The endocannabinoid 2-AG controls skeletal muscle cell differentiation via CB1 receptor-dependent inhibition of K_v7 channels

Levels of AEA and 2-AG and mRNA Expression Profile of Genes Related to Their Metabolism During C₂C₁₂ Cell Differentiation.

To gain information about the tone of the ECS and its potential changes during in vitro myogenesis, we measured AEA and 2-AG levels in C₂C₁₂ cells by liquid chromatography atmospheric pressure chemical ionization mass spectrometry (LC-APCI-MS) in myoblasts and after 3 and 7 d of their exposure to differentiation medium (DM). We found that, during myotube formation, AEA levels tended to decrease albeit not reaching statistical significance (Fig. 1A), whereas 2-AG levels strongly declined already after 3 d of myoblast exposure to DM and remained reduced up to at least day 7 of the differentiation program (Fig. 1A). To understand whether the changes in EC levels were due to changes in the mRNA expression of EC metabolic enzymes, we determined by quantitative PCR (qPCR) the mRNA expression profile for the entire set of genes known to encode for enzymes involved in EC synthesis (such as Napepld, Abdh4, Gde1, and Ptpn22 for AEA or Dagla and Daglb for 2-AG) and degradation (such as Faah for AEA or Magl, Abdh6/12, and Faah for 2-AG) (Table S1). Our analysis revealed that during myotube formation the expression profile of genes involved in AEA metabolism was not significantly changed, except for Ptpn22 and Faah, the mRNA transcript levels of which were slightly decreased in myotubes (Fig. 1B). In contrast, among the genes related to 2-AG metabolism, the expression of Dagla and Abhd12 was slightly reduced (by less than twofold), whereas Magl expression was significantly up-regulated (by about twofold) during C₂C₁₂ cell differentiation (Fig. 1B). Accordingly, Western blot analysis revealed that, concomitant to increased myosin heavy chain (MyHC) protein expression during C₂C₁₂ cell differentiation, Dagla levels are not significantly changed, although a decreasing trend is observed at 72 h, whereas Magl levels increase steadily over time (Fig. 1C).

Potential changes in AEA and 2-AG levels were then evaluated in mouse muscle at different developmental stages. The quadriceps femoris muscle was dissected from the leg of mouse embryos or pups at embryonic day 18 (E18) or at postnatal days 4 and 14 (P4 and P14), and its purified lipid extract was subsequently subjected to LC-APCI-MS analysis. As shown in Fig. 1D, endogenous levels of AEA decreased by about fivefold from E18 to P4, but rebound by P14, whereas 2-AG levels remained constant at E18 and P4, but then decreased by about 50% at P14. Taken together, these results indicate that during muscle formation, both in vitro and in vivo, the endogenous levels of 2-AG decline with myotube formation and muscle growth, respectively.

CB1 and CB2 Expression Profile During Murine C₂C₁₂ and Primary Human Skeletal Muscle Cell Differentiation.

In our attempt to characterize the ECS and its role during skeletal muscle cell differentiation, the expression of transcripts from the Cnr1 and Cnr2 genes, encoding for CB1 and CB2 receptors, respectively, was also evaluated in C₂C₁₂ myoblasts [in growth medium (GM)] and at different time points (1–3 d) during myotube formation [in differentiation medium (DM)], by qPCR. The results obtained show that Cnr1 expression was already strong in proliferating myoblasts and further increased after 1 d of exposure to DM, reaching a plateau after 2 d (Fig. 2A). In contrast, Cnr2 expression in myoblasts (GM) was much lower, and did not significantly increase after 1 d of exposure to DM; however, an increase in Cnr2 mRNA levels was found after 2 d of cell exposure to DM, and this continued after 3 d (Fig. 2A). Likewise, in primary human skeletal muscle cells (hSkMC), we found that CNR1 mRNA expression was increased by about eightfold after 2 d of myoblast exposure to DM. In contrast, CNR2 gene expression was not modified (Fig. 2B). In C₂C₁₂ cells, in keeping with the observed Cnr1 and Cnr2 mRNA expression profile, Western blot analysis showed a >4-fold increase in CB1 receptor expression, with much smaller effects on CB2 (Fig. 2C).

In conclusion, these findings indicate that both Cnr1 and Cnr2 genes are expressed in murine C₂C₁₂ cells, with Cnr1 being the most expressed and more rapidly up-regulated within the first 24–48 h of myotube formation. Down-regulation of Cnr1 was instead observed in vivo in mice, for example, in the quadriceps femoris, during development from E18 to P4 (see below).

Effects of AEA, 2-AG, and Noladin Ether on the Early Phase of Myotube Formation.

To clarify the biological significance of the changes described so far, C₂C₁₂ myoblasts were induced to differentiate in the presence of exogenous AEA or 2-AG or of drugs able to selectively modulate EC levels or activity at receptors. After 1 d of exposure to DM (± drugs), total mRNA was isolated under each condition and analyzed by qPCR to examine the expression levels of specific skeletal muscle differentiation-related genes, such as myogenin (Myog), troponin T1 (Tnnt-1), and Krp1 (Sarcosin), which are normally up-regulated during myotube formation in C₂C₁₂ cells (Fig. S1).

As shown in Fig. 3A, exogenous 2-AG (1–3 μM), but not AEA (1–3 μM), significantly inhibited Myog mRNA expression. However, considering that FAAH and MAGL are active also on synthetic AEA and 2-AG, we evaluated the effects of these exogenous ECs in the presence of selective DAGLα, FAAH, and MAGL inhibitors, i.e., OMDM188 (1 μM), URB567 (0.6–1 μM), and JZL184 (0.6–1 μM), respectively. Inhibiting endogenous 2-AG biosynthesis or degradation with OMDM188 and JZL184 (1 μM) per se slightly enhanced and reduced Myog transcript levels, respectively, suggesting the presence of an endogenous tone of 2-AG exerting an inhibitory effect on myogenesis (Fig. 3B). Indeed, exogenous 2-AG (3 μM) when provided to C₂C₁₂ cells together with a noneffective dose of JZL184 (0.6 μM) produced a stronger inhibition of Myog expression (Fig. 3B). By contrast, exogenous AEA (1–3 μM) given alone or in combination with a noneffective dose of URB567 (0.6 μM) did not alter Myog mRNA expression. URB567 at 1 μM also produced no effect on Myog mRNA expression (Fig. 3B). Moreover, 2-arachidonylglyceryl ether (noladin ether), a nonhydrolyzable structural analog of 2-AG showing high affinity for CB1 (22, 23), inhibited in a more potent and dose-dependent manner (EC₅₀ = 0.52 μM) the expression of Myog (Fig. 3C) as well as of Tnnt-1 and Krp1 mRNA (Fig. S2A).

In summary, activation of CB1 receptors with either exogenous or endogenously elevated (2-AG but not AEA) ECs, or a more enzymatically stable 2-AG analog, inhibits C₂C₁₂ myoblast differentiation, whereas inhibition of 2-AG biosynthesis produces the opposite effect.

Pharmacological and Genetic Manipulation of CB1 and CB2 Receptors During C₂C₁₂ Cell and Primary hSkMC Differentiation.

To clarify the role played by CB1 and CB2 receptors in myogenesis, we exposed C₂C₁₂ cells to DM for 24–72 h in the presence of selective CB1 or CB2 agonists or inverse agonists. Arachidonoyl-2-chloroethylamide (ACEA), a CB1 agonist selective against CB2 receptors (24), significantly reduced the mRNA expression levels of Myog in a dose (up to 3 μM) and time-dependent manner (up to 3 d of DM exposure) (Fig. 4 A and B). Also the mRNA expression levels of Tnnt-1 and Krp1 were reduced after C₂C₁₂ cell exposure to ACEA (1–3 μM) (Fig. S2B).

A decrease in Myog mRNA expression was also found after 24 h of C₂C₁₂ cell exposure to DM in the presence of the CB1/CB2 receptor agonist WIN 55,212–2 (up to 3 μM; EC₅₀ = 0.68 μM) (Fig. S2C). Interestingly, SR141716 (rimonabant) and AM251, two selective antagonists/inverse agonists at CB1 receptors (25, 26), enhanced the mRNA expression levels of Myog in a dose-dependent manner (up to 3 μM; EC₅₀ = ∼0.8 μM for both drugs) (Fig. 4 C and D) as well as those of Tnnt-1 (Fig. S2C) after 24 h of exposure. Finally, ACEA effects were counteracted by a noneffective dose of SR141716 (0.6 μM) (Fig. 4E).

Interestingly, also in hSkMCs, we found that MYOG and TNNT1 transcript levels were reduced after 2 d of myoblast exposure to DM (2% HS) in the presence of ACEA (1 μM) or 2-AG (Fig. 4F). Conversely, both MYOG and TNNT-1 mRNA expression was slightly increased in differentiating hSkMCs after 48 h of exposure to DM in the presence of 3 μM AM251 (Fig. 4F).

To further clarify the role played by CB1 during myotube formation, we transiently overexpressed the CNR1 gene in C₂C₁₂ myoblasts, and the day following the transfection we exposed the cells to DM for 2 d. After this incubation, qPCR analysis revealed that Myog and Tnnt-1 mRNA levels were reduced by about 50% in differentiating C₂C₁₂ cells overexpressing CB1 compared with control cells transfected with the empty plasmid (Fig. 4G). As predicted, in C₂C₁₂ cells overexpressing CB1, the inhibitory effect of ACEA (1 µM) on Myog and Tnnt-1 expression was stronger than that found in nontransfected or control plasmid-transfected cells (Fig. 4 A and G).

Next, we used also the opposite strategy of inducing posttranscriptional Cnr1 gene silencing by using siRNA sequences. These were transfected into the myoblasts, and the day after the transfection, cells were exposed to DM for 48 h. Under these conditions, qPCR analysis could not detect any significant change in Myog and Tnnt-1 expression in siRNA-Cnr1 cells compared with control cells (transfected with scrambled siRNA sequences), although Myog expression showed a trend towards up-regulation (Fig. 4H). Importantly however, in CB1-silenced cells exposed to DM, the inhibitory effect of ACEA (1 µM) on Myog and Tnnt-1 expression was completely lost, thus arguing strongly in favor of a CB1-mediated mechanism.

The selective CB2 receptor agonist JWH-133 (27) also reduced Myog mRNA expression at the dose of 3 μM. However, this effect was only partially reversed in the presence of the selective CB2 antagonist/inverse agonist AM630 (28), which per se (1 μM) did not change Myog mRNA expression (Fig. S3 A and B).

Thus, taking into consideration the Cnr1 and Cnr2 gene expression profiles and the pharmacological and biochemical evidence described so far, it appears that the ECS is regulated during myogenesis in a way such that CB1 receptor expression and stimulation exerts an inhibitory effect on the differentiation process in both murine C₂C₁₂ cells and hSkMC.

Morphological Analysis of Myotube Formation During CB1 Stimulation in C₂C₁₂ Cells.

Given the above described inhibitory effects of CB1 stimulation on the expression of early markers of myogenesis (Myog, Tnnt-1), we carried out a morphological analysis to evaluate the effects of prolonged exposure to 2-AG and ACEA on myotube formation by staining C₂C₁₂ cells after 4 d of exposure to DM (±drugs) with an antibody directed against the MyHC protein. MyHC is a widely used marker of mature myotube formation (29). DAPI was also used to stain the nuclei. After 4 d of DM in the presence of vehicle (DMSO), myoblasts fused to form myotubes (Fig. 5 A, a and b). By contrast, when the DM contained 2-AG (3 μM) or ACEA (3 μM), myotube formation was strongly impeded as revealed by the decreased number and size of MyHC red-positive myotubes detected (Fig. 5 A, c and d). Quantification of these pharmacological effects, using the percentage of loss of fused polynucleated myotubes (fusion index), revealed a significant impairment in C₂C₁₂ myotube formation with 2-AG or ACEA treatment (Fig. 5A, Lower).

Next, we investigated the 2-AG effect on primary satellite cells isolated from human muscle specimens from four different donors. Normally, 6–7 d of DM exposure represents the minimum period required to begin fusion in primary satellite cells, whereas at least 1 month is necessary to form mature myotubes (30). To evaluate the effect of 2-AG on the early phase of satellite cell differentiation, after 7 d of their exposure to DM, we evaluated the fusion index by staining the cells with an antibody against desmin, a canonical skeletal differentiation marker (30), and DAPI. We found that differentiating human satellite cells treated with 2-AG (1–10 μM), after 7 d of exposure to DM, exhibit a significant and dose-dependent decrease of fused, desmin-positive cells (Fig. 5B).

In an attempt to determine the relevance of CB1 to myogenesis in vivo, we examined the gastrocnemius muscles of Cnr1^−/− mice at different developmental time points (Fig. 6). At E18, the number of MyHC fibrils per muscle fascicle was increased in Cnr1^−/− mouse tissues. At this age, we also observed a coincident increase in the number of nuclei per primary fascicle (Fig. 6 A, B, F1, and F2), whereas the cross-sectional area of the individual fibers remained constant. However, the ratio between the number of nuclei and the number of muscle fibers per muscle fascicle (Fig. 6F3) remained unchanged between wild-type and Cnr1^−/− mice. We suggest that these data provide genetic evidence of increased differentiation in topologically disorganized muscle fascicles with Cnr1 deletion.

By P5, knockout of Cnr1 was associated with a marked increase in MyHC and increased myofibril diameter, although not in the cross-sectional area (Fig. 6 C and D). Similar to the results of our analysis at E18, the muscle fiber/nucleus ratio in Cnr1^−/− mice remained unchanged relative to wild-type littermates, suggesting a primary effect of the genetic manipulation on muscle differentiation rather than on cell proliferation (Fig. 6 G1–G4). The qPCR analysis of muscle samples at E18 and P5 confirmed the presence of Cnr1 expression in wild-type controls and its absence in Cnr1^−/− littermates. Cnr2 levels remained unchanged. The expression levels of the muscle marker MyHC showed no difference between wild-type and Cnr1^−/− mice, although an increasing trend was observed at P5 in Cnr1^−/− mice (Fig. 6H).

Pharmacological, Electrophysiological, and Biochemical Evidence for a PIP2-Mediated Functional Interaction Between CB1 Receptors and K_v7 Potassium Channels in C₂C₁₂ Cells.

Given that ion channels, and K⁺ channels in particular, play a critical role in skeletal myogenesis, and in consideration of the permissive role in this process recently attributed to K_v7.4 channels (13, 14), we investigated the expression profile of both Cnr1 and K_v7.4 in vivo during muscle development. Toward this aim, we dissected the quadriceps femoris from mice at E18 and P4 and subjected them to qPCR analysis. We observed a robust increase in MyHC mRNA expression from E18 to P4, indicating increased muscle development. By P4, Cnr1 transcript levels were decreased (about 50%) compared with embryonic expression, whereas, in contrast, muscle K_v7.4 mRNA expression was increased by about 12-fold (Fig. S4). These data indicate that in vivo myogenesis is associated with a down-regulation of CB1 expression and an up-regulation of K_v7.4.

To explore the possible involvement of K_v7.4 channels in the antimyogenic effects of CB1 activation in C₂C₁₂ myoblasts, a pharmacological approach was first used, in which C₂C₁₂ cells were exposed to DM for 48 h in the absence or presence of the CB1 receptor agonist ACEA and the selective blocker of K_v7 channels XE991 (31); differentiation was assessed by qPCR quantification of myogenic marker transcript levels. When differentiating C₂C₁₂ cells were exposed to either ACEA (3 μM) or XE991 (60 μM), the mRNA expression levels of Myog and Tnnt-1 were similarly reduced by about 50% (Fig. S4). When the cells were exposed to the two drugs together, their effects were not additive (P > 0.05 vs. ACEA 3-μM group), thus arguing in favor of a common molecular mechanism of action for the two drugs (Fig. S4).

To gain direct evidence of the possible consequences of CB1 receptor activation on K_v7.4-mediated currents, we performed electrophysiological experiments in CHO cells transiently coexpressing both K_v7.4 channels and CB1 receptors. This strategy was chosen because the K_v7.4 currents could not be recorded in native C₂C₁₂ cells (14). Bath application of ACEA (5 or 10 µM) caused a concentration-dependent inhibition of K_v7.4 currents (Fig. 7 A and D). During ACEA perfusion, the addition of the CB1 antagonist SR141716A (SR1; 0.5 µM) abolished the inhibition of K_v7.4 currents induced by the CB1 agonist (Fig. 7 A and D). Inhibition of K_v7.4 currents was also observed with the endocannabinoid 2-AG (5 µM) (Fig. 7 B and E), as well as with the classical K_v7 channel blocker XE991. PLC stimulation is a critical transduction mechanism triggered upon CB1 receptor activation in a variety of cell types (32–34). Treatment with the PLC inhibitor U73122 (1 μM) fully inhibited ACEA-induced inhibition of K_v7.4 currents [K_v7.4 current inhibition was 36 ± 5% in the presence of 10 µM ACEA (n = 11) and 6 ± 3% in the presence of ACEA plus U73122 (n = 7; P < 0.05)] (Fig. 7 C and F). Fig. 7G shows population data from a large number of cells studied under each experimental condition described; data are expressed as current density (pA/pF) relative to controls (DMSO). Altogether, these results suggest that, in a heterologous expression system, CB1 stimulation triggers PLC-mediated inhibition of K_v7.4 currents.

PIP2 levels are crucial regulators of K_v7 channels activity (19), and the activation of several receptors (muscarinic, in particular) is known to suppress K_v7 channel function by depleting membrane PIP2 levels. To test whether CB1 receptor activation also suppresses K_v7.4 currents by a PLC-dependent decrease in PIP2 levels, the endogenous levels of this phospholipid were measured by ELISA in C₂C₁₂ myoblasts after 24 h of exposure to ACEA (1–3 μM). In this experiment, we decided to use undifferentiated cells to avoid the potentially confounding effects due to ACEA-induced inhibition of cell differentiation. As shown in Fig. 8A, ACEA significantly decreased PIP2 levels by about 30–40% in C₂C₁₂ myoblasts. The effect of ACEA (1–3 μM) on PIP2 levels was also evaluated by measuring the signal intensity of the green fluorescent Bodipy-FL PIP2. In these experiments, the Bodipy-FL-PIP2 was first conjugated with the Shuttle PIP carrier and then added to the cells in serum-free medium for 10–15 min to allow its incorporation into the cell (see Materials and Methods). As shown in Fig. 8B, in myoblasts exposed to the vehicle (DMSO), Bodipy-FL-PIP2 produces a strong green fluorescent signal, which was markedly reduced following 5–10 min of ACEA (1–3 μM) exposure. The effect of ACEA was quantified by measuring the fluorescence intensity signal in randomly selected cell groups under each experimental condition (Fig. 8B, Lower). Finally, to gain conclusive evidence of a CB1-sensitive association between PIP2 and K_v7 channels, we exposed C₂C₁₂ myoblasts for 24 h to ACEA or vehicle and then assayed K_v7.4 coimmunoprecipitated PIP2 levels. ACEA (1 and 3 μM) caused a robust reduction of coimmunoprecipitated PIP2, whereas K_v7.4 expression was unchanged (Fig. 8C). Carbachol (10 μM), a canonical agonist of G_q-coupled muscarinic receptors used as a positive control, caused a similar effect. In conclusion, these findings clearly support the idea that the inhibitory effect of CB1 receptor activation on myogenesis is largely dependent on the inhibition of K_v7.4 activity through G_q/PLC-mediated PIP2 hydrolysis.

Effect of 2-AG and ACEA on C₂C₁₂ Myoblast Proliferation.

Finally, to show that the increased number of fibers and nuclei per fascicle observed in the muscle of Cnr1^−/− mice was not due to a negative effect of CB1 tone on myoblast proliferation, as suggested by the fact that the ratio between nuclei and muscle fiber was not altered in these mice, we investigated the effect of CB1 activation on myoblast proliferation in vitro. To this end, we plated and synchronized the myoblast cell cycle by exposing C₂C₁₂ cells for 18–24 h to DM with low serum and then switching them to GM supplemented with [³H]thymidine in the absence or presence of drugs for a further 24 h to measure the effects on proliferation rate. As shown in Fig. 9, 2-AG (1 μM) and ACEA (1–3 μM) increased myoblast proliferation by about 30–40%. This proproliferative effect was counteracted by SR141716 (0.6 μM). By contrast, when myoblasts were exposed to a higher concentration of SR141716 (1 μM), proliferation was reduced (Fig. 9).

We next evaluated if the proproliferative effect of ACEA was still present when the compound was given to cells in combination with pertussis toxin (PTX) or D609, selective blockers of G_i/o and PLC, respectively (35, 36). Interestingly, D609 (40 μM), but not PTX (500 ng/mL), significantly counteracted ACEA-induced proliferation of C₂C₁₂ cells, thus suggesting that the proproliferative effect observed upon CB1 stimulation, like the differentiation-restricting effect described above, might be due to a G_q/PLC-mediated mechanism. D609 as well as PTX, when given alone, did not affect myoblast proliferation (Fig. 9). [³H]Thymidine incorporation was lowest in those cells where the DM, after the cell cycle synchronization time, was not replaced with fresh GM (Fig. 9). In conclusion, CB1 receptor stimulation increases the number of precursor myoblasts in vitro in a PLC-dependent manner.

EC Signaling Is Altered During Myogenesis in Vitro and in Vivo.

During the early phases of skeletal muscle formation myoblasts withdraw from the cell cycle and fuse to form large and multinucleated myotubes, which represent the functional contractile unit in the mature myofiber (8). During this transition the coordinated action of a large number of genes including those encoding for transcriptional factors, metabolic enzymes, ion channels, and receptors, is required. In this study, we described, to our knowledge for the first time, that the ECS, which plays an important role in many physiological and pathological aspects of mammalian biology, participates also in myogenesis. To define the potential functional role played by the ECS during the early phases of skeletal muscle formation, we used murine C₂C₁₂ cells as a model of myogenesis in vitro. AEA and 2-AG levels were measured during the myoblast–myotube transition, and a transcriptomic analysis of genes known to regulate EC metabolism and activity was performed. Our data show that 2-AG, but not AEA, levels significantly decline during C₂C₁₂ myotube formation in vitro and during muscle formation in vivo. In myoblasts this might be due to an increase in MAGL expression and subsequent increased 2-AG enzymatic hydrolysis. During myoblast differentiation, we also found significant increases in the expression of both CB1 and CB2 receptors, with CB1 showing the strongest degree of up-regulation within the first 24–48 h of differentiation. Interestingly, the early phase of human primary myoblast differentiation was accompanied by CB1, but not CB2, up-regulation. Importantly, sufficient expression of CB1 mRNA and protein seems to occur also in both murine and human myoblasts, thus allowing for a possible modulation by the ECS of the early steps of myogenesis and suggesting pharmacological experiments in this direction.

Functional Role of the ECS During Skeletal Muscle Proliferation and Differentiation.

We next investigated the possible function of the observed changes in EC levels and of CB1 and CB2 gene expression by pharmacological and biochemical approaches. To increase the levels of both AEA and 2-AG during the early phase of C₂C₁₂ cell differentiation and study the potential effects on myogenesis, we used both exogenous EC and selective inhibitors of FAAH and MAGL. We found that 2-AG, but not AEA, inhibits myotube formation. A similar effect was found following exposure of differentiating C₂C₁₂ cells to the selective synthetic CB1 agonist ACEA, which markedly reduced the expression of skeletal muscle differentiation markers (Myog and Tnnt-1) in a dose- and time-dependent manner and in a way antagonized by per se inactive doses of two selective CB1 blockers. These latter compounds, when given to cells at higher doses, as well as an inhibitor of 2-AG biosynthesis, instead increased Myog and Tnnt-1 expression. These data, together with the above finding that endogenous 2-AG is produced in higher amounts in myoblasts and that an inhibitor of 2-AG degradation also counteracted myogenesis, suggest that this EC exerts a tonic negative control on myogenesis by activating CB1 receptors and that such control is released to allow myogenesis by increasing the expression of the 2-AG–inactivating enzyme. On the other hand, the up-regulation of CB1 expression immediately after the start of differentiation can be seen as a potential way of compensating for the decreased levels of 2-AG and allowing for a negative feedback control of myogenesis. This hypothesis is supported by our observation that overexpression of CB1 inhibited C₂C₁₂ cell differentiation and heightened the inhibitory effect of ACEA, which instead was abolished by CB1 receptor knockdown. On the other hand, experiments with pharmacological tools selective for CB2, together with the much lower up-regulation of the latter during myogenesis, seem to rule out an important effect for this 2-AG target in myotube formation. Importantly, CB1 negative control of differentiation was observed also when using primary human myoblasts. Furthermore, when differentiation was carried out starting from human satellite skeletal muscle cells from healthy donors, 2-AG was again capable of inhibiting differentiation.

Finally, immunohistochemical and morphological analyses of skeletal muscle from CB1 null mice at E18 and P5, confirmed that CB1 receptors clearly play a role in inhibiting myotube differentiation and muscle formation also in vivo. In fact, during muscle development, both the number of fibers and nuclei in each fascicle were increased in the absence of CB1. Two lines of evidence indicate that this morphological observation can be explained with stimulation of myoblast differentiation rather proliferation by CB1 knockout. First, the ratio between nuclei and fibers in CB1 null mice at both E18 and P5 remained constant. Second, and most importantly, we found that, in vitro, CB1 antagonism, rather than enhancing, in fact reduces (whereas CB1 agonism enhances) C₂C₁₂ myoblast proliferation, an effect that might be a mere consequence of the effect on differentiation shown here, as observed for CB1 in many cell types including neuronal stem cells, adipocytes, osteoblasts, smooth-muscle cells, etc. (37–39), and not necessarily occurring in vivo. Thus, it is likely that the observed increase in fiber number in CB1 null mice is due to the absence of 2-AG/CB1 negative control over satellite cell differentiation into myoblasts (Fig. 5B) and over myoblast differentiation into myotubes (Fig. 4), rather than to the absence of a negative control on myoblast proliferation. Such effects would be sufficient to sustain an increased muscle formation even in the absence of direct effects on myoblast proliferation.

In C₂C₁₂ Cells, CB1 Stimulation Causes PIP2 Hydrolysis.

We next investigated the molecular mechanism(s) through which CB1 controls myoblast proliferation and differentiation. CB1 belongs to the metabotropic class of receptors and, although prevalently coupled to G_i/o proteins, it can also trigger G_q/11 activation and subsequent PLC activation (32–34, 40). Indeed, we provided evidence that, in C₂C₁₂ cells, ACEA reduces the endogenous levels of the PLC substrate, PIP2. We surmise that, in differentiating myoblasts, CB1 activation leads to the activation of PLC and the subsequent hydrolysis of PIP2. Accordingly, we showed that the proproliferative effect produced by ACEA on C₂C₁₂ myoblasts is abolished by a PLC inhibitor, but not by PTX. CB1-induced PIP2 hydrolysis might also release channels involved in differentiation from the positive control that PIP2 usually exerts thereupon and, at the same time, lead to the formation of diacylglycerols and IP3, with subsequent activation of protein kinase C and intracellular calcium mobilization, all of which have been implicated in both cell differentiation and proliferation (41, 42).

CB1 Receptor Activation-Induced PIP2 Hydrolysis Leads to Reduced Function of K_v7 Potassium Channels.

K_v7 represents a subclass of K⁺ channels comprised of five distinct members (K_v7.1– K_v7.5), each showing distinct subcellular localization, biophysical properties, modulation, and pharmacological profile (43). K_v7.1 subunits, together with the auxiliary subunits KCNE1, form the molecular basis of delayed rectifier K⁺ currents, which control action potential repolarization in cardiomyocytes, whereas K_v7.2, K_v7.3, K_v7.4, and K_v7.5 give rise to K⁺ currents (such as the M current) widely distributed in neuronal and primary sensory cells and acting as primary regulators of the intrinsic electrical properties of excitable cells (44). To date, the spectrum of functions attributed to K_v7 subunits has been enlarged following their description in smooth muscle cells of vascular and visceral tissues and in skeletal muscle cells. In this context, we recently showed that significant changes in the expression levels of some members of the K_v7 subclass of voltage-gated K⁺ channels, most notably K_v7.4, accompany the myoblast-to-myotube transition in murine C₂C₁₂ cells (14).

Over the past decade, many studies have described how the decrease of membrane PIP2 concentrations represents one of the main molecular mechanisms by which G_q/11-coupled metabotropic receptors inhibit the activity of K_v7 channels. Moreover, K_v7 channel open state probability also depends on its intrinsic affinity for intracellular PIP2 (45), with higher PIP2 concentrations increasing the open state probability of K_v7 channels. In view of the present data showing that CB1 stimulation in C₂C₁₂ cells reduces differentiation and stimulates proliferation—i.e., two effects that are opposite to those observed following activation of K_v7 channels (13)—we hypothesized that CB1 receptor stimulation, by reducing PIP2 levels, might indirectly inhibit K_v7 channels. By using selective pharmacological tools and biochemical and electrophysiological approaches, we have provided strong evidence in favor of this hypothesis. In fact, we showed that CB1 stimulation reduces (i) the K_v7.4-mediated K⁺ current in a PLC-dependent manner in CHO cells heterologously expressing CB1 and K_v7.4, but not in cells expressing only K_v7.4, and (ii) the amounts of PIP2 associated with K_v7.4 channels. We also investigated the potential effect of CB1 stimulation by ACEA on the IP3-dependent intracellular Ca²⁺ mobilization, an event also known to be coupled to inhibition and stimulation of myoblast differentiation and proliferation, respectively (46). However, in FURA-2:00 AM-labeled C₂C₁₂ myoblasts exposed to DM for 24 h in a Ca²⁺-free buffer, we could observe only a slight increase of [Ca²⁺]_i in response to ACEA. Interestingly, the M1 muscarinic receptor (but not the B2 bradykinin receptor) also was reported to inhibit K_v7 in the brain in a manner similar to that observed here in myoblasts for CB1—i.e., by decreasing PIP2 binding to the channel, rather than by increasing [Ca²⁺]_i and leading to calmodulin activation (15). The role of diacylglycerol and protein kinase C in CB1-mediated inhibition of K_v7.4 might therefore be an interesting subject of investigation for future studies in myoblasts.

Conclusions.

The present study suggests that 2-AG, via CB1 activation, plays a crucial role in the control of myotube formation, and highlights K_v7.4 potassium channels as one important downstream target for this function. The novel molecular mechanism described here could open new avenues for the development of endocannabinoid-based therapies against skeletal muscle diseases characterized by abnormal repair and differentiation.

Cell Culture and Drugs.

C₂C₁₂ myoblasts were cultured and differentiated as previously described (13) (SI Materials and Methods). Satellite cells were obtained from muscle tissue from orthopedic surgery, as approved by the Ethical Committee of the University Medical School of Debrecen, and differentiated as previously described (30) (SI Materials and Methods).

Measurement of Endocannabinoids AEA and 2-AG.

The extraction, purification, and quantification of EC from tissues was performed as previously described (47) (SI Materials and Methods).

Cell Transfection and Cnr1 Gene Silencing and/or Overexpression.

Myoblasts were plated in six-well culture dishes, and the next day, 4 μg of total DNA of pcDNA3.1-CNR1, or empty vector and a plasmid encoding for the enhanced green fluorescent protein (Clontech), was transfected using Lipofectamine LTX (Life Technologies) according to the manufacturer’s instructions. The following day, C₂C₁₂ myoblasts were exposed to DM for 48 h in the presence or absence of ACEA before RNA isolation and quantitation. Cnr1 silencing was obtained by transfection of myoblasts plated in six-well culture dishes with endoribonuclease-prepared siRNA sequences (EMU088771 Sigma-Aldrich) using Lipofectamine RNAimax (Life Tecnology). The following day, the cells were exposed to DM for 48 h in the presence or absence of ACEA before RNA isolation and quantitation.

RNA Extraction and qPCR.

Total RNA isolation, purification, and cDNA synthesis were performed as described (13) (SI Materials and Methods).

Cell Proliferation Assay.

C₂C₁₂ cell proliferation was assessed as previously described (13).

Western Blots.

C₂C₁₂ cells were subjected to Western blot analysis as described previously (13). For this study we used the following antibodies: mouse anti-myosin heavy chain antibody, clone A4.1025 (1:1,000; Millipore); mouse anti-CNR1 (1:500; catalog no. Y080037, Applied Biological Materials Inc.); mouse anti-CNR2 (1:400; Sigma-Aldrich); mouse anti-MGL clone N90/20 (1:4; Antibodies Inc.); and goat anti-DAGLα lot (1:1,000; Abnova). An anti–α-tubulin antibody (1:5,000; Sigma-Aldrich) was used to check for equal protein loading. Reactive bands were detected by chemiluminescence (ECL or ECL-plus; Bio-Rad). Images were analyzed on a ChemiDoc station with Quantity-one software (Bio-Rad).

Cell or Tissue Morphological Analysis.

Murine C₂C₁₂ cells were stained as previously described (14). Primary human muscle cells were stained as previously described (30). Heterozygous crosses of CB1RKO [n = 3 (E18), n = 2 (P5)] and wild-type littermates [n = 3 (E18), n = 2 (P5)] on C57Bl5/N background were processed simultaneously. In brief, embryos were harvested on E18, and the right hind paw was isolated and immersion-fixed for 48 h. Pups were decapitated on postnatal day 5 with the right hind paw processed as above. Subsequently, tissues were cryoprotected in 30% (wt/vol) sucrose in physiological saline and sectioned at a thickness of 14 μm on a cryostat microtome (Leica). The cutting plane was perpendicular to the musculus gastrocnemius pars externa (Fig. 6 A and B) and, consequently, parallel to the musculus gastrocnemius pars interna (Fig. 6 C and D). Serial sections were collected on fluorescence-free SuperFrost⁺ glass slides, air-dried, and processed for multiple labeling histochemistry according to standard protocols (48) using a mouse anti-heavy chain myelin antibody (1:500, Millipore) antibody, Alexa560-conjugated phalloidin to reveal filamentous actin, and Hoechst 33,342 as nuclear dye. Imaging was performed on a Zeiss780 laser-scanning microscope using spectral detection for the optimal separation of fluorescence signals. Quantitative morphometry included (i) measurement of the cross-sectional surface of individual muscle fibers (μm²) (Fig. 6 A, B, F, and G), determining (ii) the number of muscle fibers (Fig. 6 F1–G2) and the number of nuclei (Fig. 6 F2–G3) within primary muscle fascicles at both ages. Moreover, we measured the cross-sectional diameter (Fig. 6G1) of muscle fibers on P5. Data were expressed as means ± SEM; *P < 0.05.

ELISA and BODY-PY PIP2.

The amount of PIP2 extracted from C₂C₁₂ cells was measured by using the PIP2 Mass ELISA kit following the manufacturer’s instructions (Echelon Biosciences Inc.). PIP2 was extracted from control or ACEA-treated cells (1–3 μM) after 24 h of drug exposure in GM. Cellular PIP2 quantities were estimated by comparing the values from the standard curve, which showed a linear relationship at the range from 0.5 to 1,000 pM concentration. Data were analyzed and graphed by using Graph-prism 5 software. The BODY-PY green fluorescent PIP2 (catalog no. C-45F6) as well as the shuttle carrier (catalog no. P-9045) were purchased from Echelon Biosciences Inc.

Immunoprecipitation.

C₂C₁₂ myoblasts were plated on 100-mm-diameter petri dishes and exposed to GM in the presence of vehicle (DMSO) or ACEA (1–3 μM) and carbachol (10 μM) for 24 h. After this time, the cells were washed twice in PBS (without Mg²⁺ and Ca²⁺) and lysed in a buffer solution (Tris⋅HCl 50 mM, NaCl 150 mM, 1% Triton X-100, EDTA 2 mM, EGTA 5 mM, and protease inhibitors mixture). Total lysates were collected and maintained under agitation for 30 min at 4 °C before centrifugation for 15 min at 15,000 × g. Supernatants were recovered and quantified by a Bio-Rad DC Protein Assay Kit. Subsequently, 0.8–1 mg of each lysate sample was used for the immunoprecipitation (IP) procedure using a Dynabeads Protein G Kit (catalog no. 10007D; Life Technology) following the manufacturer’s instructions. Mouse anti-K_v7.4 (IgG1) (catalog no. 75–082, Antibodies Inc.) was used to immunoprepitate K_v7.4 channels. IP samples were recovered after heating at 100 °C for 10 min in β-mercaptoetanol sample buffer (2×) and loaded on precast polyacrylamide gels (4–12% gradient; Bolt Bis-Tris Plus gels; Life Technologies) and then transferred to a PVDF membrane. Mouse anti-PIP2 antibody (1:1,000; catalog no. sc-53412, Santa Cruz Biotechnology) was incubated overnight at 4 °C. All of the proteins not pulled down by the anti-K_v7.4 antibody/beads complex (flow-through) were loaded (at a concentration of 50 μg) on precast polyacrylamide gels mentioned before and blotted using anti-K_v7.4 and anti–α-tubulin (1:500; Sigma-Aldrich) antibodies to verify the IP procedure.

Whole-Cell Electrophysiology.

Currents from transiently transfected CHO cells (1 d posttransfection) were recorded at room temperature by whole-cell patch-clamp using an Axopatch 200A amplifier (Molecular Devices) with glass micropipettes of 3–5 MΩ resistance. The extracellular solution contained (in mM): 138 NaCl, 2 CaCl2, 5.4 KCl, 1 MgCl2, 10 glucose, and 10 Hepes (pH 7.4) with NaOH. The pipette (intracellular) solution contained (in mM): 140 KCl, 2 MgCl₂, 10 EGTA, 10 Hepes, 5 Mg-ATP (pH 7.4) with KOH. The pCLAMP software (version 10.2; Molecular Devices) was used for data acquisition and analysis. Linear cell capacitance (C) was determined by integrating the area under the whole-cell capacity transient, evoked by short (5–10 ms) pulses from −80 to −75 mV with the whole-cell capacitance compensation circuit of the Axopatch 200A turned off. Current densities (expressed in pA/pF) were calculated as peak of the agonist-evoked currents divided by C. From a holding potential of −80 mV, currents were activated by 4-s square pulses from −100 to +40 mV in 20 mV increments, followed by a 4-s isopotential pulse to 0 mV, or by using voltage ramps in which the membrane voltage was changed from −100 to +40 mV in 3 s.

Data Analysis.

Data are expressed as means ± SEM of the given number of experiments (n). Data sets were compared by use of unpaired Student t tests or, if necessary, with one-way analysis of variance, followed by the Bonferroni’s test using GraphPad Prism 5 software. Statistically significant differences were accepted when the P value was at least <0.05. The dose–response curves were calculated using GraphPad Prism 5 software.