Cannabinoid receptor subtype 2 (CB2R) agonist, GW405833 reduces agonist-induced Ca2+ oscillations in mouse pancreatic acinar cells

Emerging evidence demonstrates that the blockade of intracellular Ca2+ signals may protect pancreatic acinar cells against Ca2+ overload, intracellular protease activation, and necrosis. The activation of cannabinoid receptor subtype 2 (CB2R) prevents acinar cell pathogenesis in animal models of acute pancreatitis. However, whether CB2Rs modulate intracellular Ca2+ signals in pancreatic acinar cells is largely unknown. We evaluated the roles of CB2R agonist, GW405833 (GW) in agonist-induced Ca2+ oscillations in pancreatic acinar cells using multiple experimental approaches with acute dissociated pancreatic acinar cells prepared from wild type, CB1R-knockout (KO), and CB2R-KO mice. Immunohistochemical labeling revealed that CB2R protein was expressed in mouse pancreatic acinar cells. Electrophysiological experiments showed that activation of CB2Rs by GW reduced acetylcholine (ACh)-, but not cholecystokinin (CCK)-induced Ca2+ oscillations in a concentration-dependent manner; this inhibition was prevented by a selective CB2R antagonist, AM630, or was absent in CB2R-KO but not CB1R-KO mice. In addition, GW eliminated L-arginine-induced enhancement of Ca2+ oscillations, pancreatic amylase, and pulmonary myeloperoxidase. Collectively, we provide novel evidence that activation of CB2Rs eliminates ACh-induced Ca2+ oscillations and L-arginine-induced enhancement of Ca2+ signaling in mouse pancreatic acinar cells, which suggests a potential cellular mechanism of CB2R-mediated protection in acute pancreatitis.


Animals.
Mice used for this study were adult (4-6 month old), male, CD1 mice (Charles River Laboratories International, Inc., Wilmington, MA, USA). In addition, WT, CB 1 RKO 11 , and CB 2 RKO mice 12 with C57BL/6J genetic backgrounds were initially provided by Dr. Zheng-Xiong Xi at the National Institute on Drug Abuse (NIDA; Bethesda, MD, USA), and were then bred in animal facilities at the Barrow Neurological Institute, which are accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. Genotyping was performed at the NIDA Intramural Research Program before experiments were begun. All animals used in the experiments were matched for age (8-14 weeks) and weight (25-35 grams).

Mouse Pancreatic Acinar Cell Preparation.
Acute isolated pancreatic cells were prepared as previously described [13][14][15] . In brief, pancreatic glands were taken from isoflurane-anesthetized mice, and fragments of the tissue were minced and digested using collagenase (200 U/mL, 25-30 min, 37 °C; Wako Pure Chemicals, Osaka, Japan) in the presence of 1-mM Ca 2+ . After collagenase digestion, the cell suspension was gently pipetted to obtain further separation of the cells, and then washed with physiological saline. A 100-μ L volume of cell suspension was then poured into extracellular solution in a 2-mL experimental bath. The isolated cells usually adhered to the bottom within 15-20 min and were used for recording within 3 h after preparation. All experiments were performed at room temperature (22 ± 1 °C).

Whole-Cell Patch-Clamp Recording and Perforated-Patch Recording.
Conventional whole-cell patch-clamp recording was used to record the Ca 2+ -activated Cl − currents for monitoring intracellular Ca 2+ signal oscillations, as reported previously 13,14 . The recording pipettes, made from borosilicate glass capillaries, had a resistance of 3-5 MΩ when filled with pipette solution. After a GΩ seal was established between the cell membrane and the pipette, a whole-cell configuration was achieved by brief negative suction. Transmembrane currents were recorded with a patch-clamp amplifier (Axopatch 200B; Molecular Devices, Sunnyvale, CA, USA) at a holding potential (V H ) of − 30 mV. For perforated-patch recording, amphotericin B (200 μ g/mL) was dissolved into the pipette solution. In these studies, we did not compensate for series resistance. Drug Application. A stream of standard extracellular solution was continuously perfused over the cell during recording. A computer-controlled U-tube system was used for drug application 16 . For intracellular drug application, the drug was added into pipette solution, and establishment of a whole-cell configuration allowed the drug to diffuse into the recorded cell.
Amylase Estimation. Serum amylase activity was measured using the AMS assay kit (Nanjing Jiancheng Corp., Nanjing, China) and a microplate reader, following the manufacturer's recommendations.

Myeloperoxidase Estimation.
To measure myeloperoxidase (MPO) activity, lung tissues were immediately homogenized on ice in 10 volumes of normal saline. MPO activity was measured using the MPO assay kit (Nanjing Jiancheng Corp., Nanjing, China) and a microplate reader, following the manufacturer's recommendations. CB 2 R Immunoblot Assay. WT, CB 1 R-KO, and CB 2 R-KO mice (3 mice for each group) were anesthetized and quickly perfused with saline to flash all blood cells. Both whole striatum and spleen tissue were dissected out, snap frozen, and kept on dry ice. All the tissues were homogenized in cell lysis buffer (Cell Signaling Technology, Inc., Danvers, MA, USA) using a sonicator and centrifuge at 15,000 rpm for 15 min at 4 °C to get supernatant. The protein concentration for each sample was quantified with a Bio-Rad Protein Assay (Bio-Rad Laboratories, Hercules, CA, USA). A total of 20-μ g protein (spleen) or 40-μ g protein (striatum) were loaded and separated by SDS-PAGE in a 4-15% gradient gel for the detection of endogenous calnexin (Enzo Life Sciences, SPA865) and CB 2 R (NIDA-5633) by using Invitrogen blotting and transferring modules (Grand Island, NY, USA). Membranes were blocked for 2 h at room temperature with Licor Odyssey blocking buffer (LI-COR Biosciences, Lincoln, NE, USA) after washing 3 times with phosphate-buffered saline containing 0.1% Tween-20. Membranes were first incubated with either anti-CB 2 (1:500 NIDA-5633 Ab) or anti-calnexin (1:1,000) antibody overnight at 4 °C. After washing 3 times, the membranes were incubated with goat anti-rabbit IgG (IRDye 680CW) (1:2,500) for 1.5 h at Scientific RepoRts | 6:29757 | DOI: 10.1038/srep29757 room temperature. Then the membranes were washed 3 times and then scanned in a Licor Odyssey Sa Imaging System (LI-COR Biosciences).
Immunohistochemistry. Sections were first blocked in 5% bovine serum albumin (BSA) and 0.5% Triton X-100 in phosphate buffer (PB) for 2 h at room temperature. Then, sections were incubated with 1:500 NIDA-5633 mCB 2 R antibody (Genemed Synthesis Inc, San Antonio, TX, USA) at 4 °C overnight. After washing 3 times with 0.1 M PB, sections were incubated with Alexa Fluor 488 goat anti-rabbit secondary antibody (Invitrogen, Carlsbad, CA, USA) in 5% BSA and 0.5% Triton X-100 PB for 2 h at room temperature. Sections were then washed, mounted, and cover slipped. Images were taken with a fluorescence microscope (Nikon Eclipse 80i) equipped with a digital camera (Nikon Instruments Inc., Melville, NY, USA).

Statistics.
For patch-clamp experiments, the Ca 2+ -activated Cl − current responses were presented as the current charge (current area/Cm/min), and then the drug-induced changes were compared to the baseline level of charge (induced by ACh). When data were obtained from the same recorded cell and the changes of ACh response were compared before, during, and after testing drug exposure, a paired Student t test was used. To compare the effect of the tested drug between 2 groups of animals (e.g., saline group and L-arginine group), the unpaired Student t test was used. To analyze multiple effects, one-way analysis of variance (ANOVA) with Tukey's post hoc tests were used.

Results
CB 2 Rs Are Expressed on Mouse Pancreatic Acinar Cells. Under the acutely dissociated acinar cell protocol, the isolated cells exhibited a typical kidney shape with secretion granules in the central area of the cells (Supplemental Fig. 2), suggesting the purity of the acinar cells as previously reported 13,15,17 . Figure 1A shows the results of the immunoblot assays, illustrating that a CB 2 -positive band was detected (at ~40 kD) in both the spleen and striatal tissues of WT and CB 1 -KO (CB 1 −/− ), while the densities of this band in CB 2 -KO mice (CB 2 −/− ) were substantially reduced in CB 2 -rich spleen tissues and almost undetectable in striatal tissues. Figure 1B shows CB 2 R immunostaining with the CB 2 R antibody (NIDA-5633), illustrating that the high densities of CB 2 R immunostaining were detected in the majority of spleen cells of WT mice. In contrast, a very low density of CB 2 -like staining was detected in a minority of spleen cells in CB 2 R KO mice, suggesting that the NIDA-5633 antibody used is highly mouse CB 2 R-specific. We then used this antibody to detect CB 2 R expression in single isolated acinar cells. Figure 1C demonstrates the photographs taken in bright field (Ca), mouse CB 2 R antibody (mCB 2 -ir, Cb), DAPI (Cc), and merged mCB 2 -ir and DAPI (Cd). We found high densities of CB 2 R immunolabeling in pancreatic acinar cells (Fig. 1Cb,d). These results suggest that CB 2 R protein is expressed in dissociated mouse pancreatic acinar cells.

Effects of GW405833 on ACh-Induced Ca 2+
Oscillations. In acutely dissociated pancreatic acinar cells, low nanomolar concentrations of ACh induced intracellular Ca 2+ signal oscillations, which can be detected using patch-clamp recording and Ca 2+ imaging as previously reported 13,14,[18][19][20] . Our initial series of experiments was designed to test the effects of the CB 2 R agonist, GW405833 (GW), on ACh-induced Ca 2+ oscillations. Figure 2A demonstrates an experimental protocol, in which the ACh (e.g., 10 nM) is continuously perfused to the recorded cell through a bath (U-tube) to get Ca 2+ oscillation response (as a baseline). Then, the GW is added to the bath perfusion in the presence of ACh. Finally, the GW is washed out with the same concentration of ACh. With this protocol, the ACh is continuously perfused throughout the recording period, and we can compare the change of ACh-induced Ca 2+ oscillations before GW perfusion (baseline), during GW perfusion, and after GW washout in the same recorded cell. For statistical analysis of the effects of GW on ACh-induced Ca 2+ oscillations, we measured baseline oscillations as the charge (current area/Cm/min) 18 and compared the changes of Ca 2+ oscillations during GW perfusion and after washout of GW to the baseline. Our data showed that in the continuous presence of 10 nM ACh, 10 μ M GW reduced Ca 2+ oscillations, and this inhibitory effect was reversed after washout (Fig. 2B). A similar inhibitory effect by GW (100 μ M) was also observed on 100 nM ACh-induced Ca 2+ oscillations using confocal Ca 2+ imaging (Fig. 2C). Statistical analysis of the Ca 2+ oscillation signal from 8 cells tested showed that GW significantly reduced ACh-induced Ca 2+ oscillations from baseline level of − 4.69 ± 0.32 to − 1.68 ± 0.32 nC/min (the level after GW exposure, n = 8, paired t test p < 0.001, Fig. 2D). Ca 2+ imaging experiments also showed a similar inhibition of Ca 2+ oscillations by GW (n = 66, paired t test p < 0.001, Fig. 2E). After washout of GW, Ca 2+ oscillations were partially recovered in both patch recording and Ca 2+ imaging. These results suggest that activation of CB 2 R by GW inhibits ACh-induced intracellular Ca 2+ signals in freshly isolated pancreatic acinar cells.

GW Inhibits ACh-Induced Ca 2+ Oscillations by a Selective Action on CB 2 Rs.
To address the question of whether GW inhibition of ACh-induced Ca 2+ oscillations is mediated through CB 2 Rs, we designed three sets of experiments. 1) We tested the effect of a selective CB 2 R antagonist (AM630) on GW inhibition of Ca 2+ oscillations. 2) We examined GW inhibitory effects on pancreatic acinar cells prepared from CB 1 -KO and CB 2 -KO mice. 3) We evaluated the effects of a selective CB 1 R agonist (ACEA) on ACh-induced Ca 2+ oscillations.
The results of these experiments demonstrated that GW inhibition of ACh-induced Ca 2+ oscillations was presented in WT (Fig. 4A) and CB 1 R-KO mice (Fig. 4B), but was absent in CB 2 R-KO mice (Fig. 4C). Figure 4D summarizes pooled data demonstrating the effect of GW on 30 nM ACh-induced Ca 2+ oscillations in WT (p < 0.01, n = 5), CB 1 R-KO (p < 0.001, n = 6), and CB 2 R-KO (p > 0.05, n = 8) mice. Furthermore, co-application of AM630 (0.1 μ M) and GW (10 μ M) abolished the inhibitory effect of GW on 10 nM ACh-induced Ca 2+ oscillations (baseline vs. AM630 + GW p > 0.05, n = 10), while AM630 alone had no affect (baseline vs. AM630, p > 0.05, n = 10, Fig. 4E). Finally, we found that CB 1 R agonist, ACEA (10 μ M) also reduced ACh-induced Ca 2+ oscillations but this effect was likely mediated through ethanol that was used to dissolve ACEA (Supplemental Fig. 3). Together, these results suggest that GW inhibits ACh-induced intracellular Ca 2+ signaling through the action of CB 2 Rs. GW Inhibits ACh-Induced Ca 2+ Oscillations through Membrane CB 2 Rs. Our data clearly demonstrated that GW inhibited ACh-induced intracellular Ca 2+ oscillations. However, it remained unclear whether Western blot assay shows that a high-density CB 2 -immunoreactive band is detected in both spleen and striatal tissues in WT and CB 1 -KO mice, but is undetectable in striatal tissues or substantially reduced in CB 2 -rich spleen tissues in CB 2 -KO mice. (B) Immunohistochemical assays show high densities of CB 2 -immunostaining in spleen slices of WT mice, which are undetectable or substantially diminished in CB 2 -KO mice. (C) Immunocytochemical assays use mouse CB 2 R antibody (NIDA-5633). The bright field photograph (Ca) shows freshly dissociated pancreatic acinar cells. CB 2 -immunostaining (mCB 2 -ir) in single dissociated pancreatic acinar cells illustrates the high densities of CB 2 R proteins (Cb). DAPI staining demonstrates cell nucleus (Cc). The "Merged" image shows superimposed mCB2-ir and DAPI images (Cd).
GW inhibition was mediated through extracellular or intracellular CB 2 Rs. GW could act on extracellular membrane CB 2 Rs and/or modulate muscarinic receptors, or GW could affect intracellular CB 2 Rs, and then modulate signal molecules such as G-protein and/or inositol 1,4,5-trisphosphate (IP 3 ) receptors 21 . To distinguish among these possibilities, we designed two experiments, in which, either the CB 2 R agonist (GW) or antagonist (AM630) was applied internally or in which, IP 3 was applied internally. When GW (100 μ M) was added into the recording electrode and a perforated whole-cell recording (amphotericin B) was performed, bath-application of 10 nM ACh induced Ca 2+ oscillations. When the recording mode was switched from perforated to conventional Columns indicate the mean of current charge ± SEM (left) and the mean DF/DO ± SEM (right) as compared to the baseline level. * * * Indicates p < 0.001 for the value compared to baseline level. Statistic comparison between the levels of baseline and washout of GW showed significance (p < 0.05) in patch-clamp data (Fig. 2D left panel) and in Ca 2+ imaging data (p < 0.01, Fig. 2D right panel). whole-cell recording by a brief suction, GW was infused into the recorded cell, and no detectable inhibitory effect on ACh-induced Ca 2+ oscillations was present (Fig. 5A,D). Using the same experimental protocol, we applied AM630 (1 μ M) intracellularly and found that internal AM630 failed to prevent bath-applied GW-induced inhibition in the ACh-induced Ca 2+ oscillations (Fig. 5B,D). In the presence of intracellularly applied IP 3 (30 μ M), which causes IP 3 -induced Ca 2+ oscillations, GW produced little inhibitory effect on the IP 3 -induced Ca 2+ oscillations (Fig. 5C,D). These data suggest that GW inhibition of ACh-induced Ca 2+ oscillations is not mediated through intracellular IP 3 receptors. Together, these results suggest that GW inhibition of intracellular Ca 2+ oscillations is mediated through CB 2 Rs on the surface of the cytoplasmic membrane.

Effects of GW on CCK-Induced Ca 2+
Oscillations. Data presented thus far demonstrate that GW inhibited ACh-induced Ca 2+ oscillations through cell membrane CB 2 Rs, perhaps through CB 2 Rs and muscarinic receptor cross talk. To test this possibility, we applied CCK to induce Ca 2+ oscillations, which occurs through different receptor signaling pathway than muscarinic receptor, and examined the effects of GW on the CCK-induced Ca 2+ oscillations. As shown in Fig. 6, bath application of 10 pM CCK induced Ca 2+ oscillation responses, which were not affected by bath application of GW (100 μ M, Fig. 6A). In the same recorded cell, bath application of GW (100 μ M) dramatically inhibited 10 nM ACh-induced Ca 2+ oscillations (Fig. 6B). Figure 6C summarizes pooled data from 4 cells tested, and no significant effect of GW on CCK-induced Ca 2+ oscillations was found (p > 0.05, n = 4, Ca), but GW inhibited ACh-induced Ca 2+ oscillations in the same recorded cell (p < 0.01, n = 4, Cb).

L-arginine Potentiates ACh-Induced Ca 2+
Oscillations. L-arginine is used to induce acute pancreatitis in rodents 22 . In dissociated pancreatic acinar cells, bath-application of L-arginine for 10 min enhanced ACh-induced Ca 2+ oscillations from baseline level of 4.93 ± 0.39 to 10.34 ± 1.83 nC/min (Fig. 7Aa,b), which was not reversible after washout for 10 min (Ca 2+ oscillations between L-arginine exposure and washout groups p > 0.05, n = 6, Fig. 7Ac). Statistical analysis revealed that L-arginine significantly enhanced ACh-induced Ca 2+ oscillations (p < 0.05) in an irreversible manner (Fig. 7B). GW Prevents L-arginine-Enhanced Ca 2+ Oscillations. Next, we sought to determine whether GW could eliminate L-arginine-induced enhancement of Ca 2+ oscillations. We showed that either pre-treatment with GW (Fig. 8A), or co-administration of GW (10 μ M) and L-arginine (Fig. 8B), abolished L-arginine-induced enhancement of Ca 2+ oscillations (Fig. 8C,D), suggesting that selective activation of acinar cell CB 2 Rs significantly eliminates L-arginine-induced enhancement of intracellular Ca 2+ signals in mouse pancreatic acinar cells.

GW Improves L-arginine-Induced Pathology. Finally, we tested whether systemic injection of GW
can prevent L-arginine-induced elevation of Ca 2+ oscillations, and subsequent pathological changes including enhancement of pancreatic amylases (AMS) and pulmonary peritoneal macrophages (MPO) levels, which are two major effects present in early-stage of acute pancreatitis. We injected L-arginine (4.0 g/kg, i.p.) to establish an acute pancreatitis model 23,24 , and dissociated pancreatic acinar cells 24 hours later, then compared ACh-induced Ca 2+ oscillations between saline-and L-arginine-treated groups using Ca 2+ imaging. Systemic L-arginine injection enhanced ACh-induced Ca 2+ oscillations compared to systemic saline injection, but GW and L-arginine co-injected showed similar level of ACh-induced Ca 2+ oscillations (Fig. 9A). Compared to the ACh-induced Ca 2+ oscillations in saline-treated mice, the acinar cells prepared from L-arginine-treated mice showed a significant increase in Ca 2+ oscillation response (saline vs. L-arginine group, p < 0.01), while co-injection of GW and L-arginine reduced L-arginine's effect (saline vs. L-arginine + GW group, p > 0.05). These results suggest that the activation of pancreatic acinar cell CB 2 Rs may prevent early pathogenesis of acute pancreatitis through the inhibition of enhanced intracellular Ca 2+ signals. In addition, co-injection of GW (10 mg/kg, i.p.) and L-arginine (4 g/kg, i.p.) also significantly reduced pancreatic L-arginine-induced enhancement of AMS (saline vs. L-arginine, p < 0.05, and saline vs. GW + L-arginine, p > 0.05; Fig. 9C) and pulmonary MPO levels (saline vs. L-arginine, p < 0.05, and saline vs. GW + L-arginine, p > 0.05; Fig. 9D). These results suggest that the activation of pancreatic acinar cell CB 2 Rs may prevent early pathogenesis of acute pancreatitis through the inhibition of intracellular Ca 2+ signals.

Discussion
The novel findings of this study are that the activation of membrane CB 2 Rs by GW reduces ACh-, but not CCK-induced intracellular Ca 2+ oscillations, and GW induced reduction of Ca 2+ oscillations in a concentration-dependent manner. The CB 2 R-mediated reduction of ACh-induced Ca 2+ oscillations is abolished by pharmacological blockade of CB 2 Rs (AM630) or is absent in CB 2 -KO mice, but not in CB 1 -KO mice. The pancreatitis inducer, L-arginine, significantly enhances ACh-induced intracellular Ca 2+ oscillations, and the CB 2 R agonist, GW, abolishes this L-arginine effect. In addition, this CB 2 R agonist also improved L-arginine-induced pathological changes. Collectively, our data demonstrate that CB 2 R agonist GW reduces ACh-enhanced intracellular Ca 2+ signals in mouse pancreatic acinar cells, and this may underlie an important cellular mechanism for a CB 2 R agonist to serve as a new candidate for treating acute pancreatitis. CB 2 R Expression in Mouse Pancreatic Acinar Cells. Previously, in rodent pancreatic acinar cells, CB 2 R protein expression was found using immunohistochemical staining and Western blot 10,25 . In mouse pancreatic tissue, both CB 1 R and CB 2 R mRNA were identified using real-time RT-PCR and immunohistochemical staining 10 . In the present study, we confirmed that CB 2 R proteins were expressed in freshly isolated mouse pancreatic acinar cells, which is consistent with previous report 10 . Our data demonstrate that CB 2 Rs are expressed in mouse pancreatic acinar cells and they may play an important role in modulating acinar cells function.

CB 2 R Agonist Reduces ACh-Induced Ca 2+ Oscillations in Mouse Pancreatic Acinar Cells.
Mouse pancreatic acinar cells have been used as an excellent cell model of agonist-induced Ca 2+ oscillations for studying pancreatitis 26 . We examined whether a selective CB 2 R agonist, GW, affected ACh-induced Ca 2+ oscillations in the isolated pancreatic acinar cells through CB 2 Rs. Using both patch-clamp recording and confocal Ca 2+ imaging techniques, we found that GW significantly reduced ACh-induced Ca 2+ oscillations, and this inhibition is GW-concentration dependent. We also tested another selective CB 2 R agonist, JWH-133, on the ACh-induced Ca 2+ oscillations, and found a similar inhibition (Supplemental Fig. 4), but the inhibitory effect of JWH-133 was weaker (a higher concentration of JWH-133 was needed compared with GW to induce the same inhibition). It was reported that GW acts as a potent and selective partial agonist for CB 2 R with an EC 50 of 0.65 nM and selectivity of around 1200× for CB 2 R over CB 1 R 27,28 , while JWH-133 has an EC 50 of 3.4 nM and selectivity of around 200× for CB 2 R over CB 1 R 29 . These findings may explain why GW is more potent than JWH-133 for ACh-induced Ca 2+ oscillations.
Accumulating evidence demonstrates a complex relationship between the cannabinoid ligand (and receptors) and intracellular Ca 2+ signals in different types of cells. For example, on one hand, activation of cannabinoid CB 1 R or CB 2 R increased (initiated) intracellular Ca 2+ levels in endothelia cells 30 , submandibular acinar cells 31 , canine kidney cells 32 , and bladder cancer cells 33 . On the other hand, in pancreatic beta cells, the activation of either CB 1 R 34 or CB 2 R 35 reduced glucose-induced intracellular Ca 2+ oscillations and insulin release. It has been reported that anandamide reduced intracellular Ca 2+ concentration through the suppression of a Na + /Ca 2+ exchanger current in rat cardiac myocytes 36 . To our knowledge, ours is the first report that a selective CB 2 R agonist reduces intracellular Ca 2+ signals in mouse pancreatic acinar cells. Considering that Ca 2+ plays an important role in cellular function, especially enzyme secretion in pancreatic acinar cells, our data suggest that CB 2 R modulates an important aspect of pancreatic acinar cell physiology and pathophysiology. CB 2 R Agonist Reduces ACh-Induced Ca 2+ Oscillations through Membrane CB 2 Rs. Cannabinoid ligands exert their pharmacological effects through CB 1 R or CB 2 R, but in some cases they also can act on non-cannabinoid targets 37 . We determined whether GW modulated intracellular Ca 2+ signals through a cell membrane or cytosolic CB 2 Rs. First, we examined the effects of pharmacological manipulations of CB 1 R and CB 2 R and found that the CB 2 R selective antagonist AM630 abolished GW-induced reduction of Ca 2+ oscillations, suggesting that GW modulates ACh-induced Ca 2+ oscillations through the CB 2 Rs. Then, we genetically manipulated cannabinoid receptors and compared the effects of GW on Ca 2+ oscillations between WT and CB 2 R-KO mice, and also WT and CB 1 R-KO mice. We found that in CB 2 R-KO but not CB 1 R-KO mice, GW lost its inhibitory effect, further confirming that CB 2 R is the key target for mediating GW-induced reduction in Ca 2+ oscillations.
In a group of cells tested, we found that a CB 1 R agonist, ACEA (dissolved by ethanol; 10-μ M ACEA solution contained 7.3-mM ethanol) reduced ACh-induced Ca 2+ oscillations (Supplemental Fig. 3); however, the control experiments using the same concentration of ethanol (7.3 mM) also reduced ACh-induced Ca 2+ oscillations, and the inhibitory effect of ACEA was not absent in the acinar cells dissociated from CB 1 R-KO mice, suggesting a non-specific effect, likely caused by ethanol. In addition, we also tested the effects of DMSO (GW was dissolved by DMSO to 100 mM stock solution), and found that 1 μ M DMSO itself did not affect ACh-induced Ca 2+ oscillations (Supplemental Fig. 5). Together, our data support the conclusion that GW selectively acts on acinar cell CB 2 Rs and reduces ACh-induced Ca 2+ oscillations. Finally, we asked where the CB 2 Rs are located (membrane or cytosolic CB 2 Rs). To address this question, we designed three experiments. We first examined the effect of bath-applied GW on the Ca 2+ oscillations induced by intracellular application of IP 3 , and found that GW did not affect IP 3 -induced Ca 2+ oscillations, suggesting that the target that mediated GW-induced inhibition in Ca 2+ oscillations is located in the signal pathway before IP 3 receptors, and not on the IP 3 receptor itself. We then intracellularly applied GW through a recording electrode to examine the effect of intracellular administration of GW on bath ACh-induced Ca 2+ oscillations, and found that intracellular infusion of GW (even at 100 μ M) did not alter ACh-induced Ca 2+ oscillations. Finally, we intracellularly applied AM630 through a recording electrode to examine the effect of bath-applied GW on ACh-induced Ca 2+ oscillations. Our data showed that intracellular infusion of AM630 did not prevent bath-applied GW-induced reduction of Ca 2+ oscillations. Collectively, our data support the conclusion that GW modulates intracellular Ca 2+ signaling through the membrane CB 2 Rs in pancreatic acinar cells.

Possible Mechanisms of GW-Induced Reduction in ACh-Induced Ca 2+
Oscillations. The precise mechanism by which GW modulates intracellular Ca 2+ signals is unclear. Our data show that membrane CB 2 Rs are necessary for mediating GW's effect. GW's action in ACh-induced Ca 2+ oscillations should occur at the G-protein-mediated signal pathway between muscarinic receptor (M 3 ) activation and IP 3 production because GW did not affect IP 3 -induced Ca 2+ oscillations. We also demonstrated that GW failed to affect ACh-induced Ca 2+ oscillations in pancreatic acinar cells prepared from CB 2 R-KO mice, suggesting that GW likely did not affect muscarinic receptor function. In addition, we found that bath-applied GW failed to inhibit CCK-induced Ca 2+ oscillations even at 100 μ M, suggesting that GW selectively modulates muscarinic receptor-mediated G-protein signaling 38 . Therefore, the possible mechanisms for GW-induced modulation of ACh-induced Ca 2+ oscillations may involve cross talk between muscarinic receptor-and CB 2 R-mediated G-protein signal pathways, such as homologous and/or heterologous desensitization of G-protein coupled receptors (GPCRs) 39 . For example, in the case of homologous desensitization of GPCRs, the activation of one type of GPCR can rapidly terminate another GPCR signaling through the internalization of receptors after binding, phosphorylation of G-protein coupled receptor kinases, and formation of complexes with β -arresting 39,40 . In addition, the activation of a GPCR may also result in temporary inhibition of another GPCR signal through a heterologous desensitization, which oscillations between the acinar cells prepared from L-Arg (4 g/kg, i.p.) and normal saline (NS) treated mice (after injection for 24 h). These demonstrate an enhancement of ACh-induced Ca 2+ oscillations in L-Argtreated mice compared to NS-treated mice. This enhanced effect is prevented by co-injection of GW and L-Arg (GW + L-Arg). (B) In these studies, we measured Ca 2+ responses as ∆ F/F 0 , where F refers to the current Fluo signal intensity, F 0 refers to the background Fluo signal intensity, and ∆ F/F 0 refers to the change of F/F 0 . Using the same measurement, we compared ACh-induced Ca 2+ responses from pancreatic acinar cells collected from three groups of mice: control (saline-treated mice), L-Arg, and L-Arg plus GW. Compared to the ACh-induced Ca 2+ oscillations in saline group, there is a significant enhancement of ACh-induced Ca 2+ oscillations in L-Arg group (* * Indicates p < 0.01), and there is no statistically significant difference between saline and GW + L-Arg groups ( # indicates p > 0.05), suggesting a prevention of L-Arg-induced enhanced effect by GW. In addition, GW also prevents L-Arg-induced elevation of pancreatic AMS (C) and pulmonary MPO (D). The number of cells tested is stated for each condition. Bars represent mean ± SEM. In parts C and D, * Indicates p < 0.05 between saline and L-Arg groups, but there is no statistically significant difference between saline and L-Arg + GW group ( # indicates p > 0.05).
does not involve receptor internalization, but activation of several signal transduction pathways, particularly protein kinase C (PKC)-and PKA-dependent signaling pathways 38,41 . It has been reported that intracellular cyclic AMP-generated substances play an important role in regulation of IP 3 and Ca 2+ responses to ACh in rat submandibular acini. Investigators found that intracellular cAMP increased IP 3 formation in response to ACh, while blocking PKA by H89 reduced IP 3 formation 41 . Because it is well known that the activation of CB 2 Rs significantly reduces intracellular cAMP levels, we thus postulated that GW may activate CB 2 Rs, reduce cAMP, and in turn reduce intracellular IP 3 production, and lead to a reduction of ACh-induced Ca 2+ oscillations. Our findings warrant further testing of this hypothesis.
Clinical Significance of CB 2 R-Mediated Reduction of Ca 2+ Oscillations in Pancreatic Acinar Cells. Pancreatic acinar cells are functional units of the exocrine pancreas. They synthesize, store, and secrete inactive preforms of digestive enzymes into the lumen of the acinus. The activity of pancreatic acinar cells is crucially modulated by the secretagogues ACh and CCK; both can act on their specific membrane receptors (muscarinic and CCK receptor, respectively) and then induce an elevation in cytoplasmic calcium. If high concentrations of intracellular Ca 2+ persist, intracellular signaling is disrupted, cell damage occurs, and acute pancreatitis forms. Emerging evidence suggests that the earliest abnormalities of acute pancreatitis arise by aberrant elevation of intracellular Ca 2+ within acinar cells because the sustained intracellular Ca 2+ elevation activates intracellular digestive proenzymes resulting in necrosis and inflammation, and pharmacological blockade of store-operated or Ca 2+ release-activated Ca 2+ channels would prevent sustained elevation of intracellular Ca 2+ , and consequence protease activation and necrosis 3 . In the present study, we provide the first evidence that the CB 2 R agonist, GW, reduces ACh-induced Ca 2+ oscillations, abolishes L-arginine-induced enhancement of Ca 2+ oscillations and prevents L-arginine-induced elevation of both pancreatic AMS and pulmonary MPO levels. These results suggest that a CB 2 R agonist may serve as a novel therapeutic strategy to prevent and/or treat acute pancreatitis. This conclusion is consistent with previous report that a CB 2 R agonist exhibits a protective effect on pathogenesis in an acute pancreatitis animal model 10 . Our data showing a reduction of intracellular Ca 2+ signaling by GW also provide a new target to interpret the role of CB 2 R agonists in treating acute pancreatitis in addition to CB 2 R-mediated anti-inflammation.