SEARCH

SEARCH BY CITATION

Keywords:

  • Muscle development;
  • Satellite cells;
  • Skeletal muscle;
  • MyoD1 myogenic differentiation protein;
  • CCAAT-enhancer binding protein beta

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

Upon injury, muscle satellite cells become activated and produce skeletal muscle precursors that engage in myogenesis. We demonstrate that the transcription factor CCAAT/enhancer binding protein beta (C/EBPβ) is expressed in the satellite cells of healthy muscle. C/EBPβ expression is regulated during myogenesis such that C/EBPβ is rapidly and massively downregulated upon induction to differentiate. Furthermore, persistent expression of C/EBPβ in myoblasts potently inhibits differentiation at least in part through the inhibition of MyoD protein function and stability. As a consequence, myogenic factor expression, myosin heavy chain expression, and fusogenic activity were reduced in C/EBPβ-overexpressing cells. Using knockout models, we demonstrate that loss of Cebpb expression in satellite cells results in precocious differentiation of myoblasts in growth conditions and greater cell fusion upon differentiation. In vivo, loss of Cebpb expression in satellite cells resulted in larger muscle fiber cross-sectional area and improved repair after muscle injury. Our results support the notion that C/EBPβ inhibits myogenic differentiation and that its levels must be reduced to allow for activation of MyoD target genes and the progression of differentiation. STEM CELLS 2012;30:2619–2630


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

Muscle satellite cells (SCs) are thought to be the primary source of regenerative capacity for skeletal muscle and are found between the sarcolemma and the basement membrane of mature muscle fibers [1]. These cells can be activated to both proliferate and differentiate in response to external stimuli, most importantly muscle injury and exercise, and are defined by their protein marker expression, namely CD34+, α7β1 integrin+, Pax3+, and Pax7+ [2–4]. As SCs become activated, they progressively lose expression of Pax7 and express in a coordinated fashion the myogenic basic helix-loop-helix factors MyoD, myogenin, and MRF4 that are responsible for the induction of myocyte-specific genes [5].

Significant functional redundancy exists among the myogenic factors. While loss of either Myod1 or Myf5 is without effect on muscle development, deletion of both genes results in a complete lack of skeletal muscle and myoblasts, suggesting that MyoD and Myf5 can functionally compensate for one another [6, 7]. However, MyoD does play an important role in the maintenance of muscle mass in the adult as following injury, Myod1−/− muscle is unable to repair [5, 8]. Indeed Myod1−/− mice crossed into a dystrophin-deficient background die prematurely due to the exacerbation of muscle wasting [8]. When Myod1−/− myoblast gene expression profiles were determined and compared to that of wild-type (WT) controls, it was noted that Myod1−/− cells, while expressing higher levels of Myf5, were deficient in other myogenic markers such as desmin and MRF4 and expressed the stem cell markers Sca-1 and CD34 [7, 9]. Despite low expression levels of myogenin and myosin heavy chain (MyHC) in Myod1−/− myoblasts, these cells failed to differentiate efficiently and displayed a dramatic reduction in fusion [10, 11].

Notwithstanding its important role in myogenesis, little is known about the factors that regulate Myod1 expression. Since inhibition of MyoD expression and/or function can be linked to the inhibition of adult myogenesis, regulatory pathways that control MyoD expression are of great interest. Nuclear Factor-kB (NF-kB) can regulate Myod1 mRNA expression through increased turnover [12, 13], while the E3 ubiquitin ligase MAFbx/atrogin-1 has also been associated with increased degradation of MyoD protein and the development of muscle wasting [14, 15]. Overexpression of paired domain homeobox transcription factor Pax7 has also been shown to reduce MyoD protein levels and to block MyoD function independent of its mRNA expression, resulting in decreased myogenin expression and inhibited myogenesis [16].

Appearing first in the dermomyotome, Pax7 expression becomes localized to mononucleated cells of the trunk and limb muscles between E12.5 and E16.5 [17]. While Pax7−/− animals have normal fetal myogenesis, their postnatal myogenesis is severely compromised [18, 19]. As such, Pax7 is thought to participate in the maintenance of the SC undifferentiated state by blocking differentiation and promoting self-renewal [5, 20]. Conditional depletion of Pax7-expressing cells abrogates muscle regeneration following injury [21], although loss of Pax7 expression after postnatal day 21 is without effect on muscle repair, suggesting that Pax7 is dispensable in older animals [22].

C/EBPs form a family of bzip transcription factors of which CCAAT/enhancer binding protein beta (C/EBPβ) is involved in many regulatory and differentiation processes as both an activator and a repressor. For example, it is required for liver regeneration, acts as a potent commitment factor for adipocyte differentiation, and regulates the acute phase response of the immune system [23–27]. While C/EBPβ is expressed in muscle, it does not appear to be required for embryonic myogenesis as loss of Cebpb in mice results in no overt defects in muscle histology, although increased muscle insulin sensitivity was observed [28]. C/EBPβ-expressing infiltrating macrophages have been shown to be necessary for efficient repair of an acute muscle injury, while loss of Cebpb expression in the muscle fiber itself was without impact on repair [29]. C/EBPβ expression in myonuclei has been associated with the upregulation of atrogin-1 expression in the muscle fiber in cancer cachexia [30].

Our own work has demonstrated that C/EBPβ is a major regulator of mesenchymal stem cell fate where it acts as an activator of adipogenesis and a repressor of osteoblastogenesis [31–33]. Since C/EBPβ's role in muscle stem cell function remains unknown, we sought to determine if C/EBPβ participated in adult myogenesis.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

Constructs

The C/EBPβ expression vector and retroviral expression vector have been described previously [33]. The Pax7-luc reporter construct was kindly provided by Dr. L. Shen [57]. The −2 kb myogenin-luc reporter construct was a gift from Dr. Alexandre Blais [58].

Retroviral Infection and Cellular Differentiation

Replication incompetent pLXSN-based retroviruses (Clontech, Palo Alto, CA, http://www.clontech.com) were generated in Phoenix Ampho packaging cells (ATCC, Manassas, VA, http://www.atcc.org) as described [33]. Following infection, cells were selected in media containing 400 μg ml−1 G418 for 7 days prior to differentiation to ensure expression in all cells. To stimulate skeletal muscle differentiation, 70% confluent C2C12 cells were treated with Dulbecco's modified Eagle's medium (DMEM) containing 1% horse serum (HS). For Giemsa staining, cells were washed with phosphate-buffered saline (PBS), fixed with ice-cold methanol for 15 minutes, and stained with 10% Giemsa for 1 hour. Photomicrographs are representative of a minimum of three independent experiments.

Isolation and Differentiation of Skeletal Muscle Precursor Cells

Skeletal muscle precursor cells were isolated essentially as described [8]. Lower hind limb muscles from C57BL/6 female mice aged 6–8 weeks (Charles River laboratories, Wilmington, MA, http://www.criver.com) were dissected and digested with dispase and collagenase (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com). Isolated cells were plated on Matrigel-coated dishes and allowed to grow in DMEM containing 20% FBS, 10% HS (Invitrogen, Carlsbad, CA, http://www.invitrogen.com), with penicillin and streptomycin (Wisent) in the presence of 10 ng/ml basic FGF and 2 ng/ml HGF (Peprotech, Rocky Hill, NJ, http://www.peprotech.com) as indicated in figure legends. Differentiation was achieved by changing the medium of 70% confluent myoblasts cultures to DMEM containing 2% FBS and 10% HS. To observe the natural changes in C/EBPβ during the spontaneous differentiation of isolated SCs, cultures were maintained in DMEM containing 10% FBS, and differentiated in DMEM containing 2% HS.

Western Analysis

To assess expression of myoblast and myotube markers, the following antibodies were used: anti-MyoD (M-318 and 5.8a), anti-Myf-5 (C-20), anti-myogenin (M-225), anti-MyHC (H-300), anti-C/EBPβ (C-19), and anti-tubulin (B-7) (all Santa Cruz Biotechnology, Santa Cruz, CA, http://www.scbt.com), anti-cyclophilin B (Abcam ab16045, Cambridge, U.K., http://www.abcam.com), anti-Pax7, anti-myogenin, and anti-β-tubulin (DSHB, Iowa City, IA, http://www.dshb.biology.uiowa. edu). Chemiluminescence images were captured using the Luminescent Image Analyzer LAS-4000 (Fujifilm Life Science, Tokyo, Japan, http://www.fugifilm.com).

RT-qPCR

For RT-qPCR, RNA was extracted using the RNeasy kit (Qiagen, Hilden, Germany, http://www1.qiagen.com), remaining DNA was digested with DNase (Ambion, Austin, TX, http://www.invitrogen.com/site/us/en/home/brands/ambion.html), and RNA was reverse transcribed using iScript kit (BioRad, Hercules, CA, http://www.bio-rad.com) according to manufacturer's instructions. Real-time PCR reactions were performed with Quantitect SYBR green (Qiagen) on a Mx3005p thermocycler (Stratagene, La Jolla, CA, http://www.stratagene.com). Primers were designed to span an intron when possible and sequences are available upon request. Relative fold induction was determined using the ΔΔCt method [59] following normalization with 18S rRNA.

Indirect Immunofluorescence Staining

Indirect immunofluorescence was performed on paraffin-embedded and frozen sections of muscle, fixed in 4% paraformaldehyde. Cell cultures were fixed with ice-cold methanol. Detection was performed according to standard procedures using the following antibodies: anti-C/EBPβ (C-19), (Santa Cruz Biotechnology), anti-Pax7 and anti-MF20 (DSHB), anti-mouse-Cy5 (Invitrogen) and DyLight 549 (Piercenet, Rockford, IL, http://piercenet.com), anti-rabbit DyLight 488 (Piercenet) conjugates (Donkey Anti-Mouse or anti-rabbit IgG [H+L]) (Thermo Scientific, Ottawa, Canada, http://thermoscientific.com).

Analysis of Reporter Gene Expression

For Pax7 reporter assays, C2C12 cells were transfected with 3.8 kb Pax7 promoter-luciferase reporter construct [57] and a constitutively active RSV-β-galactosidase reporter in the presence or absence of mammalian expression plasmids for C/EPBβ using FuGene HD (Promega, Madison, WI, http://www.promega.com). Cells were grown under growth conditions for 24 hours, then collected for luciferase assays or switched to differentiation media for another 24 hours. Luciferase assays were performed according to standard protocol and corrected for transfection efficiency with β-gal enzyme activity. Error bars represent the SEM of a minimum of three experiments.

Myogenin promoter activity was measured by transfecting C2C12 cells with a −2.0 kb Myogenin-luciferase reporter construct [58] and a constitutively active RSV-β-galactosidase reporter in the presence or absence of mammalian expression plasmids for C/EPBβ and/or MyoD using FuGene HD (Promega). Cells were growth in growth conditions for 24 hours then switched to differentiation media for another 24 hours (DM). Samples were collected, and luciferase assays were performed according to standard protocol and corrected for transfection efficiency with β-gal enzyme activity. Error bars represent the SEM of a minimum of three experiments.

ChIP Assay

Freshly isolated SCs were purified by selective plating and were grown in growth media containing 10% FBS for 4 days, and ChIP was performed as described [33] using C/EBPβ (C-19) (Santa Cruz Biotechnology) for precipitation or a type matched nonspecific antibody at 4°C overnight. DNA fragments were purified using the Qiaquick PCR purification kit (Qiagen) and amplified by PCR using primers to amplify −695 to −465 of the murine Pax7 promoter using qPCR. The primer sequences used were: forward 5′-CCCGAACTGGCCCCCTTTCC-3′ and reverse 5′-TCCCCCGGAGGACTGGAACG-3′. Error bars represent the SEM of three experiments.

Myogenic Conversion Assay

C2C12 myoblasts retrovirally transduced with empty vector (pLXSN) or C/EBPβ were transiently transfected with a mammalian expression plasmid for MyoD and pEGFPplasmid using FuGene HD (Promega), as indicated in the figure legend. Cells were then induced to differentiate in 2% HS for 5 days. Cells were then fixed and stained for MyHC (MF20; DSHB) or harvested for protein. Error bars represent the SEM of three experiments.

C57BL/6 and C/EBPβ Conditional Knockout Mice

C57BL/6 female mice aged 6–8 weeks were obtained through Charles River laboratories. A mouse bearing C/EBPβ-floxed allele was created previously [38] and homozygous progeny (C/EBPβfl/fl) were obtained by breeding heterozygous progenitors. C/EBPβfl/fl mice were crossed with mice bearing the Pax7-CreERtm allele [37]. All animals were maintained in a controlled facility at 22°C with 30% relative humidity on a 12 hours light/dark cycle and provided food and water ad libitum. In vivo activation of CreERtm was achieved by five daily i.p. injections of 1.5 mg of tamoxifen (Sigma-Aldrich, Oakville, Canada, http://sigmaaldrich.com) dissolved in corn oil. Mice were aged 2–3 months, and no weight differences were noted between same-sex littermates. In utero activation of CreERtm was achieved by a single gavage of 2.5 mg of tamoxifen of pregnant dams when pregnancy was at E15.5. In culture, activation of CreERtm was achieved by a 72 hours treatment with 2 μM of 4-OH tamoxifen (Sigma).

For BaCl2 injury, mice were anesthetized with isofluorane before the procedure. Legs were shaved and washed with an antiseptic solution, after which 50 μl of 1.2% BaCl2 in PBS or PBS alone was injected intramuscularly into the TA. For all animal work, following sacrifice, hind limb muscles from homozygous null animals and WT littermates aged 5 weeks were flash frozen in isopentane and sectioned for indirect immunofluorescence or processed for isolation of primary myoblasts. For histological analysis, a minimum of 400 fibers from the TA muscle were measured on a minimum of two cross-sections separated by at least 50 μm. All animal handling procedures conformed to the guidelines established by the University of Ottawa Animal Care Service and the Canadian Council on Animal Care.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

C/EBPβ Is Expressed in Muscle SCs In Vivo

We first sought to quantify C/EBPβ expression in muscle extracts. As differential initiation of translation of the Cebpb mRNA results in three proteins with identical carboxy termini and variable amino termini, we investigated the expression of the two activating isoforms of C/EBPβ, known as LAP* (liver activating protein, full-length isoform) and LAP, which lacks the first 21 amino acids but contains all of the activation domains, as well as that of the dominant negative isoform liver inhibitory protein (LIP) in whole muscle extracts. Western analysis of C/EBPβ expression in C57BL/6 mouse extensor digitorum longus and soleus muscles revealed C/EBPβ expression in both muscles, with the LAP isoform most predominant in both muscles (Fig. 1A).

thumbnail image

Figure 1. C/EBPβ is expressed in muscle satellite cells. (A): Western analysis of C/EBPβ expression in C57BL/6 mouse EDL and Sol. muscle. Positions of the LAP*, LAP, and LIP isoforms of C/EBPβ are indicated. CyPB expression is used as a loading control. (B): Indirect immunofluorescence of Pax7 and C/EBPβ expression in C57BL/6 mouse TA muscle. DAPI staining reveals nuclei, satellite cell, as determined by Pax7 staining is indicated with a white arrow. Data are representative of three independent experiments performed on a minimum of three male C57BL/6 mice aged 4–6 weeks. Scale bar = 50 μm. (C): Expression of Pax7 and C/EBPβ in primary myoblasts cultured in growth medium as detected by indirect immunofluorescence. Nuclei are visualized with DAPI stain. Scale bar = 50 μm. Failure to add primary antibodies to the samples did not result in any fluorescent signal. (D): Western analysis of C/EBPβ, Pax7, and myogenic marker expression in skeletal muscle satellite cells 3 days postisolation cultured in GM lacking FGF/HGF and during differentiation following transfer to low serum conditions on day 4. The expected migration of the three isoforms of C/EBPβ (LAP*, LAP, and LIP) are indicated with arrowheads. CyPB is a loading control. (E): Relative Cebpb mRNA expression in SCs isolated from female C57BL/6 mouse hind limb (age 6–8 weeks) cultured in GM in the absence of growth factors (day 3 post-isolation) or after incubation in DM for 3 days. Error bars are the SEM for three independent experiments. p-value was calculated using a Student's t test assuming equal variance. Abbreviations: C/EBPβ, CCAAT/enhancer binding protein β; CyPB, Cyclophilin B; DAPI, 4′,6-diamidino-2-phenylindone; DM, differentiation medium; EDL, extensor digitorum longus; GM, growth medium; LAP, liver activating protein; LIP, liver inhibitory protein; MyoG, myogenin; MyHC, myosin heavy chain; Sol. Soleus.

Download figure to PowerPoint

Further analysis by indirect immunofluorescence indicated that C/EBPβ expression was not equally distributed throughout the muscle fiber but rather was localized to Pax7+ cells found at the perimeter of muscle fibers, corresponding to muscle SCs (Fig. 1B). When primary myoblasts were harvested and cultured in growth medium in the presence of both fibroblast growth factor (FGF) and hepatocyte growth factor (HGF) to prevent differentiation, we noted that all of the Pax7+ cells were also positive for C/EBPβ expression by indirect immunofluorescence, suggesting that C/EBPβ expression was present in myogenic precursor cells (Fig. 1C).

To explore the expression of C/EBPβ during activation and differentiation of SCs, SCs were harvested from hind limb muscle by enzymatic digestion and selective plating and cultured in growth medium without addition of growth factors for several days, allowing cells to adhere and proliferate. This method was chosen to allow us to observe changes in C/EBPβ expression in the premyoblastic cell, the myoblast, and the myocyte as the cells spontaneously progressed through these stages. Distinct waves of myoblast and myocyte marker expression were observed using this method. Four days after isolation (day 0 = day of harvest), cultures were switched to differentiation medium and allowed to differentiate for 3 days. Western analysis revealed that SC cultures initially expressed the LAP* and LAP isoforms of C/EBPβ, which decreased as cells progressed through differentiation (Fig. 1D). This pattern of downregulation mirrored the loss of Pax7 expression as the cells progressed through differentiation (Fig. 1D). Indeed, the decline in C/EBPβ and Pax7 expression coincided with an increase in both MyoD and myogenin expression upon switching to differentiation medium on day 4 (Fig. 1D). Upregulation of these myogenic factors resulted in the robust induction of MyHC expression on day 5 (Fig. 1D). As differentiation progressed, MyoD levels were observed to decrease and to return to baseline (Fig. 1D).

In accordance with these results, mRNA reverse transcription followed by quantitative real time polymerase chain reaction (RT-qPCR) analysis demonstrated that Cebpb mRNA expression in cells cultured in differentiation medium (low serum) for 3 days (7 days post-isolation) decreased by approximately 60% as compared to isolated SCs cultured in growth medium on day 3 post-isolation (Fig. 1E). Consistent with our results, microarray analyses comparing the gene expression profiles of quiescent SCs to activated differentiating cells revealed that Cebpb expression was reduced by 92% in activated cells [34].

C/EBPβ Inhibits the Differentiation of Myoblasts

To evaluate the roles of C/EBPβ expression in SCs during differentiation, we retrovirally transduced freshly isolated SCs from the hind limb of C57BL/6 mice aged 6–8 weeks (n = 3) to express C/EBPβ or with empty virus (pLXSN) and allowed these cells to differentiate in low serum conditions for 3 days. Ectopic expression of C/EBPβ was confirmed by Western blotting, and it was noted that the LAP isoform was the most prominent one expressed (Fig. 2A). We then analyzed myogenic gene expression in these cultures following differentiation. Ectopic C/EBPβ potently inhibited the differentiation of these cells as evidenced by a decrease in MyoD, myogenin, and MyHC expression, without affecting Myf5 levels (Fig. 2B), suggesting that the cells remained committed to the myogenic lineage but failed to differentiate efficiently. The undifferentiated phenotype observed in C/EBPβ-overexpressing cells was further supported by the increased expression of the SC and myoblast marker Pax7 (Fig. 2B). Given that C/EBPβ has been implicated in the regulation of cell proliferation in other systems, and that reduced cell proliferation could impact the efficiency of differentiation and fusion, we measured cell number in proliferating cultures using crystal violet (supporting information Fig. S1A) [35]. Equal cell numbers were plated for both control and test lines and cell density was evaluated 48 and 72 hours after plating. We did not measure any differences in cell number.

thumbnail image

Figure 2. C/EBPβ inhibits the differentiation of myoblasts. (A): Western analysis of C/EBPβ isoform expression in satellite cells isolated from three male C57BL/6 mice in three independent trials and retrovirally transduced to express full-length C/EBPβ or with empty vector (pLX). Cells were induced to differentiate in low serum for 3 days before protein analysis. CyPB is shown as a loading control. (B): Western analysis of myogenic protein expression in primary cells harvested, transduced, and differentiated as in (A). CyPB is shown as a loading control. (C): Indirect immunofluorescence of MyHC expression (as detected using MF-20 antibody) in primary myoblasts cultures transduced and differentiated as in (A). Scale bar = 100 μm. (D): DI (#myonuclei/# total nuclei) from five random fields of cells differentiated and stained as in (C) **, p < .01, n = 3. (E): Fusion index (FI, #myonuclei/#myotubes) from experiment as in (C). (F): Western analysis of C/EBPβ isoform expression in C2C12 cells retrovirally transduced to express C/EBPβ or with empty virus (pLX) and induced to differentiate in low serum conditions for 5 days. B-tubulin expression is used as a loading control. (G): Western analysis of myogenic protein expression in C2C12 cultures transduced and induced to differentiate as in (F). CyPB is shown as a loading control. (H): Bright-field images of Giemsa-stained (top) and MyHC immunostained C2C12 cells transduced and differentiated as in (F). Scale bar = 100 μm. (I): DI of cells in (H) and calculated as in (D), ***, p < .001, n = 3. (J): Fusion index from cells differentiated and stained as in (H), **, p < .01, n = 3. Abbreviations: C/EBPβ, CCAAT/enhancer binding protein beta; CyPB, Cyclophilin B; DI, differentiation index; LAP, liver activating protein; LIP, liver inhibitory protein; MyoG, myogenin; MyHC, myosin heavy chain.

Download figure to PowerPoint

Evaluation of MyHC expression by immunofluorescence revealed that there were fewer MyHC positive cells in C/EBPβ-overexpressing cultures as compared to empty virus controls with a fourfold reduction in the differentiation index (#myonuclei/total nuclei) 3 days after induction to differentiate (Fig. 2C, 2D). Of the MyHC-positive cells present in the C/EBPβ-overexpressing cultures, the myotubes were smaller such that the fusion index was reduced by approximately 10-fold in C/EBPβ-overexpressing cells as compared to empty virus controls (Fig. 2C, 2D). Taken together, these results support the notion that ectopic C/EBPβ induces a blockade of differentiation.

The effect of overexpressing C/EBPβ in C2C12 cells was similar to that of primary myoblasts. Retroviral transduction to express ectopic C/EBPβ in C2C12 myoblasts resulted in a robust increase in both LAP* and LAP isoforms of C/EBPβ protein levels 5 days after the induction of differentiation (Fig. 2F). The inhibitory LIP isoform was not detected in control cultures, although it was detected in C/EBPβ-overexpressing cultures, in contrast to our results in SCs (Fig. 2A, 2F). In this model, following 5 days in differentiation medium, we noted an increase in Pax7 expression and a concomitant decrease in MyoD, myogenin, and MyHC expression (Fig. 2G). Ectopic expression of C/EBPβ also inhibited myoblast fusion as evidenced by Giemsa staining (Fig. 2H). While empty vector controls displayed numerous purple-stained myotubes, C/EBPβ-overexpressing cells had only a few small myocyte-like cells (Fig. 2H). Immunostaining for MyHC revealed that while some MyHC positive cells existed in C/EBPβ-overexpressing cells, their fusion was limited (Fig. 2H). While approximately 20% of the empty virus control culture cells underwent differentiation to become MyHC+, only 5% of C/EBPβ overexpressing cells did (Fig. 2I). Furthermore, the fusion index of differentiated cells was also reduced by C/EBPβ overexpression, consistent with our results in SCs (Fig. 2J). As in SC cultures, cell growth was not perturbed by C/EBPβ expression in C2C12 myoblasts (supporting information Fig. S1B).

Given that the phenotype produced in C/EBPβ-overexpressing myoblasts closely resembled that of the Myod1 null [10], we sought to determine if the blockade in differentiation could be rescued by ectopic MyoD expression. C2C12 myoblasts retrovirally transduced with empty vector or to express C/EBPβ were transiently transfected following selection to express MyoD along with green fluorescent protein (GFP) and then induced to differentiate for 4 days in low serum conditions (supporting information Fig. S2A). Under these conditions, overexpression of C/EBPβ resulted in a 50% reduction in the differentiation index with ectopic expression of MyoD restoring the differentiation index to the level of controls (supporting information Fig. S2B). Furthermore, transient overexpression of MyoD resulted in a rescue of myogenin expression (supporting information Fig. S2C), suggesting that C/EBPβ acts to inhibit differentiation at the level of MyoD expression and/or activity. Of note, transfection efficiencies were extremely low, resulting in less than 20% of the cells being transfected (supporting information Fig. S2D, S2E). However, despite inefficient transfection of MyoD into C/EBPβ-overexpressing cultures, MyoD was able to rescue the differentiation blockade imposed by C/EBPβ, suggesting that ectopic C/EBPβ inhibits myogenesis at the level of MyoD.

Pax7 Is a Target of C/EBPβ and Represses Myogenic Gene Expression

Since C/EBPβ inhibits MyoD protein expression and myogenesis in both primary myoblasts and C2C12 cells, we next analyzed myogenic transcript levels by RT-qPCR in C2C12 cells retrovirally transduced to express C/EBPβ and differentiated for 5 days in low serum conditions. Cebpb transcript levels increased 20-fold in these cells as compared to empty vector controls (Fig. 3A). Despite an important reduction of MyoD protein expression in C/EBPβ overexpressing cultures, Myod1 mRNA levels were unaffected, as was Myf5 expression. However, significant decreases in Myog (myogenin) and both the embryonic (Myh3) and perinatal/adult (Myh) MyHC isoforms 1, 2, 8, and 13 were measured in C/EBPβ overexpressing cells (Fig. 3B). We also observed a significant increase in Fbxo32/atrogin-1 expression in C/EBPβ-overexpressing cells, an E3 ubiquitin ligase known to be regulated by C/EBPβ in myofibers and to promote the degradation of MyoD protein by the 26S proteasome [14, 30] (Fig. 3B). Thus, we hypothesized that the increase in atrogin-1 expression contributed to the loss of MyoD protein seen in differentiating myoblast cultures overexpressing C/EBPβ. To test this, we treated C/EBPβ-overexpressing C2C12 cells and empty virus controls with the proteasome inhibitor MG132 for 2 hours prior to harvest and evaluated MyoD levels by Western blotting (supporting information Fig. S3A). Despite the increase in Fbxo32 expression, blocking proteasome activity with MG132 resulted in only modest and variable increases in MyoD expression under differentiation conditions, suggesting that while atrogin-1 may contribute to reducing MyoD expression in C/EBPβ-overexpressing cells, it is unlikely to be the principal mechanism (supporting information Fig. S3).

thumbnail image

Figure 3. C/EBPβ regulates myogenic gene expression. (A): RT-qPCR analysis of Cebpb expression in C2C12 cells retrovirally transduced to express C/EBPβ or with empty virus and differentiated in low serum conditions for 5 days. *, p < .01 as determined using a paired t test, error bars are the SEM. (B): RT-qPCR analysis of myogenic gene expression in C2C12 transduced and differentiated as in (A). Error bars are the SEM. *, p < .05; **, p < .01 as determined using a paired t test. Myh represents an amplicon common to Myh1, 2, 8, and 13. (C): RT-qPCR analysis of Pax7 expression in C2C12 transduced and differentiated as in (A), *, p < .05 as determined by a paired t test. (D): RT-qPCR analysis of Cebpb expression in C2C12 retrovirally transduced to express C/EBPβ or with empty virus (pLX) and cultured in growth conditions. (E): RT-qPCR analysis of myogenic gene expression in C2C12 cells transduced as in (D), *, p < .05; **, p < .01. (F): RT-qPCR analysis of Pax7 expression in C2C12 transduced as in (D). (G): Western analysis of myogenic marker expression in C2C12 cells transduced as in (D) cultured in growth conditions. Only the LAP* (liver activating protein, full-length isoform) and LAP isoforms for C/EBPβ are shown (see Fig. 2F). (H): ChIP analysis of C/EBPβ occupancy on the mouse Pax7 promoter region −695/−465 in freshly isolated satellite cells as determined by gel analysis (top) and by qPCR (bottom). qPCR data are shown as enrichment over pulldown with a type-matched nonspecific antibody (IgG). *, p < .05 as determined using a paired t test. (I):Pax7 promoter activity in a transient transcription assay where the mouse –3800/+21 Pax7 promoter drives the expression of luciferase. The reporter construct and a mammalian expression vector for C/EBPβ were transiently transfected into C2C12 cells, and luciferase activity was measured in GM and in DM. **, p < .01, n = 3. (J): Myogenin promoter activity as measured in (I) under DM conditions using a −2 kb myogenin-luc reporter construct. *, p < .05, n = 3. Abbreviations: C/EBPβ, CCAAT/enhancer binding protein beta; ChIP, chromatin immunoprecipitation; DM, differentiation medium; GM, growth medium; qPCR, quantitative real time polymerase chain reaction.

Download figure to PowerPoint

In addition to induction of Fbxo32 expression, Pax7 expression was also increased 40-fold in C/EBPβ stable cell lines (Fig. 3C), consistent with our Western analysis (Fig. 2G). High levels of Pax7 expression have been correlated with the self-renewal of SCs and the inhibition of differentiation and thus could explain the phenotype evoked by forced C/EBPβ expression in myoblasts [36]. However, given that Pax7 levels drop as differentiation progresses, the high Pax7 levels could also represent a culture that fails to differentiate as in the case of C/EBPβ-overexpressing cells. We thus assessed myogenic marker expression in C/EBPβ-overexpressing cultures under growth conditions, where differentiation is inhibited (Fig. 3D–3G). In growth medium, RT-qPCR analysis revealed that Cebpb was expressed 40-fold over empty vector controls (Fig. 3D), and both Myf5 and Myog expression was significantly decreased (Fig. 3E). Despite a modest increase in Myod1 expression, this effect was not statistically significant (Fig. 3E). Fbxo32 expression was also unaffected by C/EBPβ overexpression under growth conditions (Fig. 3E), which corresponded with no change in MyoD protein expression (Fig. 3G). Thus, while overexpression of C/EBPβ can promote loss of MyoD protein under differentiation conditions, it did not in growth medium. These results suggest that C/EBPβ activity could be sensitive to serum levels resulting in differential regulation of protein expression in growth and differentiation media.

Despite some differences in gene expression in growth and differentiation media, the expression of Pax7 was still significantly increased 12-fold in C2C12 cells stably expressing C/EBPβ (Fig. 3F). Western analysis of myogenic protein markers echoed the RT-qPCR results, with increased Pax7 expression, and unchanged MyoD expression in C/EBPβ overexpressing cells (Fig. 3G). Interestingly, Myf5 levels were only moderately decreased by C/EBPβ overexpression (Fig. 3G).

To determine if the regulation of Pax7 expression by C/EBPβ was direct, we first performed in silico analysis of the Pax7 proximal promoter to identify potential C/EBP response elements, revealing one putative element (TTGCACA) at position −590/−583. Less stringent algorithms also predicted elements in the following regions: −3,006/−2,808, −2,646/−1,955, and −1,845/−933. Chromatin immunoprecipitation (ChIP) analysis performed in freshly isolated SCs in growth conditions (where endogenous C/EBPβ levels are high) revealed that C/EBPβ did occupy the Pax7 promoter at position −695/−465 (containing the putative element at −590/−583) but none of the other predicted sites (Fig. 3H). Furthermore, transient reporter assays performed in C2C12 myoblasts in both growth and differentiation conditions revealed that C/EBPβ can activate the Pax7 promoter, most robustly in low serum conditions (Fig. 3I). Thus, the activation of Pax7 expression by C/EBPβ likely contributes, at least in part, to the C/EBPβ-dependent inhibition of myogenesis.

In addition to compromising myogenesis through the activation of both atrogin-1 and Pax7 expression in myoblasts, C/EBPβ also interferes with MyoD transcriptional activity. Reporter assays using the −2 kb myogenin promoter revealed that coexpression of C/EBPβ with MyoD blocked the activation of the myogenin promoter by MyoD (Fig. 3J). It remains unclear if this inhibition is mediated by a direct effect of C/EBPβ on MyoD or rather via the previously described inhibition of MyoD activity by high levels of Pax7 [16]. Taken together, these results suggest that C/EBPβ expression can abrogate myogenesis through at least two converging mechanisms at the level of MyoD expression and function.

Loss of Cebpb Expression Results in Precocious Differentiation Under High Serum Conditions

To evaluate the importance of changes in Cebpb expression during the differentiation of primary myoblasts, we generated a conditional knockout mouse model by breeding a floxed mouse (Cebpbfl/fl) with a Pax7CreERtm mouse [37] (Fig. 4A). The floxed Cebpb mouse was bred to heterozygosity for the Pax7CreERtm allele, and stud males were crossed with Cebpbfl/fl females (Fig. 4B). The Pax7CreERtm allele expresses, by virtue of a bicistronic expression cassette, a tamoxifen-inducible Cre under the control of the Pax7 promoter, thereby not affecting Pax7 expression [37, 38]. Both experimental (Cebpb−/−Pax7CreER/+) and control mice (Cebpbfl/flPax7+/+) were generated at Mendelian ratios. Cebpb excision was accomplished in CreER-expressing floxed animals by daily i.p. injections of tamoxifen for 5 days. This treatment reduced the percentage of Pax7+/C/EBPβ+ double positive cells by approximately 74% 1 week after the end of treatment (Fig. 4C). The percentage of Pax7+ nuclei in muscle sections was unaffected by Cebpb excision at the time of harvest (fl/fl: 1.83% ± 0.52%; −/−: 1.81% ± 0.99%). This excision rate is consistent with other reports using this strain of Cre mouse [37]. Following isolation of SCs, floxed and null cultures were maintained in growth medium including FGF/HGF, and under these conditions only minimal differentiation was observed (Fig. 4D, top). However, Cebpb−/− cultures underwent more differentiation than floxed controls, although still only restricted to a small fraction of cells. Withdrawal from growth factors in high serum conditions resulted in the stimulation of differentiation of the Cebpbfl/fl control cultures as measured by immunostaining for MyHC, while the Cebpb−/− cultures differentiated robustly producing large multinucleated fibers (Fig. 4D, bottom). Furthermore, both the differentiation and fusion indices were significantly increased in null cells as compared to floxed controls upon withdrawal of growth factors (Fig. 4E, 4F). In accordance with these observations, Western analysis of myogenic markers revealed that withdrawal from the growth factors decreased C/EBPβ expression in floxed controls (Fig. 4G). Furthermore, reduced C/EBPβ expression correlated with a decrease in Pax7 expression and a concomitant increase in myogenin expression (Fig. 4G). Interestingly, loss of C/EBPβ expression in complete growth medium resulted in an increase in MyoD protein expression, supporting our hypothesis that C/EBPβ is a regulator of MyoD expression (Fig. 4G). MyoD levels were lowest in the most differentiated of cultures, Cebpb−/− cells withdrawn from growth factors (Fig. 4G).

thumbnail image

Figure 4. Loss of Cebpb expression in Pax7+ cells results in precocious differentiation under high serum conditions. (A): Schematic representation of the transgenic animals used to generate a tamoxifen-inducible excision of Cebpb in Pax7+ cells. (B): Breeding strategy used to generate Cebpb conditional knockout progeny for experimentation. (C): Excision efficiency in Cre+ satellite cells (SCs) isolated from 11.5-week-old mice following five i.p. injections of tamoxifen 1 week prior to sacrifice, calculated as the percentage of Pax7+ cells isolated by enzymatic digestion from skeletal muscle that are also C/EBPβ+ by immunocytochemistry in floxed controls (fl/fl) and Cre+ (−/−) cells. Error bars are the SD. **, p < .01 (D): Immunostaining of myosin heavy chain expression in primary myoblasts cells isolated as in (C) and cultured in complete growth medium or the absence of growth factors (−GF). Images are representative of three independent trials from two pairs of mice. Scale bar = 100 μm. (E): Differentiation index in growth medium without growth factors calculated from three random images per trial and three trials of the images in (D). *, p < .05, n = 3. Error bars are the SD. (F): Fusion index in growth medium without growth factors from experiment as in (D). *, p < .05, n = 3. (G): Western analysis of C/EBPβ and myogenic marker expression in primary myoblasts derived from conditional nulls (−/−) and from floxed littermate controls (fl/fl) grown in growth medium in the presence (GM) or absence (GM-GF) of fibroblast growth factor/hepatocyte growth factor. Abbreviations: C/EBPβ, CCAAT/enhancer binding protein beta; GM, growth medium; GF, growth factor; myoG, myogenin G.

Download figure to PowerPoint

Among the growth factors included in growth media to prevent early myogenesis, HGF has been shown to regulate C/EBPβ expression [39]. Withdrawal of HGF from the growth medium for 48 hours decreased Cebpb expression by 50% (supporting information Fig. S4A). This decrease in Cebpb expression correlated with a modest yet significant decrease in Pax7 mRNA expression and an increase in Myog expression, suggesting enhanced differentiation in the withdrawn cultures consistent with our results in the Cebpb conditional null cultures (supporting information Fig. S4A). At the protein level, C/EBPβ and Pax7 protein levels were decreased following withdrawal from HGF (supporting information Fig. S4B). Taken together, these data suggest that HGF may act to restrain differentiation of primary myoblast cultures in part by maintaining the expression of C/EBPβ.

Loss of Cebpb Expression in Pax7+ Cells Results in Increased Cell Fusion Under Differentiation Conditions

SCs were isolated from Cebpbfl/flPax7CreER+/− and Cebpbfl/flPax7+/+ muscle and were maintained in culture in growth medium containing 20% fetal bovine serum (FBS) and FGF/HGF to inhibit differentiation. To induce excision of Cebpb, cells were treated with 4-hydroxy-tamoxifen for 3 days to activate the Cre recombinase. 4-OH-tamoxifen treatment resulted in a 45% excision of Cebpb in isolated genomic DNA and in a robust decrease in C/EBPβ LAP protein and mRNA expression (Fig. 5A, 5D, 5E). To induce differentiation, cultures were switched to low serum conditions for 2 days, after which immunostaining for MyHC revealed enhanced myotube size in Cebpb−/− cells as compared to the floxed controls with a 50% increase in fusion index (Fig. 5C, top). Loss of Cebpb expression also resulted in only a mild (17%), although not significant increase in differentiated index (Fig. 5C, bottom). Indeed, the levels of differentiation achieved in these experiments surpassed 85%, making the determination of enhanced differentiation difficult. Nonetheless, in accordance with these observations, Western analysis of myogenic marker expression 2 days following the induction to differentiate indicated that null cells expressed comparable levels of myogenic proteins (MyoD, myogenin, and MyHC) as controls (Fig. 5D). Analysis of relative mRNA expression by RT-qPCR revealed that while Cebpb levels were dramatically reduced in null cells, all other genes tested in Cebpb−/− cells were comparable to floxed controls.

thumbnail image

Figure 5. Loss of Cebpb expression in Pax7+ cells results in increased cell fusion under differentiation conditions. Primary myoblasts from female Cebpbfl/flCreER−/− and Cebpbfl/flCreER+/- hind limb muscle (aged 8–12 weeks) treated in culture with 2 μM 4-hydroxy-tamoxifen for 3 days following isolation to excise Cebpb. Only the Cre-expressing cells excised Cebpb (−/−) with an efficiency of approximately 55%. (A): Confirmation of efficient excision of Cebpb as measured by gel densitometry with amplification of Cebpb from genomic DNA. (B): Immunocytochemical analysis of MyHC expression in myotubes generated from primary myoblasts derived from control (fl/fl) and excised (−/−) littermates following differentiation in low serum conditions for 2 days. Scale bar = 100 μm. (C): Fusion and differentiation indices (shown relative to floxed controls) calculated from cells differentiated as in (D). **, p < .01, n = 4 as determined using a paired t test. (D): Western analysis of C/EBPβ and myogenic marker expression in primary myoblasts differentiated as in (B). CyPB is used as a loading control. (E): RT-qPCR analysis of myogenic marker expression in control (fl/fl) and excised (−/−) satellite cell cultures differentiated as in (B). Myh represents an amplicon common to Myh1, 2, 8, and 13. ***, p < .001, n = 4. Abbreviations: C/EBPβ, CCAAT/enhancer binding protein beta; CyPB, cyclophilin B; MyHC, myosin heavy chain.

Download figure to PowerPoint

SC-Specific Loss of Cebpb Expression Results in Muscle Fiber Hypertrophy

Given the robust precocious differentiation of Cebpb−/− SCs in culture and the enhanced fusion observed, we investigated the histology of Cebpb−/− muscle following excision of Cebpb in utero induced by a single tamoxifen gavage of pregnant dams at E15.5. Mice were sacrificed at postnatal day 21, and the muscle phenotype was characterized. Excision, as measured by dual immunofluorescence staining of tibialis anterior (TA) muscle sections for Pax7 and C/EBPβ indicated a significant 50% decrease in double positive cells (Fig. 6A). The total number of Pax7+ cells in the muscle was unaffected in null mice at this time point (Fig. 6B). Histologically, no apparent differences were observed in female Cebpb−/− TA muscle as compared to sex-matched littermates, while fiber size was perceptibly larger in the male nulls (Fig. 6C, 6D). Consistent with these observations, the average fiber cross-sectional area of female mice muscle fibers was not different from littermate controls, whereas the conditional knockout males had cross-sectional areas one-third larger than controls (Fig. 6D). The increase in cross-sectional area was not accompanied by changes in overall fiber number (Fig. 6E). When fiber cross-sectional areas were plotted as a distribution, it was noted that for both female and male mice, there was a shift of the distribution toward larger fiber sizes in null animals as compared to littermate controls (Fig. 6F, 6G). Indeed, despite no difference in average cross-sectional area in female mice, there was both an increase in larger fibers and a significant decrease in smaller fibers, suggesting that there is a modest fiber hypertrophy in the null female animals (Fig. 6F). The shift in fiber size distribution was even more pronounced in male mice (Fig. 6G). As the mice aged to postnatal day 56, the trend toward larger fiber sizes in null mice became more marked, although the sample sizes (two animals per genotype) were small (supporting information Fig. S5). These results indicate that loss of Cebpb expression in SCs promotes fiber hypertrophy, which may result from enhanced differentiation and fusion.

thumbnail image

Figure 6. Cebpb conditional knockouts display fiber hypertrophy. (A): Percent of Pax7+ cells also positive for C/EBPβ in Cebpbfl/fl and Cebpb−/− in 21-day-old mice following a single 2.5 mg tamoxifen gavage of pregnant dams at E15.5 as determined by fluorescent IHC on cross-sections of the tibialis anterior (TA) muscle. n ≥ 3 animals of each genotype. (B): Percentage of Pax7+ nuclei relative to total nuclei in sections from TA muscle derived from 21-day-old Cebpbfl/fl and Cebpb−/− mice following maternal gavage at E15.5. Error bars are the SD. (C): Representative bright-field images of TA cross-sections from control female and male Cebpbfl/fl (fl/fl) and conditional null Cebpb−/−Pax7Cre/+ (−/−) animals at postnatal day 21 stained with hematoxylin and eosin. Scale bar = 100 μm. (D): Average cross-sectional areas of muscle fibers from female (♀) and male (♂) control (fl/fl) and conditional null (−/−) animals. For each group n ≥ 3. Error bars are the SD. *, p < .02. (E): Average fiber number for the TA muscles of control and conditional null mice (n ≥ 3 for each group). (F): Frequency distribution of fiber size in female control and conditional null mice. *, p < .05; **, p < .02. Error bars are the SD. (G): Frequency distribution of fiber size in male control and conditional null mice. *, p < .05; **, p < .02; ***, p < .01. Error bars are the SD. Abbreviations: C/EBPβ, CCAAT/enhancer binding protein beta; XSA, cross-sectional area.

Download figure to PowerPoint

Loss of Cebpb Expression in SCs Promotes Healing After Muscle Injury

To evaluate the regenerative capacity of conditional null muscle, we induced the excision of Cebpb in floxed animals by daily i.p. injections of tamoxifen for 5 days. Excision was confirmed in 74% of Pax7+ cells (Fig. 4C). One week following completion of treatment, male mice were subjected to a single BaCl2 injection to the left TA muscle and allowed to recover for 1 week. After sacrifice, muscle sections were analyzed for the extent of repair. This short time frame did not permit the development of fiber hypertrophy in uninjured Cre-expressing muscle (Fig. 7A). Examination under light microscopy revealed that the healing process, as determined by the presence of centrally located nuclei, was more advanced in null animals as compared to littermate controls not expressing the Cre (Fig. 7A). Indeed, the average cross-sectional area of regenerating fibers (as defined by the presence of a centrally located nuclei) was increased by 30% in conditional null animals as compared to floxed controls (Fig. 7B). Frequency distribution of regenerating fiber cross-sectional areas revealed that there was a pronounced shift in fiber size toward larger fibers in the conditional nulls as compared to littermate controls (Fig. 7C), suggesting that repair was more advanced in the null animals. These observations are consistent with our culture work, and taken together, support the notion that the loss of Cebpb in the muscle SCs promotes a more robust differentiation program.

thumbnail image

Figure 7. Muscle repair following BaCl2 injury is enhanced in Cebpb conditional knockout animals. Four male Cebpbfl/flPax7+/+ and Cebpb−/−Pax7Cre/+ mice were induced to excise Cebpb by five daily i.p. injections of tamoxifen (see Fig. 4C). Mice were injured with BaCl2 1 week after excision. (A): Representative bright-field images of injured (BaCl2) and uninjured (vehicle) tibialis anterior (TA) muscle from control (fl/fl) and conditional null (−/−) mice 7 days postinjury. Data are representative of four experimental pairs aged 2–3 months old. Scale bar = 100 μm. (B): Average cross-sectional area of regenerating fibers from injured TA muscle in control (fl/fl) and conditional null (−/−) male mice. Error bars are the SEM, n = 4 animals per group, p < .05. (C): Distribution of cross-section areas of regenerating fibers (defined as those with centrally located nuclei) from injured TA of mice treated as in (A). *, p < .05; **, p < .01, n = 4 per group. Error bars are the SEM. Abbreviation: XSA, cross-sectional area.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

Ectopic expression of C/EBPβ produces a differentiation defect that partially recapitulates the phenotype of Myod1−/− skeletal muscle precursor cells by impinging on MyoD expression and activity through at least two converging mechanisms: the stimulation of proteasomal degradation of MyoD, the upregulation of Pax7 expression, and the interference with MyoD transcriptional activity [8, 10]. In both Myod1−/− myoblasts and myoblasts overexpressing C/EBPβ, myogenic marker expression is limited and fusion inhibited. In culture, loss of C/EBPβ expression results in precocious differentiation in high serum conditions which normally restrains differentiation. In vivo, we observe larger caliber fibers in the conditional null mice as compared to their littermate controls and more efficient repair. These observations define a novel role for C/EBPβ as a transcription factor important not only for the commitment of stem cells to adipose or osteoblast lineages, but also in myogenesis as a tissue SC marker, a regulator of Pax7 and MyoD expression and an inhibitor of differentiation [40, 41].

Based on our results, we propose that C/EBPβ expression in SCs acts to maintain the undifferentiated state. Indeed, many similarities exist between the myogenic precursor marker Pax7 and C/EBPβ expression and function. Both factors are expressed in SCs and their expression is downregulated as cells progress toward the differentiated state. Expression of either factor correlates with the maintenance of the undifferentiated state, and our results indicate that the ectopic expression of C/EBPβ in myoblasts results in increased Pax7 expression in both growth and differentiation conditions. Thus, the parallel course of C/EBPβ and Pax7 expression may be due to direct regulation of Pax7 expression by C/EBPβ. Given that Pax7 is only required until juvenile age for proper functioning of SCs, it is interesting to speculate that C/EBPβ may play a role in the maintenance of SCs in the adult, a hypothesis with many important implications for the development of treatments for muscular atrophies [22, 42]. Treatments that would prevent the loss of C/EBPβ in primary myoblasts would be predicted to maintain the undifferentiated state of these cells and could conceivably have tremendous therapeutic potential by improving myoblast engraftment and SC niche repopulation.

It is interesting to note that C/EBPβ has a different effect on myogenic regulatory factor expression, particularly MyoD, under high serum (GM) versus low serum (DM) conditions. It has previously been shown that HGF induces Cebpb expression and C/EBPβ DNA binding activity in hepatocytes and that this induction is enhanced by serum [39, 43]. In addition to being regulated by HGF, C/EBPβ has also been shown to be rapidly and transiently phosphorylated by growth hormone (GH). GH-induced phosphorylation of C/EBPβ was required for its transcriptional activity in fibroblasts cells [44]. It would be interesting to investigate serum- and growth factor-dependent regulation of C/EBPβ expression and post-translationnal modifications during the process of myogenesis. This could lead to a better understanding of the mechanisms by which C/EBPβ regulates SC function.

Misexpression of C/EBPβ in SCs in pathological conditions would be predicted to impair normal repair mechanisms. C/EBPβ expression can be induced by lipopolysaccharides, interleukin (IL)-6, and IL-1, and its nuclear localization is promoted by Tumor necrosis factor-α [45–47]. Thus, systemic inflammation could result in increases in C/EBPβ expression in muscle SCs. Given that the activation of MyoD expression in SCs is required for their differentiation and regeneration, our data suggest that C/EBPβ levels must first be downregulated for differentiation to occur and thus signals that prevent this downregulation event could trigger a failure of muscle regeneration. Indeed, both aging (sarcopenia) and sepsis can trigger increases in muscle C/EBPβ expression and muscle wasting [48–50]. Assessment of engraftment efficiency, self-renewal, and differentiation of muscle stem cells expressing ectopic C/EBPβ into skeletal muscle would address many of these questions.

Loss of C/EBPβ expression in Pax7+ SCs not only permitted the differentiation of myoblasts in high serum conditions in the absence of growth factors but also enhanced their fusion in culture and resulted in muscle fiber hypertrophy in vivo, suggesting that the expression of C/EBPβ may, in addition to controlling the efficiency of differentiation, contribute to the inhibition of fusion. Kruppel-like factors 2 and 4 have been shown to have critical functions for fusion during myogenesis, as knockdown of both these factors in primary myoblasts abrogated the fusion of myocytes [51]. Interestingly, C/EBPβ is a known repressor of Klf4 expression in preadipocytes, thereby linking the loss of Cebpb expression to the control of myocyte fusion that should be further explored [52].

With this new role as an inhibitor of myogenesis, C/EBPβ emerges as a central regulator of mesenchymal differentiation in the postnatal organism, acting to promote the formation of fat mass at the expense of lean mass [27, 31, 32]. Thus signaling pathways that impinge on C/EBPβ activity could force lineage decisions in multipotent stem cells in this manner. Forced expression of C/EBPβ in muscle SCs and C2C12 myoblasts did not promote adipoconversion in our experiments but rather restrained myogenesis. This may be due to the unaffected Myf5 expression observed, which would act to maintain commitment to the myogenic lineage. However, persistent expression of C/EBPβ in SCs coupled with a permissive environment such as signaling that promotes C/EBPβ transcriptional activity could conceivably force the differentiation of SCs into adipocytes [53, 54]. Indeed, many examples of muscle wasting including sarcopenia and Duchenne's muscular dystrophy are characterized by an abnormal accumulation of fat tissue within the muscle [55, 56].

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

This work is supported by a grant from the Canadian Institutes of Health Research (CIHR) and the Cancer Research Society with funds from the University of Ottawa's Research Development Program. F.M. is supported by an Ontario Graduate Scholarship. The authors wish to thank Dr. Esta Sterneck for the C/EBPβfl/fl mice, Dr. L. Shen for the Pax7 reporter construct, Dr. Ilona Skerjanc and Dr. Alexandre Blais for interesting discussion and for DNA constructs, Dr. Bernard Jasmin for critical reading of this manuscript, John Lunde, Émilie Lamarche, and Loretta Cheung for technical assistance, and Dr. David Lohnes for research support. We benefitted from the use of several antibodies, listed in the Methods section, that were obtained from the Developmental Studies Hybridoma Bank (DSHB) developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biology, Iowa City, IA, USA.

DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

Dr. Charles Keller has received honoraria from Novartis within the last 12 months. Dr. Charles Keller holds intellectual property rights and ownership interests in Numira Biosciences. Dr. Charles Keller has received research funding from Eli Lilly, Johnson & Johnson, and Blueprint within the last 12 months.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL C ONFLICTS OF INTEREST
  9. REFERENCES
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
sc-12-0136_sm_SupplFigure1.tif956KFigure S1. Overexpression of C/EBPβ in myoblasts does not affect cell growth. (A) Crystal violet assay to determine cell number of satellite cell cultures retrovirally transduced to express C/EBPβ or with empty vector (pLX) and cultured for 48 hours in growth medium. Error bars are the SEM, n=3. (B) Crystal violet assay to determine cell number of C2C12 cells retrovirally transduced to express C/EBPβ or with empty vector (pLX) and cultured for 48 hours in growth medium. Error bars are the SEM, n=4.
sc-12-0136_sm_SupplFigure2.pdf96KFigure S2. Restoration of MyoD protein expression rescues the differentiation defect in C/EBPβ-overexpressing C2C12 cells. (A) Indirect immunostaining for MHC expression in C2C12 cells retrovirally transduced to express C/EBPβ or with empty vector (pLXSN), and transiently transfected to express GFP and MyoD as indicated. Cells were differentiated in low serum conditions for 4 days prior to immunostaining. DAPI staining reveals nuclei. Scale bar = 50μm. (B) Differentiation index of cultures treated as in (A). *p<0.05, n=3. Error bars are the standard deviation. (C) Western analysis of C/EBPβ, MyoD and myogenin expression in C2C12 cultures transfected and differentiated as in (A) β-tubulin is used as a loading control. (D) Images of GFP expression, merged with phase contrast images, to determine transfection efficiency. (E) Transfection efficiencies, defined as the percentage of GFP+ cells in the culture 24 hours after transfection, for each transfection condition. One way ANOVA analysis revealed a significant decrease (p<0.05) between conditions labeled a and b.
sc-12-0136_sm_SupplFigure3.tif1080KFigure S3. Loss of MyoD expression is partially rescued by the inhibition of the proteasome. (A) Western analysis of C2C12 myoblasts retrovirally transduced to express C/EBPβ or with empty vector (pLXSN) and cultured in 2% horse serum for 36 hours. Two hours prior to harvest, cells were treated with 50 μM MG132. Cyclophilin B is used as a loading control. (B) Western analysis of C2C12 myoblasts retrovirally transduced to express C/EBPβ or with empty vector (pLXSN) and cultured in growth medium pretreated with MG132 for 2 hours prior to harvest.
sc-12-0136_sm_SupplFigure4.tif535KFigure S4. Withdrawal of HGF from growth medium reduces Cebpb expression and promotes differentiation. (A) RT-qPCR analysis of Cebpb, Pax7 and Myog expression in WT satellite cell cultures grown in complete growth medium or growth medium lacking HGF for 48 hours. *p<0.05, **p<0.01, n=4. (B) Western analysis of C/EBPβ and Pax7 expression in SCs cultured and treated as in (A).
sc-12-0136_sm_SupplFigure5.pdf113KFigure S5. Fiber hypertrophy increased at day P56 in conditional knockout mice. (A) Representative bright field images of tibialis anterior cross sections from control Cebpbfl/fl and conditional null Cebpb-/-Pax7Cre/+ animals at postnatal day 56 stained with hematoxylin and eosin. Scale bar = 100 μm. (B) Average cross-sectional areas of muscle fibers from control and conditional null animals. For each group n=2 animals (1 male and 1 female). Error bars are the standard deviation. (C) Relative fiber number in the tibialis anterior muscles of control and conditional null mice (n=2 for each group). Control animals are set arbitrarily at 100. (D) Frequency distribution of fiber size in control and conditional null mice. Error bars are the standard deviation.
sc-12-0136_sm_SupplMethods.pdf76KSupplementary Data

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.