Myostatin Is an Inhibitor of Myogenic Differentiation

Am J Physiol Cell Physiol 282: C993–C999, 2002. First published December 12, 2001; 10.1152/ajpcell.00372.2001.

Myostatin is an inhibitor of myogenic differentiation

RAMO´ NRIÓS, ISABEL CARNEIRO, VIĆTOR M. ARCE, AND JESU´ S DEVESA Departamento de Fisiologıá, Facultad de Medicina, Universidad de Santiago de Compostela, San Francisco s/n, 15705 Santiago de Compostela, Spain Received 3 August 2001; accepted in final form 6 December 2001

Rıós, Ramo´ n, Isabel Carneiro, Vıćtor M. Arce, and the myotome compartment of developing somites and Jesu´ s Devesa. Myostatin is an inhibitor of myogenic differ- continues to be limited to the myogenic lineage at later entiation. Am J Physiol Cell Physiol 282: C993–C999, 2002. stages of development and in adult animals. Several First published December 12, 2001; 10.1152/ajpcell.00372. murine (15, 27, 33) and bovine (4–6, 17) genetic models 2001.—Myostatin (MSTN), a transforming growth factor have clearly established the role of MSTN as a negative (TGF)-␤ superfamily member, has been shown to negatively regulate muscle growth by inhibiting muscle precursor cell regulator of muscle fiber number. A reduction in muscle fiber number can result from either a decrease in proliferation. Here, we stably transfected C2C12 cells with mouse MSTN cDNA to investigate its possible role in myo- myoblast proliferation or a delay in myoblast differen- blast differentiation. We found that MSTN cDNA overexpres- tiation. The ability of MSTN to inhibit myoblast pro- sion reversibly inhibits the myogenic process by downregu- gression through the cell cycle has been recently dem- lating mRNA levels of the muscle regulatory factors myoD onstrated. Recombinant MSTN has been shown to and myogenin, as well as the activity of their downstream reversibly inhibit C2C12 murine myoblast proliferation target creatine kinase. Taking into consideration that MSTN by arresting cells in the G1 and G2/M stages of the cell expression during development is restricted to muscle, our cycle (29). This arrest is probably mediated by the results suggest that MSTN probably regulates myogenic dif- upregulation of the cyclin-dependent kinase (cdk) in- ferentiation by an autocrine mechanism. hibitor p21cip1. In keeping with these findings, Taylor muscle differentiation; muscle regulatory factors; transform- et al. (28) showed that MSTN inhibited proliferation, ing growth factor-␤ superfamily [3H]thymidine incorporation, and protein synthesis in C2C12 cells. Using a different approach, we have shown that transient transfection of C2C12 myoblasts DURING MAMMALIAN DEVELOPMENT, skeletal muscle cells with an expression vector encoding mouse MSTN arise from pluripotential mesenchymal precursors that cDNA not only inhibited cell proliferation but also become committed to the myogenic lineage upon ex- reduced differentiation-associated cell death (25), prob- pression of the muscle-specific basic helix-loop-helix ably by a mechanism involving also the upregulation of (bHLH) transcription factors myoD and myf5. In re- p21cip1, which has been previously shown to dramati- sponse to differentiation signals, myogenin and muscle cally decrease the apoptotic rate of differentiating myo- regulatory factor 4, also belonging to the bHLH family, blasts (31). execute the differentiation program that leads to the Although MSTN expression has been reported to expression of muscle-specific proteins and myocyte fu- correlate with differentiation in several chicken mus- sion into multinucleated myotubes (18, 23). During cles (10) and in C2C12 myoblasts (25), it is currently this process, more cells have to be constantly generated unknown whether MSTN plays any role in the regula- to keep pace with embryonic growth, so muscle growth tion of the myogenic process. Moreover, such a role has results from a balance between proliferation of precur- been demonstrated for several other members of the sor cells and their subsequent differentiation into mus- TGF-␤ superfamily, such as TGF-␤1 itself (14), activin cle fibers. This process is tightly regulated in vivo (12), and bone morphogenetic protein (BMP)-2 (7), through mechanisms that involve cell-cell and cell- which have been shown to inhibit the differentiation of matrix interactions, as well as extracellular secreted C2C12 myoblasts. An inhibitory effect on muscle devel- factors. Among the latter, several members of the opment in vivo has been suggested for BMP-4, a close transforming growth factor (TGF)-␤ superfamily of BMP-2 homolog (24). In all cases, the underlying mech- growth and differentiation factors have been shown to anism explaining the inhibition of the myogenic pro- be potent regulators of muscle growth (1). gram by TGF-␤ superfamily members involves the One of the TGF-␤ superfamily members that plays downregulation of the myogenic bHLH transcription an essential role in regulating skeletal muscle growth factors myoD and myogenin. is myostatin (MSTN) (15). During skeletal muscle de- In this report we show that the stable transfection of velopment, MSTN expression is restricted initially to MSTN cDNA in C2C12 cells efficiently inhibits the

Address for reprint requests and other correspondence: J. Devesa, The costs of publication of this article were defrayed in part by the Departamento de Fisiologı´a, Facultad de Medicina, Universidad de payment of page charges. The article must therefore be hereby Santiago de Compostela, San Francisco s/n, 15705 Santiago de Com- marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734 postela, Spain (E-mail: [email protected]). solely to indicate this fact. http://www.ajpcell.org 0363-6143/02 $5.00 Copyright © 2002 the American Physiological Society C993 C994 MYOSTATIN INHIBITS C2C12 MYOGENESIS formation of multinucleated myotubes, reduces the RT-PCR analysis. The effects of MSTN cDNA overexpres- mRNA levels of myoD and myogenin, and inhibits the sion on the expression of myoD and myogenin were assayed activity of the myoD and myogenin downstream target, by RT-PCR. Cells (2 ϫ 106) were seeded in 60-mm plates. creatine kinase (CK). Therefore, we propose that After a 24-h incubation in 10% FBS-DMEM, differentiation MSTN negatively regulates muscle mass not only by was induced by shifting the medium to 1% FBS-DMEM. To test whether the effect of MSTN was reversible, we incubated decreasing the proliferation rate of myoblasts but also cells in the presence of the antibody raised against the by inhibiting its terminal differentiation. COOH-terminal region of MSTN (C-20; Santa Cruz Biotech- MATERIALS AND METHODS nology). An antibody directed against the pro-region of MSTN (N-19; Santa Cruz Biotechnology) was used as a

Generation of stable clones of C2C12 cells expressing MSTN control. Cells were harvested at the indicated times, and cDNA. The cloning of murine MSTN cDNA into the pBlue- total RNA was extracted by means of the Trizol reagent script KSϩ vector (Stratagene, San Diego, CA) has been (Invitrogen). Total RNA (1 ␮g) was reverse transcribed for previously described (25). The cDNA was further subcloned 1hat37°C with 200 units of MMLV reverse transcriptase into the expression vector pcDNA 3.1 Zeo (Invitrogen, Bar- (Invitrogen), followed by 5 min at 95°C, in a 30-␮l reaction celona, Spain). The generated construct was named pcDNA- mixture containing 50 mM Tris⅐HCl (pH 8.3), 75 mM KCl, MSTN. 5.5 mM MgCl2, 0.5 mM each dNTP, 40 units of RNaseOUT The mouse myoblast C2C12 cell line was cultured in Dul- recombinant ribonuclease inhibitor (Invitrogen), and 1.7 becco’s modified Eagle’s medium (DMEM) containing 10% ␮g/␮l random primers (Invitrogen). Three microliters of the fetal bovine serum (FBS), 2 mM glutamine, 100 IU/ml peni- RT reaction were amplified by PCR with 1.25 units of Taq cillin, and 100 ␮g/ml streptomycin in a humidified 5% CO2 polymerase (Invitrogen) in fifty microliters of a reaction mix- atmosphere. All cell culture reagents were purchased from ture containing 20 mM Tris⅐HCl, pH 8.4, 50 mM KCl, 2 mM Life Technologies (Invitrogen). Cells were transfected by MgCl2, 0.2 mM each dNTP, and 0.4 ␮M each oligonucleotide means of the Lipofectamine reagent (Invitrogen) according to primer. The housekeeping gene hypoxanthine guanine phos- the manufacturer’s protocol with 1 ␮g of either the pcDNA- phoribosyl transferase (HPRT) was used as a load control. The MSTN plasmid or the control plasmid (pcDNA 3.1 Zeo alone). oligonucleotide sequences (with product length and GenBank Clones were selected in medium supplemented with 250 accession nos. for murine sequences) were as follows: HPRT ␮g/ml Zeocine (Invitrogen). MSTN overexpression was as- (139 bp; NM013556), upper 5Ј-CAGTCCCAGCGTCGTGATTA- sessed by RT-PCR (25). To verify whether MSTN was being 3Ј, lower 5Ј-AGCAAGTCTTTCAGTCCTGTC-3Ј; myoD (528 bp; processed and secreted properly, conditioned media from M84918), upper 5Ј-GATGGCATGATGGATTACAGC-3Ј, lower control and MSTN cDNA transfected clones were separated 5Ј-GACTATGTCCTTTCTTTGGGG-3Ј; myogenin (424 bp; by 15% SDS-PAGE under reducing conditions and trans- M95800), upper 5Ј-GCTCAGCTCCCTCAACCAG-3Ј, lower ferred onto a nitrocellulose membrane by electroblotting. The 5Ј-ATGTGAATGGGGAGTGGGGA-3Ј. All the oligonucleotide membrane was then blocked overnight at 4°C in Tris-buff- primer pairs were designed to amplify a region including at ered saline (TBS) buffer with 0.1% Tween 20 and 0.2% casein least one intron (assuming conservation of exon-intron junc- and then incubated for1hatroom temperature with an tions between murine myoD and myogenin) so that am- anti-MSTN antibody (1:500) raised against the COOH termi- plimers arising from genomic DNA contamination could be nus of MSTN (C-20 antibody; Santa Cruz Biotechnology, easily distinguished from those originated from genuine Heidelberg, Germany). The membrane was washed with cDNA amplification. The conditions of the PCR reactions TBS-Tween 0.1% and further incubated with protein were 28 cycles of 94°C for 1 min, 64°C for 1 min, and 72°C for A-horseradish peroxidase (HRP) conjugate (Amersham Phar- 1 min, followed by a final amplification step of 72°C for 10 macia Biotech, Freiburg, Germany) at 1:1,000 dilution for 1 h min. The PCR products were resolved on 2% agarose gels at room temperature. After washing, HRP activity was de- stained with ethidium bromide. tected by using the enhanced chemiluminescence detection CK activity. Cells (3.5 ϫ 105) were seeded in triplicate in kit (ECL; Amersham Pharmacia Biotech). 35-mm plates, incubated for 24 h in 10% FBS-DMEM, and Cell proliferation and apoptosis assays. For the prolifera- then shifted to medium containing 1% FBS to induce its tion assay, 4 ϫ 104 cells were seeded in triplicate in 35-mm differentiation. The cells were harvested by trypsinization at diameter dishes. Cells were cultured in DMEM supple- the indicated times, washed with PBS, centrifuged, and mented with 10% FBS. At 24, 48, and 72 h, cells were washed stored as a pellet at Ϫ20°C until the assay was performed. with phosphate-buffered saline buffer (PBS), harvested after Briefly, the cells were sonicated in saline buffer, and CK a 5-min incubation with 0.25% trypsin (Invitrogen), and activity was measured by a modification of the spectrophoto- counted on a Neubauer chamber. metric method described by Olivier (22) and Rosalki (26) in a For the quantification of apoptosis, 6.5 ϫ 104 cells were Dimension Clinical Chemistry System (Dade Behring, New- seeded in 15-mm dishes (n ϭ 6). Cells were incubated for 24 h ark, NJ). The amount of total protein in the samples was in DMEM containing 10% FBS and then changed to 1% FBS. determined with the pyrogallol red/molybdenum method, de- After an additional period of 72 h, cells were stained with 50 scribed by Fujita et al. (2). nM Hoechst 33258 (Sigma, St. Louis, MO). Three random Statistical analysis. A Mann-Whitney test was used to fields of each of the six replicates were photographed at a ϫ40 evaluate the statistical significance of the data. magnification with a DP10 microscope digital camera (Olym- pus Optical, Tokyo, Japan). Hoechst-positive condensed nu- RESULTS clei were counted in each of the fields. A parallel experiment Transfected C C cells overexpress a biologically ac- was performed to assay apoptosis with the Cell Death Detec- 2 12 tion ELISA (Roche Molecular Biochemicals, Mannheim, Ger- tive MSTN. To investigate the autocrine effects of many), following manufacturer’s instructions. The assay is MSTN on the myogenic process in vitro, we generated based on the quantitative determination of oligonucleosomes stable clones of C2C12 cells overexpressing murine released into the cytoplasm of apoptotic cells with monoclo- MSTN cDNA. Stable transfectants were tested for nal antibodies directed against DNA and histones. MSTN overexpression by RT-PCR analysis of total

AJP-Cell Physiol • VOL 282 • MAY 2002 • www.ajpcell.org MYOSTATIN INHIBITS C2C12 MYOGENESIS C995

RNA. As shown in Fig. 1A, the amount of MSTN D and E) in MSTN-transfected C2C12 cells. Because mRNA is considerably higher in cells transfected with these effects have been previously attributed to MSTN the pcDNA-MSTN plasmid than in control cells. Ac- (25, 28, 29), it is likely that they are produced by the cording to our previously published results (25), only a overexpressed peptide. Furthermore, the effect of faint band can be detected in cells transfected with the MSTN on the proliferation rate of the myoblasts is also pcDNA 3.1 Zeo control plasmid. Moreover, Western consistent with the fact that double-muscled cattle, blot analysis of the conditioned medium of transfected where the phenotype for disruption of the mstn gene cells showed a band migrating at the predicted size (12 was ﬁrst observed, present increased muscle ﬁber kDa) for the monomeric processed form of MSTN (Fig. number. 1B), thus demonstrating that C2C12 cells are capable of Overexpression of MSTN in C2C12 cells inhibits myo- secreting and proteolytically processing MSTN. Fi- genic differentiation. The ability of MSTN to inhibit nally, to further validate our model we determined myoblast differentiation is shown in Fig. 2. Incubation whether the MSTN overexpressed in C2C12 cells was of C2C12 cells transfected with the control plasmid for 5 biologically active by investigating its effects on cell days in differentiation medium resulted in their fusion proliferation and cell survival. Whereas proliferation into postmitotic syncytial myotubes. In contrast, when was inhibited (Fig. 1C), survival was enhanced (Fig. 1, C2C12 cells transfected with the pcDNA-MSTN plas-

Fig. 1. Transfected C2C12 cells overexpress a biologically active myostatin (MSTN). A: RT-PCR analysis of MSTN mRNA in C2C12 cells stably transfected with an empty vector (C2 1.7) or a vector encoding MSTN cDNA (C2 2.8). Hypoxanthine guanine phosphoribosyl transferase (HPRT) was used as a load control. The results presented correspond to these 2 representative clones. The results obtained with C2 1.7 cells were essentially the same as those obtained with parental C2C12 cells. Similar results to those presented for clone C2 2.8 were obtained with at least 2 additional clones overexpressing MSTN cDNA. B: Western blot analysis of conditioned media from control and MSTN-transfected cells. Media were resolved in a 15% SDS-PAGE under reducing conditions, transferred onto a nitrocellulose membrane by electroblotting, and probed with an anti-MSTN antibody (1:500) raised against the 4 COOH terminus of MSTN. C: cell proliferation assay of C2 1.7 and C2 2.8 C2C12 clones. Cells (4 ϫ 10 ) were seeded in 35-mm diameter dishes, harvested at the indicated times, and counted on a Neubauer chamber (*P Յ 0.05 vs. control). D: apoptosis assay of C2 1.7 and C2 2.8 clones. Cells (6.5 ϫ 104) were seeded in 15-mm dishes, and after incubation in 10% FBS-DMEM for 24 h, the medium was shifted to 1% FBS-DMEM. Apoptosis was assayed after 72 h by Hoechst dye staining. Cells were scored for apoptosis by their nuclear morphology (shrinkage, condensa- tion, and fragmentation) and the higher intensity of blue ﬂuorescence. Results are expressed as percentages of Hoechst-positive cells (*P Յ 0.01 vs. control). E: enrichment of nucleosomes in the cytoplasm of C2C12 cells. Cells were cultured as described in C, and apoptosis was detected with a cell death detection ELISA kit (*P Յ 0.01 vs. control). ABS, absorption.

AJP-Cell Physiol • VOL 282 • MAY 2002 • www.ajpcell.org C996 MYOSTATIN INHIBITS C2C12 MYOGENESIS

Fig. 2. Overexpression of MSTN inhibits myotube formation in C2C12 cells. Representative microphotographs show C2 1.7 and C2 2.8 clones incubated for 5 days in differentiation medium. Lens magniﬁcation, ϫ100.

mid were incubated under the same conditions, the ampliﬁcation (Fig. 3A). As Fig. 3B clearly indicates, formation of multinucleated myotubes was suppressed. MSTN cDNA overexpression decreased the mRNA lev- The effect of MSTN overexpression on myoblast dif- els of both transcription factors at the two time points ferentiation was further investigated by assessing its tested. Note that when the medium was shifted to effects on two early markers of myogenic differentia- DMEM containing 1% FBS (differentiation medium), tion: the bHLH family transcription factors myoD and the cells had been incubated in growth medium myogenin. Myoblasts were induced to differentiate by (DMEM ϩ 10% FBS) for 24 h, so MSTN could also lowering mitogen concentration in media from 10% to exert its effect at time 0. As has been previously re- 1%. RT-PCRs of different cycle number were performed ported, myoD mRNA levels do not change with the to ensure that all the assays were in the linear range of induction of differentiation, whereas myogenin mRNA

Fig. 3. Overexpression of MSTN inhibits myoD and myogenin gene expression in C2C12 cells. A: amplification plots for HPRT, myoD, and myogenin. Samples were amplified for 20, 25, 30, 35, or 40 cycles. At 28 cycles, all the PCRs were within the linear range of amplification. B: RT-PCR analysis of the effects of MSTN cDNA overexpression on the amount of myoD and myogenin transcripts. Cells (2 ϫ 106) were seeded in 60-mm plates, and after a 24-h incubation in 10% FBS-DMEM, differentiation was induced by shifting the medium to 1% FBS-DMEM. Cells were harvested at the indicated times, and RT-PCR (28 cycles) was performed with 1 ␮g of total RNA The housekeeping gene HPRT was used as a control. Note that the MSTN effect at time 0 is due to its action during the 24-h incubation in 10% FBS-DMEM. C: effect of MSTN was counter- acted in 2 additional clones by incubating the cells with an antibody against the COOH-terminal region of MSTN (anti C-MSTN). The antibody was added after 24 h of incubation in 1% FBS-DMEM, and cells were harvested 24 h later. An antibody raised against the NH2-terminal propeptide (anti N-MSTN) was used as a control. RT-PCR was performed as described in B.

AJP-Cell Physiol • VOL 282 • MAY 2002 • www.ajpcell.org MYOSTATIN INHIBITS C2C12 MYOGENESIS C997

DISCUSSION To our knowledge, this is the first report in the literature demonstrating the inhibitory effect of MSTN on myogenesis. In the same direction, Oldham et al. (21) recently reported that myoD expression is increased in double-muscled cattle fetuses. This increase is probably caused by the lack of an active biological form of MSTN in these animals, thus suggesting a role for MSTN in the negative regulation of the differentiation process in vivo. Surprisingly, no upregulation of myogenin was found in this case. The reason for this discrepancy is unknown. Two different ways of inhibiting the myogenic pro- Fig. 4. Overexpression of MSTN inhibits creatine kinase (CK) activ- cess have been described. One is the use of mitogens, ity in C2C12 cells. CK activity of control (C2 1.7) and MSTN cDNA- which tend to delay myogenic differentiation by retain- transfected (C2 2.8) clones of C2C12 cells was determined. Cells ing the cells in a proliferative state (19). On the other 5 (3.5 ϫ 10 ) were seeded in 35-mm plates, incubated for 24 h in 10% hand, the factors belonging to the TGF-␤ superfamily, FBS-DMEM, and then shifted to medium containing 1% FBS to induce their differentiation. CK activity is expressed as nmol of along with MSTN, inhibit both proliferation and differ- NADPH produced per minute and is normalized to the amount of entiation (14). However, some striking differences exist protein (*P Յ 0.05 vs. control). among members of the superfamily. Although both TGF-␤ and activin, on one side, and BMP-2 and related factors, on the other, inhibit myotube formation in levels increase as differentiation takes place. To test C2C12 myoblasts, the latter can also induce this cell whether the effect of MSTN overexpression was revers- line to express osteoblast phenotypes, such as alkaline ible, two additional clones of C2C12 cells transfected phosphatase (ALP) activity (7). According to our own either with the empty vector (1.8) or myostatin cDNA unpublished observations, MSTN is unable to induce (2.5) were incubated for 24 h in low-serum medium and ALP activity in C2C12 cells, not a surprising finding then treated with an antibody directed against the considering the phenotype of mstn-null mice, which COOH-terminal myostatin peptide. After an additional show no defects in osteogenesis (15). Differences also 24-h incubation, the expression of myoD and myogenin exist between TGF-␤/activin and BMP with regard to was assessed by RT-PCR. As shown in Fig. 3C, the the regulation of adipogenic differentiation. MSTN has inhibition of overexpressed myostatin with the anti- also been reported to be expressed in adipose tissue body against the COOH-terminal region almost com- (15) and, along with TGF-␤1 and TGF-␤2, strongly pletely reversed its negative regulatory effects on inhibited adipogenesis (9, 30). In contrast, both BMP-2 myoD and myogenin transcription. In contrast, when and BMP-4 have been shown to induce adipogenic cells were treated with an antibody against the NH2- conversion in the pluripotential mesenchymal cell line terminal region of myostatin (this region is not con- 10T1/2 at lower concentrations than those needed for tained in the mature processed form of the protein), the transdifferentiation into osteoblasts (30). Again, these inhibition of the differentiation process was not re- results indicate that the effect of MSTN is similar to verted. that of TGF-␤/activin and different from that of BMPs. The inhibitory action of MSTN cDNA overexpression The major difference in signaling between TGF-␤/ on C2C12 cell differentiation was also evaluated by activin and BMP occurs at the level of the receptor- investigating its effect on the activity of the muscle- regulated Smad (R-Smad) that is activated by type I specific gene CK. Myoblast differentiation was induced receptors before forming a complex with the common as explained above, and CK activity was determined. Smad4 and translocating to the nucleus, where they As Fig. 4 shows, MSTN cDNA overexpression resulted regulate the transcription of target genes. R-Smads 2 in decreased CK activity at all time points studied. and 3 transduce TGF-␤ and activin signals, whereas

Fig. 5. A model for the role of MSTN in muscle development. In the course of myogenic differentiation, an autocrine loop is established to limit the muscle ﬁber number in response to MSTN signaling. MSTN actions would be ex- erted at the level of inhibition of both committed myoblast proliferation and myocyte differentiation.

AJP-Cell Physiol • VOL 282 • MAY 2002 • www.ajpcell.org C998 MYOSTATIN INHIBITS C2C12 MYOGENESIS

BMP signaling uses R-Smads 1, 5, and 8 (reviewed in This work was supported by grants from Xunta de Galicia (XUGA- Ref. 13). This differential activation of R-Smads seems 1999) and Fondo de Investigaciones Sanitarias [FIS/1180, Ministerio de Educacioń y Cultura, Spain]. R. R´ıos is a recipient of a Formacioń to mediate the differential effects of both pathways on de Profesorado Universitario fellowship from the Ministerio de Edu- myogenic differentiation, since forced expression of R- cacioń y Cultura, Spain. I. Carneiro is a recipient of a Beca de Smads 1, 5, and 8 or R-Smads 2 and 3 mimics the Formacioń en Investigacioń fellowship from the Ministerio de effects of BMPs and TGF-␤/activin, respectively, on Sanidad y Consumo, Spain. C2C12 cells (8, 32). Although MSTN signaling pathway REFERENCES has not been elucidated at the biochemical level, sev- 1. Brand-Saberi B and Christ B. Genetic and epigenetic control eral lines of evidence, apart from their common effects of muscle development in vertebrates. Cell Tissue Res 296: 199– on myogenesis and adipogenesis, suggest that MSTN 212, 1999. may share TGF-␤/activin signaling pathway. First, 2. Fujita Y, Mori I, and Kitano S. Color reaction between pyro- MSTN clusters with TGF-␤s and activins in a phylo- gallol red-molybdenum (VI) complex and protein. Bunsei Kagaku genetic tree of all known TGF-␤ superfamily members 32: 379–386, 1983. 3. Gamer LW, Cox KA, Small C, and Rosen V. GDF-11 is a (20). Second, it has been recently demonstrated that negative regulator of chondrogenesis and myogenesis in the MSTN is able to bind activin receptor type IIb (Ac- developing chick limb. Dev Biol 229: 407–420, 2001. tRIIb) and that transgenic mice overexpressing a dom- 4. Grobet L, Royo LJ, Poncelet D, Pirottin D, Brouwers B, inant negative form of ActRIIb under the control of a Riquet J, Schoeberlein A, Dunner S, Meńissier F, Massa- banda J, Fries R, Hanset R, and Georges M. A deletion in the muscle-specific promoter exhibit a dramatic increase in bovine myostatin gene causes the double-muscled phenotype in muscle mass (11), similar to that of mice lacking MSTN cattle. Nat Genet 17: 71–74, 1997. (15). In keeping with this hypothesis, a high degree of 5. Kambadur R, Sharma M, Smith TPL, and Bass JJ. Muta- similarity also exists between the phenotypes of mice tions in myostatin (GDF8) in double-muscled Belgian Blue and Piedmontese cattle. Genome Res 7: 910–915, 1997. either lacking ActRIIb or growth and differentiation 6. Karim L, Coppieters W, Grobet L, Valentini A, and factor (GDF)-11, thus suggesting that GDF-11 and Georges M. Convenient genotyping of six myostatin mutations activin may as well share a common receptor and causing double-muscling in cattle using a multiplex oligonucle- signaling pathway (16). GDF-11 has a 90% identity otide ligation assay. Anim Genet 31: 396–399, 2000. with MSTN in its COOH-terminal region, and it has 7. Katagiri T, Yamaguchi A, Komaki M, Abe E, Takahashi N, Ideda T, Rosen V, Wozney JM, Fujisawa-Sehara A, and also been shown to inhibit myogenesis in the develop- Suda T. Bone morphogenetic protein-2 converts the differentia- ing chick limb by a mechanism involving downregulation pathway of C2C12 myoblasts into the osteoblast lineage. tion of myoD expression (3). J Cell Biol 127: 1755–1766, 1994. A major difference exists between MSTN and the 8. Kawai S, Faucheu C, Gallea S, Spinella-Jaegle S, Atfi A, ␤ Baron R, and Roman S. Mouse Smad8 phosphorylation down- rest of the TGF- superfamily factors with regard to its stream of BMP receptors ALK-2, ALK-3, and ALK-6 induces its restricted expression pattern during development. association with Smad4 and transcriptional activity. Biochem Whereas MSTN expression is confined to developing Biophys Res Commun 271: 682–687, 2000. skeletal muscles, other members of the TGF-␤ super- 9. Kim HS, Liang L, Dean RG, Hausman DB, Hartzell DL, and Baile CA. Inhibition of preadipocyte differentiation by myosta- family that negatively regulate myogenesis are ex- tin treatment in 3T3-L1 cultures. Biochem Biophys Res Commun pressed in neighboring tissues. This type of communi- 281: 902–906, 2001. cation, characteristic of BMPs and GDF-11, seems to 10. Kocamis H, McFarland DC, and Killefer J. Temporal expres- be involved in the establishment of boundaries be- sion of growth factor genes during myogenesis of satellite cells derived from the biceps femoris and pectoralis major muscles of tween adjacent territories. For instance, GDF-11, the chicken. J Cell Physiol 186: 146–152, 2001. which inhibits avian limb muscle differentiation, is 11. Lee SJ and McPherron AC. Regulation of myostatin activity expressed in nonmuscle progenitor mesenchymal cells and muscle growth. Proc Natl Acad Sci USA July 17, 2001; from the progress zone, which will later differentiate 10.1073/pnas.151270098. into skeletal structures (3). Similarly, BMP-4 signals 12. Link BA and Nishi R. Opposing effects of activin A and fol- listatin on developing skeletal muscle cells. Exp Cell Res 233: emanate from the dorsal ectoderm, the neural tube, 350–362, 1997. and the lateral plate mesoderm, and its negative effect 13. Massague´ J. TGF-␤ signal transduction. Annu Rev Biochem 67: on myoD expression is counterbalanced in the dorso- 753–791, 1998. medial lip of the dermomyotome (where muscle pro- 14. Massague´ J, Cheifetz S, Endo T, and Nadal-Ginard B. Type beta transforming growth factor is an inhibitor of myogenic genitors first initiate the expression of myogenic bHLH differentiation. Proc Natl Acad Sci USA 83: 8206–8210, 1986. transcription factors) by noggin (24). Considering the 15. McPherron AC, Lawler AM, and Lee SJ. Regulation of skel- differences stated above, together with our findings, it etal muscle mass in mice by a new TGF-␤ superfamily member. is likely that MSTN regulates myogenic differentiation Nature 234: 83–89, 1997. 16. McPherron AC, Lawler AM, and Lee SJ. Regulation of an- via an autocrine/paracrine mechanism. We therefore terior/posterior patterning of the axial skeleton by growth/differ- propose a model, depicted in Fig. 5, in which MSTN entiation factor 11. Nat Genet 22: 260–264, 1999. expressed in the course of myogenic differentiation 17. McPherron AC and Lee SJ. Double muscling in cattle due to regulates the deposit of muscle mass during develop- mutations in the myostatin gene. Proc Natl Acad Sci USA 94: 12457–12461, 1997. ment by inhibiting both the proliferation of myoblasts 18. Molkentin JD and Olson EN. Defining the regulatory net- and their differentiation into multinucleated myo- works for muscle development. Curr Opin Genet Dev 6: 445–453, tubes. 1996. 19. Nadal-Ginard B. Commitment, fusion and biochemical differ- We are grateful to Dr. Jose´ Carlos Tutor and Dr. Jose´ Antonio entiation of a myogenic cell line in the absence of DNA synthesis. Casal for valuable help in the creatine kinase activity experiments. Cell 15: 855–864, 1978.

AJP-Cell Physiol • VOL 282 • MAY 2002 • www.ajpcell.org MYOSTATIN INHIBITS C2C12 MYOGENESIS C999

20. Newfeld SJ, Wisotzkey RG, and Kumar S. Molecular evolu- (Cmpt) hypermuscular mutation in mice. Mamm Genome 9: tion of a developmental pathway: phylogenetic analyses of trans- 671–672, 1998. forming growth factor-beta family ligands, receptors and Smad 28. Taylor WE, Bhasin S, Artaza J, Byhower F, Azam M, Wil- signal transducers. Genetics 152: 783–795, 1999. lard DH, Kull FC, and Gonza´ lez-Cadavid NF. Myostatin 21. Oldham JM, Martyn JAK, Sharma M, Jeanplong R, Kam- inhibits cell proliferation and protein synthesis in C2C12 cells. badur R, and Bass JJ. Molecular expression of myostatin and Am J Physiol Endocrinol Metab 280: E221–E228, 2001. MyoD is greater in double-muscled than normal-muscled cattle 29. Thomas M, Langley B, Berry C, Sharma M, Kirk S, Bass fetuses. Am J Physiol Regulatory Integrative Comp Physiol 280: JJ, and Kambadur R. Myostatin, a negative regulator of mus- R1488–R1493, 2001. cle growth, functions by inhibiting myoblast proliferation. J Biol 22. Oliver IT. A spectrophotometric method for the determination Chem 275: 40235–40243, 2000. of creatine phosphokinase and myokinase. Biochem J 61: 116, 30. Wang EA, Israel DI, Kelly S, and Luxenberg DP. Bone 1955. morphogenetic protein-2 causes commitment and differentiation 23. Rawls D and Olson EN. MyoD meets its maker. Cell 89: 5–8, in C3H10T1/2 and 3T3 cells. Growth Factors 9: 57–71, 1993. 1997. 24. Reshef R, Maroto M, and Lassar AB. Regulation of dorsal 31. Wang J and Walsh K. Resistance to apoptosis conferred by Cdk somitic cell fates: BMPs and noggin control the timing and inhibitors during myocyte differentiation. Science 273: 351–354, pattern of myogenic regulator expression. Genes Dev 12: 290– 1996. 303, 1998. 32. Yamamoto N, Akiyama S, Katagiri T, Maniki M, Kurokawa 25. Rı´os R, Carneiro I, Arce VM, and Devesa J. Myostatin T, and Suda T. Smad1 and Smad5 act downstream of intracel- regulates cell survival during C2C12 myogenesis. Biochem Bio- lular signalings of BMP-2 that inhibits myogenic differentiation phys Res Commun 19: 561–566, 2001. and induces osteoblast differentiation in C2C12 myoblasts. Bio- 26. Rosalki SB. An improved procedure for serum creatine phos- chem Biophys Res Commun 238: 574–580, 1997. phokinase determination. J Lab Clin Med 69: 696–705, 1967. 33. Zhu J, Hadhazy M, Wehling M, Tidball G, and McNally 27. Szabo G, Dallmann G, Muller G, Patthy L, Soller M, and EM. Dominant negative myostatin produces hypertrophy with- Varga L. A deletion in the myostatin gene causes the compact out hyperplasia in muscle. FEBS Lett 474: 71–75, 2000.

AJP-Cell Physiol • VOL 282 • MAY 2002 • www.ajpcell.org