点击显示 收起
【关键词】 Cannabinoid Receptor
In central neurons, the cell-surface distribution of cannabinoid receptor subtype-1 (CB1) is highly polarized toward axons and is associated with synaptic terminals, in which it is well-positioned to modulate neurotransmitter release. It has been suggested that high levels of constitutive activity mediate CB1 receptor axonal targeting, leading to domain-specific endocytosis. We have investigated further the mechanisms that underlie CB1 receptor axonal polarization in hippocampal neurons and found that constitutive activity is not an essential requirement for this process. We demonstrate that the cell-surface distribution of an N-terminally tagged, fluorescent CB1 receptor fusion-protein is almost exclusively localized to the axon when expressed in cultured hippocampal neurons. Inhibition of endocytosis by cotransfection with a dominant-negative dynamin-1 (K44A) mutant traps both recombinant and endogenous CB1 receptors at the somatodendritic cell surface. However, this effect could not be mimicked by inhibiting constitutive activity or receptor activation, either by expressing mutant receptors that lack these properties or by treatment with CB1 receptor antagonists possessing inverse agonist activity. These data are consistent with a revised model in which domain-specific endocytosis regulates the functional polarization of CB1 receptors, but this process is distinct from constitutive activity.
The cannabinoid receptor subtype-1 (CB1) is the most highly expressed G protein-coupled receptor (GPCR) in the mammalian central nervous system and is present at presynaptic terminals and axonal fibers in many brain areas, where it is believed to inhibit neurotransmitter release after agonist binding (Freund et al., 2003). In cultured hippocampal neurons, the cell-surface distribution of endogenous CB1 receptors is polarized with high levels of expression on GABA-expressing axons, in which they are associated with presynaptic terminals (Irving et al., 2000). A recent study suggested that selective targeting of CB1 receptors to axons involves differences in the rate of endocytosis between dendritic and axonal regions, leading to a net accumulation within the axonal compartment, with this process driven by constitutive activation of the CB1 receptor (Leterrier et al., 2006). Thus, CB1 receptor antagonists, which are useful therapeutic agents, especially for the treatment of obesity (Van Gaal et al., 2005), may potentially lead to mistargeting of the receptor to the somatodendritic domain and aberrant endocannabinoid function upon cessation of treatment.
In this study, we used N-terminally tagged enhanced green fluorescent protein (GFP)-CB1 chimeras and immunocytochemical labeling of endogenous surface receptors to further elucidate the mechanisms underlying the polarized distribution of CB1 receptors in hippocampal neurons. For both endogenous CB1 receptors and GFP-CB1 receptor chimeras, polarized surface expression within the axon is achieved by preferential removal from the somatodendritic domain by endocytosis. Using CB1 receptor antagonists with inverse agonist properties and expression of mutant receptors, we demonstrate that constitutive activity of the receptor is not essential for this effect and that this basal endocytosis represents a distinct cellular process. A revised model is therefore proposed to account for the CB1 receptor axonal targeting.
Constructs and Reagents. Papain and antimicrotubule-associated protein 2 (MAP2) were obtained from Sigma-Aldrich (Poole, Dorset, UK). Antibodies to glutamate decarboxylase type 65 (GAD65) from Chemicon (Hampshire, UK) and the Developmental Studies Hybridoma Bank (GAD-6 clone; Iowa City, Iowa), respectively. Cell culture reagents were sourced from Invitrogen, (Paisley, UK). Indocarbocyanine- and indodicarbocyanine-tagged secondary antibodies were obtained from Stratech Scientific Ltd. (Cambridge, UK), and rabbit anti-GFP was from Abcam (Cambridge, UK). AM281, Win55212-2, and HU210 and forskolin were obtained from Tocris Cookson (Avonmouth, UK). SR141716A was a kind gift from Sanofi-Aventis (Montpelier, France). N-terminal [amino acids (aa) 84-99] CB1 receptor antibody was obtained from Alomone (Jerusalem, Israel). All restriction enzymes were from Promega (Southampton, UK). Zero Blunt TOPO PCR Cloning Kit and cell culture reagents were obtained from Invitrogen. The original rat CB1 receptor cDNA and the N-terminal (aa 1-77) CB1 receptor antibody were kind gifts from Dr. Ken Mackie (University of Washington, Seattle, WA). Cannabinoids were made as 1000x stock solutions in dimethyl sulfoxide (DMSO) and diluted accordingly in culture medium. Dynamin-1 constructs [hemagglutinin (HA)-tagged wild type and K44A] were a kind gift from Dr. Marc Caron (Duke University, Durham, NC) to Dr. Connolly.
Generation of Constructs. GFP was introduced at the N terminus (between residues 25-26) of the cloned, rat CB1 receptor cDNA and expressed in a pcDNA1/AMP plasmid. Replacement of residues 1 to 25 with an optimized signal sequence (SS, amino acids 1-33; MATGSPTSLLLAFGLLCLPWLQEGSARDPPVAT) derived from the human growth hormone (Connolly et al., 1994) enabled efficient surface expression. The human growth hormone-derived signal sequence was amplified by polymerase chain reaction and cloned into pcDNA1 as an HindIII/BamHI fragment to produce SS-pcDNA1. GFP was amplified from pGFP-N1 by polymerase chain reaction and cloned as BamHI/XhoI fragment into SS-pcDNA1 to generate SS-GFP-pcDNA1. Amino acids 26 to 472 of rat CB1 were amplified by polymerase chain reaction and cloned as a blunt-ended fragment into the TOPO vector. CB1 26 to 472 was then digested out of TOPO vector and cloned as a SalI/EcoRI fragment into SS-GFP-pcDNA1 to produce the GFP-CB1 construct. Point mutation (D164N) and truncation (14) constructs were generated from this construct and subcloned into a PRK5 vector.
Hippocampal Cell Culture. Primary cultures of rat hippocampal neurons were prepared using a protocol modified from Coutts et al., (2001). In brief, rat pups (1-3 days old) were sacrificed according to United Kingdom Home Office guidelines. Dissected hippocampi were treated with papain (1.5 mg/ml) in HEPES-buffered saline for 25 min at 37°C. Dissociated cells were plated onto coverslips, pretreated with poly(L-lysine) (15 µg/ml). Cultures were incubated in Neurobasal A medium with B27 supplement (2%) or minimum essential medium with serum replacement 2 (Sigma) in a humidified atmosphere of 5% CO2 at 37°C. These culture conditions did not affect CB1 receptor expression and axonal polarity. Cells were used from 3 to 14 days in vitro (DIV), with polarity studies carried out using neurons older than 7 DIV, which exhibit axonal expression of native CB1 receptors. Treatment with cannabinoid antagonists or DMSO vehicle controls was carried out in culture medium at 37°C.
Neuronal Transfection. Constructs were recombinantly expressed in primary cultures of hippocampal neurons using a CaPO4 transfection protocol adapted from Jiang et al. (2004). Plasmid DNA (1 µg of GFP-CB1 construct per coverslip) was diluted in Tris-EDTA transfection buffer (10 mM Tris-HCl and 2.5 mM EDTA, pH 7.3). CaCl2 solution (2.5 M in 10 mM HEPES) was then added, drop-wise, to the DNA solution to give a final concentration of 250 mM CaCl2. This was then added to an equivalent volume of HEPES-buffered transfection solution (274 mM NaCl, 10 mM KCl, 1.4 mM Na2HPO4, 11 mM dextrose, and 42 mM HEPES, pH 7.2). A fraction (1/8th) of the DNA solution was added to the HEPES-buffered transfection solution. This was then vortexed gently for 2 to 3 s, and the precipitate was allowed to develop at room temperature for 30 min, protected from the light. Then, 50 µl of precipitate was added, drop-wise, to each coverslip, and the cultures were incubated with precipitate for 1 to 3 h in the presence of kynurenic acid (2 mM). Each coverslip was transferred to a fresh well of the 24-well plate containing 1 ml of culture medium with kynurenic acid (2 mM), acidified by equilibration in a 10% CO2/90% O2 incubator for 24 h, and the plate was returned to a 37°C/5% CO2/95% O2 incubator for 15 to 20 min. Each coverslip was then transferred to a fresh well of the 24-well plate containing reserved, conditioned medium. The cells were then returned to a 37°C 5% CO2/95% incubator to allow expression of the transfected constructs. Cells were typically analyzed 16 to 48 h after transfection. This procedure resulted in rates of 5 to 15% transfection efficiency
HEK293 and COS-7 Cell Culture and Transfection. COS-7 cells (American Type Culture Collection, Manassas, VA) and HEK293 cells (Invitrogen) were maintained in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% heat-inactivated neonatal calf serum (Invitrogen), 2 mM glutamine, 1 mM sodium pyruvate, 100 µg/ml streptomycin, and 100 U/ml penicillin in an atmosphere of 5% CO2. Suspensions of exponentially growing cells (2 x 106 cells), detached after trypsin exposure, were transfected by electroporation (400 V, infinity resistance, 125 µF, Gene Electropulser II; Bio-Rad Laboratories, Hemel Hempstead, Hertfordshire, UK) in the presence of 10 µg of DNA. These were then plated onto poly(lysine)-coated coverslips (15 µg/ml). In some studies, cells growing on coverslips were transfected with Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Cells were analyzed 12 to 24 h after transfection.
cAMP Immunoassay. Inhibition of forskolin-stimulated cAMP production by the cannabinoid agonist HU210 was assessed using the Catchpoint 96 Fluorescent cAMP immunoassay kit (Molecular Devices, Wokingham, UK). Exponentially growing HEK293 cells were transfected by electroporation with either wild-type CB1 or GFP-CB1 and plated onto a 96-well dish coated with poly(D-lysine) (0.15 mg/ml in borate buffer, pH 8.2) Confluent cells were incubated with the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (100 nM) for 10 min. The stimulation was initiated by the addition of 5 µM forskolin and the appropriate concentration of HU210 and was incubated for 15 min at 37°C. The reaction was stopped by removing the culture medium and the addition of lysis buffer [diluted to 1x in phosphate-buffered saline (PBS)], after which 40 µl was removed for the cAMP fluorescent immunoassay. The assay was carried out according to manufacturer's instructions in triplicate.
Cell-Surface Enzyme-Linked Immunosorbent Assay. HEK293 cells were fixed in 3% paraformaldehyde in PBS and then washed twice in 50 mM NH4Cl (in PBS) and blocked (10% fetal bovine serum, 0.5% bovine serum albumin in PBS) for 1 h. Subsequent washes were performed in block. Cell-surface detection was then carried out using a primary antibody raised against full-length GFP (1:50; Santa Cruz Biotechnology, Santa Cruz, CA) followed by a secondary antibody conjugated to horseradish peroxidase. To quantify the total amount of receptors present, cells from a separate, matched sample of cells from the same transfection were treated with the detergent Triton X-100 (0.5%, 15 min) to permeabilize the cell membranes before treatment with primary antibody. Receptor expression was determined using a horseradish peroxidase-conjugated secondary antibody and assayed using a spectrophotometer (VersaMax; Molecular Devices,), with 3,3',5,5'-tetramethylbenzidine (Sigma-Aldrich) as the substrate, with detection at 450 nm 30 min after the addition of 0.5 M H2SO4. n Values represent the number of pooled replicates from three independent experiments, normalized to vehicle control.
Immunofluorescence, Confocal Microscopy, and Trafficking Studies. Neurons were fixed in 3% paraformaldehyde (in PBS), washed twice in 50 mM NH4Cl (in PBS), and blocked (10% fetal bovine serum and 0.5% bovine serum albumin in PBS) for 30 min. Subsequent washes and antibody dilutions were performed in PBS containing 10% fetal bovine serum and 0.5% bovine serum albumin at room temperature. In some experiments, including those for native CB1 receptor expression, primary antibody incubations (in HEPES-buffered saline) to label surface receptors were carried out before fixation on live cells. This protocol resulted in a degree of antibody-induced receptor clustering but did not affect the ratio of axonal-dendritic receptor expression. Primary antibodies raised against the N terminus of the CB1 receptor (dilution 1:50-1:1000; 40-60 min) were used for endogenous CB1 receptor labeling or against full-length GFP (1:500-1:2000; 60 min) for recombinant GFP-CB1. Dual-labeling immunohistochemistry was performed after permeabilization (0.5% Triton X-100; 10 min) using monoclonal anti-MAP2 (somatodendritic marker) or anti-GAD65 (GABAergic cell marker) antibodies. For comparison of endogenous cell-surface and intracellular CB1 receptor-labeling, live cells were labeled with N-terminal (aa 84-99) antibody (1:100), fixed, then probed with secondary antibody (Alexa488 or indodicarbocyanine conjugations). The remaining total CB1 receptor immunoreactivity was assessed after permeabilization (1:500-1000 dilution of anti-CB1). Live antibody labeling was carried out in the presence of compounds used for cell treatment protocols (DMSO, AM281, and/or Win55212-2). A LSM510 confocal imaging system (Carl Zeiss, Jena, Germany) was used for image acquisition and processing. Images of multiple labeling were obtained in multitracking mode using a 40x water immersion objective. Z-projections were generated from a series of confocal sections (1.5- to 2.5-µm intervals) using a corresponding iris aperture (3-5 µm).
Receptor trafficking studies in neurons and COS7 cells were carried out using a Zeiss LSM510 confocal imaging system at 30°C using a heated stage. Full Z-stack images (2.5-µm sections) were obtained every 10 min using minimal 488 nm excitation (1.9-s scan speed at 521 x 521 pixels; 1-3% maximal argon laser power). For agonist-induced trafficking, low-noise images (7.9-s scan speed) were obtained at the beginning and the end of the experiment (90 min after agonist exposure).
Measurement of Surface Polarity and Data Analysis. The degree of surface polarization for fluorescent CB1 receptor chimeras and endogenous CB1 was quantified for reconstructed Z-projections of neurons, analyzed using Zeiss (Laser Sharp) software. The average pixel intensity of surface labeling (indocarbocyanine) was obtained for five regions per axonal or dendritic compartment (20 µm long x 1 pixel wide, GFP channel or intracellular, endogenous CB1 signal), traced at random. Averaged data with background values subtracted was used to generate a dendritic/axonal (D/A) polarity index. A small population of highly expressing neurons exhibited a cell-surface distribution of GFP-CB1 that was nonpolarized and were excluded from analysis (<5%). Interneurons expressing endogenous high levels of CB1 receptor were identified by their increased levels of somatodendritic CB1 receptor immunoreactivity, which was often vesicular (N-terminal antibody; aa 84-99). For GFP-CB1-expressing neurons, axons and soma/dendrites were analyzed within the same cell; for endogenous CB1 receptors, dendritic labeling was compared with axonal fluorescence derived from multiple neurons. Data were compared using appropriate statistical tests (as stated in the results or figure legends using Instat 3; GraphPad Software, San Diego, CA). Significance was noted at the level of P < 0.05. Data are presented as mean ± S.E.M., and n values reflect the number of cells or fields (for endogenous CB1) analyzed from at least two to three different experiments/culture preparations.
Characterization of N-Terminal GFP-CB1 Receptor Chimeras. The N-terminally tagged GFP-CB1 functioned normally when expressed in HEK293 cells, with EC50 values for inhibition of forskolin-stimulated cAMP production by the classic agonist HU210 in HEK293 cells similar to those of the wild-type CB1 receptor, being 2.16 nM (confidence interval 1.28-3.63 nM) for wild-type CB1 and 0.712 nM (confidence interval = 0.23-2.17) for GFP-CB1. In addition, this construct exhibited agonist-induced internalization, as measured by cell-surface enzyme-linked immunosorbent assay when expressed in HEK293 cells. Treatment with HU210 (1 µM; n = 9) for 2 h significantly reduced cell-surface expression of GFP-CB1 compared with vehicle-treated cells (41.5 ± 5.6% of vehicle; n = 7; unpaired t test, two-tailed test; P < 0.001).
Distribution of GFP-CB1 in Cultured Hippocampal Neurons. Next, the functional GFP-CB1 fusion-protein was transfected into cultured hippocampal neurons (7+ DIV), and its cellular localization was examined in live cells after 16 to 24 h. GFP-CB1 fluorescence was distributed throughout the cell soma and processes (Fig. 1A) but was excluded from the nucleus with somatodendritic regions exhibiting punctate, vesicular fluorescence. It is noteworthy that axonal fluorescence was less punctate, being present throughout axonal shafts and at growth cones. Time-lapse recordings of live, cultured hippocampal neurons expressing GFP-CB1 demonstrated bidirectional movement of dendritic puncta, showing that they represent mobile, dynamic structures (Fig. 1B). These data suggest that the CB1 receptor is not actively excluded from the somatodendritic compartment through selective sorting into axon-specific cargo vesicles.
Fig. 1. Surface expression of GFP-CB1 is restricted to the axon when expressed in cultured hippocampal neurons. Aa, representative Z-projection image of a live, cultured hippocampal neuron (9 DIV) expressing the N-terminally tagged GFP-CB1 construct, the axon is indicated with an arrow. Zoomed images comparing punctate fluorescence within dendrites (Ab) with more uniform axonal labeling and association with filopodia are illustrated (Ac) Scale bar, 20 µm. B, time lapse recordings of an area of dendrites (confocal section) from a cultured hippocampal neuron (9 DIV) expressing GFP-CB1. Numbers refer to the time in minutes. Note that puncta are bidirectionally mobile. Scale bar, 2 µm. C, representative Z-projection of a fixed hippocampal neuron (9 DIV) expressing GFP-CB1 (Ca) and probed for surface expression of GFP (Cb) and intracellular MAP2 (Cc). Note that surface expression of GFP-CB1 is restricted to the MAP2-negative axon and is more uniformly distributed. Cd, merged color image illustrating GFP-CB1 fluorescence (green), surface expression (red), and MAP2 labeling (blue). Scale bars, 20 µm.
Cell-Surface Polarity of GFP-CB1 Is Retained in Primary Hippocampal Culture. Cell-surface expression of GFP-CB1 was assessed using an antibody directed against GFP and was compared with labeling for the somatodendritic marker MAP2 after permeabilization (Fig. 1C). Although GFP-CB1 fluorescence was clearly present in the cell soma and throughout the full extent of MAP2-positive processes, surface labeling was restricted to the MAP2-negative axon. The polarity index (D/A ratio) was 0.074 ± 0.017 (n = 12), reflecting a distribution that is highly polarized toward the axon.
Targeting of GFP-CB1 to the Axon Is Not Dependent on Hippocampal Neuron Subtype. In the hippocampus, the CB1 receptor is endogenously expressed on the majority of GABAergic interneurons (Irving et al., 2000). Cell-surface labeling for GFP-CB1 was detected in transfected mature, cultured hippocampal neurons, and interneurons were subsequently identified by GAD65 expression. The extent of cell-surface polarization of GFP-CB1 expressed in GABAergic and non-GABAergic neurons was not significantly different (p > 0.05) and is therefore not dependent on hippocampal neuron subtype (Fig. 2).
Fig. 2. Surface expression of GFP-CB1 is restricted to the axon when expressed in both GABAergic and non-GABAergic cultured hippocampal neurons. A, surface expression of GFP-CB1 is restricted to the axon when expressed in GABAergic cultured hippocampal neurons. Representative Z-projection image of a GABAergic hippocampal neuron (8 DIV) expressing GFP-CB1 (Aa) and probed for surface expression of GFP (Ab) and GAD65 (Ac). Merged image (Ad) illustrating GFP-CB1 fluorescence (green), surface expression (red), and GAD65 labeling (blue). Scale bar, 20 µm. B, representative Z-projection of a non-GABAergic hippocampal neuron (8 DIV) expressing GFP-CB1 (Ba) and probed for surface expression of GFP (Bb) and GAD65 (Bc). Merged image (Bd) illustrating GFP-CB1 fluorescence (green), surface expression (red), and GAD65 labeling (blue). Scale bar, 20 µm. C, the polarity index of dendritic to axonal surface expression (D/A ratio) of GFP-CB1 was quantified for GABAergic and non-GABAergic neurons expressing the construct. A D/A ratio of <1 reflects a cell-surface distribution that is polarized toward the axon, and a ratio >1 reflects a cell-surface distribution that is polarized toward the dendrites. There was no significant difference (Unpaired t test, two-tailed test; P > 0.05).
Cotransfection with the Dominant-Negative Dynamin-1 K44A Mutant Leads to a Loss of CB1 Receptor Cell-Surface Polarity. Previous studies have suggested that the axonal polarization of a C-terminally tagged, fluorescent CB1 receptor chimera involves dynamin-dependent endocytosis (Leterrier et al., 2006). We have therefore investigated whether this is also the case with our N-terminally tagged GFP-CB1 construct and endogenous CB1 receptors using a dominant-negative dynamin-1 mutant (K44A), which inhibits clathrin-dependent endocytosis (Wilbanks et al., 2002).
Mature, cultured hippocampal neurons were cotransfected with GFP-CB1 and HA-tagged dominant-negative dynamin-1 or wild-type dynamin-1. After 16 to 24 h of expression, the neurons were fixed and the surface distribution of GFP-CB1 was determined. Cotransfection with HA-tagged dynamin-1 did not alter GFP-CB1 cell-surface targeting (Fig. 3A). It is striking that, in dominant-negative dynamin-1 (K44A) cotransfected neurons, cell-surface expression of GFP-CB1 was polarized toward the somatodendritic compartment, although axonal surface expression was still observed (Fig. 3B). The polarity index for GFP-CB1 in neurons cotransfected with dominant-negative dynamin-1 was significantly different from control neurons (P < 0.001), whereas no difference was observed with neurons cotransfected with HA-tagged dynamin-1 (P > 0.05; Fig. 3C).
Fig. 3. Preferential endocytosis in the somatodendritic compartment leads to an axonally polarized cell-surface distribution of the CB1 receptor in cultured hippocampal neurons. A, representative Z-projection of a hippocampal neuron (8 DIV) expressing GFP-CB1 and HA-tagged, wild-type dynamin-1 (dyn) (Aa) and probed for surface expression of GFP (Ab) and HA-dynamin-1 (Ac). Merged image (Ad) with GFP-CB1 fluorescence (green), surface expression (red), and HA-dynamin-1 (blue). Note that overexpression of HA-dynamin-1 does not affect axonal polarity. B, cotransfection with an HA-tagged, dominant-negative dynamin-1 (K44A) construct (2:1 ratio) leads to cell-surface expression of GFP-CB1 on the somatodendritic compartment. Representative Z-projection of hippocampal neuron (8 DIV) expressing GFP-CB1 (Ba) and probed for surface expression of GFP (Bb) and dominant-negative dynamin-1 (Bc). Note that surface GFP-CB1 fluorescence is present throughout the neuron. Merged image (Bd) with GFP-CB1 (green), surface expression (red), and HA-dynamin-1 (K44A) (blue). C, the relative intensity of dendritic to axonal surface expression (D/A ratio) of GFP-CB1 was quantified in neurons expressing GFP-CB1 alone or GFP-CB1 cotransfected with either HA-tagged functional or HA-dynamin-1 (K44A). No significant difference was observed between the control and wild-type dynamin-1 (HA-Dyn) cotransfected groups. Cotransfection with the dominant-negative dynamin-1 (HA-Dyn K44A) construct leads to a marked change in the cell-surface polarity of GFP-CB1 [P < 0.001; one-way analysis of variance (ANOVA) with Bonferroni post hoc test]. D, cultured hippocampal neurons (3 DIV) were transfected with dominant-negative dynamin-1 and allowed to express for 48 h before detecting surface expression of endogenous CB1 receptor at 5 DIV with a N-terminal antibody. A representative Z-projection of a CB1 positive neuron is illustrated. Da, membrane-associated CB1 surface labeling on the soma and dendrites. Db, labeling for the HA-dynamin-1 (K44A). Dc, merged image with CB1 receptor surface labeling (green) and HA-dynamin-1 (K44A) (blue). Scale bars, 20 µm.
Next, we investigated the effect of inhibiting endocytosis in neurons expressing endogenous CB1 receptors. Young cultures of hippocampal neurons (3 DIV, when CB1 receptor cell-surface expression is first observed; Irving et al., 2000) were transfected with the HA-tagged dominant-negative dynamin-1 (K44A) and allowed to express for 48 h before being fixed at 5 DIV and probed for endogenous CB1 receptor cell-surface expression. In neurons transfected with the dynamin-1 mutant and endogenously expressing the CB1 receptor, membrane-associated, cell-surface immunoreactivity for the CB1 receptor was observed on the cell soma and multiple neurites (Fig. 3D).
Constitutive Activity Does Not Drive CB1 Receptor Axonal Polarization in Hippocampal Neurons. In some systems, the CB1 receptor displays constitutive activity (Pertwee, 2005; D'Antona et al., 2006). This is suggested to drive domain-specific endocytosis in cultured hippocampal neurons, in which treatment with an antagonist/inverse agonist alters CB1 receptor polarity and is associated with an up-regulation of somatodendritic cell-surface expression (Leterrier et al., 2006). We therefore tested for the involvement of constitutive activity in driving GFP-CB1 receptor axonal targeting using two approaches: either treatment with CB1 receptor antagonists with inverse agonist properties (AM281 and SR1417167A), or expression of recombinant mutant receptors. CB1 receptors with a D164N point mutation or a 14 amino acid C-terminal truncation (14) do not undergo agonist-induced internalization (Hsieh et al., 1999; Roche et al., 1999), and the D164N mutant also lacks constitutive activity (Nie and Lewis 2001).
Mature, cultured hippocampal neurons were transfected with GFP-CB1 and allowed to express for 24 h in the presence of AM281 (1-10 µM) or SR141716A (1 µM) or equivalent vehicle-control (1:1000 DMSO). In both vehicle-control and antagonist-treated neurons, cell-surface expression of GFP-CB1 was still restricted to the axon (Fig. 4A). In addition, shorter periods of exposure (3 h) to a high concentration of antagonists AM281 (10 µM) or SR141716A (10 µM) after normal expression of GFP-CB1 also did not alter polarity (Fig. 4D).
Fig. 4. Constitutive activity is not required for removal of the CB1 receptor from the somatodendritic cell-surface. A, antagonist treatment does not affect the cell-surface polarity of GFP-CB1 when it is expressed in cultured hippocampal neurons. Representative Z-projection images of neurons (8 DIV) transfected with GFP-CB1 and allowed to express in the presence of either vehicle (1: 1000 DMSO; A, a, b, and c) or 10 µM AM281 (A, d, e, and f). GFP-CB1 fluorescence (A, a and d) is present throughout the neuron, but surface expression of GFP-CB1 (A, b and e) is restricted to the axon in both treatments. Merged images (A, c and f) with GFP-CB1 (green) and surface labeling (red) B, agonist-induced trafficking of mutant constructs was assessed in COS7 cells. Cells expressing GFP-CB1 (B, a and b) GFP-CB1(D164N) (B, c and d) GFP-CB1(14) (B, e and f) were exposed to Win55212-2 (100 nM) for 90 min. Note that agonist-induced trafficking was only observed in cells transfected with GFP-CB1 (arrow). B and C, the cell-surface polarity of GFP-CB1 mutants was assessed in hippocampal neurons. A representative Z-projection image of a neuron (8 DIV) transfected with GFP-CB1 (D164N) (C, a-c) GFP-CB1 (14) (C, d-f) and probed for surface expression (C, b and e). Merged image (C, c and f) illustrating GFP-CB1 (green) and surface labeling (red). Note that in these studies, neurons were probed for surface expression using an antibody directed against GFP before fixation, which resulted in a small degree of receptor clustering. However, this labeling was still restricted to the axon. Scale bars, 20 µm. D and E, the D/A ratio for surface GFP-CB1 was quantified for different antagonist treatment protocols (D) and GFP-CB1 mutants (E). No significant difference in the D/A ratio for any of the antagonist treatments or mutant receptors was observed (P > 0.05; one-way ANOVA with Bonferroni post hoc test). Scale bars, 20 µm for neurons (A and C) and 50 µm for COS7 cells (B).
Next, we studied the effects of mutant constructs lacking agonist-induced internalization/constitutive activity. In control experiments with GFP-CB1 expressed in COS7 cells, Win55212-2 treatment resulted in the appearance of pronounced perinuclear, vesicular fluorescence within the cytoplasm after 90-min exposure (30°C), which is consistent with agonist-induced trafficking (Hsieh et al., 1999). However, no trafficking was observed in cells expressing CB1(D164N) or the GFP-CB1 (14) mutants after 90-min exposure to 100 nM Win55212-2 (Fig. 4B). When expressed in neurons, the mutant constructs GFP-CB1(D164N) (n = 8) and GFP-CB1(14) (n = 11) also exhibited normal axonal targeting with respect to GFP-CB1 (Fig. 4C). Thus, the polarity indexes for GFP-CB1 in vehicle and AM281-treated cells and for the GFP-CB1 mutants were not significantly different (P > 0.05; Fig. 4E).
Treatment with CB1 receptor antagonists also had no effect on the axonal polarization of endogenous CB1 receptors detected using an N-terminal antibody. Exposure of cultures to AM281 (10 µM; Fig. 5) or SR141716A (1 µM; Fig. 5) for 17 to 18 h did not result in a significant up-regulation of endogenous somatodendritic CB1 receptor cell-surface expression, and CB1 receptor surface immunoreactivity remained highly polarized toward the axon. Prolonged treatment with agonist (Win55212-2; 17 h) resulted in the appearance of pronounced punctate labeling within somatodendritic regions (Coutts et al., 2001) and a marked reduction in cell-surface labeling within the axon and dendrites (Fig. 5D). This was associated with a small increase in the polarity index, reducing axonal surface labeling to a slightly greater extent than soma/dendrites (Fig. 5F). The observation that weak dendritic cell-surface labeling was also reduced in the presence of Win55212-2 suggests that low levels of surface receptor expression can be detected within this compartment. All of the effects of Win55212-2 were prevented in the presence of AM281 (10 µM) (Fig. 5).
Fig. 5. Antagonist treatment did not alter surface expression of endogenous CB1 receptors. Representative confocal images (Z-projections) of CB1 receptor-positive neurons from a culture treated for 17 h with vehicle, (A, a, e, and i; n = 16), Win55212-2 (100 nM; A, b, f, and j; n = 12), Win55212-2 (100 nM) and AM281 (10 µM) (A, c, g, and k; n = 16) or AM281 (10 µM; A, d, h, and l; n = 9) alone. Cultures were then probed for CB1 receptor surface expression (A, e, f, g, and h) and then for additional CB1 receptor labeling after fixation and permeabilization (Total CB1; A, a, b, c, and d). Merged images illustrate CB1 receptor surface expression (green) and total CB1 (green) (A, i, j, k, and l). Scale bars, 20 µm. Cell-surface CB1 expression levels were quantified for axons (B) and dendrites (C) and D/A ratios for surface endogenous CB1 (D) were determined for the different treatment protocols. No significant difference in the D/A ratio was observed with AM281 (P > 0.05); however, treatment with Win55212-2 decreased surface expression on dendritic regions and on the axon, resulting in a slight change in polarity toward soma and dendrites (P < 0.05). E, D/A ratios for surface endogenous CB1 were quantified after 18-h exposure to SR141716A (SR; 1 µM) relative to vehicle control. No significant difference in the D/A ratio was observed (P > 0.05; one-way ANOVA with Bonferroni post hoc test).
In cultured hippocampal neurons, cell-surface immunoreactivity for the CB1 receptor is markedly polarized toward the axon (Irving et al., 2000; Leterrier et al., 2006). In this study, we used an N-terminally tagged GFP-CB1 receptor fusion-protein together with immunohistochemically labeled endogenous CB1 receptors to investigate the mechanisms underlying CB1 receptor axonal targeting. We find that CB1 receptors are delivered to both axonal and somatodendritic plasma membranes but that surface polarity is achieved by selective, constitutive endocytosis/basal receptor removal. However, this process does not require constitutive receptor activity.
Axonal Targeting. As with endogenous CB1 receptors, the cell-surface expression of recombinant GFP-CB1 is highly polarized toward the axon when expressed in mature, cultured hippocampal neurons. Although hippocampal cultures contain a mixed population of cells, endogenous CB1 receptors are expressed at high levels within a subset of GABAergic interneurons (Irving et al., 2000). However, the cell-surface expression of GFP-CB1 was independent of the subtype of hippocampal neuron expressing the receptor. This is consistent with the receptor containing axonal-surface targeting information and the conservation of trafficking pathways between the different hippocampal cell types.
Preferential Endocytosis of the CB1 Receptor in the Somatodendritic Compartment Drives Its Functional Polarization. Intracellular GFP-CB1 receptors were observed throughout the neuron, suggesting that active exclusion of the CB1 receptor from the dendritic compartment does not occur. Consistent with recent work using hippocampal cultures (Leterrier et al., 2006), preferential endocytosis of the CB1 receptor from the somatodendritic plasma-membrane underlies axonally polarized cell-surface CB1 expression. This was demonstrated using transfection with a dominant-negative dynamin-1 mutant, which competitively inhibits clathrin-mediated endocytosis. It is noteworthy that endogenous dynamin levels did not seem to be limiting in the somatodendritic endocytosis of the CB1 receptor, because overexpression of wild-type dynamin-1 did not decrease the polarity index of GFP-CB1. In addition to demonstrating a change in recombinant surface receptor expression by inhibiting endocytosis, we also demonstrated this effect for native CB1 receptors expressed in interneurons. A similar role of endocytosis in the axonal cell-surface polarization of vesicle-associated membrane protein 2 (Sampo et al., 2003) and neuroglia cell adhesion molecule (Sampo et al., 2003; Wisco et al., 2003) has been proposed. It is likely that proteins polarized to the axon will, at least initially, also be trafficked to the soma and dendrites (Stowell and Craig, 1999), and domain-specific endocytosis seems to be increasingly recognized as important in limiting inappropriate surface expression of axonal proteins within this region (Wisco et al., 2003).
The precise mechanisms that underlie the preferential endocytosis of the CB1 receptor within the somatodendritic compartment are unclear at present. Differences in the internalization machinery may play a role; for example, the expression of dynamin subtypes varies between axonal and somatodendritic compartments (Gray et al., 2003). Endocytosis within axons may also be spatially restricted to synapses, with nonsynaptic CB1 receptors more resistant to this (Leterrier et al., 2006). Alternatively, specific anchoring proteins present in axons may bind CB1 receptors and stabilize them within the plasma membrane. Studies in polarized cell model systems suggest that anchoring to scaffolding proteins and/or the cytoskeleton underlies the retention of membrane proteins at specialized plasma membrane domains. For example, the integrity of the actin cytoskeleton is required for the retention of resident apical proteins at the apical plasma membrane in polarized WIF-B hepatic cells, and it is proposed that actin-based scaffolds actively exclude these proteins from endocytosis (Tuma et al., 2002).
Constitutive Activity Is Not a Prerequisite for Axonal Targeting. Previous studies in HEK293 cells suggest that the CB1 receptor exhibits constitutive activity, which leads to a pronounced intracellular localization under control conditions by driving receptor endocytosis (Leterrier et al., 2004; D'Antona et al., 2006). Leterrier et al. (2006) have suggested that this is analogous to endocytosis of the CB1 receptor within the somatodendritic compartment of hippocampal neurons. However, in contrast to this, we find that constitutive activation of the CB1 receptor is not required for its axonal polarization. Constitutive activity is generally believed to reflect the periodic spontaneous entry of the receptor into its active conformation in the absence of agonist (Bond and IJzerman, 2006). In certain systems, tonic activation can drive constitutive endocytosis of the CB1 receptor via a mechanistically similar process to agonist-induced internalization (Leterrier et al., 2004; D'Antona et al., 2006). However, the CB1 receptor does not exhibit constitutive activity in all systems, and an increasing number of studies suggest that it may not be tonically active in central neurons (Coutts et al., 2001; Savinainen et al., 2003; Hentges et al., 2005; Zhu and Lovinger, 2005; Neu et al., 2006), and this may reflect the expression of modulatory proteins that block this activity.
It was proposed by Leterrier et al. (2006) that the CB1 receptor antagonist/inverse agonist AM281 can change the polarity of CB1 expression by trapping constitutively active receptors at the somatodendritic plasma membrane. However, we saw no evidence for this effect with either AM281 or SR141716A, even at high concentrations (10 µM) at which inverse agonism will be more apparent (Pertwee, 2005). Indeed, we have reported previously that overnight treatment with SR141716A does not up-regulate native CB1 receptor expression, and the cellular distribution of surface receptors remained unchanged (Coutts et al., 2001). This observation was corroborated in the present study for both endogenous and recombinant CB1 receptors. Because incubation with CB1 receptor antagonists does not disrupt the CB1 receptor cell-surface polarity, this also suggests that it is not dependent on the basal release of endocannabinoids. Further evidence for a mechanism distinct from constitutive activity comes from our findings with mutant chimeras that are reported to prevent agonist-induced endocytosis (Hsieh et al., 1999; Roche et al., 1999) and constitutive activity (D164N; Roche et al., 1999; Nie and Lewis, 2001) but do not affect CB1 receptor surface polarity.
Our data are consistent with a model in which the constitutive endocytic removal of the CB1 receptor from the somatodendritic plasma membrane occurs through a cellular process distinct from constitutive activation-driven endocytosis. As such, this is likely to involve motifs/conformational states different from those used by agonist-induced internalization and may also involve different endocytic proteins. The significant down-regulation of dendritic CB1 cell-surface expression by agonist in the absence of a reciprocal up-regulation by antagonist further suggests that the receptor is constitutively removed from the plasma membrane independently of receptor activation. It is interesting that differences in structural or conformational requirements for constitutive and agonist-promoted endocytosis have been identified in other GPCR systems (Whistler et al., 2002; Waldhoer et al., 2003).
The marked contrasts between the present study and that of Leterrier et al. (2006) are probably multifactorial in their explanation. For example, truncations of the C-terminal tail of the CB1 receptor increases constitutive activity (Nie and Lewis. 2001), and tagging the CB1 receptor at its C terminus (Leterrier et al., 2004) might prevent this effect, leading to higher levels of constitutive activity. Differences in the developmental age of the cultures or culture conditions (including endocannabinoid contaminants or by inducing their generation) could also influence the apparent expression of constitutively precoupled receptors.
Our data are consistent with a revised model in which the CB1 receptor is delivered to the cell-surface membrane in both the axonal and somatodendritic compartments but is removed from the somatodendritic plasma membrane by constitutive endocytosis, which is a process that is distinct from agonist-driven endocytosis. Moreover, constitutive activation of the receptor does not drive endocytosis in cultured hippocampal neurons or influence axonal polarity. Antagonists of the CB1 receptor are of considerable therapeutic interest, and treatment is likely to be long-term (Van Gaal et al., 2005). Our findings suggest that such compounds are unlikely to lead to a marked redistribution of CB1 receptors from their correct functional sites within the brain.
ABBREVIATIONS: CB1, cannabinoid receptor subtype-1; GPCR, G protein-coupled receptor; GFP, enhanced green fluorescent protein; MAP2, microtubule-associated protein 2; GAD65, glutamate decarboxylase type 65; AM281, 1-(2,4-dichlorophenyl)-5-(4-iodophenyl)-4-methyl-N-4-morpholinyl-1H-pyrazole-3-carboxamide; Win55212-2, (R)-(+)-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl) pyrrolo[1,2,3-de]-1,4-benzoxazin-6-yl]-1-naphthalenylmethanone mesylate; HU210, (6aR)-trans-3-(1,1-dimethylheptyl)-6a,7,10,10a-tetrahydro-1-hydroxy-6,6-dimethyl-6H-dibenzo[b,d]pyran-9-methanol; SR141716A, 5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-N-(piperidin-1-yl)-1H-pyrazole-3-carboxamide; aa, amino acids; HA, hemagglutinin; DMSO, dimethyl sulfoxide; SS, signal sequence; DIV, days in vitro; PBS, phosphate-buffered saline; ANOVA, analysis of variance; D/A, dendritic/axonal; HEK, human embryonic kidney.
【参考文献】
Bond RA and IJzerman AP (2006) Recent developments in constitutive receptor activity and inverse agonism, and their potential for GPCR drug discovery. Trends Pharmacol Sci 27: 92-96.
Connolly CN, Futter CE, Hopkins CR, and Cutler D (1994) Transport into and out of the Golgi complex studied by transfecting cells with cDNAs encoding horseradish peroxidase. J Cell Biol 1 27: 641-652.
Coutts AA, Anavi-Goffer S, Ross RA, MacEwan DJ, Mackie K, Pertwee RG, and Irving AJ (2001) Agonist-induced internalization and trafficking of cannabinoid CB1 receptors in hippocampal neurons. J Neurosci 21: 2425-2433.[Abstract/Free Full Text]
D'Antona AM, Ahn KH, and Kendall DA (2006) Mutations of CB1 T210 produce active and inactive receptor forms: correlations with ligand affinity, receptor stability, and cellular localization. Biochemistry 45: 5606-56517.
Freund TF, Katona I, and Piomelli D (2003) Role of endogenous cannabinoids in synaptic signaling. Physiol Rev 83: 1017-1066.[Abstract/Free Full Text]
Gray NW, Fourgeaud L, Huang B, Chen J, Cao H, Oswald BJ, Hemar A, and McNiven MA (2003) Dynamin 3 is a component of the postsynapse, where it interacts with mGluR5 and Homer. Curr Biol 13: 510-515.
Hentges ST, Low MJ, and Williams JT (2005) Differential regulation of synaptic inputs by constitutively released endocannabinoids and exogenous cannabinoids. J Neurosci 25: 9746-9751.[Abstract/Free Full Text]
Hsieh C, Brown S, Derleth C, and Mackie K (1999) Internalization and recycling of the CB1 cannabinoid receptor. J Neurochem 73: 493-501.
Irving AJ, Coutts AA, Harvey J, Rae MG, Mackie K, Bewick GS, and Pertwee RG (2000) Functional expression of cell surface cannabinoid CB1 receptors on presynaptic inhibitory terminals in cultured rat hippocampal neurons. Neuroscience 98: 253-262.
Jiang M, Deng L, and Chen G (2004) High Ca2+-phosphate transfection efficiency enables single neuron gene analysis. Gene Ther 11: 1303-1311.
Leterrier C, Bonnard D, Carrel D, Rossier J, and Lenkei Z (2004) Constitutive endocytic cycle of the CB1 cannabinoid receptor. J Biol Chem 279: 36013-36021.[Abstract/Free Full Text]
Leterrier C, Laine J, Darmon M, Boudin H, Rossier J, and Lenkei Z (2006) Constitutive activation drives compartment-selective endocytosis and axonal targeting of type 1 cannabinoid receptors. J Neurosci 26: 3141-3153.[Abstract/Free Full Text]
Neu A, Foldy C, and Soltesz I (2006) Postsynaptic origin of CB1-dependent tonic inhibition of GABA release at CCK-positive basket cell to pyramidal cell synapses in the CA1 region of the rat hippocampus. J Physiol 578: 233-247.
Nie J and Lewis DL (2001) Structural domains of the CB1 cannabinoid receptor that contribute to constitutive activity and G-protein sequestration. J Neurosci 21: 8758-8764.[Abstract/Free Full Text]
Pertwee RG (2005) Inverse agonism and neutral antagonism at cannabinoid CB1 receptors. Life Sci 76: 1307-1324.
Roche JP, Bounds S, Brown S, and Mackie K (1999) A mutation in the second transmembrane region of the CB1 receptor selectively disrupts G protein signaling and prevents receptor internalization. Mol Pharmacol 56: 611-618.[Abstract/Free Full Text]
Savinainen JR, Saario SM, Niemi R, Jarvinen T, and Laitinen JT (2003) An optimized approach to study endocannabinoid signaling: evidence against constitutive activity of rat brain adenosine A1 and cannabinoid CB1 receptors. Br J Pharmacol 140: 1451-1459.
Sampo B, Kaech S, Kunz S, and Banker G (2003) Two distinct mechanisms target membrane proteins to the axonal surface. Neuron 37: 611-624.
Stowell JN and Craig AM (1999) Axon/dendrite targeting of metabotropic glutamate receptors by their cytoplasmic carboxy-terminal domains. Neuron 22: 525-536.
Tuma PL, Nyasae LK, and Hubbard AL (2002) Nonpolarized cells selectively sort apical proteins from cell surface to a novel compartment, but lack apical retention mechanisms. Mol Biol Cell 13: 3400-3415.[Abstract/Free Full Text]
Van Gaal LF, Rissanen AM, Scheen AJ, Ziegler O, Rossner S, and the RIO-Europe Study Group (2005) Effects of the cannabinoid-1 receptor blocker rimonabant on weight reduction and cardiovascular risk factors in overweight patients: 1-year experience from the RIO-Europe study. Lancet 365: 1389-1397.
Waldhoer M, Casarosa P, Rosenkilde MM, Smit MJ, Leurs R, Whistler JL, and Schwartz TW (2003) The carboxyl terminus of human cytomegalovirus-encoded 7 transmembrane receptor US28 camouflages agonism by mediating constitutive endocytosis. J Biol Chem 278: 19473-19482.[Abstract/Free Full Text]
Whistler JL, Gerber BO, Meng EC, Baranski TJ, von Zastrow M, and Bourne HR (2002) Constitutive activation and endocytosis of the complement factor 5a receptor: evidence for multiple activated conformations of a G protein-coupled receptor. Traffic 3: 866-877.
Wilbanks AM, Laporte SA, Bohn LM, Barak LS, and Caron MG (2002) Apparent loss-of-function mutant GPCRs revealed as constitutively desensitized receptors. Biochemistry 41: 11981-11989.
Wisco D, Anderson ED, Chang MC, Norden C, Boiko T, Folsch H, and Winckler B (2003) Uncovering multiple axonal targeting pathways in hippocampal neurons. J Cell Biol 162: 1317-1328.[Abstract/Free Full Text]
Zhu PJ and Lovinger DM (2005) Retrograde endocannabinoid signaling in a postsynaptic neuron/synaptic bouton preparation from basolateral amygdala. J Neurosci 25: 6199-6207.[Abstract/Free Full Text]
作者单位：Neurosciences Institute, Division of Pathology & Neuroscience, Ninewells Hospital & Medical School, University of Dundee, Dundee, Scotland, United Kingdom