Ceramide 1-phosphate stimulates proliferation of C2C12 myoblasts

Abstract

Recent studies have established specific cellular functions for different bioactive sphingolipids in skeletal muscle cells. Ceramide 1-phosphate (C1P) is an important bioactive sphingolipid that has been involved in cell growth and survival. However its possible role in the regulation of muscle cell homeostasis has not been so far investigated. In this study, we show that C1P stimulates myoblast proliferation, as determined by measuring the incorporation of tritiated thymidine into DNA, and progression of the myoblasts through the cell cycle. C1P induced phosphorylation of glycogen synthase kinase-3β and the product of retinoblastoma gene, and enhanced cyclin D1 protein levels. The mitogenic action of C1P also involved activation of phosphatidylinositol 3-kinase/Akt, ERK1/2 and the mammalian target of rapamycin. These effects of C1P were independent of interaction with a putative G_i-coupled C1P receptor as pertussis toxin, which maintains G_i protein in the inactive form, did not affect C1P-stimulated myoblast proliferation. By contrast, C1P was unable to inhibit serum starvation- or staurosporine-induced apoptosis in the myoblasts, and did not affect myogenic differentiation. Collectively, these results add up to the current knowledge on cell types targeted by C1P, which so far has been mainly confined to fibroblasts and macrophages, and extend on the mechanisms by which C1P exerts its mitogenic effects. Moreover, the biological activities of C1P described in this report establish that this phosphosphingolipid may be a relevant cue in the regulation of skeletal muscle regeneration, and that C1P-metabolizing enzymes might be important targets for developing cellular therapies for treatment of skeletal muscle degenerative diseases, or tissue injury.

Highlights

► In this study we show that C1P stimulates C2C12 myoblast proliferation. ► C1P does not affect myogenic differentiation or survival of myoblasts. ► The mitogenic action of C1P is mediated by PI3K/Akt, ERK1/2 and mTOR activation. ► C1P effects are pertussis toxin insensitive.

1. Introduction

Skeletal muscle is mainly involved in active force production, resulting in the movement of the skeletal system. Under normal biological conditions adult skeletal muscle is an extremely stable tissue. However, upon damage due to specific diseases, trauma or strong physical exercise, skeletal muscle exhibits a remarkable capacity of self-repair. Repair and maintenance of skeletal muscle is attributed to the skeletal muscle stem cell pool, represented by the satellite cells. In response to a number of stimuli, quiescent satellite cells become activated, start to proliferate and differentiate into myoblasts, which then fuse to pre-existing myofibres to regenerate the tissue [1,2]. Although in the last decade much effort has been dedicated to the comprehension of this key biological process, improved understanding of the molecular mechanisms implicated in skeletal muscle myogenesis is critical for the development of therapies to reduce the loss of muscular mass associated with muscle degenerative diseases, injuries, or aging.

In recent years sphingolipids have emerged as highly versatile molecules acting both as structural components of the membrane lipid bilayer and precursors of powerful bioactive signaling molecules. Of note, sphingolipid mediators are capable of regulating fundamental biological processes in a variety of tissues, including skeletal muscle (for detailed reviews see [3,4]). In this regard, it was reported that the plasma membrane content of sphingomyelin (SM), which is the main sphingolipid in plasma membranes, correlates with the activation state of muscle satellite cells, supporting the concept that SM hydrolysis produces signals that are necessary for their activation [5]. A key bioactive sphingolipid metabolite is ceramide, which can be formed via de novo synthesis or through activation of sphingomyelinases and the subsequent degradation of SM. Ceramide appears to be negatively associated with myoblast differentiation, since inhibition of its de novo synthesis enhances the onset of the differentiated phenotype in cultured rat myoblasts [6]. Moreover, a number of recent studies support the view that sphingosine 1-phosphate (S1P), another powerful sphingoid mediator that can be generated by further metabolism of ceramide, exerts a critical role in myogenesis. Indeed, S1P has been found to mediate the entry of satellite cells into the cell cycle, suggesting that the accelerated degradation of SM observed upon cell activation leads to S1P production [7]. In addition, challenging cultured C2C12 myoblasts with S1P enhanced their differentiation into myotubes without affecting their proliferative state [8]. In agreement with this observation, sphingosine kinase, the enzyme that produces S1P, has been described as pro-myogenic for myoblasts [9], and was implicated in cytokine-regulated myogenic differentiation [10].

Another key sphingolipid metabolite is ceramide 1-phosphate (C1P), which is generated via direct phosphorylation of ceramide by ceramide kinase (CERK). It was first demonstrated that C1P stimulated DNA synthesis in cultured rat fibroblasts [11,12] as well as in mouse macrophages [13]. Additionally, macrophage motility and survival are robustly enhanced by treatment with C1P [14,15]. Also, this sphingolipid plays a key role in inflammatory responses. In particular, C1P binds to and activates cytosolic phospholipase A₂, thus favoring prostanoid biosynthesis [16].

In this study, we demonstrate for the first time that C1P stimulates myoblast proliferation, an action that may be essential for controlling muscle regeneration. This process implicates activation of a whole panel of signal transduction pathways including phosphatidylinositol 3-kinase (PI3K)/Akt, ERK1/2, and mammalian target of rapamycin (mTOR). Furthermore, although C1P promotes macrophage survival, we here demonstrate that it does not inhibit myoblast apoptosis, and that it is not involved in myoblast differentiation. This study highlights skeletal muscle as a novel target tissue for C1P, and suggests that this phosphosphingolipid may play an important role in skeletal muscle regeneration.

2. Materials and methods

2.1. Materials

Biochemicals, cell culture reagents, Dulbecco’s modified Eagle’s medium (DMEM), fetal calf serum (FCS), protease inhibitor cocktail, monoclonal anti-skeletal fast myosin heavy chain (MHC) (clone MY-32), bovine serum albumin (BSA), staurosporine, LPA (synthesized l-α-monooleoyl phosphatidic acid), C8-C1P, C16-C1P and natural C1P were purchased from Sigma (St. Louis, MO, USA). Mouse skeletal muscle C2C12 cells were obtained from the American Type Culture Collection, (Manassas, VA, USA). The ceramide-d-erythro-1-phosphate [N-stearoyl-9,10-³H] (15 μM) was obtained from American Radiolabeled Chemicals (Saint Louis, MO, USA) and [γ-³³P]ATP was from Perkin Elmer (Waltham, MA, USA). Propidium iodide and the ceramide kinase antibody were from Calbiochem (San Diego, CA, USA). Pertussis toxin (PTx) was obtained from Alexis Corporation (San Diego, CA, USA). Monoclonal anti-caveolin-3 was from Transduction Laboratories (Lexington, KY, USA). LY294002 hydrochloride, 10-DEBC hydrochloride, PD98059, U0126 and rapamycin were from Tocris Cookson Limited (Bristol, UK). TGX-221 was a kind gift of Prof. E. Hirsch (University of Turin, Italy). Antibodies against myogenin (F5D), β-actin (C-11), cyclin D1 (72-13G), phosphorylated product of retinoblastoma gene (pRb) (Ser 795), phosphorylated glycogen synthase kinase-3β (GSK-3β) (Ser 9), pan-Akt (H-136), phospho-pan-Akt (Ser 473), anti-mouse, anti-rabbit, and anti-goat immunoglobulin G1 conjugated to horseradish peroxidase and Blotto (nonfat dry milk) were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Phospho-ERK1/2, pan ERK1/2, phosphorylated mTOR (Ser 2448) antibodies were from Cell Signaling Technology, Inc. (Beverly, MA, USA). Enhanced chemiluminescence (ECL) reagents was obtained from Amersham Pharmacia Biotech (Uppsala, Sweden). [³H]thymidine (20 Ci/mmol) was from Perkin Elmer (Waltham, MA, USA). Fluorescein-conjugated horse anti-mouse secondary antibody was obtained from Vector (Burlingame, CA, USA). The caspase-3 substrate Ac-DEVD-AFC was from Biomol Research Laboratories Inc. (PA, USA). Annexin V-FITC apoptosis detection kit and phosphorylated anti-GSK-3β (Tyr 216) were supplied by BD Biosciences (San Jose, CA, USA). Specific CERK inhibitor NVP-231 [17] and the related inactive compound 4 were kindly provided by Dr. F. Bornancin.

2.2. Cell culture

C2C12 mouse myoblasts were routinely grown in DMEM supplemented with 10% FCS. For proliferation experiments, cells were seeded in 12- or 6-well plates and used when ∼40% confluent. For differentiation experiments, cells were seeded in 6-well plates and, when confluent, were shifted to DMEM without serum containing 1 mg/ml BSA. For apoptosis experiments myoblasts were seeded in 6-well plates and used when subconfluent.

2.3. Delivery of C1P to cells in culture

An aqueous dispersion of long-chain natural C1P (from bovine brain, mainly containing stearic and nervonic fatty acids) was added to cultured myoblasts as previously described [11]. Stock solutions were prepared by sonicating C1P (1 mg) in sterile nanopure water (1 ml) on ice using a probe sonicator until a clear dispersion was obtained. Final concentration of the stock solution was approximately 1.47 mM.

2.4. Cell treatment with inhibitors

Specific inhibitors of PI3K, Akt, MEK and mTOR were administered to the cells 30 min before agonist addition. Preliminarily, the specific effect of various inhibitors was tested by Western Blot analysis, incubating lysates from LY294002-, 10-DEBC-, TGX-221-, PD98059-, U0126- or rapamycin-treated cells with anti-ERK1/2, anti-phospho-ERK1/2, anti-Akt, anti-phospho-Akt or anti-phospho-mTOR antibodies to verify the actual inhibition of the kinase activity. To investigate G_i-coupled events, C2C12 cells were incubated with 0.5 μg/ml PTx for 16 h before C1P challenge.

2.5. Cell proliferation measurement

To evaluate [³H]thymidine incorporation, proliferating myoblasts, previously serum-starved for 24 h, were challenged with various concentrations of C1P for 16 h. In some instances cells were pre-incubated for 30 min in the presence of appropriate inhibitors before being stimulated with C1P for 16 h. [³H]thymidine (0.5 μCi/well) was added for the last 2 h of incubation. Cells were washed twice in ice-cold PBS before addition with 500 μl 10% trichloroacetic acid for 5 min at 4 °C. Cells were washed again in ice-cold PBS, and 250 μl of ethanol:ether (3:1 v/v) was added to the insoluble material. Samples were then lysed in 0.25 N NaOH for 1 h at 37 °C. Incorporation of [³H]thymidine was measured by scintillation counting.

Alternatively, proliferation was evaluated by cell counting. Briefly, proliferating myoblasts were serum-starved for 24 h and then challenged with 15 μM C1P for 24 h before being trypsinized and counted by a hemocytometer.

2.6. Cell cycle analysis

To investigate the cell cycle distribution, C2C12 myoblasts were challenged with C1P for 16 h. Cell cycle distribution was determined by the propidium iodide-hypotonic citrate method with a FACSCanto instrument (Becton–Dickinson, San Jose, CA, USA) essentially as previously described [18]. Briefly, cells were washed twice in PBS and collected by scraping in a solution containing 50 μg/ml propidium iodide, 0.1% sodium citrate and 0.1% Nonidet, and analyzed 30 min later. The software used was ModFit LT for Windows, Proliferation Protocol, Verity Software House Inc. (Topsham, ME, USA).

2.7. Western blot analysis

Cells were lysed for 30 min at 4 °C in a buffer containing 50 mM Tris, pH 7.5, 120 mM NaCl, 1 mM EDTA, 6 mM EGTA, 15 mM Na₄P₂O₇, 20 mM NaF, 1% Nonidet, and protease inhibitor cocktail (1.04 mM AEBSF, 0.08 μM aprotinin, 0.02 mM leupeptin, 0.04 mM bestatin, 15 μM pepstatin A, 14 μM E-64) essentially as described [8]. To prepare total cell lysates, cell extracts were centrifuged for 15 min at 10,000 × g at 4 °C. Protein aliquots (30 μg) from lysates were resuspended in Laemmli’s sodium dodecylsulfate- (SDS) sample buffer. Samples were subjected to SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and Western analysis as previously described [19]. Bound antibodies were detected using ECL reagents.

2.8. Cell immunofluorescence assay

Cells were seeded on microscope slides, pre-coated with 2% gelatine, and then treated or not with C1P. After 72 h cells were fixed in 2% paraformaldehyde in PBS for 20 min and permeabilized in 0.1% Triton X-100-PBS for 30 min. Cells were then blocked in 3% BSA for 1 h and incubated with anti-MHC antibody for 2 h and fluorescein-conjugated anti-mouse secondary antibody for 1 h. To stain nuclei, the specimen was incubated with 50 μg/ml propidium iodide in PBS for 15 min. Images were obtained using a Leica SP5 laser scanning confocal microscope with 40× objective. To quantify the fusion of C2C12 cells after treatments, we calculated the fusion index as the average number of nuclei in MHC-positive cells with at least three nuclei above total number of nuclei.

2.9. Measurement of apoptosis

C2C12 myoblasts were seeded at a density of approximately 1 × 10⁵ cells/well and employed for experiments after 24 h. For serum starvation-induced apoptosis, cells were incubated in serum-free medium for 24 h. In these experiments, C1P was administered 30 min and 18 h after serum starvation. Staurosporine (0.5 μM) was added for the last 4 h of incubation to cells serum-starved for 24 h, treated or not at 30 min and 18 h incubation with C1P. Thapsigargin (3 μM) or etoposide (200 μM) were added for the last 8 h of incubation to cells serum-starved for 24 h, treated or not at 30 min and 18 h incubation with C1P. To measure caspase-3 activity cells were washed twice with PBS and then lysed for 20 min at 4 °C in 20 mM Tris–HCl buffer, pH 7.4, containing 250 mM NaCl, 2 mM EDTA, 0.1% Triton X-100, 5 μg/ml aprotinin, 5 μg/ml leupeptin, 0.5 mM phenylmethylsulfonyl fluoride, 4 mM sodium vanadate, and 1 mM dithiothreitol (DTT) essentially as previously described [20]. Cell lysis was completed by sonication, and the total protein content was determined using the Coomassie Blue reagent. Aliquots of protein (50 μg) were diluted in 50 mM HEPES-KOH buffer (pH 7.0) containing 10% glycerol, 0.1% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propane sulfonate, 2 mM EDTA, and 10 mM DTT. Caspase-3 activity was determined by incubating protein samples for 2 h at 37 °C with the fluorescence probe Ac-DEVD-AFC (30 μM) (excitation 400 nm, emission 505 nm) as previously described [21]. To determine non specific substrate degradation, the assays were also performed by preincubating total protein samples for 15 min at 37 °C with or without the specific caspase inhibitor (200 nM Ac-DEVD-CHO) before substrate addition. Cell apoptosis was also measured by using an annexin V-FITC apoptosis detection kit according to the manufacturer’s instructions (BD Biosciences). With this procedure, healthy cells remained unstained whereas annexin V-FITC stained early apoptotic cells, propidium iodide and annexin V-FITC stained late apoptotic cells and propidium iodide stained necrotic cells. Samples were analyzed by flow cytometry with an air-cooled 488 nm argon-ion laser (FACSCalibur, BD Biosciences) and CellQuest software (BD Bioscences), essentially as described [22].

2.10. Evaluation of C1P metabolism in C2C12 cells

To examine C1P metabolism, C2C12 cells were treated with 50 nCi/ml of ceramide-d-erythro-1-phosphate [N-stearoyl-9,10-³H] (15 μM). Then the cells were collected at different time points to chase for the possible generation of C1P metabolites, mainly ceramide, fatty acids, or sphingomyelin. The possible generation of sphingosine 1-phosphate was evaluated using [³³P]C1P (50 nCi/ml, 15 μM) which was synthesized according to a previously established procedure [23]. Cells were washed twice with ice-cold calcium-free PBS and scraped into 0.5 ml methanol. They were then washed with a further 0.5 ml methanol, and the two methanol samples were combined and mixed with 0.5 ml chloroform. Lipids were extracted by separation of phases with a further 0.5 ml chloroform and 0.9 ml of a solution containing 2 M KCl and 0.2 M HCl. Chloroform phases were dried under a stream of nitrogen and lipids were separated by thin-layer chromatography using silica gel 60-coated glass plates. The plates were developed in a three solvent system containing chloroform/methanol/ammonium hydroxide (65:35:7.5, v/v/v); chloroform/methanol/acetic acid (9:1:1, v/v/v); and butanol/acetic acid/water (3:1:1, v/v/v). The position of the different lipids was identified after staining with I₂ vapor by comparison with authentic standards. Radioactivity was quantified by scraping the spots from the plates by liquid scintillation counting.

This work has been carried out in accordance with Uniform Requirements for manuscripts submitted to Biomedical journals.

3. Results

3.1. C1P stimulates proliferation of C2C12 myoblasts but does not affect cell differentiation or apoptosis

The aim of the present study was to examine whether C1P could play a regulatory role of myoblast biological parameters. For this purpose, we first tested whether C1P was able to stimulate myoblast proliferation. Fig. 1 shows that C1P stimulated DNA synthesis, as determined by measuring the incorporation of [³H]thymidine into DNA in C2C12 cells. This effect was concentration-dependent and resulted to be optimum at 15 μM C1P. A similar effect on cell proliferation was observed when pure C16-C1P was employed instead of the C1P from bovine brain (see Materials and methods section) (data not shown). In this study, we also used a synthetic short-chain C1P (C8-C1P), which is more soluble in aqueous solutions than long-chain C1P. Fig. 1 shows that the DNA synthesis elicited by C8-C1P was optimum at 10 μM. Given the efficacy of all these C1P species at stimulating DNA synthesis, the natural C1P mixture was employed in all subsequent experiments.

Fig. 1.

Effect of C1P on [³H]thymidine incorporation. C2C12 myoblasts approximately 40% confluent were serum-starved for 24 h. Cells were treated for 16 h with natural C1P (solid square) or C8-C1P (open circle) at the indicated concentrations (0–30 μM). [³H]thymidine incorporation into DNA was measured as described in the Methods section. Data are means ± SEM of at least three independent experiments performed in triplicate. The effect of C1P was statistically significant by Student’s t test (*P < 0.05, **P < 0.01).

To exclude that the observed mitogenic effect exerted by C1P in the myoblasts was mediated by a bioactive metabolite generated following the addition of C1P to the cells, the metabolism of [³H]C1P was examined at various time-intervals. Results presented in Fig. 2 show that [³H]C1P was taken up by the myoblasts within minutes reaching maximal incorporation after about 60 min of incubation. Metabolism of C1P was very slow, with only traces of free ceramide, fatty acids, or SM being detected up to at least 4 h of incubation. These results are consistent with previous work by Chalfant and co-workers [24] showing that C1P was rapidly taken up by A549 cells with little metabolism to other compounds. Also, similar results were obtained in fibroblasts treated with [³H]C8-C1P or with natural long-chain C1P [11,12]. Furthermore, by using [³³P]C1P it was evident that C1P was not deacylated to S1P up to at least 4 h of incubation (data not shown). These latter results are in agreement with the well documented anti-mitogenic effect exerted by S1P in myoblasts [8,18], in contrast to the mitogenic effect of C1P observed in this work. Moreover, we found no significant amount of radioactivity in inorganic phosphate in the culture medium of cells incubated with [³³P]C1P for up to 4 h, thereby ruling out any possible formation of ceramide by dephosphorylation of [³³P]C1P, in agreement with our previous work [11,12,25].

Fig. 2.

Metabolism of labeled C1P in C2C12 cells. C2C12 myoblasts, approximately 40% confluent, were serum-starved for 24 h and then exposed to 15 μM C1P (100000 dpm/well) for the indicated times. Cell lipids were extracted and analyzed by TLC as described in the Methods section. Results are expressed as the percentage of the radioactivity present in the indicated lipids, ceramide (Cer), fatty acid (FA), C1P and sphingomyelin (SM), compared to that in total lipids and are the mean ± SEM of six independent experiments performed in triplicate.

Following up on the demonstration that the effect of C1P on DNA synthesis was elicited by C1P per se, its mitogenic action was further investigated by performing cell cycle analysis. Results presented in Fig. 3A show that in agreement with the studies on [³H]thymidine incorporation, treatment of myoblasts with 15 μM C1P approximately doubled the percentage of cells in S-phase, thereby reinforcing the notion that C1P is a potent mitogenic agent for myoblasts. Moreover, as shown in Fig. 3B, C1P was capable of increasing myoblast number, thus behaving as a complete mitogen. Interestingly, as shown in Fig. 3C, treatment of myoblasts with 15 μM C1P caused rapid phosphorylation of GSK-3β at Ser9 leading to inactivation of its kinase activity. This effect could be detected as early as 3 min and reached maximum value after about 30 min of treatment with C1P. In addition, phosphorylation of GSK-3β at Tyr 216, which is responsible for stimulation of its kinase activity, was moderately increased by C1P at 3 and 5 min, whereas it was remarkably decreased at 90–120 min. Both pRb and cyclin D1 are known to be maximally expressed throughout the G₁ phase of the cell cycle upon prevention of its GSK-3β-dependent proteolytic degradation [26]. In accordance with the observed phosphorylation and inactivation of GSK-3β, challenge with C1P provoked hyperphosphorylation of pRb and prevented cyclin D1 from being degraded in the myoblasts (Fig. 3D).

Fig. 3.

Effect of C1P on cell cycle and cell number. C2C12 myoblasts approximately 40% confluent were serum-starved for 24 h. A) To assess cell cycle distribution serum-starved C2C12 myoblasts were treated or not with 15 μM C1P for 16 h. Cellular DNA was stained with propidium iodide (50 μg/ml) and cell cycle analysis was performed by flow cytometry. Data are means ± SEM of three independent experiments. The effect of C1P was statistically significant by Student’s t test (*P < 0.05). B) Serum-starved myoblasts were stimulated with 15 μM C1P for 24 h before being counted by a hemocytometer. Data are means ± SEM of three independent experiments performed in duplicate. The effect of C1P was statistically significant by Student’s t test (*P < 0.05). C2C12 cells were incubated for the indicated times in the presence of 15 μM C1P. Cell lysates were then subjected to immunoblotting to detect phospho-GSK-3β (Ser9) and phospho-GSK-3β (Tyr216) (C), phospho-pRb and cyclin D1 (D). Equally loaded protein was checked by expression of the nonmuscle-specific β isoform of actin. Blots representative of at least three independent experiments are shown. The histograms represent band intensity of phospho-GSK-3β (Ser9), phospho-GSK-3β (Tyr216), P-pRb and cyclin D1 normalized to β-actin and reported as mean ± SEM of three independent experiments, fold change over control set as 1. The effect of C1P was statistically significant by Student’s t test (*P < 0.05).

Although myoblasts are endowed with high mitogenic potential, there are conditions when they can stop growing, initiate their differentiation program, and become myotubes, multinucleated cells. Therefore, we examined whether C1P could affect myoblast differentiation. To this end, the expression of myogenic markers such as MHC, myogenin, and caveolin-3 was evaluated by Western blotting analysis of whole cell lysates that were prepared from confluent myoblasts. Fig. 4A shows that the expression of these myogenic markers in unchallenged myoblasts was increased in a time-dependent manner as a consequence of their progressive myogenic differentiation [8]. Administration of 15 μM C1P to the myoblasts did not affect the expression of any of the myogenic markers at any time of incubation. Similar results were obtained when 5 or 30 μM C1P were used (data not shown). Moreover, as shown in the same Figure, C1P was unable to affect the fusion index, which is used to assess the extent of myotube formation as a result of the differentiation process.

Fig. 4.

Effect of C1P on C2C12 myoblast differentiation (A) and apoptosis (B). A) Confluent C2C12 myoblasts were incubated in medium supplemented with 0.1% BSA for the indicated period of time in the absence (−) or in the presence (+) of 15 μM C1P. Upper panel: Western blot analysis of myogenic marker expression. The content of myogenin, myosin heavy chain (MHC), and caveolin-3 (cav-3) was analyzed in cell lysates (30 μg) by Western Blot analysis. Equally loaded protein was checked by expression of the nonmuscle-specific β isoform of actin. A blot representative of four independent experiments with analogous results is shown. The histograms represent band intensity of MHC, myogenin, and cav-3 normalized to β-actin and reported as mean ± SEM of four independent experiments, fold change over control (time 24 h, no addition) set as 1. The effect of C1P was statistically significant by Student’s t test (*P < 0.05). Lower panel: Representative immunofluorescence images of C2C12 myoblasts treated for 72 h with 15 μM C1P stained with anti-MHC antibody and propidium iodide are shown. Fusion index represented in the histogram was calculated as described in the Methods section. Data are means ± SEM of four independent experiments. B) The apoptotic response was induced by 24 h serum starvation or 4 h treatment with 0.5 μM staurosporine in subconfluent myoblasts treated or not with 15 μM C1P. Left panel: caspase-3 activity was measured employing a fluorimetric method. Data are means ± SEM of three independent experiments performed in duplicate and are reported as fold increase in caspase-3 activity relative to the value measured in control cells set as 1. Right panel: cells were examined for necrosis or early or late apoptosis by flow cytometry analysis as described in the Methods section. Results are expressed as percentage of cells and are the mean ± SEM of three independent experiments performed in duplicate.

It is known that C1P is also capable of protecting macrophages from entering apoptosis [14,27,28]. Therefore, we examined whether C1P could exert an anti-apoptotic action in C2C12 cells incubated under apoptotic conditions (absence of serum in the culture medium or long-term incubation with staurosporin). Unexpectedly, the results presented in Fig. 4B clearly show that C1P (15 μM) was unable to reduce the extent of caspase-3 activation in serum-starved myoblasts or in cells that were treated with the pro-apoptotic agent staurosporine (0.5 μM) for 4 h. Similarly, C1P was ineffective to protect myoblasts from apoptosis induced by thapsigargin (3 μM) or by treatment with etoposide (200 μM), as assessed by measuring caspase-3 activity (data not shown). In agreement with these results, treatment with C1P did not influence the percentage of cells in early or late apoptosis induced by serum deprivation. This was assessed using cells that were labeled with propidium iodide and annexin V-FITC in combination with flow cytometry (Fig. 4B).

3.2. C1P elicits its mitogenic effect in myoblasts via stimulation of PI3K/Akt, ERK1/2 and mTOR

To gain insight into the mechanism by which C1P stimulates myoblast proliferation, we examined whether this phosphosphingolipid could activate signaling pathways specifically implicated in the regulation of cell growth. Western blotting analyses of phospho-Akt, phospho-ERK1/2 and phospho-mTOR performed on lysates obtained from myoblasts treated with 15 μM C1P for various time-intervals demonstrate that all of the investigated signaling pathways were transiently activated by C1P. In particular, Akt and mTOR phosphorylation was maximal at 5 min, while ERK1/2 were maximally activated at 10 min (Fig. 5A). To examine the possible role of these pathways in C1P-mediated myoblast proliferation, [³H]thymidine incorporation experiments were performed in myoblasts previously incubated in the presence of selective inhibitors of these pathways. As depicted in Fig. 5B, inhibition of PI3K with 5 μM LY294002 or blockade of Akt with 1 μM 10-DEBC, abolished the mitogenic action of C1P; similarly, this sphingolipid was unable to stimulate myoblast proliferation when ERK1/2 activation was prevented by treatment with the MEK inhibitors UO126 (5 μM) or PD98059 (5 μM), or when mTOR activation was blocked by treatment with rapamycin (10 nM). The data presented in Fig. 5C show that inhibition of MEK/ERK1/2, PI3K/Akt, or mTOR by UO126, LY294002, or rapamycin, respectively, also blunted the enhancement of cyclin D1 levels elicited by C1P at 8 h of incubation. Given that the activation of these signaling pathways is often detected following membrane receptor(s) engagement and that C1P was shown to stimulate migration of RAW 264.7 macrophages in a PTx-sensitive manner [15], we examined whether the mitogenic action of C1P could be mediated by a G_i protein-coupled receptor. Myoblast treatment with 0.5 μg/ml PTx for 16 h did not affect the mitogenic response to 15 μM C1P (Fig. 6A), nor did it alter the extent of Akt or ERK1/2 phosphorylation (Fig. 6B). In contrast, as shown in the inset of Fig. 6A, the anti-myogenic effect of 10 μg/ml LPA, mediated through interaction with a G_i protein-coupled receptor [29], is significantly reduced by the pre-treatment with 0.5 μg/ml PTx for 16 h. Therefore, it is unlikely that C1P-stimulated myoblast growth is mediated by a receptor of this kind.

Fig. 5.

Role of C1P-induced activation of Akt, ERK1/2 and mTOR (A) on DNA synthesis (B), and cyclin D1 expression (C) in C2C12 myoblasts. C2C12 myoblasts approximately 40% confluent were serum-starved for 24 h and incubated with or without 15 μM C1P for the indicated time-intervals. A) Cell lysates were separated by SDS-PAGE and immunoblotted using specific anti-phospho-Akt, anti-pan Akt, anti-phospho-ERK1/2, anti-pan ERK1/2, anti-phospho-mTOR and anti-β-actin antibodies. Blots representative of at least three independent experiments are shown. In the histograms band intensity corresponding to phosphorylated protein was normalized to its total content or to β-actin and reported as mean ± SEM of three independent experiments, fold change over control set as 1. The effect of C1P was statistically significant by Student’s t test (*P < 0.05). B) C2C12 cells were pre-treated for 30 min with PI3K inhibitor (5 μM LY294002) or Akt inhibitor (1 μM 10-DEBC) or MEK inhibitor (5 μM U0126 or 5 μM PD98059) or mTOR inhibitor (10 nM rapamycin) before being challenged with 15 μM C1P for 16 h. [³H]thymidine incorporation into DNA was measured as described in the Methods section. Data are means ± SEM of three independent experiments performed in triplicate. The effect of C1P was statistically significant by Student’s t test (**P < 0.01); the effect of the inhibitors was statistically significant by Student’s t test (#P < 0.05). C) Western Blot analysis of cyclin D1 expression. C2C12 cells were pre-treated 30 min with PI3K/Akt specific inhibitor (5 μM LY294002) or MEK/ERK1/2 specific inhibitor (5 μM U0126) or mTOR inhibitor (10 nM rapamycin) before being challenged (+) or not (−) with 15 μM C1P for 8 h. Cell extracts were subjected to immunoblotting. Equally loaded protein was checked by expression of the nonmuscle-specific β isoform of actin. A blot representative of at least three independent experiments is shown. The histogram represents band intensity of cyclin D1 normalized to β-actin and reported as mean ± SEM of three independent experiments, fold change over control set as 1. The effect of C1P was statistically significant by Student’s t test (*P < 0.05); the effect of the inhibitors was statistically significant by Student’s t test (#P < 0.05).

Fig. 6.

Effect of PTx on C1P mitogenic effect in C2C12 myoblasts. A) C2C12 myoblasts approximately 40% confluent were serum-starved for 24 h. C2C12 cells were pre-incubated for 16 h with PTx (0.5 μg/ml) before being challenged with 15 μM C1P. [³H]thymidine incorporation into DNA after 16 h of treatment with 15 μM C1P was measured as described in the Methods section. Data are means ± SEM of three independent experiments performed in triplicate. The effect of C1P was statistically significant by Student’s t test (**P < 0.01). Inset: confluent C2C12 myoblasts were pre-treated 16 h with PTx (0.5 μg/ml) before being challenged (+) or not (−) with 10 μg/ml LPA for 48 h. The content of myogenin and caveolin-3 (cav-3) was analyzed in cell lysates (30 μg) by Western Blot analysis. Equally loaded protein was checked by expression of the nonmuscle-specific β isoform of actin. A blot representative of three independent experiments with analogous results is shown. B) Effect of PTx on C1P-induced activation of Akt and ERK1/2. Serum-starved myoblasts were pre-incubated for 16 h with PTx (0.5 μg/ml) before being challenged with 15 μM C1P for 10 min. Cell lysates were analyzed by Western Blotting as described in Methods section. A representative blot of three independent experiments is shown. Histograms represent densitometric quantification of phosphorylated protein normalized to its total content and reported as mean ± SEM of three independent experiments, fold change over control set as 1. The effect of C1P was statistically significant by Student’s t test (*P < 0.05).

In an attempt to establish a possible cross-talk among the various signaling pathways that were activated by C1P and found to be implicated in its pro-mitogenic action, the effect of C1P on the individual signaling pathways in the presence of different inhibitors was examined. Results presented in Fig. 7 show that pre-incubation with 5 μM UO126 prevented C1P-induced ERK1/2 phosphorylation at 10 min of incubation. Moreover, this inhibitor did not reduce Akt phosphorylation indicating that ERK1/2 signaling is not upstream of the PI3K/Akt pathway. Interestingly, the blockade of ERK1/2 pathway abolished the enhancement of mTOR phosphorylation, demonstrating that in this setting ERK1/2 is upstream of mTOR activation, in keeping with the key role of this kinase in the inhibition of its main suppressor, TSC2 [30]. The same Figure shows that the blockade of PI3K with LY294002 blunted Akt phosphorylation and largely decreased ERK1/2 and mTOR activation, strongly suggesting that PI3K/Akt is partially upstream of ERK1/2. To establish whether mTOR activation by C1P is downstream of PI3K/Akt signaling pathway, TGX-221, a selective inhibitor of PI3Kβ, which does not affect mTOR activity was employed [31]. Interestingly, myoblast treatment with 1 μM TGX-221 resulted in a nearly complete inhibition of Akt phosphorylation, demonstrating that PI3Kβ is the major PI3K isoform implicated in the signaling cascade triggered by C1P; moreover, in this experimental condition, C1P was not capable of further increasing mTOR phosphorylation, whose basal levels were enhanced, therefore proving that PI3Kβ is upstream of mTOR activation by this sphingolipid.

Fig. 7.

C1P-induced activation of mTOR is downstream of PI3K/Akt and ERK1/2 signaling pathways in C2C12 myoblasts. C2C12 myoblasts approximately 40% confluent were serum-starved for 24 h and pre-incubated for 30 min with PI3K inhibitor (5 μM LY294002) or MEK inhibitor (5 μM U0126) or PI3Kβ inhibitor (1 μM TGX-221) before being challenged with 15 μM C1P for 10 min. Cell lysates were separated by SDS-PAGE and immunoblotted using specific anti-phospho-Akt, anti-pan Akt anti-phospho-ERK1/2, anti-pan ERK1/2, anti-phospho-mTOR and anti-β-actin antibodies. Blots representative of at least three independent experiments are shown. Histograms represent densitometric quantification of phosphorylated protein normalized to its total content or to β-actin and reported as mean ± SEM of three independent experiments, fold change over control set as 1. The effect of C1P was statistically significant by Student’s t test (*P < 0.05); the effect of the inhibitors was statistically significant by Student’s t test (#P < 0.05).

Finally, the role of endogenous C1P in C2C12 myoblasts was investigated. Cell treatment with the CERK inhibitor NVP-231 (100 nM) reduced [³H]thymidine incorporation by approximately 30%, highlighting the involvement of CERK in C2C12 cell proliferation (Fig. 8A). However, the protein content of CERK was not altered during myoblast proliferation or differentiation (Fig. 8B) excluding a role for the transcriptional regulation of this enzyme in these biological processes.

Fig. 8.

Role of endogenous C1P on C2C12 myoblast proliferation and differentiation. A) C2C12 myoblasts approximately 40% confluent were serum-starved for 24 h before being treated with CERK inhibitor (100 nM NVP-231) or the related inactive compound (100 nM compound 4). [³H]thymidine incorporation into DNA was measured as described in the Methods section. Data are means ± SEM of three independent experiments performed in triplicate. The effect of CERK inhibitor was statistically significant by Student’s t test (*P < 0.05). B) C2C12 myoblasts approximately 40% confluent were serum-starved for 24 h and incubated with or without 10% FBS for the indicated time-intervals (upper panel) or grown to confluence and incubated in medium supplemented with 0.1% BSA for the indicated period of time (lower panel). Cell lysates were separated by SDS-PAGE and immunoblotted using specific anti-CERK and anti-β-actin antibodies. Blots representative of three independent experiments are shown. The histograms represent band intensity of CERK normalized to β-actin and reported as mean ± SEM of three independent experiments, fold change over control set as 1.

4. Discussion

The proliferation of skeletal muscle progenitor cells such as satellite cells and myoblasts is an essential step in the repair of damaged tissue since increases in the number of these cells ensures an adequate repopulation of the tissue. Although there has been a considerable attention to this subject, the bioactive molecules implicated in the regulation of this key event have not been fully characterized [32]. In this work we have identified C1P as positive regulator of mouse skeletal myoblast proliferation. Several pieces of experimental evidence are here provided for the mitogenic effect of C1P in these cells. This phosphosphingolipid potently stimulated [³H]thymidine incorporation into DNA, cell cycle progression and cell number. Moreover, C1P enhanced GSK3-β and pRb phosphorylation as well as cyclin D1 protein content. However, other critical biological functions of myoblasts, such as myogenic differentiation or cell survival, appear to be unaffected by C1P. This phosphosphingolipid was previously found to promote cell survival in macrophages (reviewed in [28,33]); therefore, the present report supports the notion that C1P evokes a specific response in myoblasts. Notably, the mitogenic action of C1P in myoblasts was elicited by the bioactive compound per se and could not be attributable to any possible mediator formed by C1P metabolism in these cells.

So far only few cell types (mainly different types of macrophages) are known to respond to natural (long-chain) C1P [28,33]. The present demonstration that C1P enhances cell growth of mouse myoblasts extends the number of target cells that are responsive to C1P and supports the notion that C1P, besides acting as a mediator of the inflammatory response, can also participate in other key biological processes such as myogenesis. Intriguingly, other sphingolipids are known to regulate specific aspects of skeletal muscle cell physiology. In particular, S1P is known to stimulate myoblast differentiation toward myotubes [8] and to inhibit their chemotactic response [34]. By contrast sphingosine, its immediate precursor, was reported to inhibit chicken myoblast fusion [35], and de novo synthesized ceramide was found to be responsible for inhibition of myogenic differentiation of rat myoblasts [6]. Thus, C1P appears to be the only sphingolipid endowed with mitogenic properties in myoblasts at the present time, providing strong evidence that sphingolipids are critical for regulation of skeletal muscle development and regeneration, and reinforcing the notion that individual bioactive sphingolipids play specific and distinct roles in the control of this highly orchestrated biological process.

In this study, the molecular mechanisms by which C1P stimulates C2C12 cell proliferation were also investigated. The PI3K/Akt and ERK1/2 signaling pathways were found both necessary for the mitogenic response to C1P, since selective inhibition of these kinases abrogated the biological action of this sphingolipid. Interestingly, the mitogenic response elicited by C1P was equally prevented when mTOR was selectively blocked, supporting the view that this kinase, which was here identified downstream of PI3K/Akt and ERK1/2, plays a master regulatory role in myoblast growth. Thus, these results further consolidate the well-established role exerted by PI3K/Akt, mTOR and ERK1/2 in the regulation of myoblast proliferation [36–41] and highlight C1P as a critical metabolite implicated in the control of myoblast proliferation as key regulator of these kinases. In accordance with the findings presented in this report, the PI3K/Akt and ERK1/2 pathways were also identified as mediators of the mitogenic effect of C1P in bone marrow-derived macrophages [13], suggesting a common mechanism of action of C1P in macrophages and myoblasts. By contrast, C1P was found to exert a mitogenic action in fibroblasts via an ERK-independent mechanism [12], reinforcing the notion that this bioactive sphingolipid exploits distinct signaling pathways in different cellular settings for eliciting its biological response. Also, contrary to the pro-survival role of C1P in macrophages and although C1P stimulates Akt, and ERK1/2 in myoblasts, C1P did not inhibit apoptosis in these cells.

The rapid time-course of signaling cascades triggered by C1P in myoblasts might in principle, be consistent with a receptor-mediated event. In this connection, C1P was shown to stimulate macrophage migration through interaction with a G_i protein-coupled receptor [15]; however, the involvement of such a receptor in the mitogenic effect of C1P in myoblasts was ruled out by the inability of PTx to block C1P-dependent ERK1/2 or Akt phosphorylation, and cell proliferation. In agreement with these observations, PTx also failed to prevent the mitogenic and anti-apoptotic responses elicited by C1P in bone marrow-derived macrophages [14], where the mitogenic effect of C1P seems to be associated to the generation of intracellular C1P [42]. Nonetheless, a possible involvement of receptors other than G_i protein-coupled receptors in the mitogenic effect of C1P in myoblasts remains to be determined.

It is well established that endogenous C1P is formed intracellularly by ceramide kinase-catalyzed phosphorylation of ceramide [43,44], although other pathways may also exist [45]. The potential ability of mouse myoblasts to synthesize C1P by their own, which would act as an endogenous regulator of their own proliferation, was not explored here; however, the described mitogenic response of myoblasts to C1P added exogenously strongly supports the hypothesis that C1P released by other cells, such as macrophages, which are recruited to the sites of muscle damage [46–48], might participate to the beneficial effect of muscle growth and regeneration.

In summary, we show here that C1P promotes proliferation of mouse myoblasts, strongly suggesting the involvement of this sphingolipid in the regulation of muscle regeneration and repair.