Truth Seeker
New Member
Abstract
Cannabinoids activate several members of the mitogen-activated protein kinase superfamily including p44 and p42 extracellular signal-regulated kinase (ERK). We used N1E-115 neuroblastoma cells and the cannabinoid receptor agonist WIN 55,212-2 (WIN) to examine the signal transduction pathways leading to the activation of ERK. ERK phosphorylation (activation) was measured by Western blot. The EC50 for stimulation of ERK phosphorylation was 10 nM, and this effect was blocked by pertussis toxin and the CB1 (cannabinoid) receptor antagonist SR141716A. The MEK inhibitors PD 98059 and U0126 blocked ERK phosphorylation, as did the adenylate cyclase activator forskolin. The phosphatidylinositol (PI) 3-kinase inhibitor LY 294002 and the Src kinase inhibitor PP2 partially occluded the response but also decreased basal levels of phospho-ERK. The PI 3-kinase and Src pathways are known to promote cell survival in many systems; therefore, MTT (1-(4,5-dimethylthiazol-2-yl)-3,5-diphenylformazan) conversion was used to examine the effects of these inhibitors on cellular viability. LY 294002 decreased the number of viable cells after 18 h of treatment; therefore, the inhibition of ERK by this inhibitor is probably because of cytotoxicity. Forskolin blocked ERK phosphorylation with an EC50 of <3 μM, and the protein kinase A (PKA) inhibitor H-89 enhanced ERK phosphorylation. c-Raf phosphorylation at an inhibitory PKA-regulated site (Ser259) was also reduced by WIN. This is probably due to constitutive phosphatase activity because WIN did not directly stimulate PP1 or PP2A activity when measured using 6,8-difluoro-4-methylumbelliferyl phosphate as a fluorogenic substrate. These data implicate the inhibition of PKA as the predominant pathway for ERK activation by CB1 receptors in N1E-115 cells. PI 3-kinase and Src appear to contribute to ERK activation by maintaining activation of kinases, which prime the pathway and maintain cellular viability.
Although cannabinoids have been used both medicinally and recreationally for thousands of years, only recently have the receptors and signaling cascades responsible for the physiological effects of these drugs and the endogenous brain cannabinoids been elucidated. This is primarily because the receptors were only identified recently and selective potent agonists and antagonists have only been available for a short time. Three endogenous lipid-derived ligands (endocannabinoids) have been characterized to date. These include anandamide (1), noladin ether (2), and 2-arachidonoyl glycerol (3), although there are probably more endocannabinoids yet to be characterized. Two cannabinoid receptors (CB1 and CB2)1 have been cloned to date (4, 5), whereas a third subtype has been characterized in the vasculature and in the brains of CB1 receptor knock-out mice (6, 7). Both CB1 and CB2 couple to pertussis toxin-sensitive G-proteins (8), although non-receptor-mediated forms of cannabinoid signaling have been reported previously (9). Cannabinoids and endocannabinoids regulate synaptic transmission at both excitatory and inhibitory synapses and participate in long term plasticity (reviewed in Ref. 10). Recent studies suggest that cannabinoids and cannabinoid receptor antagonists may also have therapeutic potential in disorders including Parkinson's disease, multiple sclerosis, human immunodeficiency virus, and stroke (reviewed in Ref. 11).
Inhibition of adenylate cyclase (AC) through Gi coupling was the first intracellular signal transduction pathway implicated in CB1 receptor signaling (8). Regulation of AC can occur through both α and βγ subunits depending on the isoforms of AC expressed by the cells (12, 13). Cannabinoids increase cAMP production in CHO cells expressing AC II, IV, or VII through βγ subunits but decrease activity in cells expressing AC I, V, VI, and VIII (14). In addition, cannabinoids have been linked to sphingomyelin hydrolysis and ceramide generation in glial and glioma cells (15). Cannabinoids also inhibit calcium channel function (16, 17) and increase K+ conductance (18), which can decrease neuronal excitability and may contribute to synaptic plasticity.
The mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) pathway is also activated by cannabinoids (19). ERK regulates proliferation, differentiation, and synaptic plasticity and is activated through a myriad of diverse signals. Growth factors frequently use tyrosine kinases to signal, resulting in Ras-dependent activation of the Ser/Thr kinase Raf (20). G-protein-coupled receptors can "hijack" this tyrosine kinase pathway by protease-mediated transactivation (21, 22) or may employ G-protein-stimulated protein kinases to directly phosphorylate signaling intermediates, including Shc and Raf. G-protein βγ-subunits can also regulate ERK through arrestin and Src-dependent pathways (23). In addition, βγ subunits may activate ERK through PI 3-kinase as a signaling intermediate. This Src-independent pathway was recently described in human astrocytoma cells stimulated with CB1 receptor agonists (24).
Raf represents a point of convergence for multiple signaling pathways that activate ERK. Raf phosphorylates the dualspecificity kinase MEK, which in turn phosphorylates ERK at tyrosine and threonine residues in the activation domain (25). Some of these effects appear to be cell type-specific. cAMP activates ERK in cells expressing B-Raf (26) while inhibiting ERK activation in cells expressing c-Raf/Raf-1 (27, 28). In addition to PKA, PKC, Src, AKT, casein kinase I, and p21-activated kinase are also capable of phosphorylating Raf (29). Although the contribution of each of the multiple phosphorylations is not completely understood, dephosphorylation of the PKA site Ser259 appears to be absolutely required for c-Raf activation (30—33). CB1 receptors couple negatively to AC, which may mediate ERK activation in cells expressing c-Raf (e.g. relief of tonic inhibition produced by basal PKA activity).
The present experiments were performed to determine the intracellular signaling pathways responsible for ERK activation by cannabinoids in N1E-115 mouse neuroblastoma cells. Pharmacological and biochemical approaches were used to examine the role of PKA, PI 3-kinase, Src, and PKC in CB1-stimulated ERK phosphorylation. These data indicate that multiple signals generated by constitutively activated kinases and phosphatases contribute to CB1-stimulated ERK phosphorylation in N1E-115 neuroblastoma cells. Basal activation of Src and PI 3-kinase contribute to ERK activation, but adenylate cyclase activation completely abolishes CB1 signaling to ERK. These data implicate inhibition of adenylate cyclase and dephosphorylation of c-Raf in ERK activation by cannabinoids in cells of neural origin.
EXPERIMENTAL PROCEEDURES
Reagents–Anti-phospho ERK, anti-phospho AKT, and anti-phospho-c-Raf Ser259 were obtained from Cell Signaling (Beverly, MA). Polyclonal anti-ERK and horseradish peroxidase anti-rabbit were purchased from Sigma. WIN 55,212-2, DPDPE, PP2, and GF 109203X were purchased from Tocris (Avonmouth, Bristol, United Kingdom). SR141716A was from the National Institute of Mental Health chemical synthesis program. Protease inhibitors, phosphatase inhibitors, MTT, and phorbol 12-myristate 13-acetate were obtained from Sigma. Forskolin, PD 98059, and U0126 were purchased from Calbiochem. Pertussis toxin was purchased from List Biological Laboratories (Campbell, CA). Cell culture reagents were obtained from Invitrogen.
Cell Culture–N1E-115 cells (ATCC CRL-2263) were maintained at 37 °C, 5% CO2 in Dulbecco's modified Eagle's medium (high glucose with pyridoxine-HCl and sodium pyruvate, Invitrogen catalogue 10313) supplemented with 2 mM L-glutamine, 10% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin. Cells were subcultured at weekly intervals.
Kinase Activation Assays–Cells (105) were plated in 24-well cell culture dishes, incubated for 24—48 h, and serum-starved for 12—24 h prior to assay. Cells were treated with WIN in Dulbecco's modified Eagle's medium containing 0.1% bovine serum albumin as a carrier and incubated at 37 °C, 5% CO2 for 5—60 min. In experiments where antagonists were used, cells were pretreated with the antagonist for 5 min prior to stimulation. Cells were pretreated for 30 min with LY 294002, PP2, U0126, PD 98059, forskolin, and GF 109203X. For experiments with pertussis toxin inactivation of Gi/o, cells were treated for 12 h with 100 ng/ml pertussis toxin in complete medium. Cells were then serumstarved in the presence of pertussis toxin for an additional 12 h. All of the stock solutions were prepared in Me2SO, and dilutions were prepared with Me2SO concentrations kept constant (0.1—0.2%). The assay was terminated by aspirating the medium, by placing the cells on dry ice, and with the addition of lysis buffer (phosphatase inhibitor cocktails I and II (Sigma), protease inhibitor mixture (Sigma), 1 mM NaF, 1% Triton X-100, 25 mM Tris, pH 6.8). Cells were lysed by freezethawing and briefly sonicated to shear the DNA. Protein levels were determined using the BCA assay (Pierce, Rockford, IL). Samples were diluted with 2× sample buffer (0.5 mg/ml bromphenol blue, 2.5% SDS, 25% glycerol, 6% 2-mercaptoethanol, 100 mM Tris, pH 6.8) (34) and boiled for 5 min before loading.
Protein (10 μg) was fractionated on SDS-polyacrylamide gels using either Daiichi precast gels or a triple-wide electrophoresis apparatus (CBS Scientific, Del Mar, CA). Samples from replicate experiments (Figs. 4 and 5) were analyzed on the same Western blot to reduce variability. Parallel gels were stained with Coomassie Blue to verify loading, sample integrity, and protein separation (data not shown). Proteins were transferred overnight (50 mA) to polyvinylidene difluoride membranes for immunodetection (35). Membranes were blocked for 1 h with 5% nonfat powdered milk in Tris-buffered saline, 0.05% Tween 20, 25 mM Tris-HCl, 150 mM NaCl, 0.05% Tween 20, pH 7.3 and probed at 4 °C overnight (phospho-ERK, 1:1000, total ERK, 1:40,000, c-Raf Ser259, 1:1000, phospho-AKT, 1:1000). Horseradish peroxidase-conjugated anti-rabbit secondary antibody was used for detection at a dilution of 1:2000. Secondary antibody incubations were for 2 h, and membranes were washed three times in Tris-buffered saline, 0.05% Tween 20, between antibody incubations. Peroxidase activity was detected using chemiluminescence (Western Lightning, PerkinElmer Life Sciences). Chemiluminescence was imaged using a Kodak Image Station 1000, and net intensity values were plotted using Origin (Microcal).
MTT Conversion Assay–Viability was assayed by MTT conversion (36). Cells were plated in 96-well Costar plates (3000/well for log-phase experiments, 6000/well for experiments performed on confluent cultures) and allowed to adhere overnight or grown to confluence where indicated. Cells were treated with inhibitors, forskolin, or Me2SO in serum-free medium. After 30 min (to simulate pretreatment conditions used for activation assays) or 18 h of incubation, MTT was added to each well (0.5 mg/ml, final concentration) and plates were incubated for 1 h at 37 °C. Medium was aspirated with a beveled needle, and formazan product was solubilized in 100 μl of isopropyl alcohol. MTT conversion (A550—A690) was measured using a microplate spectrophotometer (SpectraMAX Plus, Molecular Devices).
Phosphatase Activity–PP1 and PP2A activity were measured using 6,8-difluoro-4-methylumbelliferyl phosphate (DiFMUP, Molecular Probes, Eugene, OR). Cells were treated as described above and lysed on ice with protease inhibitors, 0.1% Tween 20, and 100 mM Tris, pH 7.0. Membrane proteins were prepared by homogenization on ice with a Dounce homogenizer in 250 mM sucrose in the presence of protease inhibitors. Nuclei were removed by centrifugation at 4 °C and 1000 × g for 10 min. The supernatant was collected and centrifuged at 100,000 × g for 30 min. Membrane proteins were resuspended in protease inhibitors, 0.1% Tween 20, 100 mM Tris, pH 7.0. 10 μg of protein was added to equal volumes of 2× reaction buffer and requisite cofactors (2 mM dithiothreitol, 200 μM MgCl2 for PP1 or 1 mM NiCl2 for PP2A) in the substrate-coated assay wells. Samples were incubated for 30 min at room temperature. Phosphatase activity was determined fluorometrically using excitation/emission wavelengths of 358/452 nm, respectively, with a SpectraMAX Gemini microplate spectrofluorometer (Molecular Devices). Preliminary experiments were performed with PP1- and PP2A-selective inhibitors to validate the assay, and the inhibitors blocked phosphatase activity (data not shown).
Data Analysis–Data were normalized to the control samples from the same cell culture plate for presentation; however, statistics were performed on the net intensity or raw absorbance values. Data were analyzed by ANOVA using Origin (Microcal) and considered statistically significant when the p value was <0.05. Concentration-response data were fit to a logistic equation, and EC50 values were calculated using Origin.
RESULTS
The cannabinoid receptor agonist WIN 55,212-2 is a potent activator of ERK in N1E-115 cells. Serum-starved cells were incubated with increasing concentrations of WIN, and ERK phosphorylation was measured by Western blot using a phospho-specific antibody. ERK was weakly stimulated at 5 min and showed signs of desensitization after 60 min (Fig. 1A). A 20-min treatment with WIN produced robust ERK activation but was not linear with respect to WIN concentration. ERK phosphorylation was linear with respect to WIN concentration after 10 min of stimulation (Fig. 1A); therefore, all of the subsequent experiments were performed at this time point.
The rising portion of the concentration-response curve (up to 100 nM) was fit to a logistic equation, and the EC50 for ERK activation was ∼10 nM with a steep slope for activation that is characteristic of activation by a G-protein-coupled receptor (Fig. 1B). Phosphorylated ERK could also be detected by gel shift in the same samples run on parallel gels run under continuous acrylamide and pH conditions (Fig. 1B, bottom panel). The amount of hyperphosphorylated (gel-shifted) ERK was in agreement with that detected using the phospho-specific antibody. The linear response occurred between 1 and 100 nM and decreased at concentrations above 1 μM, suggesting desensitization.
ERK activation by WIN was abolished by preincubation with the CB1 receptor antagonist SR141716A (Fig. 2A), implicating the CB1 receptor in the response. SR141716A also reduced basal levels of active ERK (Fig. 2A). N1E-115 cells produce the endocannabinoid 2-arachidonoyl glycerol (data not shown); therefore this inhibition by SR141716A is probably because of autocrine activation of the ERK pathway by endocannabinoids. Pretreatment (18 h) with 100 ng/ml pertussis toxin also blocked ERK activation by WIN, indicating that a G-protein of the Gi/o class is responsible for signaling to ERK (Fig. 1B).
Consistent with recent observations (24), ERK activation by 10 nM WIN is partially abolished by pretreatment with the PI 3-kinase inhibitor LY 294002 (50 μM, Fig. 3A). However, this inhibitor also reduced basal levels of active ERK in serumstarved cells (control versus LY 294002 alone, p < 0.05 by ANOVA), resulting in a 1.7-fold ERK activation by WIN in the presence of LY 294002. Inhibitors of MEK (U0126 and PD 98059, 1 and 50 μM, respectively) completely blocked WIN-stimulated ERK phosphorylation (Fig. 3, B and C), which suggests that ERK activation occurs through a traditional Raf-MEK-ERK cascade and not through inhibition of an ERK phosphatase. The Src inhibitor PP2 also partially inhibited ERK phosphorylation (Fig. 3D); however, the PKC inhibitor GF 109203X did not block WIN-stimulated ERK phosphorylation (Fig. 3E). ERK activation was completely abolished by pretreatment with the adenylate cyclase activator forskolin and enhanced in the presence of the PKA inhibitor H-89 (Fig. 3, F and G).
Inhibition by LY 294002 and PP2 was incomplete when cells were treated with 10 nM WIN, and these inhibitors alone reduced basal ERK phosphorylation. We therefore reasoned that the effect of these inhibitors may be on basal ERK phosphorylation, which would produce a rightward shift in the doseresponse curve in the linear portion. We tested both PP2 and LY 294002 at 10 and 50 μM, respectively, against concentrations of WIN up to 100 nM. ERK activation was not completely abolished by either inhibitor (Figs. 4 and 5, A and B), and the data suggest that the dose-response curve is shifted because basal activity is decreased. Cannabinoids are known to activate the PI 3-kinase-dependent AKT pathway (37). We did not observe CB1-stimulated phosphorylation of AKT Thr308 in N1E-115 cells (data not shown); however, AKT Ser473 phosphorylation was completely blocked by LY 294002 pretreatment (Fig. 4C), indicating that the concentration of inhibitor used was sufficient to block PI 3-kinase activity. Therefore, PI 3-kinase and Src activities seem to be required for basal priming of the signaling pathway but full activation requires a second signal.
Initial experiments with pharmacological inhibitors indicated that forskolin can inhibit and H-89 can enhance CB1-stimulated ERK activation. We examined this effect on the initial linear portion of the WIN concentration-response curve. Forskolin (10 μM) completely abolished ERK activation at all of the concentrations of WIN tested while H-89 (10 μM) enhanced ERK activation at 10, 30, and 100 nM (Fig. 6A). Dideoxy forskolin did not change WIN-stimulated ERK activation (data not shown). Forskolin also abolished activation of ERK by the Gi/o-coupled δ-opioid receptor agonist DPDPE (Fig. 6B). We determined the EC50 for forskolin inhibition of WIN-stimulated ERK activation (Fig. 6C). Concentrations of <3 μM were sufficient to inhibit WIN-stimulated ERK activation with near-maximal inhibition at 10 μM. We then examined the ability of H-89 to block forskolin-induced inhibition of WIN-stimulated ERK activation. H-89 was able to reverse the effect of forskolin (Fig. 6D), suggesting that PKA is responsible for inhibition of ERK activation. Because c-Raf is a PKA substrate and phosphorylation of c-Raf by PKA negatively regulates activity, we used a phospho-specific antibody to probe for dephosphorylation of Ser259. Dephosphorylation at this site is required for c-Raf activation (30—33), and we observed a concentration-dependent decrease in Ser259 phosphorylation with WIN stimulation that was inversely related to ERK activation (Fig. 6E).
Cannabinoids are known to inhibit cAMP production in N1E-115 cells (38) and in the closely related NG108—15 cell line (8), but the effect of cannabinoids on the activity of phosphatases that remove the Ser259 phosphate have not been described. We examined PP1 and PP2A activity using 6,8-difluoro-4-methylumbelliferyl phosphate as a substrate. Activity was measured in total protein preparations (Fig. 7A) and in membrane preparations (Fig. 7B). Treatment with WIN (100 nM) did not change either PP1 or PP2A activity. These data indicate that CB1 receptor stimulation does not directly activate PP1 or PP2A but may perturb the balance between constitutively active PKA and PP1 or PP2A.
We next examined the effects of the pharmacological agents used in ERK activation experiments on cellular viability because the decrease in basal ERK activation observed with LY 294002 and PP2 could induce cell death or cause cell cycle arrest. Whereas the cells used in the ERK assays were only treated for 30 min with the inhibitors, it is probable that the initial signal that would subsequently lead to cell death would be generated in that time. Cells were treated for either 30 min or 18 h with LY 294002, U0126, PD 98059, forskolin, and PP2. Viability was modestly enhanced (5%) by forskolin and significantly reduced by LY 294002 (Fig. 8B) at 18 h, but there was no acute change in metabolic activity after 30 min of inhibitor treatment (Fig. 8A). PP2 (10 μM) and the MEK inhibitor U0126 (1 μM) also reduced viability at 18 h, but this is probably the result of decreased proliferation and not cell death because the viability of confluent cells was not affected by PP2 or U0126 treatment (Fig. 8C). LY 294002 was also toxic to confluent cells (Fig. 8C). Because LY 294002 showed significant toxicity at the 50 μM concentration, we performed the experiment again with lower drug concentrations (Fig. 8D). Concentrations as low as 5 μM significantly reduced cell viability, indicating that N1E-115 cells are dependent on PI 3-kinase for survival.
DISCUSSION
ERK activation by the cannabinoid receptor agonist WIN 55212-2 is mediated by CB1 receptor signaling through Gi/o proteins in N1E-115 mouse neuroblastoma cells. A schematic diagram of this pathway is shown in Fig. 9. Multiple basally activated pathways contribute to ERK activation including PI 3-kinase, Src, and protein phosphatases, but receptor-stimulated inhibition of adenylate cyclase/PKA is absolutely required for ERK activation.
Basal PI 3-kinase activity is a prerequisite for ERK activation by WIN. Although WIN-stimulated ERK phosphorylation was occluded by LY 294002 at lower WIN concentrations, levels were near control at the 100 nM concentration. PI 3-kinase activates the pro-survival AKT pathway (39), and N1E-115 cells are dependent on PI 3-kinase-dependent signaling for viability (Fig. 8). The concentration of LY 294002 used was sufficient to completely block WIN-stimulated AKT phosphorylation and induce cell death within 18 h. Therefore, PI 3-kinase appears to be required for WIN stimulation of ERK in N1E-115 cells in as much as the pathway provides a priming activity present in healthy viable cells.
Inhibition of Src activity decreased basal ERK phosphorylation, and WIN stimulation increased this over basal levels. However, unlike cells treated with LY 294002, the amount of activation never reached the level observed with WIN alone at the same concentration. This suggests that Src activity may be required to maximally activate ERK. We did not find any evidence for WIN-stimulated Src activity when assayed using phospho-specific antibodies against both the stimulatory and inhibitory Src phosphorylation sites (data not shown). Galve-Roperh et al. (24) described a similar effect of Src inhibition on ERK activation, and they concluded that Src was not required for ERK activation by cannabinoids. Using hippocampal slices, Derkinderen et al. (40) arrived at a similar conclusion, suggesting a scaffolding, non-catalytic role for the Src family kinase Fyn in CB1-stimulated ERK activation in hippocampal slices. Src regulates multiple basal cellular functions including proliferation and adhesion. Transformed cells are frequently mutated in pathways that control proliferation and differentiation, and this may be the source of high basal Src activity. Src appears to be required for cell cycle progression in N1E-115 cells, because Src inhibition decreased MTT conversion in log-phase cells but not in confluent cultures. Adhesion and focal adhesion kinase activation are another source of active Src (41). Focal adhesion kinase can converge with G-protein-coupled receptors (42). Therefore, adherent, transformed cells have basal Src activity that may augment or cooperate with CB1 receptors in ERK activation. This activity may be required for maximal ERK activation by CB1 receptors.
Although all of these pathways provide basal or priming activity of the pathway, receptor-stimulated inhibition of the cAMP/PKA is required for ERK activation. Inhibition of PKA has been implicated in CB1-mediated retraction of neurites (38) in N1E-115 cells and in ERK activation in hippocampal slices (40). The PKA pathway is known to antagonize growth factor-stimulated ERK activation in several systems, whereas increasing adenylate cyclase activity activates ERK in cells of neural origin (26, 43). This is because of a high level of expression of B-Raf in neurons. However, the involvement of PKA in negative regulation of ERK activation by cannabinoids may not apply only to cells expressing c-Raf. ERK activation by cannabinoids is also antagonized by cAMP analogues in neurons (40) where forskolin is known to activate ERK (29, 43) through B-Raf. ERK activation was also increased basally by H-89 and adenosine 3′,5′ cyclic phosphothioate-Rp in hippocampal slices (40), but this study did not examine the additive effects of inhibition of PKA and cannabinoid receptor stimulation. Activation of ERK by cannabinoids may therefore occur through inhibition of adenylate cyclase in neuronal and non-neuronal cells regardless of the predominant Raf isoform expressed.
Raf represents an attractive target for signal integration. c-Raf is phosphorylated by PKA at three residues: Ser43, Ser259, and Ser621. Phosphorylation of these residues inhibits enzyme activity (27—31). Ser338 is phosphorylated by p21-activated kinase, and the p21-activated kinase/PKC/Src-phosphorylated 338SSYY341 region of c-Raf regulates its association with MEK (29). Inhibition of basal tyrosine phosphorylation at this site by treating cells with PP2 may be partially responsible for the decrease we observed in WIN-stimulated ERK activation in the presence of PP2. We examined Ser259 because multiple studies implicate dephosphorylation of this site as an absolute requirement for c-Raf activation (30—33). Decreased phosphorylation of this site, combined with high basal phosphatase activity, appears to be essential for ERK activation by WIN.
These data are in contrast to observations in human astrocytoma cells (24) where a Src-, epidermal growth factor receptor-, and platelet-derived growth factor receptor-independent but PI 3-kinase-dependent pathway leading to ERK activation was described. Cannabinoids are less potent inhibitors of adenylate cyclase isoforms expressed by C6 glioma cells than those expressed by NG108—15 neuroblastoma cells or N18TG2 neuroblastoma cells (8, 14), suggesting that inhibition of AC by the CB1 receptor may not be the predominant pathway for ERK activation in cells of glial origin. Galve-Roperh et al. (24) also used the synthetic agonist HU-210, which may activate different G-protein coupled pathways than those activated by WIN (44).
ERK can modulate transcription, translation, synaptic vesicle fusion, and cytoskeletal dynamics, suggesting a role for cannabinoid-stimulated ERK in transcription-dependent and independent forms of plasticity. CB1 receptor stimulation in striatum and hippocampus activates ERK and leads to the phosphorylation of downstream transcription factors (40, 45), suggesting that ERK activation by cannabinoids may regulate transcription-dependent forms of synaptic plasticity. A complete understanding of the cellular signal transduction pathways used by cannabinoids and endocannabinoids will lead to a more thorough understanding of how retrograde messengers regulate neurotransmission.
Source, Graphs and Figures: A Predominant Role for Inhibition of the Adenylate Cyclase/Protein Kinase A Pathway in ERK Activation by Cannabinoid Receptor 1 in N1E-115 Neuroblastoma Cells
Cannabinoids activate several members of the mitogen-activated protein kinase superfamily including p44 and p42 extracellular signal-regulated kinase (ERK). We used N1E-115 neuroblastoma cells and the cannabinoid receptor agonist WIN 55,212-2 (WIN) to examine the signal transduction pathways leading to the activation of ERK. ERK phosphorylation (activation) was measured by Western blot. The EC50 for stimulation of ERK phosphorylation was 10 nM, and this effect was blocked by pertussis toxin and the CB1 (cannabinoid) receptor antagonist SR141716A. The MEK inhibitors PD 98059 and U0126 blocked ERK phosphorylation, as did the adenylate cyclase activator forskolin. The phosphatidylinositol (PI) 3-kinase inhibitor LY 294002 and the Src kinase inhibitor PP2 partially occluded the response but also decreased basal levels of phospho-ERK. The PI 3-kinase and Src pathways are known to promote cell survival in many systems; therefore, MTT (1-(4,5-dimethylthiazol-2-yl)-3,5-diphenylformazan) conversion was used to examine the effects of these inhibitors on cellular viability. LY 294002 decreased the number of viable cells after 18 h of treatment; therefore, the inhibition of ERK by this inhibitor is probably because of cytotoxicity. Forskolin blocked ERK phosphorylation with an EC50 of <3 μM, and the protein kinase A (PKA) inhibitor H-89 enhanced ERK phosphorylation. c-Raf phosphorylation at an inhibitory PKA-regulated site (Ser259) was also reduced by WIN. This is probably due to constitutive phosphatase activity because WIN did not directly stimulate PP1 or PP2A activity when measured using 6,8-difluoro-4-methylumbelliferyl phosphate as a fluorogenic substrate. These data implicate the inhibition of PKA as the predominant pathway for ERK activation by CB1 receptors in N1E-115 cells. PI 3-kinase and Src appear to contribute to ERK activation by maintaining activation of kinases, which prime the pathway and maintain cellular viability.
Although cannabinoids have been used both medicinally and recreationally for thousands of years, only recently have the receptors and signaling cascades responsible for the physiological effects of these drugs and the endogenous brain cannabinoids been elucidated. This is primarily because the receptors were only identified recently and selective potent agonists and antagonists have only been available for a short time. Three endogenous lipid-derived ligands (endocannabinoids) have been characterized to date. These include anandamide (1), noladin ether (2), and 2-arachidonoyl glycerol (3), although there are probably more endocannabinoids yet to be characterized. Two cannabinoid receptors (CB1 and CB2)1 have been cloned to date (4, 5), whereas a third subtype has been characterized in the vasculature and in the brains of CB1 receptor knock-out mice (6, 7). Both CB1 and CB2 couple to pertussis toxin-sensitive G-proteins (8), although non-receptor-mediated forms of cannabinoid signaling have been reported previously (9). Cannabinoids and endocannabinoids regulate synaptic transmission at both excitatory and inhibitory synapses and participate in long term plasticity (reviewed in Ref. 10). Recent studies suggest that cannabinoids and cannabinoid receptor antagonists may also have therapeutic potential in disorders including Parkinson's disease, multiple sclerosis, human immunodeficiency virus, and stroke (reviewed in Ref. 11).
Inhibition of adenylate cyclase (AC) through Gi coupling was the first intracellular signal transduction pathway implicated in CB1 receptor signaling (8). Regulation of AC can occur through both α and βγ subunits depending on the isoforms of AC expressed by the cells (12, 13). Cannabinoids increase cAMP production in CHO cells expressing AC II, IV, or VII through βγ subunits but decrease activity in cells expressing AC I, V, VI, and VIII (14). In addition, cannabinoids have been linked to sphingomyelin hydrolysis and ceramide generation in glial and glioma cells (15). Cannabinoids also inhibit calcium channel function (16, 17) and increase K+ conductance (18), which can decrease neuronal excitability and may contribute to synaptic plasticity.
The mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) pathway is also activated by cannabinoids (19). ERK regulates proliferation, differentiation, and synaptic plasticity and is activated through a myriad of diverse signals. Growth factors frequently use tyrosine kinases to signal, resulting in Ras-dependent activation of the Ser/Thr kinase Raf (20). G-protein-coupled receptors can "hijack" this tyrosine kinase pathway by protease-mediated transactivation (21, 22) or may employ G-protein-stimulated protein kinases to directly phosphorylate signaling intermediates, including Shc and Raf. G-protein βγ-subunits can also regulate ERK through arrestin and Src-dependent pathways (23). In addition, βγ subunits may activate ERK through PI 3-kinase as a signaling intermediate. This Src-independent pathway was recently described in human astrocytoma cells stimulated with CB1 receptor agonists (24).
Raf represents a point of convergence for multiple signaling pathways that activate ERK. Raf phosphorylates the dualspecificity kinase MEK, which in turn phosphorylates ERK at tyrosine and threonine residues in the activation domain (25). Some of these effects appear to be cell type-specific. cAMP activates ERK in cells expressing B-Raf (26) while inhibiting ERK activation in cells expressing c-Raf/Raf-1 (27, 28). In addition to PKA, PKC, Src, AKT, casein kinase I, and p21-activated kinase are also capable of phosphorylating Raf (29). Although the contribution of each of the multiple phosphorylations is not completely understood, dephosphorylation of the PKA site Ser259 appears to be absolutely required for c-Raf activation (30—33). CB1 receptors couple negatively to AC, which may mediate ERK activation in cells expressing c-Raf (e.g. relief of tonic inhibition produced by basal PKA activity).
The present experiments were performed to determine the intracellular signaling pathways responsible for ERK activation by cannabinoids in N1E-115 mouse neuroblastoma cells. Pharmacological and biochemical approaches were used to examine the role of PKA, PI 3-kinase, Src, and PKC in CB1-stimulated ERK phosphorylation. These data indicate that multiple signals generated by constitutively activated kinases and phosphatases contribute to CB1-stimulated ERK phosphorylation in N1E-115 neuroblastoma cells. Basal activation of Src and PI 3-kinase contribute to ERK activation, but adenylate cyclase activation completely abolishes CB1 signaling to ERK. These data implicate inhibition of adenylate cyclase and dephosphorylation of c-Raf in ERK activation by cannabinoids in cells of neural origin.
EXPERIMENTAL PROCEEDURES
Reagents–Anti-phospho ERK, anti-phospho AKT, and anti-phospho-c-Raf Ser259 were obtained from Cell Signaling (Beverly, MA). Polyclonal anti-ERK and horseradish peroxidase anti-rabbit were purchased from Sigma. WIN 55,212-2, DPDPE, PP2, and GF 109203X were purchased from Tocris (Avonmouth, Bristol, United Kingdom). SR141716A was from the National Institute of Mental Health chemical synthesis program. Protease inhibitors, phosphatase inhibitors, MTT, and phorbol 12-myristate 13-acetate were obtained from Sigma. Forskolin, PD 98059, and U0126 were purchased from Calbiochem. Pertussis toxin was purchased from List Biological Laboratories (Campbell, CA). Cell culture reagents were obtained from Invitrogen.
Cell Culture–N1E-115 cells (ATCC CRL-2263) were maintained at 37 °C, 5% CO2 in Dulbecco's modified Eagle's medium (high glucose with pyridoxine-HCl and sodium pyruvate, Invitrogen catalogue 10313) supplemented with 2 mM L-glutamine, 10% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin. Cells were subcultured at weekly intervals.
Kinase Activation Assays–Cells (105) were plated in 24-well cell culture dishes, incubated for 24—48 h, and serum-starved for 12—24 h prior to assay. Cells were treated with WIN in Dulbecco's modified Eagle's medium containing 0.1% bovine serum albumin as a carrier and incubated at 37 °C, 5% CO2 for 5—60 min. In experiments where antagonists were used, cells were pretreated with the antagonist for 5 min prior to stimulation. Cells were pretreated for 30 min with LY 294002, PP2, U0126, PD 98059, forskolin, and GF 109203X. For experiments with pertussis toxin inactivation of Gi/o, cells were treated for 12 h with 100 ng/ml pertussis toxin in complete medium. Cells were then serumstarved in the presence of pertussis toxin for an additional 12 h. All of the stock solutions were prepared in Me2SO, and dilutions were prepared with Me2SO concentrations kept constant (0.1—0.2%). The assay was terminated by aspirating the medium, by placing the cells on dry ice, and with the addition of lysis buffer (phosphatase inhibitor cocktails I and II (Sigma), protease inhibitor mixture (Sigma), 1 mM NaF, 1% Triton X-100, 25 mM Tris, pH 6.8). Cells were lysed by freezethawing and briefly sonicated to shear the DNA. Protein levels were determined using the BCA assay (Pierce, Rockford, IL). Samples were diluted with 2× sample buffer (0.5 mg/ml bromphenol blue, 2.5% SDS, 25% glycerol, 6% 2-mercaptoethanol, 100 mM Tris, pH 6.8) (34) and boiled for 5 min before loading.
Protein (10 μg) was fractionated on SDS-polyacrylamide gels using either Daiichi precast gels or a triple-wide electrophoresis apparatus (CBS Scientific, Del Mar, CA). Samples from replicate experiments (Figs. 4 and 5) were analyzed on the same Western blot to reduce variability. Parallel gels were stained with Coomassie Blue to verify loading, sample integrity, and protein separation (data not shown). Proteins were transferred overnight (50 mA) to polyvinylidene difluoride membranes for immunodetection (35). Membranes were blocked for 1 h with 5% nonfat powdered milk in Tris-buffered saline, 0.05% Tween 20, 25 mM Tris-HCl, 150 mM NaCl, 0.05% Tween 20, pH 7.3 and probed at 4 °C overnight (phospho-ERK, 1:1000, total ERK, 1:40,000, c-Raf Ser259, 1:1000, phospho-AKT, 1:1000). Horseradish peroxidase-conjugated anti-rabbit secondary antibody was used for detection at a dilution of 1:2000. Secondary antibody incubations were for 2 h, and membranes were washed three times in Tris-buffered saline, 0.05% Tween 20, between antibody incubations. Peroxidase activity was detected using chemiluminescence (Western Lightning, PerkinElmer Life Sciences). Chemiluminescence was imaged using a Kodak Image Station 1000, and net intensity values were plotted using Origin (Microcal).
MTT Conversion Assay–Viability was assayed by MTT conversion (36). Cells were plated in 96-well Costar plates (3000/well for log-phase experiments, 6000/well for experiments performed on confluent cultures) and allowed to adhere overnight or grown to confluence where indicated. Cells were treated with inhibitors, forskolin, or Me2SO in serum-free medium. After 30 min (to simulate pretreatment conditions used for activation assays) or 18 h of incubation, MTT was added to each well (0.5 mg/ml, final concentration) and plates were incubated for 1 h at 37 °C. Medium was aspirated with a beveled needle, and formazan product was solubilized in 100 μl of isopropyl alcohol. MTT conversion (A550—A690) was measured using a microplate spectrophotometer (SpectraMAX Plus, Molecular Devices).
Phosphatase Activity–PP1 and PP2A activity were measured using 6,8-difluoro-4-methylumbelliferyl phosphate (DiFMUP, Molecular Probes, Eugene, OR). Cells were treated as described above and lysed on ice with protease inhibitors, 0.1% Tween 20, and 100 mM Tris, pH 7.0. Membrane proteins were prepared by homogenization on ice with a Dounce homogenizer in 250 mM sucrose in the presence of protease inhibitors. Nuclei were removed by centrifugation at 4 °C and 1000 × g for 10 min. The supernatant was collected and centrifuged at 100,000 × g for 30 min. Membrane proteins were resuspended in protease inhibitors, 0.1% Tween 20, 100 mM Tris, pH 7.0. 10 μg of protein was added to equal volumes of 2× reaction buffer and requisite cofactors (2 mM dithiothreitol, 200 μM MgCl2 for PP1 or 1 mM NiCl2 for PP2A) in the substrate-coated assay wells. Samples were incubated for 30 min at room temperature. Phosphatase activity was determined fluorometrically using excitation/emission wavelengths of 358/452 nm, respectively, with a SpectraMAX Gemini microplate spectrofluorometer (Molecular Devices). Preliminary experiments were performed with PP1- and PP2A-selective inhibitors to validate the assay, and the inhibitors blocked phosphatase activity (data not shown).
Data Analysis–Data were normalized to the control samples from the same cell culture plate for presentation; however, statistics were performed on the net intensity or raw absorbance values. Data were analyzed by ANOVA using Origin (Microcal) and considered statistically significant when the p value was <0.05. Concentration-response data were fit to a logistic equation, and EC50 values were calculated using Origin.
RESULTS
The cannabinoid receptor agonist WIN 55,212-2 is a potent activator of ERK in N1E-115 cells. Serum-starved cells were incubated with increasing concentrations of WIN, and ERK phosphorylation was measured by Western blot using a phospho-specific antibody. ERK was weakly stimulated at 5 min and showed signs of desensitization after 60 min (Fig. 1A). A 20-min treatment with WIN produced robust ERK activation but was not linear with respect to WIN concentration. ERK phosphorylation was linear with respect to WIN concentration after 10 min of stimulation (Fig. 1A); therefore, all of the subsequent experiments were performed at this time point.
The rising portion of the concentration-response curve (up to 100 nM) was fit to a logistic equation, and the EC50 for ERK activation was ∼10 nM with a steep slope for activation that is characteristic of activation by a G-protein-coupled receptor (Fig. 1B). Phosphorylated ERK could also be detected by gel shift in the same samples run on parallel gels run under continuous acrylamide and pH conditions (Fig. 1B, bottom panel). The amount of hyperphosphorylated (gel-shifted) ERK was in agreement with that detected using the phospho-specific antibody. The linear response occurred between 1 and 100 nM and decreased at concentrations above 1 μM, suggesting desensitization.
ERK activation by WIN was abolished by preincubation with the CB1 receptor antagonist SR141716A (Fig. 2A), implicating the CB1 receptor in the response. SR141716A also reduced basal levels of active ERK (Fig. 2A). N1E-115 cells produce the endocannabinoid 2-arachidonoyl glycerol (data not shown); therefore this inhibition by SR141716A is probably because of autocrine activation of the ERK pathway by endocannabinoids. Pretreatment (18 h) with 100 ng/ml pertussis toxin also blocked ERK activation by WIN, indicating that a G-protein of the Gi/o class is responsible for signaling to ERK (Fig. 1B).
Consistent with recent observations (24), ERK activation by 10 nM WIN is partially abolished by pretreatment with the PI 3-kinase inhibitor LY 294002 (50 μM, Fig. 3A). However, this inhibitor also reduced basal levels of active ERK in serumstarved cells (control versus LY 294002 alone, p < 0.05 by ANOVA), resulting in a 1.7-fold ERK activation by WIN in the presence of LY 294002. Inhibitors of MEK (U0126 and PD 98059, 1 and 50 μM, respectively) completely blocked WIN-stimulated ERK phosphorylation (Fig. 3, B and C), which suggests that ERK activation occurs through a traditional Raf-MEK-ERK cascade and not through inhibition of an ERK phosphatase. The Src inhibitor PP2 also partially inhibited ERK phosphorylation (Fig. 3D); however, the PKC inhibitor GF 109203X did not block WIN-stimulated ERK phosphorylation (Fig. 3E). ERK activation was completely abolished by pretreatment with the adenylate cyclase activator forskolin and enhanced in the presence of the PKA inhibitor H-89 (Fig. 3, F and G).
Inhibition by LY 294002 and PP2 was incomplete when cells were treated with 10 nM WIN, and these inhibitors alone reduced basal ERK phosphorylation. We therefore reasoned that the effect of these inhibitors may be on basal ERK phosphorylation, which would produce a rightward shift in the doseresponse curve in the linear portion. We tested both PP2 and LY 294002 at 10 and 50 μM, respectively, against concentrations of WIN up to 100 nM. ERK activation was not completely abolished by either inhibitor (Figs. 4 and 5, A and B), and the data suggest that the dose-response curve is shifted because basal activity is decreased. Cannabinoids are known to activate the PI 3-kinase-dependent AKT pathway (37). We did not observe CB1-stimulated phosphorylation of AKT Thr308 in N1E-115 cells (data not shown); however, AKT Ser473 phosphorylation was completely blocked by LY 294002 pretreatment (Fig. 4C), indicating that the concentration of inhibitor used was sufficient to block PI 3-kinase activity. Therefore, PI 3-kinase and Src activities seem to be required for basal priming of the signaling pathway but full activation requires a second signal.
Initial experiments with pharmacological inhibitors indicated that forskolin can inhibit and H-89 can enhance CB1-stimulated ERK activation. We examined this effect on the initial linear portion of the WIN concentration-response curve. Forskolin (10 μM) completely abolished ERK activation at all of the concentrations of WIN tested while H-89 (10 μM) enhanced ERK activation at 10, 30, and 100 nM (Fig. 6A). Dideoxy forskolin did not change WIN-stimulated ERK activation (data not shown). Forskolin also abolished activation of ERK by the Gi/o-coupled δ-opioid receptor agonist DPDPE (Fig. 6B). We determined the EC50 for forskolin inhibition of WIN-stimulated ERK activation (Fig. 6C). Concentrations of <3 μM were sufficient to inhibit WIN-stimulated ERK activation with near-maximal inhibition at 10 μM. We then examined the ability of H-89 to block forskolin-induced inhibition of WIN-stimulated ERK activation. H-89 was able to reverse the effect of forskolin (Fig. 6D), suggesting that PKA is responsible for inhibition of ERK activation. Because c-Raf is a PKA substrate and phosphorylation of c-Raf by PKA negatively regulates activity, we used a phospho-specific antibody to probe for dephosphorylation of Ser259. Dephosphorylation at this site is required for c-Raf activation (30—33), and we observed a concentration-dependent decrease in Ser259 phosphorylation with WIN stimulation that was inversely related to ERK activation (Fig. 6E).
Cannabinoids are known to inhibit cAMP production in N1E-115 cells (38) and in the closely related NG108—15 cell line (8), but the effect of cannabinoids on the activity of phosphatases that remove the Ser259 phosphate have not been described. We examined PP1 and PP2A activity using 6,8-difluoro-4-methylumbelliferyl phosphate as a substrate. Activity was measured in total protein preparations (Fig. 7A) and in membrane preparations (Fig. 7B). Treatment with WIN (100 nM) did not change either PP1 or PP2A activity. These data indicate that CB1 receptor stimulation does not directly activate PP1 or PP2A but may perturb the balance between constitutively active PKA and PP1 or PP2A.
We next examined the effects of the pharmacological agents used in ERK activation experiments on cellular viability because the decrease in basal ERK activation observed with LY 294002 and PP2 could induce cell death or cause cell cycle arrest. Whereas the cells used in the ERK assays were only treated for 30 min with the inhibitors, it is probable that the initial signal that would subsequently lead to cell death would be generated in that time. Cells were treated for either 30 min or 18 h with LY 294002, U0126, PD 98059, forskolin, and PP2. Viability was modestly enhanced (5%) by forskolin and significantly reduced by LY 294002 (Fig. 8B) at 18 h, but there was no acute change in metabolic activity after 30 min of inhibitor treatment (Fig. 8A). PP2 (10 μM) and the MEK inhibitor U0126 (1 μM) also reduced viability at 18 h, but this is probably the result of decreased proliferation and not cell death because the viability of confluent cells was not affected by PP2 or U0126 treatment (Fig. 8C). LY 294002 was also toxic to confluent cells (Fig. 8C). Because LY 294002 showed significant toxicity at the 50 μM concentration, we performed the experiment again with lower drug concentrations (Fig. 8D). Concentrations as low as 5 μM significantly reduced cell viability, indicating that N1E-115 cells are dependent on PI 3-kinase for survival.
DISCUSSION
ERK activation by the cannabinoid receptor agonist WIN 55212-2 is mediated by CB1 receptor signaling through Gi/o proteins in N1E-115 mouse neuroblastoma cells. A schematic diagram of this pathway is shown in Fig. 9. Multiple basally activated pathways contribute to ERK activation including PI 3-kinase, Src, and protein phosphatases, but receptor-stimulated inhibition of adenylate cyclase/PKA is absolutely required for ERK activation.
Basal PI 3-kinase activity is a prerequisite for ERK activation by WIN. Although WIN-stimulated ERK phosphorylation was occluded by LY 294002 at lower WIN concentrations, levels were near control at the 100 nM concentration. PI 3-kinase activates the pro-survival AKT pathway (39), and N1E-115 cells are dependent on PI 3-kinase-dependent signaling for viability (Fig. 8). The concentration of LY 294002 used was sufficient to completely block WIN-stimulated AKT phosphorylation and induce cell death within 18 h. Therefore, PI 3-kinase appears to be required for WIN stimulation of ERK in N1E-115 cells in as much as the pathway provides a priming activity present in healthy viable cells.
Inhibition of Src activity decreased basal ERK phosphorylation, and WIN stimulation increased this over basal levels. However, unlike cells treated with LY 294002, the amount of activation never reached the level observed with WIN alone at the same concentration. This suggests that Src activity may be required to maximally activate ERK. We did not find any evidence for WIN-stimulated Src activity when assayed using phospho-specific antibodies against both the stimulatory and inhibitory Src phosphorylation sites (data not shown). Galve-Roperh et al. (24) described a similar effect of Src inhibition on ERK activation, and they concluded that Src was not required for ERK activation by cannabinoids. Using hippocampal slices, Derkinderen et al. (40) arrived at a similar conclusion, suggesting a scaffolding, non-catalytic role for the Src family kinase Fyn in CB1-stimulated ERK activation in hippocampal slices. Src regulates multiple basal cellular functions including proliferation and adhesion. Transformed cells are frequently mutated in pathways that control proliferation and differentiation, and this may be the source of high basal Src activity. Src appears to be required for cell cycle progression in N1E-115 cells, because Src inhibition decreased MTT conversion in log-phase cells but not in confluent cultures. Adhesion and focal adhesion kinase activation are another source of active Src (41). Focal adhesion kinase can converge with G-protein-coupled receptors (42). Therefore, adherent, transformed cells have basal Src activity that may augment or cooperate with CB1 receptors in ERK activation. This activity may be required for maximal ERK activation by CB1 receptors.
Although all of these pathways provide basal or priming activity of the pathway, receptor-stimulated inhibition of the cAMP/PKA is required for ERK activation. Inhibition of PKA has been implicated in CB1-mediated retraction of neurites (38) in N1E-115 cells and in ERK activation in hippocampal slices (40). The PKA pathway is known to antagonize growth factor-stimulated ERK activation in several systems, whereas increasing adenylate cyclase activity activates ERK in cells of neural origin (26, 43). This is because of a high level of expression of B-Raf in neurons. However, the involvement of PKA in negative regulation of ERK activation by cannabinoids may not apply only to cells expressing c-Raf. ERK activation by cannabinoids is also antagonized by cAMP analogues in neurons (40) where forskolin is known to activate ERK (29, 43) through B-Raf. ERK activation was also increased basally by H-89 and adenosine 3′,5′ cyclic phosphothioate-Rp in hippocampal slices (40), but this study did not examine the additive effects of inhibition of PKA and cannabinoid receptor stimulation. Activation of ERK by cannabinoids may therefore occur through inhibition of adenylate cyclase in neuronal and non-neuronal cells regardless of the predominant Raf isoform expressed.
Raf represents an attractive target for signal integration. c-Raf is phosphorylated by PKA at three residues: Ser43, Ser259, and Ser621. Phosphorylation of these residues inhibits enzyme activity (27—31). Ser338 is phosphorylated by p21-activated kinase, and the p21-activated kinase/PKC/Src-phosphorylated 338SSYY341 region of c-Raf regulates its association with MEK (29). Inhibition of basal tyrosine phosphorylation at this site by treating cells with PP2 may be partially responsible for the decrease we observed in WIN-stimulated ERK activation in the presence of PP2. We examined Ser259 because multiple studies implicate dephosphorylation of this site as an absolute requirement for c-Raf activation (30—33). Decreased phosphorylation of this site, combined with high basal phosphatase activity, appears to be essential for ERK activation by WIN.
These data are in contrast to observations in human astrocytoma cells (24) where a Src-, epidermal growth factor receptor-, and platelet-derived growth factor receptor-independent but PI 3-kinase-dependent pathway leading to ERK activation was described. Cannabinoids are less potent inhibitors of adenylate cyclase isoforms expressed by C6 glioma cells than those expressed by NG108—15 neuroblastoma cells or N18TG2 neuroblastoma cells (8, 14), suggesting that inhibition of AC by the CB1 receptor may not be the predominant pathway for ERK activation in cells of glial origin. Galve-Roperh et al. (24) also used the synthetic agonist HU-210, which may activate different G-protein coupled pathways than those activated by WIN (44).
ERK can modulate transcription, translation, synaptic vesicle fusion, and cytoskeletal dynamics, suggesting a role for cannabinoid-stimulated ERK in transcription-dependent and independent forms of plasticity. CB1 receptor stimulation in striatum and hippocampus activates ERK and leads to the phosphorylation of downstream transcription factors (40, 45), suggesting that ERK activation by cannabinoids may regulate transcription-dependent forms of synaptic plasticity. A complete understanding of the cellular signal transduction pathways used by cannabinoids and endocannabinoids will lead to a more thorough understanding of how retrograde messengers regulate neurotransmission.
Source, Graphs and Figures: A Predominant Role for Inhibition of the Adenylate Cyclase/Protein Kinase A Pathway in ERK Activation by Cannabinoid Receptor 1 in N1E-115 Neuroblastoma Cells
