Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Canonical Wnt signaling induces BMP-4 to specify slow myofibrogenesis of fetal myoblasts

Canonical Wnt signaling induces BMP-4 to specify slow myofibrogenesis of fetal myoblasts Background: The Wnts are secreted proteins that play important roles in skeletal myogenesis, muscle fiber type diversification, neuromuscular junction formation and muscle stem cell function. How Wnt proteins orchestrate such diverse activities remains poorly understood. Canonical Wnt signaling stabilizes β-catenin, which subsequently translocate to the nucleus to activate the transcription of TCF/LEF family genes. Methods: We employed TCF-reporter mice and performed analysis of embryos and of muscle groups. We further isolated fetal myoblasts and performed cell and molecular analyses. Results: We found that canonical Wnt signaling is strongly activated during fetal myogenesis and weakly activated in adult muscles limited to the slow myofibers. Muscle-specific transgenic expression of a stabilized β-catenin protein led to increased oxidative myofibers and reduced muscle mass, suggesting that canonical Wnt signaling promotes slow fiber types and inhibits myogenesis. By TCF-luciferase reporter assay, we identified Wnt-1 and Wnt-3a as potent activators of canonical Wnt signaling in myogenic progenitors. Consistent with in vivo data, constitutive overexpression of Wnt-1 or Wnt-3a inhibited the proliferation of both C2C12 and primary myoblasts. Surprisingly, Wnt-1 and Wnt-3a overexpression up-regulated BMP-4, and inhibition of BMP-4 by shRNA or recombinant Noggin protein rescued the myogenic inhibitory effect of Wnt-1 and Wnt-3a. Importantly, Wnt-3a or BMP-4 recombinant proteins promoted slow myosin heavy chain expression during myogenic differentiation of fetal myoblasts. Conclusions: These results demonstrate a novel interaction between canonical Wnt and BMP signaling that induces myogenic differentiation towards slow muscle phenotype. Keywords: Skeletal muscle, Fetal myoblasts, Canonical Wnt signaling, BMP4 signaling, Differentiation, Slow muscle specification Background key transcription factors involved in myogenesis [6-10]. Skeletal muscles of the trunk and limb, except for some Importantly, genetic knockout studies have clearly dem- craniofacial and esophageal muscles, are derived from so- onstrated the requirement of several Wnt molecules and mites during embryonic development [1-3]. Specification β-catenin in the normal development of skeletal muscles of somitic cells into myogenic lineages is regulated by [11,12]. These diverse functions of Wnt are mediated by positive and negative signals from the surrounding tissues. both a canonical signaling pathway that requires sta- Wnt signaling induced by Wnt-1, -3a, -4, -6, -7a and 11 bilization and nuclear translocation of β-catenin, and non- from dorsal neural tube or ectoderm is critical for the in- canonical pathway that is independent of β-catenin [5]. duction, initiation and progression of myogenesis in the Therefore, canonical and non-canonical Wnt signaling presomitic mesoderm and early somites (Reviewed in pathways play multiple essential roles in embryonic [4,5]). Within the embryonic myogenic progenitors, Wnt myogenesis. also regulate the expression of Pax3/7, MyoD and Myf5, Wnt signaling is also involved in the regulation of post- natal satellite cell function and skeletal muscle rege- * Correspondence: mrudnicki@ohri.ca neration. Satellite cells are muscle resident stem cells Sprott Center for Stem Cell Research, Ottawa Hospital Research Institute, responsible for postnatal regeneration of injured muscles. 501 Smyth Road, Ottawa, ON K1H 8L6, Canada Full list of author information is available at the end of the article © 2013 Kuroda et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Kuroda et al. Skeletal Muscle 2013, 3:5 Page 2 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 During muscle regeneration, Wnt-5a and Wnt-7a induce the function of canonical Wnt signaling in muscle. We muscle resident CD45 stem cells to undergo myogenic found that Wnt signaling is highly activated in prenatal specification and differentiation [13]. Wnt-7a also acts muscle and rapidly declines in postnatal muscle, with through the non-canonical Wnt signaling pathway to some residual activation limited to the slow myofibers. stimulate the symmetric expansion of satellite cells and Interestingly, Wnt signaling is highly activated proximal promote skeletal muscle hypertrophy [14,15]. Similarly, to motor endplates of slow myofibers. Wnt activation canonical Wnt activation was shown to induce satellite induced by stabilized β-catenin inhibits muscle differen- cell proliferation during skeletal muscle regeneration [16]. tiation and promotes slow muscle determination. At the However, contradictory results showing that activation of molecular level, canonical Wnt signaling induces BMP-4 canonical Wnt signaling is necessary to counteract Notch , which promotes expression of slow MyHC. These novel signaling to induce myogenic differentiation were also results demonstrate that interplay between Wnt and reported [17]. Furthermore, in the aged niche, elevated BMP signaling regulates skeletal myogenesis and muscle systemic Wnt molecules impede myogenic differentiation fiber type. and facilitate satellite cell fate conversion to fibroblastic cell lineages [18]. Methods Adult skeletal muscles contain heterogeneous types of Plasmids muscle fibers that can be broadly divided into slow- and Wnts cDNA subcloned from pLNCX-HA-Wnt constructs fast-twitch myofibers [19]. In the limb muscles, the slow- (kindly gifted from Dr. Jan Kitajewski) to pHAN-puro twitch myofibers express type I myosin heavy chain retrovirus vector by PCR [27,28]. Wnt-10b and TCF-4B (MyHC), while the fast-twitch myofibers express type IIa, cDNA were obtained from Thermo Fisher Scientific Open IIx and IIb MyHC isoforms [19]. During early stages Biosystems (Waltham, MA, USA). Super TOPFlash was of myofiber generation in chicks, Wnt-11 promotes fast kindly gifted from Dr. Moon [29]. The mouse BMP-4 myofiber formation, whereas Wnt-5 enhances slow myo- shRNA pLKO.1-puro (RMM3981-9595140) was obtained fiber generation [20]. Wnt-4 similarly stimulated fast from Openbiosystems. Scramble shRNA pLKO.1-puro myofiber formation in chicks [21]. Activation of β-catenin (Addgene Plasmid #1864), MD2.G (Addgene Plasmid induced myofiber hypertrophy followed by degeneration #12259) and psPAX2 (Addgene Plasmid #12260) were of fast myofibers in zebrafish [22]. In mice, depletion of obtained from Addgene (Cambridge, MA, USA). β-catenin during myogenesis in Pax7-lineage cells led to reduced slow myofibers and overall reduction of muscle C2C12 myoblasts, single myofibers and isolation of mass [23]. Interestingly, expression of stabilized β-catenin primary myoblasts in Pax7 lineage cells also led to reduced myogenesis but C2C12 cells are purchased from American Type Culture increased slow myofibers [23]. These studies suggest that Collection (ATCC) (Manassas, VA, USA) and cultured in Wnt signaling plays diverse roles in regulating slow- ver- DMEM with 10% FBS and antibiotics. Cells are induced sus fast-twitch myofiber formation during development. to differentiation upon 80% confluence by serum with- However, embryonic lethality of β-catenin deficient mice drawal (DMEM with 2% horse serum). Primary myoblasts or over-expressing mice precludes analysis of Wnt signa- were isolated from the hind limb of two- to three–month- ling in postnatal muscles. In addition, how canonical Wnt old wild type mice as previously described [27]. Single signaling regulates myofiber types has remained unclear. myofibers were isolated from flexor digitorum longus Myofibers in postnatal skeletal muscles retain an adap- (EDL) muscle by collagenase digestion. The single fibers tive capacity to switch between slow- and fast-twitch were immediately fixed in 4% paraformaldehyde (PFA), properties that largely depend on motoneuron activity permeabilized by Triton-X100, stained with primary anti- [19]. Wnt signaling also regulates the establishment and bodies for β-gal (Life Technologies, Carlsbad, CA, USA), maintenance of neuromuscular junctions that connect MyHC (MF-20, Developmental Studies Hybridoma Bank motor neurons and myofibers. In Drosophila and mice, [DSHB], Iowa, Iowa, USA), slow MyHC (DSHB) and Wnt secreted by presynaptic motoneurons interact with FITC-conjugated α-bungarotoxin (BTX) (Sigma-Aldrich, Agrin-MusK to induce assembly of postsynaptic endplates St. Louis, MO, USA). Nuclei were counter stained with (Reviewed in [24]). There is evidence that neuromuscular 4',6-diamidino-2-phenylindole (DAPI). junctions are phenotypically and functionally distinct in Fetal myoblasts were prepared from the legs of Myf5- fast and slow muscles [25,26]. Whether Wnt signaling is Cre/ROSA26-YFP embryos at E14.5. The legs are differentially activated in slow and fast myofibers is com- minced and digested by collagenase, dispase and DNaseI, pletely unknown. as previously described for satellite cells isolation [30]. In this study, we used transgenic Wnt-reporter mice After collagenase procedure, collected cells were stained and mice with constitutively activated canonical Wnt by alpha7-integrin antibody and anti-mouse IgG1- signaling specifically within skeletal muscle to investigate Alexa648. Fetal myoblast were isolated by MoFlo (Dako, Kuroda et al. Skeletal Muscle 2013, 3:5 Page 3 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Glostrup, Denmark). Fetal myoblast were isolated by tained inside a barrier facility, and experiments were per- magnetic-antibody cell sorting (MACS) (Miltenyi Biotec, formed in accordance with the University of Ottawa regu- Bergish Gladbach, Germany). The isolated cells were incu- lations for animal care and handling. bated in culture dishes at 37°C for one hour to remove ad- herent cells, and the nonadherent cells were collected. Immunofluorescence staining The isolated fetal myoblasts cultured in DMEM/F10 C2C12 cells were incubated at 37°C for two hours with medium with 20% FBS and bFGF. 10 μM BrdU, then washed with PBS(−) and fixed with 2% PFA/PBS. The fixed cells were stained with DAPI for Retrovirus and lentivirus infection 15 minutes at room temperature, washed with PBS(−), To prepare ecotropic retrovirus, Phoenix-eco packaging and refixed with 2% PFA for 5 minutes at room tem- (kindly gifted from Dr. Gally Nolan) cells were trans- perature. The refixed cells were treated with 2N HCl for fected with retrovirus vectors using GeneJuice (Novagen 20 minutes at room temperature (RT), neutralized by EMD Chemicals, Madison, WI, USA). Viral supernatants washing with 0.1 M borate buffer pH 8.5. The cells were were harvested 30 hours post transfection and used to permeablized with 0.2% Triton X-100 PBS(−), blocked infect C2C12 cells in the presence of polybrene (Sigma, with broking buffer, incubated with anti-BrdU antibody 8 mg/ml) for 12 hours. Infected C2C12 cells were then for 2 to 12 hours. After staining with the primary anti- washed twice with phosphate-buffered saline (PBS), body, cells were washed with PBS(−), stained with anti- maintained in growth media and were selected 24 hours mouse IgG1-Alexa488, washed with PBS(−) and moun- post-infection with puromycin (1.5 micro g/ml, Sigma). ted on slide glass with Dako mounting buffer. Myoblasts Lentivirus was packaged in 293T cells (ATCC CRL- were fixed in 4% PFA/PBS (−) for 5 minutes, blocked 11268) transfected with the mouse BMP-4 or scramble with 10% goat serum/PBS (−) for 10 minutes, and shRNA pLKO.1-puro plasmids in addition to the pMD2. stained with MyHC slow, MyHC fast, MyHC pan, anti- G and psPAX2 plasmids. Lentivirus was concentrated by mouse IgG1-Alexa568, anti-mouseIgG2b-Alexa648 (Life ultracentrifuge and resolved in PBS(−) for virus infec- Technologies) and DAPI (Sigma). Slides were mounted tion. Infected C2C12 and primary myoblast cells were in SlowFade Light antifade Kit Component A (Molecular then washed twice with PBS, and maintained in growth Probes) and analyzed with a Bio-Rad confocal laser scan- media. ning microscope (model MRC-1024) (Bio-Rad Labora- tories, Hercules, CA, USA). Gene expression analysis Total RNAs were prepared from C2C12 and myoblast ALP, X-gal and NADH-TR staining cells by TRIzol (Life Technologies). RNA samples were Cells were fixed 2% PFA and washed with PBS(−), then reverse transcribed using random hexamer and oligo dT stained by alkaline phosphatase (ALP) buffer (100 mM mixed primers with SuperScriptII enzyme (Life Techono- Tris–HCl pH 9.5, 100 mM NaCl, 50 mM MgCl ) logies) according to the manufacturer’s instructions. Re- containing 4.5 μl nitro-blue tetrazolium chloride (NBT) verse transcription reactions were diluted (1:10) with 10 and 3.5 μl 5-bromo-4-chloro-3-indolyl phosphate (BCIP) mM Tris, pH 8.0, yielding master samples of reverse- per 1 ml of ALP buffer at 37°C. Isolated whole embryos transcribed products. Real-time PCR reactions are previ- and tissue were fixed in 2% PFA for 3 hours at 4°C and ously described [27]. Real-time data were gathered using a permeablized with X-gal staining buffer (0.1 M phoshate system (MX4000; Agilent Technologies, Santa Clara, CA, buffer (pH 7.3), 2 mM MgCl , 0.01% sodium deoxy- USA) over 40 cycles (30 s at 90°C, 60 s at 58°C and 30 s at cholate, 0.02% Nonidet P-40) for 2 hours at 37°C and 72°C) followed by a denaturation curve from 54°C to 94°C stained by X-gal staining buffer with 0.1% X-gal, 5 mM in 30-s increments of 0.5°C to ensure amplification specifi- potassium ferricyanide and 5 mM potassium ferrocya- city. Threshold cycle (Ct) values were calculated with the nide at 37°C for 2 to 24 hours. MX4000 software (Agilent Technologies, Santa Clara, CA, The isolated muscle tissues were placed directly into USA) by using moving window aver-aging and an adap- optimal cutting temperature (OCT) compound and fro- tive baseline. Fold changes, other calculations and chart zen in deep cold isopentane with ethanol and dry ice. plotting were performed in Microsoft Excel (Redmond, The muscle tissues were cut 16 μm by cryostat and dried WA, USA). The sequence of PCR primers is listed in in room temperature. The muscle sections were incu- Additional file 1: Table S1. bated with NADH-TR staining solution (0.8 mg/ml NADH and 1 mg/ml NBT in 50 mM Tris–HCl (pH 7.6) Animal care at 37°C. The stained muscle sections were washed with Myf5-Cre [31] heterozygous mice were bred with deionized water, unbound NBT was removed by acetone ROSA26-YFP [32]. TCF-lacZ [33] and Ctnn1 exon3 floxed solution, then the sections were re-washed with deio- [34] mice were previously described. All mice are main- nized water and mounted with a coverslip. Kuroda et al. Skeletal Muscle 2013, 3:5 Page 4 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Luciferase assays The number of muscles labeled with X-GAL was re- Myoblasts in 24-well plates were transfected with the plas- duced in the forelimb and hand limb in P0 fetus mids indicated and 50 ng pRL-PGK using Lipofectamine (Figure 1E, F). In the adult, β-gal activity was undetect- (Invitrogen). Transfected cells were harvested around 24 able in EDL and TA, muscles that are predominantly hours after transfection, and luciferase activities in the cell enriched with fast type myofibers, but was readily extracts were measured according to the manufacturer's detected in part of the diaphragm (Figure 1G) and soleus instructions (Promega, Fitcburg, WI, USA) in a lumino- (Figure 1H), muscles that are known to be enriched with meter (Microplate luminometer LB96V (EG&G Berthold slow myofibers. These data indicate that canonical Wnt Technologies, Bad Wildbad, Germany). Luciferase acti- signaling is strongly activated during fetal myogenesis vities as indicated by arbitrary unit were normalized by and declined in postnatal muscles with some residual ac- sea urchin luciferase activities in each sample. All experi- tivity in slow myofibers. ments were repeated at least three times, and the averages Interestingly, the β-gal activity in postnatal slow of more than three independent experiments with stan- muscle was especially strong at neuromuscular junctions dard deviations are shown as bars [35]. (dark dotted staining patterns in the mid-belly of the muscles; Figure 1G-H). To further examine this Western blots phenomenon, we isolated single myofiber from the so- Infected C2C12 cells were grown in 60-mm dishes, leus muscles (containing about 50% slow and 50% fast washed twice with PBS and lysed in 100 mL radioim- myofibers) of Tcf-lacZ mice. The isolated single munoprecipitation assay (RIPA) buffer (50 mM Tris HCl, myofibers were fixed and stained with α-bungarotoxin pH 7.5; 150 mM NaCl; 0.5% Nonidet P-40; 0.1% (BTX) and β-gal antibodies, together with slow-MyHC deoxycholate) containing protease inhibitor cocktail or pan-MyHC antibodies (Figure 2A-D). Consistently, (Roche Applied Science, Penzberg, Germany) [35]. Cell strong β-gal immunoreactivity was detected proximal to extracts were collected and spun in a microcentrifuge at motor endplates located within BTX stained neuromus- 13,000 rpm for 5 minutes. Total proteins (5 mg) were sep- cular junctions of type I fibers (Figure 2A, B), but not in arated on 10% SDS-PAGE and transferred to Immobilon- type II myofibers (Figure 2C-D). Co-labeling whole P (EMD Millipore Corporation, Billerica, MA, USA). The mount muscle with slow and fast myosin heavy antibody membranes were probed with primary antibodies, fol- confirmed the specific activation of β-gal in the in slow lowed by horseradish peroxidase (HRP)-conjugated se- myofibers (Figure 2E, F). Overall, 97% of the β-gal condary antibodies at 1:5,000 (Bio-Rad Laboratories), and myofibers co-expressed the slow MyHC, where only 8% ™ - developed using ECL Plus (GE Healthcare, Chalfont St. of the β-gal myofibers co-expressed slow MyHC (Figure Giles, United Kingdam). Membranes were exposed to 2G). These data indicate that canonical Wnt signaling is BIOMAX film (Eastman Kodak, Rochester, NY, USA). Pri- highly activated at the neuromuscular junction area spe- mary antibodies used in this work: anti-MyoD (5.8A, BD cifically in slow myofibers of adult skeletal muscles. Bioscience [San Jose, CA, USA]), anti-myogenin (F5D, DSHB), anti-MHC (MF-20, DSHB), anti-GAPDH (6C5, Canonical Wnt signaling promotes formation of slow Life Technologies) and anti-.alpha;-tublin (Sigma-Aldrich). myofibers in vivo To confirm the role of canonical Wnt signaling in Results muscle fiber type specification in vivo, we took advan- lox(ex3) Canonical Wnt signaling is activated during fetal tage of the Ctnnb transgenic mice in which the myogenesis and reduced in adult muscle exon 3 of β-catenin (Ctnnb) gene flanked by LoxP sites As the first step to investigate the function of Wnt sig- [34]. The exon 3 encodes serine and threonine residues naling in myogenesis, we used the TCF-lacZ transgenic that are normally phosphorylated GSK3β, leading to the reporter mouse to examine the activity of the canonical proteasomal degradation of β-catenin. Upon Cre- ΔEx3 Wnt signaling pathway in embryonic and adult muscles. mediated excision of Ctnnb exon 3, β-catenin is The promoter of the LacZ transgene (encoding β- prevented from degradation (stabilized) and, therefore, galactosidase, β-gal) contains multimerized TCF binding constitutively active. ΔEx3 sites [33], a key downstream effector of canonical Wnt We first used Myf5-Cre to induce β-catenin expres- signaling. We analyzed β-gal activity, in embryonic and sion in myogenic progenitor cells. Myf5 is an early myo- adult muscles by X-GAL staining (Figure 1). β-gal activ- genic commitment marker during embryonic myogenesis lox(ex3) ity was detected in many muscles in E14.5 embryos [36]. Myf5-Cre/Ctnnb mice die at E15.5 with (Figure 1A). Among those labeled, β-gal activity was par- extremely reduced muscle mass (data not shown), thus ticularly intense in both forelimb and hind limb muscles precluding further analysis of myofiber types. We next ΔEx3 (Figure 1B), ventral body wall muscles (Figure 1C), dor- used MCK-Cre to drive β-catenin expression only in sal spinotrapezius (Figure 1D) and intercostal muscles. differentiated muscle cells. As expected, we detected Kuroda et al. Skeletal Muscle 2013, 3:5 Page 5 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Figure 1 Activation of canonical Wnt signaling in embryonic, neonatal and adult skeletal muscles. Tcf-lacZ reporter mice were used to report activation of TCF promoter, the nuclear target of canonical Wnt signaling. X-gal staining (in blue) was used to reveal LacZ (b-gal) activity. (A) Whole mount staining of an E14.5 embryo. (B) Forelimbs and hind limbs at E14.5. (C) Ventral and (D) dorsal view of an E14.5 embryo revealing intensive staining in some muscles. (E) Hind limb and (F) forelimb at P0 (postnatal Day 0). (G) Diaphragm and (H) soleus muscles of adult mouse showing staining in a subset of myofibers with intensive signals at the neuromuscular junction area. increased activation of canonical Wnt signaling in vivo in oxidative myofibers was increased in wild type TA muscles lox(ex3) transgenic mice carrying MCK-Cre, Ctnnb and TCF- (n = 3, 28.3 ± 7.7%) compared relative to the mutant litter- LacZ alleles. β-gal activity was detectable in TA muscles of mate (Figure 3C-E). As slow muscles are known to con- lox(ex3) the MCK-Cre/Ctnnb /TCF-LacZ mice, but not the tain mainly oxidative myofibers and fast muscles are lox(ex3) Ctnnb /Tcf-lacZ littermate controls (Figure 3A, B). mainly low oxidative and glycolytic [19], these data pro- In addition, we analyzed the oxidative activity of skeletal vide in vivo evidence that canonical Wnt signaling pro- muscle myofibers using NADH-tetrazolium (NADH-TR) motes slow myofiber phenotype in the postnatal skeletal staining. Strikingly, the high and middle oxidative muscle. myofibers were increased in TA muscles of MCK-Cre/ lox(ex3) Ctnnb mice (n = 3, 65.4 ± 8.4% and 34.6 ± 8.4%, re- Canonical Wnt signaling inhibits proliferation of C2C12 spectively) compared to the same muscle of wild type lit- and primary myoblasts termates (n = 3, 52.7 ± 5.3% and 19.0 ± 3.8%, respectively) To identify the Wnt molecules that activate the canon- (Figure 3C-E). By contrast, the number of the low ical Wnt signaling pathway in muscle, we co-transfected Kuroda et al. Skeletal Muscle 2013, 3:5 Page 6 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Figure 2 Activation of canonical Wnt signaling in adult slow myofibers. (A-D) Soleus myofibers from adult Tcf-lacZ mice were stained with antibodies for pan-myosin heavy chain (White) or slow myosin heavy chain (White), and β-gal (Red) and FITC conjugated BTX (Green). (E-F) Whole mount muscles co-stained with antibodies for slow (Blue) and fast (Red) myosin heavy together with beta-gal (Green) and BTX (Red). (G) Percentage of β-gal positive (n = 69) and negative (n = 60) myofibers that co-express slow myosin heavy chain. candidate Wnt plasmids with the TCF/LEF reporter Super We next examined the function of both canonical and TopFlash. We found that Wnt-1 and Wnt-3a strongly ac- non-canonical Wnt signaling in cultured C2C12 myo- tivated (>400 times increase in luciferase activity), Wnt-2 blasts using retrovirus expressing various Wnts. Com- and Wnt-10b moderately activated (>20 times increase in pared to mock controls, C2C12 cells overexpressing luciferase activity), and Wnt-2b and Wnt-4 weakly acti- canonical Wnts (Wnt-1, -2, -2b, -3a, -10b) had a signifi- vated (>2 times increase in luciferase activity) the canon- cantly decreased cell number after 96 h in culture ical Wnt reporter (Additional file 2: Figure S1A). Other (Additional file 2: Figure S1B). By contrast, C2C12 cells Wnts (Wnt-5a, Wnt-5b, Wnt-6, Wnt-7a, Wnt-7b, Wnt- overexpressing non-canonical Wnts (Wnt-5a, -5b, -6, -7b, 10a, Wnt-11) had no effect on the activation of the Super -11) had a moderately increased cell number compared to TopFlash reporter (Additional file 2, Figure S1A). We, the control treated cells (Additional file 2: Figure S1B). therefore, used Wnt-1 and Wnt-3a to activate the canon- A reduced cell number in C2C12 cells overexpressing ical Wnt signaling in the following studies. Wnt-1, -2, -2b, -3a and -10b indicates that canonical Kuroda et al. Skeletal Muscle 2013, 3:5 Page 7 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 decreased in the presence of Wnt-3a protein. These re- sults indicate that canonical Wnt signaling suppressed the proliferation of C2C12 cells and adult primary myoblasts. To test if Wnt signaling has any effect on myogenic differentiation, we induced C2C12 myoblasts to differen- tiate. Upon serum withdrawal, control myoblasts exited the cell cycle and fused to form myotubes with uniform morphology (Additional file 4: Figure S3A). By contrast, Wnt-1 and Wnt-3a expressing C2C12 cells formed very few myosin heavy chain (MyHC) positive myotubes (Additional file 4: Figure S3B-C). In addition, these myotubes were morphologically abnormal: they were short and chubby (Additional file 4: Figure S3B-C). As previously reported, non-canonical Wnt7a overexpres- sion led to the formation of large myotubes resembling the muscle hypertrophy phenotype (Additional file 4: Figure S3D). We further examined by Western blotting the expression of two myogenic differentiation markers, myogenin and MyHC. Compared to the control, Wnt-1 and -3a robustly inhibited the expression of Myogenin and MyHC at 48 hours, 72 hours and 96 hours post- induction of differentiation (Additional file 4: Figure Figure 3 Muscle-specific constitutive activation of canonical S3E). Intriguingly, the non-canonical Wnt-7a not only Wnt signaling promotes oxidative myofiber phenotype. The lox(ex3) MCK-Cre/Ctnnb mice were used to express constitutively increased Myogenin and MyHC protein levels, but also Δex3 active β-catenin , which mimics canonical Wnt signaling. The induced their earlier expression at 24 hours (Additional MCK promoter-drive Cre expression limits Wnt activation only in file 4: Figure S3E). However, recombinant Wnt-3a pro- mature skeletal muscles. (A-B) X-Gal staining (blue signal) of whole lox(ex3) tein at 50 ng/ml had no effect on the differentiation of mount TA muscles of Ctnnb /Tcf-LacZ (WT) and MCK-Cre lox(ex3) neither C2C12 myoblasts (not shown) nor fetal primary /Ctnnb /Tcf-LacZ (Mut) mice. Blue signal, indicative of canonical Wnt signaling, is only detectable in the Mut TA muscles (A and B myoblasts (see Figure 4H). Thus, these data indicate that represent the dorsal and ventral view of the same muscles). (C-D) prolonged high-level constitutive activation of Wnt-1 NADH-TR staining of TA muscle sections from the WT (C) and Mut and Wnt-3a in C2C12 cells suppresses myogenic (D)mice. (E) Percentage of high oxidative (black), middle oxidative (gray) differentiation. and low oxidative (white) myofibers of WT and Mut mice (n = 3, each). Scale bar: 1 mm. Canonical Wnt signaling activates BMP signaling in C2C12 myoblasts Wnt signaling inhibits cell proliferation. To confirm this, The reduced myogenic differentiation of Wnt-1 and we analyzed the expression of Ki-67 (Additional file 2: Wnt-3a overexpressing C2C12 myoblasts prompted us Figure S1 C-E), a nuclear antigen specifically expressed to analyze the alternative differentiation fate of these in S, G2 and M phase cells. The Wnt-1 and -3a ex- cells. Previous studies show that C2C12 cells can also pressing C2C12 cells had decreased Ki-67 expression differentiate into osteogenic lineage [37], which express (Additional file 2: Figures S1 D-E, and 4I). In addition, ALP. Surprisingly, Wnt-1, Wnt-3a and Wnt-10b overex- we examined incorporation of BrdU, a thymidine analog pression strongly induced ALP immunochemical signals that is incorporated into proliferating cells during the S (Additional file 5: Figure S4A-F). By contrast, expression phase. Proliferating C2C12 cells were incubated with of non-canonical Wnt-5a and Wnt-7a did not affect BrdU for one hour and then fixed for BrdU staining ALP expression (Additional file 5: Figure S4D-E). Quan- (Additional file 2: Figure S1F-H). Consistently, BrdU in- titative analysis confirms that Wnt-1 and Wnt-3a corporation was reduced in the Wnt-1 and -3a express- increased ALP enzyme activity by 15- and 30-fold, re- ing C2C12 cells (Additional file 2: Figure S1G-H, J). spectively (Additional file 5: Figure S4G). Furthermore, we analyzed proliferation of skeletal To examine if Wnt-induced ALP activity was depen- muscle-derived primary myoblasts cultured with recom- dent on β-catenin, which mediates canonical Wnt sig- binant Wnt-3a protein (50 ng/ml) for 24 hours. Both Ki- naling, we transduced C2C12 cells with a dominant 67 expression (Additional file 3: Figure S2A-C) and negative TCF-4b (DN-TCF-4b) that lacks the β-catenin BrdU uptake (Additional file 3: Figure S2D-F) were binding site. As expected, DN-TCF-4b suppressed both Kuroda et al. Skeletal Muscle 2013, 3:5 Page 8 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Figure 4 Canonical Wnt signaling induces BMP-4 in C2C12 myoblasts. (A) qPCR analysis of the relative expression of various genes in C2C12 myoblasts overexpressing Wnt-1 and Wnt-3a under growth condition. (B) BMP-4 knockdown reduced the expression of alkaline phosphatase (Akp2) but not the Axin2 gene. (C-F) BMP-4 shRNA rescued the myogenic differentiation of Wnt-1 and Wnt-3a infected C2C12 cells. (C-D) Wnt-1 and Wnt-3a overexpressing myoblasts treated with Scramble shRNA; (E-F) Wnt-1 and Wnt-3a overexpressing myoblasts treated with BMP-4 shRNA. Black signaling is myosin heavy chain antibody staining reacted with 3, 3'-diaminobenzidine (DAB) substrate. (G) Western blotting showing myosin heavy chain protein expression after BMP-4 shRNA treatment. (H) qPCR analysis of myosin heavy chain gene expression. (I-L) BMP-4 antagonist Noggin-Fc dose-dependently rescued the myogenic differentiation of Wnt-3a infected C2C12 cells. The myosin heavy chain was labeled in black and nuclei were labeled in white. Wnt-1 and Wnt-3a induced ALP activity in C2C12 cells expression was increased by more than five-fold in Wnt- (Additional file 5: Figure S4H). These data indicate that 1 and Wnt-3a overexpressing C2C12 cells (Figure 4A). canonical Wnt signaling induces ALP expression via a In comparison, Myf5 and MyoD expression was not af- β-catenin/TCF-dependent pathway in C2C12 cells. fected by Wnt-1 and Wnt-3a (Figure 4A). To ensure that That canonical Wnt signaling induces osteogenic ALP canonical Wnt signaling is involved in the induction of expression suggests a potential interaction between Wnt BMP-4, we examined Axin2, a transcriptional target of and BMP signaling pathways. We first examined BMP-4 β-catenin and canonical Wnt signaling [38]. Wnt-1 and gene expression using quantitative RT-PCR (qPCR) Wnt-3a overexpression led to over 40X increase in the given its role in osteogenesis. Indeed, BMP-4 mRNA expression Axin2 (Figure 4A). Kuroda et al. Skeletal Muscle 2013, 3:5 Page 9 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Figure 5 Wnt-3a induces slow myosin heavy chain expression in fetal myoblasts via BMP-4. Fetal myoblasts were isolated by fluorescence- activated cell sorting (FACS) from E14.5-15.5 embryos of Myf5-Cre/Rosa26-YFP mice. FACS isolated fetal myoblasts were cultured and induced to differentiate for three days. (A-D) Immunofluorescence showing slow myosin heavy chain (green) expression in myotubes treated with vehicle control medium or Wnt-3a recombinant protein (50 ng/ml). (E) Relative expression of myosin heavy chain and BMP-4 genes based on qPCR analysis. (F) Western blotting showing slow myosin heavy chain protein expression. (G-J) Immunofluorescence showing slow- (green) and pan- (red) myosin heavy chain expression in myotubes treated with vehicle control medium, Wnt-3a, and BMP-4 recombinant protein at concentration shown. (K) qPCR analysis showing relative expression of myosin heavy chain isoform genes. Scale bars: 40 μm. We next asked if BMP-4 is necessary for pro- myotubes in the C2C12 cells overexpressing Wnt-1 osteogenic effect of canonical Wnt signaling. We used (Figure 4C) and Wnt-3a (Figure 4C). Knockdown of BMP- lentiviral shRNA mediated knockdown of BMP-4 in 4 remarkably increased the numbers of MyHC positive C2C12 cells. This approach resulted in nearly 80% reduc- myotubes in Wnt-1 (Figure 4E) and Wnt-3a (Figure 4F) tion of BMP-4 transcripts (Figure 4B). Importantly, BMP- expressing C2C12 cells. Consistent with this observation, 4 knockdown reduced Wnt-1 and Wnt-3a induced Akp4 BMP-4 shRNA rescued MyHC protein expression in Wnt- (ALP gene) expression by more than 70% (Figure 4B). By 1 and Wnt-3a overexpressing C2C12 cells (Figure 4G). contrast, Axin2 mRNA levels were not decreased follow- Moreover, BMP-4 shRNA increased the mRNA levels of ing knockdown of BMP-4 (Figure 4B), suggesting that Myogenin, MyHC-IIa, MyHC-IIb and MyHC-I (Figure 4H). BMP-4 signaling does not affect canonical Wnt signaling. To confirm the above observations, we used recombinant We further investigated if the anti-myogenic effect of ca- Noggin protein, an antagonist of BMP, to block BMP activ- nonical Wnt signaling is mediated by BMP-4 using the ity. Noggin dose-dependently increased the number of same shRNA knockdown approach. In the control groups MyHC positive myotubes in Wnt-3a expressing C2C12 (scrambled shRNA), there were only a few MyHC positive cells (Figure 4I-L). These data provide compelling evidence Kuroda et al. Skeletal Muscle 2013, 3:5 Page 10 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 that canonical Wnt signaling inhibits myogenic differen- BMP-4 to induce slow MyHC expression during embry- tiation through inducing BMP-4 signaling. onic myoblast differentiation. Discussion Canonical Wnt signaling induces slow MyHC expression In this study, we use genetic, cell culture and molecular through BMP-4 biology approaches to dissect the function of canonical To directly examine the function of canonical Wnt signa- Wnt signaling in myogenic differentiation and skeletal ling in muscle fiber type specification, we isolated em- myofiber types. We show that the canonical Wnt signal- bryonic myoblasts from E14.5 embryos by FACS. We ing is most active during perinatal myogenesis and only employed positive selection for Myf5 and α7-integrin ex- activated in slow myofibers with high activity at the pression of myogenic cells from Myf5-Cre/ROSA-YFP em- neuromuscular junction area in mature muscles. Con- bryos (Additional file 6: Figure S5). The purity of isolated stitutive activation of β-catenin, the canonical Wnt sig- fetal myoblasts was confirmed by immunostaining for naling effector, leads to impaired myogenesis and an Pax7 and desmin (Additional file 6: Figure S5). Embryonic increased proportion of oxidative myofibers in the post- myoblasts were cultured for one day before being induced natal muscles. Importantly, Wnt-1 and Wnt-3a mediated to differentiate with or without the addition of Wnt-3a downstream signaling activates BMP-4, which inhibits protein (50 ng/ml). In the absence of Wnt-3a, embryonic the overall proliferation of myoblasts and promotes myoblasts differentiated into myotubes that expressed fast myogenic differentiation towards slow muscle pheno- MyHC (Figure 5A), but not slow MyHC (Figure 5C) in type. These results establish a novel interaction between agreement with previous studies [39]. By contrast, Wnt-3a Wnt and BMP signaling that regulates muscle fiber type treated fetal myoblasts differentiated into myotubes specification and maintenance. that expressed both fast (Figure 5B) and slow MyHC The TCF-LacZ reporter mouse has been widely used (Figure 5D). in reporting activation of canonical Wnt signaling in To confirm this result, we analyzed MyHC-I mRNA various tissues/cells [11,40]. Strong LacZ expression in expression by qPCR. The MyHC-I mRNA level was up- specific muscles during embryonic and fetal myogenesis regulated 5.5-fold by Wnt-3a compared to control ve- indicates activation of canonical Wnt signaling. Interes- hicle treatment (Figure 5E). Wnt3a also robustly induced ting, several muscles (spinotrapezius, body wall muscle the expression of slow MyHC-I at the protein level and diaphragm) with high β-gal activity are known to be (Figure 5F). These results indicate that canonical Wnt enriched with slow myofibers [19]. These results suggest signaling induces slow MyHC expression in fetal a role of Wnt signaling in slow muscle generation and myoblasts. maintenance. Our in vivo results are consistent with pre- As canonical Wnt signaling induced BMP-4, we fur- vious studies in chick and fish embryos, in which cano- ther examined the role of BMP-4 in muscle fiber type nical Wnt signaling was shown to promote slow muscle specification. Consistent with our previous results in fate [20-22]. C2C12 cells (Figure 4A), Wnt-3a treatment of embry- Our analysis of canonical Wnt signaling in adult ma- onic myoblasts induced a seven-fold increase in BMP-4 ture muscles reveal several interesting points. First, β-gal mRNA expression (Figure 5E). Next, we added recom- activity is only detectable in muscles known to contain binant BMP-4 to fetal myoblast cultures during slow myofibers. This result confirms our observation in differentiation. In the control treated with vehicle the developing embryonic muscles. Second, in contrast medium, newly formed MyHC myotubes seldom to embryonic muscle, where β-gal activity is evenly dis- expressed slow MyHC after three days of differentia tributed within myofibers, the highest β-gal activity was tion (Figure 5G). In the presence of 2.5 to 5 ng/ml within the slow myofibers was proximal to the motor BMP-4, slow-MyHC myotubes were abundantly visible endplate (Figure 1G-H). This observation suggests that (Figure 5I-J). The level of slow MyHC immunofluores- whereas in embryonic muscle the Wnt molecules are cence induced by 5 ng/ml BMP-4 is similar to that in- released from surrounding tissues [5], Wnt signaling duced by 50 ng/ml Wnt-3a (Figure 5H), suggesting in adult slow myofibers is most likely initiated by Wnt BMP-4 more potently induces slow MyHC expression. molecules from motor neurons that innervate these Western blotting showed that the slow MyHC protein myofibers. In support of the notion that the motor expression level was indeed increased in the presence of neuron supplies Wnt molecules, we found that β-gal BMP-4 (Figure 5F). In addition, qPCR analysis indicated immunoreactivity was no longer detectable in slow that BMP-4 not only induced the slow MyHC-I gene ex- myofibers after three days of suspended culture in vitro pression, but also robustly suppressed the fast MyHC-IIb in the absence of neural innervation (data not shown). gene expression (Figure 5K). Collectively, these results Previous studies demonstrate that Wnt molecules indicate that canonical Wnt signaling acts through released by motor neurons play key roles in the Kuroda et al. Skeletal Muscle 2013, 3:5 Page 11 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 development of neuromuscular junctions. Specifically, Interestingly, embryonic and fetal myoblasts seem to interaction of Wnt and LRP is necessary for clustering have different responses to canonical Wnt signaling [23], of postsynaptic acetylcholine receptors (AchR) [24]. Our suggesting that the role of Wnt signaling, even in the new results demonstrate that Wnt signaling is further same cell lineage, is also context dependent. Consistent required for the maintenance of neuromuscular junction with this notion, we show that the growth and differenti- in slow myofibers. Future study is needed to examine ation of fetal primary myoblasts are not inhibited by re- the functional significance of Wnt signaling in slow ver- combinant Wnt-3a protein in culture. The osteogenic sus fast muscle fibers and identify the Wnt molecules re- fate choice of Wnt-1 and Wnt-3a overexpressing C2C12 leased by slow and fast motor neurons. cells is in line with a recent report demonstrating fibro- Using Cre-inducible transgenic mice that express sta- blastic lineage differentiation of satellite cells in response bilizes β-catenin, we investigated the role of canonical to high level of systemic Wnt molecules [18]. Thus, the Wnt signaling in embryonic myogenesis and postnatal observed effect of Wnt-1 and Wnt-3 in myogenic cell muscle maintenance. When Myf5-Cre is used as the proliferation and differentiation is largely consistent with driver mouse, which is expressed in embryonic myogenic information in the literature. progenitors, we detected abnormal muscle development We discovered a novel interaction between Wnt and and perinatal lethality. This observation is consistent to BMP signaling in myoblasts. Bone morphogenetic pro- recent studies using Pax7-Cre or Myogenin-Cre to teins (BMPs) are multi-functional proteins belonging to stabilize β-catenin, which also results in lethality at P0 the transforming growth factor beta (TGFβ) superfamily. [23,41]. In these studies, constitutive activation of β- In zebrafish and frogs, BMP signaling inhibits the diffe- catenin in the myogenic progenitors and committed rentiation of muscle precursors in the dermomyotome myocytes resulted in a shift of fetal myofibers to slow and controls the number of myogenic cells [42,43]. Du- muscle phenotype, and reduced myofiber size. However, ring late myogenesis of mice, BMP signaling regulates the the perinatal lethality of the Myf5-Cre, Myogenin-Cre and number of fetal myoblasts and satellite cells [44]. This ac- Pax7-Cre drive β-catenin activation precludes analysis of tion is through preventing the premature activation of Lox(ex3) postnatal muscles. Using the Mck-Cre/ctnnb mice, MyoD while maintaining Pax3 expression. Therefore, we found that adult muscles indeed have higher canonical BMPs may function to establish a sufficient number of Wnt activity based on the TCF-LacZ reporter assay. This myogenic progenitors before terminal differentiation. verifies the utility of our mouse model. Our cell culture results indicate that Wnt signaling in- Importantly, we found the adult fast (TA) muscles duces BMP4, and BMP4 inhibition rescues the inhibitory exhibited features of slow muscle phenotype as increased effect of Wnt-1 and Wnt-3a on myogenesis. This result is oxidative capacity. We also examined myosin heavy in line with the above results in vivo.We further identify chain expression by immunohistochemistry but did not an unexpected role for BMP-4 in promoting slow muscle find any overt changes in myofiber types (data not fate during fetal myogenesis. It is important to mention shown). This result suggests that although Wnt signaling that low concentrations (1 to 5 ng/ml) of BMP-4 protein affects metabolic properties in the adult muscles, it is were used in our study. Non-physiological, high concen- not sufficient to switch myosin heavy chain expression trations of BMP will probably generate completely diffe- in the adult. This is expected since other factors, such as rent effects [45]. Future studies should illustrate how BMP hormones and neural activity, can also influence myosin signaling regulates myosin gene expression. Lox(ex3) expression [19]. Together, our Mck-Cre/ctnnb mo- Interestingly, in Drosophila larval neuromuscular junc- del bypasses the premature lethality and provides novel tions, retrograde BMP signaling controls synaptic growth insights of canonical Wnt signaling in regulating the oxi- [46]. The muscle-derived BMP modulates cytoskeletal dy- dative capacity of adult muscles. namics and structural changes at presynaptic terminals. The observation that C2C12 cells overexpressing Wnt-1 This forms a feedback system in which canonical Wnt and Wnt-3a exhibited reduced proliferation and myo- molecules secreted from motor neurons not only induce genic differentiation is quite intriguing. It could suggest formation of neuromuscular junctions, but also activate BMP-4 expression in the muscle. The muscle derived that canonical Wnt/β-catenin inhibits the proliferation and differentiation of myogenic cell lineages. This pos- BMP-4 subsequently promotes development of presynaptic sibility would explain the reduced muscle mass phe- motor neuron terminals. Indeed, β-catenin stabilization in skeletal muscles (not limited to the neuromuscular junc- notypes of the Myf5-Cre, Myogenin-Cre and Pax7-Cre induced stabilized β-catenin mice [23,41]. Alternati- tion area) results in increased motor axon number and ex- vely, the result may also suggest that expression of cessive intramuscular nerve defasciculation and branching [41]. Taken together, our experiments have identified a Wnt-1 and Wnt-3a in the cell, independent of Frizzled receptor activation, is detrimental to cell growth and novel interaction between canonical Wnt and BMP signa- differentiation. ling that plays a role in myofiber type specification. Kuroda et al. Skeletal Muscle 2013, 3:5 Page 12 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Conclusion bFGF: Basic fibroblast growth factor; PFA: Paraformaldehyde; PGK: Phosphoglycerate kinase; GSK: Glycogen synthase kinase; MCK: Muscle Our study demonstrates that canonical Wnt-signaling creatine kinase; BrdU: Bromodeoxyuridine. controls the development of skeletal muscles via BMP-4 expression. High concentrations of BMP-4 have been Competing interests The authors declare no competing interests. previously established to inhibit myogensis and induce osteogenesis. We found that isolated fetal myoblasts do Authors’ contributions not normally form slow myofibers during myogenic dif- KK and MAR designed the research and wrote paper. SK performed the ferentiation in vitro. Strikingly, canonical Wnt-signaling histology and tissue staining, and helped with paper writing. TMM provided the Ctnn1 exon3 floxed mice. All authors read and approved the final induced low level BMP-4 expression that act to induce manuscript. slow myofibergenesis. Therefore, we conclude that ca- nonical Wnt and BMP signaling plays a hitherto un- Acknowledgments We thank Dr. Daniel Dufort for TCF-lacZ mice, Dr. Randall Moon and Dr. appreciated role in myofiber type specification during Valerie Wallace for DNA constructs. We thank members of the Rudnicki fetal myogenesis. laboratory for their technical assistance and helpful discussions, and Dr. Makoto Sato for critical reading of the manuscript. KK was supported by a Postdoctoral Fellowship from Training Program in Regenerative Medicine in Additional files Canada. Additional file 1: Table S1. The sequence of PCR primers for qPCR Author details analysis is listed. Sprott Center for Stem Cell Research, Ottawa Hospital Research Institute, 501 Smyth Road, Ottawa, ON K1H 8L6, Canada. Present Address: Additional file 2: Figure S1. Canonical Wnt signaling inhibits the Department of Animal Sciences, Purdue University, 174B Smith Hall, 901 W growth and proliferation of C2C12 myoblast cells. (A) Relative luciferase State St, West Lafayette, IN 47907, USA. Department of Pharmacology, Kyoto activity of C2C12 cells overexpressing various Wnt genes together with University Graduate School of Medicine, Konoe, Yoshida, Sakyo, Kyoto the SuperTop Flash reporter plasmid. (B) Graph of C2C12 cells transduced 606-8501, Japan. Present address: Division of Cell Biology and Neuroscience, with retroviral Wnt plasmids. (C-E) The Ki-67 antibody staining and (F-H) Department of Morphological and Physiological Sciences, Faculty of Medical BrdU incorporation of control, Wnt-1 and Wnt-3a transduced cells. (I) Sciences, University of Fukui, 23-3, Matsuoka-Shimoaizuki, Eiheiji, Fukui Percentage of Ki-67 positive cells in control, Wnt-1 and Wnt-3a 910-1193, Japan. transduced cells. (J) Percentage of BrdU incorporated cells in Control, Wnt-1 and Wnt-3a retrovirus infected C2C12 cells. Received: 29 October 2012 Accepted: 15 February 2013 Additional file 3: Figure S2. Wnt-3a inhibits proliferation of adult Published: 5 March 2013 primary myoblasts. (A-B) Ki-67 antibody staining of primary myoblasts treated with vehicle control (A) and Wnt-3a recombinant protein (50 ng/ References ml) (B). (C) Percentage of Ki67 positive cells. (D-E) BrdU incorporation of 1. Biressi S, Molinaro M, Cossu G: Cellular heterogeneity during vertebrate primary myoblast cells treated with control (D) and Wnt-3a protein (E). (F) skeletal muscle development. Dev Biol 2007, 308:281–293. Percentage of cells incorporated BrdU. 2. Buckingham M, Vincent SD: Distinct and dynamic myogenic populations Additional file 4: Figure S3. Overexpression of Wnt-1 and Wnt-3a in the vertebrate embryo. Curr Opin Genet Dev 2009, 19:444–453. inhibits the myogenic differentiation of C2C12 myoblasts. (A-D) 3. Christ B, Ordahl CP: Early stages of chick somite development. Anat Immunofluorescence showing sarcomeric myosin heavy chain protein Embryol 1995, 191:381–396. expression in C2C12 cells transduced with retrovirus overexpressing (A) 4. Cossu G, Borello U: Wnt signaling and the activation of myogenesis in Mock, (B) Wnt-1, (C) Wnt-3a and (D) Wnt-7a after four days of induced mammals. EMBO J 1999, 18:6867–6872. myogenic differentiation. (E) Western blotting analysis showing 5. von Maltzahn J, Chang NC, Bentzinger CF, Rudnicki MA: Wnt signaling in expression of myosin heavy chain (MyHC), myogenin and GAPDH in myogenesis. Trends Cell Biol 2012, 22:602–609. control, Wnt-1, Wnt-3a and Wnt-7a overexpression C2C12 myoblasts at 6. Abu-Elmagd M, Robson L, Sweetman D, Hadley J, Francis-West P, different time points after induced to differentiate. GAPDH: Munsterberg A: Wnt/Lef1 signaling acts via Pitx2 to regulate somite Glyceraldehyde 3-phosphate dehydrogenase. Scale bar: 100 μm. myogenesis. Dev Biol 2010, 337:211–219. Additional file 5: Figure S4. Canonical Wnts (−1, -3a and 10b) induces 7. Tajbakhsh S, Borello U, Vivarelli E, Kelly R, Papkoff J, Duprez D, Buckingham alkaline phosphatase (ALP) activity in C2C12 cells. (A-F) ALP staining M, Cossu G: Differential activation of Myf5 and MyoD by different Wnts (Purple signal) of C2C12 cells transduced retrovirus expressing empty in explants of mouse paraxial mesoderm and the later activation of vector (Mock) or Wnt vectors as indicated. (G) Relative ALP activity. (H) myogenesis in the absence of Myf5. Development 1998, 125:4155–4162. The dominant negative (DN) TCF-4 suppressed ALP activity in Wnt-1 and 8. Borello U, Berarducci B, Murphy P, Bajard L, Buffa V, Piccolo S, Buckingham Wnt-3a expressing C2C12 cells. M, Cossu G: The Wnt/beta-catenin pathway regulates Gli-mediated Myf5 expression during somitogenesis. Development 2006, 133:3723–3732. Additional file 6: Figure S5. Isolation of fetal myoblasts from E14.5 9. Ikeya M, Takada S: Wnt signaling from the dorsal neural tube is required embryos of Myf5-Cre/ROSA26-YFP mice by fluorescence activated cell for the formation of the medial dermomyotome. Development 1998, sorting (FACS). (A-C) YFP expression in whole embryo (A), forelimb (B) 125:4969–4976. and hindlimb (C) of E14.5-15.5 Myf5-Cre/ROSA26-YFP embryo. (D) 10. Otto A, Schmidt C, Patel K: Pax3 and Pax7 expression and regulation in Strategy for isolating fetal myoblasts by FACS. Whole limbs of Myf5-Cre the avian embryo. Anat Embryol (Berl) 2006, 211:293–310. /ROSA26-YFP embryos were minced and digested by collagenase and 11. Aoki K, Taketo MM: Tissue-specific transgenic, conditional knockout and dispase. Single cells were stained with alpha7-integrin antibody and knock-in mice of genes in the canonical Wnt signaling pathway. Methods selected by YFP and alpha7-integrin expression. The sorted YFP and Mol Biol 2008, 468:307–331. alpha7-integrin double positive cells were stained with antibody for Pax7 12. van Amerongen R, Berns A: Knockout mouse models to study Wnt signal and desmin, makers of fetal myoblasts (nuclei were counterstained by transduction. Trends Genet 2006, 22:678–689. DAPI in blue). 13. Seale P, Polesskaya A, Rudnicki MA: Adult stem cell specification by Wnt signaling in muscle regeneration. Cell Cycle 2003, 2:418–419. Abbreviations 14. Le Grand F, Jones AE, Seale V, Scime A, Rudnicki MA: Wnt7a activates the BMP: Bone morphogenetic protein; TCF: T-cell factor; LEF: Lymphoid planar cell polarity pathway to drive the symmetric expansion of enhancer-binding factor; Pax: Paired box protein; FBS: Fetal bovine serum; satellite stem cells. Cell Stem Cell 2009, 4:535–547. Kuroda et al. Skeletal Muscle 2013, 3:5 Page 13 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 15. von Maltzahn J, Bentzinger CF, Rudnicki MA: Wnt7a-Fzd7 signalling differentiation of the pluripotent mesenchymal precursor cell line directly activates the Akt/mTOR anabolic growth pathway in skeletal C2C12. J Biol Chem 1998, 273:1872–1879. muscle. Nat Cell Biol 2012, 14:186–191. 38. Jho EH, Zhang T, Domon C, Joo CK, Freund JN, Costantini F: Wnt/beta- 16. Otto A, Schmidt C, Luke G, Allen S, Valasek P, Muntoni F, Lawrence-Watt D, catenin/Tcf signaling induces the transcription of Axin2, a negative Patel K: Canonical Wnt signalling induces satellite-cell proliferation regulator of the signaling pathway. Mol Cell Biol 2002, 22:1172–1183. during adult skeletal muscle regeneration. J Cell Sci 2008, 121:2939–2950. 39. Hagiwara N, Yeh M, Liu A: Sox6 is required for normal fiber type 17. Brack AS, Conboy IM, Conboy MJ, Shen J, Rando TA: A temporal switch differentiation of fetal skeletal muscle in mice. Dev Dyn 2007, 236:2062–2076. from Notch to Wnt signaling in muscle stem cells is necessary for 40. Barolo S: Transgenic Wnt/TCF pathway reporters: all you need is Lef? normal adult myogenesis. Cell Stem Cell 2008, 2:50–59. Oncogene 2006, 25:7505–7511. 18. Brack AS, Conboy MJ, Roy S, Lee M, Kuo CJ, Keller C, Rando TA: Increased 41. Liu Y, Sugiura Y, Wu FF, Mi WT, Taketo MM, Cannon S, Carroll T, Lin WC: Wnt signaling during aging alters muscle stem cell fate and increases beta-Catenin stabilization in skeletal muscles, but not in motor neurons, fibrosis. Science 2007, 317:807–810. leads to aberrant motor innervation of the muscle during neuromuscular development in mice. Dev Biol 2012, 366:255–267. 19. Schiaffino S, Reggiani C: Fiber types in mammalian skeletal muscles. 42. Patterson SE, Bird NC, Devoto SH: BMP regulation of myogenesis in Physiol Rev 2011, 91:1447–1531. zebrafish. Dev Dyn 2010, 239:806–817. 20. Anakwe K, Robson L, Hadley J, Buxton P, Church V, Allen S, Hartmann C, 43. Daughters RS, Chen Y, Slack JM: Origin of muscle satellite cells in the Harfe B, Nohno T, Brown AM, Evans DJ, Francis-West P: Wnt signalling Xenopus embryo. Development 2011, 138:821–830. regulates myogenic differentiation in the developing avian wing. 44. Wang H, Noulet F, Edom-Vovard F, Tozer S, Le Grand F, Duprez D: Bmp Development 2003, 130:3503–3514. signaling at the tips of skeletal muscles regulates the number of fetal 21. Takata H, Terada K, Oka H, Sunada Y, Moriguchi T, Nohno T: Involvement of muscle progenitors and satellite cells during development. Dev Cell 2010, Wnt4 signaling during myogenic proliferation and differentiation of 18:643–654. skeletal muscle. Dev Dyn 2007, 236:2800–2807. 45. Nakashima A, Katagiri T, Tamura M: Cross-talk between Wnt and bone 22. Tee JM, van Rooijen C, Boonen R, Zivkovic D: Regulation of slow and fast morphogenetic protein 2 (BMP-2) signaling in differentiation pathway of muscle myofibrillogenesis by Wnt/beta-catenin and myostatin signaling. C2C12 myoblasts. J Biol Chem 2005, 280:37660–37668. PLoS One 2009, 4:e5880. 46. Ball RW, Warren-Paquin M, Tsurudome K, Liao EH, Elazzouzi F, Cavanagh C, 23. Hutcheson DA, Zhao J, Merrell A, Haldar M, Kardon G: Embryonic and fetal An BS, Wang TT, White JH, Haghighi AP: Retrograde BMP signaling limb myogenic cells are derived from developmentally distinct controls synaptic growth at the NMJ by regulating trio expression in progenitors and have different requirements for beta-catenin. Genes Dev motor neurons. Neuron 2010, 66:536–549. 2009, 23:997–1013. 24. Henriquez JP, Salinas PC: Dual roles for Wnt signalling during the doi:10.1186/2044-5040-3-5 formation of the vertebrate neuromuscular junction. Acta Physiol (Oxf) Cite this article as: Kuroda et al.: Canonical Wnt signaling induces BMP-4 2012, 204:128–136. to specify slow myofibrogenesis of fetal myoblasts. Skeletal Muscle 2013 25. Chakkalakal JV, Nishimune H, Ruas JL, Spiegelman BM, Sanes JR: Retrograde 3:5. influence of muscle fibers on their innervation revealed by a novel marker for slow motoneurons. Development 2010, 137:3489–3499. 26. Ogata T: Structure of motor endplates in the different fiber types of vertebrate skeletal-muscles. Arch Histol Cytol 1988, 51:385–424. 27. Ishibashi J, Perry RL, Asakura A, Rudnicki MA: MyoD induces myogenic differentiation through cooperation of its NH2- and COOH-terminal regions. J Cell Biol 2005, 171:471–482. 28. Soneoka Y, Cannon PM, Ramsdale EE, Griffiths JC, Romano G, Kingsman SM, Kingsman AJ: A transient three-plasmid expression system for the production of high titer retroviral vectors. Nucleic Acids Res 1995, 23:628–633. 29. Veeman MT, Axelrod JD, Moon RT: A second canon. Functions and mechanisms of beta-catenin-independent Wnt signaling. Dev Cell 2003, 5:367–377. 30. Kuang SH, Kuroda K, Le Grand F, Rudnicki MA: Asymmetric self-renewal and commitment of satellite stem cells in muscle. Cell 2007, 129:999– 31. Tallquist MD, Weismann KE, Hellstrom M, Soriano P: Early myotome specification regulates PDGFA expression and axial skeleton development. Development 2000, 127:5059–5070. 32. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, Costantini F: Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol 2001, 1:4. 33. Mohamed OA, Clarke HJ, Dufort D: beta-catenin signaling marks the prospective site of primitive streak formation in the mouse embryo. Dev Dyn 2004, 231:416–424. Submit your next manuscript to BioMed Central 34. Harada N, Tamai Y, Ishikawa T, Sauer B, Takaku K, Oshima M, Taketo MM: and take full advantage of: Intestinal polyposis in mice with a dominant stable mutation of the beta-catenin gene. EMBO J 1999, 18:5931–5942. • Convenient online submission 35. Kuroda K, Han H, Tani S, Tanigaki K, Tun T, Furukawa T, Taniguchi Y, Kurooka • Thorough peer review H, Hamada Y, Toyokuni S, Honjo T: Regulation of marginal zone B cell development by MINT, a suppressor of Notch/RBP-J signaling pathway. • No space constraints or color figure charges Immunity 2003, 18:301–312. • Immediate publication on acceptance 36. Tajbakhsh S, Rocancourt D, Cossu G, Buckingham M: Redefining the • Inclusion in PubMed, CAS, Scopus and Google Scholar genetic hierarchies controlling skeletal myogenesis: Pax-3 and Myf-5 act upstream of MyoD. Cell 1997, 89:127–138. • Research which is freely available for redistribution 37. Nishimura R, Kato Y, Chen D, Harris SE, Mundy GR, Yoneda T: Smad5 and DPC4 are key molecules in mediating BMP-2-induced osteoblastic Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Skeletal Muscle Springer Journals

Canonical Wnt signaling induces BMP-4 to specify slow myofibrogenesis of fetal myoblasts

Loading next page...
 
/lp/springer-journals/canonical-wnt-signaling-induces-bmp-4-to-specify-slow-myofibrogenesis-aI8zEgIUQZ
Publisher
Springer Journals
Copyright
Copyright © 2013 by Kuroda et al.; licensee BioMed Central Ltd.
Subject
Life Sciences; Cell Biology; Developmental Biology; Biochemistry, general; Systems Biology; Biotechnology
eISSN
2044-5040
DOI
10.1186/2044-5040-3-5
pmid
23497616
Publisher site
See Article on Publisher Site

Abstract

Background: The Wnts are secreted proteins that play important roles in skeletal myogenesis, muscle fiber type diversification, neuromuscular junction formation and muscle stem cell function. How Wnt proteins orchestrate such diverse activities remains poorly understood. Canonical Wnt signaling stabilizes β-catenin, which subsequently translocate to the nucleus to activate the transcription of TCF/LEF family genes. Methods: We employed TCF-reporter mice and performed analysis of embryos and of muscle groups. We further isolated fetal myoblasts and performed cell and molecular analyses. Results: We found that canonical Wnt signaling is strongly activated during fetal myogenesis and weakly activated in adult muscles limited to the slow myofibers. Muscle-specific transgenic expression of a stabilized β-catenin protein led to increased oxidative myofibers and reduced muscle mass, suggesting that canonical Wnt signaling promotes slow fiber types and inhibits myogenesis. By TCF-luciferase reporter assay, we identified Wnt-1 and Wnt-3a as potent activators of canonical Wnt signaling in myogenic progenitors. Consistent with in vivo data, constitutive overexpression of Wnt-1 or Wnt-3a inhibited the proliferation of both C2C12 and primary myoblasts. Surprisingly, Wnt-1 and Wnt-3a overexpression up-regulated BMP-4, and inhibition of BMP-4 by shRNA or recombinant Noggin protein rescued the myogenic inhibitory effect of Wnt-1 and Wnt-3a. Importantly, Wnt-3a or BMP-4 recombinant proteins promoted slow myosin heavy chain expression during myogenic differentiation of fetal myoblasts. Conclusions: These results demonstrate a novel interaction between canonical Wnt and BMP signaling that induces myogenic differentiation towards slow muscle phenotype. Keywords: Skeletal muscle, Fetal myoblasts, Canonical Wnt signaling, BMP4 signaling, Differentiation, Slow muscle specification Background key transcription factors involved in myogenesis [6-10]. Skeletal muscles of the trunk and limb, except for some Importantly, genetic knockout studies have clearly dem- craniofacial and esophageal muscles, are derived from so- onstrated the requirement of several Wnt molecules and mites during embryonic development [1-3]. Specification β-catenin in the normal development of skeletal muscles of somitic cells into myogenic lineages is regulated by [11,12]. These diverse functions of Wnt are mediated by positive and negative signals from the surrounding tissues. both a canonical signaling pathway that requires sta- Wnt signaling induced by Wnt-1, -3a, -4, -6, -7a and 11 bilization and nuclear translocation of β-catenin, and non- from dorsal neural tube or ectoderm is critical for the in- canonical pathway that is independent of β-catenin [5]. duction, initiation and progression of myogenesis in the Therefore, canonical and non-canonical Wnt signaling presomitic mesoderm and early somites (Reviewed in pathways play multiple essential roles in embryonic [4,5]). Within the embryonic myogenic progenitors, Wnt myogenesis. also regulate the expression of Pax3/7, MyoD and Myf5, Wnt signaling is also involved in the regulation of post- natal satellite cell function and skeletal muscle rege- * Correspondence: mrudnicki@ohri.ca neration. Satellite cells are muscle resident stem cells Sprott Center for Stem Cell Research, Ottawa Hospital Research Institute, responsible for postnatal regeneration of injured muscles. 501 Smyth Road, Ottawa, ON K1H 8L6, Canada Full list of author information is available at the end of the article © 2013 Kuroda et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Kuroda et al. Skeletal Muscle 2013, 3:5 Page 2 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 During muscle regeneration, Wnt-5a and Wnt-7a induce the function of canonical Wnt signaling in muscle. We muscle resident CD45 stem cells to undergo myogenic found that Wnt signaling is highly activated in prenatal specification and differentiation [13]. Wnt-7a also acts muscle and rapidly declines in postnatal muscle, with through the non-canonical Wnt signaling pathway to some residual activation limited to the slow myofibers. stimulate the symmetric expansion of satellite cells and Interestingly, Wnt signaling is highly activated proximal promote skeletal muscle hypertrophy [14,15]. Similarly, to motor endplates of slow myofibers. Wnt activation canonical Wnt activation was shown to induce satellite induced by stabilized β-catenin inhibits muscle differen- cell proliferation during skeletal muscle regeneration [16]. tiation and promotes slow muscle determination. At the However, contradictory results showing that activation of molecular level, canonical Wnt signaling induces BMP-4 canonical Wnt signaling is necessary to counteract Notch , which promotes expression of slow MyHC. These novel signaling to induce myogenic differentiation were also results demonstrate that interplay between Wnt and reported [17]. Furthermore, in the aged niche, elevated BMP signaling regulates skeletal myogenesis and muscle systemic Wnt molecules impede myogenic differentiation fiber type. and facilitate satellite cell fate conversion to fibroblastic cell lineages [18]. Methods Adult skeletal muscles contain heterogeneous types of Plasmids muscle fibers that can be broadly divided into slow- and Wnts cDNA subcloned from pLNCX-HA-Wnt constructs fast-twitch myofibers [19]. In the limb muscles, the slow- (kindly gifted from Dr. Jan Kitajewski) to pHAN-puro twitch myofibers express type I myosin heavy chain retrovirus vector by PCR [27,28]. Wnt-10b and TCF-4B (MyHC), while the fast-twitch myofibers express type IIa, cDNA were obtained from Thermo Fisher Scientific Open IIx and IIb MyHC isoforms [19]. During early stages Biosystems (Waltham, MA, USA). Super TOPFlash was of myofiber generation in chicks, Wnt-11 promotes fast kindly gifted from Dr. Moon [29]. The mouse BMP-4 myofiber formation, whereas Wnt-5 enhances slow myo- shRNA pLKO.1-puro (RMM3981-9595140) was obtained fiber generation [20]. Wnt-4 similarly stimulated fast from Openbiosystems. Scramble shRNA pLKO.1-puro myofiber formation in chicks [21]. Activation of β-catenin (Addgene Plasmid #1864), MD2.G (Addgene Plasmid induced myofiber hypertrophy followed by degeneration #12259) and psPAX2 (Addgene Plasmid #12260) were of fast myofibers in zebrafish [22]. In mice, depletion of obtained from Addgene (Cambridge, MA, USA). β-catenin during myogenesis in Pax7-lineage cells led to reduced slow myofibers and overall reduction of muscle C2C12 myoblasts, single myofibers and isolation of mass [23]. Interestingly, expression of stabilized β-catenin primary myoblasts in Pax7 lineage cells also led to reduced myogenesis but C2C12 cells are purchased from American Type Culture increased slow myofibers [23]. These studies suggest that Collection (ATCC) (Manassas, VA, USA) and cultured in Wnt signaling plays diverse roles in regulating slow- ver- DMEM with 10% FBS and antibiotics. Cells are induced sus fast-twitch myofiber formation during development. to differentiation upon 80% confluence by serum with- However, embryonic lethality of β-catenin deficient mice drawal (DMEM with 2% horse serum). Primary myoblasts or over-expressing mice precludes analysis of Wnt signa- were isolated from the hind limb of two- to three–month- ling in postnatal muscles. In addition, how canonical Wnt old wild type mice as previously described [27]. Single signaling regulates myofiber types has remained unclear. myofibers were isolated from flexor digitorum longus Myofibers in postnatal skeletal muscles retain an adap- (EDL) muscle by collagenase digestion. The single fibers tive capacity to switch between slow- and fast-twitch were immediately fixed in 4% paraformaldehyde (PFA), properties that largely depend on motoneuron activity permeabilized by Triton-X100, stained with primary anti- [19]. Wnt signaling also regulates the establishment and bodies for β-gal (Life Technologies, Carlsbad, CA, USA), maintenance of neuromuscular junctions that connect MyHC (MF-20, Developmental Studies Hybridoma Bank motor neurons and myofibers. In Drosophila and mice, [DSHB], Iowa, Iowa, USA), slow MyHC (DSHB) and Wnt secreted by presynaptic motoneurons interact with FITC-conjugated α-bungarotoxin (BTX) (Sigma-Aldrich, Agrin-MusK to induce assembly of postsynaptic endplates St. Louis, MO, USA). Nuclei were counter stained with (Reviewed in [24]). There is evidence that neuromuscular 4',6-diamidino-2-phenylindole (DAPI). junctions are phenotypically and functionally distinct in Fetal myoblasts were prepared from the legs of Myf5- fast and slow muscles [25,26]. Whether Wnt signaling is Cre/ROSA26-YFP embryos at E14.5. The legs are differentially activated in slow and fast myofibers is com- minced and digested by collagenase, dispase and DNaseI, pletely unknown. as previously described for satellite cells isolation [30]. In this study, we used transgenic Wnt-reporter mice After collagenase procedure, collected cells were stained and mice with constitutively activated canonical Wnt by alpha7-integrin antibody and anti-mouse IgG1- signaling specifically within skeletal muscle to investigate Alexa648. Fetal myoblast were isolated by MoFlo (Dako, Kuroda et al. Skeletal Muscle 2013, 3:5 Page 3 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Glostrup, Denmark). Fetal myoblast were isolated by tained inside a barrier facility, and experiments were per- magnetic-antibody cell sorting (MACS) (Miltenyi Biotec, formed in accordance with the University of Ottawa regu- Bergish Gladbach, Germany). The isolated cells were incu- lations for animal care and handling. bated in culture dishes at 37°C for one hour to remove ad- herent cells, and the nonadherent cells were collected. Immunofluorescence staining The isolated fetal myoblasts cultured in DMEM/F10 C2C12 cells were incubated at 37°C for two hours with medium with 20% FBS and bFGF. 10 μM BrdU, then washed with PBS(−) and fixed with 2% PFA/PBS. The fixed cells were stained with DAPI for Retrovirus and lentivirus infection 15 minutes at room temperature, washed with PBS(−), To prepare ecotropic retrovirus, Phoenix-eco packaging and refixed with 2% PFA for 5 minutes at room tem- (kindly gifted from Dr. Gally Nolan) cells were trans- perature. The refixed cells were treated with 2N HCl for fected with retrovirus vectors using GeneJuice (Novagen 20 minutes at room temperature (RT), neutralized by EMD Chemicals, Madison, WI, USA). Viral supernatants washing with 0.1 M borate buffer pH 8.5. The cells were were harvested 30 hours post transfection and used to permeablized with 0.2% Triton X-100 PBS(−), blocked infect C2C12 cells in the presence of polybrene (Sigma, with broking buffer, incubated with anti-BrdU antibody 8 mg/ml) for 12 hours. Infected C2C12 cells were then for 2 to 12 hours. After staining with the primary anti- washed twice with phosphate-buffered saline (PBS), body, cells were washed with PBS(−), stained with anti- maintained in growth media and were selected 24 hours mouse IgG1-Alexa488, washed with PBS(−) and moun- post-infection with puromycin (1.5 micro g/ml, Sigma). ted on slide glass with Dako mounting buffer. Myoblasts Lentivirus was packaged in 293T cells (ATCC CRL- were fixed in 4% PFA/PBS (−) for 5 minutes, blocked 11268) transfected with the mouse BMP-4 or scramble with 10% goat serum/PBS (−) for 10 minutes, and shRNA pLKO.1-puro plasmids in addition to the pMD2. stained with MyHC slow, MyHC fast, MyHC pan, anti- G and psPAX2 plasmids. Lentivirus was concentrated by mouse IgG1-Alexa568, anti-mouseIgG2b-Alexa648 (Life ultracentrifuge and resolved in PBS(−) for virus infec- Technologies) and DAPI (Sigma). Slides were mounted tion. Infected C2C12 and primary myoblast cells were in SlowFade Light antifade Kit Component A (Molecular then washed twice with PBS, and maintained in growth Probes) and analyzed with a Bio-Rad confocal laser scan- media. ning microscope (model MRC-1024) (Bio-Rad Labora- tories, Hercules, CA, USA). Gene expression analysis Total RNAs were prepared from C2C12 and myoblast ALP, X-gal and NADH-TR staining cells by TRIzol (Life Technologies). RNA samples were Cells were fixed 2% PFA and washed with PBS(−), then reverse transcribed using random hexamer and oligo dT stained by alkaline phosphatase (ALP) buffer (100 mM mixed primers with SuperScriptII enzyme (Life Techono- Tris–HCl pH 9.5, 100 mM NaCl, 50 mM MgCl ) logies) according to the manufacturer’s instructions. Re- containing 4.5 μl nitro-blue tetrazolium chloride (NBT) verse transcription reactions were diluted (1:10) with 10 and 3.5 μl 5-bromo-4-chloro-3-indolyl phosphate (BCIP) mM Tris, pH 8.0, yielding master samples of reverse- per 1 ml of ALP buffer at 37°C. Isolated whole embryos transcribed products. Real-time PCR reactions are previ- and tissue were fixed in 2% PFA for 3 hours at 4°C and ously described [27]. Real-time data were gathered using a permeablized with X-gal staining buffer (0.1 M phoshate system (MX4000; Agilent Technologies, Santa Clara, CA, buffer (pH 7.3), 2 mM MgCl , 0.01% sodium deoxy- USA) over 40 cycles (30 s at 90°C, 60 s at 58°C and 30 s at cholate, 0.02% Nonidet P-40) for 2 hours at 37°C and 72°C) followed by a denaturation curve from 54°C to 94°C stained by X-gal staining buffer with 0.1% X-gal, 5 mM in 30-s increments of 0.5°C to ensure amplification specifi- potassium ferricyanide and 5 mM potassium ferrocya- city. Threshold cycle (Ct) values were calculated with the nide at 37°C for 2 to 24 hours. MX4000 software (Agilent Technologies, Santa Clara, CA, The isolated muscle tissues were placed directly into USA) by using moving window aver-aging and an adap- optimal cutting temperature (OCT) compound and fro- tive baseline. Fold changes, other calculations and chart zen in deep cold isopentane with ethanol and dry ice. plotting were performed in Microsoft Excel (Redmond, The muscle tissues were cut 16 μm by cryostat and dried WA, USA). The sequence of PCR primers is listed in in room temperature. The muscle sections were incu- Additional file 1: Table S1. bated with NADH-TR staining solution (0.8 mg/ml NADH and 1 mg/ml NBT in 50 mM Tris–HCl (pH 7.6) Animal care at 37°C. The stained muscle sections were washed with Myf5-Cre [31] heterozygous mice were bred with deionized water, unbound NBT was removed by acetone ROSA26-YFP [32]. TCF-lacZ [33] and Ctnn1 exon3 floxed solution, then the sections were re-washed with deio- [34] mice were previously described. All mice are main- nized water and mounted with a coverslip. Kuroda et al. Skeletal Muscle 2013, 3:5 Page 4 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Luciferase assays The number of muscles labeled with X-GAL was re- Myoblasts in 24-well plates were transfected with the plas- duced in the forelimb and hand limb in P0 fetus mids indicated and 50 ng pRL-PGK using Lipofectamine (Figure 1E, F). In the adult, β-gal activity was undetect- (Invitrogen). Transfected cells were harvested around 24 able in EDL and TA, muscles that are predominantly hours after transfection, and luciferase activities in the cell enriched with fast type myofibers, but was readily extracts were measured according to the manufacturer's detected in part of the diaphragm (Figure 1G) and soleus instructions (Promega, Fitcburg, WI, USA) in a lumino- (Figure 1H), muscles that are known to be enriched with meter (Microplate luminometer LB96V (EG&G Berthold slow myofibers. These data indicate that canonical Wnt Technologies, Bad Wildbad, Germany). Luciferase acti- signaling is strongly activated during fetal myogenesis vities as indicated by arbitrary unit were normalized by and declined in postnatal muscles with some residual ac- sea urchin luciferase activities in each sample. All experi- tivity in slow myofibers. ments were repeated at least three times, and the averages Interestingly, the β-gal activity in postnatal slow of more than three independent experiments with stan- muscle was especially strong at neuromuscular junctions dard deviations are shown as bars [35]. (dark dotted staining patterns in the mid-belly of the muscles; Figure 1G-H). To further examine this Western blots phenomenon, we isolated single myofiber from the so- Infected C2C12 cells were grown in 60-mm dishes, leus muscles (containing about 50% slow and 50% fast washed twice with PBS and lysed in 100 mL radioim- myofibers) of Tcf-lacZ mice. The isolated single munoprecipitation assay (RIPA) buffer (50 mM Tris HCl, myofibers were fixed and stained with α-bungarotoxin pH 7.5; 150 mM NaCl; 0.5% Nonidet P-40; 0.1% (BTX) and β-gal antibodies, together with slow-MyHC deoxycholate) containing protease inhibitor cocktail or pan-MyHC antibodies (Figure 2A-D). Consistently, (Roche Applied Science, Penzberg, Germany) [35]. Cell strong β-gal immunoreactivity was detected proximal to extracts were collected and spun in a microcentrifuge at motor endplates located within BTX stained neuromus- 13,000 rpm for 5 minutes. Total proteins (5 mg) were sep- cular junctions of type I fibers (Figure 2A, B), but not in arated on 10% SDS-PAGE and transferred to Immobilon- type II myofibers (Figure 2C-D). Co-labeling whole P (EMD Millipore Corporation, Billerica, MA, USA). The mount muscle with slow and fast myosin heavy antibody membranes were probed with primary antibodies, fol- confirmed the specific activation of β-gal in the in slow lowed by horseradish peroxidase (HRP)-conjugated se- myofibers (Figure 2E, F). Overall, 97% of the β-gal condary antibodies at 1:5,000 (Bio-Rad Laboratories), and myofibers co-expressed the slow MyHC, where only 8% ™ - developed using ECL Plus (GE Healthcare, Chalfont St. of the β-gal myofibers co-expressed slow MyHC (Figure Giles, United Kingdam). Membranes were exposed to 2G). These data indicate that canonical Wnt signaling is BIOMAX film (Eastman Kodak, Rochester, NY, USA). Pri- highly activated at the neuromuscular junction area spe- mary antibodies used in this work: anti-MyoD (5.8A, BD cifically in slow myofibers of adult skeletal muscles. Bioscience [San Jose, CA, USA]), anti-myogenin (F5D, DSHB), anti-MHC (MF-20, DSHB), anti-GAPDH (6C5, Canonical Wnt signaling promotes formation of slow Life Technologies) and anti-.alpha;-tublin (Sigma-Aldrich). myofibers in vivo To confirm the role of canonical Wnt signaling in Results muscle fiber type specification in vivo, we took advan- lox(ex3) Canonical Wnt signaling is activated during fetal tage of the Ctnnb transgenic mice in which the myogenesis and reduced in adult muscle exon 3 of β-catenin (Ctnnb) gene flanked by LoxP sites As the first step to investigate the function of Wnt sig- [34]. The exon 3 encodes serine and threonine residues naling in myogenesis, we used the TCF-lacZ transgenic that are normally phosphorylated GSK3β, leading to the reporter mouse to examine the activity of the canonical proteasomal degradation of β-catenin. Upon Cre- ΔEx3 Wnt signaling pathway in embryonic and adult muscles. mediated excision of Ctnnb exon 3, β-catenin is The promoter of the LacZ transgene (encoding β- prevented from degradation (stabilized) and, therefore, galactosidase, β-gal) contains multimerized TCF binding constitutively active. ΔEx3 sites [33], a key downstream effector of canonical Wnt We first used Myf5-Cre to induce β-catenin expres- signaling. We analyzed β-gal activity, in embryonic and sion in myogenic progenitor cells. Myf5 is an early myo- adult muscles by X-GAL staining (Figure 1). β-gal activ- genic commitment marker during embryonic myogenesis lox(ex3) ity was detected in many muscles in E14.5 embryos [36]. Myf5-Cre/Ctnnb mice die at E15.5 with (Figure 1A). Among those labeled, β-gal activity was par- extremely reduced muscle mass (data not shown), thus ticularly intense in both forelimb and hind limb muscles precluding further analysis of myofiber types. We next ΔEx3 (Figure 1B), ventral body wall muscles (Figure 1C), dor- used MCK-Cre to drive β-catenin expression only in sal spinotrapezius (Figure 1D) and intercostal muscles. differentiated muscle cells. As expected, we detected Kuroda et al. Skeletal Muscle 2013, 3:5 Page 5 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Figure 1 Activation of canonical Wnt signaling in embryonic, neonatal and adult skeletal muscles. Tcf-lacZ reporter mice were used to report activation of TCF promoter, the nuclear target of canonical Wnt signaling. X-gal staining (in blue) was used to reveal LacZ (b-gal) activity. (A) Whole mount staining of an E14.5 embryo. (B) Forelimbs and hind limbs at E14.5. (C) Ventral and (D) dorsal view of an E14.5 embryo revealing intensive staining in some muscles. (E) Hind limb and (F) forelimb at P0 (postnatal Day 0). (G) Diaphragm and (H) soleus muscles of adult mouse showing staining in a subset of myofibers with intensive signals at the neuromuscular junction area. increased activation of canonical Wnt signaling in vivo in oxidative myofibers was increased in wild type TA muscles lox(ex3) transgenic mice carrying MCK-Cre, Ctnnb and TCF- (n = 3, 28.3 ± 7.7%) compared relative to the mutant litter- LacZ alleles. β-gal activity was detectable in TA muscles of mate (Figure 3C-E). As slow muscles are known to con- lox(ex3) the MCK-Cre/Ctnnb /TCF-LacZ mice, but not the tain mainly oxidative myofibers and fast muscles are lox(ex3) Ctnnb /Tcf-lacZ littermate controls (Figure 3A, B). mainly low oxidative and glycolytic [19], these data pro- In addition, we analyzed the oxidative activity of skeletal vide in vivo evidence that canonical Wnt signaling pro- muscle myofibers using NADH-tetrazolium (NADH-TR) motes slow myofiber phenotype in the postnatal skeletal staining. Strikingly, the high and middle oxidative muscle. myofibers were increased in TA muscles of MCK-Cre/ lox(ex3) Ctnnb mice (n = 3, 65.4 ± 8.4% and 34.6 ± 8.4%, re- Canonical Wnt signaling inhibits proliferation of C2C12 spectively) compared to the same muscle of wild type lit- and primary myoblasts termates (n = 3, 52.7 ± 5.3% and 19.0 ± 3.8%, respectively) To identify the Wnt molecules that activate the canon- (Figure 3C-E). By contrast, the number of the low ical Wnt signaling pathway in muscle, we co-transfected Kuroda et al. Skeletal Muscle 2013, 3:5 Page 6 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Figure 2 Activation of canonical Wnt signaling in adult slow myofibers. (A-D) Soleus myofibers from adult Tcf-lacZ mice were stained with antibodies for pan-myosin heavy chain (White) or slow myosin heavy chain (White), and β-gal (Red) and FITC conjugated BTX (Green). (E-F) Whole mount muscles co-stained with antibodies for slow (Blue) and fast (Red) myosin heavy together with beta-gal (Green) and BTX (Red). (G) Percentage of β-gal positive (n = 69) and negative (n = 60) myofibers that co-express slow myosin heavy chain. candidate Wnt plasmids with the TCF/LEF reporter Super We next examined the function of both canonical and TopFlash. We found that Wnt-1 and Wnt-3a strongly ac- non-canonical Wnt signaling in cultured C2C12 myo- tivated (>400 times increase in luciferase activity), Wnt-2 blasts using retrovirus expressing various Wnts. Com- and Wnt-10b moderately activated (>20 times increase in pared to mock controls, C2C12 cells overexpressing luciferase activity), and Wnt-2b and Wnt-4 weakly acti- canonical Wnts (Wnt-1, -2, -2b, -3a, -10b) had a signifi- vated (>2 times increase in luciferase activity) the canon- cantly decreased cell number after 96 h in culture ical Wnt reporter (Additional file 2: Figure S1A). Other (Additional file 2: Figure S1B). By contrast, C2C12 cells Wnts (Wnt-5a, Wnt-5b, Wnt-6, Wnt-7a, Wnt-7b, Wnt- overexpressing non-canonical Wnts (Wnt-5a, -5b, -6, -7b, 10a, Wnt-11) had no effect on the activation of the Super -11) had a moderately increased cell number compared to TopFlash reporter (Additional file 2, Figure S1A). We, the control treated cells (Additional file 2: Figure S1B). therefore, used Wnt-1 and Wnt-3a to activate the canon- A reduced cell number in C2C12 cells overexpressing ical Wnt signaling in the following studies. Wnt-1, -2, -2b, -3a and -10b indicates that canonical Kuroda et al. Skeletal Muscle 2013, 3:5 Page 7 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 decreased in the presence of Wnt-3a protein. These re- sults indicate that canonical Wnt signaling suppressed the proliferation of C2C12 cells and adult primary myoblasts. To test if Wnt signaling has any effect on myogenic differentiation, we induced C2C12 myoblasts to differen- tiate. Upon serum withdrawal, control myoblasts exited the cell cycle and fused to form myotubes with uniform morphology (Additional file 4: Figure S3A). By contrast, Wnt-1 and Wnt-3a expressing C2C12 cells formed very few myosin heavy chain (MyHC) positive myotubes (Additional file 4: Figure S3B-C). In addition, these myotubes were morphologically abnormal: they were short and chubby (Additional file 4: Figure S3B-C). As previously reported, non-canonical Wnt7a overexpres- sion led to the formation of large myotubes resembling the muscle hypertrophy phenotype (Additional file 4: Figure S3D). We further examined by Western blotting the expression of two myogenic differentiation markers, myogenin and MyHC. Compared to the control, Wnt-1 and -3a robustly inhibited the expression of Myogenin and MyHC at 48 hours, 72 hours and 96 hours post- induction of differentiation (Additional file 4: Figure Figure 3 Muscle-specific constitutive activation of canonical S3E). Intriguingly, the non-canonical Wnt-7a not only Wnt signaling promotes oxidative myofiber phenotype. The lox(ex3) MCK-Cre/Ctnnb mice were used to express constitutively increased Myogenin and MyHC protein levels, but also Δex3 active β-catenin , which mimics canonical Wnt signaling. The induced their earlier expression at 24 hours (Additional MCK promoter-drive Cre expression limits Wnt activation only in file 4: Figure S3E). However, recombinant Wnt-3a pro- mature skeletal muscles. (A-B) X-Gal staining (blue signal) of whole lox(ex3) tein at 50 ng/ml had no effect on the differentiation of mount TA muscles of Ctnnb /Tcf-LacZ (WT) and MCK-Cre lox(ex3) neither C2C12 myoblasts (not shown) nor fetal primary /Ctnnb /Tcf-LacZ (Mut) mice. Blue signal, indicative of canonical Wnt signaling, is only detectable in the Mut TA muscles (A and B myoblasts (see Figure 4H). Thus, these data indicate that represent the dorsal and ventral view of the same muscles). (C-D) prolonged high-level constitutive activation of Wnt-1 NADH-TR staining of TA muscle sections from the WT (C) and Mut and Wnt-3a in C2C12 cells suppresses myogenic (D)mice. (E) Percentage of high oxidative (black), middle oxidative (gray) differentiation. and low oxidative (white) myofibers of WT and Mut mice (n = 3, each). Scale bar: 1 mm. Canonical Wnt signaling activates BMP signaling in C2C12 myoblasts Wnt signaling inhibits cell proliferation. To confirm this, The reduced myogenic differentiation of Wnt-1 and we analyzed the expression of Ki-67 (Additional file 2: Wnt-3a overexpressing C2C12 myoblasts prompted us Figure S1 C-E), a nuclear antigen specifically expressed to analyze the alternative differentiation fate of these in S, G2 and M phase cells. The Wnt-1 and -3a ex- cells. Previous studies show that C2C12 cells can also pressing C2C12 cells had decreased Ki-67 expression differentiate into osteogenic lineage [37], which express (Additional file 2: Figures S1 D-E, and 4I). In addition, ALP. Surprisingly, Wnt-1, Wnt-3a and Wnt-10b overex- we examined incorporation of BrdU, a thymidine analog pression strongly induced ALP immunochemical signals that is incorporated into proliferating cells during the S (Additional file 5: Figure S4A-F). By contrast, expression phase. Proliferating C2C12 cells were incubated with of non-canonical Wnt-5a and Wnt-7a did not affect BrdU for one hour and then fixed for BrdU staining ALP expression (Additional file 5: Figure S4D-E). Quan- (Additional file 2: Figure S1F-H). Consistently, BrdU in- titative analysis confirms that Wnt-1 and Wnt-3a corporation was reduced in the Wnt-1 and -3a express- increased ALP enzyme activity by 15- and 30-fold, re- ing C2C12 cells (Additional file 2: Figure S1G-H, J). spectively (Additional file 5: Figure S4G). Furthermore, we analyzed proliferation of skeletal To examine if Wnt-induced ALP activity was depen- muscle-derived primary myoblasts cultured with recom- dent on β-catenin, which mediates canonical Wnt sig- binant Wnt-3a protein (50 ng/ml) for 24 hours. Both Ki- naling, we transduced C2C12 cells with a dominant 67 expression (Additional file 3: Figure S2A-C) and negative TCF-4b (DN-TCF-4b) that lacks the β-catenin BrdU uptake (Additional file 3: Figure S2D-F) were binding site. As expected, DN-TCF-4b suppressed both Kuroda et al. Skeletal Muscle 2013, 3:5 Page 8 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Figure 4 Canonical Wnt signaling induces BMP-4 in C2C12 myoblasts. (A) qPCR analysis of the relative expression of various genes in C2C12 myoblasts overexpressing Wnt-1 and Wnt-3a under growth condition. (B) BMP-4 knockdown reduced the expression of alkaline phosphatase (Akp2) but not the Axin2 gene. (C-F) BMP-4 shRNA rescued the myogenic differentiation of Wnt-1 and Wnt-3a infected C2C12 cells. (C-D) Wnt-1 and Wnt-3a overexpressing myoblasts treated with Scramble shRNA; (E-F) Wnt-1 and Wnt-3a overexpressing myoblasts treated with BMP-4 shRNA. Black signaling is myosin heavy chain antibody staining reacted with 3, 3'-diaminobenzidine (DAB) substrate. (G) Western blotting showing myosin heavy chain protein expression after BMP-4 shRNA treatment. (H) qPCR analysis of myosin heavy chain gene expression. (I-L) BMP-4 antagonist Noggin-Fc dose-dependently rescued the myogenic differentiation of Wnt-3a infected C2C12 cells. The myosin heavy chain was labeled in black and nuclei were labeled in white. Wnt-1 and Wnt-3a induced ALP activity in C2C12 cells expression was increased by more than five-fold in Wnt- (Additional file 5: Figure S4H). These data indicate that 1 and Wnt-3a overexpressing C2C12 cells (Figure 4A). canonical Wnt signaling induces ALP expression via a In comparison, Myf5 and MyoD expression was not af- β-catenin/TCF-dependent pathway in C2C12 cells. fected by Wnt-1 and Wnt-3a (Figure 4A). To ensure that That canonical Wnt signaling induces osteogenic ALP canonical Wnt signaling is involved in the induction of expression suggests a potential interaction between Wnt BMP-4, we examined Axin2, a transcriptional target of and BMP signaling pathways. We first examined BMP-4 β-catenin and canonical Wnt signaling [38]. Wnt-1 and gene expression using quantitative RT-PCR (qPCR) Wnt-3a overexpression led to over 40X increase in the given its role in osteogenesis. Indeed, BMP-4 mRNA expression Axin2 (Figure 4A). Kuroda et al. Skeletal Muscle 2013, 3:5 Page 9 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Figure 5 Wnt-3a induces slow myosin heavy chain expression in fetal myoblasts via BMP-4. Fetal myoblasts were isolated by fluorescence- activated cell sorting (FACS) from E14.5-15.5 embryos of Myf5-Cre/Rosa26-YFP mice. FACS isolated fetal myoblasts were cultured and induced to differentiate for three days. (A-D) Immunofluorescence showing slow myosin heavy chain (green) expression in myotubes treated with vehicle control medium or Wnt-3a recombinant protein (50 ng/ml). (E) Relative expression of myosin heavy chain and BMP-4 genes based on qPCR analysis. (F) Western blotting showing slow myosin heavy chain protein expression. (G-J) Immunofluorescence showing slow- (green) and pan- (red) myosin heavy chain expression in myotubes treated with vehicle control medium, Wnt-3a, and BMP-4 recombinant protein at concentration shown. (K) qPCR analysis showing relative expression of myosin heavy chain isoform genes. Scale bars: 40 μm. We next asked if BMP-4 is necessary for pro- myotubes in the C2C12 cells overexpressing Wnt-1 osteogenic effect of canonical Wnt signaling. We used (Figure 4C) and Wnt-3a (Figure 4C). Knockdown of BMP- lentiviral shRNA mediated knockdown of BMP-4 in 4 remarkably increased the numbers of MyHC positive C2C12 cells. This approach resulted in nearly 80% reduc- myotubes in Wnt-1 (Figure 4E) and Wnt-3a (Figure 4F) tion of BMP-4 transcripts (Figure 4B). Importantly, BMP- expressing C2C12 cells. Consistent with this observation, 4 knockdown reduced Wnt-1 and Wnt-3a induced Akp4 BMP-4 shRNA rescued MyHC protein expression in Wnt- (ALP gene) expression by more than 70% (Figure 4B). By 1 and Wnt-3a overexpressing C2C12 cells (Figure 4G). contrast, Axin2 mRNA levels were not decreased follow- Moreover, BMP-4 shRNA increased the mRNA levels of ing knockdown of BMP-4 (Figure 4B), suggesting that Myogenin, MyHC-IIa, MyHC-IIb and MyHC-I (Figure 4H). BMP-4 signaling does not affect canonical Wnt signaling. To confirm the above observations, we used recombinant We further investigated if the anti-myogenic effect of ca- Noggin protein, an antagonist of BMP, to block BMP activ- nonical Wnt signaling is mediated by BMP-4 using the ity. Noggin dose-dependently increased the number of same shRNA knockdown approach. In the control groups MyHC positive myotubes in Wnt-3a expressing C2C12 (scrambled shRNA), there were only a few MyHC positive cells (Figure 4I-L). These data provide compelling evidence Kuroda et al. Skeletal Muscle 2013, 3:5 Page 10 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 that canonical Wnt signaling inhibits myogenic differen- BMP-4 to induce slow MyHC expression during embry- tiation through inducing BMP-4 signaling. onic myoblast differentiation. Discussion Canonical Wnt signaling induces slow MyHC expression In this study, we use genetic, cell culture and molecular through BMP-4 biology approaches to dissect the function of canonical To directly examine the function of canonical Wnt signa- Wnt signaling in myogenic differentiation and skeletal ling in muscle fiber type specification, we isolated em- myofiber types. We show that the canonical Wnt signal- bryonic myoblasts from E14.5 embryos by FACS. We ing is most active during perinatal myogenesis and only employed positive selection for Myf5 and α7-integrin ex- activated in slow myofibers with high activity at the pression of myogenic cells from Myf5-Cre/ROSA-YFP em- neuromuscular junction area in mature muscles. Con- bryos (Additional file 6: Figure S5). The purity of isolated stitutive activation of β-catenin, the canonical Wnt sig- fetal myoblasts was confirmed by immunostaining for naling effector, leads to impaired myogenesis and an Pax7 and desmin (Additional file 6: Figure S5). Embryonic increased proportion of oxidative myofibers in the post- myoblasts were cultured for one day before being induced natal muscles. Importantly, Wnt-1 and Wnt-3a mediated to differentiate with or without the addition of Wnt-3a downstream signaling activates BMP-4, which inhibits protein (50 ng/ml). In the absence of Wnt-3a, embryonic the overall proliferation of myoblasts and promotes myoblasts differentiated into myotubes that expressed fast myogenic differentiation towards slow muscle pheno- MyHC (Figure 5A), but not slow MyHC (Figure 5C) in type. These results establish a novel interaction between agreement with previous studies [39]. By contrast, Wnt-3a Wnt and BMP signaling that regulates muscle fiber type treated fetal myoblasts differentiated into myotubes specification and maintenance. that expressed both fast (Figure 5B) and slow MyHC The TCF-LacZ reporter mouse has been widely used (Figure 5D). in reporting activation of canonical Wnt signaling in To confirm this result, we analyzed MyHC-I mRNA various tissues/cells [11,40]. Strong LacZ expression in expression by qPCR. The MyHC-I mRNA level was up- specific muscles during embryonic and fetal myogenesis regulated 5.5-fold by Wnt-3a compared to control ve- indicates activation of canonical Wnt signaling. Interes- hicle treatment (Figure 5E). Wnt3a also robustly induced ting, several muscles (spinotrapezius, body wall muscle the expression of slow MyHC-I at the protein level and diaphragm) with high β-gal activity are known to be (Figure 5F). These results indicate that canonical Wnt enriched with slow myofibers [19]. These results suggest signaling induces slow MyHC expression in fetal a role of Wnt signaling in slow muscle generation and myoblasts. maintenance. Our in vivo results are consistent with pre- As canonical Wnt signaling induced BMP-4, we fur- vious studies in chick and fish embryos, in which cano- ther examined the role of BMP-4 in muscle fiber type nical Wnt signaling was shown to promote slow muscle specification. Consistent with our previous results in fate [20-22]. C2C12 cells (Figure 4A), Wnt-3a treatment of embry- Our analysis of canonical Wnt signaling in adult ma- onic myoblasts induced a seven-fold increase in BMP-4 ture muscles reveal several interesting points. First, β-gal mRNA expression (Figure 5E). Next, we added recom- activity is only detectable in muscles known to contain binant BMP-4 to fetal myoblast cultures during slow myofibers. This result confirms our observation in differentiation. In the control treated with vehicle the developing embryonic muscles. Second, in contrast medium, newly formed MyHC myotubes seldom to embryonic muscle, where β-gal activity is evenly dis- expressed slow MyHC after three days of differentia tributed within myofibers, the highest β-gal activity was tion (Figure 5G). In the presence of 2.5 to 5 ng/ml within the slow myofibers was proximal to the motor BMP-4, slow-MyHC myotubes were abundantly visible endplate (Figure 1G-H). This observation suggests that (Figure 5I-J). The level of slow MyHC immunofluores- whereas in embryonic muscle the Wnt molecules are cence induced by 5 ng/ml BMP-4 is similar to that in- released from surrounding tissues [5], Wnt signaling duced by 50 ng/ml Wnt-3a (Figure 5H), suggesting in adult slow myofibers is most likely initiated by Wnt BMP-4 more potently induces slow MyHC expression. molecules from motor neurons that innervate these Western blotting showed that the slow MyHC protein myofibers. In support of the notion that the motor expression level was indeed increased in the presence of neuron supplies Wnt molecules, we found that β-gal BMP-4 (Figure 5F). In addition, qPCR analysis indicated immunoreactivity was no longer detectable in slow that BMP-4 not only induced the slow MyHC-I gene ex- myofibers after three days of suspended culture in vitro pression, but also robustly suppressed the fast MyHC-IIb in the absence of neural innervation (data not shown). gene expression (Figure 5K). Collectively, these results Previous studies demonstrate that Wnt molecules indicate that canonical Wnt signaling acts through released by motor neurons play key roles in the Kuroda et al. Skeletal Muscle 2013, 3:5 Page 11 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 development of neuromuscular junctions. Specifically, Interestingly, embryonic and fetal myoblasts seem to interaction of Wnt and LRP is necessary for clustering have different responses to canonical Wnt signaling [23], of postsynaptic acetylcholine receptors (AchR) [24]. Our suggesting that the role of Wnt signaling, even in the new results demonstrate that Wnt signaling is further same cell lineage, is also context dependent. Consistent required for the maintenance of neuromuscular junction with this notion, we show that the growth and differenti- in slow myofibers. Future study is needed to examine ation of fetal primary myoblasts are not inhibited by re- the functional significance of Wnt signaling in slow ver- combinant Wnt-3a protein in culture. The osteogenic sus fast muscle fibers and identify the Wnt molecules re- fate choice of Wnt-1 and Wnt-3a overexpressing C2C12 leased by slow and fast motor neurons. cells is in line with a recent report demonstrating fibro- Using Cre-inducible transgenic mice that express sta- blastic lineage differentiation of satellite cells in response bilizes β-catenin, we investigated the role of canonical to high level of systemic Wnt molecules [18]. Thus, the Wnt signaling in embryonic myogenesis and postnatal observed effect of Wnt-1 and Wnt-3 in myogenic cell muscle maintenance. When Myf5-Cre is used as the proliferation and differentiation is largely consistent with driver mouse, which is expressed in embryonic myogenic information in the literature. progenitors, we detected abnormal muscle development We discovered a novel interaction between Wnt and and perinatal lethality. This observation is consistent to BMP signaling in myoblasts. Bone morphogenetic pro- recent studies using Pax7-Cre or Myogenin-Cre to teins (BMPs) are multi-functional proteins belonging to stabilize β-catenin, which also results in lethality at P0 the transforming growth factor beta (TGFβ) superfamily. [23,41]. In these studies, constitutive activation of β- In zebrafish and frogs, BMP signaling inhibits the diffe- catenin in the myogenic progenitors and committed rentiation of muscle precursors in the dermomyotome myocytes resulted in a shift of fetal myofibers to slow and controls the number of myogenic cells [42,43]. Du- muscle phenotype, and reduced myofiber size. However, ring late myogenesis of mice, BMP signaling regulates the the perinatal lethality of the Myf5-Cre, Myogenin-Cre and number of fetal myoblasts and satellite cells [44]. This ac- Pax7-Cre drive β-catenin activation precludes analysis of tion is through preventing the premature activation of Lox(ex3) postnatal muscles. Using the Mck-Cre/ctnnb mice, MyoD while maintaining Pax3 expression. Therefore, we found that adult muscles indeed have higher canonical BMPs may function to establish a sufficient number of Wnt activity based on the TCF-LacZ reporter assay. This myogenic progenitors before terminal differentiation. verifies the utility of our mouse model. Our cell culture results indicate that Wnt signaling in- Importantly, we found the adult fast (TA) muscles duces BMP4, and BMP4 inhibition rescues the inhibitory exhibited features of slow muscle phenotype as increased effect of Wnt-1 and Wnt-3a on myogenesis. This result is oxidative capacity. We also examined myosin heavy in line with the above results in vivo.We further identify chain expression by immunohistochemistry but did not an unexpected role for BMP-4 in promoting slow muscle find any overt changes in myofiber types (data not fate during fetal myogenesis. It is important to mention shown). This result suggests that although Wnt signaling that low concentrations (1 to 5 ng/ml) of BMP-4 protein affects metabolic properties in the adult muscles, it is were used in our study. Non-physiological, high concen- not sufficient to switch myosin heavy chain expression trations of BMP will probably generate completely diffe- in the adult. This is expected since other factors, such as rent effects [45]. Future studies should illustrate how BMP hormones and neural activity, can also influence myosin signaling regulates myosin gene expression. Lox(ex3) expression [19]. Together, our Mck-Cre/ctnnb mo- Interestingly, in Drosophila larval neuromuscular junc- del bypasses the premature lethality and provides novel tions, retrograde BMP signaling controls synaptic growth insights of canonical Wnt signaling in regulating the oxi- [46]. The muscle-derived BMP modulates cytoskeletal dy- dative capacity of adult muscles. namics and structural changes at presynaptic terminals. The observation that C2C12 cells overexpressing Wnt-1 This forms a feedback system in which canonical Wnt and Wnt-3a exhibited reduced proliferation and myo- molecules secreted from motor neurons not only induce genic differentiation is quite intriguing. It could suggest formation of neuromuscular junctions, but also activate BMP-4 expression in the muscle. The muscle derived that canonical Wnt/β-catenin inhibits the proliferation and differentiation of myogenic cell lineages. This pos- BMP-4 subsequently promotes development of presynaptic sibility would explain the reduced muscle mass phe- motor neuron terminals. Indeed, β-catenin stabilization in skeletal muscles (not limited to the neuromuscular junc- notypes of the Myf5-Cre, Myogenin-Cre and Pax7-Cre induced stabilized β-catenin mice [23,41]. Alternati- tion area) results in increased motor axon number and ex- vely, the result may also suggest that expression of cessive intramuscular nerve defasciculation and branching [41]. Taken together, our experiments have identified a Wnt-1 and Wnt-3a in the cell, independent of Frizzled receptor activation, is detrimental to cell growth and novel interaction between canonical Wnt and BMP signa- differentiation. ling that plays a role in myofiber type specification. Kuroda et al. Skeletal Muscle 2013, 3:5 Page 12 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 Conclusion bFGF: Basic fibroblast growth factor; PFA: Paraformaldehyde; PGK: Phosphoglycerate kinase; GSK: Glycogen synthase kinase; MCK: Muscle Our study demonstrates that canonical Wnt-signaling creatine kinase; BrdU: Bromodeoxyuridine. controls the development of skeletal muscles via BMP-4 expression. High concentrations of BMP-4 have been Competing interests The authors declare no competing interests. previously established to inhibit myogensis and induce osteogenesis. We found that isolated fetal myoblasts do Authors’ contributions not normally form slow myofibers during myogenic dif- KK and MAR designed the research and wrote paper. SK performed the ferentiation in vitro. Strikingly, canonical Wnt-signaling histology and tissue staining, and helped with paper writing. TMM provided the Ctnn1 exon3 floxed mice. All authors read and approved the final induced low level BMP-4 expression that act to induce manuscript. slow myofibergenesis. Therefore, we conclude that ca- nonical Wnt and BMP signaling plays a hitherto un- Acknowledgments We thank Dr. Daniel Dufort for TCF-lacZ mice, Dr. Randall Moon and Dr. appreciated role in myofiber type specification during Valerie Wallace for DNA constructs. We thank members of the Rudnicki fetal myogenesis. laboratory for their technical assistance and helpful discussions, and Dr. Makoto Sato for critical reading of the manuscript. KK was supported by a Postdoctoral Fellowship from Training Program in Regenerative Medicine in Additional files Canada. Additional file 1: Table S1. The sequence of PCR primers for qPCR Author details analysis is listed. Sprott Center for Stem Cell Research, Ottawa Hospital Research Institute, 501 Smyth Road, Ottawa, ON K1H 8L6, Canada. Present Address: Additional file 2: Figure S1. Canonical Wnt signaling inhibits the Department of Animal Sciences, Purdue University, 174B Smith Hall, 901 W growth and proliferation of C2C12 myoblast cells. (A) Relative luciferase State St, West Lafayette, IN 47907, USA. Department of Pharmacology, Kyoto activity of C2C12 cells overexpressing various Wnt genes together with University Graduate School of Medicine, Konoe, Yoshida, Sakyo, Kyoto the SuperTop Flash reporter plasmid. (B) Graph of C2C12 cells transduced 606-8501, Japan. Present address: Division of Cell Biology and Neuroscience, with retroviral Wnt plasmids. (C-E) The Ki-67 antibody staining and (F-H) Department of Morphological and Physiological Sciences, Faculty of Medical BrdU incorporation of control, Wnt-1 and Wnt-3a transduced cells. (I) Sciences, University of Fukui, 23-3, Matsuoka-Shimoaizuki, Eiheiji, Fukui Percentage of Ki-67 positive cells in control, Wnt-1 and Wnt-3a 910-1193, Japan. transduced cells. (J) Percentage of BrdU incorporated cells in Control, Wnt-1 and Wnt-3a retrovirus infected C2C12 cells. Received: 29 October 2012 Accepted: 15 February 2013 Additional file 3: Figure S2. Wnt-3a inhibits proliferation of adult Published: 5 March 2013 primary myoblasts. (A-B) Ki-67 antibody staining of primary myoblasts treated with vehicle control (A) and Wnt-3a recombinant protein (50 ng/ References ml) (B). (C) Percentage of Ki67 positive cells. (D-E) BrdU incorporation of 1. Biressi S, Molinaro M, Cossu G: Cellular heterogeneity during vertebrate primary myoblast cells treated with control (D) and Wnt-3a protein (E). (F) skeletal muscle development. Dev Biol 2007, 308:281–293. Percentage of cells incorporated BrdU. 2. Buckingham M, Vincent SD: Distinct and dynamic myogenic populations Additional file 4: Figure S3. Overexpression of Wnt-1 and Wnt-3a in the vertebrate embryo. Curr Opin Genet Dev 2009, 19:444–453. inhibits the myogenic differentiation of C2C12 myoblasts. (A-D) 3. Christ B, Ordahl CP: Early stages of chick somite development. Anat Immunofluorescence showing sarcomeric myosin heavy chain protein Embryol 1995, 191:381–396. expression in C2C12 cells transduced with retrovirus overexpressing (A) 4. Cossu G, Borello U: Wnt signaling and the activation of myogenesis in Mock, (B) Wnt-1, (C) Wnt-3a and (D) Wnt-7a after four days of induced mammals. EMBO J 1999, 18:6867–6872. myogenic differentiation. (E) Western blotting analysis showing 5. von Maltzahn J, Chang NC, Bentzinger CF, Rudnicki MA: Wnt signaling in expression of myosin heavy chain (MyHC), myogenin and GAPDH in myogenesis. Trends Cell Biol 2012, 22:602–609. control, Wnt-1, Wnt-3a and Wnt-7a overexpression C2C12 myoblasts at 6. Abu-Elmagd M, Robson L, Sweetman D, Hadley J, Francis-West P, different time points after induced to differentiate. GAPDH: Munsterberg A: Wnt/Lef1 signaling acts via Pitx2 to regulate somite Glyceraldehyde 3-phosphate dehydrogenase. Scale bar: 100 μm. myogenesis. Dev Biol 2010, 337:211–219. Additional file 5: Figure S4. Canonical Wnts (−1, -3a and 10b) induces 7. Tajbakhsh S, Borello U, Vivarelli E, Kelly R, Papkoff J, Duprez D, Buckingham alkaline phosphatase (ALP) activity in C2C12 cells. (A-F) ALP staining M, Cossu G: Differential activation of Myf5 and MyoD by different Wnts (Purple signal) of C2C12 cells transduced retrovirus expressing empty in explants of mouse paraxial mesoderm and the later activation of vector (Mock) or Wnt vectors as indicated. (G) Relative ALP activity. (H) myogenesis in the absence of Myf5. Development 1998, 125:4155–4162. The dominant negative (DN) TCF-4 suppressed ALP activity in Wnt-1 and 8. Borello U, Berarducci B, Murphy P, Bajard L, Buffa V, Piccolo S, Buckingham Wnt-3a expressing C2C12 cells. M, Cossu G: The Wnt/beta-catenin pathway regulates Gli-mediated Myf5 expression during somitogenesis. Development 2006, 133:3723–3732. Additional file 6: Figure S5. Isolation of fetal myoblasts from E14.5 9. Ikeya M, Takada S: Wnt signaling from the dorsal neural tube is required embryos of Myf5-Cre/ROSA26-YFP mice by fluorescence activated cell for the formation of the medial dermomyotome. Development 1998, sorting (FACS). (A-C) YFP expression in whole embryo (A), forelimb (B) 125:4969–4976. and hindlimb (C) of E14.5-15.5 Myf5-Cre/ROSA26-YFP embryo. (D) 10. Otto A, Schmidt C, Patel K: Pax3 and Pax7 expression and regulation in Strategy for isolating fetal myoblasts by FACS. Whole limbs of Myf5-Cre the avian embryo. Anat Embryol (Berl) 2006, 211:293–310. /ROSA26-YFP embryos were minced and digested by collagenase and 11. Aoki K, Taketo MM: Tissue-specific transgenic, conditional knockout and dispase. Single cells were stained with alpha7-integrin antibody and knock-in mice of genes in the canonical Wnt signaling pathway. Methods selected by YFP and alpha7-integrin expression. The sorted YFP and Mol Biol 2008, 468:307–331. alpha7-integrin double positive cells were stained with antibody for Pax7 12. van Amerongen R, Berns A: Knockout mouse models to study Wnt signal and desmin, makers of fetal myoblasts (nuclei were counterstained by transduction. Trends Genet 2006, 22:678–689. DAPI in blue). 13. Seale P, Polesskaya A, Rudnicki MA: Adult stem cell specification by Wnt signaling in muscle regeneration. Cell Cycle 2003, 2:418–419. Abbreviations 14. Le Grand F, Jones AE, Seale V, Scime A, Rudnicki MA: Wnt7a activates the BMP: Bone morphogenetic protein; TCF: T-cell factor; LEF: Lymphoid planar cell polarity pathway to drive the symmetric expansion of enhancer-binding factor; Pax: Paired box protein; FBS: Fetal bovine serum; satellite stem cells. Cell Stem Cell 2009, 4:535–547. Kuroda et al. Skeletal Muscle 2013, 3:5 Page 13 of 13 http://www.skeletalmusclejournal.com/content/3/1/5 15. von Maltzahn J, Bentzinger CF, Rudnicki MA: Wnt7a-Fzd7 signalling differentiation of the pluripotent mesenchymal precursor cell line directly activates the Akt/mTOR anabolic growth pathway in skeletal C2C12. J Biol Chem 1998, 273:1872–1879. muscle. Nat Cell Biol 2012, 14:186–191. 38. Jho EH, Zhang T, Domon C, Joo CK, Freund JN, Costantini F: Wnt/beta- 16. Otto A, Schmidt C, Luke G, Allen S, Valasek P, Muntoni F, Lawrence-Watt D, catenin/Tcf signaling induces the transcription of Axin2, a negative Patel K: Canonical Wnt signalling induces satellite-cell proliferation regulator of the signaling pathway. Mol Cell Biol 2002, 22:1172–1183. during adult skeletal muscle regeneration. J Cell Sci 2008, 121:2939–2950. 39. Hagiwara N, Yeh M, Liu A: Sox6 is required for normal fiber type 17. Brack AS, Conboy IM, Conboy MJ, Shen J, Rando TA: A temporal switch differentiation of fetal skeletal muscle in mice. Dev Dyn 2007, 236:2062–2076. from Notch to Wnt signaling in muscle stem cells is necessary for 40. Barolo S: Transgenic Wnt/TCF pathway reporters: all you need is Lef? normal adult myogenesis. Cell Stem Cell 2008, 2:50–59. Oncogene 2006, 25:7505–7511. 18. Brack AS, Conboy MJ, Roy S, Lee M, Kuo CJ, Keller C, Rando TA: Increased 41. Liu Y, Sugiura Y, Wu FF, Mi WT, Taketo MM, Cannon S, Carroll T, Lin WC: Wnt signaling during aging alters muscle stem cell fate and increases beta-Catenin stabilization in skeletal muscles, but not in motor neurons, fibrosis. Science 2007, 317:807–810. leads to aberrant motor innervation of the muscle during neuromuscular development in mice. Dev Biol 2012, 366:255–267. 19. Schiaffino S, Reggiani C: Fiber types in mammalian skeletal muscles. 42. Patterson SE, Bird NC, Devoto SH: BMP regulation of myogenesis in Physiol Rev 2011, 91:1447–1531. zebrafish. Dev Dyn 2010, 239:806–817. 20. Anakwe K, Robson L, Hadley J, Buxton P, Church V, Allen S, Hartmann C, 43. Daughters RS, Chen Y, Slack JM: Origin of muscle satellite cells in the Harfe B, Nohno T, Brown AM, Evans DJ, Francis-West P: Wnt signalling Xenopus embryo. Development 2011, 138:821–830. regulates myogenic differentiation in the developing avian wing. 44. Wang H, Noulet F, Edom-Vovard F, Tozer S, Le Grand F, Duprez D: Bmp Development 2003, 130:3503–3514. signaling at the tips of skeletal muscles regulates the number of fetal 21. Takata H, Terada K, Oka H, Sunada Y, Moriguchi T, Nohno T: Involvement of muscle progenitors and satellite cells during development. Dev Cell 2010, Wnt4 signaling during myogenic proliferation and differentiation of 18:643–654. skeletal muscle. Dev Dyn 2007, 236:2800–2807. 45. Nakashima A, Katagiri T, Tamura M: Cross-talk between Wnt and bone 22. Tee JM, van Rooijen C, Boonen R, Zivkovic D: Regulation of slow and fast morphogenetic protein 2 (BMP-2) signaling in differentiation pathway of muscle myofibrillogenesis by Wnt/beta-catenin and myostatin signaling. C2C12 myoblasts. J Biol Chem 2005, 280:37660–37668. PLoS One 2009, 4:e5880. 46. Ball RW, Warren-Paquin M, Tsurudome K, Liao EH, Elazzouzi F, Cavanagh C, 23. Hutcheson DA, Zhao J, Merrell A, Haldar M, Kardon G: Embryonic and fetal An BS, Wang TT, White JH, Haghighi AP: Retrograde BMP signaling limb myogenic cells are derived from developmentally distinct controls synaptic growth at the NMJ by regulating trio expression in progenitors and have different requirements for beta-catenin. Genes Dev motor neurons. Neuron 2010, 66:536–549. 2009, 23:997–1013. 24. Henriquez JP, Salinas PC: Dual roles for Wnt signalling during the doi:10.1186/2044-5040-3-5 formation of the vertebrate neuromuscular junction. Acta Physiol (Oxf) Cite this article as: Kuroda et al.: Canonical Wnt signaling induces BMP-4 2012, 204:128–136. to specify slow myofibrogenesis of fetal myoblasts. Skeletal Muscle 2013 25. Chakkalakal JV, Nishimune H, Ruas JL, Spiegelman BM, Sanes JR: Retrograde 3:5. influence of muscle fibers on their innervation revealed by a novel marker for slow motoneurons. Development 2010, 137:3489–3499. 26. Ogata T: Structure of motor endplates in the different fiber types of vertebrate skeletal-muscles. Arch Histol Cytol 1988, 51:385–424. 27. Ishibashi J, Perry RL, Asakura A, Rudnicki MA: MyoD induces myogenic differentiation through cooperation of its NH2- and COOH-terminal regions. J Cell Biol 2005, 171:471–482. 28. Soneoka Y, Cannon PM, Ramsdale EE, Griffiths JC, Romano G, Kingsman SM, Kingsman AJ: A transient three-plasmid expression system for the production of high titer retroviral vectors. Nucleic Acids Res 1995, 23:628–633. 29. Veeman MT, Axelrod JD, Moon RT: A second canon. Functions and mechanisms of beta-catenin-independent Wnt signaling. Dev Cell 2003, 5:367–377. 30. Kuang SH, Kuroda K, Le Grand F, Rudnicki MA: Asymmetric self-renewal and commitment of satellite stem cells in muscle. Cell 2007, 129:999– 31. Tallquist MD, Weismann KE, Hellstrom M, Soriano P: Early myotome specification regulates PDGFA expression and axial skeleton development. Development 2000, 127:5059–5070. 32. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, Costantini F: Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol 2001, 1:4. 33. Mohamed OA, Clarke HJ, Dufort D: beta-catenin signaling marks the prospective site of primitive streak formation in the mouse embryo. Dev Dyn 2004, 231:416–424. Submit your next manuscript to BioMed Central 34. Harada N, Tamai Y, Ishikawa T, Sauer B, Takaku K, Oshima M, Taketo MM: and take full advantage of: Intestinal polyposis in mice with a dominant stable mutation of the beta-catenin gene. EMBO J 1999, 18:5931–5942. • Convenient online submission 35. Kuroda K, Han H, Tani S, Tanigaki K, Tun T, Furukawa T, Taniguchi Y, Kurooka • Thorough peer review H, Hamada Y, Toyokuni S, Honjo T: Regulation of marginal zone B cell development by MINT, a suppressor of Notch/RBP-J signaling pathway. • No space constraints or color figure charges Immunity 2003, 18:301–312. • Immediate publication on acceptance 36. Tajbakhsh S, Rocancourt D, Cossu G, Buckingham M: Redefining the • Inclusion in PubMed, CAS, Scopus and Google Scholar genetic hierarchies controlling skeletal myogenesis: Pax-3 and Myf-5 act upstream of MyoD. Cell 1997, 89:127–138. • Research which is freely available for redistribution 37. Nishimura R, Kato Y, Chen D, Harris SE, Mundy GR, Yoneda T: Smad5 and DPC4 are key molecules in mediating BMP-2-induced osteoblastic Submit your manuscript at www.biomedcentral.com/submit

Journal

Skeletal MuscleSpringer Journals

Published: Mar 5, 2013

References