Histone deacetylase 6 (HDAC6) is an essential factor for oocyte maturation and asymmetric division in mice

Tubastatin A (Tub-A), a highly selective histone deacetylase 6 (HDAC6) inhibitor, has been widely used as a cytotoxic anticancer agent, or for the treatment of patients with asthma. However, the potential toxicity of Tub-A on oocyte maturation and asymmetric division is still unclear. Therefore, the present study was designed to examine the effect and potential regulatory role of Tub-A on the meiotic maturation of oocytes. We observed that Tub-A treatment induced an increased level of the acetylation of α-tubulin, and a failure of spindle migration and actin cap formation. Based on the spindle structure, most Tub-A treated oocytes were arrested in an MI-like or a GVBD-like stage and exhibited decondensed chromosomes in a dose dependent manner. Moreover, Tub-A treatment decreased the protein expression of mTOR, a factor responsible for spindle formation, and the expression of mDia1, an inhibitor of actin assembly, in an HDAC6 expression-dependent manner. Importantly, following Tub-A supplementation, most oocytes failed to extrude the first polar body, which indicates that these defects are closely linked to abnormal oocyte maturation. Taken together, our data demonstrates that HDAC6 is one of the essential factors for oocyte maturation and asymmetric division via the HDAC6/mTOR or mDia1 pathway in mice.

mice successfully undergo oocyte meiosis and asymmetric division. This is very important issue for the successful generation of homozygous offspring.
Tubastatin A (Tub-A) is a potent and highly selective HDAC6 inhibitor 29 . Daily intraperitoneal (i.p.) injection of 25 mg/kg Tub-A into mice for 20 days neither affected brain morphology, brain/body weight mass, liver enzyme measurements, nor kidney function 30 . In vitro, even though Tub-A has no significant side effects on normal cells, a previous study demonstrated that the selective inhibition of HDAC6 can promote hyperacetylation of α-tubulin and decrease cell motility 10,11 . This area of study is important because histone deacetylase inhibitors can be used as a treatment for airway remodelling in patients with asthma 10 . However, little is known regarding the potential toxicity of Tub-A on oocyte maturation and asymmetric division in the context of animal studies. Therefore, this study was aimed to investigate the influence of HDAC6 on oocyte meiotic maturation and asymmetric division following treatment with Tub-A.

Results
Tubastatin A blocks polar body extrusion in mouse oocytes. We first examined the effects of tubastatin A (Tub-A) on mouse oocyte maturation and asymmetric division, as visualised by light and confocal microscopy. To examine the functional roles of HDAC6 during oocyte meiotic maturation, we treated germinal vesicle (GV) oocytes with five different Tub-A concentrations for 12 h (Fig. 1a). Most control oocytes extruded the polar body and developed to the MII stage (Fig. 1b, left), whereas, as shown in the right panel of Fig. 1b and in Fig. 1c, treatment with 20 μM Tub-A resulted in the failure of polar body extrusion. As shown in Fig. 1c and Supplementary Figure 1a, the ratio of polar body extrusion in control oocytes was 72.57 ± 3.74% (n = 476). When oocytes were treated with Tub-A at a concentration of 1 µM, 5 µM, 10 µM, 15 µM, and 20 µM, the ratio dramatically decreased from 65.03 ± 4.86% (n = 165; p = 0.011193) to 57.28 ± 10.02% (n = 177; p = 0.083494), 46.37 ± 3.41% (n = 143; p = 0.001941), 13.57 ± 5.95% (n = 162; p = 0.004428), and 1.17 ± 0.76% (n = 447; p = 0.000547), respectively. Therefore, our observations suggest that Tub-A treatment reduces polar body extrusion in a dose-dependent manner.
To identify the underlying mechanism of Tub-A function, 20 μM Tub-A-treated oocytes were subjected to further study. Of note, Western blotting analysis using anti-HDAC6 antibody showed that Tub-A treated oocytes for 12 h significantly decreased the expression of HDAC6 protein (Fig. 1d), indicating that HDAC6 directly target HDAC6 gene itself to regulate its expression 31 . To further define the cellular events in which HDAC6 is involved during the meiotic maturation, we examined its localisation at different stages of the mouse oocyte maturation by indirect immunofluorescence microscopy (Fig. 1e). In control oocytes, HDAC6 expression in GV-stage oocytes uniformly resided in the cytoplasm of oocytes. At the GVBD stage, the HDAC6 staining patterns started to migrate in the nuclear area respect to the surrounding the chromosomes. As the oocytes entered metaphase, HDAC6 staining is gradually increased along with the spindle region. During the anaphase and telophase stages, intense fluorescence signals of HDAC6 are detected into a spindle like pattern and then move to the cortex region. However, localization of HDAC6 expression after Tub-A treatment was mainly limited to nucleus including weakly and sparsely spot of cytoplasm at GVBD stage and spindle at MI and ATI (Fig. 1e). Likewise, the Tub-A treated group showed very strong protein aggregation (PA) staining, indicating that Tub-A treatment may result in cell death due to the significant accumulation of PA (Fig. 1f).
Tub-A causes the failure of cytokinesis in oocyte meiosis. In mammalian oocytes, spindle migration is driven by actin [32][33][34] . Using confocal microscopy, we examined the spindle and actin morphologies in oocytes after 12 h of in vitro maturation. In the control group, a small polar body and a large MII oocyte had been formed. Most oocytes in metaphase II presented typical barrel-shaped spindles, which were located under the region of the cortex where the actin cap had been formed (Fig. 2a). In contrast, in the Tub-A treated group, spindle defects were readily observed at high frequency (Fig. 2b), and were characterised by MI-like stage (Type I) and GVBD-like stage spindles (Type II) (Fig. 2a,c). Compared with the control oocytes (15.7 ± 3.93% and 11.73 ± 1.99%, respectively), the rates of Type I and Type II in Tub-A treated oocytes were significantly increased (46.82 ± 1.79% and 52.02 ± 1.45%; p = 0.003181 and 0.001038, respectively). In most of oocytes, the spindle failed to migrate to the cortex after treatment with Tub-A (Fig. 2d). In contrast to control oocytes, in which condensed metaphase chromosomes aligned along the equatorial plane, exposure to Tub-A induced the formation of elongated chromosomes (Fig. 2a,e).
As shown in Fig. 2f, actin staining using phalloidin showed that control oocytes could form an actin cap. However, Tub-A treatment caused an apparent alteration in the arrangement of the actin cytoskeleton. Immunofluorescence analysis showed that the actin signals in the Tub-A treated oocytes displayed different patterns when compared with control oocytes (Fig. 2g). These results indicate that spindle migration and actin cap formation were disrupted after treatment with Tub-A.

Tub-A increased the α-tubulin acetylation level in oocytes.
It has been reported that Tub-A treatment leads to an increase in acetylated α-tubulin in vitro 29 . α-tubulin acetylation serves as a marker for the presence of stable microtubules, and may affect the activity of microtubule-associated proteins and microtubule-based motors [35][36][37][38] . It is possible to speculate that HDAC6 regulates spindle function through the direct deacetylation of tubulin. To test this, we examined the effects of Tub-A on tubulin acetylation by staining oocytes with an antibody against acetylated-tubulin. As expected, we found that the acetylation levels of α-tubulin were significantly increased in the Tub-A treated oocytes compared with controls (p = 0.002) (Fig. 3). As shown in Fig. 3a, the abnormal bipolar spindle in Type I oocytes and the single round-shaped spindle in Type II oocytes showed very strong α-tubulin acetylation staining patterns, whereas the acetylated α-tubulin in control oocytes showed very weak staining around the bipolar spindle. These results indicate that the abnormal spindle morphology in Tub-A treated oocytes is closely associated with a high level of acetylated α-tubulin.
HDAC6 functions in several pathways during mouse oocyte maturation. To obtain further insights into the cellular pathways affected by the down-regulation of HDAC6 expression during the meiosis of mouse oocytes, we examined the expression levels of key regulatory factors involved in oocyte meiosis and asymmetric division. As shown in Fig. 1c and Supplementary Figure 1a, the effect of Tub-A on maturating oocytes is clearly dose-dependent. At 20 μM dose, oocyte maturation is completely arrested. RT-qPCR revealed that the mRNA expression levels for mTOR and mDia1 following treatment with Tub-A were significantly decreased. Western blot analysis also confirmed that the mTOR and mDia1 protein expression levels were significantly reduced after Tub-A treatment (Fig. 4a,b). Of note, key genes and proteins expression involved in actin assembly and spindle formation, such as Arp2/3 and RhoA, were significantly down-regulated following Tub-A treatment ( Fig. 4c and d, respectively). These results may be caused by signaling pathway involved in downregulation of HDAC6 because HDAC6 could directly interact with HDAC6 genes and as a result, downregulate the expression of HDAC6 protein. However, the protein expression levels of PI3 kinase, p-AKT, and AKT were not substantially different between the Tub-A treated and the control group ( Supplementary Fig. 1b,c). Also, the data showed that Tub-A treatment for 12 h did not block the extracellular signal, which regulates Kinase-1 and -2 (ERK1/2) phosphorylation and/or activation (Supplementary Fig. 1d). Therefore, we propose a key underlying mechanism for explaining the failure of meiosis in Tub-A treated oocytes (Fig. 4d). Taken together, our results suggest that HDAC6 might be essential for the regulation of actin assembly and spindle formation during oocyte meiosis via mTOR and mDia1 pathway.

Discussion
This study demonstrated that the selective HDAC6 inhibitor, Tub-A, disrupted the spindle migration, actin cap formation, and asymmetric division during oocyte maturation. Moreover, we first demonstrated that HDAC6 expression is an essential factor for mouse oocyte maturation and provided direct evidence that HDAC6 is critically involved in the asymmetric division of the oocyte.
Previous study demonstrated that the endogenous HDAC6 expression in murine somatic cells such as FM3A, MEL, B16 cells, or Balb/c3T3 fibroblasts is mainly located in the cytoplasm 39 . However; its localisation has remained unclear during oocyte maturation. In the current study, our data showed that HDAC6 expression at the GV stage of control oocytes is mainly located in the cytoplasm. Our observation is line with the previous study found that HDAC6 protein expression is localised to the cytoplasm of the germinal vesicle 40 . At the GVBD stage, however, the HDAC6 protein formed aggregates and surrounded each chromosome. From the MI stage onwards, the HDAC6 protein expression was located to the spindle structure (Fig. 1e). Therefore, we hypothesised that HDAC6 is a critical factor for spindle formation during oocyte meiosis and the asymmetric oocyte division. To test our hypothesis, Tub-A, a potent and highly selective HDAC6 inhibitor 29 , were supplemented into culture medium for culturing of GV stage oocytes and oocyte maturation was counted at 12 h after Tub-A supplementation. As expected, the majority of Tub-A treated oocytes lost or showed a decrease in HDAC6 activity. Furthermore, these oocytes failed to extrude the oocyte polar body. Of note, even though we thoroughly washed Tub-A-treated oocytes with fresh medium to remove Tub-A, most oocytes pretreated with Tub-A for 12 h were not able to develop to the MII stage. Instead, the majority of oocytes pretreated with TubA were arrested at an MI-like or GVBD-like stage. Taken together, our observations strongly indicate that HDAC6 is a critical factor for spindle formation during oocyte meiosis and the asymmetric oocyte division.
To examine the quality of MI-like stage or GVBD-like stage oocytes, we examined the aggregates of ubiquitinated proteins (aggresomes) in these oocytes using a ProteoStat Aggresome Detection Kit. As shown in Fig. 1f, most Tub-A treated oocytes led to very strong aggresome staining patterns, whereas control oocytes showed very weak signals. Generally, protein aggregates do not accumulate in normal cells despite their continued production, because of the existence of a cellular 'quality control' machinery 15,41 . As described in a previous review, the bulk of protein aggregates in intracellular and extracellular lesions are closely associated with cell death in many degenerative diseases 41,42 . Lee et al. reported that HDAC6 promotes the fusion of autophagosomes and lysosomes 14 . In this study, Tub-A treated oocytes significantly decreased HDAC6 expression. Consistent with a previous finding 43 , Tub-A treated oocytes significantly reduced HDAC6 protein expression.
It is conceivable to speculate that control oocytes may remove the bulk of accumulated aggresomes through an autophagic route to protect the oocytes, whereas treatment with Tub-A triggers a significant accumulation of protein aggregates, indicating that HDAC6 inhibition in oocytes is irreversible. Taken together, we conclude that the accumulation of aggresomes in Tub-A treated oocytes might interfere with essential functions in oocyte meiosis, such as cytoskeletal organisation and the asymmetric oocyte division.
To determine the reason behind polar body extrusion defects following HDAC6 inhibition, we examined the actin filament distribution, which plays essential roles in oocyte polarity formation and cytokinesis 44 . Moreover, it is well known that the Arp2/3 complex 45 and mDia1 46 are involved in actin organisation during oocyte maturation in mice. The results of this study showed that the mDia1 mRNA and protein expression levels were significantly decreased, compared to those in the control (Fig. 4b). Further, following the inhibition of HDAC6 after Tub-A treatment, actin failed to form the cortical actin cap (Fig. 2d). Destaing et al. 47 reported that Rho interferes with the osteoclast maturation process by controlling the level of microtubule acetylation and actin organisation through mDIA2 and HDAC6. Therefore, we tested and confirmed that RhoA was also inhibited by Tub-A (Fig. 4d). In conclusion, we suggest that mDia1 and RhoA dysregulation via the inhibition of HDAC6 may represent a possible pathway underlying the actin defects during oocyte maturation.
Previously, Lee et al. 48 reported that the failure to form the actin cap disrupt spindle migration and lead to an abnormal asymmetric division during the meiotic maturation of oocytes, which was caused by a low expression of mTOR. In this study, we also detected a low level of mTOR expression in Tub-A treated oocytes. Previous study reported that HDAC6 protects neurons from toxicity of prion peptide, and that this protection occurs at through the regulation of the PI3k-Akt-mTOR axis 49 . However, the inhibition of HDAC6 during oocyte maturation did not alter PI3K and AKT protein expression ( Supplementary Figure 1b and c). Of note, our results indicate that Tub-A treatment also decreased mTOR expression at both mRNA and protein levels by HDAC6 downregulation. The impact of HDAC6 on mTOR signaling could be linked to targeting of HDAC6 gene itself or its effect on α-tubulin acetylation. Taken together, these results suggest that dichotomous effects of HDAC6 on the HDAC6/ mTOR or HDAC6/mDia1-mediated signaling pathways might inhibit spindle migration and asymmetric division of oocytes. Further, we observed a striking reduction in oocyte numbers with asymmetric division upon 12 h of HDAC6 inhibition that correlated well with the increase in α-tubulin acetylation. In conclusion, we believe that HDAC6 is an essential factor for cytokinesis and chromosome condensation in normal mouse oocytes. Thus, this study can provide important information for the development of safe and non-toxic HDAC6 inhibitors for animals and human beings. Western blotting. A total of 15-50 oocytes were lysed in RIPA buffer (GenDOPET, Texas, USA) containing protease inhibitors and heated for 5 min at 100 °C. Total oocyte proteins were subjected to electrophoresis on a 10% SDS-PAGE gel. The separated proteins were transferred to PVDF membranes, which were pretreated with methanol. The membranes were blocked in 5% skim milk and incubated with primary antibodies as follows: HDAC6 (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), Acetyl-α-tubulin (Thermo Fisher Scientifc, Rockford, IL, USA), mDia1 (BD Biosciences, San Jose, CA), and mTOR (2983), PI3 Kinase p110α (4249), Phospho-Akt (4060), Akt (4685), Phospho-p44/42 (9101) and p44/42 (9102), all purchased from Cell Signaling Technology (Beverly, MA, USA). After three washes in TBST, the blots were then incubated with anti-rabbit or anti-mouse IgG antibody conjugated to horseradish peroxidase for 1 h. The protein bands were visualised using an SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientifc, Rockford, IL, USA). The membrane was then washed and reblotted with an actin antibody as an internal control. Densitometric quantification was performed using the ImageJ software (NIH, Bethesda, Maryland).

Methods
Immunostaining and confocal imaging. Ten oocytes were fixed in 4% paraformaldehyde for 30 min and permeabilised for 30 min with PBS containing 0.1% Triton X-100. Permeabilised oocytes were blocked for 1 h at room temperature in 1% bovine serum albumin (BSA) and 0.1% Triton X-100 in PBS before overnight incubation at 4 °C with the primary antibodies for anti-α-tubulin (Cell Signaling Technology, Beverly, MA, USA), anti-HDAC6 and anti-acetyl-α-tubulin. The oocytes were washed several times in 0.05% Tween 20 in PBS (PBST), transferred to a secondary antibody mixture of Alexa Fluor 568 goat anti-mouse and Alexa Fluor 488 goat anti rabbit (Molecular Probes, USA), and incubated at room temperature for 30 min. Aggregates of ubiquitinated proteins (aggresomes) were detected in oocytes after treatment with Tub A using a ProteoStat Aggresome Detection Kit (Enzo Life Sciences, Inc., USA), and confocal images using the TO-PRO-3 fluorescent dye were acquired using an Olympus FV1000 Confocal microscope (Tokyo, Japan), and were processed using the FV10-ASW 2.0 Viewer software (Olympus, Tokyo, Japan). Fluorescent images were acquired using an Olympus BX-UCB microscope and were processed using a DP controller software (Olympus, Tokyo, Japan). The