SIRT1 activators promote membrane resealing in C2C12 cells.
Initially, we examined whether Rsv, a SIRT1 activator, promotes membrane resealing in C2C12 cells. C2C12 myoblasts were incubated with 50 µM Rsv for 12 h. Rsv induced the deacetylation of histone H3 at lysine 9 (H3K9) in C2C12 myoblasts, causing SIRT1 activation (Fig. 1a). Membrane resealing can be monitored by an influx of the fluorescent dye FM1 − 43 in cells after laser irradiation. Pretreatment with 10 µM of cytochalasin D (CyD), actin polymerization inhibitor, enhanced FM1 − 43 uptake, as previously reported (Figure S1; 18). In contrast, when cells were treated with Rsv, FM1 − 43 uptake by cells was significantly limited (Fig. 1b, Video S2) compared with those of control cells (Fig. 1b, Video S1). Similar to C2C12 myoblasts, Rsv also promoted membrane resealing after laser irradiation in C2C12 myotubes (Fig. 1c, Video S3 and 4). Additionally, 10 mM of nicotinamide mononucleotide (NMN), a SIRT1 activator, promoted membrane resealing in C2C12 myoblasts and myotubes (Figure S2). These data indicated that SIRT1 activators promote membrane resealing.
Rsv promotes membrane resealing via SIRT1 function in C2C12 myoblasts.
To assess whether Rsv promotes membrane resealing through SIRT1 activation, we examined the effects of SIRT1 knockdown. C2C12 myoblasts were treated with control siRNA or Sirt1 siRNA for 48 h. Knockdown of SIRT1 did not change morphology (data not shown) or induce apoptosis (19). Treatment with Sirt1 siRNA reduced SIRT1 protein levels to approximately 20% compared to that in control siRNA-treated cells (Fig. 2a). When cells were treated with Rsv, the influx of FM1 − 43 was significantly reduced in control siRNA-treated cells (Fig. 2b, Videos S5 and S6). In contrast, Sirt1 siRNA-treated cells showed the persistent intracellular entry of FM1 − 43 dye after laser injury, where Rsv failed to promote membrane resealing (Fig. 2c, Videos S7 and S8). This indicated that Rsv promotes membrane resealing through SIRT1 activation.
SIRT1 is necessary for the reorganization of the cytoskeleton after membrane injury.
Cytoskeleton reorganization occurs during the early phase of membrane resealing (20). After laser damage, F-actin is enriched at the membrane disruption sites observed in cultured muscle cells and oocytes (21, 22). To determine whether SIRT1 regulates cytoskeleton reorganization, we used C2C12 myoblasts expressing GFP-tagged actin (actin-GFP) and induced membrane damage by laser irradiation. The fluorescence intensity of actin-GFP in the dotted semicircle (Fig. 3) was measured as the actin concentration. Upon laser injury, actin slightly accumulated at the lesion immediately under the sarcolemma (Fig. 3a arrow on the left upper panel, Video S9), as previously reported (21, 22). Interestingly, Rsv-treated cells showed a stronger actin accumulation than control cells (Fig. 3a arrow on the left lower panel, Video S10). These findings indicate that SIRT1 plays a key role in cytoskeletal reorganization following membrane damage. However, when cells were treated with 10 µM of CyD accumulation of actin was not observed, and the membrane was disarranged regardless of treatment with Rsv (Fig. 3b, Video S11 and 12). We next examined the effects of SIRT1 specific inhibitor, Ex527, and the knockdown of SIRT1 on actin reorganization. Ex527 at 30 µM or siRNA was treated for 12 h or 48 h respectively; then laser injury was induced. The results showed that Ex527 and Sirt1 siRNA treatment inhibited actin accumulation at the injury site (Fig. 3c and d, Video S13–16).
Cortactin is required for membrane resealing via accumulation of actin at the injury site.
Since small membrane protrusions were frequently observed at the injury site and appeared as small lamellipodia (Figure S3), we focused on the function of CTTN. To confirm the interaction between CTTN and F-actin, we performed a pull-down assay using phalloidin, a mushroom toxin that binds to F-actin. F-actin and its complexes were pulled down with phalloidin-XX-biotin. Immunoblotting showed that the CTTN protein was present in the phalloidin-pulldown fraction (Fig. 4a). Since, we could not get enough florescence intensity of mCherry-tagged CTTN (CTTN-mCherry) in C2C12 cells, we used COS7 cells for evaluating the transport of these proteins upon laser injury. Upon injury, COS7 cells expressing only CTTN-mCherry did not show CTTN accumulation at the injury site (data not shown). However, COS7 cells co-expressing actin-GFP and CTTN-mCherry showed an accumulation of actin and CTTN at the lesion immediately under the sarcolemma (Fig. 4b arrows, Video S17). This indicates that CTTN plays a role in cytoskeleton reorganization after membrane injury. To assess the function of CTTN in membrane resealing, we used siRNA to knockdown CTTN in C2C12 cells. Cttn siRNA reduced the CTTN protein levels by approximately 50% (Fig. 4c). Control siRNA-treated C2C12 myoblasts showed actin accumulation at the injury site (Fig. 4d arrow, Video S18). Treatment with Cttn siRNA attenuated actin accumulation (Fig. 4d, Video S19). Additionally, FM1 − 43 uptake upon membrane damage was enhanced in CTTN-knockdown C2C12 cells compared to control cells (Fig. 4e, Video S20 and S21).
SIRT1 deacetylates CTTN and promotes CTTN accumulation at the injury site.
Since knockdown of SIRT1 and CTTN resulted in less actin accumulation at the injury site, we assumed that SIRT1 regulates CTTN function required for membrane resealing since SIRT1 deacetylates CTTN and promotes cell migration (17). To confirm the interaction between SIRT1 and CTTN, COS7 cells expressing GFP-tagged SIRT1 (SIRT1-GFP) and Flag-tagged CTTN (CTTN-Flag) were used. COS7 cells expressing CTTN-Flag and either GFP or SIRT1-GFP were lysed, and further immunoprecipitation was performed using an anti-Flag antibody conjugated with beads. Immunoprecipitation of Flag pulled down CTTN-Flag, and the anti-Flag antibody co-immunoprecipitated SIRT1-GFP but not GFP (Fig. 5a). Additionally, acetylated CTTN-Flag was not detected in COS7 cells co-expressing SIRT1-GFP (Fig. 5a). In COS7 cells, CTTN-Flag was pulled down with phalloidin but was detectable in the phalloidin-pulldown fraction in lysates from cells expressing SIRT1-GFP (Fig. 5b). We used mutant SIRT1 (H355Y), lacking the deacetylation activity of normal SIRT1 (23), to examine whether the deacetylation of CTTN is necessary for binding to F-actin. Lysates from COS7 cells expressing CTTN-Flag and H355Y-GFP were immunoprecipitated with an anti-Flag antibody, and we found that H355Y-GFP co-immunoprecipitated with CTTN-Flag (Fig. 5a). However, CTTN was not deacetylated by co-expression with H355Y-GFP (Fig. 5a) and was not pulled down with phalloidin in cells expressing H355Y-GFP. These results indicate that the deacetylation of CTTN by SIRT1 promotes the binding of CTTN to F-actin.
Next, we examined whether co-expression of SIRT1-GFP induces CTTN accumulation at the injury site (Fig. 5c, arrows). Similar to CTTN-mCherry, SIRT1-GFP did not accumulate at the injury site in COS7 cells expressing only SIRT1-GFP (data not shown). In cells co-expressing CTTN-mCherry and GFP, CTTN-mCherry was not accumulated at the injury site (Fig. 5c upper panels and 5d, Video S22). When SIRT1-GFP was co-expressed, CTTN-mCherry was accumulated significantly (Fig. 5c middle panels and 5d, Video S23). In contrast, H355Y-GFP did not promote CTTN accumulation (Fig. 5c lower panels and 5d, Video S24).
Rsv induces deacetylation of CTTN and promotes membrane resealing.
Our data indicate that SIRT1 regulates membrane resealing through the deacetylation of CTTN. To evaluate the effects of Rsv on CTTN, C2C12 myoblasts were treated with 50 µM Rsv, and further immunoblotting was performed for acetylated CTTN. CTTN was deacetylated by Rsv treatment (Fig. 6a). The phalloidin-pulldown fraction from Rsv-treated C2C12 myoblasts contained more CTTN than that from control cells, indicating that Rsv enhanced the binding of CTTN to F-actin (Fig. 6b). In COS7 cells expressing CTTN-mCherry, Rsv treatment induced the accumulation of CTTN at the injury site, while no accumulation was found in control cells (Fig. 6c, Video S25 and S26). These results indicate that Rsv promotes the deacetylation of CTTN, which enhances the binding of CTTN to F-actin needed for membrane resealing. To evaluate the role of CTTN in the effect of Rsv on membrane resealing, we knocked down CTTN expression in C2C12 cells and monitored membrane resealing by an influx of the fluorescent dye FM1 − 43. Rsv treatment improved membrane resealing in control siRNA-treated C2C12 myoblasts (Fig. 6d, Videos S27 and S28). Although treatment with Rsv tended to suppress the entry of FM1 − 43 dye into CTTN-knockdown cells, there was no statistically significant difference compared to vehicle (dimethyl sulfoxide; DMSO)-treated cells (Fig. 6e, Video S29 and S30).
Rsv improves membrane resealing of single fiber myotube ex vivo.
To confirm the function of SIRT1 and effects of Rsv on membrane resealing ex vivo, we separated single fiber myotubes from the flexor digitorum brevis (FDB) of mice. First, we took FDB single fiber myotubes from SIRT1-mKO mice and their litters (WT), and membrane resealing was monitored, as in the case of C2C12 myotubes. Unexpectedly, SIRT1-mKO mice showed significantly limited influx of FM1 − 43 compared to WT (Fig. 7a). Because SIRT1-mKO mice have a mild dystrophic phenotype (15), we evaluated protein levels of other membrane repair proteins, caveolin3 (Cav3) and mitsugumin 53 (MG53; 24). Cav3 tended to be higher in SIRT1-mKO mice tibialis anterior muscle than that in WT mice, and MG53 was significantly higher in SIRT1-mKO mice than WT mice (Fig. 7b).
Next, WT mice were fed a control diet or diet including a 0.4 g Rsv/kg for 5 days. FM1 − 43 uptake in single fiber myotubes from Rsv-fed WT mice was significantly limited (Fig. 7C, Video S32) compared to that in control WT mice (Fig. 7c, Video S31). In contrast, Rsv-diet failed to promote membrane resealing in SIRT1-mKO mice (Fig. 7d), indicating that Rsv promoted membrane resealing via SIRT1 ex vivo.