Nebulin and N-WASP Cooperate to Cause IGF-1–Induced Sarcomeric Actin Filament Formation

See allHide authors and affiliations

Science  10 Dec 2010:
Vol. 330, Issue 6010, pp. 1536-1540
DOI: 10.1126/science.1197767


Insulin-like growth factor 1 (IGF-1) induces skeletal muscle maturation and enlargement (hypertrophy). These responses require protein synthesis and myofibril formation (myofibrillogenesis). However, the signaling mechanisms of myofibrillogenesis remain obscure. We found that IGF-1–induced phosphatidylinositol 3-kinase–Akt signaling formed a complex of nebulin and N-WASP at the Z bands of myofibrils by interfering with glycogen synthase kinase-3β in mice. Although N-WASP is known to be an activator of the Arp2/3 complex to form branched actin filaments, the nebulin–N-WASP complex caused actin nucleation for unbranched actin filament formation from the Z bands without the Arp2/3 complex. Furthermore, N-WASP was required for IGF-1–induced muscle hypertrophy. These findings present the mechanisms of IGF-1–induced actin filament formation in myofibrillogenesis required for muscle maturation and hypertrophy and a mechanism of actin nucleation.

IGF-1–induced muscle maturation and hypertrophy necessitate not only protein synthesis but also myofibrillogenesis. IGF-1 causes protein synthesis by activating phosphatidylinositol 3-kinase (PI3K)–Akt signaling (1, 2). In contrast, the signaling mechanisms of myofibrillogenesis are unclear. Although the proteins inducing actin filament formation in nonmuscle cells have been elucidated (37), those responsible for actin filament formation from the Z bands of myofibrils remain unknown. Thus, we asked whether N-WASP, which causes rapid actin polymerization by activating the Arp2/3 complex in nonmuscle cells, is involved in myofibrillar actin filament formation.

N-WASP was localized to the Z bands in mature skeletal muscle and became confined to the Z bands during muscle regeneration (Fig. 1A and fig. S1, A and B). Then we examined the location of N-WASP in the skeletal muscle of IGF-1–administered mice, which were prefasted for 48 hours. Although N-WASP was diffusely distributed throughout the myofibers in the fasted muscle, it was mobilized to the Z bands within 30 min after IGF-1 stimulation (Fig. 1B). N-WASP was associated with the Z bands and colocalized with α-actinin by 2 hours after the stimulation, but became diffusely distributed again by 4 hours. α-actinin was restricted to the Z bands in both the fasted and the stimulated muscles, indicating that the Z bands remained intact during these periods.

Fig. 1

IGF-1 stimulation induces mobilization of N-WASP to the Z bands and actin polymerization. (A) Immunofluorescent localization of N-WASP at the Z bands in longitudinal cryosections of the tibialis anterior (TA) muscle. Scale bar indicates 10 μm. (B) Recruitment of N-WASP to the Z bands by IGF-1 stimulation. IGF-1 was administered to fasted mice, and the TA muscles were isolated at the indicated time points. Intensity, average fluorescence intensities. Scale bar, 2 μm. (C) Mobilization of EGFP–α-actin to the Z bands and its subsequent distribution throughout the thin filament length induced by IGF-1 stimulation.

A plasmid encoding enhanced green fluorescent protein (EGFP)–tagged α-actin was introduced into the fasted muscle. Although the EGFP–α-actin was diffusely distributed in the fasted muscle myofibers in the absence of stimulation, it was mobilized to the Z bands within 30 min after IGF-1 stimulation (Fig. 1C). It was distributed gradually in broader stripes away from the Z bands and eventually almost throughout the entire length of the thin filaments, which were detected by phalloidin staining by 4 hours. Thus, IGF-1 signaling induces the recruitment of N-WASP to the Z bands, and N-WASP may be involved in the formation of actin thin filaments from the Z bands without the Arp2/3 complex (fig. S1, C and D).

To identify the protein that recruits N-WASP to the Z bands in response to IGF-1 stimulation, we searched for N-WASP–binding protein(s) associated with the Z bands. In a binding (pull-down) assay with glutathione S-transferase (GST)–tagged N-WASP, nebulin bound to N-WASP (Fig. 2A). Nebulin is a giant (500 to 900 kD) sarcomeric protein spanning almost the entire length of thin filaments (811). Its C-terminal region, which is associated with the Z band, contains the Src homology 3 (SH3) domain (fig. S2, A and B). N-WASP has a proline (Pro)-rich region (fig. S2C), where several SH3 domain–containing proteins, such as Grb2, bind to activate N-WASP by inducing the open conformation (3). We thus examined whether the N-WASP Pro-rich region [N-WASP(Pro)] bound to the SH3 domain–containing C-terminal region of nebulin [Neb(C)]. N-WASP and N-WASP(Pro) but not Pro-rich region–deleted N-WASP [N-WASP(ΔPro)] bound to Neb(C) (Fig. 2B and fig. S3, A and B). Furthermore, N-WASP(Pro) competed for the binding of GST–N-WASP and nebulin in skeletal muscle lysates (Fig. 2C). Accordingly, N-WASP binds through its Pro-rich region to the SH3 domain in the C-terminal region of nebulin in vitro. When N-WASP was immunoprecipitated with the antibody against N-WASP (anti–N-WASP) from muscle lysates of fasted mice, nebulin was scarcely coprecipitated with N-WASP (Fig. 2D). Administration of IGF-1 to the mice, however, resulted in coprecipitation of nebulin with N-WASP (Fig. 2D and fig. S3C). Therefore, IGF-1 stimulation allows N-WASP to bind to nebulin in vivo, and N-WASP is recruited to the Z bands through the binding to nebulin.

Fig. 2

N-WASP binds to the nebulin C-terminal region by IGF-1 stimulation. (A) Binding of nebulin in skeletal muscle lysates to N-WASP, detected by a GST–N-WASP pull-down assay. Shown are SDS–polyacrylamide gel electrophoresis patterns and corresponding immunoblots (IB) with an antibody against nebulin (anti-nebulin). MyHC, myosin heavy chain. (B) Binding of N-WASP and N-WASP(Pro) to Neb(C), detected by a GST–Neb(C) pull-down assay. Hemagglutinin (HA)–tagged N-WASP and its mutants were expressed in COS-1 cells. (C) Binding of nebulin in muscle lysates to N-WASP, detected by a GST–N-WASP pull-down assay. N-WASP(Pro) was added as a binding competitor. (D) Induction of the binding of N-WASP to nebulin by IGF-1 stimulation, detected by coimmunoprecipitation with N-WASP. Muscle lysates prepared 60 min after IGF-1 administration to fasted mice were used for immunoprecipitation (IP).

Next, we addressed how IGF-1 signaling induced the binding of N-WASP to nebulin. IGF-1 stimulation activates two signaling pathways, PI3K-Akt signaling and the Ras–extracellular signal–regulated kinase (ERK) pathway. When COS-1 cells were stimulated with IGF-1, activating phosphorylation of both Akt and ERK occurred (Fig. 3A). The phosphorylation of Akt and ERK was prevented by treatment with the PI3K inhibitor LY294002 and the mitogen-activated protein kinase kinase (MEK) inhibitor U0126, respectively. Analyses with these inhibitors indicated that PI3K-Akt signaling but not the Ras-ERK pathway participated in the binding of N-WASP to Myc-Neb(C) (Fig. 3A and fig. S4, A and B).

Fig. 3

IGF-1–induced PI3K-Akt signaling causes nebulin–N-WASP interaction by suppressing GSK-3β. (A) Induction of nebulin–N-WASP interaction by PI3K-Akt signaling. Myc–Neb(C)-transfected COS-1 cells were stimulated with IGF-1 in the presence of the PI3K inhibitor LY294002, its noninhibitory analog LY303511, or the MEK inhibitor U0126. Activating phosphorylation of Akt and ERK1/2 was detected with phospho-antibodies. The binding of N-WASP to Myc–Neb(C) was detected by coimmunoprecipitation with the monoclonal antibody (mAb) to Myc (anti-Myc). (B) Facilitation of the binding of N-WASP to Neb(C) by inhibiting GSK-3β. Myc-Neb(C)–transfected COS-1 cells were stimulated with IGF-1 or treated with GSK-3β inhibitors, LiCl and SB216763. The binding of N-WASP to Myc-Neb(C) was detected by coimmunoprecipitation with anti-Myc. (C) Abrogation of the binding of N-WASP to Neb(C) phosphorylated by GSK-3β, detected by a GST-Neb(C) pull-down assay. Phosphorylation of GST-Neb(C) was analyzed by autoradiography (ARG). U, units. (D) Binding of N-WASP to unphosphorylatable Neb(C) but not to pseudo-phosphorylated Neb(C) expressed in the muscle. The TA muscles were transfected with Myc-Neb(C)(S1A/S2A) or Myc-Neb(C)(S1D/S2E). The localization of Myc–Neb(C) and N-WASP was detected in the fasted mouse muscle. Scale bar, 2 μm.

One of the target proteins of Akt is glycogen synthase kinase-3β (GSK-3β), and phosphorylation by Akt inhibits its kinase activity (12). When COS-1 cells transfected with Myc-Neb(C) were treated with a GSK-3β inhibitor, LiCl or SB216763, N-WASP was co-precipitated with Myc-Neb(C), even without IGF-1 stimulation (Fig. 3B). Moreover, GSK-3β phosphorylated GST-Neb(C) dose-dependently in vitro (Fig. 3C and fig S4C). Unphosphorylated Neb(C) strongly bound N-WASP, but the binding level was reduced according to the degree of phosphorylation. Thus, the phosphorylation of Neb(C) by GSK-3β inhibits the interaction of nebulin with N-WASP. Indeed, there are two consensus serine residues (Ser1 and Ser2) phosphorylatable by GSK-3β in the Ser-rich region adjacent to the SH3 domain of nebulin (fig. S2B). To determine whether GSK-3β phosphorylates them, we mutated the Ser (S) to Ala (A) [Neb(C)(S1A) and Neb(C)(S1A/S2A)]. GSK-3β efficiently phosphorylated GST-tagged wild-type Neb(C) in vitro, whereas it phosphorylated Neb(C)(S1A) less efficiently and did not phosphorylate Neb(C)(S1A/S2A) (fig. S4, D to G).

Administration of IGF-1 to fasted mice caused the activating phosphorylation of Akt and the inhibiting phosphorylation of GSK-3β by Akt in skeletal muscle (fig. S5A). Both Ser1 and Ser2 in nebulin were phosphorylated at high levels in the fasted mouse muscle, but IGF-1 stimulation prominently reduced their phosphorylation levels. Endogenous N-WASP was located to the Z bands when Myc-Neb(C)(S1A/S2A) was expressed, whereas it was diffusely distributed when Myc-Neb(C)(S1D/S2E) (where D indicates Asp and E indicates Glu) was expressed (Fig. 3D). These results and those in fig. S5, B to F, imply that IGF-1–induced PI3K-Akt signaling suppresses GSK-3β by phosphorylation and consequently prevents the phosphorylation of Ser1 and Ser2 in nebulin by GSK-3β on the Z bands. Unphosphorylated nebulin but not phosphorylated nebulin can bind N-WASP and recruit it to the Z bands.

Because the Arp2/3 complex was not associated with myofibrils (fig. S1, C and D), we investigated the possibility that N-WASP induces myofibrillar actin filament formation in the absence of the Arp2/3 complex. N-WASP has two WASP homology 2 (WH2) domains (fig. S2C), which are actin monomer–binding motifs present in several actin nucleators and elongators (3, 57, 13). On the other hand, nebulin contains 185 tandem modules (fig. S2A), which contain an actin monomer–binding motif in each repeat (8, 14). We thus examined whether the two WH2 domains of N-WASP cooperate with module 182 (M182) to M185 of nebulin to induce actin nucleation or elongation (fig. S6, A and B). In a fluorometric actin polymerization assay in the absence of N-WASP, Neb(C)2M [Neb(C) with M184 and M185] (fig. S6A) facilitated actin polymerization only weakly, Neb(C)3M facilitated moderately, and Neb(C)4M more efficiently (Fig. 4A and fig. S6C). In the presence of N-WASP, all of the Neb(C)Ms polymerized actin much more efficiently than in the absence of N-WASP. In contrast, neither N-WASP(ΔPro), which cannot bind to Neb(C), nor N-WASP(ΔWCA), which lacks the WCA region involved in actin monomer recruitment, enhanced actin polymerization by Neb(C)Ms (fig. S6D). Accordingly, Neb(C)Ms and N-WASP cooperate to polymerize actin.

Fig. 4

The nebulin–N-WASP complex induces actin filament formation, and N-WASP is essential for muscle hypertrophy. (A) Promotion of actin polymerization by Neb(C)Ms with N-WASP, detected by a pyrene-actin polymerization assay. a.u., arbitrary units. (B) Facilitation of barbed-end formation by Neb(C)Ms with N-WASP, calculated from the results of total internal reflection fluorescence (TIRF) microscopy. (C) Suppression of α-actin incorporation into the Z bands and thin filaments by N-WASP RNAi. Muscles were transfected with the siRNA expression vectors and EGFP–α-actin. IGF-1 was administered to fasted mice, and the muscles were isolated at 2 hours. Scale bar, 2 μm. (D) Suppression of natural and IGF-1–induced muscle hypertrophy by N-WASP RNAi. Muscles were transfected with the siRNA expression vectors and pEGFP-C1 vector to monitor the siRNA-expressing myofibers. IGF-1 was administered every 2 days for 14 days to mice (from 35 days to 49 days old). Cross sections of the muscle are shown. Scale bar, 20 μm.

The number of unbranched actin filament seeds increased proportionally to the concentration of Neb(C)Ms in the presence of N-WASP (movies S1 to S10 and fig. S7, A and B). Barbed end formation and nucleation rates by Neb(C)Ms were facilitated by N-WASP (Fig. 4B and fig. S7, C to E). The results suggest that nebulin modules and the N-WASP WH2 domains cooperate to nucleate an unbranched actin filament. Then the actin filament might elongate along the nebulin modules from the Z band toward the center of a sarcomere (fig. S8A).

We assessed the requirement of N-WASP for IGF-1–induced actin filament formation in myofibrils and muscle hypertrophy by RNA interference (RNAi) (fig. S9). EGFP–α-actin coexpressed with control small interfering RNA (siRNA) in the fasted mouse muscle was diffusely distributed without IGF-1 stimulation but located to the Z bands and thin filaments within 2 hours after the stimulation (Fig. 4C and fig. S9D). In contrast, EGFP–α-actin coexpressed with siRNA1 or 2 (fig. S11A) remained diffusely distributed after the stimulation. Therefore, N-WASP is indispensable for the recruitment of α-actin to the Z bands and for myofibrillar actin filament formation.

IGF-1 administration to mice caused muscle hypertrophy owing to the increase in myofiber volume. The expression of siRNA1 or 2 reduced the cross-sectional area of the myofibers regardless of IGF-1 administration (Fig. 4D and figs. S10 and S11). Thus, N-WASP plays essential roles in both age-dependent natural hypertrophy and administered IGF-1–induced hypertrophy. N-WASP seems to participate in myofiber hypertrophy by inducing myofibrillar actin filament formation through the nebulin–N-WASP complex. This notion is consistent with the observation that Neb-deficient mice develop a muscle atrophy–like phenotype (15, 16).

We elucidated the signaling of IGF-1–induced myofibrillar actin filament formation from the Z bands (fig. S8B) and a mechanism of actin nucleation [supporting online material (SOM) text]. The Neb–N-WASP complex formed by the signaling can explain actin filament formation arising from the Z bands. These findings may provide insights into the mechanisms of muscular diseases, such as nemaline myopathy, caused by NEB gene mutations (17). The actin filament formation together with myosin filament assembly, which might also induced by IGF-1 signaling, results in myofibrillogenesis required for muscle maturation and hypertrophy.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 to S11


Movies S1 to S10

References and Notes

  1. We thank N. Watanabe and T. Itoh for comments. This work was supported by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan and by the Research Grants (17A-10 and 20B-13) for Nervous and Mental Disorders from the Ministry of Health, Labor, and Welfare of Japan.

Stay Connected to Science

Navigate This Article