Skeletal muscle atrophy is a prominent and disabling feature of chronic wasting diseases. Prevention or reversal of muscle atrophy by administration of skeletal muscle growth (hypertrophy)-stimulating agents such as insulin-like growth factor I (IGF-I) could be an important therapeutic strategy in these diseases. To elucidate the IGF-I signal transduction responsible for muscle formation (myogenesis) during muscle growth and regeneration, we applied IGF-I to differentiating C2C12 myoblasts and evaluated the effects on phosphatidylinositol 3-kinase (PI3K)/Akt/glycogen synthase kinase-3β (GSK-3β) signaling and myogenesis. IGF-I caused phosphorylation and inactivation of GSK-3β activity via signaling through the PI3K/Akt pathway. We assessed whether pharmacological inhibition of GSK-3β with lithium chloride (LiCl) was sufficient to stimulate myogenesis. Addition of IGF-I or LiCl stimulated myogenesis, evidenced by increased myotube formation, muscle creatine kinase (MCK) activity, and troponin I (TnI) promoter transactivation during differentiation. Moreover, mRNAs encoding MyoD, Myf-5, myogenin, TnI-slow, TnI-fast, MCK, and myoglobin were upregulated in myoblasts differentiated in the presence of IGF-I or LiCl. Importantly, blockade of GSK-3β inhibition abrogated IGF-I- but not LiCl-dependent stimulation of myogenic mRNA accumulation, suggesting that the promyogenic effects of IGF-I require GSK-3β inactivation and revealing an important negative regulatory role for GSK-3β in myogenesis. Therefore, this study identifies GSK-3β as a potential target for pharmacological stimulation of muscle growth.
- insulin-like growth factor I
- muscle hypertrophy
skeletal muscle atrophy is a prominent and disabling feature of, and an independent predictor of mortality in (37, 44), many chronic wasting diseases, including chronic obstructive pulmonary disease (19, 51), chronic heart failure (5), and cancer (7). Prevention or reversal of muscle atrophy by stimulating skeletal muscle growth (hypertrophy) can be an important therapeutic strategy in these diseases.
Skeletal muscle hypertrophy is characterized by an increase in the cross-sectional area of the muscle fibers, which is associated with increased protein accumulation (47). Simultaneously, the number of myonuclei may increase during hypertrophy (myonuclear accretion) to maintain the myonuclear domain (i.e., the ratio between myofiber size and myonuclear number) (3). Myonuclear accretion requires activation and subsequent proliferation of satellite cells (45), which are quiescent mononuclear precursor cells that reside in the basal lamina of terminally differentiated muscle fibers. Upon activation, satellite cells proliferate (now called myoblasts) either to become quiescent again and repopulate the satellite cell pool or to differentiate and fuse with existing muscle fibers or form new myofibers (3).
The growth-promoting properties of insulin-like growth factor I (IGF-I) have been well documented in various tissues and include an important role for this polypeptide in muscle hypertrophy responses (26). Studies in mice deficient for IGF-I or IGF-I signaling have revealed that IGF-I is required for muscle growth and development (43, 49). Conversely, various groups have shown that overexpression or local infusion of IGF-I in skeletal muscle results in increased muscle mass (2, 46, 53), illustrating that IGF-I positively modulates skeletal muscle hypertrophy.
On the basis of these promising reports, several clinical trials have investigated the efficacy of pharmaceutical stimulation of circulating IGF-I levels on skeletal muscle hypertrophy, but only modest effects on lean mass and functional status have been reported (13). In addition, high circulating levels of IGF-I were associated with increased risk for certain types of cancer (14, 29) and insulin resistance (12), illustrating that elucidation of the effector molecules of IGF-I-dependent muscle hypertrophy is essential for successful therapeutic application.
After the binding of IGF-I to its receptor, two major signaling pathways are activated in skeletal muscle: the mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3K)/Akt pathway. The proliferative response of myoblasts to IGF-I is mediated by the MAPK pathway, whereas the PI3K/Akt pathway mediates muscle differentiation (16), although recent reports suggest interactions between these pathways during differentiation (28).
Although myoblast proliferation is important in muscle growth, most evidence indicates that IGF-I-mediated muscle hypertrophy relies on increased (muscle) protein accretion (10) and stimulation of myogenic differentiation (23), occurring through PI3K/Akt signaling (34). Activation of this pathway by IGF-I requires recruitment of insulin receptor substrate-1 to the receptor, resulting in increased enzymatic activity of PI3K and subsequent phosphorylation and activation of Akt. This serine/threonine kinase regulates increased protein synthesis during muscle hypertrophy through activation of mammalian target of rapamycin (mTOR) and p70S6K, which control mRNA translation (10). In addition, glycogen synthase kinase-3β (GSK-3β), a dual-specificity kinase involved in many signaling and metabolic pathways, including transcriptional modulation, glycogen accumulation, and negative regulation of mRNA translation (21), is inhibited after phosphorylation by Akt (18).
Recently, GSK-3β has been implicated in the negative regulation of both cardiac and skeletal muscle hypertrophy (6, 30, 50, 56). These skeletal muscle studies, however, were performed in fully differentiated myotubes, which does not permit investigators to address the role of GSK-3β on myonuclear accretion and myogenic differentiation during (IGF-I dependent) hypertrophy. Because we hypothesized that IGF-I-mediated hypertrophy involves inhibition of GSK-3β, we addressed GSK-3β regulation during IGF-I-stimulated myogenesis and compared the effects of IGF-I and GSK-3β inhibition on myoblast differentiation and myotube formation.
MATERIALS AND METHODS
The murine skeletal muscle cell line C2C12 was obtained from the American Type Culture Collection (ATCC CRL1772; Manassas, VA). These cells are able to undergo differentiation into spontaneously contracting myotubes after growth factor withdrawal (58). Myoblasts were cultured in growth medium (GM) composed of low-glucose Dulbecco's modified Eagle's medium (DMEM) containing antibiotics (50 U/ml penicillin and 50 μg/ml streptomycin) and 9% (vol/vol) fetal bovine serum (FBS) (all from GIBCO, Rockville, MD). Cells were plated at 104 cells/cm2 and cultured in GM for 24 h. To induce differentiation, we replaced GM with differentiation medium (DM), which contained DMEM with 1.0% heat-inactivated FBS and antibiotics. C2C12 cells were grown on Matrigel (BD Biosciences, Bedford, MA)-coated (1:50 in DMEM) dishes as described previously (39). Murine IGF-I (Calbiochem, La Jolla, CA) or lithium chloride (LiCl; Sigma, St. Louis, MO) was added directly after induction of differentiation and again 24 h later, when the cells were provided with fresh DM. Where indicated, LY-294002 (Calbiochem) dissolved in DMSO was added 30 min before the IGF-I stimulation.
Stable cell line.
For the assessment of transcriptional activation of the troponin I (TnI) gene during differentiation, a stable cell line was created that carries a TnI promoter-luciferase reporter gene in its genome (38). To determine luciferase activity, we lysed cells in 1× luciferase lysis buffer (Promega, Madison, WI) and stored them at −80°C. Luciferase activity was measured according to the manufacturer's instructions and expressed relative to total protein in the soluble fraction.
May-Grunwald Giemsa staining.
As a morphological parameter of differentiation, the myogenic index was defined as the number of nuclei residing in cells containing three or more nuclei, divided by the total number of nuclei in May-Grunwald Giemsa-stained cells. Cells were grown on Matrigel-coated 60-mm dishes, and after IGF-I or LiCl treatment, cells were washed twice in PBS (room temperature), fixed in methanol, and stained in May-Grunwald Giemsa (Sigma) according to the manufacturer's instructions.
Western blot analysis.
Phosphorylated Akt (p-Akt), Akt, phosphorylated GSK-3β (p-GSK-3β), and GSK-3β protein abundance was evaluated using Western blot analysis. Adherent cells were washed in PBS, and whole cell lysates were prepared by the addition of lysis buffer [20 mM Tris, 150 mM NaCl, 1% (vol/vol) Nonidet P-40, 1 mM DTT, 1 mM Na3VO4, 1 mM PMSF, 10 μg/ml leupeptin, and 1% (vol/vol) aprotinin]. Lysates were incubated on ice for 30 min, followed by 30 min of centrifugation at 16,000 g. A fraction of the supernatant was saved for protein determination, and 4× Laemmli sample buffer [0.25 M Tris·HCl, pH 6.8, 8% (wt/vol) SDS, 40% (vol/vol) glycerol, 0.4 M DTT, and 0.04% (wt/vol) bromphenol blue] was added, followed by boiling of the samples for 5 min and storage at −20°C. Total protein was assessed using the Bio-Rad DC protein assay kit (Hercules, CA) according to the manufacturer's instructions, and 10 μg of protein were loaded per lane and separated on a 10% polyacrylamide gel (Mini Protean 3 system; Bio-Rad), followed by transfer to a 0.45-μm nitrocellulose membrane (Bio-Rad) by electroblotting. The membrane was blocked for 1 h at room temperature in 5% (wt/vol) nonfat dry milk. Nitrocellulose blots were washed in PBS-Tween 20 (0.05%), followed by overnight incubation (4°C) with a polyclonal antibody specific for p-Akt, Akt, p-GSK-3β, GSK-3β (Cell Signaling, Beverly, MA), or a monoclonal antibody specific for GSK-3β (BD Biosciences, Lexington, KY). After three wash steps of 20 min each, the blots were probed with a peroxidase-conjugated secondary antibody (Vector Laboratories, Burlingame, CA) and visualized using SuperSignal West Pico chemiluminescent substrate (Pierce Biotechnology, Rockford, IL) according to the manufacturer's instructions.
GSK-3β immunoprecipitation and kinase activity assay.
C2C12 homogenates were prepared as described in Western blotting, with the exception that the lysis buffer consisted of 50 mM Tris, 150 mM NaCl, 10% (vol/vol) glycerol, 0.5% (vol/vol) Nonidet P-40, 1 mM EDTA, 1 mM DTT, 500 μM Na3VO4, 500 μM NaF, 100 μM β-glycerophosphate, 100 μM Natrium-pyro-PO4, 10 μg/ml leupeptin, and 1% (vol/vol) aprotinin. An input of 100 μg of total protein was precleared with protein G-Sepharose (Invitrogen, Carlsbad, CA) and subsequently incubated with 1 μg of monoclonal anti-GSK-3β under gentle agitation for 1 h at 4°C. The immune complexes were isolated by the addition of 25 μl of 50% slurry of protein G-Sepharose and incubation under gentle agitation overnight at 4°C. Immunoprecipitates were washed twice in lysis buffer and divided in two. To assess equal GSK-3β content, we used one half of the immunoprecipitate for Western blot analysis, as described above, and the other half for GSK-3β activity measurement. For this purpose, the immunoprecipitate was washed twice in kinase reaction buffer [8 mM MOPS, pH 7.4, 0.2 mM EDTA, 10 mM magnesium acetate, 1 mM Na3VO4, 1 mM DTT, 2.5 mM β-glycerophosphate, 10 μg/ml leupeptin, and 1% (vol/vol) aprotinin]. The kinase activity assays were performed in 40 μl of total reaction buffer containing 62.5 μM GSK-3β substrate (BioMol, Plymouth, PA), 20 mM MgCl2, 125 μM ATP, and 10 μCi [γ-32P]ATP. The reaction mixture was allowed to proceed for 30 min at 30°C, and 25 μl of the supernatant were spotted onto Whatmann P81 phosphocellulose paper. The filter squares were washed five times for 5 min each in 0.75% phosphoric acid. Next, the filters were briefly rinsed in acetone, dried at room temperature, and subjected to liquid scintillation counting.
Muscle creatine kinase activity.
Myogenic differentiation was assessed biochemically via determination of muscle creatine kinase (MCK) activity. Cells were grown on Matrigel-coated 60-mm dishes. After IGF-I or LiCl treatment, cells were washed twice in cold PBS, lysed in 0.5% Triton X-100, and scraped off the dish with a rubber policeman. Lysates were centrifuged for 2 min at 16,000 g, and the supernatant was stored in separate aliquots at −80°C for determination of protein content or MCK activity. MCK activity was measured using a spectrophotometry-based (55) kit from Sigma Diagnostics (St. Louis, MO). Specific MCK activity was calculated after correction for total protein, which was assessed using the Bradford method (11).
Cells were grown on 60-mm dishes. After treatment, C2C12 cells were washed twice with PBS and lysed using 1 ml of solution D (6.4 M guanidine thiocyanate, 40 mM sodium citrate, 0.8% sarcosyl, 100 mM β-mercaptoethanol) per 60-mm dish. RNA was extracted by adding 0.1 vol of 2 M sodium acetate, pH 5.0, 1 vol of water-saturated phenol, and 0.2 vol of chloroform-isoamyl alcohol (50:1). After vigorous vortexing and centrifugation, RNA was precipitated with an equal volume of isopropanol. The pellet was resuspended in 0.5 ml of solution D and precipitated again with isopropanol. After a wash step in 70% ethanol, the RNA was dissolved in nuclease-free water and stored at −80°C in small aliquots.
RNase protection assay.
Panels containing multiple muscle-specific RNase protection assay (RPA) probes described previously (40) were prepared by mixing EcoRI-linearized constructs in equal concentrations [the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probe was prepared separately] and then subjected to a transcription reaction using T7 RNA polymerase in the presence of excess unlabeled nucleotides and 50 μCi [α-32P]UTP (800 Ci/mmol), according to the manufacturer's MaxiScript kit instructions (Ambion, Houston, TX). The protection reaction was performed after 18 h of hybridization at 42°C with a mix containing 10 μg of RNA per sample, 8 × 104 cpm per probe with the exception of GAPDH (3.2 × 105 cpm), after 3 min of denaturation at 95°C, according to the manufacturer's RPA III kit instructions (Ambion). The amount of probe used was determined to be saturating in optimization experiments performed for each individual mRNA species. Subsequently, unprotected fragments were removed by RNase digestion (1:100 diluted RNase A/RNase T1 cocktail), followed by inactivation and precipitation of the protected fragments. Precipitated protected fragments were dissolved in 5 μl of loading buffer, denatured at 95°C, and separated on a 5% denaturing polyacrylamide gel. Gels were dried and exposed to film (Biomax MR-1; Kodak) or an imaging screen for quantification in a Personal Molecular Imager FX (Bio-Rad), and band intensity was analyzed using Quantity One software (Bio-Rad).
Semiquantitative RT-PCR was used to assess MCK mRNA levels. Total RNA was isolated as indicated and treated with DNase (New England Biolabs, Ipswich, MA), and 1 μg was reverse transcribed to cDNA using the Reverse iT first-strand synthesis kit (ABgene, Rochester, NY) according to the manufacturer's instructions. PCR reactions were performed using PCR master mix (ABgene). MCK transcripts were amplified using the following primers: forward primer, 5′-CACCTCCACAGCACAGACAG-3′; reverse primer, 5′-ACCTTGGCCATGTGATTGTT-3′, yielding a 160-bp product. MCK mRNA abundance was normalized to β-actin mRNA within each individual reaction by using QuantumRNA technology (Ambion) according to the manufacturer's instructions.
All values are means ± SE. Raw data were entered into SPSS (version 11) for statistical analysis. All differences between groups were determined using one-way ANOVA. A 0.05 level was used to represent statistical significance for all effects.
IGF-I promotes morphological differentiation.
C2C12 myoblasts were cultured in DM, and extensive myotube formation was observed after 72 h. IGF-I addition to the culture medium resulted in an increase in the size and number of myotubes (Fig. 1A, bottom left) This increase in size and number of myotubes was reflected in a doubling of the myogenic index (Fig. 1B), which is defined by the fraction of nuclei that reside in myotubes (cells containing 3 or more nuclei), in response to IGF-I.
IGF-I induces rapid phosphorylation of Akt and GSK-3β.
Next, we evaluated whether IGF-I treatment affected the phosphorylation status of Akt. The addition of IGF-I to differentiating myoblasts (Fig. 2A) or myotubes (data not shown) resulted in rapid phosphorylation of Akt at serine residue 473. Fifteen minutes after IGF-I administration, phosphorylation of Akt was maximal, whereas Akt phosphorylation returned to baseline 24 h after IGF-I treatment (data not shown). Addition of the PI3K inhibitor LY-294002 inhibited Akt phosphorylation by IGF-I (Fig. 2A). Activation of Akt results in the phosphorylation and inactivation of GSK-3β (18). IGF-I treatment caused rapid phosphorylation of GSK-3β (Fig. 2B), with a maximal response after 15 min corresponding to the kinetics of Akt phosphorylation (data not shown). In line with the effects on Akt phosphorylation, inhibition of PI3K activity abrogated GSK-3β phosphorylation by IGF-I (Fig. 2B). These results show that both Akt and GSK-3β are rapidly phosphorylated in response to IGF-I in differentiating myoblasts and that these effects require PI3K activity.
IGF-I and LiCl suppress GSK-3β kinase activity.
In contrast to IGF-I, LiCl, a well-characterized inhibitor of GSK-3β (21), did not affect Akt or GSK-3β phosphorylation status (data not shown). As expected, increased GSK-3β phosphorylation by IGF-I was accompanied by a decrease in GSK-3β kinase activity in C2C12 cells (Fig. 2C). Importantly, compared with IGF-I, LiCl administration inhibited GSK-3β activity even more strongly (Fig. 2C). To ensure equal GSK-3β loading, we subjected immunoprecipitation lysates used for GSK-3β activity assay to Western blot analysis. Although IGF-I and LiCl (36) inhibit GSK-3β through different mechanisms, both resulted in a significant reduction of GSK-3β activity (Fig. 2C).
Inhibition of GSK-3β by LiCl promotes morphological differentiation.
To investigate whether GSK-3β inhibition promotes myotube formation, we added LiCl to the culture medium of differentiating myoblasts for 72 h. In response to LiCl, myotube number and size were increased (Fig. 1A, bottom right), which was reflected by a doubling of the myogenic index (Fig. 1B). Interestingly, compared with the control, the mean number of nuclei per myotube demonstrated a stronger increase in response to IGF-I than to LiCl (control: 5.53 ± 0.15, IGF-I: 7.36 ± 0.17, and LiCl: 6.10 ± 0.07; P < 0.05, LiCl vs. control; P < 0.001, IGF-I vs. control and IGF-I vs. LiCl). A more detailed evaluation of the myogenic index by examination of the nuclear distribution in the myotubes revealed that IGF-I caused a shift in the number of cells fused into large myotubes (>15 nuclei/myotube), whereas LiCl increased the number of smaller myotubes (3–8 nuclei/myotube) as well as bigger myotubes (>15 nuclei/myotube) compared with control (Fig. 1C). Our data demonstrate that IGF-I or inhibition of GSK-3β by LiCl stimulates myotube formation by increasing myonuclear accretion through enhanced myoblast fusion.
IGF-I treatment or inhibition of GSK-3β by LiCl stimulates muscle-specific protein expression.
Because both IGF-I and LiCl administration promoted myogenic differentiation based on the increase in myogenic index, we next evaluated whether muscle-specific protein expression was enhanced after IGF-I or LiCl administration. The activity of creatine kinase, an enzyme specifically expressed in mature skeletal and cardiac muscle, was assessed. Muscle-specific enzyme activity (U/mg) of MCK was significantly upregulated in a dose-dependent fashion in response to IGF-I or LiCl treatment (Fig. 3A). This increase was more apparent after LiCl treatment compared with IGF-I administration. Further evaluation of these data demonstrated that IGF-I and LiCl induced a comparable increase in overall MCK activity (Fig. 3B, left and right, respectively). However, a parallel increase in total protein accumulation in response to IGF-I (Fig. 3B, left) attenuated the increase in specific MCK activity by IGF-I (Fig. 3A). In contrast, LiCl increased total protein accumulation to a lesser extent than IGF-I (Fig. 3B, right), resulting in a stronger increase in specific MCK enzyme activity (Fig. 3A).
To further investigate whether IGF-I or GSK-3β inhibition by LiCl stimulated muscle-specific gene expression, we used a stable C2C12 cell line with a TnI promoter reporter construct integrated in the genome (38). LiCl treatment resulted in a significant dose-dependent increase in TnI luciferase activity (RLU/mg) expressed as a percentage of control (72 h DM), which was not observed for IGF-I (Fig. 4A). However, further evaluation of these results showed that uncorrected TnI luciferase activity was elevated after IGF-I treatment (Fig. 4B, left) but that because of an even stronger increase in protein accumulation (Fig. 4B, left), specific TnI luciferase activity did not increase in response to IGF-I (Fig. 4A). In contrast, inhibition of GSK-3β with LiCl stimulated TnI luciferase activity (RLU/ml) to a higher extent than total protein (Fig. 4B, right), resulting in a sustained increase of specific TnI luciferase activity (Fig. 4A). The MCK and TnI-luciferase data suggest that inhibition of GSK-3β by LiCl stimulated myogenic differentiation by selectively increasing muscle-specific protein expression, whereas stimulation of myogenesis by IGF-I also appeared to involve a general increase in protein synthesis in differentiating myoblasts.
IGF-I and LiCl increase myogenic mRNA accumulation.
To assess whether the effects of IGF-I or LiCl on myogenesis also were apparent at the mRNA level, we cultured differentiating C2C12 myoblasts in the presence or absence of IGF-I or LiCl and evaluated the accumulation of various muscle-specific mRNA species using RPA. Because histone 3.3 (His 3.3) remains constant during myogenic differentiation (57), data were normalized to His 3.3. Similar responses were obtained after normalization to GAPDH as housekeeping gene (not shown).
Under control conditions, mRNA levels of genes encoding the muscle-specific transcripts MyoD, myogenin, TnI-slow, TnI-fast, MCK, and myoglobin increased as myogenesis progressed after the induction of differentiation (Fig. 5A). In contrast, His 3.2 (which is expressed in proliferating cells) decreased during differentiation and was not affected by IGF-I or LiCl (Fig. 5I). Upregulation of muscle-specific mRNA levels was markedly enhanced in myoblasts differentiated in the presence of IGF-I or LiCl (Fig. 5, B–H). MyoD and Myf-5, important transcription factors in early myogenesis, were markedly increased, respectively, by IGF-I and LiCl: MyoD (2- and 3-fold at 48 h, Fig. 5B) and Myf-5 (2- and 4-fold at 24 h, Fig. 5C). Compared with control, IGF-I or LiCl increased mRNA abundance of the following muscle-specific genes as of 48 h: myogenin (2- and 4-fold, Fig. 5D); TnI-slow (2- and 9-fold, Fig. 5E), TnI-fast (2- and 7-fold, Fig. 5F), and MCK mRNA (2- and 12-fold, Fig. 5G), whereas myoglobin mRNA levels were 1.5-fold elevated after 72 h in the presence of IGF-I or LiCl (Fig. 5H). Although both IGF-I and LiCl increased muscle mRNA expression, muscle-specific mRNA accumulation was generally stimulated to a higher extent by LiCl compared with IGF-I. In particular, the expression of the muscle regulatory factors (MRFs) MyoD (Fig. 5B), Myf-5 (Fig. 5C), and myogenin (Fig. 5D) appeared to be accelerated by LiCl. The strong elevation in muscle-specific mRNA abundance observed after IGF-I or LiCl treatment suggests that stimulation of myogenic differentiation relies on inhibition of GSK-3β activity and may be regulated at the transcriptional level.
Inhibition of GSK-3β activity and stimulation of muscle gene expression by IGF-I require PI3K activity.
Both IGF-I and LiCl inhibit GSK-3β activity (Fig. 2C) and stimulate muscle-specific mRNA accumulation during differentiation (Fig. 5). To evaluate whether inhibition of GSK-3β activity by IGF-I requires signaling through PI3K/Akt, as was shown for GSK-β phosphorylation by IGF-I (Fig. 2B), we added LY-294002 to the cells before application of IGF-I and determined kinase activity. Inhibition of GSK-3β activity by IGF-I required PI3K activity, because incubation with LY-294002 restored GSK-3β activity to control levels in the presence of IGF-I (Fig. 6A). In contrast, inhibition of GSK-3β activity by LiCl was not significantly affected by LY-294002, indicating that LiCl inhibited GSK-3β downstream of PI3K/Akt signaling.
Accordingly, stimulation of MCK expression by LiCl was refractory to PI3K inhibition, whereas MCK mRNA accumulation could not be stimulated by IGF-I when LY-294002 was present (Fig. 6B). These data suggest that stimulation of myogenic transcript accumulation by IGF-I requires signaling through the PI3K/Akt pathway to inhibit GSK-3β and demonstrate that inhibition of GSK-3β activity is sufficient to promote myogenic differentiation.
Skeletal muscle formation during muscle hypertrophy or regeneration involves myogenic differentiation of myoblasts, which relies on muscle-specific gene expression and fusion of myoblasts into myotubes or with existing myofibers, resulting in myonuclear accretion (4). The essential role of IGF-I in developmental and postnatal muscle growth is well established, but the pathways mediating IGF-I signaling in myogenic differentiation and myonuclear accretion are not fully elucidated.
Our study demonstrates that IGF-I promotes general protein accretion, muscle-specific gene expression, and myoblast fusion during myogenic differentiation. Moreover, our data suggest that stimulation of these processes by IGF-I relies on negative regulation of GSK-3β, because increased GSK-3β phosphorylation and a subsequent decrease in kinase activity were observed in response to IGF-I. The stimulatory effects of IGF-I on muscle-specific gene expression and myoblast fusion were reproduced by inhibiting GSK-3β using LiCl, indicating that inhibition of GSK-3β is sufficient to promote myogenic differentiation.
Most studies aimed at elucidation of the signal transduction pathways utilized by IGF-I to induce muscle growth have been performed in fully differentiated myotubes (41, 50, 56). However, with the use of cultured myotubes, the role of IGF-I in myonuclear accretion could not be addressed, because this requires proliferation, differentiation, and fusion of myoblasts. Administration of IGF-I to differentiating myoblasts was previously reported to stimulate myotube formation (22), although this effect was not quantitatively and qualitatively assessed except for one study (32). In line with the findings in that report, we have shown that IGF-I administration during differentiation stimulates the fusion of myoblasts into myotubes, based on an increase in the mean number of nuclei per myotube and myogenic index.
Because various studies have reported mitogenic effects of IGF-I on myoblasts (23), we assessed whether increased myotube formation could be explained by an increase in the number of myoblasts available for fusion. No increases were observed in the total number of nuclei per field when the myogenic index was determined after 72 h of differentiation (data not shown). In addition, mRNA abundance of the proliferation marker His 3.2 was not elevated in myoblasts differentiating in the presence of IGF-I compared with control (Fig. 5I). These findings support a mechanism of IGF-I-stimulated myotube formation, which relies on increased myonuclear accretion through enhanced myoblast fusion, independently of myoblast proliferation.
Because we hypothesized that IGF-I-mediated hypertrophy involved the negative regulation of GSK-3β during myonuclear accretion, the effects of GSK-3β inactivation by LiCl on myogenic differentiation were compared with those of IGF-I. LiCl administration caused an increase in the myogenic index equivalent to IGF-I, illustrating that inactivation of GSK-3β was sufficient to stimulate myonuclear accretion and suggesting that IGF-I may promote myotube formation through inactivation of GSK-3β.
Despite a similar increase in the myogenic index, subtle differences in myotube formation after IGF-I or LiCl application were apparent. Compared with control, the mean number of nuclei per myotube was particularly increased in response to IGF-I, suggesting that IGF-I stimulated the formation of large myotubes, which was confirmed by analysis of the nuclear distribution in myotubes. In contrast, LiCl stimulated the formation of both large and small myotubes compared with control, which explained the relatively small increase in the mean number of nuclei per myotube in response to LiCl.
Although these seemingly different modes by which IGF-I and LiCl stimulate myotube formation suggest the contribution of other signaling events besides inhibition of GSK-3β, they may also reflect differences in the extent and kinetics of GSK-3β inhibition by IGF-I and LiCl. In support of this notion, LiCl reduced GSK-3β activity more effectively than IGF-I did (Fig. 2C). Moreover, GSK-3β was inhibited transiently by IGF-I, because GSK-3β phosphorylation returned to baseline levels 24 h after IGF-I application (data not shown), whereas the continuous presence of LiCl likely resulted in sustained inhibition of GSK-3β activity.
In addition to the accretion of nuclei, myotube formation during myogenic differentiation is also characterized by increased protein synthesis and activation of a muscle-specific gene expression program. Consistent with previous reports (24), IGF-I treatment resulted in increased MCK activity. However, because of a parallel increase in total protein accretion, IGF-I caused only a moderate increase in specific MCK enzyme activity (corrected for total protein). Similar results were found for TnI promoter activity in a stable cell line, which appeared to be stimulated by IGF-I, but correction for total protein masked this effect, potentially as a result of the marked increase in general protein synthesis. Indeed, additional experiments using transient transfection of the TnI reporter construct and cotransfection of a constitutive active β-galactosidase reporter gene revealed a twofold increase of TnI promoter transactivation after β-galactosidase correction (data not shown).
Many investigators have reported the stimulatory effect of IGF-I on mRNA translation (54), which generally relies on enhanced activity of eukaryotic translation initiation factors (eIF) resulting from signaling through the PI3K/Akt pathway. At least three eIFs under positive control by IGF-I can be distinguished. S6- and eIF-4E-dependent translation is increased by IGF-I through Akt-dependent activation of mTOR (42). In a parallel but mTOR-independent pathway, Akt phosphorylates and inactivates GSK-3β, thereby promoting eIF-2B-dependent translation by IGF-I (33). In contrast to IGF-I, the actions of LiCl with respect to protein translation are restricted to promoting eIF-2B-dependent translation, because it inhibits GSK-3β independently of Akt. This may account for the relatively small increase in total protein observed in response to LiCl compared with the striking increase in protein accretion caused by IGF-I. A consequent robust increase in specific enzyme activity was maintained after protein correction for MCK and TnI-luciferase in response to LiCl, suggesting that inactivation of GSK-3β stimulates myogenesis by promoting muscle-specific gene expression.
Because the striking increase in general protein synthesis potentially concealed elevated muscle-specific gene expression in response to IGF-I, myogenic mRNA abundance was determined. Compared with control, IGF-I clearly enhanced myogenic mRNA accumulation, which is in line with previous reports demonstrating increased transcript levels of the MRFs myogenin and MyoD (15, 24). In addition, the expression of Myf-5, another MRF, as well as that of TnI-slow, TnI-fast, MCK, and myoglobin, was elevated in response to IGF-I, which was not yet reported. Importantly, the increased mRNA abundance of MCK, TnI-slow, and TnI-fast observed with IGF-I at 48 and 72 h after induction of differentiation demonstrated that elevated MCK activity and TnI-luciferase activity levels in response to IGF-I were not solely due to a general increase in protein translation but clearly involved pretranslational regulation.
In line with its more potent inhibition of GSK-3β, LiCl stimulated muscle-specific gene expression even more than IGF-I. Inhibition of GSK-3β enzymatic activity by IGF-I corresponded with the phosphorylation of serine 9 of GSK-3β, and both occurred in a PI3K/Akt-dependent fashion (Figs. 2B and 6A). Importantly, blockade of GSK-3β inhibition by LY-294002 abrogated the stimulatory effect of IGF-I on MCK mRNA accumulation (Fig. 6B). In contrast, GSK-3β inhibition and stimulation of MCK expression by LiCl were not affected by LY-294002, suggesting that inhibition of GSK-3β following PI3K/Akt signaling is required for the stimulatory effect of IGF-I on muscle-specific gene expression and revealing a negative regulatory role of GSK-3β in myogenesis.
Negative regulation of muscle-specific gene expression by GSK-3β may occur at the transcriptional level, and transcription factors that are controlled by GSK-3β include nuclear factor of activated T cells (NFAT) and β-catenin (1, 9). A number of studies have proposed a role for NFAT signaling in hypertrophy responses of skeletal muscle (52). NFATc1 nuclear translocation was observed during hypertrophy in C2C12 myotubes in response to IGF-I (53), and Myf-5 expression was found to be regulated by a NFAT-dependent pathway (25). NFAT has also been implicated in the regulation of myonuclear accretion. NFATc2 was recently postulated to promote myotube formation through the induction of IL-4 production by nascent myotubes, which selectively promoted myoblast-myotube fusion (31). This NFATc2-dependent process may have been stimulated by IGF-I through GSK-3β inhibition, resulting in the preferential stimulation of large myotube formation observed in our study. The stringent and sustained inhibition of GSK-3β by LiCl may have caused premature NFATc2-dependent secretion of IL-4, resulting in a general stimulation of fusion, with a consequent increase in both small and large myotubes.
In addition to NFAT, ample evidence suggests an important role for β-catenin in myogenesis (17), which relies on regulation of MyoD and Myf-5 expression (35, 48). β-Catenin is rapidly degraded upon phosphorylation by GSK-3β (1). β-Catenin abundance increases during muscle differentiation (27), and considering its essential role in cell-cell adhesion and cell morphology (20), a regulatory function of β-catenin in myotube formation is conceivable. This notion is further supported by the reported stabilization of β-catenin protein after LiCl treatment (8) and by our data demonstrating increased myotube formation when myoblasts are differentiated in the presence of LiCl. Therefore, increased myotube formation and muscle-specific gene expression observed in response to GSK-3β inhibition may occur in a NFAT- or β-catenin-dependent fashion.
One limitation of the current study is that all data regarding inhibition of GSK-3β were obtained by pharmacological inhibition using LiCl. Although LiCl is one of the best-characterized inhibitors of GSK-3β (21) and was shown to significantly decrease GSK-3β activity in this study, we controlled for nonspecific effects of LiCl by performing parallel experiments using equimolar concentrations of NaCl. Compared with control, addition of NaCl during differentiation did not affect any of the parameters described in the present study (data not shown), illustrating that the effects of LiCl were not due to osmolarity changes in the medium.
In conclusion, our study suggests that IGF-I promotes muscle-specific gene expression and myonuclear accretion during myogenic differentiation through the inactivation of GSK-3β and identifies GSK-3β as a potential target for pharmacological stimulation of muscle growth.
This work was supported by a research grant of the Netherlands Asthma Foundation (NAF 3.2.02.63).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society