Summary
The structure and function of microtubules (MTs) are regulated by post-translational modifications of tubulin subunits, such as acetylation of the Lys40 residue of α-tubulin. Regulation of the organization and dynamics of MTs is essential for the precise formation of the mitotic spindle. Spindle MTs are highly acetylated, but the mechanism regulating this acetylation is largely unknown. Furry (Fry) is an evolutionarily conserved protein that binds to MTs and colocalizes with acetylated MTs in the mitotic spindle. In this study, we examined the role of Fry in the acetylation of MTs in the mitotic spindle. Depletion of Fry significantly reduced the level of MT acetylation in the mitotic spindle. Expression of the N-terminal fragment of Fry induced hyperacetylation of MTs in both mitotic and interphase cells. These results indicate that Fry promotes MT acetylation in the mitotic spindle. We also found that Fry binds to the tubulin deacetylase SIRT2, preferentially in mitotic cells. Cell-free experiments revealed that the N-terminal region of Fry is the domain responsible for binding to and inhibiting the tubulin-deacetylase activity of SIRT2. AGK2, a specific inhibitor of SIRT2, increased the level of MT acetylation in the mitotic spindle, indicating that SIRT2 is involved in the deacetylation of spindle MTs. Furthermore, AGK2 reversed the decrease in MT acetylation induced by Fry depletion. In summary, these results suggest that Fry plays a crucial role in promoting the level of MT acetylation in the mitotic spindle by inhibiting the tubulin-deacetylase activity of SIRT2.
Introduction
Microtubules (MTs) are cytoskeletal filaments that play essential roles in diverse cellular functions, including intracellular transport, organelle positioning, chromosome segregation, neurite outgrowth, ciliogenesis, cell migration and cell morphogenesis. MTs are composed of heterodimers of α- and β-tubulin subunits, which are diversified by post-translational modifications such as acetylation, detyrosination, polyglutamylation and polyglycylation. These post-translational modifications influence the interaction of MTs with a diverse set of microtubule-associated proteins (MAPs) and motor proteins, and thereby alter the dynamic properties and cellular functions of MTs. Thus, post-translational modifications of MTs generate ‘tubulin codes’ that define the organization and diverse functions of MTs (Janke and Bulinski, 2011; Verhey and Gaertig, 2007).
Acetylation of the Lys40 residue in α-tubulin (hereafter referred to as ‘tubulin acetylation’) is a well-defined MT modification that was originally discovered in Chlamydomonas (L'Hernault and Rosenbaum, 1985). Unlike other post-translational modifications, the acetylation site is positioned on the luminal side of MTs (Nogales et al., 1999). Tubulin acetylation is induced by stabilization of MTs, which can be achieved by treatment with taxol, and is therefore regarded as a marker of long-lived MTs (Piperno et al., 1987). It is currently unknown whether tubulin acetylation itself affects MT stability (Matsuyama et al., 2002; Palazzo et al., 2003; Zilberman et al., 2009). Although neither a defect nor an excess of tubulin acetylation causes a severe phenotype in Tetrahymena, nematodes and mice (Fukushige et al., 1999; Gaertig et al., 1995; Zhang et al., 2008), tubulin acetylation does modulate the ability of MTs to bind to MAPs and motor proteins, and may regulate MT stability and function (Dompierre et al., 2007; Reed et al., 2006; Sudo and Baas, 2010). Recent studies showed that tubulin acetylation restricts the number of protofilaments in nematode touch receptor neurons, suggesting that tubulin acetylation is involved in the organization of MTs (Cueva et al., 2012; Topalidou et al., 2012). However, the existence of a similar regulatory mechanism in other cell types and organisms remains unknown.
MTs are highly acetylated in mitotic cells. During metaphase, acetylated tubulin (Ac-tubulin) is enriched at interpolar and kinetochore MTs, but not at astral MTs, and Ac-tubulin becomes concentrated on the midbody during telophase and cytokinesis (Cha et al., 1998; Piperno et al., 1987). However, the lack of current knowledge about regulation of tubulin acetylation during mitosis precludes an understanding of the functional significance of this post-translational modification. The enzymes responsible for tubulin deacetylation and acetylation have been identified; histone deacetylase 6 (HDAC6) and Sirtuin 2 (SIRT2) are tubulin deacetylases, and α-tubulin acetyl transferase (αTAT, also termed MEC-17) is an acetyltransferase (Akella et al., 2010; Hubbert et al., 2002; North et al., 2003; Shida et al., 2010). SIRT2 is accumulated on the mitotic spindle and the midbody, which are both structures that contain highly acetylated MTs (North and Verdin, 2007a), indicating that the tubulin-deacetylase activity of SIRT2 is strictly controlled during mitosis. However, the mechanism of regulation of SIRT2 activity during mitosis is still unknown.
Furry (Fry) is an evolutionarily conserved protein of ∼300 kDa that contains multiple HEAT/Armadillo-like repeats in the N-terminal region (Gallegos and Bargmann, 2004). Drosophila Fry and its orthologs (Tao3p in budding yeast, Mor2p in fission yeast, and Sax-2 in nematode) genetically and physically interact with the nuclear Dbf2-related (NDR) family of Ser/Thr kinases and are implicated in the regulation of cell division, morphogenesis, polarized growth, neurite outgrowth and dendritic tiling (Cong et al., 2001; Du and Novick, 2002; Emoto et al., 2004; Hergovich et al., 2006; Hirata et al., 2002). We previously showed that mammalian Fry plays crucial roles in chromosome alignment and bipolar spindle organization during mitosis by promoting the activities of NDR and polo-like kinase 1 (Plk1) (Chiba et al., 2009; Ikeda et al., 2012). We also showed that Fry is phosphorylated by the mitotic kinases, Cdk1, Plk1 and Aurora A, and that Fry binds to MTs and colocalizes with Ac-tubulin on the mitotic spindle and the midbody during mitosis (Chiba et al., 2009; Ikeda et al., 2012). However, the potential ability of Fry to affect the acetylation, organization and functions of MTs during mitosis is currently unknown.
In this study, we show that Fry promotes acetylation of MTs in the mitotic spindle by inhibiting the tubulin-deacetylase activity of SIRT2. These results may explain how MT acetylation is maintained at the mitotic spindle, despite the existence of SIRT2 at this subcellular location.
Results
Knockdown of Fry reduces acetylation of MTs in the mitotic spindle
We previously showed that Fry is a MT-binding protein that localizes to the mitotic spindle in HeLa cells (Chiba et al., 2009). We also showed that Fry colocalizes with acetylated MTs in the spindle during metaphase and the midbody during telophase (Chiba et al., 2009). To examine the role of Fry in the acetylation of MTs in the mitotic spindle, HeLa cells were transfected with control siRNA or two independent Fry-specific siRNAs, both of which effectively suppressed expression of endogenous Fry (Fig. 1A), and then exposed to the proteasome inhibitor MG-132 to synchronize them in metaphase. Cells were then fixed and immunostained using antibodies targeting Ac-tubulin or α-tubulin. Compared with cells transfected with the control siRNA, the fluorescence intensity of Ac-tubulin in the mitotic spindle was decreased in cells transfected with the Fry-specific siRNAs (Fig. 1B). Quantitative analyses of the fluorescence intensities of Ac-tubulin and α-tubulin within the spindle area revealed that the level of Ac-tubulin, but not α-tubulin, was significantly (P<0.05) decreased in Fry-depleted cells, and thus the relative ratio of Ac-tubulin to α-tubulin was decreased in Fry-depleted cells (Fig. 1C). Previous studies showed that MG-132-induced metaphase arrest can cause chromosome scattering because of unscheduled chromatid separation and defects in spindle pole organization (Daum et al., 2011; Stevens et al., 2011), which raises the possibility that the decrease in spindle MT acetylation might be the side effects derived from MG-132 treatment. To test this possibility, we analyzed the effect of Fry depletion on spindle MT acetylation in mitotic cells prepared without MG-132 treatment. HeLa cells were synchronized at early S phase by the thymidine block method and released for 10 hours, and then the cells were fixed and analyzed by immunostaining with antibodies against Ac-tubulin and α-tubulin. The level of spindle MT acetylation was significantly (P<0.05) reduced upon Fry depletion in mitotic cells that were not exposed to MG-132 (supplementary material Fig. S1), indicating that the change in spindle MT acetylation is not due to the side effect of MG-132 treatment. Fry depletion causes defects in chromosome alignment (Chiba et al., 2009), which may alter MT stability and thereby indirectly affect MT acetylation. To define the role of Fry in MT acetylation, we sorted the data in Fig. 1C into two categories; the cells with the metaphase spindle consisting of fully aligned chromosomes and the cells with the spindle in which chromosomes were not fully aligned. In both types of spindles, Fry depletion significantly (P<0.05) decreased the level of spindle MT acetylation (supplementary material Fig. S2). These results suggest that Fry promotes MT acetylation in the mitotic spindle and this function is not simply through the action of Fry on chromosome alignment.
Knockdown of Fry reduces MT acetylation in mitotic spindles. (A) Validation of the Fry-specific siRNAs. HeLa cells transfected with control or two different Fry-specific siRNAs were cultured for 60 hours and then cell lysates were immunoblotted with anti-Fry and anti-α-tubulin antibodies. (B) Knockdown of Fry reduces MT acetylation in mitotic spindles. HeLa cells transfected with control or Fry-specific siRNAs were synchronized in metaphase by MG-132 treatment, fixed with methanol, and then stained with anti-α-tubulin (green) and anti-Ac-tubulin (red) antibodies. DNA was stained with DAPI (blue). Scale bar: 5 µm. (C) Quantitative analysis of Ac-tubulin and α-tubulin in the mitotic spindle. The left and middle panels show the relative fluorescence intensities of Ac-tubulin and α-tubulin, respectively. The right panel shows the relative ratio of Ac-tubulin to α-tubulin. Values are the means ± s.e.m. from four independent experiments (at least 40 cells for each experiment). *P<0.05; n.s., not significant, by one-way ANOVA followed by Dunnett's test. (D) Immunoblot analyses of the level of MT acetylation in mitotic cells. HeLa cells transfected with control or Fry-specific siRNAs were synchronized in metaphase by MG-132 treatment and then collected by the mechanical shake-off procedure. Cell lysates were analyzed by immunoblotting with the indicated antibodies. The relative ratios of Ac-tubulin to α-tubulin were measured by densitometric analysis of the α-tubulin and Ac-tubulin immunoblot data. Values are the means ± s.e.m. from four independent experiments. *P<0.05. (E) Immunoblot analyses of the level of MT acetylation in asynchronous cells. HeLa cells transfected with control or Fry-specific siRNAs were asynchronously cultured for 48 hours. Cell lysates were analyzed by immunoblotting with anti-Ac-tubulin and anti-α-tubulin antibodies. The relative ratios of Ac-tubulin to α-tubulin were measured as in D. Values are the means ± s.e.m. from three independent experiments. n.s., not significant.
The reduction in the level of Ac-tubulin in Fry-depleted cells was also confirmed by immunoblot analyses of lysates from mitotic HeLa cells (Fig. 1D). Densitometric analysis showed that, compared with control cells, the ratio of Ac-tubulin to α-tubulin was significantly (P<0.05) reduced in Fry-depleted cells (Fig. 1D). Because tubulin detyrosination and polyglutamylation are known to occur in the mitotic spindle (Gundersen and Bulinski, 1986; Wolff et al., 1992), the effect of Fry-knockdown on these modifications was also investigated. In contrast to the effect on tubulin acetylation, depletion of Fry had no effect on the level of tubulin detyrosination and polyglutamylation (Fig. 1D), indicating that Fry is specifically involved in acetylation of MTs.
We also examined the effect of Fry depletion on MT acetylation in interphase cells. HeLa cells were transfected with control siRNA or Fry siRNA and the lysates of asynchronous cells were analyzed by immunoblotting with antibodies against Ac-tubulin and α-tubulin. Fry depletion had no effect on the ratio of Ac-tubulin to α-tubulin (Fig. 1E), indicating that Fry is preferentially involved in acetylation of MTs in mitotic cells.
The N-terminal fragment of Fry induces MT acetylation
To further examine the role of Fry in MT acetylation, deletion mutants of murine Fry were constructed and the effects of expression of these mutants on the level of MT acetylation in interphase HeLa cells were analyzed. Cells were transfected with (Myc+His)-tagged full-length Fry (Fry-FL) or the Fry-N (amino acids 1–730), Fry-M (amino acids 718–1575), or Fry-C (amino acids 1550–3020) deletion mutants (Fig. 2A), and then immunostained with antibodies targeting Ac-tubulin and Myc (Fig. 2B) and the relative fluorescence intensity of Ac-tubulin in each cell was quantified (Fig. 2C). Control cells were transfected with GFP. Expression of Fry-FL, Fry-M and Fry-C had no significant effect on the level of MT acetylation, whereas expression of Fry-N induced a significant (P<0.05) increase in the level of MT acetylation (Fig. 2B,C). When we compared the level of expression of Fry-FL and Fry-N, together with the level of tubulin acetylation, the level of expression of Fry-N was lower than that of Fry-FL, but the level of Ac-tubulin in Fry-N-expressing cells was higher than that in Fry-FL-expressing cells (supplementary material Fig. S3). These results indicate that the N-terminal region (amino acids 1–730) of Fry has the ability to increase the level of MT acetylation in cells. Since Fry-FL did not promote MT acetylation, the activity of the N-terminal region of Fry may be masked in the full-length protein in interphase cells.
The N-terminal fragment of Fry induces MT hyperacetylation. (A) Schematic illustration of mouse Fry and its deletion mutants. The effects of each construct on MT hyperacetylation is shown. Numbers indicate the amino acid residues. (B) Fry-N induces MT hyperacetylation in interphase HeLa cells. Cells were transfected with GFP, or (Myc+His)-tagged full-length (FL) Fry or its deletion mutants, and then fixed and stained with anti-Ac-tubulin (red) and anti-Myc (green) antibodies. DNA was stained with DAPI (blue). Scale bar: 10 µm. (C) Quantitative analysis of the relative Ac-tubulin fluorescence intensity. Values are the means ± s.e.m. from four independent experiments (more than 10 cells for each experiment). *P<0.05; n.s., not significant, by one-way ANOVA followed by Dunnett's test. (D) Fry-N localizes to MTs in interphase HeLa cells and induces MT bundling. Cells transfected with (Myc+His)-tagged Fry deletion mutants were permeabilized with 0.1% Triton X-100, and then fixed and stained with anti-α-tubulin (red) and anti-Myc (green) antibodies. DNA was stained with DAPI (blue). Scale bar: 10 µm.
We also examined the effect of expression of the deletion mutants of Fry on MT organization and their localization by immunostaining cells with antibodies targeting α-tubulin and Myc. To more clearly detect the positions of the Fry proteins on MTs, cells were permeabilized with Triton X-100 before immunostaining. Expression of Fry-N induced drastic changes in MT organization; aberrant bundling of MTs around the nucleus was observed and Fry-N was localized to the MT bundles (Fig. 2D). By contrast, the expression of Fry-M or Fry-C had no effect on MT organization (Fig. 2D). Additionally, MT co-sedimentation assays revealed that Fry-N co-precipitated with polymerized MTs more strongly than Fry-FL (supplementary material Fig. S4). Taken together, these results indicate that the N-terminal region of Fry is the domain responsible for binding to and inducing acetylation and bundling of MTs.
We next examined the effect of Fry-N expression on MT acetylation in mitotic spindles. HeLa cells transfected with (Myc+His)-tagged Fry-N were synchronized in metaphase by MG-132 treatment, and then immunostained with antibodies against Ac-tubulin and α-tubulin. Expression of Fry-N significantly (P<0.05) increased the ratio of Ac-tubulin to α-tubulin, compared with control cells (Fig. 3). Neither the mitotic spindle organization nor the localization of endogenous Fry on the spindle was affected by the expression of Fry-N (Fig. 3; supplementary material Fig. S5). These results indicate that Fry-N has the potential to promote spindle MT acetylation in mitotic cells. We also tried to analyze the effect of overexpression of Fry-FL on spindle MT acetylation, but the trial was unsuccessful because sufficient amounts of Fry-FL-expressing mitotic cells were not obtained under our experimental conditions.
The N-terminal fragment of Fry increases the level of MT acetylation in mitotic spindles. (A) Effect of Fry-N expression on the level of MT acetylation in mitotic spindles. HeLa cells transfected with (Myc+His)-tagged Fry-N or Fry-N(Δ519–540) were synchronized in metaphase by MG-132 treatment, fixed with methanol, and then stained with anti-α-tubulin (green) and anti-Ac-tubulin (red) antibodies. DNA was stained with DAPI (blue). Scale bar: 5 µm. (B) Quantitative analysis of the relative ratio of Ac-tubulin to α-tubulin in the mitotic spindle. Values are the means ± s.e.m. from four independent experiments (more than 20 cells for each experiment). *P<0.05; n.s., not significant, by one-way ANOVA followed by Dunnett's test.
Effects of taxol and monastrol on the Fry knockdown-induced decrease in spindle MT acetylation
Fry knockdown causes chromosome misalignment and spindle disorganization (Chiba et al., 2009; Ikeda et al., 2012). Because defects in spindle integrity can affect the stability of MTs and the level of MT acetylation in the spindle, the Fry knockdown-induced decrease in spindle MT acetylation may simply reflect the MT instability in Fry knockdown cells. To address this issue, we examined the effect of taxol, a drug stabilizing MTs, on the Fry depletion-induced changes in spindle MT acetylation. Taxol treatment resulted in MT asters in mitotic cells, as previously reported (De Brabander et al., 1981). The ratio of Ac-tubulin to α-tubulin in mitotic asters was significantly reduced (P<0.05) in Fry-depleted cells, compared with that in control cells (Fig. 4A). In contrast, the ratio of polyglutamylated tubulin to α-tubulin in mitotic asters was not affected by Fry depletion (Fig. 4B). These results suggest that Fry depletion specifically decreases the level of spindle MT acetylation even after MTs are stabilized by taxol treatment and that the decrease in MT acetylation upon Fry depletion is not simply caused by a decrease in MT stability.
Effects of taxol and monastrol on the Fry knockdown-induced decrease in MT acetylation in mitotic cells. (A,B) Effects of Fry knockdown on MT acetylation and polyglutamylation in taxol-treated mitotic cells. HeLa cells transfected with control or Fry-specific siRNAs were treated with 100 nM taxol for 1 hour, fixed with methanol, and then stained with anti-α-tubulin (green) and anti-Ac-tubulin or anti-polyGlu-tubulin (red) antibodies. DNA was stained with DAPI (blue). Scale bar: 5 µm. The right panels show the relative ratio of the fluorescence intensity of Ac-tubulin to α-tubulin (A) or polyGlu-tubulin to α-tubulin (B) in the mitotic spindle. Values are the means ± s.e.m. from four independent experiments (more than 30 cells for each experiment). *P<0.05; n.s., not significant, by one-way ANOVA followed by Dunnett's test. (C) Effect of Fry depletion on MT acetylation in monastrol-induced monopolar spindles. HeLa cells transfected with control or Fry-specific siRNAs were treated with 100 µM monastrol for 4 hours, fixed with methanol, and then stained with anti-α-tubulin (green) and anti-Ac-tubulin (red) antibodies. DNA was stained with DAPI (blue). Scale bar: 5 µm. The right panel shows the relative ratio of the fluorescence intensity of Ac-tubulin to α-tubulin in the monopolar spindle. Values are the means ± s.e.m. from four independent experiments (more than 40 cells for each experiment). *P<0.05, by one-way ANOVA followed by Dunnett's test.
We also examined the effect of Fry depletion on spindle MT acetylation after the treatment of cells with monastrol, a drug that inhibits the kinesin Eg5 and produces the monoaster spindle resulting in the reduction in the tension of kinetochore MTs (Kapoor et al., 2000). Monastrol treatment formed the monopolar spindle in mitotic cells. Fry depletion caused the significant decrease (P<0.05) in the ratio of Ac-tubulin to α-tubulin in monoasters in monastrol-treated mitotic cells (Fig. 4C). This further suggests that Fry depletion decreases the level of MT acetylation even in monopolar spindles where the tension of kinetochore MTs is decreased.
Fry binds to SIRT2 tubulin deacetylase in mitotic cells
Next, we investigated the mechanism of Fry-mediated acetylation of MTs. SIRT2 is an NAD+-dependent tubulin deacetylase (North et al., 2003). To explore the possibility that Fry increases MT acetylation by inhibiting the tubulin-deacetylase activity of SIRT2, we examined the ability of Fry to bind to SIRT2. HEK293T cells expressing GFP-tagged SIRT2 were synchronized in S phase by a double-thymidine block or in the early mitotic phase by nocodazole treatment, and then cell lysates were subjected to immunoprecipitation with an anti-Fry antibody. GFP–SIRT2 co-precipitated with endogenous Fry in lysates from the mitotic cells, but not in lysates from the S-phase cells (Fig. 5A), indicating that Fry preferentially binds to SIRT2 in the mitotic phase.
Fry binds to SIRT2 in the mitotic phase and via the N-terminal region. (A) Fry binds to SIRT2 in mitotic cells. HEK293T cells transfected with GFP-SIRT2 were synchronized in S phase by a double-thymidine block or in M phase by treatment with nocodazole. Cell lysates (1.5×106 cells) were immunoprecipitated with an anti-Fry antibody and the portions (40%) of the immunoprecipitates were analyzed by immunoblotting with anti-SIRT2 and anti-Fry antibodies. Cell lysates were analyzed by immunoblotting with anti-SIRT2, anti-MPM2 and anti-α-tubulin antibodies. (B) Schematic illustration of Fry and its deletion mutants. Results of SIRT2 binding assays are also shown on the right. (C) Mapping the SIRT2-binding region of Fry. HEK293T cells were transfected with (Myc+His)-tagged Fry-FL or the indicated deletion mutants (N, M, C). Cell lysates were subjected to pull-down assays using GST- or GST-SIRT2-conjugated beads. Precipitates were immunoblotted with an anti-Myc antibody. (D) Co-precipitation assays. HEK293T cells were co-transfected with FLAG-SIRT2 and GFP-tagged fragments of Fry. Cell lysates were immunoprecipitated with an anti-FLAG antibody. Immunoprecipitates were immunoblotted with anti-GFP and anti-FLAG antibodies. (E) Amino acids 519–540 of Fry are required for SIRT2 binding. HEK293T cells were transfected with (Myc+His)-tagged Fry-N or the Fry-N(Δ519–540) deletion mutant and cell lysates were subjected to GST pull-down assays as described in B.
The N-terminal region of Fry binds to SIRT2
To map the SIRT2-binding region of Fry, (Myc+His)-tagged Fry-FL and its deletion mutants (Fry-N, Fry-M and Fry-C; Fig. 5B) were expressed in HEK293T cells and the cell lysates were subjected to pull-down assays using GST–SIRT2-conjugated beads. Fry-N was pulled down with GST–SIRT2, but Fry-M and Fry-C were not (Fig. 5C), indicating that the N-terminal region (amino acids 1–730) of Fry binds to SIRT2. Fry-FL was not pulled down with GST–SIRT2. Together with the result of the co-precipitation assays (Fig. 5A), these data suggest that the N-terminal region of Fry is responsible for its SIRT2-binding activity and that this activity in the full-length protein may be masked in interphase and unmasked in the mitotic phase.
To further define the SIRT2-binding region of Fry, GFP-tagged fragments of Fry (fragments I, II, III, IV; Fig. 5B) were coexpressed with FLAG-tagged SIRT2 in HEK293T cells and the cell lysates were subjected to immunoprecipitation with an anti-FLAG antibody. The Fry-I (amino acids 1–540) and Fry-II (amino acids 519–1500) fragments co-precipitated with FLAG–SIRT2, but the Fry-III (amino acids 1501–2400) and Fry-IV (amino acids 2401–3020) fragments did not (Fig. 5D). Since Fry-I and Fry-II share a common overlapping sequence (amino acids 519–540), a deletion mutant of Fry-N [(Fry-N(Δ519–540)], in which amino acids 519–540 were removed, was constructed and the ability of the mutant protein to bind to SIRT2 was examined. GST pull-down assays revealed that Fry-N bound to GST–SIRT2, whereas Fry-N(Δ519–540) did not (Fig. 5E), indicating that the N-terminal region of Fry binds to SIRT2 and the region of amino acids at positions 519–540 is necessary for this interaction.
The SIRT2-binding activity of Fry-N is required for MT acetylation
To examine whether the SIRT2-binding activity of the N-terminal region of Fry is required for its ability to induce MT acetylation, the Fry-N and Fry-N(Δ519–540) deletion mutants were expressed in HeLa cells and the level of MT acetylation was measured in interphase and mitotic cells. In contrast to the increased MT acetylation induced by the expression of Fry-N, Fry-N(Δ519–540) did not induce MT acetylation in either interphase cells (Fig. 2B,C) or mitotic cells (Fig. 3A,B). Therefore, it is likely that the binding activity of Fry-N to SIRT2 is required for the induction of MT acetylation in cells.
Fry inhibits the tubulin-deacetylase activity of SIRT2
Next, we examined whether Fry inhibits the tubulin-deacetylase activity of SIRT2. Whereas deleted in breast cancer gene 1 protein (DBC1; gene name KIAA1967) has been reported to be an inhibitor of SIRT1, an NAD+-dependent protein deacetylase that targets diverse proteins (Kim et al., 2008; Zhao et al., 2008), a protein responsible for inhibition of SIRT2 has not yet been identified. Binding of DBC1 to SIRT1 is prevented by the mutation of His363 to Tyr in the catalytic domain of SIRT1, suggesting that DBC1 inhibits SIRT1 by directly binding to its catalytic domain and competitively blocking access to substrates (Kim et al., 2008; Zhao et al., 2008). GST pull-down assays were performed to determine the ability of Fry-N to bind to a SIRT2 mutant in which the conserved His187 in the catalytic domain was replaced by Tyr (H187Y) (Frye, 1999). Fry-N bound to wild-type GST–SIRT2 but not the H187Y mutant (Fig. 6A), suggesting that Fry binds to the catalytic domain of SIRT2 in a manner analogous to the interaction between DBC1 and SIRT1.
Fry binds to the catalytic domain of SIRT2 and inhibits its tubulin-deacetylase activity. (A) Mutation (H187Y) of SIRT2 in the catalytic domain abolishes its ability to bind to Fry-N. HEK293T cells were transfected with Fry-N-(Myc+His) and cell lysates were subjected to pull-down assays using wild-type (WT) or the H187Y mutant of GST–SIRT2. Precipitates were immunoblotted with an anti-Myc antibody. (B) Fry inhibits SIRT2 activity. FLAG-tagged SIRT2 expressed in HEK293T cells was precipitated with an anti-FLAG antibody or GST–FRY-II and the precipitates were incubated for 2 hours with cell lysates containing Ac-tubulin, together with 1 mM NAD+ or 5 mM nicotinamide (NAM). Reaction mixtures were analyzed by immunoblotting with anti-Ac-tubulin, anti-α-tubulin and anti-FLAG antibodies. (C) Densitometric measurements of the immunoblots shown in B. The relative intensities of Ac-tubulin and FLAG–SIRT2 staining in lanes 4 and 5 of B are shown. Values are the means ± s.e.m. from five independent experiments. *P<0.05; n.s., not significant, by unpaired two-tailed Student's t-test.
To examine the effect of Fry binding on the MT-deacetylating activity of SIRT2, FLAG-tagged SIRT2 expressed in HEK293T cells was precipitated with an anti-FLAG antibody or GST–FRY-II and subjected to an in vitro MT deacetylation assay (North et al., 2003). As expected, FLAG-tagged SIRT2 that was precipitated with the anti-FLAG antibody exhibited dose-dependent tubulin-deacetylating activity in the presence of NAD+ and this activity was inhibited by nicotinamide (NAM), a specific inhibitor of class III HDACs (Luo et al., 2001). By contrast, the tubulin-deacetylating activity of FLAG–SIRT2 was substantially reduced when SIRT2 was bound to GST–Fry-II (compare lanes 4 and 5 in Fig. 6B). Quantitative analysis of the level of Ac-tubulin (lanes 4 and 5 in Fig. 6B) revealed that SIRT2 bound to the Fry-II fragment exhibited significantly (P<0.05) lower tubulin-deacetylating activity than Fry-II-unbound SIRT2 (Fig. 6C). Measurements of the time-courses of MT deacetylation reactions further confirmed that the MT-deacetylating activity of SIRT2 was attenuated by complexing with GST-Fry-II (supplementary material Fig. S6). Together these results suggest that Fry attenuates the tubulin-deacetylase activity of SIRT2 by binding of the N-terminal region of Fry to the catalytic domain of SIRT2.
Fry promotes acetylation of spindle MTs through suppression of SIRT2
Since both Fry and SIRT2 localize to the mitotic spindle (Chiba et al., 2009; North and Verdin, 2007a), we speculated that the observed reduction in MT acetylation in Fry-depleted cells is caused by the failure of Fry-mediated suppression of SIRT2 activity on the mitotic spindle. To investigate this possibility, we examined whether inhibition of SIRT2 restores the decrease in acetylation of MTs on the mitotic spindle in Fry-depleted cells. HeLa cells were synchronized in metaphase using MG-132, and then fixed and stained with anti-Ac-tubulin and anti-α-tubulin antibodies. The SIRT2-specific inhibitor AGK2 was used to inhibit SIRT2 activity (Outeiro et al., 2007). In control cells, exposure to AGK2 considerably increased the level of MT acetylation on the mitotic spindle, indicating that SIRT2 acts as the tubulin-deacetylase responsible for reducing acetylation of spindle MTs in vivo. Knockdown of Fry with specific siRNAs decreased the level of MT acetylation, but this was recovered to levels similar to that in control cells, when AGK2 was added (Fig. 7A,B). These results suggest that Fry plays a crucial role in increasing the level of spindle MT acetylation by inhibiting the MT-deacetylating activity of SIRT2.
The decrease in spindle MT acetylation induced by depletion of Fry is restored by inhibition of SIRT2. (A) Inhibition of SIRT2 increases the level of MT acetylation in the mitotic spindle in control and Fry-depleted cells. HeLa cells transfected with control or Fry-specific siRNAs (#1 and #2) were treated with 30 µM AGK2 at the time of the release from single thymidine block. Cells were then synchronized in metaphase by treatment with MG-132, fixed with methanol, and immunostained with α-tubulin (green) and Ac-tubulin (red) antibodies. DNA was stained with DAPI (blue). Scale bar: 5 µm. (B) Quantitative analysis of the relative ratio of Ac-tubulin to α-tubulin in the mitotic spindle. Values are the means ± s.e.m. from four independent experiments (more than 15 cells for each experiment). *P<0.05; n.s., not significant, by one-way ANOVA followed by Dunnett's test. **P<0.05, analyzed by unpaired two-tailed Student's t-test.
Discussion
Fry plays a crucial role in the formation of the bipolar spindle during mitosis. We previously demonstrated that Fry-induced activation of NDR1 and Plk1 is required for the alignment of chromosomes on the metaphase plate and spindle pole integrity during early mitosis (Chiba et al., 2009; Ikeda et al., 2012). However, the ability of Fry to directly bind to MTs and colocalize with acetylated MTs in the mitotic spindle raises the possibility that Fry is directly involved in the regulation of MT organization and function. In this study, we investigated the role of Fry in acetylation of MTs in the mitotic spindle. Our results showed that: (1) depletion of Fry reduces the level of MT acetylation in the mitotic spindle; (2) expression of the N-terminal fragment of Fry enhances MT acetylation in both interphase and mitotic cells; (3) Fry binds to SIRT2 during mitosis and this interaction occurs between the N-terminal region of Fry and the catalytic domain of SIRT2; (4) Fry inhibits the tubulin-deacetylase activity of SIRT2; and (5) the reduction in the level of MT acetylation in the mitotic spindle caused by depletion of Fry is restored by SIRT2 inhibition. Based on these observations, we propose that Fry plays a crucial role in the promotion of MT acetylation in the mitotic spindle by inhibiting the tubulin-deacetylase activity of SIRT2.
Acetylation of spindle MTs is closely related to their stability. It was reported that depletion of MAPs causes MT instability and loss of MT acetylation in spindles, along with a reduction in spindle width and MT density (Cha et al., 1999). We showed that Fry-N has the potential to induce MT bundling, which raises the possibility that the decrease in MT acetylation in Fry-depleted spindles may simply reflect MT instability. In this respect, however, we showed that knockdown of Fry reduced the level of MT acetylation but did not affect the levels of MT polyglutamylation and detyrosination in mitotic cells, indicating that Fry has a specific role in the regulation of MT acetylation. In addition, Fry depletion decreased the level of spindle MT acetylation even in the cells treated with taxol or monastrol. Furthermore, we provided evidence that Fry is capable of suppressing the MT-deacetylating activity of SIRT2. These results indicate that the decrease in MT acetylation in the spindle of Fry-depleted cells is not simply caused by a decrease in MT stability, but is principally caused by the loss of Fry-mediated suppression of SIRT2, which leads to excessive activation of SIRT2 and subsequent MT deacetylation.
During interphase, Fry-N localized to MTs, induced MT acetylation and bundling, and bound to SIRT2; however, Fry-FL did not exhibit these activities. In addition, endogenous Fry bound to SIRT2 in the mitotic phase, but not in interphase. Furthermore, Fry is diffusely distributed in the cytoplasm during interphase but is localized to spindle MTs during mitosis (Chiba et al., 2009). These results suggest that the abilities of the N-terminal region of Fry to bind to and bundle MTs, as well as bind to and inhibit SIRT2 are masked in the full-length protein during interphase, possibly as a result of intramolecular autoinhibition, and that this inhibition is relieved during the mitotic phase. In accord with this, depletion of Fry had no effect on MT acetylation in interphase cells. We previously showed that Fry is highly phosphorylated during mitosis and that the C-terminal region of Fry is effectively phosphorylated by the mitotic kinases Cdk1, Plk1 and Aurora A (Ikeda et al., 2012). It is possible that phosphorylation of the C-terminal region of Fry is involved in unmasking and promoting the MT- and SIRT2-binding activities of the N-terminal region during mitosis.
This study describes a novel mechanism of SIRT2 regulation by a direct protein–protein interaction. DBC1, a negative regulator of SIRT1, directly binds to the catalytic domain of SIRT1 and competitively blocks binding of substrates (Kim et al., 2008; Zhao et al., 2008). Similar to the interaction between DBC1 and SIRT1, a mutation (H187Y) in the catalytic domain of SIRT2 abolished the interaction between Fry and SIRT2, suggesting that Fry binds to the catalytic domain of SIRT2. Since both Fry and SIRT2 localize to spindle MTs (Chiba et al., 2009; North and Verdin, 2007a), we propose that Fry interacts with and specifically attenuates the tubulin-deacetylating activity of SIRT2 at the mitotic spindle.
Members of the HDAC family, including SIRT2, regulate mitotic progression by targeting both histone and non-histone substrates. Expression of the SIRT2 protein is upregulated during mitosis and SIRT2 is involved in the regulation of chromatin condensation during the G2–M transition by deacetylating Lys16 in histone H4 (H4K16) and the control of anaphase-promoting complex (also called cyclosome) activity by deacetylating its coactivators, Cdh1 and Cdc20 (Kim et al., 2011; Vaquero et al., 2006). Although SIRT2 localizes to the mitotic spindle, the spindle MTs are highly acetylated; therefore, the role of SIRT2 in the deacetylation of spindle MTs during mitosis was unclear. Cyclin-dependent kinases (Cdks) phosphorylate and inhibit the deacetylase activity of SIRT2 (Dryden et al., 2003; North and Verdin, 2007b; Pandithage et al., 2008). The cyclin-A–Cdk2-complex-mediated phosphorylation and inhibition of SIRT2 cause hyperacetylation of spindle MTs (Orpinell et al., 2010), indicating that SIRT2 is involved in the deacetylation of spindle MTs. In this study, AGK2, a specific inhibitor of SIRT2, increased the level of MT acetylation in the mitotic spindle, which provides direct evidence that SIRT2 is involved in spindle MT deacetylation during mitosis.
Although SIRT2 is localized to the spindle and deacetylates several substrates (H4K16, Cdh1 and Cdc20) during mitosis, the spindle MTs are still highly acetylated, suggesting that SIRT2 activity is spatially and temporally controlled and the activity on spindle MTs is specifically suppressed. In this study, we showed that Fry localizes to the mitotic spindle and inhibits SIRT2, and that MT deacetylation induced by depletion of Fry is recovered by treatment of cells with AGK2. These data suggest that Fry plays a critical role in maintaining the high level of MT acetylation in the spindle through suppression of SIRT2 activity. Because Fry accumulates on the spindle MTs during metaphase, we propose that Fry specifically blocks SIRT2 activity on the spindle MTs and thereby inhibits SIRT2-mediated deacetylation of spindle MTs without affecting SIRT2-mediated deacetylation of other substrates. In summary, Fry plays a crucial role in controlling SIRT2 activity and MT acetylation in the spindle.
The data presented in this study indicate that Fry promotes acetylation of MTs in the mitotic spindle by inhibition of SIRT2. The role of MT acetylation in spindle organization and integrity is still unclear. Although changes in the level of MT acetylation do not affect MT stability directly (Palazzo et al., 2003), insufficient MT acetylation in Fry-depleted cells might alter the architecture and functions of MTs through effects on MT-binding proteins. Recent studies of neuronal cells and fibroblasts demonstrated that tubulin acetylation influences the activity of kinesin motors and the MT-severing protein katanin, both of which are essential for the proper formation of mitotic spindles (Dompierre et al., 2007; Reed et al., 2006; Sudo and Baas, 2010). Therefore, future studies are required to elucidate the roles of post-translational modifications of tubulin in bipolar spindle organization and mitotic progression.
Materials and Methods
Reagents and antibodies
Nocodazole, taxol, monastrol, MG-132, thymidine, NAD+ and nicotinamide were purchased from Sigma. Rabbit antibodies specific to human Fry were raised against C-terminal peptides of Fry (FC-14) (Chiba et al., 2009) or amino acids 2437–2455. Monoclonal antibodies against c-Myc (9E10; Roche), α-tubulin (B-5-1-2; Sigma), Ac-tubulin (6-11B-1; Sigma), polyGlu-tubulin (GT335; Enzo Life Sciences, New York), MPM2 (05-368; Millipore) and FLAG (F3165; Sigma), and polyclonal antibodies against SIRT2 (09-843; Millipore), deTyr-tubulin (AB3201; Millipore), GFP (A6455; Molecular Probes) and c-Myc (562; Medical and Biological Laboratories, Nagoya, Japan), and Alexa-Fluor-488-conjugated α-tubulin (B-5-1-2; Millipore) were purchased from the specific suppliers listed.
Plasmid construction
The cDNA coding for mouse Fry was cloned as described previously (Chiba et al., 2009; Ikeda et al., 2012). The plasmids coding for GFP–SIRT2 WT and FLAG–SIRT2 WT were constructed by subcloning PCR-amplified SIRT2 cDNA into expression vectors. The SIRT2 point mutant (H187Y) was constructed using the QuickChange Site-Directed Mutagenesis Kit (Stratagene). Plasmids expressing truncated mutants of Fry were constructed as described previously (Ikeda et al., 2012). For protein expression in the baculovirus system, cDNAs coding for GST-tagged proteins were subcloned into a pFastBac1 vector (Invitrogen).
RNA interference
The Stealth siRNA sequences targeting human Fry were as follows: siRNA #1, 5′-UUUACUUCCCGGAGCAGGAAGUUGG-3′; and siRNA #2, 5′-UAAAGUAUCCCGCUCUAGGGCUCCA-3′ (Invitrogen). A Stealth RNAi negative control (Invitrogen) was used as the control siRNA.
Cell culture, transfection and synchronization
HeLa and HEK293T cells were cultured in DMEM supplemented with 10% FCS. Transfections were performed using FuGENE HD, FuGENE6 (Promega) or XtremeGENE HP (Roche). Transfection of cells with duplexed siRNAs was performed by incubation of cells with 50 nM Fry-specific siRNA or negative control siRNA in Lipofectamine RNAi MAX reagent (Invitrogen). To examine the effects of knockdown of Fry on the level of spindle MT acetylation, HeLa cells transfected with siRNAs were cultured in DMEM for 12 hours and then thymidine-containing medium for 24 hours, followed by release from thymidine for 8 hours and culture in medium containing 10 µM MG-132 for a further 2 hours. To examine the effect of Fry-N expression on spindle MT acetylation, HeLa cells transfected with plasmid DNAs were cultured in DMEM for 24 hours and then in medium containing 10 µM MG-132 for a further 2 hours. To examine the effects of taxol or monastrol, HeLa cells were cultured in thymidine-containing medium for 24 hours, followed by release from thymidine for 9 hours and culture in medium containing 100 nM taxol for 1 hour, or release from thymidine for 6 hours and culture in medium containing 100 µM monastrol for 4 hours. To collect mitotic cells for immunoblot analyses, HeLa cells transfected with siRNAs were synchronized in metaphase by MG-132 treatment and mitotic cells were selectively collected by the mechanical shake-off procedure. To collect asynchronous cells, HeLa cells transfected with siRNAs were cultured for 48 hours and then harvested. To induce cell cycle arrest in the S and M phases, HEK293T cells were cultured in medium containing thymidine for 15 hours. Cells were released from thymidine exposure, transfected with plasmids, and incubated for 8 hours. Then, they were incubated for 17 hours with either thymidine, to induce S phase block, or nocodazole, to induce M phase block.
Immunofluorescence staining
HeLa cells grown on poly-L-lysine-coated coverslips were washed with PBS and then fixed with either 4% formaldehyde for 20 minutes at 37°C followed by methanol for 10 minutes at −20°C, or methanol alone. For immunostaining of Fry mutants and MTs, cells were washed with PHEM buffer [100 mM PIPES (pH 6.8), 20 mM HEPES (pH 6.9), 4 M glycerol, 5 mM EGTA and 2 mM MgCl2] and permeabilized with 0.1% Triton X-100 in PHEM buffer for 1 minute. Cells were then blocked with 2% FCS in PBS for 30 minutes and incubated with primary antibodies for 90 minutes at room temperature or overnight at 4°C, and then secondary antibodies for 90 minutes at room temperature. DNA was stained with 0.5 µg/ml DAPI.
Whole cell extraction
To examine the effect of Fry depletion on tubulin acetylation by immunoblot analysis, whole cell lysates were prepared. Mitotic or asynchronous HeLa cells were extracted with extraction buffer [1% SDS, 50 mM Tris-HCl (pH 7.5) and 1 mM DTT] at 100°C for 20 minutes. After addition of SDS sample buffer, extracts were centrifuged and supernatants were subjected to immunoblot analysis.
Microscopy and image analysis
To measure the fluorescence intensity in mitotic spindles, z-stack images were acquired at 0.75 µm intervals using a Zeiss LSM 710 confocal imaging system (Carl Zeiss) equipped with a PL Apo 63× oil-immersion objective lens (NA 1.4) driven by LSM ZEN 2009 software (Carl Zeiss). The z-stack images were average-intensity projected and the mean fluorescence intensity in each manually defined spindle area was measured. To measure the fluorescence intensity in interphase cells, fluorescent images were acquired using a Leica DMI6000B fluorescence microscope equipped with a PL Apo 63× oil-immersion objective lens (NA 1.3) and cooled charge-coupled device camera (Cool SNAP HQ; Roper Scientific) driven by LAS AF Imaging Software (Leica). The mean fluorescence intensity in each cell area was measured. Z-stack projection and measurement of fluorescence intensity were performed using ImageJ software (NIH).
Immunoprecipitation and immunoblotting
Immunoprecipitation was performed as described previously (Toshima et al., 2001). Lysates of HEK293T cells expressing GFP–SIRT2 or FLAG–SIRT2 and GFP–Fry-I, –Fry-II, –Fry-III and –Fry-IV were incubated with protein-A–Sepharose and anti-Fry (FC-14) antiserum or an anti-FLAG antibody. After centrifugation at 200 g for 3 minutes, the pellets were analyzed by immunoblotting as described previously (Amano et al., 2002).
Pull-down assay
GST and GST–SIRT2 were expressed in Sf21 cells using a baculovirus system and then purified using glutathione–Sepharose. Cell lysates were pre-cleared with protein-A–Sepharose and the supernatants were incubated with GST-tagged proteins bound to glutathione–Sepharose beads. After centrifugation, the beads were washed three times and samples were subjected to immunoblot analyses.
In vitro MT deacetylation assay
In vitro MT deacetylation assays were performed as described previously (North et al., 2003). FLAG–SIRT2 was precipitated with an anti-FLAG antibody (Sigma) or GST–FRY-II (amino acids 519–1500). Precipitates were washed twice with HDAC buffer [50 mM Tris-HCl (pH 9.0), 4 mM MgCl2 and 0.2 mM DTT], resuspended in HDAC buffer containing HEK293T cell lysates as a substrate and 1 mM NAD or 5 mM nicotinamide, and then incubated for 2 hours at room temperature with constant agitation. The reactions were stopped by adding SDS sample buffer and samples were subjected to immunoblot analyses.
Microtubule co-sedimentation assay
Microtubule co-sedimentation assay was performed as previously described (Jeong et al., 2007). HEK293T cells were transfected with (Myc+His)-tagged Fry-FL or Fry-N and treated with 1 µM taxol for 1 hour. Cells were washed with PEM buffer [100 mM PIPES (pH 6.9), 1 mM EGTA, 1 mM MgCl2] and then lysed with MT-stabilizing lysis buffer [100 mM PIPES (pH 6.9), 1 mM EGTA, 1 mM MgCl2, 20 µM taxol, 1% NP-40, 1 mM GTP, 1 mM PMSF, 10 µg/ml leupeptin] for 1 minute at room temperature. Cell lysates were centrifuged at 16,000 g for 5 minutes at room temperature. The supernatant was clarified by ultracentrifugation at 100,000 g for 30 minutes. Equal amounts of supernatant and pellet fractions were analyzed by immunoblotting with anti-Myc and anti-α-tubulin antibodies.
Statistical analysis
Data are presented as the means ± s.e.m. of more than three independent experiments. All statistical analyses were performed using Prism 6 (GraphPad Software). The P-values were calculated using unpaired two-tailed Student's t-tests for pairwise data comparisons (in Fig. 6C; Fig. 7B) or using one-way ANOVA followed by Dunnett's test for multiple data set comparisons (in Fig. 1C,D,E; Fig. 2C; Fig. 3B; Fig. 4A,B,C; Fig. 7B). P<0.05 was considered to be significant.
Acknowledgments
We thank Dr K. Ohashi and Dr A. Yasui for their helpful suggestions.
Footnotes
Author contributions
T.N. designed the study, performed experiments, analyzed and discussed the data and wrote the manuscript; M.I. and S.C. performed experiments and analyzed the data; S.-I.K. provided expertise; K.M. designed the study, discussed the data and wrote the manuscript.
Funding
This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan [grant number 21370086 and 24657126 to K.M.]; and the Uehara Memorial Foundation to K.M.
Supplementary material available online at http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.127209/-/DC1
- Accepted July 5, 2013.
- © 2013. Published by The Company of Biologists Ltd