Report

Regulation of Distinct Stages of Skeletal Muscle Differentiation by Mitogen-Activated Protein Kinases

See allHide authors and affiliations

Science  14 Nov 1997:
Vol. 278, Issue 5341, pp. 1288-1291
DOI: 10.1126/science.278.5341.1288

Abstract

The signal transduction pathway or pathways linking extracellular signals to myogenesis are poorly defined. Upon mitogen withdrawal from C2C12 myoblasts, the mitogen-activated protein kinase (MAPK) p42Erk2 is inactivated concomitant with up-regulation of muscle-specific genes. Overexpression of MAPK phosphatase-1 (MKP-1) inhibited p42Erk2 activity and was sufficient to relieve the inhibitory effects of mitogens on muscle-specific gene expression. Later during myogenesis, endogenous expression of MKP-1 decreased. MKP-1 overexpression during differentiation prevented myotube formation despite appropriate expression of myosin heavy chain. This indicates that muscle-specific gene expression is necessary but not sufficient to commit differentiated myocytes to myotubes and suggests a function for the MAPKs during the early and late stages of skeletal muscle differentiation.

C2C12 myoblasts proliferate in response to mitogens and upon mitogen withdrawal differentiate into multinucleated myotubes (1). Although much progress has been made in defining the mechanisms governing myogenesis at the transcriptional level (1, 2), the signal transduction pathways involved in myogenesis are poorly defined. Differentiation of C2C12 myoblasts is inhibited by serum, growth factors (3), oncogenic tyrosine kinases (4), and oncogenic forms of the small guanine nucleotide-binding protein Ras (5). These observations implicate the Ras pathway in relaying to the nucleus extracellular signals that repress myogenesis. One of the effector pathways of Ras leads to activation of MAPKs, which have been implicated directly in regulating proliferation and differentiation by phosphorylation of transcription factors in a variety of cellular systems (6,7). However, the function of the MAPKs during skeletal muscle differentiation has not been elucidated.

We induced C2C12 myoblasts to differentiate by transfer from growth medium (GM) to differentiation medium (DM) (8). Within 24 hours in DM, p42Erk2 activity (9) was reduced by 50%, whereas the activity of the related MAPK, c-Jun NH2-terminal kinase (JNK), remained unaffected (Fig.1A). Concomitantly, expression of the skeletal muscle-specific basic helix-loop-helix transcription factors MyoD and myogenin and the cell cycle inhibitor p21 (10) were increased, marking the onset of myogenesis and cell cycle withdrawal (Fig. 1B). By day 3, multinucleated myotubes had formed (11), correlating with expression of the terminal differentiation marker myosin heavy chain (MHC) (Fig. 1B). Between days 2 and 3 of differentiation, p42Erk2 activity remained low, whereas JNK activity increased by 1.8-fold (Fig. 1A). The amount of p42Erk2 and JNK proteins remained constant throughout differentiation (11).

Figure 1

Activity of p42Erk2 and JNK during C2C12 differentiation (A) C2C12 myoblasts were induced to differentiate by transfer from GM to DM. At the times indicated, p42Erk2 (closed bars) and JNK kinase (open bars) activities were assessed. Results are representative of the mean ± SEM from five separate experiments for p42Erk2 and four separate experiments for JNK. (B) C2C12 myoblasts were assessed for their progression through differentiation by immunoblot analysis for expression of MyoD, myogenin, MHC, and p21 at the indicated times.

These results suggest that inactivation of p42Erk2 might be required for C2C12 myoblasts to initiate myogenesis. If so, ectopic expression of a dual-specificity protein phosphatase with substrate specificity toward the MAPKs should facilitate myogenesis in the presence of mitogens. We sought to overexpress the MAPK phosphatase-1 (MKP-1), which inactivates the MAPKs selectively by dephosphorylation (12). We generated stable C2C12 lines expressing the first 314 residues of MKP-1 fused at the COOH-terminus to the Myc epitope (MKP-1–Myc) under the control of a tetracycline-repressible promoter (13). Ten clones were isolated that inducibly express MKP-1–Myc (14). Results presented were obtained from one such clone, C2MKP-1; similar results were obtained from two other clones.

C2MKP-1 myoblasts cultured for 3 days in GM in the absence of tetracycline expressed amounts of MKP-1–Myc similar to those of endogenous MKP-1 (Fig. 2A). Maximal MKP-1–Myc expression was achieved within 3 days after transfer of C2MKP-1 myoblasts to tetracycline-free GM (11), resulting in the inactivation of p42Erk2 and JNK (Fig. 2B) and inhibition of C2MKP-1 myoblast proliferation (Fig. 2C). The activities of p42Erk2 and JNK in parental C2C12 myoblasts were unaffected by tetracycline (Fig.2B). These data indicate that MKP-1–Myc expression is sufficient to interfere with endogenous MAPK activity and myoblast growth. We tested whether increased expression of MKP-1–Myc in C2MKP-1 myoblasts could facilitate muscle-specific gene expression in the presence of mitogens. C2MKP-1 myoblasts cultured in GM containing tetracycline failed to express detectable levels of MyoD, myogenin, or MHC (Fig.3A). However, when C2MKP-1 myoblasts were cultured in GM in the absence of tetracycline, MyoD and myogenin were detected by day 4, and by day 5, expression of MHC was detected (Fig.3A). MyoD expression is linked to cell cycle withdrawal through MyoD-induced transcriptional activation of p21 (15). We investigated whether expression of p21 was also facilitated by MKP-1–Myc. C2MKP-1 myoblasts cultured in GM in the absence of tetracycline expressed p21 after 3 to 4 days (Fig. 3A). Control C2MKP-1 myoblasts cultured in GM containing tetracycline expressed p21 more slowly; this likely occurs because these cells proliferate more rapidly than C2MKP-1 myoblasts grown in the absence of tetracycline and consequently achieve density-dependent arrest. In contrast, the cell cycle inhibitor p27, which is not regulated at the transcriptional level by MyoD (16), remained unaffected (Fig. 3A). These results demonstrate that inhibition of p42Erk2 or JNK, or both, by MKP-1 in C2MKP-1 myoblasts, in the presence of mitogens, is sufficient to initiate the process of muscle-specific gene expression, analogous to that of mitogen withdrawal. Because p42Erk2 but not JNK activity declines during normal C2C12 differentiation induced by mitogen withdrawal, p42Erk2 likely participates in mediating the inhibitory effects of mitogens on myoblast differentiation. This notion was supported by the observation that treatment of C2C12 myoblasts in GM with an inhibitor (PD 098059) of the dual-specificity kinases MEK [or MAPK/extracellular signal-regulated kinase (ERK) kinases] 1 and 2, which activate the Erks, also facilitated expression of MHC (11).

Figure 2

Inhibition of p42Erk2 and JNK activity and C2MKP-1 myoblast proliferation by overexpression of MKP-1. (A) C2MKP-1 myoblasts were cultured in GM in either the absence (–) or presence (+) of tetracycline (2 μg/ml) for 3 days. Cell lysates were prepared, and MKP-1 was immunoprecipitated. The resulting immunocomplexes were resolved by 10% SDS-PAGE, transferred to PVDF membrane, and immunoblotted with anti-MKP-1. Molecular size standards (Sigma) in kilodaltons are shown on the left. (B) Parental C2C12 and C2MKP-1 myoblasts were grown in either the presence or absence of tetracycline for 3 days. Lysates were prepared from C2C12 and C2MKP-1 myoblasts, and p42Erk2 (closed bars) and JNK (open bars) activities were assessed. Immunoblots from the respective kinase assays confirmed that equal amounts of p42Erk2 and JNK were precipitated. Results presented show the mean ± SEM of four independent experiments. Statistical significance was derived using a Student's t test (one-way): *, P < 0.01 and **, P < 0.05. (C) C2MKP-1 myoblasts were cultured in GM in either the presence (squares) or absence (triangles) of tetracycline (2 μg/ml) for 5 days. At the times indicated, C2MKP-1 myoblasts were trypsinized, and viable cells that excluded trypan blue were counted. The data are representative of four independent experiments done in duplicate.

Figure 3

Increased muscle-specific gene expression after overexpression of MKP-1 in GM. C2MKP-1 myoblasts were cultured in GM in either the absence or presence of tetracycline (2 μg/ml). At the indicated times, lysates were prepared and expression of (A) MyoD, myogenin, MHC, p21, and p27 and (B) cyclin D1 and Cdk4 was assessed by immunobloting. (C) NIH 3T3 cells were cotransfected with the indicated expression plasmids, and 48 hours after transfection, cells were harvested and the luciferase and β-galacatosidase activities were measured. Data are representative of four separate experiments and represent the mean ± SEM from triplicate transfections expressed as luciferase units/β-galactosidase units.

Our data indicate that p42Erk2 may couple extracellular signaling events to the process of myogenesis; however, the mechanism or mechanisms of p42Erk2 regulation of muscle-specific gene expression is unclear. The amounts of cyclin D1 decline early during C2C12 differentiation (17, 18), and cyclin D1 negatively regulates the transcriptional activity of MyoD (17, 19). Therefore, we tested the possibility that overexpression of MKP-1 in C2MKP-1 myoblasts could decrease expression of cyclin D1. C2MKP-1 myoblasts grown in the absence of tetracycline expressed reduced amounts of cyclin D1 between days 4 and 6, compared with C2MKP-1 myoblasts grown in the presence of tetracycline (Fig. 3B). This is consistent with maximal MKP-1–Myc expression by day 3 and inhibition of cell proliferation in MKP-1-Myc–expressing myoblasts (Fig. 2, B and C). Tetracycline did not induce detectable changes in the amounts of Cdk4 protein (Fig. 3B). The effects of MKP-1 and p42Erk2 on transcriptional activation of cyclin D1 were assessed with the human cyclin D1 promoter fused to the luciferase gene [CycD1-Luc (20, 21)]. Cotransfection of constitutively active MEK [MEK(EE)] but not wild-type MEK [MEK(WT)] increased CycD1-Luc activity by about threefold (Fig. 3C). When MEK(EE) was cotransfected with an amount of MKP-1 cDNA that did not affect basal CycD1-Luc activity, the MEK(EE)-inducible component of CycD1-Luc activity was inhibited (Fig.3C). Similar effects of p42Erk2 on cyclin D1 transcription have been described in CCL39 cells (22). These data support the notion that early during myogenesis, inactivation of p42Erk2 is coupled to the transcriptional repression of cyclin D1. Therefore, the decision between either proliferation or differentiation in C2C12 myoblasts is coordinated by the amount of p42Erk2 activity. Consistent with this, we observed that MEK(EE) was also sufficient to repress the transcriptional activity of MyoD (23).

By overexpressing MKP-1, we demonstrated that the p42Erk2/MAPK pathway may supress the early stages of myogenesis and thus couple extracellular signals to regulate skeletal muscle differentiation. However, expression of endogenous MKP-1 declines in differentiated multinucleated myotubes (Fig. 4A), suggesting that decreased MKP-1 activity might serve to modulate the MAPKs later during myogenesis. To assess the functional significance of MKP-1 down-regulation during skeletal muscle differentiation, we induced C2MKP-1 myoblasts to express MKP-1–Myc during differentiation initiated by mitogen withdrawal. Ectopic expression of MKP-1–Myc before and during C2MKP-1 myoblast differentiation did not affect the expression of MyoD, myogenin, p21, or MHC (Fig. 4B). Despite the appropriate expression of MHC, C2MKP-1 myocytes expressing MKP-1–Myc during differentiation (Fig. 5A) failed to undergo appropriate cell-cell fusion to form multinucleated myotubes (Fig. 5B). In contrast, C2MKP-1 myoblasts grown in the presence of tetracycline differentiated and fused to form multinucleated myotubes (Fig. 5B). Under these conditions, both p42Erk2 and JNK activities were inhibited by ∼50% in myocytes expressing MKP-1–Myc in DM (11). Readdition of tetracycline to repress expression of MKP-1–Myc (Fig. 5A) rescued the fusion defect (Fig. 5B). Thus, inhibition of myotube formation by MKP-1 overexpression is reversible. The formation of multinucleated myotubes upon readdition of tetracycline to differentiated C2MKP-1 myocytes was not accompanied by changes in the amounts of MHC (Fig. 5C), indicating that the formation of multinucleated myotubes upon readdition of tetracycline stemmed from the preexisting population of differentiated myocytes.

Figure 4

Decreased expression of MKP-1 in multinucleated myotubes, but lack of effect of ectopic MKP-1–Myc expression on muscle-specific gene regulation during differentiation. (A) C2C12 myoblasts were cultured in GM and then transferred to DM for 3 days. MKP-1 protein expression was measured by immunoblotting. Molecular size markers (in kilodaltons) are shown on the left. (B) C2MKP-1 myoblasts were cultured in GM either in the presence or absence of tetracycline for 3 days to induce MKP-1–Myc expression. Differentiation was then induced by transferring cultures to DM for 3 days in either the continued absence or presence of tetracycline. Lysates prepared from C2MKP-1 myoblasts at 3 days in GM and DM were assessed by immunoblotting for the expression of MyoD, myogenin, MHC, p21, and MKP-1–Myc.

Figure 5

Prevention of multinucleated myotube formation by overexpression of MKP-1. (A) C2MKP-1 myoblasts were cultured in GM in either the presence or absence of tetracycline for 3 days and were then transferred to DM for 3 days in either the continued absence or presence of tetracycline. On day 3, myoblasts cultured in DM in the absence of tetracycline were transferred to DM containing tetracycline, and MKP-1–Myc expression was assessed 2 and 3 days later. MKP-1–Myc expression was assessed by immunoprecipitation with anti-MKP-1 [preimmune (PI) anti-serum was used as control], followed by immunoblot analysis with antibody to Myc (9E10). (B) C2MKP-1 myoblasts were cultured in GM for 3 days in the presence of tetracycline, then transferred to DM for 3 days in the presence of tetracycline (left); myoblasts were cultured in GM for 3 days in the absence of tetracycline, then transferred to DM for 3 days in the absence of tetracycline (middle); and myocytes at day 6 after addition of tetracycline to myocytes (from middle) cultured in DM at day 3 in the absence of tetracycline (right). Light photomicrographs were obtained with a Zeiss Axiovert 10 light microscope at 100× magnification. (C) Immunoblot analysis of MHC expression in C2MKP-1 myoblasts cultured as in (A) for the indicated times in the presence or absence of tetracycline. C2MKP-1 myoblasts cultured in DM in the absence of tetracycline were transferred on day 3 to DM containing tetracycline; MHC expression was assessed 3 days later.

These observations indicate that down-regulation of MKP-1 late during myogenesis is functionally required for the process of myotube formation. Thus, expression of the muscle-specific genes is necessary but not sufficient to commit terminally differentiated myocytes irrevocably to the formation of multinucleated myotubes. The uncoupling of muscle-specific gene activation from myotube formation suggests that a distinct pathway involving a member or members of the MAPK family regulates the formation of multinucleated myotubes.

  • * To whom correspondence should be addressed. E-mail: tonks{at}cshl.org

REFERENCES AND NOTES

View Abstract

Navigate This Article