N-WASP and WAVE2 acting downstream of phosphatidylinositol 3-kinase are required for myogenic cell migration induced by hepatocyte growth factor.

During skeletal muscle regeneration caused by injury, muscle satellite cells proliferate and migrate toward the site of muscle injury. This migration is mainly induced by hepatocyte growth factor (HGF) secreted by intact myofibers and also released from injured muscle. However, the intracellular machinery for the satellite cell migration has not been elucidated. To examine the mechanisms of satellite cell migration, we utilized satellite cell-derived mouse C2C12 skeletal muscle cells. HGF induced reorganization of actin cytoskeleton to form lamellipodia in C2C12 myoblasts. HGF treatment facilitated both nondirectional migration of the myoblasts in phagokinetic track assay and directional chemotactic migration toward HGF in a three-dimensional migration chamber assay. Endogenous N-WASP and WAVE2 were concentrated in the lamellipodia at the leading edge of the migrating cells. Moreover, exogenous expression of wild-type N-WASP or WAVE2 promoted lamellipodial formation and migration. By contrast, expression of the dominant-negative mutant of N-WASP or WAVE2 and knockdown of N-WASP or WAVE2 expression by the RNA interference prevented the HGF-induced lamellipodial formation and migration. When the cells were treated with LY294002, an inhibitor of phosphatidylinositol 3-kinase, the HGF-induced lamellipodial formation and migration were abrogated. These results imply that both N-WASP and WAVE2, which are activated downstream of phosphati-dylinositol 3-kinase, are required for the migration through the lamellipodial formation of C2C12 cells induced by HGF.

Migration of myogenic cells is essential for myogenesis in vertebrates during embryonic development. Particularly, muscle precursor cells delaminate from hypaxial dermomyotome and migrate long distance to sites of muscle formation including limbs, tongue, and diaphragm (1)(2)(3). Myogenic cell migra-tion is not restricted to the embryonic stage but required for muscle regeneration in adult. Adult skeletal muscle contains satellite cells, a specialized population of myogenic stem cells (4 -8). The satellite cells are situated adjacently to a mature myofiber and surrounded by basal lamina together with the myofiber. They are mitotically quiescent in normal adult muscle but activated to proliferate in response to diverse stimuli such as injury, denervation, exercise, or stretching. These cells migrate to the injured regions of myofibers by chemotaxis within or beyond the basal lamina and proliferate. Then they fuse each other or with preexisting myotubes or myofibers to regenerate muscle.
The mechanisms of satellite cell activation, proliferation, and migration have been under examination. Several growth factors and cytokines, which are released from damaged myofibers, infiltrating macrophages, or satellite cells by themselves, are involved in particular steps of a series of these processes during regeneration (5,6). Among them, hepatocyte growth factor (HGF) 1 /scatter factor have been shown to induce multiple these steps, i.e. satellite cell activation (9,10), proliferation (11,12), and chemotactic migration (12,13). HGF is present in uninjured myofibers and may be secreted from them to the extracellular matrix surrounding myofibers. Upon muscle injury, it may be released from myofibers and the extracellular matrix (5,9). In addition, it is likely to be secreted at least in part by satellite cells and quickly acts on satellite cells in an autocrine manner in response to muscle injury (9,14). The quiescent satellite cells already express the HGF receptor Met, whereas the MyoD family muscle regulatory factors, Myf5 and MyoD, are expressed after the satellite cell activation, and myogenin and MRF4 are expressed after the beginning of their differentiation program (15).
HGF induces via its receptor Met a variety of cellular responses including cell proliferation, chemotactic migration, invasion, branching morphogenesis, and cell survival in diverse or particular types of cells (16,17). Molecular aspects of the HGF-Met signaling pathways for specific cellular responses have been defined (18 -20). When HGF binds to Met receptor tyrosine kinase, its cytoplasmic Tyr residues are autophosphorylated and bind the scaffolding adaptor protein Gab1 as well as the adaptor protein Grb2. For migration or scattering of MDCK epithelial cells, the signaling pathways lead to the activation of phosphatidylinositol 3-kinase (PI3K) and Ras-ERK mitogen-activated protein kinase cascade (21)(22)(23). The signaling results in the activation of the Rho family small GTPases, Rac1 and Cdc42, although signaling pathways leading to the activation of Cdc42 have not been elucidated (24 -26).
Cdc42 as well as phosphatidylinositol 4,5-bisphosphate (PI(4,5)P 2 ) binds to its target protein N-WASP and induces rapid actin polymerization through activating Arp2/3 complex, resulting in filopodial formation in migrating cells (27)(28)(29). By contrast, Rac1 activates WAVE2, a ubiquitously expressed member of WAVE proteins (30), indirectly via the target protein IRSp53 (31) or by binding to the PIR121/Sra1-Nap1-Abi1-WAVE2 complex (32). Rac1 can also activate WAVE1 via this huge protein complex (33). The binding to WAVE2 of phosphatidylinositol 3,4,5-trisphosphate (PI(3,4,5)P 3 ), which is produced by PI3K, is necessary for the lamellipodial formation, probably by recruiting WAVE2 to the polarized plasma membrane (34). Activated WAVE proteins cause branched actin filament formation through the activation of Arp2/3 complex, leading to the lamellipodial formation (30,31,(35)(36)(37)(38). The formation of filopodia and that of lamellipodia are both required for cell migration of certain types of cells such as fibroblasts under two-dimensional culture conditions (39 -41). Although cell migration in vivo occurs in most cases in threedimensional extracellular matrix, the mechanisms of cell migration under three-dimensional conditions have not been well examined. Recent investigations have revealed that N-WASP plays important roles in HGF-induced three-dimensional migration, invasion, and tubulogenesis in MDCK cells (42). In mouse embryonic fibroblasts, WAVE2 is essential for the lamellipodial formation at the leading edge for directional chemotactic migration induced by platelet-derived growth factor (PDGF) regardless of the presence of extracellular matrix, whereas WAVE1 is required for the migration in extracellular matrix (43).
Although the roles of HGF and Met in satellite cells during muscle regeneration are evident, the signaling mechanisms for each step in regeneration remain to be determined. To elucidate the signal transduction pathways for the migration of satellite cells, we applied the mouse satellite cell-derived C2C12 cells. The myoblasts treated with HGF formed lamellipodia at the leading edges and migrated toward HGF by chemotaxis. Here we show that both N-WASP and WAVE2, which act downstream of PI3K, play essential roles in the lamellipodial formation and migration of C2C12 cells induced by HGF.

EXPERIMENTAL PROCEDURES
Cell Culture-Mouse skeletal muscle cell line C2C12 (44) is a subclone of C2 cells (45) derived from satellite cells. The cells were maintained as proliferating myoblasts in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS) (growth medium). For the stimulation with HGF, the cells were first subjected to serum starvation by culturing for 24 h in the Dulbecco's modified Eagle's medium containing 0.5% FBS (low serum medium) and then treated with 50 ng/ml recombinant human HGF (PeproTech, Inc.). The PI3K-specific inhibitor LY294002 (Sigma) and its noninhibitory analogue LY303511 (Sigma) were added to the medium at the concentration of 10 or 30 M together with HGF.
Phagokinetic Track Assay-Phagokinetic track assay was conducted as described previously (46). Gold colloid particles were prepared and glass coverslips were coated with the particles according to Albrecht-Buehler (47). The serum-starved C2C12 cells were plated on the gold particle-coated glass coverslips and cultured for 24 h in the low serum medium supplemented with or without HGF. The specimens were observed and photographed by dark field microscopy. The area of migration was calculated by Openlab software (Improvision).
Three-dimensional Migration Chamber Assay-Three-dimensional migration was assayed with Cell Culture Insert (6-mm diameter, 8 m pore size, BD Falcon). The upper surface of the membrane was coated with 50 g/ml collagen (Cellgen, Koken). C2C12 cells were serumstarved as stated above, and 1 ϫ 10 4 cells were replated on the membrane. The well contained the low serum medium supplemented with or without 50 ng/ml HGF. After the incubation for 5 h at 37°C, nonmigrated cells on the upper surface of the membrane were removed with a cotton swab. The migrated cells through the pores were fixed with 4% paraformaldehyde in PBS for 15 min and stained with 0.4% Crystal Violet in 10% ethanol for 30 min. They were examined with a Zeiss Axiovert 200 microscope, and their number was counted for quantitation.
Immunoblotting-Cultured cells were washed twice with cold PBS, lysed with the lysis buffer (150 mM NaCl, 4 mM MgCl 2 , 50 mM Tris-HCl, pH 7.5, 1% Nonidet P-40, 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride), and clarified by centrifugation at 15,000 rpm for 15 min. The supernatant was treated with SDS sample buffer and subjected to SDS-polyacrylamide gel electrophoresis. Proteins were transferred to Immobilon-P transfer membranes (Millipore). They were incubated with the anti-N-WASP, anti-WAVE1, and anti-WAVE2 polyclonal antibodies (pAbs) (48,49), which do not crossreact with one another, as well as with the anti-␤-tubulin monoclonal antibody (mAb) E7 (50) (Developmental Studies Hybridoma Bank). They were then incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG or goat anti-mouse IgG (Bio-Rad). Immunological reactions were detected with Western Lightning Chemiluminescence Reagent Plus (PerkinElmer Life Sciences).
Immunofluorescence Microscopy-Cells were fixed with 4% paraformaldehyde in PBS for 15 min and permeabilized with 0.2% Triton X-100 in PBS. They were incubated with anti-N-WASP and anti-WAVE2 pAbs and then with Alexa 488-conjugated goat anti-rabbit IgG (Molecular Probes, Inc.). Actin filaments were detected by the staining with rhodamine-phalloidin (Molecular Probes, Inc.). The specimens were observed with a Zeiss Axioskop microscope equipped with phase-contrast and fluorescence optics or with a Zeiss LSM 410 confocal laser scanning microscope.
Transduction with Adenoviral Vectors-Recombinant adenoviral vectors containing cDNAs of wild-type (wt) N-WASP and dominantnegative N-WASP(⌬cof) in addition to lacZ gene were constructed by homologous recombination between the expression cosmid cassette pAxCAwt and the parental viral genome as described previously (51,52). Recombinant adenoviral vectors containing the cDNAs of wt WAVE2 and dominant-negative WAVE2(⌬V) together with that of enhanced green fluorescent protein (EGFP) were constructed by use of the AdEasy system (49,53). The vectors were amplified in HEK293 cells and recovered from the cells according to He et al. (53). C2C12 cells grown on culture dishes were infected with the viral particles at a multiplicity of infection of 50 -100 in the low serum medium for 3 h. Twenty-four hours after the transduction, the cells were used for each assay.
RNA Interference (RNAi)-RNAi of N-WASP and WAVE2 was conducted with pSuper vector (54) (Oligo Engine) as described previously (43). The target sequences of mouse N-WASP and WAVE2 were 5Ј-AAGACGAGATGCTCCAAATGG-3Ј (nucleotides 434 -454 from the start codon) and 5Ј-AAGTGCCTTTGCCTCCCGAGT-3Ј (nucleotides 174 -194 from the start codon), respectively. Cells were transfected with the pSuper vectors targeting these sequences by using FuGENE 6 Transfection Reagent (Roche Applied Science). Twenty-four hours after the first transfection, the cells were retransfected to elevate the transfection efficiency. To monitor the short interfering RNA (siRNA)-expressing cells, one-tenth of the amount of pEGFP-C1 vector (BD Biosciences Clontech) encoding EGFP was cotransfected with the pSuper vectors.

HGF Induces Lamellipodial Formation and Migration in C2C12
Cells-During skeletal muscle regeneration, satellite cells are activated and migrate by chemotaxis to injured regions of myofibers, from which HGF is released. To elucidate the intracellular machinery for the satellite cell migration, we utilized satellite cell-derived C2C12 cells. When the myoblasts were cultured for 24 h under a low serum condition in 0.5% FBS, most cells were flattened and elongated and only ϳ10% of the cells had lamellipodial processes (Fig. 1A, panels a and c). Treatment of these serum-starved myoblasts with 50 ng/ml HGF for 30 min induced polarized cell shape and lamellipodial formation in ϳ40% of the cells (Fig. 1A, panels b and c). Next, we conducted phagokinetic track assay to examine whether the HGF treatment led to nondirectional cell migration. The serum-starved cells were replated on gold particle-coated coverslips and cultured for 24 h in the low serum medium supplemented with or without HGF (Fig. 1B, panels a and b). The tracks free of particles generated by phagokinetic ingestion or attachment of the particles on the cell surface correspond to migration trails. The area of migration was increased ϳ1.6-fold by the treatment with HGF (Fig. 1B, panel c). We further analyzed by three-dimensional migration chamber assay whether the HGF treatment was also responsible for the directional cell migration or chemotaxis. The serum-starved cells were replated on the collagen-coated membrane of the migration chamber and cultured for 5 h in the low serum medium supplemented with or without HGF supplied in the well (Fig.  1C, panels a and b). The number of cells migrated through the pores and toward the opposite side (HGF side) of the membrane was elevated ϳ3-fold by the incubation with HGF (Fig. 1C, panel c). Taken together, these results indicate that HGF induce both nondirectional and directional cell migration in C2C12 cells probably by forming lamellipodia at the leading edges.
N-WASP and WAVE2 Are Concentrated in the Lamellipodia-N-WASP and WAVE proteins play essential roles in actin polymerization to form filopodia and lamellipodia, respectively, leading to cell migration in at least certain types of cells (29,42,43,49). In MDCK epithelial cells, N-WASP is required for HGF-induced three-dimensional migration, invasion, and tubulogenesis (42), although implication of WAVEs in these functions has not been examined. In addition, it remains to be clarified whether N-WASP and WAVEs are involved in HGFinduced migration in other types of cells. Thus, we addressed whether N-WASP and WAVEs were implicated in HGF-induced lamellipodial formation and migration in C2C12 cells. First, we determined which proteins among the N-WASP and WAVEs were present in the myoblasts. Immunoblotting with specific antibodies showed that N-WASP and WAVE2 were present both in serum-starved C2C12 myoblasts and in those stimulated with HGF for 1 h in equivalent amounts as well as in the cells cultured in the growth medium containing 10% FBS ( Fig. 2A). In contrast, WAVE1 was scarcely detected in the cells under either condition. These results are consistent with the previous reports that N-WASP and WAVE2 are ubiquitously expressed in a variety of tissues and cells, whereas WAVE1 is predominantly expressed in brain (30,48,55,56). Therefore we concentrated on N-WASP and WAVE2 for the analyses of lamellipodial formation and migration of C2C12 cells hereafter.
Confocal optical sections (XY planes) and three-dimensional reconstitutions of the orthogonal planes derived from these sections (XZ planes) showed that lamellipodia formed at the leading edges of HGF-treated migrating cells contained actin filament networks (Fig. 2B, panels a and e). Both N-WASP and WAVE2 were highly concentrated in the lamellipodia (Fig. 2B, panels b and f) and colocalized with the actin filament networks in the  (Fig. 2B, panels c and g), in addition to their diffuse distribution throughout the cytoplasm and nuclei. Although N-WASP was well colocalized with the actin filament networks even at a higher magnification (Fig. 2B, panel d), WAVE2 was located also at the very front of the lamellipodia, 0.2-0.5 m wide, where actin filament networks were absent (Fig. 2B, panel  h, arrows). These results suggest that both N-WASP and WAVE2 are involved in lamellipodial formation and migration in HGFtreated C2C12 cells by acting sequentially or cooperating with each other in slightly different locations.
N-WASP and WAVE2 Are Required for the Lamellipodial Formation and Cell Migration-To examine whether N-WASP and WAVE2 actually participate in the lamellipodial formation and cell migration in HGF-treated C2C12 cells, the cells were transduced with cDNAs encoding wt and dominant-negative mutants of N-WASP and WAVE2 via adenoviral vectors and then treated with HGF. Mock transduction with the viral vectors harboring lacZ or EGFP interfered with the lamellipodial formation to some degree, but transduction with either wt N-WASP or WAVE2 facilitated the lamellipodial formation 1.6 -1.8-fold (Fig. 3, A and C). By contrast, transduction of the dominant-negative N-WASP(⌬cof) (28,52,57), which lacks its cofilin homology domain involved in the binding of Arp2/3 complex, suppressed the lamellipodial formation to ϳ30% of the level induced by wt N-WASP expression (Fig. 3, A and C). Furthermore, transduction of the dominant-negative WAVE2(⌬V) (49), which lacks its verprolin homology domain required for the binding of G-actin, markedly abrogated the lamellipodial formation to ϳ17% of the level induced by wt WAVE2 expression (Fig. 3, A and C).
The number of migrated cells toward HGF increased ϳ1.5fold by the transduction with either wt N-WASP or WAVE2 in three-dimensional migration chamber assay (Fig. 3, B and D). On the contrary, transduction of N-WASP(⌬cof) and WAVE2(⌬V) reduced the number of migrated cells to ϳ40 and ϳ30% of that of the wt N-WASP-and WAVE2-transduced cells, respectively (Fig. 3, B and D). The degree of promotion of lamellipodial formation and that of chemotaxis by these wt proteins were similar (Fig. 3, C and D). In addition, the profiles of their suppression by the dominant-negative mutants of these proteins also showed close similarity (Fig. 3, C and D). Accordingly, the lamellipodial formation induced by HGF seems to be closely correlated with the chemotaxis toward HGF, and they are likely to be under the regulation of both N-WASP and WAVE2.
We further investigated the role of N-WASP and WAVE2 in the formation of lamellipodia by silencing their expression through siRNA expression vector-mediated RNAi. As detected by immunoblotting, the amounts of endogenous N-WASP and WAVE2 proteins were down-regulated by the expression of respective siRNAs, whereas that of ␤-tubulin was not affected (Fig. 4A). The degree of silencing of WAVE2 expression was higher than that of N-WASP expression. To analyze the effects of RNAi of N-WASP and WAVE2, EGFP expression vector was cotransfected with the siRNA expression vectors, and cells with EGFP fluorescence were regarded as the cells expressing siR-NAs. Although Ͼ30% of mock transfected cells treated with HGF formed lamellipodia, the cells with lamellipodia decreased to ϳ20% and Ͻ10% by the expression of N-WASP and WAVE2 siRNAs, respectively (Fig. 4, B and D).
The RNAi of N-WASP also moderately interfered with the chemotaxis toward HGF, whereas that of WAVE2 considerably retarded it (Fig. 4, C and E). The suppression of lamellipodial formation and chemotaxis by WAVE2 RNAi was more remarkable than by N-WASP RNAi. This is presumably because the extent of knockdown of WAVE2 expression was higher than that of N-WASP expression. Again, the profiles of suppression of lamellipodial formation and those of chemotaxis by N-WASP and WAVE2 RNAi showed close similarity (Fig. 4, D and E). Thus, these results also indicate the close correlation between the lamellipodial formation and the chemotaxis and essential roles of both N-WASP and WAVE2 in these phenomena in C2C12 cells.
PI3K Is Essential for the Lamellipodial Formation and Cell Migration-Activation of PI3K is required for the HGF-induced adherence junction disassembly, lamellipodial formation, and subsequent migration or scattering of MDCK epithelial cells (21-23, 25, 26). To examine whether PI3K is also involved in HGF-induced lamellipodial formation and chemotaxis in C2C12 cells, the cells were treated with the PI3K inhibitor LY294002 and with its noninhibitory analogue LY303511 as a negative control of LY294002. By treating with 50 ng/ml HGF for 30 min, Ͼ35% of the serum-starved cells formed actinenriched lamellipodia (Fig. 5, A (panel a) and C). These cells treated with 30 M LY303511 also formed the lamellipodia at equivalent efficiency (Fig. 5, A (panels b and c) and C). When 10 M LY294002 was added together with HGF, however, cells with lamellipodial formation were retarded to ϳ10%, and consequently polarized cell shape was lost in these cells (Fig. 5, A  (panels d and g) and C). Instead, many of these cells had peripheral thick stress fibers. By the addition of 30 M LY294002, the lamellipodial formation was further reduced to ϳ5% (Fig. 5C). Although N-WASP and WAVE2 were colocalized with the actin filament networks in the lamellipodia in the HGF-treated cells (see Fig. 2), they were barely distributed in the cell peripheries but rather concentrated in the cell center and the nuclei in the LY294002-treated cells (Fig. 5A (panels e, f, h, and i). Furthermore, migrated cells toward HGF were reduced to about a half and to ϳ15% by the treatment with 10 and 30 M LY294002, respectively (Fig. 5, B (panels b and c) and D). In contrast, the treatment with 30 M LY303511 did not significantly affect the migration (Fig. 5, B (panel d) and D). These results imply that PI3K activity is necessary for the HGF-induced lamellipodial formation and chemotactic migration toward HGF in C2C12 cells as has been shown in MDCK epithelial cells. They further suggest that PI3K plays crucial roles in the recruitment of N-WASP and WAVE2 to the cell peripheries or polarized plasma membranes.  Fig. 1A. The difference between the mock transduced cells and the cells transduced with wt or dominant-negative mutants is statistically significant (* and ** represent p Ͻ 0.05 and p Ͻ 0.01, respectively). D, relative number of migrated cells. Values were assessed as described in the legend to Fig. 1C. cell migration toward HGF, we have utilized C2C12 cells derived from satellite cells. HGF induced reorganization of actin cytoskeleton leading to polarized cell shape and formation of lamellipodia. HGF treatment also facilitated nondirectional two-dimensional migration and directional three-dimensional migration toward HGF. Since polarized shape with lamellipo- dia is generally essential for cell migration (39 -41), these morphological alterations are likely to be intimately coupled with both the two-dimensional and three-dimensional migration of C2C12 cells.
The formation of lamellipodia or ruffles is induced by WAVE proteins activated indirectly by Rac1 (31-33). In contrast, filop-odia are formed by N-WASP activated directly by Cdc42 and PI(4,5)P 2 (27,28). For the first step to address the mechanisms of HGF-induced lamellipodial formation and cell migration, we examined the expression of not only WAVEs but also N-WASP in C2C12 myoblasts. N-WASP and WAVE2 proteins were constantly present in the cells regardless of the stimulation with HGF, whereas WAVE1 was barely detected. Although the lamellipodia were peripherally formed in C2C12 cells, dorsal circular ruffles, which are formed on the dorsal surface of fibroblasts treated with PDGF (43,58), were not formed in C2C12 cells treated with HGF or other growth factors (date not shown). WAVE2 is responsible for the formation of the peripheral ruffles or lamellipodia, whereas WAVE1 is essential for the dorsal ruffle formation (43). The presence of WAVE1 at the marginal level in C2C12 cells may account for the lack of dorsal ruffle formation.
Both N-WASP and WAVE2 were concentrated in the lamellipodia at the leading edges and colocalized with the actin filament networks in the lamellipodia. WAVE2 was located also at the very front of the lamellipodia, where actin filament networks were absent, as has been shown in mouse embryonic fibroblasts treated with PDGF (43). These locations suggest the participation of not only WAVE2 but also N-WASP in the lamellipodial formation in HGF-treated C2C12 cells. Recent analyses by fluorescence resonance energy transfer have shown that activated N-WASP is localized together with active Cdc42 and PI(4,5)P 2 in lamellipodia at the leading edges of migrating COS-7 and carcinoma cells stimulated with epidermal growth factor (59,60). These findings strongly support the above postulation that N-WASP as well as WAVE2 is involved in the lamellipodial formation at the leading edges at least in certain types of cells under particular conditions. Both the shared and separate locations of N-WASP and WAVE2 in the lamellipodia of C2C12 cells may imply that they act sequentially or cooperating with each other for the formation of lamellipodia.
Both lamellipodial formation by HGF and directional threedimensional migration toward HGF were facilitated by the overexpression of wt N-WASP and WAVE2. These results imply that both N-WASP and WAVE2 are activated by HGFinduced signaling pathways mediated by the activation of Cdc42 and Rac1. Alternatively, some other signaling molecules that activate N-WASP and WAVE2 independently of these small GTPases may be produced by the HGF signaling. These molecules include PI(4,5)P 2 (28,48) and Src homology 3 (SH3) domain-containing proteins such as Grb2 (48,61) and WISH (62) for N-WASP activation. On the other hand, PI(3,4,5)P 3 recruits WAVE2 to polarized plasma membrane to form lamellipodia (34). The HGF-induced lamellipodial formation and the migration were prevented by the expression of the dominantnegative mutants of N-WASP and WAVE2. These results confirm the possibility that both N-WASP and WAVE2 are involved in the lamellipodial formation and the migration in these cells. This notion is corroborated by the results that the lamellipodial formation and the migration were abrogated by the knockdown of the endogenous expression of either N-WASP or WAVE2 through RNAi. To our knowledge, this is the first evidence that WAVE2 is activated by HGF and responsible for the HGF-induced lamellipodial formation and cell migration regardless of cell type.
Although N-WASP activated by Cdc42 is generally believed to be responsible for the filopodial formation rather than lamellipodial formation, the results in the present study clearly indicate the involvement of N-WASP in the lamellipodial formation and subsequent migration. This is possibly due to the HGF-induced lamellipodial formation in C2C12 cells may require actin filament reorganization achieved by sequential activation of N-WASP and WAVE2. Indeed, filopodial formation is accompanied by lamellipodial formation by the sequential activation of Cdc42 and Rac1 (63). In addition, HGF-induced lamellipodial formation, spreading, and scattering of MDCK cells are inhibited by a dominant-negative Cdc42 (26). The more proximal location of N-WASP and more distal location of WAVE2 in lamellipodia shown in this and other studies (43) might support the sequential activation of N-WASP and WAVE2. Moreover, N-WASP is only temporally activated at the initiation stage of invadopodial formation at the base of invadopodia, which are cellular protrusions responsible for the invasive migration of carcinoma cells into extracellular matrix (60). This may also favorable for the above notion. It is necessary to investigate whether the involvement of N-WASP in lamellipodial formation and migration is specific for particular types of cells including C2C12 cells.
PI3K is required for HGF-induced lamellipodial formation and subsequent migration or scattering in MDCK cells (21-23, 25, 26). PI3K was also indispensable for C2C12 cells to take polarized cell shapes, to form lamellipodia in response to HGF, and to migrate toward HGF because inhibition of PI3K by LY294002 in the HGF-treated C2C12 cells abrogated these responses. The inhibition of PI3K impaired the localization of N-WASP and WAVE2 in the cell peripheries and instead led to their accumulation in the cell center and the nuclei. This is probably because PI(3,4,5)P 3 produced by PI3K recruits WAVE2 to the polarized plasma membrane and this recruitment is essential for the lamellipodial formation at the leading edge as has been shown with PDGF-treated mouse fibroblasts (34). If PI3K is inhibited, WAVE2 cannot be recruited from the cell center and the nucleus to the plasma membrane. PI(3,4,5)P 3 not only binds to WAVE2 but also to N-WASP (34). Thus, PI(3,4,5)P 3 seems to be also involved in the recruitment of N-WASP to the leading edge of the polarized plasma membrane to form membrane protrusion leading to lamellipodia.
FIG. 6. Putative HGF-Met signaling networks for the migration of C2C12 cells. HGF-Met activates PI3K via the scaffolding adaptor protein Gab1 and Grb2-Sos-Ras. A PI3K product PI(3,4,5)P 3 induces activation of WAVE2 by activating Rac1. It also recruits WAVE2 and possibly N-WASP as well to the polarized plasma membrane for their activation. N-WASP is activated by Cdc42, which is activated by unknown mechanisms, and Grb2 associated with Met. N-WASP and WAVE2 are involved in migration by inducing cellular protrusions including lamellipodia. PI(3,4,5)P 3 plays essential roles also in translocation to the plasma membrane and activation of the guanine nucleotide exchange factors, Tiam1, Vav, and Sos, which activate Rac1 (64 -66). In turn, activated Rac1 induces the activation of WAVE2. Although the mechanisms remain to be elucidated, HGF stimulation of MDCK cells also activates Cdc42 independently of PI3K, and this activation is implicated in lamellipodial formation and cell scattering (26) probably by activating N-WASP. Furthermore, activated Met can associate with Grb2 (18 -20), which may activate N-WASP (48,61) as well as PI3K via the Sos-Ras pathway. If these signals are induced in HGFtreated C2C12 cells, HGF-induced lamellipodial formation and migration are possibly regulated by the signaling networks summarized in Fig. 6.
Here we have shown that C2C12 cells are responsive to HGF to exhibit nondirectional and directional migration. This implies that C2C12 cells represent a model system for satellite cell migration regulated by HGF signaling. Isolation of enough number of pure satellite cell population for biochemical and molecular biological analyses is difficult because satellite cells easily cease from proliferation and instead contaminated fibroblasts vigorously proliferate. Thus, C2C12 cells substitute for satellite cells for these analyses, as do MDCK cells for epithelial cells.
Increasing evidence has been accumulated that N-WASP and WAVE proteins are involved in a variety of cell migration both in vitro and in vivo (42,43,49,60) This study has first presented the possibility that N-WASP and WAVE2 as well as PI3K play essential roles in chemotactic migration of satellite cells to injured regions during muscle regeneration. However, intracellular mechanisms of migration during development have scarcely been elucidated, including migration of muscle precursor cells from hypaxial dermomyotome to sites of muscle formation such as limbs, tongue, and diaphragm (1-3) as well as that of neural crest cells from dorsal neural tube to a variety of regions (67,68). It is important to examine whether N-WASP and WAVE proteins are necessary for the migration of these cells during development.