Reduction in Intracellular Calcium Levels Inhibits Myoblast Differentiation*

In myocytes, calcium plays an important role in intracellular signaling and contraction. However, the ability of calcium to modulate the differentiation of striated muscle cells is poorly understood. To examine this issue we studied C2C12 cells, which is a myoblast cell line that differentiates in vitro. First, we observed that the L-type calcium channel blockers nifedipine and verapamil effectively inhibited electrically induced calcium transients. Next, C2C12 cells were exposed to these agents during conditions that induce myocyte differentiation. In the presence of nifedipine and verapamil, myoblasts failed to form myotubes. Dantrolene and thapsigargin, which decrease intracellular calcium by different mechanisms, also inhibited differentiation. In addition, nifedipine and verapamil inhibited the expression of myosin heavy chain and myogenin, two markers of skeletal myoblast differentiation. In contrast, levels of the transcriptional factor Myf5, which is expressed in undifferentiated myoblasts, did not decline. Calcium channel blockade also prevented the expression of a reporter driven by the skeletal muscle α-actin promoter. These data demonstrate that lowering intracellular calcium levels inhibits the differentiation of skeletal myoblasts into mature myotubes.

In myocytes, calcium plays an important role in intracellular signaling and contraction. However, the ability of calcium to modulate the differentiation of striated muscle cells is poorly understood. To examine this issue we studied C2C12 cells, which is a myoblast cell line that differentiates in vitro. First, we observed that the L-type calcium channel blockers nifedipine and verapamil effectively inhibited electrically induced calcium transients. Next, C2C12 cells were exposed to these agents during conditions that induce myocyte differentiation. In the presence of nifedipine and verapamil, myoblasts failed to form myotubes. Dantrolene and thapsigargin, which decrease intracellular calcium by different mechanisms, also inhibited differentiation. In addition, nifedipine and verapamil inhibited the expression of myosin heavy chain and myogenin, two markers of skeletal myoblast differentiation. In contrast, levels of the transcriptional factor Myf5, which is expressed in undifferentiated myoblasts, did not decline. Calcium channel blockade also prevented the expression of a reporter driven by the skeletal muscle ␣-actin promoter. These data demonstrate that lowering intracellular calcium levels inhibits the differentiation of skeletal myoblasts into mature myotubes.
Mammalian myogenic differentiation is a complex process that involves maturation of myoblasts into mature myocytes (1,2). These events lead to the activation of muscle-specific genes and the repression of genes required for cell proliferation, both of which are necessary for differentiation (1)(2)(3). This genetic program requires the action of a number of transcriptional factors and nuclear proteins (1)(2)(3)(4). Our understanding of the signaling molecules that link extracellular signals to changes in transcriptional factor activity and subsequent gene expression is beginning to evolve.
Calcium is an intracellular messenger that is important in muscle cells for excitation-contraction coupling (reviewed in Refs. 3,5). First, action potentials open voltage gated, L-type calcium channels to allow small quantities of calcium to enter the cell (so called "sparks"). This triggers calcium release from the sarcoplasmic reticulum (so called "calcium transients").
This larger wave of intracellular calcium leads to direct activation of the contractile apparatus, myosin cross-bridge cycling, and muscle contraction. Second, through the action of calcium/calmodulin-dependent protein kinase II (CaMK II) 1 and other kinases and phosphatases, calcium can indirectly alter the activity of proteins involved in contraction, calcium release, and calcium reuptake (3).
In addition to regulating cellular contraction, calcium acts as an intracellular messenger to modulate many aspects of cell physiology. Calcium-signaling pathways modulate the expression of muscle-specific genes during differentiation by altering the phosphorylation and activity of a number of transcriptional factors (reviewed in Ref. 3,4). More specifically, calcium regulates the function of phosphatases and kinases (e.g. calcineurin, CaMK and protein kinase C) that can alter the localization and function of transcriptional activating factors such as cAMP response element-binding protein, MEF2, Myf5, MyoD, myogenin, nuclear factor of activated T-cells, and serum response factor (3, 4, 6 -11).
Based on these observations, it is possible that intracellular calcium may play a role in the differentiation of striated myocytes by altering gene expression. To examine this issue, we studied the effects of decreased intracellular calcium levels in C2C12 cells, which are a favored model of skeletal muscle differentiation (12)(13)(14). Using three independent indices of myoblast differentiation, we now demonstrate that intracellular calcium levels play a crucial role in muscle cell maturation.

EXPERIMENTAL PROCEDURES
Materials-Medium and bovine serum were obtained from American Type Tissue Collection (Manassas, VA). All other culture supplements were from Invitrogen. Fluo-3 acetoxymethyl ester (Fluo-3 AM) and pluronic F-127 were from Molecular Probes (Eugene, OR). All other chemicals were obtained from Sigma. Anti-myosin heavy chain antibody (MF-20) was developed by Donald Fischman and obtained from the Developmental Studies Hybridoma Bank, University of Iowa (15).
Cell Culture-C2C12 cells (ATCC no. CRL-1772) were grown on plastic or glass culture substrates as described (14). Briefly, cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 1 M pyruvate, 100 units/ml penicillin G sodium, and 100 g/ml streptomycin sulfate until 90 -100% confluence. Cells were induced to differentiate using a low serum differentiation medium (Dulbecco's modified Eagle's medium, pyruvate, penicillin, streptomycin, and 2% horse serum). Upon induction of differentiation, medium containing vehicle or drugs was changed daily. Cells were cultured at 37°C in a humidified incubator containing 5% CO 2 . Unless otherwise indicated, cells were harvested after 4 days in differentiation medium.
To induce acute calcium release (calcium transients), samples were stimulated using a Grass SD5 stimulator (Quincy, MA). Platinum electrodes were placed 1.7 cm apart, and current flowed roughly parallel to the long axis of the myotube. Initial voltage response curves demonstrated that calcium transients occurred with a stimulation threshold of 15-30 V and a plateau of 50 V. Therefore, all further stimulations were performed at 30 -40 V for 100 ms. Samples were stimulated, and images were recorded three times for each condition with 30 s between each stimulation. Specimens were first stimulated before the addition of any drug (0 M). Specimens were next treated with increasing concentrations of drug by adding successive 1-l aliquots of 1000ϫ concentrated drug. After a 2-min incubation at each drug concentration, electrically stimulated calcium transients were recorded. Individual experiments lasted less than 1 h after the addition of Fluo-3 AM.
Images of calcium signal were obtained using an Olympus Fluoview confocal laser scanning microscope (IX70, Melville, NY), an UPlanFl 20ϫ objective (numerical aperture 0.5) and Fluoview image processing software (v2.1). The same area of the cell was analyzed during each stimulation. Images were scanned at a frame rate of 0.3 s per frame and 30 frames were obtained to record a total of 9 s per stimulation.
Image intensity within an outlined area was scanned in Fluoview and saved to a spreadsheet (Microsoft Excel 2000, Microsoft, Inc., Redmond, WA). The three curves for each condition were averaged and normalized to the three measurements prior to stimulation. Baseline and peak values for each treatment were combined and analyzed for statistical significance using GraphPad Prism (v3.0, GraphPad Software, Inc, San Diego, CA). Concentration-response curves were performed at least five times for each condition.
Calcium Imaging Movies-Images of Fluo-3 fluorescence versus time were saved as single .tiff files. Movies were created by copying sequential images as layers into ImageReady (v3.0, Adobe, San Jose, CA). For easier viewing, movies were saved as QuickTime files at 0.1 s per frame.
Immunolabeling-After 4 days in differentiation medium, nuclear and myosin heavy chain labeling were performed as described (17,18). Briefly, cells were incubated for 30 min with 6 g/ml Hoechst nuclear dye and fixed with 100% methanol for 15 min at Ϫ20°C. After rehydration with Dulbecco's phosphate-buffered saline (Invitrogen) plus 10 mM sodium azide and blocking with this buffer plus 2% bovine serum albumin, specimens were labeled with 1:20 anti-myosin heavy chain antibody (15). Finally, samples were washed and labeled with Cy-3 rabbit anti-mouse secondary antibody (Jackson ImmunoResearch, West Grove, PA).
Phase contrast and indirect immunofluorescence images were captured using an Olympus IX70 epifluorescence microscope equipped with 4ϫ (UPlanFl, numerical aperture 0.13) and 10ϫ (CPlan, numerical aperture 0.25) objectives. Images were captured with a DEI 750 T Color charge-coupled device camera (Optronics Engineering, Goleta, CA) and analyzed using ImagePro software (v4.1, Media Cybernetics, Silver Spring, MD). Phase images were obtained prior to Hoechst staining. Images were processed for presentation using Adobe PhotoShop (v4.0) and Adobe Illustrator (v9.0).
For measurement of the differentiation index, Hoechst-stained nuclei were counted in each image. Images of Hoechst and myosin heavy chain staining were then merged, and nuclei associated with myosinpositive cells were counted. The differentiation index for each image pair is defined as the percent of all nuclei present in MF-20 stained cells. For each separate experiment, three image pairs were taken per condition.
Northern Blot Analysis-RNA was isolated from C2C12 cells using Trizol (Invitrogen) and quantified by spectrophotometry. 1-5 g of total RNA was blotted to an Ambion BrightStar-Plus nylon membrane and cross-linked (UV Stratalinker 2400, Stratagene). Northern blotting was performed using Ambion NorthernMax reagents at 42°C per the manufacturer's instructions. Images of labeled blots were obtained using a phosphorimaging system (GS-525 Molecular Imaging System, Bio-Rad, Hercules, CA).
Luciferase Expression Assays-Reporter assays using a luciferase expression vector driven by the skeletal ␣-actin promoter (SK-Luc, provided by Dr. Robert Schwartz) were performed as described (14). Briefly, C2C12 cells were grown in 12-well plates to 50 -75% confluence and were transfected with Sk-Luc (1 g/well) using LipofectAMINE 2000 (Invitrogen). To control for transfection efficiency, cells were cotransfected with pSV-␤-galactosidase (35 ng/well, Promega, Madison, WI). When confluent, samples were then differentiated in the presence of vehicle or drug for 4 days as above. Specimens were processed for luciferase and ␤-galactosidase expression using the luciferase and ␤-galactosidase assay systems with reporter lysis buffer (Promega; per manufacturer's instructions). Luciferase measurements were made using a luminometer (Turner Designs, Model TD-20/20, Sunnyvale, CA), and ␤-galactosidase was measured using a Bio-Rad Model 550 microplate reader.
Statistics-Data are presented as means and standard errors of the mean (S.E.). Except as noted, at least three separate experiments were  performed for each condition. All quantifiable data were saved to Microsoft Excel or GraphPad Prism. Statistically significant differences between groups were evaluated using one-way or repeated measures ANOVA with Dunnett's post hoc testing using GraphPad Prism (p Ͻ 0.05).

Calcium Channel Blockade Decreases C2C12 Cellular
Calcium Levels-To begin to examine the role of calcium on myocyte differentiation we first examined the effects of L-type calcium channel blockade on intracellular calcium levels in C2C12 cells. Changes in intracellular calcium concentrations were assessed by confocal microscopy using the calcium-sensitive dye, Fluo-3 AM. Cultures were studied after 4 days in differentiation medium.
To define patterns of Fluo-3 response under normal conditions, cells were stimulated at different voltages (5-50 V) to generate voltage-response curves. In all cases the threshold for a response in Fluo-3 signal was 15-30 V. At 50 V, we saw maximal responses in all cells. Thus, in all subsequent studies cells were stimulated at 30 -40 V.
Next, to determine the effects of drugs on calcium release, samples were examined in the presence or absence of increasing concentrations of nifedipine and verapamil, two antagonists of L-type calcium channels. Both agents decreased the peakcalciumresponsetoelectricalstimulationinaconcentrationdependent manner, although nifedipine was more effective at the concentrations used ( Figs. 1 and 2). Control experiments using vehicle demonstrated that the effects were not due to nonspecific effects of the experimental protocol over time (not shown).
Decreasing Intracellular Calcium Levels Inhibits Morphologic Differentiation of C2C12 Myoblasts-Having demonstrated that nifedipine and verapamil decrease intracellular calcium levels, we next examined the effects of reduced intracellular calcium concentrations on myocyte differentiation. In the presence of differentiation medium, C2C12 cells mature into multinucleated myotubes with a functional contractile apparatus (12,20). Therefore, confluent cultures of C2C12 cells were incubated in differentiation medium in the presence of vehicle, nifedipine, or verapamil for 4 days. The morphology of each specimen was first assessed using phase microscopy. Cul-  tures were then stained for nuclei to assess for the presence of multinucleated myotubes and for myosin heavy chain to examine the maturation of the contractile apparatus.
Vehicle-treated cultures contained few undifferentiated myoblasts and numerous multinucleated, myosin heavy chainpositive myotubes (Fig. 3, A and F). In contrast, treatment with either nifedipine or verapamil inhibited the ability of these cultures to differentiate (Fig. 3, B-E, G-J). Calcium channel blockade decreased both the number and the size of differentiated myotubes. These effects were concentration-dependent, and the effective concentrations correlated with those required to attenuate electrically stimulated calcium transients (Figs. 1  and 2).
When the effects of nifedipine and verapamil on differentiation were quantified using a differentiation index, these visual impressions were confirmed. Control experiments demonstrated the emergence of multinucleated myotubes during incubation in differentiation medium with 67% of the nuclei present in differentiated cells after 4 days (Fig. 4A). In contrast, both nifedipine and verapamil significantly decreased the ability of C2C12 cells to differentiate (Fig. 4B). This effect was concentration-dependent, with 100 M nifedipine and 100 M verapamil reducing the differentiation index to 25.1% and 7.6% of control values, respectively.
To further demonstrate that these effects on differentiation are due to decreased calcium levels, we next treated C2C12 cells with agents known to decrease intracellular calcium by mechanisms other than calcium channel blockade. Differentiating C2C12 cells were treated with dantrolene, which inhibits calcium release from the sarcoplasmic reticulum, or thapsigargin, which inhibits calcium reuptake by the sarcoplasmic reticulum and eventually depletes total cellular calcium (21). The doses used are known to effectively decrease intracellular calcium levels and to alter gene expression in skeletal myotubes (21). Treatment with either agent led to a qualitative and quantitative inhibition of myocyte differentiation in a concentration-dependent manner (Figs. 3  and 4).
Calcium Channel Blockade Alters the Expression of Musclespecific Transcriptional-Activating Factors-After observing that nifedipine and verapamil inhibited cellular differentiation, we next examined the effects of these agents on the expression of two muscle-specific transcriptional activating factors, myogenin and Myf5. During myoblast differentiation the expression of myogenin is increased, while that of Myf5 is decreased (1,2,22). To address this issue, Northern blotting experiments were performed using RNA from C2C12 cells treated with nifedipine or verapamil during differentiation. Myogenin and Myf5 levels were quantified using a phosphorimaging device.
As expected, we observed that myogenin gene expression increased during differentiation (not shown) (1, 2, 22). In contrast, nifedipine and verapamil inhibited the expression of myogenin in a concentration-dependent manner (Fig. 5, A  and B).
Myf5 expression decreased dramatically during differentiation (not shown) as expected (1,2,22). However, cultures treated with nifedipine or verapamil maintained high levels of Myf5 expression, although the levels of Myf5 expression in cell treated with verapamil were not statistically significant from control (Fig. 5, A and C). As with myogenin, the effects of these agents on Myf5 expression were concentration-dependent (Fig.  5C).
Skeletal ␣-Actin Expression Is Inhibited by Calcium Channel Blockade-The previous experiments showed that calcium channel blockade inhibits morphologic changes and alters transcriptional factor expression during myoblast differentiation. Next, to assess if decreased intracellular calcium levels alter the expression of structural genes that are normally up-regulated during myocyte differentiation, the effects of nifedipine and verapamil on ␣-actin gene expression were examined.
Cultures of C2C12 cells were transfected with SK-Luc, a luciferase expression vector driven by the skeletal ␣-actin promoter that has been validated in previous studies of C2C12 cell and primary myocyte differentiation (14,22,23). Non-confluent C2C12 cultures were transfected with SK-Luc and were allowed to differentiate in medium containing vehicle, nifedipine, or verapamil. After 4 days in differentiation medium, specimens were assayed for luciferase expression.

DISCUSSION
It is becoming clear that many extracellular signals regulate the expression of genes essential for striated myocyte differentiation (7,10,11,19,24). However, the intracellular pathways that transduce these signals to the nucleus are only beginning to be understood. Previous reports have suggested that calcium-activated signaling cascades may play an important role in intracellular signaling during myocyte differentiation (7,11,13,25,26). Using C2C12 cells as a model of myocyte differentiation, we demonstrate that the maintenance of normal intracellular calcium concentrations is vital to myocyte gene expression and differentiation.
Using a calcium-sensitive dye and confocal microscopy, we first demonstrated that electrically stimulated calcium transients were inhibited by calcium channel blockade. These results are not unexpected because differentiating C2C12 cells are known to express L-type calcium channels (27). These data also agree with previous reports that calcium channel blockade decreases the frequency and amplitude of calcium transients in C2C12 cultures and in primary cultures of skeletal myocytes (25, 28 -30). It is interesting that verapamil was not as effective as nifedipine in these experiments. The binding of verapamil to L-type calcium channels is enhanced if the channel is actively opening and closing (31). Therefore, these findings were not unexpected due to the acute treatments used in these experiments. In addition, other experiments demonstrated that verapamil lowered calcium transient amplitude more effectively if specimens were stimulated at 0.5 Hz during each 2-min treatment period. 2 Having established the concentrations of nifedipine and verapamil required to effectively decrease intracellular calcium concentration, we found that calcium channel blockade profoundly inhibited myoblast differentiation. Visual observations of C2C12 cells induced to differentiate in the presence of nifedipine or verapamil demonstrated that decreasing intracellular calcium levels inhibited the formation of multinucleated myotubes. Three additional quantitative indices supported this finding. First, immunolabeling and calculation of a differentiation index quantified gross morphologic observations. Second, decreased expression of myogenin and increased expression of Myf5 in cells treated with both drugs demonstrated that the genetic program required for normal differentiation was disrupted. Third, the expression of actin and myosin, two structural genes present in the mature contractile apparatus, was inhibited by calcium channel blockade.
We then treated C2C12 cells with dantrolene and thapsigar-gin to demonstrate that these effects were due to decreased intracellular calcium levels and not due to some other effect of calcium channel blockade. Dantrolene inhibits calcium release from the sarcoplasmic reticulum and thus decreases sarcoplasmic calcium (21). In contrast, thapsigargin blocks reuptake of calcium into the sarcoplasmic reticulum and thus eventually depletes total cellular calcium (21). As with nifedipine and verapamil, treatment of C2C12 cells with these agents also prevented morphologic differentiation. Therefore, these data support the hypothesis that decreased intracellular calcium levels inhibit differentiation of muscle cells. Potential mechanisms of this calcium-mediated modulation of differentiation and gene expression include alterations in protein degradation, in mRNA stabilization, or in gene transcription. The latter mechanism appears to be most important during differentiation, but evidence exists to support a role for the former two mechanisms. For example, calpain-mediated protein degradation appears to allow cell fusion and differentiation of C2C12 cells (13). In addition, increased levels of intracellular calcium appear to stabilize mRNA encoding the inward-rectifying potassium channel 1 (26).
However, regulation of gene transcription appears to play a major role in calcium-mediated modulation of differentiation. Recent reports suggest that this occurs via the activation of a number of phosphatases and kinases, including calcineurin, CaMK isoforms, and protein kinase C. These enzymes alter the activity of transcriptional activating factors, such as cAMPresponse element-binding protein, MEF2, Myf5, myogenin, nuclear factor of activated T cells, and serum response factor (3, 4, 6 -11). Our results provide a basis for further investigation of these potential mechanisms.
In conclusion, our data suggest that mammalian skeletal myoblast differentiation is dependent upon normal intracellular calcium concentrations. Future experiments using other methods to alter intracellular calcium levels should further define the relationship between intracellular calcium concentrations and myoblast differentiation. In addition, these experiments may also help determine which calcium-dependent pathways contribute to the integration of signals that drive skeletal myofiber development.