Control of MyoD Function during Initiation of Muscle Differentiation by an Autocrine Signaling Pathway Activated by Insulin-like Growth Factor-II^*

The maintenance, repair, and regeneration of skeletal muscle is controlled by environmental cues mediated by systemic hormones, locally produced growth factors, and neuronal signals acting on an intrinsic genetic program involving muscle-restricted transcriptional regulatory proteins. The basic helix-loop-helix transcription factors of the MyoD family (MyoD, myf-5, myogenin, MRF4) play central roles in myogenesis throughout the lifespan (1, 2). These proteins bind as heterodimers in conjunction with more widely expressed E12/E47 transcription factors to DNA elements termed E-boxes that are found in the promoters of many muscle-specific genes (1-3). After the recruitment of transcriptional co-activators, this complex of proteins stimulates muscle gene expression (2). As determined by gene knock-out experiments in mice, MyoD and myf-5 are essential for myoblast specification during early embryonic development (4), myogenin is critical for differentiation (5, 6), and MRF4 plays a secondary role in muscle formation (7). MyoD is also necessary for satellite cell activation and muscle regeneration in the adult (8, 9). Other transcription factors, including members of the MEF2 family, are enriched in muscle and act in an accessory capacity to enhance the actions of MyoD proteins on muscle gene expression (3, 10).

Several families of peptide growth factors influence muscle growth, metabolism, and repair. Myostatin, a member of the transforming growth factor-β family, is a dominant inhibitor of muscle growth (11). In mice, cows, and a human with myostatin gene defects muscle hypertrophy develops (12-14), and overexpression of myostatin in mice leads to a profound decline in muscle mass (15). In contrast to myostatin, the insulin-like growth factors, IGF-I² and IGF-II, enhance muscle growth (16-18). Local expression of IGF-I in muscle in transgenic mice promotes hypertrophy, is able to combat the decline in muscle mass during aging, and is capable of preserving muscle mass and strength in experimental muscular dystrophy (19-23).

IGF-I and IGF-II act by binding to the IGF-I receptor, a ligand-stimulated transmembrane tyrosine-protein kinase that through a series of adaptor molecules activates several intracellular signaling cascades (24). Signaling through the IGF-I receptor is crucial for normal muscle development and function. Mice engineered to lack the IGF-I receptor have muscle hypoplasia and die in the neonatal period secondary to inability to inflate their lungs (25). A number of investigations also have established that the phosphatidylinositol 3-kinase-Akt pathway downstream of the IGF-I receptor is essential for maintaining myoblast viability, for promoting differentiation, and for stimulating hypertrophy (22, 26-33). Despite these well validated observations, only a few IGF-mediated steps acting beyond phosphatidylinositol 3-kinase and Akt in muscle cells have been identified. For example, muscle atrophy appears to be counteracted by IGFs in part via Akt-mediated inhibition of the activity of FoxO transcription factors, leading to a decline in expression of atrogin-1/MAFbx and MuRF1, muscle-specific E3 ubiquitin ligases that facilitate muscle protein breakdown (17, 34-36). Here, using as a model system C3H 10T1/2 mesenchymal stem cells acutely converted to the muscle lineage by MyoD, we show that IGF action is a necessary prerequisite for the full transcriptional activity of MyoD early in differentiation. In these cells inhibition of production of IGF-II attenuates MyoD-mediated gene expression by decreasing the recruitment of transcriptional co-activators to MyoD-occupied E-boxes on target gene promoters, thus limiting histone acetylation and chromatin remodeling, reducing the activation of RNA polymerase II, and preventing gene transcription. Our results define an early acting autocrine amplification cascade in muscle cells whereby IGF-II-mediated signaling positively regulates myogenic transcription factor function, thereby controlling terminal differentiation.

EXPERIMENTAL PROCEDURES

Chemicals and Reagents—Fetal calf serum, newborn calf serum, horse serum, Dulbecco's modified Eagle's medium (DMEM), and phosphate-buffered saline were from Mediatech-Cellgrow (Herndon, VA). TRIzol reagent and trypsin/EDTA solution were purchased from Invitrogen. Doxycycline (Dox; Clontech (Palo Alto, CA)) was dissolved in distilled water at a concentration of 1 mg/ml and stored at -20 °C until use at a final concentration of 1 μg/ml. Protease inhibitor tablets were purchased from Roche Applied Sciences, okadaic acid was from Alexis Biochemicals (San Diego, CA), and sodium orthovanadate was from Sigma. TransIT-LT-1 was from Mirus Corp. (Madison, WI). The BCA protein assay kit was from Pierce, Immobilon-FL was from Millipore Corp. (Billerico, MA), and AquaBlock™/EIA/WIB solution was from East Coast Biologicals (North Berwick, ME). Antarctic phosphatase was purchased from New England Biolabs (Ipswich, MA) and used according to the supplier's directions. Restriction enzymes, buffers, ligases, and polymerases were from Roche Applied Sciences, BD Biosciences (Clontech), and Fermentas (Hanover, MD). All other chemicals were reagent grade and were obtained from commercial suppliers.

Antibodies—The following hybridoma cell lines were purchased from the Developmental Studies Hybridoma Bank (Iowa City, IA): F5D (anti-myogenin, from W. E. Wright) and MF20 (anti-myosin heavy chain (MHC), from D. A. Fischman). F5D cells were grown in RPMI plus 10% fetal calf serum, and MF20 cells were grown in DMEM with 10% fetal calf serum. For both hybridoma lines, supernatants were harvested, clarified by centrifugation, and stored in aliquots at -20 °C until use. Concentrated ascites fluid for CT3 (anti-troponin T, from J. J.-C. Lin) was obtained from the Developmental Studies Hybridoma Bank. The monoclonal antibody to MyoD was from BD Biosciences and to α-tubulin was from Sigma-Aldrich. Polyclonal antibodies to Akt and phospho-Akt (Ser-473) and the acetyl-lysine monoclonal antibody were from Cell Signaling Technology (Beverly, MA). Polyclonal antibodies to E2A.E12 (V-18), MEF2 (C-21), MyoD (M-318), p300 (N-15), P/CAF (H-369), and Sp-1 (PEP2) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-acetyl-histone H3 (06-599) and anti-acetyl-histone H4 (06-866) were from Upstate Cell Signaling Solutions (Lake Placid, NY). The following antibody conjugates were purchased from Molecular Probes (Eugene, OR): goat anti-mouse IgG₁-Alexa 488, goat anti-mouse IgG_2b-Alexa 594, goat-anti-mouse IgG-Alexa 680. Goat-anti-rabbit IgG-IR800 was from Rockland Immunochemical Inc. (Gilbertsville, PA).

Cell Culture—C3H 10T1/2 mouse embryonic fibroblasts (ATCC catalog number CCL226) were incubated on gelatin-coated tissue culture dishes in growth medium (DMEM with 10% heat-inactivated fetal bovine serum) at 37 °C in humidified air with 5% CO₂. Differentiation was initiated after cells reached 95% of confluent density (typically 1 day after viral infection) as described below and previously (30).

Gene Transfer with Recombinant Adenoviruses—All recombinant adenoviruses have been described (30). Viruses were purified on discontinuous cesium chloride gradients and titered by optical density. Before infection, viruses were diluted in DMEM plus 2% fetal calf serum and filtered through a Gelman syringe filter (0.45 μm). Viruses were added to cells at 37 °C for 120 min at the following multiplicities of infection: 125 for Ad-tTA and 250 for Ad-MyoD. After the addition of an equal volume of DMEM with 20% fetal bovine serum, cells were incubated for a further 24 h. Cells then were washed twice with phosphate-buffered saline and incubated in differentiation medium (DM-DMEM plus 2% horse serum) without or with Ad-IGF-II^AS (multiplicity of infection of 500) with and without Dox. Under these conditions, ∼90% of cells were infected with all viruses. In the absence of infection with Ad-MyoD, no MyoD protein could be detected in 10T1/2 cells incubated in DM by either immunoblotting or immunocytochemistry.

Protein Extraction and Immunoblotting—Whole cell and nuclear protein extracts were prepared as described (30), and aliquots were stored at -80 °C until use. Protein concentrations were determined using the BCA protein assay kit. Aliquots (40 μg per lane for whole cell protein samples, 5 μg per lane for nuclear proteins) were separated by SDS-PAGE, transferred to Immobilon-FL, blocked in AquaBlock, and incubated with primary antibodies in buffer containing 50% AquaBlock, 50% phosphate-buffered saline, and 0.1% Tween 20 overnight at 4 °C. Membranes were rinsed with Tris-buffered saline-Tween and incubated at 23 °C for 1 h in 50% AquaBlock, 50% phosphate-buffered saline, 0.1% Tween 20, and 0.01% SDS with the appropriate secondary antibody conjugated to Alexa 680 or IR-800 at 1:5000. Membranes were washed according to a protocol from LiCoR and scanned on an Odyssey Infrared Imaging System using Version 1.2 analysis software (LiCoR Biosciences, Lincoln, NB). Antibodies were used at the following dilutions: anti-MHC supernatant (1:100), anti-myogenin supernatant (1:100), anti-troponin T (1:1000), anti-Akt (1:2000), anti-phospho-Akt (Ser-473) (1:1000), anti-MyoD (1:3000), anti-MEF2 (1:3000), anti-Id-1 (1:500), anti-E2A.E12 (1:1000), anti-p300 (1:1000), anti-P/CAF (1:1000), anti-Sp-1 (1:2000).

Analysis of MyoD by Two-dimensional Gel Electrophoresis and Immunoblotting—Two micrograms of nuclear protein extracts from 10T1/2 cells infected with Ad-MyoD, Ad-tTA, and Ad-IGF-II^AS, as described above, and incubated in DM with and without Dox for 24 h were diluted in a total volume of 125 μl of rehydration buffer (7 m urea, 2 m thiourea, 4% CHAPS, 60 mm DTT, 1× ready strip buffer, pH 3-10 (Bio-Rad), and 0.1% bromphenol blue) and incubated at 23 °C for 30 min. The denatured nuclear proteins were applied to an Amersham Biosciences 7-cm strip holder along with a pH 5-8 ReadyStrip immobilized pH gradient strip (Bio-Rad) and hydrated for 13 h at 20 °C in an Amersham Biosciences Ettan IPGphor isoelectric focusing system and focused at 4000 V for a total of 10,000 V-h. The immobilized pH gradient strips were then equilibrated in 5 ml of equilibration buffer (50 mm Tris-HCl, pH 8.8, 6 m urea, 30% glycerol, 2% SDS, and 0.002% bromphenol blue) with 20 mg/ml DTT for 15 min at 23 °C. The buffer was replaced with 5 ml of new rehydration buffer containing 50 mg/ml iodoacetamide and incubated for 15 min at 23 °C. The equilibrated immobilized pH gradient strips were then subjected to SDS-PAGE (12% separating gel) and immunoblotted as described above. For analysis of dephosphorylated proteins, nuclear protein extracts (10 μg) were precipitated with water-saturated chloroform:methanol (1:4 ratio), suspended in 90 μl of Antarctic phosphatase reaction buffer and 100 units of enzyme, and incubated for 2 h at 37°C. Proteins were concentrated by chloroform:methanol precipitation and suspended in 125 μl of rehydration buffer followed by isoelectric focusing, SDS-PAGE, and immunoblotting as described above.

Immunocytochemistry—Cells were fixed, permeabilized, blocked, and incubated with antibodies as previously described (30). Primary antibodies were added in blocking buffer for 16 h at 4 °C (anti-MHC, 1:50 dilution; anti-MyoD, 1:1000; secondary antibodies at 1:1000). Images were captured with a Roper Scientific Cool Snap FX CCD camera attached to a Nikon Eclipse T300 fluorescent microscope using IP Labs 3.5 software.

Analysis of Myogenin Gene Transcription—Nuclear RNA was isolated as described (37). RNA concentrations were determined spectrophotometrically at 260 nm and quality assessed by agarose gel electrophoresis. Nuclear RNA (2.5 μg) was reverse-transcribed in a final volume of 20 μl using the Superscript III-1^st strand synthesis kit (Invitrogen) with random hexamer primers. Each PCR reaction contained 1 μl of cDNA and one pair of the following primer sequences: myogenin, exon 1 (sense strand) 5′-GGGGACCCCTGAGCATTGTCC-3′, intron 1 (antisense) 5′-CCAAGGGCCCTGCTTTGCAC C-3′; S17, exon 2 (sense) 5′-ATCCCCAGCAAGAAGCTTCGGAACA-3′, intron 2 (antisense) 5′-GCCGTCACCAGCCCTCCTCCG-3′. The linear range of product amplification was established in pilot studies for each primer pair, and the cycle number representing the approximate midpoint was used in final experiments. This ranged from 25 to 30 cycles. Results were quantified using Quantity One Software (Bio-Rad) after electrophoresis through 1.2% agarose gels.

Transient Transfections and Luciferase Reporter Gene Assays—Reporter plasmids containing the mouse muscle creatine kinase promoter and the 4× E-box have been reported previously (31). A 249-bp fragment of the mouse myogenin proximal promoter was isolated from genomic DNA using PCR and the following primers: 5′-GGGCCAGGGGCAGGCCTGCAG-3′ (sense) and 5′-CATCAGGTCGCAAAAGGCTTG-3′ (antisense). The purified PCR product was ligated into the pGEM-T Easy vector (Promega Inc., Madison, WI), and recombinants were identified by restriction enzyme digestion and DNA sequencing. The promoter fragment was isolated by restriction enzyme digestion with EcoRI, the ends were made blunt with Klenow DNA polymerase, and the fragment was ligated into the SmaI site of pGL3 basic (Promega). Recombinant plasmids with a correctly oriented myogenin promoter were identified by restriction endonuclease analysis. For reporter gene experiments, C3H 10T1/2 cells were seeded at 1 × 10⁵/well of 12-well tissue culture dishes, and each well was transfected the following day with 100 ng of a promoter-reporter plasmid using TransIT LT-1. One day later cells were infected sequentially with recombinant adenoviruses as described above. After incubation in DM with and without Dox for 18-20 h, cell extracts were assayed for luciferase activity using a kit from Promega and a PerkinElmer Life Sciences luminometer. Enzymatic activity was normalized to protein concentration.

Chromatin Immunoprecipitation Assays—Initial steps were modified from published protocols (38). Aliquots for immunoprecipitation were derived from ∼1 × 10⁶ cells and contained 5 μg of specific antibody or 10 μl of ascites fluid. We determined that all antibodies used for chromatin immunoprecipitation (ChIP) assays detected the appropriate protein by immunoblotting of nuclear protein extracts. DNA was isolated using the Qiagen QiaQuick PCR purification kit following the steps recommended by the supplier, except that nine volumes of PB (Qiagen) were added to each sample before applying it to the spin column, and DNA was eluted with 100 μl of buffer (10 mm Tris-Cl, 1 mm EDTA, pH 8.0). PCR reactions were performed with 1 μl of input DNA or 4 μl of immunoprecipitated DNA with the following primers for the myogenin promoter: 5′-GAATCACATGTAATCCACTGG-3′ (sense strand) and 5′-CACACCAACTGCTGGGTGCCA-3′ (antisense). The linear range of product amplification was established for each primer pair in pilot studies, and the cycle number that reflected the approximate midpoint was used in final experiments. This varied from 30 to 35 cycles. Results were visualized after electrophoresis through 1.5% agarose gels and were quantified by densitometry after normalization to signals obtained from input DNA. The extent of myogenin gene transcription was assayed in parallel using nuclear RNA, as described above, and differentiation was assessed by immunoblotting for expression of muscle-specific proteins.

Statistical Analysis—Data are presented as the mean ± S.E. Statistical significance was determined using a paired Student's t test. Results were considered statistically significant when p < 0.05.

RESULTS

Defining Early Events in MyoD-mediated Muscle Differentiation at the Myogenin Gene—The critical roles of MyoD and related basic helix-loop-helix transcription factors in muscle cell determination and differentiation have been recognized for nearly two decades (for review, see Ref. 2). Many studies have demonstrated that these proteins individually can convert uncommitted mesenchymal precursors to the myoblast lineage (39), as we showed recently using 10T1/2 cells acutely infected with a recombinant adenovirus encoding mouse MyoD (Ad-MyoD) (30). Using this model we now have analyzed early events in MyoD-initiated muscle differentiation, focusing on the myogenin gene, which is a direct transcriptional target of MyoD (for review, see Ref. 2). Fig. 1 shows the results of experiments examining the kinetics and consequences of binding of MyoD to the proximal myogenin promoter in myoblasts. As assessed by ChIP in Fig. 1B, MyoD is already found on DNA in chromatin at this site at the onset of incubation of cells in DM ∼ 24 h after infection of cells with Ad-MyoD. Binding appeared to increase over the ensuing 4 h and then declined but remained measurable for up to 24 h. Similar results were observed for MEF2 proteins, which can interact directly with MyoD but which also recognize a DNA sequence adjacent to the E-box in the proximal myogenin promoter (10). By contrast, the presence of E12/E47, the hetero-dimerization partner of MyoD on DNA (2), was relatively constant. MyoD and MEF2 have been found to interact with the transcriptional co-activators p300 and P/CAF (40), and both proteins are able to facilitate gene activation in part through their acetylation of histones in chromatin (41, 42). Both p300 and P/CAF also have been shown to acetylate MyoD on lysine residues within its DNA binding domain (43, 44), thereby increasing its affinity for E-box elements (45-47). As seen in Fig. 1B, p300 and P/CAF are each recruited to the proximal myogenin promoter early in differentiation, preceding the sustained acetylation of histones H3 and H4 at this site. Among the main consequences of recruitment of a transcriptionally competent complex of activators to the myogenin promoter are first, activation of RNA polymerase II, as evidenced by its phosphorylation on serine 5 within its COOH-terminal domain repeating motifs (48), which occurs within4hof incubation of cells in DM (Fig. 1B), and second, initiation of myogenin gene transcription, which is seen as a sustained rise in nascent nuclear myogenin RNA beginning by 8 h (Fig. 1C). Thus, activation of myogenin gene transcription occurs as a very early event in MyoD-regulated muscle differentiation and is preceded by the accumulation of key regulatory proteins, including MyoD, at the proximal myogenin promoter.

IGF-II Action Is Required for MyoD-stimulated Myogenin Gene Transcription and Muscle Differentiation—In previous studies we established that inhibition of IGF-II expression in differentiating myoblasts through delivery of an IGF-II antisense RNA by a doxycycline-regulated adenovirus (Ad-IGF-II^AS) reversibly blocked MyoD-mediated muscle differentiation (30). We documented that endogenous production of IGF-II and subsequent activation of the IGF-I receptor and the phosphatidylinositol 3-kinase-Akt signaling pathway were critical components of an autocrine amplification cascade that appeared necessary for full differentiation to occur (30). We now asked if IGF action initiated by muscle-derived IGF-II was required for early events in MyoD-regulated differentiation, including induction of myogenin gene transcription. As depicted in Figs. 2, A and B, myoblasts expressing IGF-II^AS RNA did not differentiate. No multinucleated myotubes were seen, and the production of muscle proteins, including myogenin, MEF2, troponin T, and myosin heavy chain was severely attenuated. In addition, activation of Akt, as measured by its phosphorylation on serine 473, was reduced to basal background levels.

Inhibition of IGF-II expression by Ad-IGF-II^AS also prevented the activation of myogenin gene transcription seen early in differentiation, as shown by a lack of increase in abundance of nascent nuclear myogenin RNA after incubation of cells in DM (Fig. 2C) and by the 70% reduction in activity of a transfected luciferase reporter gene containing the myogenin promoter (Fig. 3A). The impairment in muscle-specific gene activation in IGF-II-deficient myoblasts was not limited to myogenin, as activity of the muscle creatine kinase promoter was inhibited by more than 75% (Fig. 3B), and expression of a minimal promoter containing 4 tandem copies of an E-box derived from the muscle creatine kinase gene (49) was reduced by 80% in the absence of IGF-II. A similar decline in activity of the 4× E-box-containing promoter also was observed in myoblasts expressing a dominant-negative version of Akt.³ In aggregate, these results demonstrate that several genes controlled by MyoD during muscle differentiation require IGF signaling through Akt for their full transcriptional potential to be reached.

Inhibition of IGF-II Does Not Alter Nuclear Expression of MyoD or E12/E47 and Does Not Prevent Post-translational Modifications of MyoD—We next addressed potential mechanisms by which inhibition of endogenous IGF-II expression blocked MyoD-mediated myogenin gene transcription and muscle differentiation. We first examined MyoD and E12/E47. As visualized in Figs. 2A and 4A, MyoD was found primarily in the nucleus whether or not IGF-II expression was prevented, although multinucleated cells were formed only when IGF-II was not inhibited. As assessed by biochemical fractionation of myoblasts in Fig. 4B, the abundance of MyoD and of E12/E47 within the nucleus was not altered by IGF-II deficiency, although myogenin and MEF2 proteins only accumulated when IGF-II antisense RNA was blocked by doxycycline. The abundance of the transcriptional co-activators p300 and P/CAF also remained relatively constant under all conditions and, as expected, so did levels of a resident nuclear protein, the transcription factor Sp-1.

The activity of MyoD and other basic helix-loop-helix transcription factors can be prevented by Id proteins, which can act as dominant inhibitors of muscle differentiation (39). The abundance of Id proteins normally declines during myoblast differentiation (50). As pictured in Fig. 4C, Id-1 levels decreased to the same extent during incubation in DM whether or not IGF-II expression was impaired. Thus, the actions of endogenous IGF-II to regulate myoblast differentiation do not affect the overall abundance or nuclear accumulation of MyoD or E12/E47 and do not influence the normal loss of Id expression.

MyoD is highly phosphorylated in cells (51) and also is acetylated at several sites within its DNA binding domain by the transcriptional co-activators p300 and P/CAF (43, 47). To assess whether IGF-II deficiency could lead to changes in the extent of post-translational modifications of MyoD, we performed immunoblotting after sequential isoelectric focusing and SDS-PAGE of nuclear protein extracts isolated from cells infected with Ad-IGF-II^AS and incubated in DM with and without Dox for 24 h. As depicted in Fig. 5A, multiple superimposable spots reactive with MyoD antibodies were detected at an apparent M_r of 42 whether or not IGF-II was expressed, indicating that the extent of phosphorylation of MyoD was not dependent on active signaling through the IGF-I receptor. We also looked at MyoD after its in vitro dephosphorylation by incubating nuclear protein extracts with Antarctic phosphatase. As expected, after phosphatase treatment the pI of MyoD increased as did its mobility on SDS-PAGE from 42 to ∼34 kDa (51), but the same three immunoreactive spots were observed whether or not IGF-II expression was inhibited (Fig. 5B, upper panel). Because immunoblotting with antibodies to acetyl-lysine also gave identical results whether or not IGF-II deficiency was imposed (Fig. 5B, lower panel), we conclude based on these results that although blocking signaling through the IGF-I receptor impairs the transcriptional actions of MyoD, it has minimal effects on its overall abundance, nuclear location, phosphorylation, or acetylation.

Signaling by IGF-II Is Required for Sustained Recruitment of MyoD-associated Proteins at the Myogenin Promoter—We next asked if IGF-II deficiency altered protein-DNA interactions at a MyoD target gene. Myoblasts were incubated in differentiation medium with and without Dox for 8 or 24 h, and recruitment of transcriptional regulatory proteins to the myogenin promoter was assessed by ChIP. As shown in Fig. 6A, binding of MyoD and E12/E47 to DNA in chromatin at the promoter was not altered by reduction of IGF-II expression, although as expected binding of MEF2 declined since its total abundance decreased (see Fig. 4B). Accumulation of co-activators p300 and P/CAF at the myogenin promoter also were reduced, as was acetylation of core histones H3 and H4. The decline in co-activator recruitment to the myogenin promoter occurred even though the abundance of these proteins in the nucleus appeared to be unchanged in IGF-II-deficient myoblasts compared with normal cells (Fig. 4B). The net outcome of this decrease in development of a transcriptionally competent complex of activators was a decline in phosphorylation of RNA polymerase II at serine 5, leading to a marked impairment of myogenin gene transcription, as illustrated in Fig. 6B (and in Fig. 2C). Thus, diminished IGF-mediated signaling blocks MyoD-regulated muscle differentiation through a reduction in recruitment of transcriptional co-activators to a MyoD-dependent gene promoter.

DISCUSSION

The central role of MyoD in skeletal muscle specification and differentiation has been recognized for nearly two decades (52). More recent experimental evidence has implicated MyoD as the key agent in orchestrating the myogenic differentiation program in cultured cells through its function as a nodal point in recruiting components of a “feed-forward” amplification circuit that interacts to specify appropriate temporal expression of muscle genes (2). In support of this idea, Tapscott and co-workers (53) have shown using an experimental model of MyoD-mediated determination and differentiation of mouse embryonic fibroblasts that although MyoD alone can activate transcription of myogenin very early in muscle differentiation, it requires assistance from MEF2D and p38 MAPK to induce other genes normally expressed at later phases. Because MEF2D gene activation and stimulation of p38 kinase activity are both MyoD-regulated events that occur relatively early in differentiation (53, 54), and because phosphorylation of MEF2D by p38 is necessary for its full transcriptional activity to be achieved (53), their results are consistent with the feed-forward hypothesis. Other studies also have provided evidence that p38 can contribute to muscle gene promoter activation through interactions with components of the Swi/Snf chromatin remodeling complex (55).

Against this background, the studies presented here argue for another, even earlier-acting feed-forward cascade in MyoD-mediated muscle differentiation. Using a model of acute MyoD-stimulated differentiation of 10T1/2 mesenchymal stem cells similar to that employed by Tapscott and co-workers (54), we find that autocrine signaling through the IGF-I receptor by muscle-derived IGF-II is necessary for the full transcriptional effects of MyoD at the myogenin promoter within the first few hours after initiation of differentiation. Blockade of this pathway inhibits stimulation of myogenin gene expression and prevents subsequent biochemical and morphological differentiation, whereas its activation contributes to the assembly and/or maintenance of a transcriptionally competent complex of proteins at the myogenin promoter that is necessary for sustained induction of myogenin gene expression. Our results, thus, define an obligatory feed-forward loop that acts early in differentiation on the transcriptional functions of MyoD.

Our current observations also complement and extend previous studies, in which we found that inhibition of endogenous IGF-II expression both impaired survival and prevented differentiation of C2 myoblasts (27, 28). In these IGF-II-deficient cells, myoblast death could be blocked and differentiation induced by ligand-mediated activation of the IGF-I receptor or by a constitutively active catalytic subunit of phosphatidylinositol 3-kinase, leading to stimulation of MyoD gene expression and protein accumulation in the nucleus (28). In contrast, forced expression of MyoD in IGF-II-deficient C2 myoblasts was insufficient to promote differentiation, although it did enhance survival through production of the cyclin-dependent kinase inhibitor, p21 (28). Thus, in a skeletal muscle cell line as well as in MyoD-converted mesenchymal stem cells, the ability of MyoD to induce differentiation by transcriptional activation of muscle gene expression is dependent on signaling through the IGF-I receptor.

Cell-based ChIP analyses revealed that MyoD along with its DNA binding partner E12/E47 is rapidly recruited to the proximal myogenin promoter within a few hours after incubation of cells in differentiation-promoting medium and is accompanied by MEF2 and the transcriptional co-activators p300 and P/CAF, with consequent acetylation of core histones H3 and H4 and the additional recruitment and activation of RNA polymerase II. The absence of signaling through the IGF-I receptor did not change the abundance of MyoD or E12/E47 in the nucleus or at the myogenin promoter and, perhaps surprisingly, did not alter the extent of post-translational modifications of MyoD, specifically its phosphorylation or acetylation. Rather, lack of IGF signaling caused a decline in binding of p300 and P/CAF at the myogenin locus without a significant change in abundance of these proteins in the nucleus, leading to loss of histone acetylation at this site and diminished recruitment of RNA polymerase II phosphorylated at serine 5. Taken together, these observations provisionally place p300 and/or P/CAF and not MyoD as downstream targets of IGF action in differentiating myoblasts. Our results are compatible with previous studies showing that deficiency of p300 but not the related transcriptional co-activator, CBP (cAMP-response element-binding protein (CREB)-binding protein), impaired myogenic differentiation of embryonic stem cells (56) and are consistent with other data demonstrating a critical role for p300 and P/CAF in the actions of MyoD (45, 46, 57).

In previous experiments we showed in this model system that Akt is activated downstream of the IGF-I receptor and found that a dominant-negative Akt prevented MyoD-mediated muscle differentiation in a manner similar to inhibition of IGF-II production (30). Akt has been linked to muscle development, regeneration, and hypertrophy (17, 32, 33) through its prevention of apoptotic cell death (26-28), through its blockade of pathways activated in muscle atrophy (22, 34, 35), and through its stimulation of muscle protein biosynthesis via the mTOR protein kinase (17, 22, 32). In mammals, the three isoforms of Akt are products of distinct genes but are nearly 85% identical to each other in amino acid sequence and are activated by similar stimuli (58, 59). In mice, Akt1 and Akt2 appear to be important for muscle development, as their combined deficiency results in muscle hypoplasia (60), a phenotype similar to targeted loss of the IGF-I receptor (25). In contrast, individual knockouts of Akt1, -2, or -3 have no effect on muscle mass or strength (61-63). It is unclear which Akt is involved in facilitating MyoD-mediated gene transcription. In cultured myoblasts only Akt2 is induced during differentiation, although expression of Akt1 is maintained at constant levels in proliferating and differentiating cells (64, 65). A connection between the actions of Akt in the cytoplasm and its effects on gene transcription in the nucleus has been explored only provisionally. Akt signaling inhibits FoxO transcription factors through a biochemical mechanism that involves their multisite phosphorylation and translocation from the nucleus to the cytoplasm (66, 67). In this way, by blocking expression of FoxO-stimulated genes encoding the muscle-enriched E3 ubiquitin ligases atrogin-1/MAFbx and MuRF1, Akt inhibits the muscle protein breakdown that culminates in atrophy (34, 35). Signaling by Akt also has been linked to the biological effects of p300, but to date only in cancer cells (68, 69). It has been shown in a human ovarian carcinoma cell line that phosphorylation of p300 by Akt was necessary to maintain its stability; when the phosphatidylinositol 3-kinase-Akt pathway was blocked, the half-life of p300 declined dramatically (68). However, in muscle cells the abundance of p300 remains constant (45) even when activation of the IGF-I receptor is prevented (see Fig. 4B). It also has been found that treatment of a human alveolar epithelial cancer cell line with tumor necrosis factor α leads to the activation and nuclear translocation of Akt, its association with p300, and the subsequent phosphorylation of p300 (69). The net effect of these events in this cancer cell line appears to be the increased interaction of p300 with P/CAF and potentially with chromatin at a tumor necrosis factor α target gene, ICAM (intercellular adhesion molecule), with the subsequent transcriptional stimulation of ICAM, although in this study the DNA sequence-specific transcription factors responsible for ICAM gene activation were not evaluated (69). Although in some reports Akt2 has been detected in muscle cell nuclei (70), in our hands both Akt1 and Akt2 are cytoplasmic,³ whereas p300 and P/CAF are exclusively nuclear. It, thus, remains to be established how p300 and/or P/CAF are regulated by IGF-stimulated signaling pathways in muscle cells and to define the components acting downstream of Akt to mediate its potential nuclear actions in differentiating myoblasts. A model outlining our observations appears in Fig. 7.

Recent studies in a variety of species have delineated the negative impact of signaling through the IGF-I receptor to accelerate aging and tissue senescence (71) and to enhance cancer risk and progression (72). Any future targeted therapeutic use of components of IGF signaling pathways to promote muscle regeneration will require more detailed understanding of the specific biochemical mechanisms through which these proteins interact with muscle transcriptional programs.