosphate-buffered saline (PBS), scraped off the plate and centrifuged at 1000g for 5 min at 4 C. They were then resuspended in hypotonic buffer containing 20 mM Hepes, 5 mM MgCl2 and 1 mM EDTA, and incubated on ice for 10 min. After a final spin at 13,000g for 20 min, the pellet was resupended in 50 mm Tris–HCl, pH 7.4. Ligand binding of [125I] porcine galanin (2200 Ci/ mmol, Perkin–Elmer Life Science, Boston, MA, USA) to CHO cells was performed in 150 lL binding buffer [50 mM Tris–Cl (pH 7.4), 5 mM MgCl2, 0.05% (w/v) bovine serum albumin, supplemented with peptidase and protease inhibitors: 50 lM leupeptin, 100 lM phenylmethanesulfonyl fluoride (PMSF) and 2 lg/mL aprotinin]. Incubations were carried out at room temperature for 45 min and were terminated by rapid vacuum filtration through glass-fiber filters (Packard, Meriden, CT, USA). The filters were washed three times with ice cold PBS (pH 7.4) containing 0.01% Triton X-100 and counted with Cobra II auto-Gamma counting systems (Packard Bioscience, Downer Grove, IL, USA). In all binding experiments, [125I] galanin concentration was 0.2 nM and unlabeled galanin (1–29) (Vulpes Ltd., Tallinn, Estonia) was used at 1 · 105– 1 · 1011 M to displace [125I] galanin binding. Each galanin concentration was performed in triplicate and data were normalized by protein content in the samples (20–30 lg protein in each assay), measured using the BCA protein assay kit from Pierce (Rockford, IL, USA). Non-specific binding of [125I] galanin was determined in the presence of 1 · 105 M galanin (2–11). Binding data were analyzed by nonlinear curve fitting
2.4. Western blot analysis of receptors
2.5. Fluorescence microscopy study of receptor internalization For live cell imaging, an Olympus Fluoview 500 laserscanning confocal microscope was used. Fluorescence images were collected by illumination using the 442 helium cadmium laser and the 488 line of an argon laser. Images were collected on a Delta Vision Optical Sectioning Microscope consisting of an Olympus IX-70 microscope equipped with a mercury arc lamp. A photometrics CH 350 cooled CCD camera and a high precision motorized XYZ stage were used to acquire images of transfected cells at timepoints 0, 60 and 120 min after addition of 1 lM galanin (1–29), 1 lM galanin (2–11), 1 lM M40, or 10 lM Forskolin [in the absence or presence of 1 lM galanin (1–29)] with a 60· immersion objective. The settings of the confocal were maintained constant for each series of experiments so that the resulting images could be analyzed by densitometry and the time dependent changes of internalized receptor could be compared. 2.6. FRET analysis of homodimerization For a detailed description of the FRET analysis system, please refer to Zal and Gascoigne (2004). Briefly, the dual-camera 3I system was controlled by the Slidebook 4.0 software (3I Corporation, Denver, CO, USA) to simultaneously capture CFP emission and YFP emission by two CoolSnap HQ cameras (Roper, Tucson, AZ, USA) attached to a Zeiss 200-M microscope through a beamsplitter (custom 510LPXR, Chroma,
538
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
Rockingham, VT, USA) and stationary bandpass filters. The optical filters were 430/25, 470/30 (CFP excitation and emission), 510/20, and 550/50 nm (YFP excitation and emission), and the JP4 dichroic mirror (Chroma, Rockingham, VT, USA). Rapid wavelength switching between CFP and YFP excitation was performed with a DG4 galvo illuminator customized with a 300 W xenon lamp (Sutter, Novato, CA, USA). YFP excitation was attenuated to 20% by appropriate positioning of the exit galvo. Cameras were typically run in 2 · 2 binning mode with software flat-field correction. The same exposure time was used for all light channels, and background was subtracted in each image based on the average reading in an area of the image devoid of cells. Images were aligned with subpixel resolution using the frequency-based function of the Slidebook software. Sensitized-emission live FRET imaging was performed on the widefield microscope and the sensitized fluorescence of acceptor was detected through an optical FRET filter set selecting acceptor emission during donor excitation (IDA image). The fluorescence of acceptor and the tail of the donor emission spectrum contaminate the IDA image. To account for this bleed-through and to render the FRET index independent of fluorescence intensity, two additional images are acquired: acceptor fluorescence during acceptor excitation (IAA) and donor fluorescence during donor excitation (IDD). Given that the crosstalk coefficients of acceptor and donor (a and d, respectively) in the FRET filter set are constant and assuming that no other crosstalk components are present, sensitized emission, Fc, can be calculated by linear unmixing of the IDA intensity leading to the equation, Fc = IDA dIDD aIAA (Zal and Gascoigne, 2004). Fc is dependent on fluorophore concentration, which leads to the determination of a ratiometric FRET indicator. FRET efficiency was calculated as E = R/(R + G), where the FRET ratio is R = (IDA dIDD aIAA)/ IDD. The values IDA, IDD, and IAA are pixel intensities in the images acquired using donor excitation acceptor emission, donor excitation donor emission, and acceptor excitation acceptor emission, respectively. The bleed-through from YFP during excitation through the 430/25 filter and imaging through the 550/50 filter was calibrated using cells expressing YFP only and was a = 8.6% of the intensity using the YFP excitation and emission filters. The bleed-through of CFP in the 550/50 filter during excitation through the 430/25 filter was calibrated using CFP-only expressing cells and was d = 63% of the intensity in the CFP excitation and emission filter set. The G-factor constant that relates sensitized emission to donor quenching was calibrated at G=3.5 for whole cells on the dual-camera 3I system using the CFP-YFP fusion protein and a photobleaching approach as previously described (Zal and Gascoigne, 2004).
2.7. Statistical analysis Statistical analyses were performed using InStat from GraphPad Software Inc. (San Diego, CA, USA). To determine statistical significance, analysis of variance was used followed by StudentÕs t-test (Figs. 4 and 5). The level of statistically significant difference was defined as p > 0.05.
3. Results 3.1. Binding of ligands to CFP- and YFP- labeled GalR1 To determine if C-terminal CFP and YFP tags alter the ligand binding properties of GalR1, equilibrium binding assays were performed on membranes prepared from CHO cells transiently transfected with GalR1-CFP and GalR1-YFP. Galanin (1–29), the endogenous ligand for galanin receptors, displaced 125I galanin with IC50 values (means ± SEM) of 1.14 ± 0.84 nM for the wild type receptor, 3.08 ± 1.62 nM for GalR1-CFP and 9.48 ± 1.52 for GalR1-YFP. Both GalR1-CFP and GalR1-YFP exhibited high affinity binding to galanin (1–29) with IC50 values that were comparable to the non-tagged receptor (Fig. 2), suggesting that the C-terminal CFP and YFP tags did not impair the ability of galanin (1–29) to bind of GalR1. 3.2. Internalization of GalR1-CFP and GalR1-YFP following agonist stimulation On a subcellular level, confocal microscopy images revealed transfected cells with strong fluorescent signals on the cell surface (Fig. 3A and D). In the presence of galanin (1–29), dose dependent internalization of GalR1-CFP and GalR1-YFP was detected by a redistribution of fluorescence signal intensity from the cell surface to the perinuclear compartment of the cell (Fig. 4). Internalization began as early as 15 min after ligand addition and continued at 120 min. Incubation of transfected cells with galanin (1–29) at concentrations higher than 500 nM led to a fivefold increase in intracellular fluorescent signal intensity and to significant drop in cell membrane fluorescent signal intensity (Fig. 4). Addition of the non-selective adenylyl cyclase activator forskolin had no effect on endocytosis of the XFPtagged GalR1. However, the combination of galanin (1–29) (1 lM) and forskolin (10 lM) led to a slower internalization than observed with galanin (1–29) alone (Fig. 5). 3.3. GalR1 homodimerization CHO cells were transiently transfected with independent FRET pairs: as a positive control, the CFP/YFP
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
pair cloned into one construct and linked via 18 amino acids, as a negative control, two constructs containing only CFP or YFP were cotransfected. FRET efficiency (E) was defined as the proportion of the excited states of the donor that becomes transferred to the acceptor. The positive control showed E = 28%, whereas the negative control was barely above detection background (E = 0–3%). GalR1 FRET efficiency was on average 14%, as measured in 10 different cells from individual experiments (Fig. 6). Fig. 7A–D show the different channels captured: CFP, YFP, merge and FRET. FRET images are shown in pseudocolor, red indicating high levels of FRET efficiency and blue indicating little to no FRET. The positive control produced high levels of FRET efficiency, making this a useful tool to compare maximal FRET signal with the FRET signal obtained in stimulated cells (Fig. 7D). As a negative control, two separate constructs for CFP and YFP were transfected into CHO cells. These fluorescent proteins are indicated as blue pseudocolor suggesting that there was no physical interaction within the cell (Fig. 7E–H). Unstimulated cells expressing GalR1 fusion proteins are indicated by red pseudocolor predominantly on the cell surface level, suggesting close proximity of these receptors as it is commonly seen with oligomerizing receptors (Fig. 8). Addition of the agonist Galanin (1–29) did not cause any increase in intracellular FRET intensity, suggesting that upon agonist stimulation, the GalR1homodimer dissociates and internalizes as a monomer (monomer/agonist complex). Fig. 9 shows the Emission Spectral Analysis of the FRET. The single transfected GalR1-CFP (CFP) and double transfected GalR1CFP/GalR1-YFP (FRET) CHO cells are shown and the increase in fluorescence intensity at 527 nm corresponds to energy transfer between the CFP-YFP fluorophores. The peak in the spectrum at 513 nm is due to Raman emission (not corrected). 3.4. Western blot analysis of GalR1-CFP and of GalR1-YFP Transiently transfected CHO cells were analyzed to determine the extent of receptor monomer and dimer formation in the living cell. Extracts were prepared to
539
isolate chimeric receptors expressed in the cellular membranes. Since the antibodies to GalR1 are not selective, we made use of an antibody directed towards the XFP fluorescent protein to detect the GalR1 receptor. Fig. 10 depicts the CHO cell-expressed CFP and YFP tagged GalR1 receptors as monomers and dimers in presence and absence of the reducing agent Dithiothreitol (20 mM). Receptor monomers and dimers migrated at molecular masses of approximately 60 and 120 kD, respectively. These molecular masses are consistent with the predicted mass of the tagged receptor. CHO cells transiently expressed with either the CFP or YFP alone indicated that the fluorescent proteins migrated at 27 kD and did not oligomerize in the cell (data not shown).
4. Discussion A number of studies have shown homo- and/or heterodimerization of neuropeptide receptors similar to these data. Neuropeptide Y receptors have been shown to homodimerize in a subtype specific manner (Dinger et al., 2003), whereas the opioid receptor subtypes can form both homo- and heterodimers (Cvejic and Devi, 1997; Jordan and Devi, 1999; Gomes et al., 2000). These oligomerizations are important regulatory mechanisms for receptor function, which influence agonist affinity, potency, and efficacy (Bouvier, 2001). Techniques used so far for dimerization studies have utilized both a biochemical and biophysical approaches including co-immunoprecipitation, and Western blot analysis, bioluminescence resonance energy transfer (BRET) and fluorescence resonance energy transfer (FRET). We tested FRET in two different systems, widefield microscopy and fluorescence spectroscopy and confirmed the homodimerization of GalR1 (GalR1-XFP) with Western blots. This study was conducted with C-terminally tagged GalR1 receptors carrying CFP or YFP (Fig. 1) to assess homodimerization and internalization of GalR1 receptors upon agonist binding. The C-terminally tagged GalR1 has retained high affinity for the endogenous agonist galanin (1–29) (Fig. 2). This is expected as the
Fig. 1. Schematic representation of FRET between fluorescently labeled GalR1. If the distance between the donor CFP and acceptor YFP is less than 10 nm, energy transfer occurs (A). Otherwise, no FRET occurs (B).
540
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
Fig. 2. Binding properties of CHO cells transiently expressing GalR1XFP fusion proteins. The two tagged receptors, GalR1-CFP and GalR1-YFP, have similar affinity to the endogenous ligand Galanin (1–29) as the wildtype receptor. IC50 values were as follows: 1.14 ± 0.84 nM for the wild type receptor, 3.08 ± 1.62 nM for GalR1-CFP and 9.48 ± 1.52 for GalR1-YFP (IC50 values are shown as means ± SEM from three independent experiments performed in triplicate).
binding of the 29 amino acids long neuropeptide has earlier been shown to involve the extracellular loops of the receptors and a few amino acid residues in the transmembrane domains 2 and 3 close to the extracellular surface (Kask et al., 1996), and no interactions with the intracellular loops or with the C-terminus that was extended with XFP. Nevertheless, allosteric interactions might have affected the ability of the C-terminally
tagged GalR1 in which case galanin binding and galanin induced effects on dimerization and internalization could be affected. This does not appear to be the case. Confocal imaging of tagged GalR1 transfected into CHO cells revealed cell surface expression (Fig. 3A and D) and substantial galanin 1–29 induced internalization (Fig. 3B, C, E and F), similar to what was observed with fluorescently labeled agonist (Nfluorescein-galanin 1–29) on GalR1 transfected CHO cells (Wang et al., 1998a) and with the fluorescently labeled GalR2 receptors heterologously expressed in PC12 cells (Xia et al., 2004). On a confocal microscope, subcellular receptor expression was seen both on the cell surface as well as intracellularly, predominantly in the perinuclear region. Unlike the observations of tagged GalR2 expressed in PC12 cells (Xia et al., 2004), there was no constitutive internalization in the absence of ligand, as observed with timelapse fluorescent imaging. Similarly, the GalR2 selective ligand galanin (2–11) did not induce endocytosis of GalR1-CFP or of GalR1-YFP confirming the pharmacological specificity of the agonist action of GalR1-XFP. However, the addition of Galanin (1–29) at concentrations of 500 nM, 1 lM and 2 lM induced internalization. The internalization process was time dependent and at room temperature peaked at 60 min after ligand addition and was still ongoing at 120 min (Fig. 4). Addition of 10 lM forskolin alone, a non-selective adenylyl cyclase activator, had no effect on endocytosis of the tagged GalR1. The combination of galanin (1–29) (1 lM) and forskolin (10 lM), however, led to a slower internalization than observed with galanin alone (Fig. 5A and B). This
Fig. 3. The endogenous agonist galanin (1–29) causes internalization of GalR1-XFP in CHO cells. Confocal microscopy of cell surface expression of single transfected CHO cells indicates basal state plasma membrane expression of GalR1-CFP (A) and GalR1-YFP (D). Agonist induced internalization of the tagged galanin receptors are observed following the addition of 1 lM galanin (1–29). Timelapse fluorescent images were taken indicating internalization of the tagged GalR1 after 60 (B and E) and 120 min (C and F). Galanin (1–29) caused a time dependent depletion of cell surface GalR1-XFP and a corresponding accumulation intracellularly as determined in independent measurements.
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
541
Fig. 4. GalR1-XFP transfected CHO cells were subject to various concentrations of galanin (1–29). Fluorescence intensity was measured in five different regions of interest intracellularly and on the cell surface. At the cell surface level, all concentrations resulted in a significant drop of fluorescence intensity, although this is more due to a general drop in fluorescent intensity due to cell motility rather than substantial internalization of the receptor ligand complex (A). Intracellularly, addition of galanin (1–29) at concentrations of 500 nM and above showed a highly significant increase in fluorescence intensity (B). Statistically significant difference from controls are indicated as *, p < 0.05.
finding is consistent with results demonstrating that PKA activation by forskolin inhibited the agonistinduced internalization of mGluR1a and mGluR1b (Mundell et al., 2004). The serine at position 144 within the second intracellular loop has been identified as a potential PKA phosphorylation site (Wang et al., 1997). We assume that cAMP acts via PKA mediated phosphorylation of
GalR1 similarly to numerous GPCRs that are phosphorylated intracellularly (Gainetdinov et al., 2004). In addition to the PKA binding site, the intracellular regions of GalR1 contain potential phosphorylation sites for G protein-coupled receptor kinases (GRKs) and protein kinase C. GPCRs are desensitized and downregulated by the binding of b-arrestins to the phosphorylated tail of the receptor thereby interfering with G
542
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
Fig. 5. GalR1-XFP transfected CHO cells were subject to various ligand stimulations. Cells were treated with 1 lM galanin (1–29), 1 lM galanin (2– 11), 1 lM M40, 10 lM forskolin and a combination of 1 lM galanin (1–29) and 10 lM forskolin. Fluorescence intensity was measured in five different regions of interest on the cell surface and intracellular. Only galanin (1–29) showed a significant decrease in cell surface fluorescence intensity. Intracellularly, both galanin (1–29) and the combination of galanin (1–29) and forskolin showed a significant increase in fluorescent intensity. Forskolin slowed the rate of galanin (1–29) induced internalization. Statistically significant difference from controls are indicated as *, p < 0.05.
protein coupling and initiating rapid localization to clathrin coated pits for internalization. Alternative roles for b-arrestins have been identified in which they can act to mediate receptor signaling, by recruiting components of MAPK cascades into protein signaling complex. Further investigation is necessary to determine the extent of and consequences of GalR1 phosphorylation and subsequent interaction with these kinases and the b-arrestins.
Altered cAMP levels affect signaling via GalR1 in multiple ways: cAMP activation of CREB increases GalR1 synthesis (Zachariou et al., 2001), galanin binding to GalR1 inhibits adenylyl cyclase and depresses cAMP levels (Wang et al., 1998b). These mechanisms form a multiple level mutual control between galaninGalR1 and cellular cAMP. The negative coupling of GalR1 to cAMP has been shown to decrease neuronal firing, suggesting a compensatory role for this receptor.
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
It was shown that peripheral administration of the GalR1 antagonist galnon attenuates the increase in cAMP in the locus coeruleus due to morphine withdrawal (Zachariou et al., 2003). These results provide a mechanism by which a direct activation of the galanin receptor leads to changes in cAMP production and altered neuronal activity. Using ratio imaging and emission spectral analysis we provide for the first time evidence of a physical association between GalR1 receptors. The expression levels of GalR1-CFP and GalR1-YFP were comparable upon cotransfection of both GalR1-CFP and GalR1-YFP constructs. To assess whether FRET occurrence was due to actual dimerization and not due to random receptor clustering (random proximity effect), cells were transfected with increasing amounts of acceptor DNA and checked for differing FRET signals. Donor/acceptor ratios between 1:2 and 1:8 did not show a difference in FRET efficiency, indicating independence of acceptor expression level and that the signal was not caused by random crowding of receptors but rather due to specific interactions between the fluorescently tagged GalR1 receptors (data not shown). FRET studies were conducted on a widefield microscope system. GalR1 transfected CHO cells constitutively showed on average a 14% FRET efficiency on the cell surface, indicating a very significant extent of dimer formation between GalR1-CFP and GalR1-YFP (Fig. 6). When treated with the agonist galanin (1–29), no increase in FRET efficiency intracellularly was observed after 60 and 120 min. This would be expected if
543
Fig. 6. FRET efficiency (E) is defined as the proportion of the excited states of the donor that becomes transferred to the acceptor. The positive control showed E = 28%, whereas the negative control was barely above detection background. GalR1 FRET efficiency was on average 16%, as measured in 10 different cells from individual experiments.
the GalR1 would internalize as a dimer (Fig. 8). Rather, it appears that the GalR1 internalizes as a monomer upon agonist stimulation, similarly to what was previously observed with the d opioid receptor, where dissociation of the receptor dimer precedes the agonist-induced internalization of the receptor (Cvejic and Devi, 1997). Western blots of GalR1-CFP and GalR1-YFP were visualized with an antibody recognizing CFP and
Fig. 7. GalR1 homodimerization as measured in FRET efficiency on a widefield microscope. As a positive control for FRET, a CFP–YFP fusion protein linked via 18 amino acids was expressed in CHO cells (A–D). As a negative control, separate CFP and YFP plasmids were co-transfected into CHO cells (E–H). Pictures on the right are colored in pseudocolor, red meaning high levels of dimerization and blue meaning low levels to none. Transiently transfected CHO cells constitutively express the GalR1 as dimers on the cell surface. After addition of galanin (1–29), the receptor seems to internalize as a monomer and aggregates intracellularly in the perinuclear region. (For interpretation of the references in color in this figure legend, the reader is referred to the web version of this article.)
544
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
Fig. 8. Timelapse FRET imaging after addition of Galanin (1–29). CHO cells were co-transfected with GalR1-CFP and GalR1-YFP and treated with the agonist galanin (1–29). No increase in FRET efficiency intracellularly was observed after 60 and 120 min. This would be expected if the GalR1 would internalize as a dimer. Rather, it appears that the GalR1 internalizes as a monomer upon agonist stimulation. (For interpretation of the references in color in this figure legend, the reader is referred to the web version of this article.)
YFP, as the antibodies to GalR1 are not selective. The expected molecular weights for the monomers of GALR1-YFP or GALR-CFP are approximately 63 kD based on the primary sequence of the covalent complex of GalR1 and YFP or CFP. We have detected a band at approximately 60 and at 120 kD in both GalR1-CFP and in GalR1-YFP transfected cells in the presence of the disulfide reducing agent DTT (Fig. 10). These data confirm the FRET data that homodimers of GalR1-CFP and GalR1-YFP were responsible for the FRET shown in Figs. 7 and 8. The GalR1 receptor has three cysteine residues in transmembrane and intracellular domains, and these appear to be involved in the formation of the GalR1 dimer as receptor monomers are observed only in the presence of DTT, while a dimer of GalR1-XFP is observed at 120 kD under nonreducing conditions (see Figs. 9 and 10). Our data illustrate that GalR1 form homodimers to a large extent at the cell surface and to some extent already in the ER. The binding of agonist galanin (1–29) does not directly affect the dimer-monomer equilibrium as no changes in FRET at the cell surface were observed but it induces internalization of the GalR1 monomer, probably with the agonist bound to it, as suggested from the study with N-fluorescein-galanin
(1–29) bound to GalR1 (Wang et al., 1998a). Thus, the agonist is indirectly shifting the dimer–monomer equilibrium towards monomers which internalize once they bind galanin. Therefore, it is likely that GalR1 dimers formed in the endoplasmic reticulum or in the plasma membrane in the absence of agonist are in equilibrium with monomers and this equilibrium is shifted through dissociation and subsequent internalization of GalR1–galanin complexes. Our data are not useful to determine whether galanin binds to the GalR1 homodimer and that the galanin occupied dimer that dissociates or whether galanin binds to monomeric GalR1 only. The rate of internalization is affected by galanin acting on the extracellular side of the GalR1 and by elevation of cAMP levels intracellularly, by forskolin in the opposite direction. The role of GalR1 homodimers in galaninergic signaling requires further studies as the present data only address the existence and the regulation of dimers in the processes of agonist binding and internalization, but does not provide data on the pharmacology of receptor oligomerization. This is the case for most GPCR dimers as we have yet to find pharmacological agents that distinguish between binding to monomeric and oligomeric GPCRs or which signal differently through these different oligomeric forms of the receptor.
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
545
Acknowledgements The authors thank Dr. Kathy Spencer for valuable assistance in confocal microscopy, Dr. Peter Vanderklish for providing the CFP–YFP construct as a positive control, Liam Palmer for excellent technical assistance in fluorescence spectroscopy and Dr. Nicolas Gascoigne for use of the widefield microscope. This study was supported by National Institutes of Health Grant ROIMH63080 and ROI NS043409 (T.B.) and a stipend from the Swiss National Science Foundation (S.W.). This article is manuscript no. 17221-NP from The Scripps Research Institute.
References
Fig. 9. Emission spectral analysis of the FRET analysis. The single transfected GalR1-CFP (CFP) and double transfected GalR1-CFP/ GalR1-YFP (FRET) CHO cells are shown in the upper panel. The lower panel indicates the spectral analysis of the CFP/YFP pair positive control cloned into one construct and linked via 18 amino acids. An increase in fluorescence intensity at 527 nm is indicated by the arrow and corresponds to energy transfer between the fluorophores. The peak at 513 nm is due to Raman emission (not corrected).
Fig. 10. Dimerization of fluorescently tagged GalR1. Membranes were prepared from CHO cells transfected with CFP- and YFP-tagged GalR1 (GalR1CFP and GalR1YFP). Representative blots of two independent experiments are shown. Chimeric receptors as detected by anti-XFP antibody exist as a dimer in the cell in the absence of reducing agent. In the presence of 20 mM DTT, monomers of the chimeric receptors are detected. Monomer and dimer of receptors are indicated by a * and * *, respectively. Mock-transfected CHO cells (CHO) are indicated in the absence and presence of DTT.
Ahren, B., Arkhammar, P., Berggren, P.O., Nilsson, T., 1986. Galanin inhibits glucose-stimulated insulin release by a mechanism involving hyperpolarization and lowering of cytoplasmic free Ca2+ concentration. Biochem. Biophys. Res. Commun. 140, 1059–1063. Bauer, F.E., Ginsberg, L., Venetikou, M., MacKay, D.J., Burrin, J.M., Bloom, S.R., 1986. Growth hormone release in man induced by galanin, a new hypothalamic peptide. Lancet 2, 192–195. Bouvier, M., 2001. Oligomerization of G-protein-coupled transmitter receptors. Nat. Rev. Neurosci. 2, 274–286. Carrillo, J.J., Lopez-Gimenez, J.F., Milligan, G., 2004. Multiple interactions between transmembrane helices generate the oligomeric alpha1b-adrenoceptor. Mol. Pharmacol. 66, 1123–1137. Carron, C., Pascal, A., Djiane, A., Boucaut, J.C., Shi, D.L., Umbhauer, M., 2003. Frizzled receptor dimerization is sufficient to activate the Wnt/beta-catenin pathway. J. Cell. Sci. 116, 2541– 2550. Cheng, Z.J., Harikumar, K.G., Holicky, E.L., Miller, L.J., 2003. Heterodimerization of type A and B cholecystokinin receptors enhance signaling and promote cell growth. J. Biol. Chem. 278, 52972–52979. Crawley, J.N., 1995. Biological actions of galanin. Regul. Pept. 59, 1– 16. Cvejic, S., Devi, L.A., 1997. Dimerization of the delta opioid receptor: implication for a role in receptor internalization. J. Biol. Chem. 272, 26959–26964. Dean, M.K., Higgs, C., Smith, R.E., Bywater, R.P., Snell, C.R., Scott, P.D., Upton, G.J., Howe, T.J., Reynolds, C.A., 2001. Dimerization of G-protein-coupled receptors. J. Med. Chem. 44, 4595–4614. Dinger, M.C., Bader, J.E., Kobor, A.D., Kretzschmar, A.K., BeckSickinger, A.G., 2003. Homodimerization of neuropeptide y receptors investigated by fluorescence resonance energy transfer in living cells. J. Biol. Chem. 278, 10562–10571. Ferguson, S.S., 2001. Evolving concepts in G protein-coupled receptor endocytosis: the role in receptor desensitization and signaling. Pharmacol. Rev. 53, 1–24. Fernando, H., Chin, C., Rosgen, J., Rajarathnam, K., 2004. Dimer dissociation is essential for interleukin-8 (IL-8) binding to CXCR1 receptor. J. Biol. Chem. 279, 36175–36178. Gainetdinov, R.R., Premont, R.T., Bohn, L.M., Lefkowitz, R.J., Caron, M.G., 2004. Desensitization of G protein-coupled receptors and neuronal functions. Annu. Rev. Neurosci. 27, 107–144. Gomes, I., Jordan, B.A., Gupta, A., Trapaidze, N., Nagy, V., Devi, L.A., 2000. Heterodimerization of mu and delta opioid receptors: a role in opiate synergy. J. Neurosci. 20, RC110. Hague, C., Uberti, M.A., Chen, Z., Hall, R.A., Minneman, K.P., 2004. Cell surface expression of alpha1D-adrenergic receptors is
546
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
controlled by heterodimerization with alpha1B-adrenergic receptors. J. Biol. Chem. 279, 15541–15549. Herrick-Davis, K., Grinde, E., Mazurkiewicz, J.E., 2004. Biochemical and biophysical characterization of serotonin 5-HT2C receptor homodimers on the plasma membrane of living cells. Biochemistry 43, 13963–13971. Holmes, F.E., Bacon, A., Pope, R.J., Vanderplank, P.A., Kerr, N.C., Sukumaran, M., Pachnis, V., Wynick, D., 2003. Transgenic overexpression of galanin in the dorsal root ganglia modulates pain-related behavior. Proc. Natl. Acad. Sci. USA 100, 6180–6185. Javitch, J.A., 2004. The ants go marching two by two: oligomeric structure of G-protein-coupled receptors. Mol. Pharmacol. 66, 1077–1082. Jordan, B.A., Devi, L.A., 1999. G-protein-coupled receptor heterodimerization modulates receptor function. Nature 399, 697–700. Jordan, B.A., Trapaidze, N., Gomes, I., Nivarthi, R., Devi, L.A., 2001. Oligomerization of opioid receptors with beta 2-adrenergic receptors: a role in trafficking and mitogen-activated protein kinase activation. Proc. Natl. Acad. Sci. USA 98, 343–348. Kask, K., Berthold, M., Kahl, U., Nordvall, G., Bartfai, T., 1996. Delineation of the peptide binding site of the human galanin receptor. EMBO J. 15, 236–244. Kniazeff, J., Bessis, A.S., Maurel, D., Ansanay, H., Prezeau, L., Pin, J.P., 2004. Closed state of both binding domains of homodimeric mGlu receptors is required for full activity. Nat. Struct. Mol. Biol. 11, 706–713. Kwok, Y.N., Verchere, C.B., McIntosh, C.H., Brown, J.C., 1988. Effect of galanin on endocrine secretions from the isolated perfused rat stomach and pancreas. Eur. J. Pharmacol. 145, 49–54. Lee, S.P., OÕDowd, B.F., Ng, G.Y., Varghese, G., Akil, H., Mansour, A., Nguyen, T., George, S.R., 2000. Inhibition of cell surface expression by mutant receptors demonstrates that D2 dopamine receptors exist as oligomers in the cell. Mol. Pharmacol. 58, 120– 128. Liu, H.X., Brumovsky, P., Schmidt, R., Brown, W., Payza, K., Hodzic, L., Pou, C., Godbout, C., Hokfelt, T., 2001. Receptor subtype-specific pronociceptive and analgesic actions of galanin in the spinal cord: selective actions via GalR1 and GalR2 receptors. Proc. Natl. Acad. Sci. USA 98, 9960–9964. Maggio, R., Vogel, Z., Wess, J., 1993. Coexpression studies with mutant muscarinic/adrenergic receptors provide evidence for intermolecular ‘‘cross-talk’’ between G-protein-linked receptors. Proc. Natl. Acad. Sci. USA 90, 3103–3107. Mandrika, I., Petrovska, R., Wikberg, J., 2005. Melanocortin receptors form constitutive homo- and heterodimers. Biochem. Biophys. Res. Commun. 326, 349–354. Mazarati, A.M., Liu, H., Soomets, U., Sankar, R., Shin, D., Katsumori, H., Langel, U., Wasterlain, C.G., 1998. Galanin modulation of seizures and seizure modulation of hippocampal galanin in animal models of status epilepticus. J. Neurosci. 18, 10070–10077. McIlhinney, R.A., 2004. Generation and use of epitope-tagged receptors. Meth. Mol. Biol. 259, 81–98. Milligan, G., 2004. G protein-coupled receptor dimerization: function and ligand pharmacology. Mol. Pharmacol. 66, 1–7. Mundell, S.J., Pula, G., McIlhinney, R.A., Roberts, P.J., Kelly, E., 2004. Desensitization and internalization of metabotropic glutamate receptor 1a following activation of heterologous Gq/11coupled receptors. Biochemistry 43, 7541–7551.
Ray, K., Hauschild, B.C., 2000. Cys-140 is critical for metabotropic glutamate receptor-1 dimerization. J. Biol. Chem. 275, 34245– 34251. Robbins, M.J., Ciruela, F., Rhodes, A., McIlhinney, R.A., 1999. Characterization of the dimerization of metabotropic glutamate receptors using an N-terminal truncation of mGluR1alpha. J. Neurochem. 72, 2539–2547. Tsien, R.Y., 1998. The green fluorescent protein. Annu. Rev. Biochem. 67, 509–544. Trettel, F., Di Bartolomeo, S., Lauro, C., Catalano, M., Ciotti, M.T., Limatola, C., 2003. Ligand-independent CXCR2 dimerization. J. Biol. Chem. 278, 40980–40988. Uberti, M.A., Hall, R.A., Minneman, K.P., 2003. Subtype-specific dimerization of alpha 1-adrenoceptors: effects on receptor expression and pharmacological properties. Mol. Pharmacol. 64, 1379– 1390. Villar, M.J., Cortes, R., Theodorsson, E., Wiesenfeld-Hallin, Z., Schalling, M., Fahrenkrug, J., Emson, P.C., Hokfelt, T., 1989. Neuropeptide expression in rat dorsal root ganglion cells and spinal cord after peripheral nerve injury with special reference to galanin. Neuroscience 33, 587–604. Vrontakis, M.E., 2002. Galanin: a biologically active peptide. Curr. Drug Targets CNS Neurol. Disord. 1, 531–541. Wager-Miller, J., Westenbroek, R., Mackie, K., 2002. Dimerization of G protein-coupled receptors: CB1 cannabinoid receptors as an example. Chem. Phys. Lipids 121, 83–89. Wang, S., Clemmons, A., Strader, C., Bayne, M., 1998a. Evidence for hydrophobic interaction between galanin and the GalR1 galanin receptor and GalR1-mediated ligand internalization: fluorescent probing with a fluorescein-galanin. Biochemistry 37, 9528–9535. Wang, S., Hashemi, T., Fried, S., Clemmons, A.L., Hawes, B.E., 1998b. Differential intracellular signaling of the GalR1 and GalR2 galanin receptor subtypes. Biochemistry 37, 6711–6717. Wang, S., He, C., Maguire, M.T., Clemmons, A.L., Burrier, R.E., Guzzi, M.F., Strader, C.D., Parker, E.M., Bayne, M.L., 1997. Genomic organization and functional characterization of the mouse GalR1 galanin receptor. FEBS Lett. 411, 225–230. Wrenn, C.C., Kinney, J.W., Marriott, L.K., Holmes, A., Harris, A.P., Saavedra, M.C., Starosta, G., Innerfield, C.E., Jacoby, A.S., Shine, J., Iismaa, T.P., Wenk, G.L., Crawley, J.N., 2004. Learning and memory performance in mice lacking the GAL-R1 subtype of galanin receptor. Eur. J. Neurosci. 19, 1384–1396. Xia, S., Kjaer, S., Zheng, K., Hu, P.S., Bai, L., Jia, J.Y., Rigler, R., Pramanik, A., Xu, T., Hokfelt, T., Xu, Z.Q., 2004. Visualization of a functionally enhanced GFP-tagged galanin R2 receptor in PC12 cells: constitutive and ligand-induced internalization. Proc. Natl. Acad. Sci. USA 101, 15207–15212. Zachariou, V., Brunzell, D.H., Hawes, J., Stedman, D.R., Bartfai, T., Steiner, R.A., Wynick, D., Langel, U., Picciotto, M.R., 2003. The neuropeptide galanin modulates behavioral and neurochemical signs of opiate withdrawal. Proc. Natl. Acad. Sci. USA 100, 9028– 9033. Zachariou, V., Georgescu, D., Kansal, L., Merriam, P., Picciotto, M.R., 2001. Galanin receptor 1 gene expression is regulated by cyclic AMP through a CREB-dependent mechanism. J. Neurochem. 76, 191–200. Zal, T., Gascoigne, N.R., 2004. Photobleaching-corrected FRET efficiency imaging of live cells. Biophys. J. 86, 3923–3939.
Homodimerization and internalization of galanin type 1 receptor in living CHO cells Sebastian A. Wirz a, Christopher N. Davis a,1, Xiaoying Lu a, Tomasz Zal b,2, Tamas Bartfai a,* a
Department of Neuropharmacology, Harold L. Dorris Neurological Research Center, The Scripps Research Institute, 10550 North Torrey Pines Road, SR 307, La Jolla, CA 92037, United States b Department of Immunology, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, United States Received 6 July 2005; accepted 2 September 2005 Available online 20 October 2005
Abstract Galanin is a 29- to 30-aa-long neuropeptide affecting feeding, cognitive, and sexual behavior. It exerts its effects through galanin receptors 1, 2 and 3, which are all seven transmembrane domain G-protein coupled receptors (GPCRs). The GPCRs have been shown to function as monomers, homodimers, heterodimers and oligomers. In this study, we examined the extent of galanin receptor 1 (GalR1) dimerization and internalization in living CHO cells using fluorescence resonance energy transfer (FRET) and time lapse confocal imaging. Ratio imaging analysis and emission spectral analysis revealed substantial homodimerization of GalR1. In addition, internalization of GalR1 after 1 h of agonist stimulation with the GalR1 agonist galanin (1–29) was observed with time lapse fluorescence imaging, whereas stimulation with the GalR2 specific agonist galanin (2–11) did not lead to internalization. Treatment of GalR1 transfected cells with the non-selective adenylyl cyclase activator forskolin influenced the rate of internalization when administered together with galanin (1–29). These results indicate that GalR1 can act as a dimer on the cell surface and that receptor desensitization and internalization was observed after stimulation with the agonist galanin (1–29). Western blots further confirm the FRET data that GalR1-XFP dimerizes and can be detected in the cell as a monomer or dimer using antibodies to XFP. Internalization and dimerization of GalR1 is shown, contributing to the regulation of galanergic signaling. 2005 Elsevier Ltd. All rights reserved. Keywords: Galanin; GPCR; Homodimerization; FRET
1. Introduction Abbreviations: GalR1, galanin receptor 1; GalR2, galanin receptor 2; FRET, fluorescence resonance energy transfer; CFP, cyan fluorescent protein; YFP, yellow fluorescent protein; GPCR, G-protein coupled receptor; CHO cells, Chinese hamster ovarian cells; PBS, phosphate-buffered saline. * Corresponding author. Tel.: +1 858 784 8404; fax: +1 858 784 9099. E-mail addresses: [email protected] (C.N. Davis), tbartfai@ scripps.edu (T. Bartfai). 1 Tel.: +1 858 784 9442. 2 Currently at M.D. Anderson Cancer Center, Department of Immunology, South Campus Bldg Unit 901, 7455 Fannin, Houston, TX 77030, United States. 0143-4179/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.npep.2005.09.001
Galanin has been shown to be important in endocrine, pain, mood and seizure regulation (Crawley, 1995; Vrontakis, 2002). Galanin exerts its effects at three galanin receptors (GalR1-3), each belonging to the superfamily of GPCRs. Galanin type 1 receptor (GalR1) mediates the effects of galanin in regulation of insulin release (Ahren et al., 1986; Kwok et al., 1988), growth hormone release (Bauer et al., 1986), pain threshold (Liu et al., 2001; Holmes et al., 2003), cognitive performance (Wrenn et al., 2004) and seizure threshold (Mazarati et al., 1998). The local concentrations of the endogenous
536
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
agonist galanin can vary broadly, e.g., it is highly inducible upon nerve injury (up to 120-fold upregulation) (Villar et al., 1989) or it is rapidly depleted by seizure activity (Mazarati et al., 1998). Oligomerization and internalization of the galanin receptors may influence receptor–agonist binding and subsequent downstream signaling and may be involved in coping with the large variation in ligand concentrations. Several seven transmembrane domain receptors that couple to the trimeric G proteins have been shown to undergo oligomerization. The past five years have provided multiple examples of GPCR homodimerization (Bouvier, 2001; Dean et al., 2001; Javitch, 2004; Milligan, 2004) among monoamine receptors such as the 5HT2C receptor (Herrick-Davis et al., 2004), a1 adrenoreceptors (Uberti et al., 2003), D2 dopamine receptor (Lee et al., 2000) and metabotropic glutamate receptors mGluR1 (Robbins et al., 1999). Neuropeptide receptor dimerization has been shown with the d opioid receptor (Cvejic and Devi, 1997), neuropeptide Y1, Y2, and Y5 receptors (Dinger et al., 2003), melanocortin receptors MC1R and MC3R (Mandrika et al., 2005). Furthermore, dimerization of CB1 cannabinoid receptor (Wager-Miller et al., 2002), the growth factor receptor Frizzled Xfz3 (Carron et al., 2003) and the cytokine receptor CXCR2 (Trettel et al., 2003) was also shown. The existence of heterodimers of GPCRs has also been demonstrated between the d and j opioid receptors (Jordan and Devi, 1999), d and l opioid receptors (Gomes et al., 2000), type A and B cholecystokinin receptors (Cheng et al., 2003), a 1D and 1B adrenergic receptors (Hague et al., 2004), b 2 adrenergic with d and j (Jordan et al., 2001) and MCR1 and MCR3 melanocortin receptors (Mandrika et al., 2005). The existence of GPCR dimers has been shown by fluorescence energy transfer (FRET) of tagged receptors (Dinger et al., 2003), by bioluminescence energy transfer and acceptor photobleaching (Herrick-Davis et al., 2004), by immunoprecipitation and Western blot analysis of the receptors (Cvejic and Devi, 1997). The structural basis of the dimerization of GPCRs is slowly emerging. For example, the long intracellular loop of the muscarinic M3 receptor has been implicated in dimerization (Maggio et al., 1993) and the large extracellular N-terminal domain of mGluR1 is involved in dimerization, respectively (Kniazeff et al., 2004). Interactions between transmembrane domains of GPCRs including the first, third and fourth transmembrane domains have been suggested to participate in the formation of dimeric GPCRs (Carrillo et al., 2004). While most reported cases of dimerization involve noncovalent binding of monomers, others such as the mGluR1 dimer is formed with a disulfide bond between the cysteine residues of each monomer (Ray and Hauschild, 2000), or in the case of the frizzled receptor Xfz3, dimer disulfide bonds are formed between the intracellular domains of the monomers (Carron et al., 2003).
GPCR homo- and heterodimerization results in altered agonist binding, rate of internalization, receptor ligand pharmacology and G protein mediated signaling (Ferguson, 2001). The most striking example so far is in the case of Interleukin-8 (IL-8) binding to CXCR1 that requires the CXCR1 dimer to dissociate, thus suggesting that the dimerization equilibrium alters agonist binding, leading to changes in receptor signaling (Fernando et al., 2004). In the case of the d opiate receptor dimer, dissociation to monomers precedes agonist binding and internalization of the receptor/agonist complex (Cvejic and Devi, 1997). Frizzled Xfz3 dimerization is sufficient to start the Wnt/b-catenin signaling (Carron et al., 2003), while agonist binding or G-a protein binding had no effect on the state of oligomerization of the NPY 1 receptors (Dinger et al., 2003). The discovery of homo- and heterodimerization of GPCRs provides an interesting explanation of differences in GPCR signaling upon ligand binding. Different agonists to the same GPCR can differ greatly in their ability to stabilize oligomers of the receptor: morphine binding to the d opiate receptor promotes the dissociation of dimers leading to reduced internalization (Cvejic and Devi, 1997), however, other d agonists promote dimerization resulting in accelerated internalization of the receptor dimer. Original studies indicate that two spectral variants of GFP, the cyan fluorescent protein (CFP) and the yellow fluorescent protein (YFP), show the most promising spectral properties for FRET (Tsien, 1998). Here, we report data from combined FRET and Western blot studies on the dimerization and internalization of the GalR1 receptors tagged with either CFP or YFP and heterologously expressed in CHO cells.
2. Materials and methods 2.1. Vector construction C-terminally tagged fusion protein constructs were generated by cloning full length murine coding sequence of galanin receptor 1 into pECFP-N1 and pEYFP-N1 vectors (Clontech Laboratories, Palo Alto, CA, USA). GalR1 cDNA was amplified using Platinum Taq High Fidelity Polymerase (Invitrogen, Carlsbad, CA, USA) with the following primers: GalR1 forward primer: 5 0 GAGCTCAAGCTTATGGAACTGGCTATGGTGAACCT – 3 0 which contained a HindIII site; GalR1 reverse primer: 5 0 – ACCGTCGACTTCACGTGGGTGCAGTTGGT – 3 0 which contained a SalI site and replaced the stop codon. PCR products were digested and cloned into pECFP-N1 and pEYFP-N1 vectors. The constructs were sequenced (The Scripps Research Sequencing Core Facility, La Jolla, CA, USA) to ensure that no mutations had occurred during the PCR and cloning process. As a positive control for FRET between CFP
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
537
and YFP, a CFP–YFP fusion protein linked via 18 amino acids was used (kindly provided by Dr. Peter Vanderklish, The Scripps Research Institute, La Jolla, CA, USA).
using Prism software (GraphPad, San Diego, CA, USA).
2.2. Cell culture and transfection
Membranes were isolated 48 h after transfection from transiently transfected CHO cells expressing CFP or YFP tagged GalR1 as previously described (McIlhinney, 2004). Membranes were resuspended in 10 mM NaHCO3 containing protease inhibitors 16 lg/ml benzamidine HCl, 10 lg/ml phenanthroline, 10 lg/ml aprotinin, 10 lg/ml leupeptin, 10 lg/ml pepstatin A, 1 mM PMSF (protease inhibitor cocktail, BD Biosciences, Palo Alto, CA, USA) and incubated at 60 C for 15 min in the absence or presence of 20 mM DTT before electrophoresis. To determine the extent of monomer and dimer formation in the cell, equal amounts of mock transfected CHO cell and chimeric receptor protein (20 lg) were separated by electrophoresis in a 7% SDS–PAGE gel in the presence of sample buffer, electrophoretically transferred to nitrocellulose membrane and probed with anti-XFP Living Colors antibody (BD Biosciences, Palo Alto, CA, USA).
Chinese hamster ovary (CHO) cells were cultured in DulbeccoÕs modified EagleÕs medium supplemented with 10% (v/v) fetal bovine serum (FBS), 4 mM L-glutamine, 2 mg/ml L-proline, 100 U/ml penicillin/streptomycin (Gibco, Invitrogen Corporation, Carlsbad, CA, USA) at 37 C and in an atmosphere of 5% CO2. Cells were grown in 35 · 10 mm culture dishes and transiently transfected with 4 lL lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) at 80% confluency with 1 lg of either GalR1-CFP, GalR1-YFP or 1 lg of CFP–YFP construct (positive control). For cotransfection experiments, 500 ng GalR1-CFP and 500 ng GalR1-YFP were used. 2.3. Membrane preparation and radioligand binding assay Forty eight hours after transfection, cells were washed three times with cold phosphate-buffered saline (PBS), scraped off the plate and centrifuged at 1000g for 5 min at 4 C. They were then resuspended in hypotonic buffer containing 20 mM Hepes, 5 mM MgCl2 and 1 mM EDTA, and incubated on ice for 10 min. After a final spin at 13,000g for 20 min, the pellet was resupended in 50 mm Tris–HCl, pH 7.4. Ligand binding of [125I] porcine galanin (2200 Ci/ mmol, Perkin–Elmer Life Science, Boston, MA, USA) to CHO cells was performed in 150 lL binding buffer [50 mM Tris–Cl (pH 7.4), 5 mM MgCl2, 0.05% (w/v) bovine serum albumin, supplemented with peptidase and protease inhibitors: 50 lM leupeptin, 100 lM phenylmethanesulfonyl fluoride (PMSF) and 2 lg/mL aprotinin]. Incubations were carried out at room temperature for 45 min and were terminated by rapid vacuum filtration through glass-fiber filters (Packard, Meriden, CT, USA). The filters were washed three times with ice cold PBS (pH 7.4) containing 0.01% Triton X-100 and counted with Cobra II auto-Gamma counting systems (Packard Bioscience, Downer Grove, IL, USA). In all binding experiments, [125I] galanin concentration was 0.2 nM and unlabeled galanin (1–29) (Vulpes Ltd., Tallinn, Estonia) was used at 1 · 105– 1 · 1011 M to displace [125I] galanin binding. Each galanin concentration was performed in triplicate and data were normalized by protein content in the samples (20–30 lg protein in each assay), measured using the BCA protein assay kit from Pierce (Rockford, IL, USA). Non-specific binding of [125I] galanin was determined in the presence of 1 · 105 M galanin (2–11). Binding data were analyzed by nonlinear curve fitting
2.4. Western blot analysis of receptors
2.5. Fluorescence microscopy study of receptor internalization For live cell imaging, an Olympus Fluoview 500 laserscanning confocal microscope was used. Fluorescence images were collected by illumination using the 442 helium cadmium laser and the 488 line of an argon laser. Images were collected on a Delta Vision Optical Sectioning Microscope consisting of an Olympus IX-70 microscope equipped with a mercury arc lamp. A photometrics CH 350 cooled CCD camera and a high precision motorized XYZ stage were used to acquire images of transfected cells at timepoints 0, 60 and 120 min after addition of 1 lM galanin (1–29), 1 lM galanin (2–11), 1 lM M40, or 10 lM Forskolin [in the absence or presence of 1 lM galanin (1–29)] with a 60· immersion objective. The settings of the confocal were maintained constant for each series of experiments so that the resulting images could be analyzed by densitometry and the time dependent changes of internalized receptor could be compared. 2.6. FRET analysis of homodimerization For a detailed description of the FRET analysis system, please refer to Zal and Gascoigne (2004). Briefly, the dual-camera 3I system was controlled by the Slidebook 4.0 software (3I Corporation, Denver, CO, USA) to simultaneously capture CFP emission and YFP emission by two CoolSnap HQ cameras (Roper, Tucson, AZ, USA) attached to a Zeiss 200-M microscope through a beamsplitter (custom 510LPXR, Chroma,
538
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
Rockingham, VT, USA) and stationary bandpass filters. The optical filters were 430/25, 470/30 (CFP excitation and emission), 510/20, and 550/50 nm (YFP excitation and emission), and the JP4 dichroic mirror (Chroma, Rockingham, VT, USA). Rapid wavelength switching between CFP and YFP excitation was performed with a DG4 galvo illuminator customized with a 300 W xenon lamp (Sutter, Novato, CA, USA). YFP excitation was attenuated to 20% by appropriate positioning of the exit galvo. Cameras were typically run in 2 · 2 binning mode with software flat-field correction. The same exposure time was used for all light channels, and background was subtracted in each image based on the average reading in an area of the image devoid of cells. Images were aligned with subpixel resolution using the frequency-based function of the Slidebook software. Sensitized-emission live FRET imaging was performed on the widefield microscope and the sensitized fluorescence of acceptor was detected through an optical FRET filter set selecting acceptor emission during donor excitation (IDA image). The fluorescence of acceptor and the tail of the donor emission spectrum contaminate the IDA image. To account for this bleed-through and to render the FRET index independent of fluorescence intensity, two additional images are acquired: acceptor fluorescence during acceptor excitation (IAA) and donor fluorescence during donor excitation (IDD). Given that the crosstalk coefficients of acceptor and donor (a and d, respectively) in the FRET filter set are constant and assuming that no other crosstalk components are present, sensitized emission, Fc, can be calculated by linear unmixing of the IDA intensity leading to the equation, Fc = IDA dIDD aIAA (Zal and Gascoigne, 2004). Fc is dependent on fluorophore concentration, which leads to the determination of a ratiometric FRET indicator. FRET efficiency was calculated as E = R/(R + G), where the FRET ratio is R = (IDA dIDD aIAA)/ IDD. The values IDA, IDD, and IAA are pixel intensities in the images acquired using donor excitation acceptor emission, donor excitation donor emission, and acceptor excitation acceptor emission, respectively. The bleed-through from YFP during excitation through the 430/25 filter and imaging through the 550/50 filter was calibrated using cells expressing YFP only and was a = 8.6% of the intensity using the YFP excitation and emission filters. The bleed-through of CFP in the 550/50 filter during excitation through the 430/25 filter was calibrated using CFP-only expressing cells and was d = 63% of the intensity in the CFP excitation and emission filter set. The G-factor constant that relates sensitized emission to donor quenching was calibrated at G=3.5 for whole cells on the dual-camera 3I system using the CFP-YFP fusion protein and a photobleaching approach as previously described (Zal and Gascoigne, 2004).
2.7. Statistical analysis Statistical analyses were performed using InStat from GraphPad Software Inc. (San Diego, CA, USA). To determine statistical significance, analysis of variance was used followed by StudentÕs t-test (Figs. 4 and 5). The level of statistically significant difference was defined as p > 0.05.
3. Results 3.1. Binding of ligands to CFP- and YFP- labeled GalR1 To determine if C-terminal CFP and YFP tags alter the ligand binding properties of GalR1, equilibrium binding assays were performed on membranes prepared from CHO cells transiently transfected with GalR1-CFP and GalR1-YFP. Galanin (1–29), the endogenous ligand for galanin receptors, displaced 125I galanin with IC50 values (means ± SEM) of 1.14 ± 0.84 nM for the wild type receptor, 3.08 ± 1.62 nM for GalR1-CFP and 9.48 ± 1.52 for GalR1-YFP. Both GalR1-CFP and GalR1-YFP exhibited high affinity binding to galanin (1–29) with IC50 values that were comparable to the non-tagged receptor (Fig. 2), suggesting that the C-terminal CFP and YFP tags did not impair the ability of galanin (1–29) to bind of GalR1. 3.2. Internalization of GalR1-CFP and GalR1-YFP following agonist stimulation On a subcellular level, confocal microscopy images revealed transfected cells with strong fluorescent signals on the cell surface (Fig. 3A and D). In the presence of galanin (1–29), dose dependent internalization of GalR1-CFP and GalR1-YFP was detected by a redistribution of fluorescence signal intensity from the cell surface to the perinuclear compartment of the cell (Fig. 4). Internalization began as early as 15 min after ligand addition and continued at 120 min. Incubation of transfected cells with galanin (1–29) at concentrations higher than 500 nM led to a fivefold increase in intracellular fluorescent signal intensity and to significant drop in cell membrane fluorescent signal intensity (Fig. 4). Addition of the non-selective adenylyl cyclase activator forskolin had no effect on endocytosis of the XFPtagged GalR1. However, the combination of galanin (1–29) (1 lM) and forskolin (10 lM) led to a slower internalization than observed with galanin (1–29) alone (Fig. 5). 3.3. GalR1 homodimerization CHO cells were transiently transfected with independent FRET pairs: as a positive control, the CFP/YFP
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
pair cloned into one construct and linked via 18 amino acids, as a negative control, two constructs containing only CFP or YFP were cotransfected. FRET efficiency (E) was defined as the proportion of the excited states of the donor that becomes transferred to the acceptor. The positive control showed E = 28%, whereas the negative control was barely above detection background (E = 0–3%). GalR1 FRET efficiency was on average 14%, as measured in 10 different cells from individual experiments (Fig. 6). Fig. 7A–D show the different channels captured: CFP, YFP, merge and FRET. FRET images are shown in pseudocolor, red indicating high levels of FRET efficiency and blue indicating little to no FRET. The positive control produced high levels of FRET efficiency, making this a useful tool to compare maximal FRET signal with the FRET signal obtained in stimulated cells (Fig. 7D). As a negative control, two separate constructs for CFP and YFP were transfected into CHO cells. These fluorescent proteins are indicated as blue pseudocolor suggesting that there was no physical interaction within the cell (Fig. 7E–H). Unstimulated cells expressing GalR1 fusion proteins are indicated by red pseudocolor predominantly on the cell surface level, suggesting close proximity of these receptors as it is commonly seen with oligomerizing receptors (Fig. 8). Addition of the agonist Galanin (1–29) did not cause any increase in intracellular FRET intensity, suggesting that upon agonist stimulation, the GalR1homodimer dissociates and internalizes as a monomer (monomer/agonist complex). Fig. 9 shows the Emission Spectral Analysis of the FRET. The single transfected GalR1-CFP (CFP) and double transfected GalR1CFP/GalR1-YFP (FRET) CHO cells are shown and the increase in fluorescence intensity at 527 nm corresponds to energy transfer between the CFP-YFP fluorophores. The peak in the spectrum at 513 nm is due to Raman emission (not corrected). 3.4. Western blot analysis of GalR1-CFP and of GalR1-YFP Transiently transfected CHO cells were analyzed to determine the extent of receptor monomer and dimer formation in the living cell. Extracts were prepared to
539
isolate chimeric receptors expressed in the cellular membranes. Since the antibodies to GalR1 are not selective, we made use of an antibody directed towards the XFP fluorescent protein to detect the GalR1 receptor. Fig. 10 depicts the CHO cell-expressed CFP and YFP tagged GalR1 receptors as monomers and dimers in presence and absence of the reducing agent Dithiothreitol (20 mM). Receptor monomers and dimers migrated at molecular masses of approximately 60 and 120 kD, respectively. These molecular masses are consistent with the predicted mass of the tagged receptor. CHO cells transiently expressed with either the CFP or YFP alone indicated that the fluorescent proteins migrated at 27 kD and did not oligomerize in the cell (data not shown).
4. Discussion A number of studies have shown homo- and/or heterodimerization of neuropeptide receptors similar to these data. Neuropeptide Y receptors have been shown to homodimerize in a subtype specific manner (Dinger et al., 2003), whereas the opioid receptor subtypes can form both homo- and heterodimers (Cvejic and Devi, 1997; Jordan and Devi, 1999; Gomes et al., 2000). These oligomerizations are important regulatory mechanisms for receptor function, which influence agonist affinity, potency, and efficacy (Bouvier, 2001). Techniques used so far for dimerization studies have utilized both a biochemical and biophysical approaches including co-immunoprecipitation, and Western blot analysis, bioluminescence resonance energy transfer (BRET) and fluorescence resonance energy transfer (FRET). We tested FRET in two different systems, widefield microscopy and fluorescence spectroscopy and confirmed the homodimerization of GalR1 (GalR1-XFP) with Western blots. This study was conducted with C-terminally tagged GalR1 receptors carrying CFP or YFP (Fig. 1) to assess homodimerization and internalization of GalR1 receptors upon agonist binding. The C-terminally tagged GalR1 has retained high affinity for the endogenous agonist galanin (1–29) (Fig. 2). This is expected as the
Fig. 1. Schematic representation of FRET between fluorescently labeled GalR1. If the distance between the donor CFP and acceptor YFP is less than 10 nm, energy transfer occurs (A). Otherwise, no FRET occurs (B).
540
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
Fig. 2. Binding properties of CHO cells transiently expressing GalR1XFP fusion proteins. The two tagged receptors, GalR1-CFP and GalR1-YFP, have similar affinity to the endogenous ligand Galanin (1–29) as the wildtype receptor. IC50 values were as follows: 1.14 ± 0.84 nM for the wild type receptor, 3.08 ± 1.62 nM for GalR1-CFP and 9.48 ± 1.52 for GalR1-YFP (IC50 values are shown as means ± SEM from three independent experiments performed in triplicate).
binding of the 29 amino acids long neuropeptide has earlier been shown to involve the extracellular loops of the receptors and a few amino acid residues in the transmembrane domains 2 and 3 close to the extracellular surface (Kask et al., 1996), and no interactions with the intracellular loops or with the C-terminus that was extended with XFP. Nevertheless, allosteric interactions might have affected the ability of the C-terminally
tagged GalR1 in which case galanin binding and galanin induced effects on dimerization and internalization could be affected. This does not appear to be the case. Confocal imaging of tagged GalR1 transfected into CHO cells revealed cell surface expression (Fig. 3A and D) and substantial galanin 1–29 induced internalization (Fig. 3B, C, E and F), similar to what was observed with fluorescently labeled agonist (Nfluorescein-galanin 1–29) on GalR1 transfected CHO cells (Wang et al., 1998a) and with the fluorescently labeled GalR2 receptors heterologously expressed in PC12 cells (Xia et al., 2004). On a confocal microscope, subcellular receptor expression was seen both on the cell surface as well as intracellularly, predominantly in the perinuclear region. Unlike the observations of tagged GalR2 expressed in PC12 cells (Xia et al., 2004), there was no constitutive internalization in the absence of ligand, as observed with timelapse fluorescent imaging. Similarly, the GalR2 selective ligand galanin (2–11) did not induce endocytosis of GalR1-CFP or of GalR1-YFP confirming the pharmacological specificity of the agonist action of GalR1-XFP. However, the addition of Galanin (1–29) at concentrations of 500 nM, 1 lM and 2 lM induced internalization. The internalization process was time dependent and at room temperature peaked at 60 min after ligand addition and was still ongoing at 120 min (Fig. 4). Addition of 10 lM forskolin alone, a non-selective adenylyl cyclase activator, had no effect on endocytosis of the tagged GalR1. The combination of galanin (1–29) (1 lM) and forskolin (10 lM), however, led to a slower internalization than observed with galanin alone (Fig. 5A and B). This
Fig. 3. The endogenous agonist galanin (1–29) causes internalization of GalR1-XFP in CHO cells. Confocal microscopy of cell surface expression of single transfected CHO cells indicates basal state plasma membrane expression of GalR1-CFP (A) and GalR1-YFP (D). Agonist induced internalization of the tagged galanin receptors are observed following the addition of 1 lM galanin (1–29). Timelapse fluorescent images were taken indicating internalization of the tagged GalR1 after 60 (B and E) and 120 min (C and F). Galanin (1–29) caused a time dependent depletion of cell surface GalR1-XFP and a corresponding accumulation intracellularly as determined in independent measurements.
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
541
Fig. 4. GalR1-XFP transfected CHO cells were subject to various concentrations of galanin (1–29). Fluorescence intensity was measured in five different regions of interest intracellularly and on the cell surface. At the cell surface level, all concentrations resulted in a significant drop of fluorescence intensity, although this is more due to a general drop in fluorescent intensity due to cell motility rather than substantial internalization of the receptor ligand complex (A). Intracellularly, addition of galanin (1–29) at concentrations of 500 nM and above showed a highly significant increase in fluorescence intensity (B). Statistically significant difference from controls are indicated as *, p < 0.05.
finding is consistent with results demonstrating that PKA activation by forskolin inhibited the agonistinduced internalization of mGluR1a and mGluR1b (Mundell et al., 2004). The serine at position 144 within the second intracellular loop has been identified as a potential PKA phosphorylation site (Wang et al., 1997). We assume that cAMP acts via PKA mediated phosphorylation of
GalR1 similarly to numerous GPCRs that are phosphorylated intracellularly (Gainetdinov et al., 2004). In addition to the PKA binding site, the intracellular regions of GalR1 contain potential phosphorylation sites for G protein-coupled receptor kinases (GRKs) and protein kinase C. GPCRs are desensitized and downregulated by the binding of b-arrestins to the phosphorylated tail of the receptor thereby interfering with G
542
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
Fig. 5. GalR1-XFP transfected CHO cells were subject to various ligand stimulations. Cells were treated with 1 lM galanin (1–29), 1 lM galanin (2– 11), 1 lM M40, 10 lM forskolin and a combination of 1 lM galanin (1–29) and 10 lM forskolin. Fluorescence intensity was measured in five different regions of interest on the cell surface and intracellular. Only galanin (1–29) showed a significant decrease in cell surface fluorescence intensity. Intracellularly, both galanin (1–29) and the combination of galanin (1–29) and forskolin showed a significant increase in fluorescent intensity. Forskolin slowed the rate of galanin (1–29) induced internalization. Statistically significant difference from controls are indicated as *, p < 0.05.
protein coupling and initiating rapid localization to clathrin coated pits for internalization. Alternative roles for b-arrestins have been identified in which they can act to mediate receptor signaling, by recruiting components of MAPK cascades into protein signaling complex. Further investigation is necessary to determine the extent of and consequences of GalR1 phosphorylation and subsequent interaction with these kinases and the b-arrestins.
Altered cAMP levels affect signaling via GalR1 in multiple ways: cAMP activation of CREB increases GalR1 synthesis (Zachariou et al., 2001), galanin binding to GalR1 inhibits adenylyl cyclase and depresses cAMP levels (Wang et al., 1998b). These mechanisms form a multiple level mutual control between galaninGalR1 and cellular cAMP. The negative coupling of GalR1 to cAMP has been shown to decrease neuronal firing, suggesting a compensatory role for this receptor.
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
It was shown that peripheral administration of the GalR1 antagonist galnon attenuates the increase in cAMP in the locus coeruleus due to morphine withdrawal (Zachariou et al., 2003). These results provide a mechanism by which a direct activation of the galanin receptor leads to changes in cAMP production and altered neuronal activity. Using ratio imaging and emission spectral analysis we provide for the first time evidence of a physical association between GalR1 receptors. The expression levels of GalR1-CFP and GalR1-YFP were comparable upon cotransfection of both GalR1-CFP and GalR1-YFP constructs. To assess whether FRET occurrence was due to actual dimerization and not due to random receptor clustering (random proximity effect), cells were transfected with increasing amounts of acceptor DNA and checked for differing FRET signals. Donor/acceptor ratios between 1:2 and 1:8 did not show a difference in FRET efficiency, indicating independence of acceptor expression level and that the signal was not caused by random crowding of receptors but rather due to specific interactions between the fluorescently tagged GalR1 receptors (data not shown). FRET studies were conducted on a widefield microscope system. GalR1 transfected CHO cells constitutively showed on average a 14% FRET efficiency on the cell surface, indicating a very significant extent of dimer formation between GalR1-CFP and GalR1-YFP (Fig. 6). When treated with the agonist galanin (1–29), no increase in FRET efficiency intracellularly was observed after 60 and 120 min. This would be expected if
543
Fig. 6. FRET efficiency (E) is defined as the proportion of the excited states of the donor that becomes transferred to the acceptor. The positive control showed E = 28%, whereas the negative control was barely above detection background. GalR1 FRET efficiency was on average 16%, as measured in 10 different cells from individual experiments.
the GalR1 would internalize as a dimer (Fig. 8). Rather, it appears that the GalR1 internalizes as a monomer upon agonist stimulation, similarly to what was previously observed with the d opioid receptor, where dissociation of the receptor dimer precedes the agonist-induced internalization of the receptor (Cvejic and Devi, 1997). Western blots of GalR1-CFP and GalR1-YFP were visualized with an antibody recognizing CFP and
Fig. 7. GalR1 homodimerization as measured in FRET efficiency on a widefield microscope. As a positive control for FRET, a CFP–YFP fusion protein linked via 18 amino acids was expressed in CHO cells (A–D). As a negative control, separate CFP and YFP plasmids were co-transfected into CHO cells (E–H). Pictures on the right are colored in pseudocolor, red meaning high levels of dimerization and blue meaning low levels to none. Transiently transfected CHO cells constitutively express the GalR1 as dimers on the cell surface. After addition of galanin (1–29), the receptor seems to internalize as a monomer and aggregates intracellularly in the perinuclear region. (For interpretation of the references in color in this figure legend, the reader is referred to the web version of this article.)
544
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
Fig. 8. Timelapse FRET imaging after addition of Galanin (1–29). CHO cells were co-transfected with GalR1-CFP and GalR1-YFP and treated with the agonist galanin (1–29). No increase in FRET efficiency intracellularly was observed after 60 and 120 min. This would be expected if the GalR1 would internalize as a dimer. Rather, it appears that the GalR1 internalizes as a monomer upon agonist stimulation. (For interpretation of the references in color in this figure legend, the reader is referred to the web version of this article.)
YFP, as the antibodies to GalR1 are not selective. The expected molecular weights for the monomers of GALR1-YFP or GALR-CFP are approximately 63 kD based on the primary sequence of the covalent complex of GalR1 and YFP or CFP. We have detected a band at approximately 60 and at 120 kD in both GalR1-CFP and in GalR1-YFP transfected cells in the presence of the disulfide reducing agent DTT (Fig. 10). These data confirm the FRET data that homodimers of GalR1-CFP and GalR1-YFP were responsible for the FRET shown in Figs. 7 and 8. The GalR1 receptor has three cysteine residues in transmembrane and intracellular domains, and these appear to be involved in the formation of the GalR1 dimer as receptor monomers are observed only in the presence of DTT, while a dimer of GalR1-XFP is observed at 120 kD under nonreducing conditions (see Figs. 9 and 10). Our data illustrate that GalR1 form homodimers to a large extent at the cell surface and to some extent already in the ER. The binding of agonist galanin (1–29) does not directly affect the dimer-monomer equilibrium as no changes in FRET at the cell surface were observed but it induces internalization of the GalR1 monomer, probably with the agonist bound to it, as suggested from the study with N-fluorescein-galanin
(1–29) bound to GalR1 (Wang et al., 1998a). Thus, the agonist is indirectly shifting the dimer–monomer equilibrium towards monomers which internalize once they bind galanin. Therefore, it is likely that GalR1 dimers formed in the endoplasmic reticulum or in the plasma membrane in the absence of agonist are in equilibrium with monomers and this equilibrium is shifted through dissociation and subsequent internalization of GalR1–galanin complexes. Our data are not useful to determine whether galanin binds to the GalR1 homodimer and that the galanin occupied dimer that dissociates or whether galanin binds to monomeric GalR1 only. The rate of internalization is affected by galanin acting on the extracellular side of the GalR1 and by elevation of cAMP levels intracellularly, by forskolin in the opposite direction. The role of GalR1 homodimers in galaninergic signaling requires further studies as the present data only address the existence and the regulation of dimers in the processes of agonist binding and internalization, but does not provide data on the pharmacology of receptor oligomerization. This is the case for most GPCR dimers as we have yet to find pharmacological agents that distinguish between binding to monomeric and oligomeric GPCRs or which signal differently through these different oligomeric forms of the receptor.
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
545
Acknowledgements The authors thank Dr. Kathy Spencer for valuable assistance in confocal microscopy, Dr. Peter Vanderklish for providing the CFP–YFP construct as a positive control, Liam Palmer for excellent technical assistance in fluorescence spectroscopy and Dr. Nicolas Gascoigne for use of the widefield microscope. This study was supported by National Institutes of Health Grant ROIMH63080 and ROI NS043409 (T.B.) and a stipend from the Swiss National Science Foundation (S.W.). This article is manuscript no. 17221-NP from The Scripps Research Institute.
References
Fig. 9. Emission spectral analysis of the FRET analysis. The single transfected GalR1-CFP (CFP) and double transfected GalR1-CFP/ GalR1-YFP (FRET) CHO cells are shown in the upper panel. The lower panel indicates the spectral analysis of the CFP/YFP pair positive control cloned into one construct and linked via 18 amino acids. An increase in fluorescence intensity at 527 nm is indicated by the arrow and corresponds to energy transfer between the fluorophores. The peak at 513 nm is due to Raman emission (not corrected).
Fig. 10. Dimerization of fluorescently tagged GalR1. Membranes were prepared from CHO cells transfected with CFP- and YFP-tagged GalR1 (GalR1CFP and GalR1YFP). Representative blots of two independent experiments are shown. Chimeric receptors as detected by anti-XFP antibody exist as a dimer in the cell in the absence of reducing agent. In the presence of 20 mM DTT, monomers of the chimeric receptors are detected. Monomer and dimer of receptors are indicated by a * and * *, respectively. Mock-transfected CHO cells (CHO) are indicated in the absence and presence of DTT.
Ahren, B., Arkhammar, P., Berggren, P.O., Nilsson, T., 1986. Galanin inhibits glucose-stimulated insulin release by a mechanism involving hyperpolarization and lowering of cytoplasmic free Ca2+ concentration. Biochem. Biophys. Res. Commun. 140, 1059–1063. Bauer, F.E., Ginsberg, L., Venetikou, M., MacKay, D.J., Burrin, J.M., Bloom, S.R., 1986. Growth hormone release in man induced by galanin, a new hypothalamic peptide. Lancet 2, 192–195. Bouvier, M., 2001. Oligomerization of G-protein-coupled transmitter receptors. Nat. Rev. Neurosci. 2, 274–286. Carrillo, J.J., Lopez-Gimenez, J.F., Milligan, G., 2004. Multiple interactions between transmembrane helices generate the oligomeric alpha1b-adrenoceptor. Mol. Pharmacol. 66, 1123–1137. Carron, C., Pascal, A., Djiane, A., Boucaut, J.C., Shi, D.L., Umbhauer, M., 2003. Frizzled receptor dimerization is sufficient to activate the Wnt/beta-catenin pathway. J. Cell. Sci. 116, 2541– 2550. Cheng, Z.J., Harikumar, K.G., Holicky, E.L., Miller, L.J., 2003. Heterodimerization of type A and B cholecystokinin receptors enhance signaling and promote cell growth. J. Biol. Chem. 278, 52972–52979. Crawley, J.N., 1995. Biological actions of galanin. Regul. Pept. 59, 1– 16. Cvejic, S., Devi, L.A., 1997. Dimerization of the delta opioid receptor: implication for a role in receptor internalization. J. Biol. Chem. 272, 26959–26964. Dean, M.K., Higgs, C., Smith, R.E., Bywater, R.P., Snell, C.R., Scott, P.D., Upton, G.J., Howe, T.J., Reynolds, C.A., 2001. Dimerization of G-protein-coupled receptors. J. Med. Chem. 44, 4595–4614. Dinger, M.C., Bader, J.E., Kobor, A.D., Kretzschmar, A.K., BeckSickinger, A.G., 2003. Homodimerization of neuropeptide y receptors investigated by fluorescence resonance energy transfer in living cells. J. Biol. Chem. 278, 10562–10571. Ferguson, S.S., 2001. Evolving concepts in G protein-coupled receptor endocytosis: the role in receptor desensitization and signaling. Pharmacol. Rev. 53, 1–24. Fernando, H., Chin, C., Rosgen, J., Rajarathnam, K., 2004. Dimer dissociation is essential for interleukin-8 (IL-8) binding to CXCR1 receptor. J. Biol. Chem. 279, 36175–36178. Gainetdinov, R.R., Premont, R.T., Bohn, L.M., Lefkowitz, R.J., Caron, M.G., 2004. Desensitization of G protein-coupled receptors and neuronal functions. Annu. Rev. Neurosci. 27, 107–144. Gomes, I., Jordan, B.A., Gupta, A., Trapaidze, N., Nagy, V., Devi, L.A., 2000. Heterodimerization of mu and delta opioid receptors: a role in opiate synergy. J. Neurosci. 20, RC110. Hague, C., Uberti, M.A., Chen, Z., Hall, R.A., Minneman, K.P., 2004. Cell surface expression of alpha1D-adrenergic receptors is
546
S.A. Wirz et al. / Neuropeptides 39 (2005) 535–546
controlled by heterodimerization with alpha1B-adrenergic receptors. J. Biol. Chem. 279, 15541–15549. Herrick-Davis, K., Grinde, E., Mazurkiewicz, J.E., 2004. Biochemical and biophysical characterization of serotonin 5-HT2C receptor homodimers on the plasma membrane of living cells. Biochemistry 43, 13963–13971. Holmes, F.E., Bacon, A., Pope, R.J., Vanderplank, P.A., Kerr, N.C., Sukumaran, M., Pachnis, V., Wynick, D., 2003. Transgenic overexpression of galanin in the dorsal root ganglia modulates pain-related behavior. Proc. Natl. Acad. Sci. USA 100, 6180–6185. Javitch, J.A., 2004. The ants go marching two by two: oligomeric structure of G-protein-coupled receptors. Mol. Pharmacol. 66, 1077–1082. Jordan, B.A., Devi, L.A., 1999. G-protein-coupled receptor heterodimerization modulates receptor function. Nature 399, 697–700. Jordan, B.A., Trapaidze, N., Gomes, I., Nivarthi, R., Devi, L.A., 2001. Oligomerization of opioid receptors with beta 2-adrenergic receptors: a role in trafficking and mitogen-activated protein kinase activation. Proc. Natl. Acad. Sci. USA 98, 343–348. Kask, K., Berthold, M., Kahl, U., Nordvall, G., Bartfai, T., 1996. Delineation of the peptide binding site of the human galanin receptor. EMBO J. 15, 236–244. Kniazeff, J., Bessis, A.S., Maurel, D., Ansanay, H., Prezeau, L., Pin, J.P., 2004. Closed state of both binding domains of homodimeric mGlu receptors is required for full activity. Nat. Struct. Mol. Biol. 11, 706–713. Kwok, Y.N., Verchere, C.B., McIntosh, C.H., Brown, J.C., 1988. Effect of galanin on endocrine secretions from the isolated perfused rat stomach and pancreas. Eur. J. Pharmacol. 145, 49–54. Lee, S.P., OÕDowd, B.F., Ng, G.Y., Varghese, G., Akil, H., Mansour, A., Nguyen, T., George, S.R., 2000. Inhibition of cell surface expression by mutant receptors demonstrates that D2 dopamine receptors exist as oligomers in the cell. Mol. Pharmacol. 58, 120– 128. Liu, H.X., Brumovsky, P., Schmidt, R., Brown, W., Payza, K., Hodzic, L., Pou, C., Godbout, C., Hokfelt, T., 2001. Receptor subtype-specific pronociceptive and analgesic actions of galanin in the spinal cord: selective actions via GalR1 and GalR2 receptors. Proc. Natl. Acad. Sci. USA 98, 9960–9964. Maggio, R., Vogel, Z., Wess, J., 1993. Coexpression studies with mutant muscarinic/adrenergic receptors provide evidence for intermolecular ‘‘cross-talk’’ between G-protein-linked receptors. Proc. Natl. Acad. Sci. USA 90, 3103–3107. Mandrika, I., Petrovska, R., Wikberg, J., 2005. Melanocortin receptors form constitutive homo- and heterodimers. Biochem. Biophys. Res. Commun. 326, 349–354. Mazarati, A.M., Liu, H., Soomets, U., Sankar, R., Shin, D., Katsumori, H., Langel, U., Wasterlain, C.G., 1998. Galanin modulation of seizures and seizure modulation of hippocampal galanin in animal models of status epilepticus. J. Neurosci. 18, 10070–10077. McIlhinney, R.A., 2004. Generation and use of epitope-tagged receptors. Meth. Mol. Biol. 259, 81–98. Milligan, G., 2004. G protein-coupled receptor dimerization: function and ligand pharmacology. Mol. Pharmacol. 66, 1–7. Mundell, S.J., Pula, G., McIlhinney, R.A., Roberts, P.J., Kelly, E., 2004. Desensitization and internalization of metabotropic glutamate receptor 1a following activation of heterologous Gq/11coupled receptors. Biochemistry 43, 7541–7551.
Ray, K., Hauschild, B.C., 2000. Cys-140 is critical for metabotropic glutamate receptor-1 dimerization. J. Biol. Chem. 275, 34245– 34251. Robbins, M.J., Ciruela, F., Rhodes, A., McIlhinney, R.A., 1999. Characterization of the dimerization of metabotropic glutamate receptors using an N-terminal truncation of mGluR1alpha. J. Neurochem. 72, 2539–2547. Tsien, R.Y., 1998. The green fluorescent protein. Annu. Rev. Biochem. 67, 509–544. Trettel, F., Di Bartolomeo, S., Lauro, C., Catalano, M., Ciotti, M.T., Limatola, C., 2003. Ligand-independent CXCR2 dimerization. J. Biol. Chem. 278, 40980–40988. Uberti, M.A., Hall, R.A., Minneman, K.P., 2003. Subtype-specific dimerization of alpha 1-adrenoceptors: effects on receptor expression and pharmacological properties. Mol. Pharmacol. 64, 1379– 1390. Villar, M.J., Cortes, R., Theodorsson, E., Wiesenfeld-Hallin, Z., Schalling, M., Fahrenkrug, J., Emson, P.C., Hokfelt, T., 1989. Neuropeptide expression in rat dorsal root ganglion cells and spinal cord after peripheral nerve injury with special reference to galanin. Neuroscience 33, 587–604. Vrontakis, M.E., 2002. Galanin: a biologically active peptide. Curr. Drug Targets CNS Neurol. Disord. 1, 531–541. Wager-Miller, J., Westenbroek, R., Mackie, K., 2002. Dimerization of G protein-coupled receptors: CB1 cannabinoid receptors as an example. Chem. Phys. Lipids 121, 83–89. Wang, S., Clemmons, A., Strader, C., Bayne, M., 1998a. Evidence for hydrophobic interaction between galanin and the GalR1 galanin receptor and GalR1-mediated ligand internalization: fluorescent probing with a fluorescein-galanin. Biochemistry 37, 9528–9535. Wang, S., Hashemi, T., Fried, S., Clemmons, A.L., Hawes, B.E., 1998b. Differential intracellular signaling of the GalR1 and GalR2 galanin receptor subtypes. Biochemistry 37, 6711–6717. Wang, S., He, C., Maguire, M.T., Clemmons, A.L., Burrier, R.E., Guzzi, M.F., Strader, C.D., Parker, E.M., Bayne, M.L., 1997. Genomic organization and functional characterization of the mouse GalR1 galanin receptor. FEBS Lett. 411, 225–230. Wrenn, C.C., Kinney, J.W., Marriott, L.K., Holmes, A., Harris, A.P., Saavedra, M.C., Starosta, G., Innerfield, C.E., Jacoby, A.S., Shine, J., Iismaa, T.P., Wenk, G.L., Crawley, J.N., 2004. Learning and memory performance in mice lacking the GAL-R1 subtype of galanin receptor. Eur. J. Neurosci. 19, 1384–1396. Xia, S., Kjaer, S., Zheng, K., Hu, P.S., Bai, L., Jia, J.Y., Rigler, R., Pramanik, A., Xu, T., Hokfelt, T., Xu, Z.Q., 2004. Visualization of a functionally enhanced GFP-tagged galanin R2 receptor in PC12 cells: constitutive and ligand-induced internalization. Proc. Natl. Acad. Sci. USA 101, 15207–15212. Zachariou, V., Brunzell, D.H., Hawes, J., Stedman, D.R., Bartfai, T., Steiner, R.A., Wynick, D., Langel, U., Picciotto, M.R., 2003. The neuropeptide galanin modulates behavioral and neurochemical signs of opiate withdrawal. Proc. Natl. Acad. Sci. USA 100, 9028– 9033. Zachariou, V., Georgescu, D., Kansal, L., Merriam, P., Picciotto, M.R., 2001. Galanin receptor 1 gene expression is regulated by cyclic AMP through a CREB-dependent mechanism. J. Neurochem. 76, 191–200. Zal, T., Gascoigne, N.R., 2004. Photobleaching-corrected FRET efficiency imaging of live cells. Biophys. J. 86, 3923–3939.