DA-Raf and the MEK inhibitor trametinib reverse skeletal myocyte differentiation inhibition or muscle atrophy caused by myostatin and GDF11 through the non-Smad Ras–ERK pathway

Abstract

Myostatin (Mstn) and GDF11 are critical factors that are involved in muscle atrophy in the young and sarcopenia in the elderly, respectively. These TGF-β superfamily proteins activate not only Smad signalling but also non-Smad signalling including the Ras-mediated ERK pathway (Raf–MEK–ERK phosphorylation cascade). Although Mstn and GDF11 have been shown to induce muscle atrophy or sarcopenia by Smad2/3-mediated Akt inhibition, participation of the non-Smad Ras–ERK pathway in atrophy and sarcopenia has not been well determined. We show here that both Mstn and GDF11 prevented skeletal myocyte differentiation but that the MEK inhibitor U0126 or trametinib restored differentiation in Mstn- or GDF11-treated myocytes. These MEK inhibitors induced the expression of DA-Raf1 (DA-Raf), which is a dominant-negative antagonist of the Ras–ERK pathway. Exogenous expression of DA-Raf in Mstn- or GDF11-treated myocytes restored differentiation. Furthermore, administration of trametinib to aged mice resulted in an increase in myofiber size or recovery from muscle atrophy. The trametinib administration downregulated ERK activity in these muscles. These results imply that the Mstn/GDF11-induced Ras–ERK pathway plays critical roles in the inhibition of myocyte differentiation and muscle regeneration, which leads to muscle atrophy. Trametinib and similar approved drugs might be applicable to the treatment of muscle atrophy in sarcopenia or cachexia.

Graphical Abstract

Open in new tabDownload slide

Abbreviations

 
• DM

  differentiation medium

 
• DMEM

  Dulbecco’s-modified Eagle’s medium

 
• GM

  growth medium

 
• mAb

  monoclonal antibody

 
• Mstn

  myostatin

 
• MyHC

  myosin heavy chain

 
• pAb

  polyclonal antibody

 
• TA muscle

  tibialis anterior muscle

 
• TnT

  troponin T

 
• WGA

  wheat germ agglutinin

1. Introduction

Muscle atrophy, or a decrease in skeletal muscle mass, is associated with ageing (sarcopenia) and poor prognosis in several diseases including muscular dystrophies, myopathies and cancer (cachexia). Muscle atrophy occurs when protein degradation exceeds protein synthesis, and several mechanisms are involved in muscle atrophy (1–3). Among TGF-β superfamily proteins, myostatin (Mstn/GDF8), GDF11 and activin A physiologically or pathophysiologically play crucial roles in muscle atrophy (4–7). Mstn is expressed in developing skeletal myocytes to negatively regulate both proliferation and differentiation of muscle progenitors and myoblasts. It is also produced by adult skeletal muscle and acts in an autocrine manner to limit muscle growth and moreover to cause muscle atrophy in the young. GDF11 and activin A are expressed in a variety of cells and tissues during development and in adult animals. They also interfere with skeletal myocyte differentiation and muscle development and provoke muscle atrophy in cachexia or sarcopenia (8–11).

Mstn, GDF11 and activin A activate Smad2/3 signalling through their shared receptors ActRIIA/B–ActRI (ALK4/5). Smad2/3 activated by phosphorylation block Akt, which is activated through insulin-like growth factor 1 (IGF-1)-induced phosphatidylinositol 3-kinase (PI3K) signalling leading to protein synthesis (5, 9, 12, 13). Akt phosphorylates and inhibits the FoxO family of transcription factors by excluding them from the nucleus. FoxO induces the transcription of the E3 ubiquitin ligases MAFbx and MuRF1, thereby causing ubiquitin–proteasome-mediated protein degradation (5, 9, 13). In addition, FoxO promotes the expression of the autophagy regulators LC3, VSP34 and Bnip3, resulting in autophagy-mediated protein degradation (1–3). In consequence, Smad2/3 signalling activated by Mstn, GDF11 and activin A is considered to induce muscle atrophy by blocking Akt-mediated protein synthesis and indirectly by provoking protein degradation at least in the young. Nevertheless, autophagy, particularly macroautophagy, maintains muscle mass and function by removing dysfunctional mitochondria and protein aggregates in the elderly. Consequently, the decline in autophagy is associated with sarcopenia (3, 14).

TGF-β superfamily proteins activate not only canonical Smad signalling but also non-Smad signalling pathways, depending on the family proteins and receptors, their existing cell types and conditions (15, 16). The entity of ActRII–ActRI-mediated non-Smad signalling has not been elucidated well compared with Smad signalling. Mstn has been shown to activate non-Smad signalling such as the Ras–ERK pathway (17), p38 MAPK pathway (18–20) and JNK pathway (21). The activation of the Ras–ERK pathway leads to prevention of skeletal myocyte differentiation (17). However, the involvement of the Mstn-induced non-Smad Ras–ERK pathway in muscle atrophy has not been reported. Furthermore, although GDF11 also activates ERK and p38 MAPK (10), it remains to be clarified whether the GDF11-induced non-Smad signalling is involved in myocyte differentiation inhibition, muscle atrophy or sarcopenia.

The small GTPase Ras-induced ERK pathway (Raf–MEK–ERK phosphorylation cascade) regulates a variety of cellular and physiological functions, including cell proliferation, differentiation, apoptosis, motility and metabolism (22). When the pathway is constitutively activated by mutants of Ras or Raf, it can lead to tumorigenesis. Moreover, germline mutations in Ras, Raf and MEK, as well as in several regulatory proteins of this pathway, cause genetic syndromes, referred to as RASopathies, which are characterized by developmental morphogenetic disorders and cancer (23–25). The activity of the pathway is strictly controlled through the coordinated action of both positive and negative regulators. A variety of negative regulators of the pathway have been identified, including Sprouty, Spred, RKIP, DiRas3 and DA-Raf1 (hereafter referred to as DA-Raf) (26).

DA-Raf is generated by alternative splicing of Araf pre-mRNA. DA-Raf consists only of the N-terminal portion of A-Raf containing the Ras-binding domain (RBD) and cysteine-rich domain (CRD) but lacks the kinase domain that phosphorylates to activate MEK. According to this structure, DA-Raf binds to active Ras and interferes with the ERK pathway in a dominant-negative fashion by competing with Raf proteins for binding to Ras (26, 27). The Ras–ERK pathway negatively regulates skeletal myocyte differentiation by preventing the expression or functions of the muscle-specific MyoD and MEF2 families of transcription factors (28–31). Endogenous DA-Raf expression is prominently induced, and MEK–ERK activity is noticeably reduced during differentiation of mouse C2C12 skeletal myocytes. Overexpression of DA-Raf facilitates myocyte differentiation, whereas knockdown of DA-Raf by RNAi suppresses differentiation (27, 32). Therefore, DA-Raf serves as a master inducer of skeletal myocyte differentiation by blocking the Ras–ERK pathway.

To elucidate whether the non-Smad Ras–ERK pathway participates in muscle atrophy and sarcopenia caused by Mstn or GDF11, we treated C2C12 cells with Mstn or GDF11 together with the MEK inhibitors U0126 and trametinib, which is a clinically approved anticancer drug. Either of the inhibitors induced DA-Raf expression and restored differentiation that was inhibited by Mstn or GDF11. Exogenous expression of DA-Raf also retrieved differentiation. Furthermore, administration of trametinib to aged mice resulted in an increase in myofiber size, or recovery from muscle atrophy. These results imply that the Ras–ERK pathway is critical for Mstn/GDF11-induced muscle atrophy and that trametinib or similar drugs might be applicable to the treatment of muscle atrophy in sarcopenia or cachexia.

2. Materials and Methods

2.1. Cell culture

Mouse skeletal muscle myoblast C2C12 cells and human 293T cell line-based Plat-E retroviral packaging cells (Cell Biolabs) were maintained in Dulbecco’s-modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (growth medium, GM). To induce differentiation in C2C12 cells, the GM was replaced with DMEM containing 5% horse serum (differentiation medium, DM). The MEK inhibitors U0126 (Promega) and trametinib DMSO solvate (ChemScene) were dissolved in DMSO and added to the culture media at concentrations of 3 μM and 10–50 nM, respectively. Recombinant human/mouse/rat Mstn and GDF11 (PeproTech) were dissolved in 4 mM HCl and 0.1% BSA at 500 μg/ml and added to the culture media at concentrations of 500 ng/ml.

2.2. Plasmid construction

To construct pMXs-Neo/EGFP, a blunt-ended NheI–PstI fragment including EGFP of pEGFP-C2 vector (Clontech) was inserted into the blunt-ended EcoRI site of pMXs-Neo retroviral vector (Cell Biolabs). pEGFP/DAraf was constructed by inserting a mouse DAraf cDNA fragment into pEGFP-C2 vector in frame with the EGFP tag. To construct pMXs-Neo/EGFP-DAraf, a blunt-ended NheI–PstI fragment including EGFP-DAraf of pEGFP/DAraf was inserted into the blunt-ended EcoRI site of pMXs-Neo vector.

2.3. Retroviral infection

Plat-E cells cultured in the antibiotics-free GM on a 100-mm dish were transfected with 19 μg of pMXs-Neo/EGFP or pMXs-Neo/EGFP-DAraf mixed with PEI MAX 40K (Polysciences) and Opti-MEM-I Reduced Serum Medium (Thermo Fisher Scientific) and cultured for 48 h. Retroviruses in the culture medium were collected and diluted with the same volume of the antibiotics-free GM, and polybrene was added to the final concentration of 8 μg/ml. C2C12 cells at ~30% confluent were cultured in this retrovirus solution for 6 h, and the medium was replaced with the normal GM after the infection.

2.4. Administration of trametinib to aged mice and muscle tissue isolation

Trametinib DMSO solvate was dissolved in 0.5% hydroxypropyl methylcellulose and 0.1% Tween 80 at a concentration of 3 mg/ml. To 92-week-old (21-month-old) male C57BL/6J-Aged mice (JAX mice strain) (Charles River Laboratories Japan), trametinib (1 μg/g body weight) or solvent was administered daily for 21 days by intraperitoneal injections. One day after the final administration, mice were euthanized by cervical dislocation. The tibialis anterior (TA), gastrocnemius and soleus muscles were isolated and frozen in isopentane precooled with liquid nitrogen for cryosectioning. Frozen tissues were stored at −85°C. All mouse care and experiments were approved by the animal care and use committee of Chiba University and carried out in accordance with the guidelines.

2.5. Fluorescence microscopy

C2C12 cells were washed with PBS, fixed with 4% paraformaldehyde in PBS for 15 min, and permeabilized with 0.2% Triton X-100 in PBS for 5 min. They were incubated with the myosin heavy chain (MyHC) mouse monoclonal antibody (mAb) A4.1025 (Developmental Studies Hybridoma Bank, DSHB) for 60 min, with Alexa-Fluor 488 (or 555) goat anti-mouse IgG (Thermo Fisher Scientific) for 60 min, and then with 1 mg/ml Hoechst 33258 for 1 min.

Transverse cryosections (5 μm-thick) of the frozen skeletal muscles were prepared with a Leica CM1850 Cryomicrotome and fixed with 4% paraformaldehyde in PBS for 15 min. They were incubated with Alexa-Fluor 555 wheat germ agglutinin (WGA) (Thermo Fisher Scientific) for 10 min.

The specimens were observed with a fluorescence microscope Axio Imager.A2 (Carl Zeiss) equipped with a CoolSNAP HQ2 CCD camera (Photometrics) operated with Micro-Manager 1.4 open source microscopy software. The fusion index was calculated as the percentage of nuclei within MyHC-positive myotubes containing ≥3 nuclei.

2.6. Immunoblotting

Cells were washed with ice-cold PBS and lysed with the lysis buffer containing 1% SDS (33). The lysates were cleared by centrifugation and treated with the SDS sample buffer. The frozen skeletal muscles for cryosectioning were treated with the SDS sample buffer and cleared by centrifugation. The samples were subjected to SDS-PAGE, transferred to PVDF membranes and analysed by immunoblotting as described previously (33). Primary antibodies used were DA-Raf rabbit polyclonal antibody (pAb) (27), ERK1/2 rabbit pAb and phospho-ERK1/2 (T202/Y204) rabbit mAb D13.14.4E (Cell Signaling Technology), MyHC mouse mAb A4.1025, troponin T (TnT) mouse mAb NT302 (34) and β-tubulin mouse mAb E7 (DSHB). The blotting bands were detected with Western Lightning Plus-ECL (PerkinElmer) by using a ChemiDoc XRS Plus System (Bio-Rad). The band intensity was densitometrically analysed with Image Lab Software Version 4.1 (Bio-Rad).

2.7. Statistical analysis

Statistical analysis was conducted using unpaired, one-tailed Student’s t-test or one-way ANOVA followed by Tukey’s post hoc test to calculate differences between treatment groups. Data were analysed using Prism GraphPad Software Version 7.0b. Values in graphs are presented as means ± SEM of three independent experiments.

3. Results

3.1. Mstn and GDF11 activate the non-Smad ERK pathway

Both Mstn and GDF11 have been shown to activate the non-Smad ERK pathway (10, 17). Accordingly, we first analysed the time-course activation of ERK by Mstn and GDF11 in mouse C2C12 myoblasts under differentiation conditions. C2C12 myoblasts at ~30% confluence were shifted to the DM containing 500 ng/ml Mstn or GDF11 and cultured for 1–72 h. The levels of activated ERK1/2, or phospho-ERK1/2, were analysed by immunoblotting (Fig. 1A and B). In the control culture, the levels of phospho-ERK1/2 gradually declined to ~25% by 24 h in the DM. In contrast, the addition of Mstn elevated the phospho-ERK1/2 levels >2.5-fold ~2 h after the addition. The amounts were maintained at high levels for >12 h but reduced to ~80% by 24 h. The addition of GDF11 also raised the phospho-ERK1/2 levels >1.5-fold 1–2 h later. The quantities were maintained at high levels for >6 h but decreased to ~50% by 24 h. Collectively, both Mstn and GDF11 activate the ERK pathway for a relatively long time, but the effect is lost by 24 h in C2C12 myoblasts in the DM.

3.2. Inhibition of myocyte differentiation by Mstn or GDF11 is reversed by MEK inhibition

Mstn has been shown to inhibit skeletal myocyte differentiation by suppressing the expression and activity of the MyoD family of transcription factors MyoD and myogenin (9, 17, 35). GDF11 also prevents the differentiation of primary human myoblasts (9, 10). The differentiation inhibition function of Mstn and GDF11 is likely to be at least in part responsible for their ability to provoke muscle atrophy or sarcopenia (4–7). However, participation of the non-Smad Ras–ERK pathway in myocyte differentiation inhibition has not been clearly demonstrated. Thus, we analysed the effects of MEK inhibitors on Mstn- or GDF11-treated mouse C2C12 myoblasts under the differentiation conditions.

C2C12 myoblasts cultured in the DM for 72 h differentiated to form multinucleated myotubes expressing muscle-specific MyHC (Fig. 2A). The addition of the MEK inhibitor U0126 facilitated differentiation and led to a more extensive formation of myotubes containing more nuclei, as indicated by the fusion index (Fig. 2A and B). Treatment of the cells with Mstn or GDF11 almost completely inhibited differentiation, as indicated by both MyHC expression and myoblast fusion in either case (Fig. 2A and B). However, the addition of U0126 to Mstn- or GDF11-treated cells retrieved myotube formation, although the myotubes formed were smaller than those in the control culture (Fig. 2A and B).

We then examined the effect of Mstn or GDF11 treatment as well as that of U0126 addition on ERK activity through phosphorylation and the expression level of DA-Raf in C2C12 cells cultured in the DM for 72 h by immunoblotting (Fig. 2C). The addition of U0126 notably reduced the level of phospho-ERK and elevated those of DA-Raf. The amount of the MyoD family of transcription factor myogenin was reduced to some degree. This is probably because its expression reaches a maximal level by 48 h in C2C12 cells in the DM (32) and because its expression level is declined in mature myofibers (36). On the other hand, the amounts of troponin T (TnT) and MyHC were increased to some extent. Treatment of the cells with Mstn or GDF11 reduced the amount of DA-Raf and those of MyHC and TnT. In contrast, the addition of U0126 to Mstn- or GDF11-treated cells also markedly diminished the levels of phospho-ERK1/2 and increased those of DA-Raf. In addition, the amounts of MyHC and TnT were raised to the levels comparable with those in the control culture. These results of muscle-specific protein levels in immunoblotting are consistent with the above microscopic data.

3.3. Inhibition of myocyte differentiation by Mstn or GDF11 is reversed with trametinib

Several MEK inhibitors have been approved for the treatment of BRAF- or RAS-mutant cancers (37, 38). Thus, we next examined whether one of the approved MEK inhibitors, trametinib (GSK112021/JTP-74057) (39–41), was also effective in restoring myocyte differentiation inhibited by Mstn or GDF11. When various concentrations of trametinib were added to C2C12 cells in the DM, 10 nM trametinib facilitated MyHC-expressing myotube formation, compared with the control culture (Fig. 3A and B). In contrast, 25 or 50 nM trametinib did not promote myotube formation but instead formed smaller myotubes or MyHC-expressing mononucleated myocytes, probably due to overdose toxicity (Fig. 3A and B). Accordingly, we added 10 nM trametinib to Mstn- or GDF11-treated cells, in which differentiation was completely blocked. It restored MyHC-expressing myotube formation to a level almost comparable with that in the control culture (Fig. 4A and B).

The addition of 10 nM trametinib to C2C12 cells in the DM prominently decreased the levels of phospho-ERK1/2 and increased that of DA-Raf (Fig. 4C). The amount of myogenin was declined in the same way that U0126 was added to the cells, whereas those of TnT and MyHC were obviously increased. Furthermore, the addition of trametinib to Mstn- or GDF11-treated cells also remarkably diminished the levels of phospho-ERK1/2 and elevated that of DA-Raf. Moreover, the amounts of MyHC and TnT were elevated to levels comparable with those in the control culture. The results of muscle-specific protein levels in immunoblotting are also consistent with the above microscopic data.

Taken together, these results indicate that the non-Smad Ras–ERK pathway induced by Mstn or GDF11 negatively regulates skeletal myocyte differentiation in both muscle-specific gene expression and myoblast fusion. Thus, the Ras–ERK pathway is, at least in part, responsible for the differentiation-inhibiting function of Mstn and GDF11, which leads to muscle atrophy or sarcopenia. Consequently, suppression of the Ras–ERK pathway by these MEK inhibitors efficiently restores differentiation.

3.4. Inhibition of myocyte differentiation by Mstn or GDF11 is reversed by DA-Raf-mediated inhibition of the ERK pathway

Because DA-Raf expression was induced by the MEK inhibitors U0126 and trametinib, and because myocyte differentiation inhibition by Mstn or GDF11 was reversed by both these inhibitors, we further investigated whether DA-Raf participated in the recovery of differentiation. EGFP-tagged DA-Raf was expressed in C2C12 cells by the retroviral expression system, and the cells were cultured in the DM for 72 h. Myotube formation was promoted in the DA-Raf-expressing cells in comparison with the control EGFP-expressing cells (Fig. 5A and C). Treatment with Mstn or GDF11 almost completely inhibited differentiation of EGFP-expressing cells in both MyHC expression and myoblast fusion. In contrast, DA-Raf-expressing cells with Mstn- or GDF11-treatment formed myotubes, although these myotubes were smaller than those in the control culture (Fig. 5A and C).

The viability of the retrovirally infected cells cultured in the DM was reduced, and formed myotubes were smaller or shorter than those without infection (Fig. 5A vs. Fig. 2A). Thus, we cultured C2C12 cells in the GM for 72 h to reduce the influence of retroviral infection. Control EGFP-expressing cells formed myotubes by 72 h even in the GM, and EGFP-tagged DA-Raf-expressing cells formed larger myotubes (Fig. 5B and D). Treatment with Mstn or GDF11 almost completely inhibited differentiation of EGFP-expressing cells. In contrast, DA-Raf-expressing cells with Mstn- or GDF11-treatment formed larger myotubes than those in the DM (Fig. 5B and D).

The exogenous expression of EGFP–DA-Raf in C2C12 cells cultured in the DM or GM for 72 h almost did not affect the endogenous DA-Raf level (Fig. 5E and F). In contrast, the exogenous DA-Raf expression reduced the phospho-ERK1/2 levels in the control culture in both the DM and GM. In the cells treated with Mstn or GDF11, the exogenous DA-Raf expression also diminished the phospho-ERK1/2 levels in both the DM and GM (Fig. 5E and F). These results imply that overexpression of DA-Raf suppresses both basal ERK activity and Mstn/DGF11-induced non-Smad ERK activity, thereby facilitating and recovering myocyte differentiation.

3.5. Trametinib ameliorates muscle atrophy in aged mice by suppressing the ERK pathway

Since suppression of the Ras–ERK pathway by trametinib efficiently restored myocyte differentiation that was inhibited by Mstn or GDF11, we further examined whether administration of trametinib to aged mice retrieved myofiber mass from sarcopenia. Myofiber atrophy is detected in 21-month-old mice (42). In a mouse xenograft model, 2.5–5 μg/g body weight trametinib can almost completely inhibit tumour growth (43; ChemScene data sheet). Thus, we administered 1 μg/g body weight trametinib to 92-week-old (21-month-old) male C57BL/6 mice daily for 21 days by intraperitoneal injections. This dose did not affect the health or viability of the aged mice at least for 21 days, compared with control mice. Three types of skeletal muscles, or the TA (predominantly fast-twitch), soleus (predominantly slow-twitch) and gastrocnemius (mixture of fast-twitch and slow-twitch) muscles, were isolated, and their transverse cryosections stained with Alexa Fluor-conjugated WGA were analysed to determine the cross-sectional area of myofibers. Administration of trametinib noticeably increased the cross-sectional area in all three muscle types (Fig. 6A and B). The distribution of cross-sectional areas in the histogram clearly showed that trametinib administration prominently increased the numbers of thicker myofibers in the three muscle types, particularly in the TA muscle (Fig. 6C). Therefore, trametinib administration is effective in ameliorating muscle atrophy in sarcopenia more pronouncedly in fast muscle fibres than in slow muscle fibres in aged mice.

We next analysed whether the trametinib administration increased the cross-sectional areas by inhibiting MEK constituting the Ras–ERK pathway. Administration of trametinib for 21 days significantly decreased the phospho-ERK1/2 levels in the TA and gastrocnemius muscles as detected by immunoblotting (Fig. 7A and B). The phospho-ERK1/2 levels also tended to be decreased in the soleus muscle. Thus, the trametinib-induced amelioration of muscle atrophy is likely to be ascribable to the inhibition of the non-Smad Ras–ERK pathway that is activated by TGF-β superfamily proteins. We further tried to assess whether the trametinib administration induced DA-Raf expression to block the Ras–ERK pathway. Although the trametinib administration tended to upregulate DA-Raf expression in some muscle samples, we could not obtain conclusive results that trametinib invariably induced DA-Raf expression to ameliorate muscle atrophy.

4. Discussion

Mstn, GDF11 and activin A pathophysiologically play critical roles in muscle atrophy in the young, cachexia or sarcopenia through their shared receptors ActRII–ActRI (4–7). Smad2/3 signalling activated by them has been demonstrated to induce muscle atrophy by blocking Akt-mediated protein synthesis and by indirectly provoking protein degradation mediated by the ubiquitin–proteasome system or autophagy at least in the young (5, 9, 12, 13). However, autophagy maintains muscle mass and function by removing dysfunctional mitochondria and protein aggregates in the elderly (3, 14). In contrast, the participation of ActRII–ActRI-mediated non-Smad signalling in muscle atrophy or sarcopenia has not been fully elucidated. We have shown here that both Mstn and GDF11 inhibit C2C12 cell differentiation through the non-Smad Ras–ERK pathway by suppressing DA-Raf expression. Moreover, we have shown that suppression of the Ras–ERK pathway retrieves Mstn- and GDF11-inhibited myocyte differentiation and induces muscle hypertrophy in aged mice (Fig. 8).

Both Mstn and GDF11 almost completely inhibited C2C12 cell differentiation represented by both muscle-specific protein expression and myoblast fusion. Muscle-specific gene expression is elicited by the MyoD and MEF2 families of transcription factors (44–48) and by MRTF–SRF transcription factors (49, 50). In addition, a variety of proteins are responsible for myoblast fusion (51, 52). Among them, the muscle-specific plasma membrane fusion proteins myomaker and myomixer, which directly participate in membrane fusion, are transcriptionally regulated by MyoD and myogenin (53, 54). Indeed, Mstn and GDF11 decreased the level of myogenin in the present study. This is consistent with the reports that Mstn inhibits C2C12 cell differentiation by suppressing the expression and activity of MyoD and myogenin (9, 17, 35). It has been shown that Mstn-activated Smad3 binds to MyoD and suppresses its expression, implying that inhibition of myocyte differentiation by Mstn is, at least in part, mediated by Smad3 (35). The results presented here indicate, however, that the MEK inhibitors restored myogenin expression and myocyte differentiation to a certain degree. These findings are supported by another report (17). Taken together, Mstn-induced myocyte differentiation inhibition is mediated by both Smad signalling and the non-Smad Ras–ERK pathway.

Although GDF11 shares the receptor with Mstn, the signalling pathway that inhibits myocyte differentiation needed to be elucidated. GDF11 has been shown to activate both ERK and p38 MAPK in addition to Smad2/3 (10). We have shown here that the MEK inhibitors also restored myocyte differentiation that was inhibited by GDF11. Therefore, the non-Smad Ras–ERK pathway is also, at least in part, responsible for the differentiation-inhibiting function of GDF11. In contrast, since p38 MAPK positively regulates the induction of myocyte differentiation in various aspects (55), the significance of the activation of p38 MAPK by GDF11 remains to be clarified.

Forced expression of DA-Raf in Mstn- or GDF11-treated C2C12 cells retrieved myocyte differentiation. DA-Raf binds to active Ras and interferes with the ERK pathway in a dominant-negative fashion not only in vitro but also in vivo (26, 27, 32, 33, 56, 57). With regard to myocyte differentiation, DA-Raf induces myocyte differentiation in both muscle-specific protein expression and myoblast fusion by blocking the Ras–ERK pathway from impairing the MyoD family of transcription factors (27, 32). The MEK inhibitors elevated the expression of DA-Raf, whereas the Mstn- or GDF11-induced Ras–ERK pathway reduced its expression. Therefore, DA-Raf and the Ras–ERK pathway function in a mutually suppressive manner. DA-Raf is generated by alternative splicing of Araf pre-mRNA through intron 6 retention (26, 27). Another A-Raf splicing variant that lacks the kinase domain, or A-Raf_short, can bind to active Ras and dominant-negatively interferes with the ERK pathway, as does DA-Raf (58). c-Myc, which is stabilized by Ras–ERK pathway-mediated phosphorylation (59), transcribes the splicing factors hnRNP H and A2 (58, 60). hnRNP H or A2 participates in the splicing of full-length A-Raf and consequently reduces the production of A-Raf_short (58, 61). It remains to be determined whether the Ras–ERK pathway abrogates DA-Raf expression through hnRNP H- or A2-mediated A-Raf splicing.

Mstn, GDF11 and activin A play crucial roles in muscle atrophy (4–7). However, several studies have reported that the muscle Mstn level decreases with age in rats (10) and that the serum or plasma Mstn level does not change or even declines in elderly people (62–64). Thus, Mstn is not likely to be involved in the induction of ageing-related sarcopenia. In contrast, the serum or plasma GDF11 level increases or does not change during ageing, and GDF11 correlates with sarcopenia and cachexia (7, 10, 11, 64, 65). Indeed, several studies have revealed that GDF11 inhibits myocyte differentiation and muscle regeneration (10, 11, 66). We have shown here that administration of trametinib to aged mice retrieved myofiber mass from sarcopenia. The trametinib administration downregulated ERK activity in the muscles. These results imply that the non-Smad Ras–ERK pathway induced by GDF11, at least in part, is implicated in muscle atrophy in aged mice possibly through the inhibition of myocyte differentiation and muscle regeneration.

Muscle atrophy in the elderly is more pronounced in Type 2 fast myofibers than in Type 1 slow myofibers (3, 67). In addition, the transition from MyHC Type 2B (fast-twitch glycolytic) to 2X (intermediate shortening velocity between 2B and 2A) fibres in fast muscles and that from Type 2A (fast-twitch oxidative glycolytic) to Type 1 (slow oxidative) fibres in slow soleus muscle occur in rats during ageing (67). We have further shown that the recovery of myofiber mass was more prominent in the TA muscle (predominantly type 2 fibres) compared with the soleus muscle (predominantly type 1 fibres) and the gastrocnemius muscle (mixture of Type 2 and Type 1 fibres). Taken together, these results imply that atrophied myofibers in aged mice may be restored to the healthy adult status by the trametinib administration. Therefore, the trametinib-induced recovery of myofiber mass is likely to be more prominent in the TA muscle than in the soleus and gastrocnemius muscles. In this context, it should be noted that the Ras–ERK pathway promotes nerve-activity-dependent differentiation of Type 1 fibres in the rat soleus muscle (68). These findings are consistent with our results that the trametinib-induced recovery of myofiber mass is less conspicuous in the soleus muscle than in the TA muscle.

Both quantitative and qualitative changes occur in satellite cells, or skeletal muscle-residing muscle stem cells, during ageing (3). The number of satellite cells is reduced, and intrinsic defects in satellite cell function, including proliferation, arise in aged skeletal muscle. Consequently, muscle regeneration capacity is impaired in aged muscle. However, induced depletion of satellite cells neither accelerates nor exacerbates sarcopenia. Satellite cells do not contribute to the maintenance of muscle mass or myofiber type composition during ageing, but their loss contributes to muscle fibrosis in aged muscle (69). Accordingly, the trametinib-induced recovery of myofiber mass may not be attributable to promotion of satellite cell number or function, even if trametinib induces such changes.

Trametinib is approved for use singly or in combination with other drugs for the treatment of BRAF- or RAS-mutant cancers (37, 38). Therefore, trametinib might be therapeutically applicable to muscle atrophy in sarcopenia and cachexia. Furthermore, we need to determine whether other MEK inhibitors or ERK pathway inhibitors approved for the treatment of cancers or other diseases are also effective for the treatment of sarcopenia and cachexia.

Acknowledgments

We thank Aoi Matsuda for helping plasmid construction.

Funding

This work was supported by the Japan Society for the Promotion of Science KAKENHI (JP15H04348); the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT) KAKENHI (JP25117706); the Intramural Research Grant (26-8 and 29-4) for Neurological and Psychiatric Disorders of National Center of Neurology and Psychiatry; The Mitsubishi Foundation; The Uehara Memorial Foundation; and the Takeda Science Foundation.

Conflict of Interest

T. Sakai obtains a patent fee on trametinib from JT Pharmaceutical Institute.

References