Involvement of Ras and Ral in Chemotactic Migration of Skeletal Myoblasts

Abstract

In skeletal myoblasts, Ras has been considered to be a strong inhibitor of myogenesis. Here, we demonstrate that Ras is involved also in the chemotactic response of skeletal myoblasts. Expression of a dominant-negative mutant of Ras inhibited chemotaxis of C2C12 myoblasts in response to basic fibroblast growth factor (bFGF), hepatocyte growth factor (HGF), and insulin-like growth factor 1 (IGF-1), key regulators of limb muscle development and skeletal muscle regeneration. A dominant-negative Ral also decreased chemotactic migration by these growth factors, while inhibitors for phosphatidylinositol 3-kinase and mitogen-activated protein kinase kinase (MEK) showed no effect. Activation of the Ras-Ral pathway by expression of an activated mutant of either Ras, the guanine-nucleotide dissociation stimulator for Ral, or Ral resulted in increased motility of myoblasts. The ability of Ral to stimulate motility was reduced by introduction of a mutation which prevents binding to Ral-binding protein 1 or phospholipase D. These results suggest that the Ras-Ral pathway is essential for the migration of myoblasts. Furthermore, we found that Ras and Ral are activated in C2C12 cells by bFGF, HGF and IGF-1 and that the Ral activation is regulated by the Ras- and the intracellular Ca²⁺-mediated pathways. Taken together, our data indicate that Ras and Ral regulate the chemotactic migration of skeletal muscle progenitors.

Migration of skeletal muscle precursor cells is a crucial step in skeletal muscle development and postlesional muscle regeneration. In vertebrate limb development, myogenic precursor cells actively migrate from the lateral lip of the dermomyotome into the limb buds, where they terminally differentiate (7, 48, 55). During migration, cells are prevented from myogenic differentiation (1, 27, 63). These processes seem to be governed by a diverse array of signaling factors that are secreted from adjacent tissues, including the limb buds (6, 15, 36). An initial determination of migratory limb muscle precursors is thought to be achieved by upregulation of several transcription factors such as Pax-3 (3). Pax-3 induces the expression of c-Met, a tyrosine kinase receptor for hepatocyte growth factor (HGF) (also known as scatter factor), leading to HGF-guided migration of myogenic precursor cells from the dermomyotome into the limb buds (5, 8, 23, 38). Fibroblast growth factors (FGFs), which regulate outgrowth and patterning of limbs (44, 52), act as chemoattractants for several kinds of limb bud cells, including myogenic precursor cells (4, 42, 72). Thus, HGF and FGFs are likely to induce the movement of myogenic precursors for the patternings of limb skeletal muscle. Besides HGF and FGFs, insulin-like growth factor 1 (IGF-1) has been suggested to be involved in limb muscle formation (12–14), although it is unknown whether IGF-1 can act as a chemoattractant. These three factors also seem to be involved in postlesional regeneration of skeletal muscle, which requires the migration of skeletal muscle satellite cells (9, 16, 19, 25, 28, 40, 41, 43, 71).

Ras (H-, K-, or N-Ras) is one of a number of small GTP-binding proteins that function as molecular switches by cycling between active GTP- and inactive GDP-bound states (32). The ratio of the two states is regulated positively by guanine-nucleotide exchange factors (GEFs) and negatively by GTPase-activating proteins (GAPs). A variety of growth factors coupled to receptor tyrosine kinases induce Ras activation. Ras activates multiple effectors, including Raf (i.e., A-Raf, B-Raf, and Raf-1), Ral GEF (i.e., RalGDS, Rlf, and Rgl), and phosphatidylinositol 3-kinase (PI3K) (69, 75), and plays an important role in several cellular processes, such as proliferation and differentiation.

Recent reports suggest the involvement of Ras in cell motility. For instance, cell migration in wound healing, FGF-stimulated motility of endothelial cells, platelet-derived growth factor-stimulated chemotaxis of fibroblasts, and HGF-induced scattering of epithelial cells are dependent on Ras activity (17, 22, 34, 60, 62). A Dictyostelium mutant lacking Ras GEF or a subtype of Ras has been reported to have a defect in cell movement (26, 68). The Caenorhabditis elegans ras homologue, let-60, mediates a gonadal signal required for the migration of sex myoblasts (65). For Drosophila melanogaster, expression of a dominant-negative Ras mutant inhibits the migration of the border cells during oogenesis (39). In addition, an activated mutant of Ras partially substitutes for breathless, a Drosophila FGF receptor homologue, in the migration of tracheal cells (59). Thus, Ras seems to function as a crucial regulator of cell motility. However, Ras involvement in migration of mammalian skeletal muscle cells has not been investigated. In this study, to examine the role of Ras in the migratory response of skeletal muscle precursors, we utilized C2C12 myoblasts (78) as an in vitro model system, since it has been demonstrated that, when implanted into mouse muscles, C2C12 cells can migrate toward areas of muscle injury and regeneration to form myofibers (71). Here, we demonstrate that Ras and its downstream molecule, Ral, are involved in the regulation of myogenic cell migration.

MATERIALS AND METHODS

Plasmids.

Complementary DNA of raf-1 was kindly provided by Martin McMahon (University of California, San Francisco). The expression plasmids pBJ/Myc-RalGDS-CAAX and pBJ/Myc-RalBP1 were generous gifts from Akira Kikuchi (Hiroshima University, Hiroshima, Japan). Plasmids pOPRSV1-H-ras^G12V and pOPRSV1-H-ras^S17N were constructed as described elsewhere (66). To construct pOPRSV1-ralA^S28N, the fragment carrying FLAG-His₆-tagged ralA^S28N was inserted into pOPRSV1-CAT (Stratagene, La Jolla, Calif.). The DNA fragments of the Ras-binding domain (RasBD) of Raf-1 (amino acids 51 to 131) and the Ral-binding domain (RalBD) of Ral-binding protein 1 (RalBP1) (amino acids 397 to 518) were generated by PCR and inserted into pGEX-2T to produce pGEX-RasBD and pGEX-RalBD, respectively. The fragments of H-ras^G12V,E37G, ral^G23V, ral^G23V,D49N, and ral^G23V,ΔN were generated by site-directed mutagenesis from H-ras^G12V or the wild-type ralA and subcloned into pCMV5 (2).

Reagents.

We purchased anti-H-Ras antibody (sc-520) from Santa Cruz Biotechnology (Santa Cruz, Calif.), anti-Myc (9E10) antibody from BabCO (Richmond, Calif.), anti-active mitogen-activated protein kinase (MAPK) antibody (V8031) and U0126 from Promega Corporation (Madison, Wis.), anti-RalA antibody (R23520) and anti-Ras antibody (R02120) from Transduction Laboratories (Lexington, Ky.), human IGF-1 from Life Technologies (Rockville, Md.), human basic FGF (bFGF) from Pepro Tech EC (London, England), HGF from Becton Dickinson Labware (Bedford, Mass.), A23187 and LY294002 from Calbiochem (San Diego, Calif.), and 1,2-bis(2-aminophenoxy)-ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester) (BAPTA-AM) from Dojindo (Kumamoto, Japan).

Cell culture and transfection.

C2C12 cells were obtained from RIKEN Cell Bank (Ibaraki, Japan). C2C12 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 15% (vol/vol) fetal bovine serum. For stable transfection, pOPRSV1-H-ras^G12V, -H-ras^S17N, or -ralA^S28N was introduced into C2C12 cells with p3′SS (Stratagene) by using Lipofectamine (Life Technologies). One day after transfection, transfectants were isolated in the culture medium containing 400 μg of G418 per ml and 200 μg of hygromycin per ml. For transient transfection, plasmids of interest (2.5 μg) were introduced into subconfluent C2C12 cells in a 60-mm dish with Lipofectamine and Lipofectamine Plus (Life Technologies).

Human normal skeletal muscle cells were obtained from BioWhittaker (Walkersville, Md.) and were cultured in skeletal muscle basal medium (BioWhittaker) supplemented with 10 ng of epidermal growth factor per ml, 100 ng of insulin per ml, 500 μg of feutin per ml, 39 μg of dexamethasone per ml, and antibiotics.

Preparation of cell lysate.

Harvested cells were lysed in lysis buffer (20 mM HEPES-NaOH [pH 7.2], 150 mM NaCl, 10% [vol/vol] glycerol, 0.5% [vol/vol] Triton X-100, 10 mM MgCl₂, 1 mM EDTA, 1 mM EGTA, 10 mM NaF, 1 mM Na₃VO₄, 25 mM β-glycerophosphate, 2 μg of aprotinin per ml, 1 μg of leupeptin per ml, and 1 μg of pepstatin A per ml). After centrifugation at 12,000 × g, the supernatant was saved as cell lysate. The protein concentration of the cell lysate was determined using protein assay CBB solution (Nakalai Tesque, Kyoto, Japan).

Pull-down assay.

Cell lysates (500 μg of protein) were incubated for 40 min at 4°C with 20 μl of glutathione-Sepharose 4B beads (Amersham Pharmacia Biotech, Uppsala, Sweden) and 10 μg of glutathione S-transferase (GST)–RasBD or GST-RalBD that had been produced in Escherichia coli and purified according to the procedure recommended by the manufacturer. Then the beads were washed three times with 20 mM HEPES-NaOH (pH 7.4)–150 mM NaCl–0.25% (vol/vol) Triton X-100–5 mM MgCl₂ and boiled in the sample buffer. Samples were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (15% acrylamide), followed by Western blotting analysis using anti-Ras or anti-RalA antibody.

Migration assay and motility assay.

Prior to the experiments, Falcon cell culture inserts (Becton Dickinson Labware) were coated with 10 μg of E-C-L Cell Attachment Matrix (Upstate Biotechnology) per ml. Trypsinized cells were washed once with the culture medium and suspended in DMEM containing 0.1% bovine serum albumin (BSA) at a concentration of 1 × 10⁵ cell/ml. DMEM containing 0.1% BSA (750 μl) was added to a Falcon companion plate (24 wells; Becton Dickinson Labware). In the migration assay, growth factors were also added to the companion plate. Immediately after the cell culture inserts were placed on the companion plate, the cell suspensions (500 μl) were added to the inserts. Samples were then incubated for 2 h at 37°C in a 10% CO₂ atmosphere. After incubation, the nonmigrated cells were wiped away from the upper surface of the membrane of the insert, and the migrated cells on the lower surface of the membrane were stained with Diff-Quik (International Reagents Corp., Kobe, Japan). The number of stained cells was counted in five microscope fields.

Adhesion assay.

Trypsinized cells were washed once and suspended in DMEM containing 0.1% BSA at a concentration of 5 × 10⁵ cells/ml. The cell suspensions (100 μl) were added to a 96-well dish which was precoated with 10 μg of E-C-L Cell Attachment Matrix per ml. After 30 min of incubation at 37°C in a 10% CO₂ atmosphere, the suspensions were aspirated and the wells were washed three times with phosphate-buffered saline. Then, adherent cells were stained with 0.1% (wt/vol) crystal violet in 20% (vol/vol) methanol for 5 min at room temperature. The stain was eluted with 100 μl of 50% (vol/vol) ethanol, and the absorbance at 595 nm was measured.

RESULTS

Chemotactic response of C2C12 myoblasts to growth factors.

In order to study the chemotactic behavior of C2C12 myoblasts in response to growth factors, we performed an in vitro migration assay with the multiwell chemotaxis chamber. In this assay, cells in an upper chamber migrate through an extracellular matrix protein (ECM)-coated nucleopore filter to a lower chamber, which includes a higher concentration of chemotactic factors. In preliminary experiments, we observed that C2C12 myoblasts migrated to bFGF, HGF, IGF-1, and platelet-derived growth factor but that they migrated hardly at all to transforming growth factor-β, interleukin-15, or bone morphogenetic protein-4 (data not shown). For further studies, we selected bFGF, HGF, and IGF-1, whose involvement in skeletal muscle development and regeneration had been indicated. As shown in Fig. 1A, bFGF, HGF, and IGF-1 significantly stimulated migration of C2C12 myoblasts. Among them, IGF-1 showed the highest chemotactic activity toward C2C12 cells. We carried out a checkerboard analysis (79) and confirmed that the observed migratory response is mainly chemotaxis, a directional migration along a gradient of factors (data not shown). Since IGF-1 has not been reported to be a chemoattractant for myoblasts, we confirmed our results using primary skeletal muscle satellite cells, skeletal muscle progenitors in adult myofibers (63). Consequently, satellite cells also showed the capability to migrate toward all the growth factors (Fig. 1B). As in the case of C2C12 myoblasts, the highest chemotactic activity was observed for IGF-1. The results suggest that IGF-1, as well as bFGF and HGF, is a chemotactic factor for skeletal muscle precursors.

FIG. 1.

Chemotaxis of C2C12 myoblasts and primary skeletal muscle satellite cells in response to bFGF, HGF, and IGF-1. A migration assay (as described in Materials and Methods) was performed to determine chemotactic migration of C2C12 cells (A) and human satellite cells (B). The concentrations of the growth factors bFGF, HGF, and IGF-1 were 10, 15, and 10 ng/ml, respectively. The data are expressed as the fold increase in the number of migrated cells relative to the number of migrated cells in the absence of factor and are the means ± the standard errors (SE) of at least three independent experiments.

Ras and Ral are involved in chemotaxis of myoblasts.

To test the possible involvement of Ras in the chemotactic migration of myoblasts, we established several C2C12 clones that express a dominant-negative mutant of H-Ras (Ras^S17N) in an isopropyl-β-d-thiogalactoside (IPTG)-dependent manner. One of these clones, designated C2-Ras^S17N, was used for further analyses, since this clone expresses Ras^S17N in excess of endogenous Ras in the presence of IPTG (Fig. 2). After induction of Ras^S17N, a migration assay was performed for bFGF, HGF, and IGF-1. Expression of Ras^S17N significantly inhibited migration of myoblasts in response to all factors (Table 1). Since cell adhesiveness toward ECM was not affected by Ras^S17N expression (Fig. 3A), the inhibition of migration by Ras^S17N was not due to the reduction of initial attachment of cells to the ECM-coated filters. Therefore, it is suggested that Ras is required for the migration of C2C12 myoblasts in response to bFGF, HGF, and IGF-1.

FIG. 2.

IPTG-induced expression of Ras^S17N and Ral^S28N in C2C12 myoblasts. C2-Ras^S17N and C2-Ral^S28N cells were cultured for 1 day in the presence (+) or absence (−) of IPTG (5 mM). Then the cells were lysed, and the lysates (25 μg of protein) were subjected to Western blotting analysis with anti-H-Ras and anti-RalA antibodies to detect Ras^S17N and Ral^S28N, respectively.

TABLE 1.

Effect of Ras^S17N, Ral^S28N, MEK inhibitor, or PI3K inhibitor on migration of C2C12 myoblasts in response to bFGF, HGF, and IGF-1

Growth factor (concn)	% Migrating cellsa
	RasS17Nb	RalS28Nb	U0126c	LY294002c
bFGF (10 ng/ml)	69.3 ± 9.0	53.6 ± 15.1	135.7 ± 15.4	108.4 ± 10.1
HGF (15 ng/ml)	52.4 ± 10.1	13.9 ± 8.4	97.0 ± 7.0	121.2 ± 17.8
IGF-1 (10 ng/ml)	56.5 ± 3.9	46.1 ± 7.8	107.0 ± 5.0	122.7 ± 7.2

^aResults are the means ± the SE of at least three separate experiments. 

^bC2-Ras^S17N and C2-Ral^S28N cells were cultured in the presence or absence of IPTG (5 mM) for 1 day and then subjected to migration assay. The cell number obtained in the absence of IPTG was set to 100%. 

^cU0126 or LY294002 was added to both lower and upper chambers at a concentration of 10 μM on initiation of the assay. The cell number obtained in the absence of U0126 or LY294002 was set to 100%. 

FIG. 3.

Expression of Ras^S17N or Ral^S28N does not affect cell adhesion. C2-Ras^S17N (A) and C2-Ral^S28N (B) cells were cultured for 1 day in the presence (+) or absence (−) of IPTG (5 mM). Then a cell adhesion assay was performed as described in Materials and Methods. The cell number obtained in the absence of IPTG was set to 100%. Results are the means ± the SE of at least three experiments.

Ras is known to activate at least three pathways, the Raf–mitogen-activated protein kinase kinase (MEK)–extracellular signal-regulated kinase (ERK), the PI3K-Akt, and the Ral GEF-Ral pathways. To examine the involvement of Raf- and PI3K-exerted signaling pathways in the chemotaxis of C2C12 cells, we utilized specific inhibitors for MEK (U0126) and PI3K (LY294002). Neither 10 μM U0126 nor 10 μM LY294002 exhibited a substantial effect on the chemotactic migration of C2C12 cells (Table 1). We confirmed that, at these concentrations, the inhibitors could abolish IGF-1-induced phosphorylation of ERK and Akt even after a 2-h treatment of cells, suggesting that the activity of ERK and Akt was suppressed throughout the migration assay (data not shown). Therefore, the Raf-ERK and the PI3K-Akt pathways may not be essential for C2C12 migration toward bFGF, HGF, and IGF-1. We next determined whether the Ral GEF-Ral pathway is involved in the chemotactic response. For this purpose, we generated a C2C12 clone (C2-Ral^S28N) that inducibly expresses a dominant-negative RalA (Ral^S28N) (Fig. 2). As shown in Table 1, overexpression of Ral^S28N reduced chemotactic migration by bFGF, HGF, and IGF-1. Cell adhesion onto ECM was not affected by Ral^S28N expression (Fig. 3B). These results suggest that activation of the Ral GEF-Ral pathway downstream of Ras is essential for the chemotactic migration of C2C12 myoblasts.

Activation of Ras and Ral stimulates motility of myoblasts.

We investigated whether activation of the Ras-Ral GEF-Ral pathway stimulates motility of C2C12 myoblasts. First, we examined the effect of an activated mutant of H-Ras (Ras^G12V) on cell motility. When Ras^G12V was transiently expressed in C2C12 cells, transfected cells showed higher motility (Fig. 4A), suggesting that Ras activation leads to an increase of cell motility in C2C12 cells. Second, we tested the ability of Ral GEF to stimulate the motility of C2C12 cells. For this, we used two mutants: Ras^G12V,E37G, an effector-domain mutant of Ras that is capable of activating Ral GEFs but not Raf and PI3K (74), and RalGDS-CAAX, an activated mutant of RalGDS (45). When these mutants were transiently expressed in C2C12, both transfected myoblasts displayed higher motility than the mock transfectant (Fig. 4B and C). Finally, we examined the effect of a constitutively activated RalA (Ral^G23V) and found that transient expression of Ral^G23V in C2C12 myoblasts also stimulated cell motility (Fig. 4D). These data suggest that activation of the Ras-Ral GEF-Ral pathway can increase the motility of C2C12 cells.

FIG. 4.

Activation of the Ras-Ral GEF-Ral pathway stimulates cell motility. C2C12 cells were transfected with pOPRSV1-H-Ras^G12V (A), pCMV5-ras^G12V,E37G (B), pBJ-Myc/RalGDS-CAAX (C), or pCMV5-ral^G23V, -ral^G23V,D49N, or -ral^G23V,ΔN (D) and subjected to migration assay. The values are expressed as the fold increase in the number of the migrated cells and are the means ± the SE of three independent experiments. As shown at the bottom of each panel, expression of each protein was confirmed by Western blotting analysis using anti-H-Ras (for panels A and B), anti-Myc (for panel C), and anti-RalA (for panel D) antibodies.

It was reported that Ral binds to its effector molecules, RalBP1 (also known as RLIP76, RIP1, or cytocentrin) (10, 31, 56, 58) and phospholipase D (PLD) (29), through the effector region and the amino-terminal 11 amino acids, respectively. Thus, we examined whether these two molecules are involved in cell motility using two Ral mutants: Ral^G23V,S49N, an effector region mutant of Ral lacking affinity to RalBP1 (10); and Ral^G23V,ΔN, which cannot bind to PLD due to a deletion of the amino-terminal 11 amino acids (29). Consequently, C2C12 myoblasts expressing either of the above two Ral mutants exhibited much lower motility than those expressing Ral^G23V (Fig. 4D). Therefore, it is possible that both RalBP1 and PLD are involved in Ral-mediated cell migration.

Ras and Ral are activated by chemotactic signals.

In order to investigate whether Ras and Ral are indeed activated in C2C12 cells during chemotactic migration, we performed a pull-down assay to measure the number of the GTP-bound forms of Ras and Ral within C2C12 cells. Figure 5 shows that Ras-GTP and Ral-GTP accumulated in C2C12 cells in response to bFGF, HGF, and IGF-1, suggesting that Ras and Ral are activated by these factors. However, the extents of Ral activation by these three factors were not parallel to those of Ras activation. Activation of Ras was strongly stimulated by bFGF but stimulated only moderately by HGF and IGF-1. On the other hand, IGF-1 showed the highest activity for Ral activation. These data raise the possibility that Ral may be activated, in part, in a Ras-independent manner.

FIG. 5.

Basic FGF, HGF, and IGF-1 induce activation of Ras and Ral. C2C12 cells were starved for 2.5 h in DMEM containing 0.1% BSA and stimulated by bFGF (50 ng/ml), HGF (50 ng/ml), or IGF-1 (50 ng/ml) for 6 min at 37°C. Then cells were lysed, and the amounts of Ras-GTP or Ral-GTP in the lysates were determined by pull-down assay, as described in Materials and Methods.

We thus examined whether Ral activation by the growth factors is mediated by Ras in C2C12 myoblasts. IGF-1 was chosen for this analysis since it possesses the greatest ability to stimulate the Ral activation, as well as the migration of C2C12 cells, among the three factors. In C2-Ras^S17N cells, IPTG-induced expression of Ras^S17N inhibited Ral activation by IGF-1, though only partially (Fig. 6A). Western blotting analysis confirmed that Ras^S17N expression did not affect the amount of endogenous Ral (Fig. 6A). The inhibition was not due to the downregulation of IGF-1 receptor by Ras^S17N, since the expression of IGF-1 receptor and the IGF-1-induced tyrosine phosphorylation of IGF-1 receptor were not affected by IPTG treatment (data not shown). Therefore, the results suggest that IGF-1-induced activation of Ral is mediated by Ras. The reason for the partial inhibition by Ras^S17N was not insufficient expression of this dominant-negative mutant, because IPTG treatment of C2-Ras^S17N cells inhibited almost completely the IGF-1-induced phosphorylation of ERK (Fig. 6A). Thus, our data suggest that there is a Ras-independent pathway, as well as the Ras-dependent pathway, downstream of the IGF-1 receptor for Ral activation.

FIG. 6.

IGF-1 activates Ral through Ras- and Ca²⁺-dependent pathways. (A) Dominant-negative Ras inhibits IGF-1-induced Ral activation. C2-Ras^S17N cells were cultured in the presence (+) or absence (−) of IPTG (5 mM) for 1 day and starved for 2.5 h. Cells were stimulated with IGF-1 (50 ng/ml), and the activity of Ral and the phosphorylation of ERK were determined by pull-down assay and by immunoblotting with anti-active MAPK antibody, respectively. The amount of endogenous RalA was verified using anti-RalA antibody. (B) Ral activation by IGF-1 is inhibited by deprivation of intracellular Ca²⁺. After starvation for 2.5 h, C2C12 cells were treated with BAPTA-AM (25 μM) for 15 min at 37°C. Then cells were stimulated with IGF-1 (50 ng/ml), and the activities of Ral (top) and Ras (bottom) were measured. (C) Ras^G12V and Ca²⁺ ionophore cooperatively activate Ral. C2-Ras^G12V cells were cultured in the presence of IPTG (5 mM) for 16 h. After starvation, cells were incubated with A23187 (500 nM) for 2.5 min, and the amount of Ral-GTP in cell lysates was determined. Relative density of the spots of Ral-GTP on the film was quantitated by the NIH Image program. (Bottom) The amount of endogenous RalA was verified using anti-RalA antibody. Results are representative of at least three independent experiments.

It has been reported that an increase in intracellular Ca²⁺ leads to Ral activation through a Ras-independent pathway (24, 49, 76, 80). Since IGF-1 induces the increase of intracellular Ca²⁺ in C2C12 cells (data not shown), we explored the possibility that Ral activation by IGF-1 is mediated by Ca²⁺. When cells were treated with an intracellular Ca²⁺ chelator, BAPTA-AM, Ral activation by IGF-1 was partially suppressed (Fig. 6B). The suppression was not through the inhibition of Ras activation, since BAPTA-AM treatment conversely augmented IGF-1-induced Ras activation (Fig. 6B, bottom). The results suggest that the IGF-1-induced Ral activation is mediated also by Ca²⁺ increase.

Based on the present data, it seems that IGF-1 activates Ral through at least two pathways: the Ras-dependent pathway and the Ca²⁺-dependent pathway. To confirm this, we investigated the effect of the activated mutant of Ras and a Ca²⁺ ionophore, A23187, on Ral activation. In C2-Ras^G12V cells, which express Ras^G12V in an IPTG-dependent manner (data not shown), either Ras^G12V or A23187 alone induced Ral activation slightly, whereas the combination of two stimuli resulted in a significant activation of Ral (Fig. 6C). We confirmed that the amount of endogenous Ral was not affected by the expression of Ras^G12V (Fig. 6C) and that A23187 treatment showed no effect on the basal activity of Ras (data not shown). Therefore, it is likely that Ras and intracellular Ca²⁺ act coordinately in the activation of Ral by IGF-1.

DISCUSSION

In this study, we identified IGF-1 as a new chemoattractant for myoblasts. The dominant-negative mutants of Ras and Ral inhibited chemotaxis of C2C12 cells in response to bFGF, HGF, and IGF-1, and activation of the Ras-Ral GEF-Ral pathway stimulated myoblast motility. All three growth factors induced activation of Ras and Ral. These results indicate that Ras and Ral are involved in the migration of skeletal muscle precursor cells towards bFGF, HGF, and IGF-1.

Previously, the role of Ras in cell migration was studied with epithelial cells. As is the case with myoblasts, Ras is required for the locomotion of epithelial cells by HGF (22, 60). However, the signaling pathways downstream of Ras for chemotaxis seem different between myoblasts and epithelial cells. Epithelial cells require activation of PI3K and ERK for HGF-stimulated migration (33, 57, 61). On the other hand, inhibitors for PI3K and MEK had little effect on the chemotactic migration of C2C12 cells (Table 1). Furthermore, a combination of two inhibitors did not further inhibit the migration of C2C12 cells expressing the dominant-negative Ral (J. Suzuki, Y. Kaziro, and H. Koide, unpublished data). These results suggest that, although PI3K and ERK may somehow be involved in chemotactic response of C2C12 cells, their contribution is not as large as that of Ral and that the signaling pathways for the Ras-mediated migration vary among cell types.

Here, we demonstrated that the dominant-negative mutant of Ral inhibits the chemotactic migration of C2C12 cells and that ectopic expression of Ral^G23V stimulates motility in the absence of chemotactic factors. Furthermore, the profile of Ral activation by FGF, HGF, and IGF-1 corresponded well to the ability of these factors to induce myoblast migration. These results strongly indicate that Ral is involved in the chemotactic migration of mammalian myoblast cells. Since Ral has been implicated in border cell migration during Drosophila oogenesis (39), Ral involvement in cell motility may be evolutionally preserved. The present data also suggest the possibility that Ral may participate in the chemotaxis of other cell types, such as lymphocytes, since Ral is activated by chemoattractants for lymphocytes, such as formyl Met-Leu-Phe, platelet-activating factor, and lysophosphatidic acid (24, 49, 76, 77).

The mechanism underlying the Ral-stimulated motility of myoblasts is presently unknown. Thus far, three types of putative effectors for Ral have been identified: RalBP1, PLD, and filamin (10, 29, 31, 51, 56, 58). Experiments with the Ral mutants suggest that RalBP1 and PLD are possibly involved in motility. RalBP1 has the GAP domain for Rac1 and Cdc42, well-known regulators of the actin cytoskeleton (10, 31, 56), and PLD activation results in actin stress fiber formation (11). Accordingly, to stimulate the cell motility of C2C12 cells, Ral may modulate actin cytoskeleton dynamics by controlling the activities of RalBP1 and PLD. On the other hand, a recent paper suggests that RalBP1 regulates endocytosis of epidermal growth factor and insulin receptors (50). Endocytosis is an essential process in cell motility, since recycling of integrins, receptors for ECM, from the rear of the cell to its leading edge requires endocytosis of integrins at the rear end (37). PLD is also known to be involved in membrane trafficking and vesicle transport (30). Therefore, it is also possible that Ral may stimulate cell motility by recycling membrane proteins, including integrins. As for filamin, its involvement in the Ral-dependent cell movement is unknown. However, since filamin is a downstream intermediate in Cdc42-mediated filopodia formation (51), Ral may also utilize filamin for cell migration.

We showed that IGF-1 utilizes both the Ras- and the Ca²⁺-mediated pathways for the regulation of Ral activity. Since HGF and bFGF are reported to induce the increase of intracellular Ca²⁺ (46, 47, 67), these growth factors may activate Ral through the Ras- and the Ca²⁺-dependent pathways. The use of the Ca²⁺-dependent pathway, in addition to the Ras-dependent pathway, for Ral activation may be the reason why the level of Ral activation by the growth factors does not correlate to Ras activation. We speculate that the biological significance of the differential activation between Ras and Ral is that, besides the regulation of cell motility through Ral, Ras plays extra roles in migrating myoblasts through the other effectors. For example, bFGF, which only slightly activates Ral in spite of a marked activation of Ras (Fig. 5), strongly inhibits skeletal muscle differentiation (54; Suzuki et al., unpublished data). Since Ras is a strong inhibitor of myogenic differentiation (53), it is possible that bFGF activates Ras not only to increase cell motility but also to inhibit myogenesis.

Several mechanisms of Ca²⁺-dependent Ral activation have been suggested. It has been reported that Rap1 activation is rapidly induced by the increase of Ca²⁺ concentration (18, 76). Since Rap1 can activate Ral GEF in a certain type of cell (80), it is possible that Ca²⁺-mediated activation of Rap1 leads to the activation of Ral in C2C12 cells. Alternatively, it is also possible that Ral is activated by interaction with Ca²⁺-calmodulin, although Ca²⁺-calmodulin is a less effective GEF for Ral than Ral GEFs are (70).

IGF-1-induced Ras activation was found to be effectively promoted by BAPTA-AM treatment (Fig. 6B). Furthermore, it was also found that the addition of BAPTA-AM to C2C12 cells by itself can induce Ras activation (Suzuki et al., unpublished data). The precise reason for the observed activation of Ras in the presence of BAPTA-AM is presently unknown. However, since Ras GAPs possess the Ca²⁺-dependent phospholipid-binding domain (20), it is possible that BAPTA-AM treatment might interfere with the activities of Ras GAPs.

We demonstrated that bFGF, HGF, and IGF-1 act as chemotactic factors for C2C12 cells and primary human muscle satellite cells. It has been reported that FGFs and HGF act as chemoattractants for myoblasts in vivo. Expression of FGFs is observed in the limb buds, including the apical ectodermal ridge and the underlying mesenchyme (44), and interaction between FGFs and their receptors is important for myogenic cell migration from somites to limb buds (27). Expression of HGF is observed in the limb bud along the migrating routes of myogenic progenitors (5). Furthermore, c-met-deficient mice show a defect in migration of myogenic progenitors into the limb buds, which causes complete loss of skeletal muscle of the limb (5). Thus, it is likely that FGFs and HGF are guide molecules for skeletal muscle precursor cells from the dermomyotome to the limb buds (38, 72). Unlike bFGF and HGF, IGF-1 has not been shown to be a chemoattractant for myoblasts, and the biological significance of IGF-1-stimulated myoblast migration is unknown. It has been reported that, like FGFs, IGF-1 is expressed in the apical ectodermal ridge and the underlying mesenchyme (13, 21, 64). In addition, IGF-1 is required for FGFs to promote limb bud outgrowth (12). Therefore, IGF-1 may induce distal migration of myogenic cells in the limb buds during skeletal muscle development. Another possibility is that IGF-1-induced migration may be involved in the postlesional regeneration of skeletal muscle. In injured skeletal muscle, several growth factors, including IGF-1, are locally produced by regenerating muscle cells and inflammatory cells to regulate proliferation and differentiation of satellite cells (9, 16, 19, 25, 28, 40, 41, 43). Thus, IGF-1 may attract satellite cells into muscle regenerating areas.