Follistatin complexes Myostatin and antagonises Myostatin-mediated inhibition of myogenesis
www.AAApeptides.com for All Your Research Peptide Needs!
Abstract
Follistatin is known to antagonise the function of several members of the TGF-β family of secreted signalling factors, including Myostatin, the most powerful inhibitor of muscle growth characterised to date. In this study, we compare the expression of Myostatin and Follistatin during chick development and show that they are expressed in the vicinity or in overlapping domains to suggest possible interaction during muscle development. We performed yeast and mammalian two-hybrid studies and show that Myostatin and Follistatin interact directly. We further show that single modules of the Follistatin protein cannot associate with Myostatin suggesting that the entire protein is required for the interaction. We analysed the interaction kinetics of the two proteins and found that Follistatin binds Myostatin with a high affinity of 5.84 × 10[SUP]−10[/SUP] M. We next tested whether Follistatin suppresses Myostatin activity during muscle development. We confirmed our previous observation that treatment of chick limb buds with Myostatin results in a severe decrease in the expression of two key myogenic regulatory genes Pax-3 and MyoD. However, in the presence of Follistatin, the Myostatin-mediated inhibition of Pax-3 and MyoD expression is blocked. We additionally show that Myostatin inhibits terminal differentiation of muscle cells in high-density cell cultures of limb mesenchyme (micromass) and that Follistatin rescues muscle differentiation in a concentration-dependent manner. In summary, our data suggest that Follistatin antagonises Myostatin by direct protein interaction, which prevents Myostatin from executing its inhibitory effect on muscle development.
Keywords
Introduction
Myostatin, a member of the transforming growth factor-beta (TGF-β) family of signalling molecules, has been implicated in determining muscle size by restricting muscle growth McPherron and Lee, 1997 and McPherron et al., 1997. During development, Myostatin is expressed at the appropriate time and positions to locally decrease the rate of muscle growth without interfering with the establishment of the muscle pattern (Amthor et al., 2002b). Myogenic cells respond to Myostatin by down-regulating the expression of key transcriptional regulators of muscle development such as Pax-3, MyoD and Myf-5, which inhibit differentiation and further growth of muscle.Follistatin, a secreted glycoprotein, antagonises numerous members of the TGF-β superfamily including Myostatin Amthor et al., 2002a, Fainsod et al., 1997, Hemmati-Brivanlou et al., 1994, Iemura et al., 1998,Michel et al., 1993 and Zimmers et al., 2002. Follistatin and Myostatin are expressed in or near developing muscle Amthor et al., 1996, Amthor et al., 1999, Amthor et al., 2002a and Amthor et al., 2002b. However, it has not yet been demonstrated whether Follistatin and Myostatin interact directly. Experimentally induced over-expression of Follistatin results in muscle enlargement, whereas the Follistatin−/− KO mouse displays muscle deficiency Lee and McPherron, 2001 and Matzuk et al., 1995. In the presence of Follistatin, Myostatin fails to bind its receptor, and Myostatin-induced muscle loss can be prevented Lee and McPherron, 2001 and Zimmers et al., 2002. Although both Myostatin and Follistatin are detected in serum, they do not associate in this medium. Instead, Myostatin circulates as a complex by associating either with the Myostatin propeptide, with FLRP, a Follistatin-related protein, or with GASP, which is a putative protease inhibitor that contains a Follistatin-like domain Hill et al., 2002 and Hill et al., 2003. This raises the question whether Follistatin and Myostatin interact directly or whether both proteins use independent signalling cascades.The Follistatin gene undergoes alternative splicing to yield either short or long forms of mRNAs. These are translated into pre-proteins and then modified to remove the signal sequence (reviewed by Patel, 1998). The short isoform yields a protein composed of 288 amino acids (FS-288), which is 8–10 times more biologically active than the product of the long isoform (FS-315) (Inouye et al., 1991).Myostatin is synthesised as a precursor protein, which consists of a N-terminal propeptide domain that harbours the signal sequence and a C-terminal domain that forms a disulfide-linked dimer and functions as the active ligand McPherron et al., 1997 and Thomas et al., 2000. After cleavage of the propeptide, a large fraction of Myostatin is still non-covalently bound to its propeptide and requires release from the propeptide to attain biological activity Lee and McPherron, 2001 and Zimmers et al., 2002. Myostatin binds the Activin Receptor Type IIB, which leads to the intracellular phosphorylation of Smad3 Langley et al., 2002, Lee and McPherron, 2001, Massague and Chen, 2000 and McPherron and Lee, 1996. Phosphorylated Smad3 can bind other Smad proteins and these complexes translocate into the nucleus, where they regulate the transcription of target genes (Massague and Chen, 2000). Additionally, phosphorylated Smad3 binds and thereby inhibits the transcriptional activity of MyoD (Liu et al., 2001).In the first part of this study, we have compared the expression pattern of Myostatin and Follistatin with a view to detect evidence that these proteins may interact during muscle development. We next determined whether Follistatin binds Myostatin using yeast and mammalian two-hybrid systems. We analysed the kinetics of the Myostatin–Follistatin interaction using surface plasmon resonance. We have subsequently tested the biological relevance of this interaction by applying recombinant Myostatin and Follistatin to developing muscle of chick embryonic limb buds both in vitro and in vivo.Materials and methods
Yeast two-hybrid studies
The DupLEX-A™ yeast two-hybrid system (Origene) was used to test protein–protein interactions. The C-terminal coding region of mouse Myostatin (bp 905–1234; Genbank accession number NM010834) encoding the processed or mature portion of Myostatin was cloned into the pEG202 bait plasmid (carrying the HIS3gene) containing the Lex-A DNA binding domain using BamH1 restriction sites (LexA-MSTNmat). The portion of mouse Follistatin (bp 88–948; Genbank accession number NM08046) encoding the Follistatin was cloned into the pJG4-5 target plasmid (carrying the TRP1 gene) containing the B42 DNA activation domain using EcoR1 restriction sites (B42-FS-288). This portion of mouse Follistatin (1–287) is analogous to the active human Follistatin-288 and therefore is referred to in this manuscript as Follistatin-288. Interaction between the encoded fusion proteins was investigated by co-transforming the bait and target plasmids together with the reporter gene (lacZ) plasmid pJK103 (carrying the URA3 gene) into the yeast strain EGY194 (MATa trp1 his3 ura3 leu2:4 LexAop-LEU2). Transformed yeast cells were plated onto medium lacking histidine, uracil and tryptophan and grown at 30°C for 3 days to select for the presence of the three plasmids. Two independent colonies were then transferred to medium lacking histidine, uracil, tryptophan and leucine for 3 days at 30°C to select for positive interactions between mature Myostatin and Follistatin-288. Positives clones were then tested for expression of the second reporter gene, lacZ, by growth on medium containing X-gal and lacking histidine, uracil and tryptophan. In the DupLEX™ system, expression of the target-B42 activation domain fusion protein is galactose-inducible and therefore galactose growth-dependence was also tested. Finally, positives clones were tested against the negative bait control pEG202max to ensure specificity.To map the binding site of Follistatin to mature Myostatin, cDNA encoding truncations of Follistatin 1–63, 64–288, 1–86, 1–100, 1–136, 1–150, 1–200 and 1–250 were cloned into pJG4-5 to create target plasmids, which were tested against the mature Myostatin bait plasmid as described above.Mammalian two-hybrid studies
The TOPO® Tools Mammalian Two Hybrid Kit (Invitrogen) was used to verify yeast two-hybrid results. The C-terminal coding region of Myostatin (bp 905–1234; Genbank accession number as above) encoding the processed portion of Myostatin was TOPO® joined to the Psv40-GAL4 5′ element and SV40 pA 3′ element according to the manufacturer's instructions to create a linear DNA template for the bait protein (GAL4-MSTNmat). Similarly, the portion of Follistatin (bp 88–948; Genbank accession number as above) encoding the active Follistatin was TOPO® joined to the Psv40-VP16 5′ element and SV40 pA 3′ element to create a linear DNA template for the prey protein (VP16-FS-288). The linear DNA templates were PCR amplified using a proofreading polymerase and the following primers: 5′-TATGTATCATACACATACGATTTAGGT-3′ and 5′-GACTCAAAGGGAACTTGTTTATTGCAGCTTATAATG-3′ and PCR products were purified.
CHO cells (American Tissue Culture Collection), a Chinese hamster ovary cell line, were maintained in Dulbecco's modified Eagle medium (DMEM)/F12 (1:1) (Invitrogen) containing 10% fetal bovine serum (Sigma), 1 × 10[SUP]5[/SUP] IU/l penicillin (Sigma) 100 mg/l streptomycin (Sigma) and 27.8 mM NaHCO[SUB]3[/SUB] (maintenance medium) at 37°C in a humidified atmosphere of 5% CO[SUB]2[/SUB].Thirty nanograms of the bait and prey linear constructs were co-transfected along with the reporter plasmid pGAL/lacZ into CHO cells seeded at 2 × 10[SUP]4[/SUP] cells per well on a 96-well plate using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. After 24 h, the medium was changed to fresh maintenance medium. After a further 24 h, cells were fixed for in situ staining of β-galactosidase activity by hydrolysis of X-gal. For in situ staining, media was discarded from wells and the cells fixed with 0.05% glutaraldehyde for 15 min at room temperature. Cells were rinsed thoroughly with three washes of PBS and incubated for 1 h at 37°C with 1 mg/ml X-gal in 35 mM K[SUB]3[/SUB]Fe(CN)[SUB]6[/SUB], 35 mM K[SUB]4[/SUB]Fe(CN)[SUB]6[/SUB]·3H[SUB]2[/SUB]0 and 2 mM MgCl[SUB]2[/SUB].Control transfections were carried out to ensure specificity of interactions. For the background control, no DNA was transfected. Mature Myostatin and active Follistatin were also tested against the pCR2.1/LgT prey control plasmid and pCR2.1/p53 bait control plasmid, respectively, proteins with which they should not interact.Surface plasmon resonance
All plasmon surface resonance experiments were performed using the BIACORE 3000. Purified recombinant human Follistatin and recombinant mouse Myostatin were purchased from R&D System (USA). Follistatin was immobilised onto the surface of a CM5 sensor chip (600 resonance units) using amine-coupling chemistry. A range of Myostatin concentrations were injected over the sensor chip surface at a flow rate of 30 μl/min at 25°C. Hepes-buffered saline (HBS: 10 mM Hepes, 150 mM NaCl, 3.4 mM EDTA, 0.005% Tween 20 pH 7.4) was used as a running buffer and for sample dilution. For controls, Myostatin was run over a derivatised sensor chip, which lacked Follistatin. All Myostatin curves were corrected by subtraction of the blank run. Biacore evaluation software was used for the mathematical fitting of experimental data.Preparation of chick embryos
Fertilised chick eggs were incubated at 38°C, and the embryos were staged according to Hamburger and Hamilton (1992). Experiments were performed on embryos at stages 22 to 24, re-incubated between 6 and 8 h, sacrificed and processed for whole-mount in situ hybridisation.Myostatin and Follistatin bead preparation and application to limb buds
Recombinant Myostatin and Follistatin protein were purchased from R&D Systems. Myostatin was applied to 80- to 120-μm Affigel beads and Follistatin to Heparin beads of same size (both Sigma, UK). The proteins were loaded onto beads as described by Cohn et al. (1995). Myostatin and Follistatin were used at 1 mg/ml concentration. For control bead implantation, beads were soaked in PBS only. For bead implantation, the dorsal ectoderm and mesenchyme of the right wing were punctured with an electrolytically sharpened tungsten needle, and beads were inserted into the punctured mesenchyme using a blunt glass needle. Beads were implanted at HH-stages stated in the text.Whole-mount in situ hybridisation
All chick embryos were washed in PBS and then fixed overnight in 4% paraformaldehyde at 4°C. Anti-sense RNA probes were labelled with digoxygenin, and whole-mount in situ hybridisation was performed as described by Nieto et al. (1996). The following probes were used in this study: Follistatin, full-length fragment, 1.1 kb (gift from Dr. Anthony Graham); Myostatin, 1 kb fragment (gift from Professor Se Jin Lee);MyoD, clone CMD9 full 1.5 kb length fragment (gift from Professor Bruce Patterson) and Pax-3, 645 bp fragment corresponding to nucleotides 468–1113 (gift from Dr Martin Goulding). Whole-mount embryos were wax or cryo-sectioned at a thickness of 15 μm for histological examination.Chick limb bud micromass assay
Micromass assays were carried out as described in Swalla and Solursh (1986). Briefly, limb buds from HH stages 21–22 chick embryos were dissected and placed into a 0.05% Trypsin-EDTA solution (Invitrogen). The ectoderm was removed using tungsten needles and a single cell suspension of the limb bud mesenchyme was obtained. Micromass cultures were plated at 2 × 10[SUP]5[/SUP] cells in 10 μl drops, allowed to adhere for 2 h and then treated with media (DMEM, 10% Fetal Calf Serum—Invitrogen) containing either Myostatin or Myostatin plus Follistatin at concentrations stated in the text (R&D Systems). Micromass cultures were then fixed in 4% PFA and processed for myosin heavy chain (MHC) immunocytochemistry. Cultures were dehydrated through a methanol series and treated with hydrogen peroxide to eliminate endogenous peroxidase activity. After rehydration, cultures were incubated with a monoclonal anti-PanMHC antibody (clone A4·1025, gift from Dr Simon Hughes) in the presence of 10% horse serum. After washing in PBS, rabbit anti-mouse biotin secondary antibody (Dako) was applied, cultures again washed, then incubated with an avidin–biotin complex (Vector Laboratories) and stained using a nickel enhanced DAB/hydrogen peroxide reaction (Vector Laboratories) according to manufacturer's protocols.Results
Co-expression of Follistatin and Myostatin related to muscle development
We have previously published independent detailed expression patterns of Follistatin and Myostatin during chick embryonic development Amthor et al., 1996, Amthor et al., 1999, Amthor et al., 2002a and Amthor et al., 2002b, which suggested an overlap in expression at sites of muscle development. Here, we directly compared the expression pattern of both genes during limb and trunk muscle development.Abutting or overlapping expression of Follistatin and Myostatin during wing bud development occurs for the first time at HH-stages 25–26. Follistatin is highly expressed in the proximal wing bud and fades distally (Fig. 1A). Myostatin is expressed in a central domain and the expression extends more distally compared toFollistatin (Fig. 1B). Transverse sections through the proximal part of the wing buds reveal expression ofFollistatin in the subectodermal mesenchyme and in a central domain of the dorsal premuscle mass (Fig. 1E). Myostatin is expressed in the mesenchymal core of the limb bud and also in a central domain of dorsal premuscle mass (Fig. 1F). However, at this stage, the expression of Follistatin and Myostatin does not encompass the entire premuscle mass as indicated by the expression of Pax-3 and MyoD (Figs. 1C and D).Follistatin expression is increased in the muscle masses up to HH-stage 31, whereas during these stages,Myostatin expression is only found in some subdomains of the developing muscle (Figs. 1G and H). Thereafter, at late embryonic stages, Follistatin is down-regulated in muscle, but expression resides in the connective tissue surrounding the muscles, whereas Myostatin is increasingly expressed in most of the muscles (data not shown, see also Amthor et al., 2002b).<dl class="figure" id="FIG1" data-t="f" style="border: 1px solid rgb(215, 215, 215); margin-right: 0px; margin-bottom: 15px; margin-left: 0px; padding: 6px; vertical-align: baseline; border-top-left-radius: 3px; border-top-right-radius: 3px; border-bottom-right-radius: 3px; border-bottom-left-radius: 3px; "><dt class="autoScroll" data-style="height:996px;width:629px;" style="border: 0px; margin: 12px 0px 0px; padding: 0px; vertical-align: baseline; overflow-y: hidden; overflow-x: auto; "></dt><dd id="labelCaptionFIG1" style="border: 0px; font-size: 0.8em; margin: 0px; padding: 0px; vertical-align: baseline; color: rgb(92, 92, 92); ">Fig. 1. Follistatin and Myostatin are expressed in or close to developing muscle. Comparison of Follistatin, Myostatin, Pax-3 andMyoD expression during wing and interlimb somite development. (A–D) HH-stage 26 wing buds, dorsal view. (E–H) Transverse sections of wing buds. (I–L) HH-stage 21 interlimb somites, lateral view and (M–P) corresponding frontal sections. (Q–T) Transverse sections of HH-stage 21 interlimb somites. (A) Strong proximal Follistatin expression, which fades distally in stage 26 wing bud. Section level indicated with broken green line. (B) Myostatin expression in a central domain of a stage 26 wing bud in contrast to the broader extension of the Pax-3-expressing (C) and MyoD-expressing (D) dorsal premuscle mass. (E) Transverse section of (A) shows Follistatin expression in a distinct location of the dorsal premuscle mass (arrow) and in the subectodermal mesenchyme (arrowhead). (F) Transverse section of (B) showsMyostatin expression in a distinct location of the dorsal premuscle mass (arrow) and in the mesenchymal core (arrowhead). (G and H) Follistatin and Myostatin expression in dorsal and ventral zeugopod muscles (arrows) at HH-stage 30. Follistatinexpression partly overlaps with Myostatin expression in distinct muscle subdomains (compare G and H, arrows). (I) StrongFollistatin expression at cranial and caudal somite edges and in a hypaxial domain in contrast to the high Myostatinexpression in the somite centre (J). Somite demarked by dotted line. (K) High Pax-3 expression in the hypaxial domain, moderate expression in the cranial and caudal domain of the dermomyotome (see also O) and no expression dorsomedially (see also S). (L) MyoD highlights the full extent of the myotome. (M) Frontal section shows high Follistatin expression in the cranial and caudal somite edges (arrowheads) and weak expression in the myotome. Brackets mark extent of somites. (N)Myostatin expression in the dermomyotomal and myotomal centre, but not at the cranial and caudal dermomyotomal edges nor in the dorsomedial and ventrolateral edges (R). (O) High Pax-3 expression in the cranial and caudal part of the dermomyotome (arrowheads). Intermediate expression in the dermomyotomal centre (arrow) and in the dorsal root ganglia (asterisks). (P) MyoD expression marks the cranio-caudal extent of the myotome. (Q) Transverse section reveals highFollistatin expression in the hypaxial domain of the somite with weak expression in the myotome. (R) Myostatin expression in the central part of the dermomyotome and myotome. (S) Highest level of Pax-3 expression in the hypaxial, moderate in the intermediate and no expression in the dorsomedial domain of the dermomyotome. Asterisk marks dorsal root ganglion. (T)MyoD expression marks the full extent of the myotome.
</dd><dd class="menuButtonLinks" style="border: 0px; font-size: 0.8em; margin: 0px; padding: 6px 0px 0px; vertical-align: baseline; text-align: right; position: relative; color: rgb(92, 92, 92); ">Figure options
</dd></dl>
Although Follistatin is expressed in interlimb somites as soon as they are formed, Myostatin is not up-regulated before HH-stage 19 (Amthor et al., 2002b). In stage 21 interlimb somites, Follistatin is strongly expressed at the cranial and caudal edges of the dermomyotome and in the hypaxial somite domain, whereas there is only a faint expression in the somite centre (Figs. 1I, M, Q). The expression of Myostatin is almost complementary to that of Follistatin as it is expressed in a central domain of the dermomyotome, but not at the somite edges or in the hypaxial domain (Figs. 1J, N, R). Highest expression level of Follistatincoincides with a high expression level of Pax-3 (Figs. 1K, O, S) but not of MyoD (Figs. 1L, P, T). At later stages, both Follistatin and Myostatin are predominantly expressed in hypaxial muscle (data not shown, see also Amthor et al., 2002b).These data show that Follistatin and Myostatin are expressed in or close to developing muscle. They are expressed partly in overlapping domains and partly in abutting regions to each other. As both Myostatin and Follistatin are secreted signalling proteins, they appear to be expressed sufficiently close to enable protein–protein interaction.
www.AAApeptides.com for All Your Research Peptide Needs!
Abstract
Follistatin is known to antagonise the function of several members of the TGF-β family of secreted signalling factors, including Myostatin, the most powerful inhibitor of muscle growth characterised to date. In this study, we compare the expression of Myostatin and Follistatin during chick development and show that they are expressed in the vicinity or in overlapping domains to suggest possible interaction during muscle development. We performed yeast and mammalian two-hybrid studies and show that Myostatin and Follistatin interact directly. We further show that single modules of the Follistatin protein cannot associate with Myostatin suggesting that the entire protein is required for the interaction. We analysed the interaction kinetics of the two proteins and found that Follistatin binds Myostatin with a high affinity of 5.84 × 10[SUP]−10[/SUP] M. We next tested whether Follistatin suppresses Myostatin activity during muscle development. We confirmed our previous observation that treatment of chick limb buds with Myostatin results in a severe decrease in the expression of two key myogenic regulatory genes Pax-3 and MyoD. However, in the presence of Follistatin, the Myostatin-mediated inhibition of Pax-3 and MyoD expression is blocked. We additionally show that Myostatin inhibits terminal differentiation of muscle cells in high-density cell cultures of limb mesenchyme (micromass) and that Follistatin rescues muscle differentiation in a concentration-dependent manner. In summary, our data suggest that Follistatin antagonises Myostatin by direct protein interaction, which prevents Myostatin from executing its inhibitory effect on muscle development.
Keywords
- Follistatin;
- Myostatin;
- Myogenesis;
- Chick;
- Embryo;
- Development;
- Pax-3;
- MyoD
Introduction
Myostatin, a member of the transforming growth factor-beta (TGF-β) family of signalling molecules, has been implicated in determining muscle size by restricting muscle growth McPherron and Lee, 1997 and McPherron et al., 1997. During development, Myostatin is expressed at the appropriate time and positions to locally decrease the rate of muscle growth without interfering with the establishment of the muscle pattern (Amthor et al., 2002b). Myogenic cells respond to Myostatin by down-regulating the expression of key transcriptional regulators of muscle development such as Pax-3, MyoD and Myf-5, which inhibit differentiation and further growth of muscle.Follistatin, a secreted glycoprotein, antagonises numerous members of the TGF-β superfamily including Myostatin Amthor et al., 2002a, Fainsod et al., 1997, Hemmati-Brivanlou et al., 1994, Iemura et al., 1998,Michel et al., 1993 and Zimmers et al., 2002. Follistatin and Myostatin are expressed in or near developing muscle Amthor et al., 1996, Amthor et al., 1999, Amthor et al., 2002a and Amthor et al., 2002b. However, it has not yet been demonstrated whether Follistatin and Myostatin interact directly. Experimentally induced over-expression of Follistatin results in muscle enlargement, whereas the Follistatin−/− KO mouse displays muscle deficiency Lee and McPherron, 2001 and Matzuk et al., 1995. In the presence of Follistatin, Myostatin fails to bind its receptor, and Myostatin-induced muscle loss can be prevented Lee and McPherron, 2001 and Zimmers et al., 2002. Although both Myostatin and Follistatin are detected in serum, they do not associate in this medium. Instead, Myostatin circulates as a complex by associating either with the Myostatin propeptide, with FLRP, a Follistatin-related protein, or with GASP, which is a putative protease inhibitor that contains a Follistatin-like domain Hill et al., 2002 and Hill et al., 2003. This raises the question whether Follistatin and Myostatin interact directly or whether both proteins use independent signalling cascades.The Follistatin gene undergoes alternative splicing to yield either short or long forms of mRNAs. These are translated into pre-proteins and then modified to remove the signal sequence (reviewed by Patel, 1998). The short isoform yields a protein composed of 288 amino acids (FS-288), which is 8–10 times more biologically active than the product of the long isoform (FS-315) (Inouye et al., 1991).Myostatin is synthesised as a precursor protein, which consists of a N-terminal propeptide domain that harbours the signal sequence and a C-terminal domain that forms a disulfide-linked dimer and functions as the active ligand McPherron et al., 1997 and Thomas et al., 2000. After cleavage of the propeptide, a large fraction of Myostatin is still non-covalently bound to its propeptide and requires release from the propeptide to attain biological activity Lee and McPherron, 2001 and Zimmers et al., 2002. Myostatin binds the Activin Receptor Type IIB, which leads to the intracellular phosphorylation of Smad3 Langley et al., 2002, Lee and McPherron, 2001, Massague and Chen, 2000 and McPherron and Lee, 1996. Phosphorylated Smad3 can bind other Smad proteins and these complexes translocate into the nucleus, where they regulate the transcription of target genes (Massague and Chen, 2000). Additionally, phosphorylated Smad3 binds and thereby inhibits the transcriptional activity of MyoD (Liu et al., 2001).In the first part of this study, we have compared the expression pattern of Myostatin and Follistatin with a view to detect evidence that these proteins may interact during muscle development. We next determined whether Follistatin binds Myostatin using yeast and mammalian two-hybrid systems. We analysed the kinetics of the Myostatin–Follistatin interaction using surface plasmon resonance. We have subsequently tested the biological relevance of this interaction by applying recombinant Myostatin and Follistatin to developing muscle of chick embryonic limb buds both in vitro and in vivo.Materials and methods
Yeast two-hybrid studies
The DupLEX-A™ yeast two-hybrid system (Origene) was used to test protein–protein interactions. The C-terminal coding region of mouse Myostatin (bp 905–1234; Genbank accession number NM010834) encoding the processed or mature portion of Myostatin was cloned into the pEG202 bait plasmid (carrying the HIS3gene) containing the Lex-A DNA binding domain using BamH1 restriction sites (LexA-MSTNmat). The portion of mouse Follistatin (bp 88–948; Genbank accession number NM08046) encoding the Follistatin was cloned into the pJG4-5 target plasmid (carrying the TRP1 gene) containing the B42 DNA activation domain using EcoR1 restriction sites (B42-FS-288). This portion of mouse Follistatin (1–287) is analogous to the active human Follistatin-288 and therefore is referred to in this manuscript as Follistatin-288. Interaction between the encoded fusion proteins was investigated by co-transforming the bait and target plasmids together with the reporter gene (lacZ) plasmid pJK103 (carrying the URA3 gene) into the yeast strain EGY194 (MATa trp1 his3 ura3 leu2:4 LexAop-LEU2). Transformed yeast cells were plated onto medium lacking histidine, uracil and tryptophan and grown at 30°C for 3 days to select for the presence of the three plasmids. Two independent colonies were then transferred to medium lacking histidine, uracil, tryptophan and leucine for 3 days at 30°C to select for positive interactions between mature Myostatin and Follistatin-288. Positives clones were then tested for expression of the second reporter gene, lacZ, by growth on medium containing X-gal and lacking histidine, uracil and tryptophan. In the DupLEX™ system, expression of the target-B42 activation domain fusion protein is galactose-inducible and therefore galactose growth-dependence was also tested. Finally, positives clones were tested against the negative bait control pEG202max to ensure specificity.To map the binding site of Follistatin to mature Myostatin, cDNA encoding truncations of Follistatin 1–63, 64–288, 1–86, 1–100, 1–136, 1–150, 1–200 and 1–250 were cloned into pJG4-5 to create target plasmids, which were tested against the mature Myostatin bait plasmid as described above.Mammalian two-hybrid studies
The TOPO® Tools Mammalian Two Hybrid Kit (Invitrogen) was used to verify yeast two-hybrid results. The C-terminal coding region of Myostatin (bp 905–1234; Genbank accession number as above) encoding the processed portion of Myostatin was TOPO® joined to the Psv40-GAL4 5′ element and SV40 pA 3′ element according to the manufacturer's instructions to create a linear DNA template for the bait protein (GAL4-MSTNmat). Similarly, the portion of Follistatin (bp 88–948; Genbank accession number as above) encoding the active Follistatin was TOPO® joined to the Psv40-VP16 5′ element and SV40 pA 3′ element to create a linear DNA template for the prey protein (VP16-FS-288). The linear DNA templates were PCR amplified using a proofreading polymerase and the following primers: 5′-TATGTATCATACACATACGATTTAGGT-3′ and 5′-GACTCAAAGGGAACTTGTTTATTGCAGCTTATAATG-3′ and PCR products were purified.
CHO cells (American Tissue Culture Collection), a Chinese hamster ovary cell line, were maintained in Dulbecco's modified Eagle medium (DMEM)/F12 (1:1) (Invitrogen) containing 10% fetal bovine serum (Sigma), 1 × 10[SUP]5[/SUP] IU/l penicillin (Sigma) 100 mg/l streptomycin (Sigma) and 27.8 mM NaHCO[SUB]3[/SUB] (maintenance medium) at 37°C in a humidified atmosphere of 5% CO[SUB]2[/SUB].Thirty nanograms of the bait and prey linear constructs were co-transfected along with the reporter plasmid pGAL/lacZ into CHO cells seeded at 2 × 10[SUP]4[/SUP] cells per well on a 96-well plate using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. After 24 h, the medium was changed to fresh maintenance medium. After a further 24 h, cells were fixed for in situ staining of β-galactosidase activity by hydrolysis of X-gal. For in situ staining, media was discarded from wells and the cells fixed with 0.05% glutaraldehyde for 15 min at room temperature. Cells were rinsed thoroughly with three washes of PBS and incubated for 1 h at 37°C with 1 mg/ml X-gal in 35 mM K[SUB]3[/SUB]Fe(CN)[SUB]6[/SUB], 35 mM K[SUB]4[/SUB]Fe(CN)[SUB]6[/SUB]·3H[SUB]2[/SUB]0 and 2 mM MgCl[SUB]2[/SUB].Control transfections were carried out to ensure specificity of interactions. For the background control, no DNA was transfected. Mature Myostatin and active Follistatin were also tested against the pCR2.1/LgT prey control plasmid and pCR2.1/p53 bait control plasmid, respectively, proteins with which they should not interact.Surface plasmon resonance
All plasmon surface resonance experiments were performed using the BIACORE 3000. Purified recombinant human Follistatin and recombinant mouse Myostatin were purchased from R&D System (USA). Follistatin was immobilised onto the surface of a CM5 sensor chip (600 resonance units) using amine-coupling chemistry. A range of Myostatin concentrations were injected over the sensor chip surface at a flow rate of 30 μl/min at 25°C. Hepes-buffered saline (HBS: 10 mM Hepes, 150 mM NaCl, 3.4 mM EDTA, 0.005% Tween 20 pH 7.4) was used as a running buffer and for sample dilution. For controls, Myostatin was run over a derivatised sensor chip, which lacked Follistatin. All Myostatin curves were corrected by subtraction of the blank run. Biacore evaluation software was used for the mathematical fitting of experimental data.Preparation of chick embryos
Fertilised chick eggs were incubated at 38°C, and the embryos were staged according to Hamburger and Hamilton (1992). Experiments were performed on embryos at stages 22 to 24, re-incubated between 6 and 8 h, sacrificed and processed for whole-mount in situ hybridisation.Myostatin and Follistatin bead preparation and application to limb buds
Recombinant Myostatin and Follistatin protein were purchased from R&D Systems. Myostatin was applied to 80- to 120-μm Affigel beads and Follistatin to Heparin beads of same size (both Sigma, UK). The proteins were loaded onto beads as described by Cohn et al. (1995). Myostatin and Follistatin were used at 1 mg/ml concentration. For control bead implantation, beads were soaked in PBS only. For bead implantation, the dorsal ectoderm and mesenchyme of the right wing were punctured with an electrolytically sharpened tungsten needle, and beads were inserted into the punctured mesenchyme using a blunt glass needle. Beads were implanted at HH-stages stated in the text.Whole-mount in situ hybridisation
All chick embryos were washed in PBS and then fixed overnight in 4% paraformaldehyde at 4°C. Anti-sense RNA probes were labelled with digoxygenin, and whole-mount in situ hybridisation was performed as described by Nieto et al. (1996). The following probes were used in this study: Follistatin, full-length fragment, 1.1 kb (gift from Dr. Anthony Graham); Myostatin, 1 kb fragment (gift from Professor Se Jin Lee);MyoD, clone CMD9 full 1.5 kb length fragment (gift from Professor Bruce Patterson) and Pax-3, 645 bp fragment corresponding to nucleotides 468–1113 (gift from Dr Martin Goulding). Whole-mount embryos were wax or cryo-sectioned at a thickness of 15 μm for histological examination.Chick limb bud micromass assay
Micromass assays were carried out as described in Swalla and Solursh (1986). Briefly, limb buds from HH stages 21–22 chick embryos were dissected and placed into a 0.05% Trypsin-EDTA solution (Invitrogen). The ectoderm was removed using tungsten needles and a single cell suspension of the limb bud mesenchyme was obtained. Micromass cultures were plated at 2 × 10[SUP]5[/SUP] cells in 10 μl drops, allowed to adhere for 2 h and then treated with media (DMEM, 10% Fetal Calf Serum—Invitrogen) containing either Myostatin or Myostatin plus Follistatin at concentrations stated in the text (R&D Systems). Micromass cultures were then fixed in 4% PFA and processed for myosin heavy chain (MHC) immunocytochemistry. Cultures were dehydrated through a methanol series and treated with hydrogen peroxide to eliminate endogenous peroxidase activity. After rehydration, cultures were incubated with a monoclonal anti-PanMHC antibody (clone A4·1025, gift from Dr Simon Hughes) in the presence of 10% horse serum. After washing in PBS, rabbit anti-mouse biotin secondary antibody (Dako) was applied, cultures again washed, then incubated with an avidin–biotin complex (Vector Laboratories) and stained using a nickel enhanced DAB/hydrogen peroxide reaction (Vector Laboratories) according to manufacturer's protocols.Results
Co-expression of Follistatin and Myostatin related to muscle development
We have previously published independent detailed expression patterns of Follistatin and Myostatin during chick embryonic development Amthor et al., 1996, Amthor et al., 1999, Amthor et al., 2002a and Amthor et al., 2002b, which suggested an overlap in expression at sites of muscle development. Here, we directly compared the expression pattern of both genes during limb and trunk muscle development.Abutting or overlapping expression of Follistatin and Myostatin during wing bud development occurs for the first time at HH-stages 25–26. Follistatin is highly expressed in the proximal wing bud and fades distally (Fig. 1A). Myostatin is expressed in a central domain and the expression extends more distally compared toFollistatin (Fig. 1B). Transverse sections through the proximal part of the wing buds reveal expression ofFollistatin in the subectodermal mesenchyme and in a central domain of the dorsal premuscle mass (Fig. 1E). Myostatin is expressed in the mesenchymal core of the limb bud and also in a central domain of dorsal premuscle mass (Fig. 1F). However, at this stage, the expression of Follistatin and Myostatin does not encompass the entire premuscle mass as indicated by the expression of Pax-3 and MyoD (Figs. 1C and D).Follistatin expression is increased in the muscle masses up to HH-stage 31, whereas during these stages,Myostatin expression is only found in some subdomains of the developing muscle (Figs. 1G and H). Thereafter, at late embryonic stages, Follistatin is down-regulated in muscle, but expression resides in the connective tissue surrounding the muscles, whereas Myostatin is increasingly expressed in most of the muscles (data not shown, see also Amthor et al., 2002b).<dl class="figure" id="FIG1" data-t="f" style="border: 1px solid rgb(215, 215, 215); margin-right: 0px; margin-bottom: 15px; margin-left: 0px; padding: 6px; vertical-align: baseline; border-top-left-radius: 3px; border-top-right-radius: 3px; border-bottom-right-radius: 3px; border-bottom-left-radius: 3px; "><dt class="autoScroll" data-style="height:996px;width:629px;" style="border: 0px; margin: 12px 0px 0px; padding: 0px; vertical-align: baseline; overflow-y: hidden; overflow-x: auto; "></dt><dd id="labelCaptionFIG1" style="border: 0px; font-size: 0.8em; margin: 0px; padding: 0px; vertical-align: baseline; color: rgb(92, 92, 92); ">Fig. 1. Follistatin and Myostatin are expressed in or close to developing muscle. Comparison of Follistatin, Myostatin, Pax-3 andMyoD expression during wing and interlimb somite development. (A–D) HH-stage 26 wing buds, dorsal view. (E–H) Transverse sections of wing buds. (I–L) HH-stage 21 interlimb somites, lateral view and (M–P) corresponding frontal sections. (Q–T) Transverse sections of HH-stage 21 interlimb somites. (A) Strong proximal Follistatin expression, which fades distally in stage 26 wing bud. Section level indicated with broken green line. (B) Myostatin expression in a central domain of a stage 26 wing bud in contrast to the broader extension of the Pax-3-expressing (C) and MyoD-expressing (D) dorsal premuscle mass. (E) Transverse section of (A) shows Follistatin expression in a distinct location of the dorsal premuscle mass (arrow) and in the subectodermal mesenchyme (arrowhead). (F) Transverse section of (B) showsMyostatin expression in a distinct location of the dorsal premuscle mass (arrow) and in the mesenchymal core (arrowhead). (G and H) Follistatin and Myostatin expression in dorsal and ventral zeugopod muscles (arrows) at HH-stage 30. Follistatinexpression partly overlaps with Myostatin expression in distinct muscle subdomains (compare G and H, arrows). (I) StrongFollistatin expression at cranial and caudal somite edges and in a hypaxial domain in contrast to the high Myostatinexpression in the somite centre (J). Somite demarked by dotted line. (K) High Pax-3 expression in the hypaxial domain, moderate expression in the cranial and caudal domain of the dermomyotome (see also O) and no expression dorsomedially (see also S). (L) MyoD highlights the full extent of the myotome. (M) Frontal section shows high Follistatin expression in the cranial and caudal somite edges (arrowheads) and weak expression in the myotome. Brackets mark extent of somites. (N)Myostatin expression in the dermomyotomal and myotomal centre, but not at the cranial and caudal dermomyotomal edges nor in the dorsomedial and ventrolateral edges (R). (O) High Pax-3 expression in the cranial and caudal part of the dermomyotome (arrowheads). Intermediate expression in the dermomyotomal centre (arrow) and in the dorsal root ganglia (asterisks). (P) MyoD expression marks the cranio-caudal extent of the myotome. (Q) Transverse section reveals highFollistatin expression in the hypaxial domain of the somite with weak expression in the myotome. (R) Myostatin expression in the central part of the dermomyotome and myotome. (S) Highest level of Pax-3 expression in the hypaxial, moderate in the intermediate and no expression in the dorsomedial domain of the dermomyotome. Asterisk marks dorsal root ganglion. (T)MyoD expression marks the full extent of the myotome.
</dd><dd class="menuButtonLinks" style="border: 0px; font-size: 0.8em; margin: 0px; padding: 6px 0px 0px; vertical-align: baseline; text-align: right; position: relative; color: rgb(92, 92, 92); ">Figure options
</dd></dl>
Although Follistatin is expressed in interlimb somites as soon as they are formed, Myostatin is not up-regulated before HH-stage 19 (Amthor et al., 2002b). In stage 21 interlimb somites, Follistatin is strongly expressed at the cranial and caudal edges of the dermomyotome and in the hypaxial somite domain, whereas there is only a faint expression in the somite centre (Figs. 1I, M, Q). The expression of Myostatin is almost complementary to that of Follistatin as it is expressed in a central domain of the dermomyotome, but not at the somite edges or in the hypaxial domain (Figs. 1J, N, R). Highest expression level of Follistatincoincides with a high expression level of Pax-3 (Figs. 1K, O, S) but not of MyoD (Figs. 1L, P, T). At later stages, both Follistatin and Myostatin are predominantly expressed in hypaxial muscle (data not shown, see also Amthor et al., 2002b).These data show that Follistatin and Myostatin are expressed in or close to developing muscle. They are expressed partly in overlapping domains and partly in abutting regions to each other. As both Myostatin and Follistatin are secreted signalling proteins, they appear to be expressed sufficiently close to enable protein–protein interaction.
Last edited: