The homeobox transcription factor Barx2 regulates plasticity of young primary myofibers

Abstract

Background: Adult mammalian muscle retains incredible plasticity. Muscle growth and repair involves the activation of undifferentiated myogenic precursors called satellite cells. In some circumstances, it has been proposed that existing myofibers may also cleave and produce a pool of proliferative cells that can re-differentiate into new fibers. Such myofiber dedifferentiation has been observed in the salamander blastema where it may occur in parallel with satellite cell activation. Moreover, ectopic expression of the homeodomain transcription factor Msx1 in differentiated C2C12 myotubes has been shown to induce their dedifferentiation. While it remains unclear whether dedifferentiation and redifferentiaton occurs endogenously in mammalian muscle, there is considerable interest in induced dedifferentiation as a possible regenerative tool.

Methodology/principal findings: We previously showed that the homeobox protein Barx2 promotes myoblast differentiation. Here we report that ectopic expression of Barx2 in young immature myotubes derived from cell lines and primary mouse myoblasts, caused cleavage of the syncytium and downregulation of differentiation markers. Microinjection of Barx2 cDNA into immature myotubes derived from primary cells led to cleavage and formation of mononucleated cells that were able to proliferate. However, injection of Barx2 cDNA into mature myotubes did not cause cleavage. Barx2 expression in C2C12 myotubes increased the expression of cyclin D1, which may promote cell cycle re-entry. We also observed differential muscle gene regulation by Barx2 at early and late stages of muscle differentiation which may be due to differential recruitment of transcriptional activator or repressor complexes to muscle specific genes by Barx2.

Conclusions/significance: We show that Barx2 regulates plasticity of immature myofibers and might act as a molecular switch controlling cell differentiation and proliferation.

Figure 1. Ectopic expression of Barx2 in differentiated myotubes causes their dedifferentiation.

A. Serum-deprived C2C12 cultures contaning many early myotubes were transfected with either Barx2/pcDNA, empty pcDNA3, or GFP-expressing plasmids. Images of maturing or apparently fragmenting myotubes were taken after 3–4 days, B. Levels of the differentiation-associated proteins myogenin and myosin heavy chain (MyHC) were measured by immunoblotting of total protein from Barx2-expressing and control cultures corresponding to those shown in A. Untransfected myoblast and myotube cultures were also analysed as a reference. C. C3H10T1/2 mesenchymal progenitor cells were transfected with either MyoD/pcDNA3 or a combination of MyoD/pcDNA3 and Barx2/pcDNA3 plasmids and differentiation was subsequently induced by serum withdrawal for 2–3 days. The proportion of cells expressing myogenin and MyHC and appearance of myotubes was assessed by immunostaining. D. MyHC protein was assessed by immunoblotting of total protein from cultures corresponding to those shown in C and D. Analysis of BrdU incorporation was performed in C3H10T1/2 cultures corresponding to those shown in C; co-expression of Barx2 with MyoD increased numbers of proliferating cells relative to MyoD alone - E.

Figure 2. Differentiating myoblasts and maturating myotubes have different levels of Barx2 expression.

A. Cultured primary myoblasts were induced to differentiate by serum withdrawal and changes in cell shape, actin remodeling and myogenin expression were monitored over the first 6 hours. B–D. Myotube maturation was examined between 24 and 72 hours post serum-withdrawal. At 24 hours myotubes were thin and appeared immature (B). Between 48 and 72 hours, thicker myotubes appeared often with local aggregations of nuclei (C and D) and frequent strong contractions. E. Observations of myotube maturation in culture presented schematically. F. Barx2 expression was measured by RT-PCR at different stages of differentiation. Scale bars represent (A–C) - 20 µm, D–50 µm.

Figure 3. Barx2 induces cleavage of mouse myotubes.

A–D. Barx2/pcDNA3 or empty pcDNA3 plasmids were mixed with Alexa Fluor®488-conjugated dextran and then injected into immature (stage 1) myotubes. Cultures were stimulated with growth medium and fluorescently labeled myotubes were monitored for signs of cleavage and appearance of labeled mononucleated cells. Scale bar represent 10 µm. E. BrdU labeling after microinjection of myotubes. Single Alexa Fluor®488-conjugated dextran-labeled cells occasionally incorporated BrdU suggesting they has re-entered the cell cycle. Scale bar represent 10 µm. F–H. Mature (stage 2) myotubes were microinjected and examined as in A-D. No myotube cleavage was observed. Scale bars represent F–10 µm, G - 50 µm, H–20 µm.

Figure 4. Regulatory connections of Barx2 during muscle differentiation.

A. Quantitative RT-PCR analysis of cyclin D1 expression in C2C12 cells stably transfected with a Barx2-expression plasmid or control plasmid. B. The SMA promoter-luciferase construct was co-transfected with combinations of Barx2, MyoD, or empty pcDNA3 expression plasmids in C2C12 cells. Barx2 increased SMA promoter activation in undifferentiated cell cultures yet inhibited activation in myotubes. C. Co-immunoprecipitation was performed after expression of both Barx2 and Hes6 proteins in COS1 cells and C2C12 cells. Barx2 co-immunoprecipitated Hes6 in both contexts. D. Expression of Hes6 was compared in C2C12 myoblasts and myotubes using RT-PCR. Hes6 is upregulated in myotubes. E. Graphical representation of the RT PCR data shown in D (n – number of experiments).