Loss of HDAC3 contributes to meiotic defects in aged oocytes

Abstract Maternal age‐related decline in oocyte quality is associated with meiotic defects, but the underlying mechanisms remain to be explored. Histone deacetylase 3 (HDAC3) has been shown to govern multiple cellular events via deacetylating diverse substrates. We previously found that HDAC3 could promote meiotic apparatus assembly in mouse oocytes. In the present study, we identified a substantial reduction in HDAC3 protein in oocytes from old mice. Importantly, overexpression of HDAC3 in old oocytes not only partially prevents spindle/chromosome disorganization, but also significantly lowers the incidence of aneuploidy. Meanwhile, we noticed the elevated acetylation level of α‐tubulin in oocytes derived from old mice. By employing site‐directed mutagenesis, we showed that acetylation‐mimetic mutant tubulin‐K40Q disrupts the kinetochore–microtubule attachments and results in the assembly failure of meiotic apparatus in mouse oocytes. Importantly, forced expression of tubulin‐K40R (nonacetylatable‐mimetic mutant) was capable of alleviating the defective phenotypes of oocytes from aged mice. To sum up, this study uncovers that loss of HDAC3 represents one potential mechanism mediating the effects of advanced maternal age on oocyte quality.


| INTRODUC TI ON
Advanced maternal age in mammals is associated with reduced fertility. Reproductive capacity in women declines dramatically beyond the mid-30s (Hamatani et al., 2004;Hassold & Hunt, 2001). In mice, there is an increased frequency of aneuploidy embryos and fetuses with maternal age (Golbus, 1981;Pan, Ma, Zhu, & Schultz, 2008).
Emerging evidence suggests that poor oocyte quality is a critical factor mediating the effects of aging on female fertility (Keefe, Kumar, & Kalmbach, 2015;Volarcik et al., 1998). In mammals, the oocytes enter meiosis during the fetal period and their development arrests in the dictyate of prophase until they resume maturation just prior to ovulation at the reproductive age (Kurahashi et al., 2012). It is worth noting that this meiotic process is error prone. Oocytes with the wrong number of chromosomes give rise to aneuploid embryos following fertilization, which in humans is a major cause of pregnancy loss and developmental disabilities (Hassold & Hunt, 2001). Recently, it has been demonstrated that aged oocytes exhibit substantially altered spindle microtubule dynamics, resulting in kinetochore-microtubule attachment defects and chromosome segregation errors (Holubcova, Blayney, Elder, & Schuh, 2015;Nakagawa & FitzHarris, 2017).
Mammalian genomes encode 11 proteins of the classical histone deacetylase family (HDAC1-11) (Yang & Seto, 2008). These proteins are grouped into class I, II, and IV, with class II being further divided into two subclasses (IIa and IIb), based on their homology to yeast Rpd3/Hda1. HDACs play roles in numerous biological process largely through their influence on transcription by modulating the acetylate state of histones or transcription factors (Haberland, Montgomery, & Olson, 2009). HDAC3, as a member of class I, was found to function in stem cell self-renewal in a transcription-independent manner (Li et al., 2006). Accumulated evidence demonstrated that HDAC3 is required for normal mitotic progression through the interaction with Aurora B kinase (Eot-Houllier, Fulcrand, Watanabe, Magnaghi-Jaulin, & Jaulin, 2008;Fadri-Moskwik et al., 2012;Li et al., 2006).
Intriguingly, we recently reported that HDAC3 depletion adversely influences meiotic maturation, especially spindle assembly and chromosome organization (Li et al., 2017), similar to the phenotypes of aged oocytes. HDAC3 has been shown to modulate the acetylation state of different substrates in multiple cell lines. For example, in an in vitro reconstituted chromatin system, an HDAC3-containing protein complex selectively deacetylated histone H3 (Vermeulen et al., 2004). Blocking HDAC3 activity could dramatically alter the tubulin acetylation in the human prostate cancer cells (Bacon et al., 2015). In the present study, we identify a substantial reduction of HDAC3 protein in oocytes from old mice. Notably, overexpression of HDAC3 is capable of alleviating the meiotic defects in oocytes from old mice. These findings indicate that HDAC3 insufficiency in the oocyte may represent a connection between oocyte quality and reproductive aging.

| Reduced HDAC3 expression in oocytes from old mice
Given that the oocytes depleted of HDAC3 displayed the similar phenotypes as those from aged mice (Li et al., 2017), we decided to determine the HDAC3 expression in oocytes from young (3 weeks of age) and old (42 weeks of age) mice by Western blot analysis.
As shown in Figure 1a- immunostaining coupled with confocal microscopy. As shown in Figure 1c-d, there was no apparent alteration in the distribution pattern of HDAC3 between young and old oocytes. However, the average intensity of HDAC3 signal was significantly decreased in old oocytes relative to young cells. In addition, we found that the levels of HDAC3 protein in different stages of oocytes are comparable when cultured in vitro (Figure 1e), indicating the absence of fluctuation in HDAC3 expression during oocyte maturation.

| HDAC3 overexpression alleviates the meiotic defects in old oocytes
To Error bars indicate ± SD. *p < .05 versus controls the results indicate that loss of HDAC3 is one of potential pathways mediating the effects of maternal aging on oocyte quality.

| Elevated acetylation levels of tubulin in old oocytes
Tubulin is one of the most abundant nonhistone proteins that is subjected to acetylation, which occurs on lysine (K) 40 of α-tubulin subunit (Zilberman et al., 2009). We previously demonstrated that HDAC3 modulates spindle/chromosome organization in oocytes by maintaining the hypoacetylation state of tubulin (Li et al., 2017).
Herein, the reduced HDAC3 expression in old oocytes prompted us to ask whether the acetylation status of tubulin was altered accordingly. To address this question, MII oocytes were isolated from young and old mice. By performing Western blot and immunostaining analysis, we found that the acetylation level of tubulin-K40 was markedly increased in old oocytes as compared to young oocytes (Figure 3a-d). This observation implies that tubulin acetylation may be involved in the deficient meiotic apparatus in oocytes from old mice.

| Acetylation of tubulin-K40 functions in kinetochore-microtubule interaction during oocyte meiosis
To test whether tubulin acetylation affects meiotic structure in oocytes, we constructed the site-specific mutants (K-to-Q and K-to-R) targeting lysine 40 (K40) of tubulin, and then the mRNA encoding the tubulin mutants was injected into fully grown oocytes for analysis.
As shown in Figure 4a-b, we found that acetylation-mimetic mutant K40Q resulted in almost threefold increase in spindle/chromosome defects compared to WT control, whereas K40R had little effects on meiotic apparatus.
Coordination between spindle maintenance and chromosome movement largely relies on kinetochore-microtubule (K-MT) interaction (Prosser & Pelletier, 2017). Hence, we further examined the effects of tubulin acetylation on K-MT attachment during meiosis. For this purpose, metaphase oocytes were immunolabeled with CREST to detect kinetochores and stained with anti-α-tubulin-FITC-conjugated antibody to visualize microtubules. As shown in  that HDAC3 localizes to the plasma membrane and forms a complex with c-Src to regulate its activity in HFK + 31 cells (Longworth & Laimins, 2006). In addition, HDAC3 was detected over the spindle in HeLa cells and mouse 3T3 fibroblasts (Ishii et al., 2008). Similarly, we found HDAC3 resides in the cytoplasm of mouse oocytes, with accumulated signals on the meiotic spindle region, promoting the assembly of meiotic apparatus (Li et al., 2017).
It has been well documented that oocytes recovered from old females show a series of changes in multiple nuclear and cytoplasmic structures (Miao, Kikuchi, Sun, & Schatten, 2009) (Nakagawa & FitzHarris, 2017).
However, the underlying molecular details remain to be elucidated.
In the present study, we identified a reduction of HDAC3 protein in oocytes from old mice (Figure 1). Importantly, we noted that overexpression of HDAC3 in old oocytes could partially rescue the spindle/ chromosomes defects and lower the incidence of aneuploidy eggs ( Figure 2). Therefore, these findings support the conclusion that loss of HDAC3 is an important factor contributing to the compromised oocyte quality of aged mice, specifically the failure to assemble the meiotic apparatus.
Microtubules are subject to a remarkable number of post-translational modifications. Thirty years ago, the conserved ε-amino group of lysine 40 of the N-terminal domain of the α-tubulin was identified as the site of acetylation (LeDizet & Piperno, 1987). The association of acetylation with long-lived microtubules indicated that stable microtubules tend to be acetylated (Schatten et al., 1988). Reed et al. further showed that loss of α-tubulin acetylation influences the binding and motility of kinesin-1 in vitro (Reed et al., 2006). Like other post-translational modifications of tubulin, acetylation on lysine 40 is under the control of balanced enzyme activities (acetyltransferases and deacetylases). αTAT1 was identified to be the major α-tubulin acetyltransferase (Kalebic et al., 2013;Shida, Cueva, Xu, Goodman, & Nachury, 2010). Recently, Bacon et al.
found that blocking HDAC3 activity modulates tubulin acetylation in the human prostate cancer line PC3 (Bacon et al., 2015). Likewise, we previously showed that the acetylation levels of α-tubulin were dramatically increased in mouse oocytes depleted of HDAC3 (Li et al., 2017). In line with this observation, here we found the elevated acetylation of tubulin in old oocytes with lowered HDAC3 expression ( Figure 3). Moreover, our data showed that nonacetylated tubulin-K40R is capable of partly rescuing the deficient phenotypes of aged oocytes ( Figure 5). These results suggest that HDAC3-dependent tubulin deacetylation functions as a protective mechanism in oocytes, which is aberrantly blunted in the aged mouse model.
Microtubule stability is regulated in part by a heterogeneous family of proteins that bind to tubulin subunits of microtubules. Related to the functions of microtubules, cyclin B1 interacts with microtubuleassociated proteins (MAPs) to regulate cell cycle progression (Ookata, Hisanaga, Okano, Tachibana, & Kishimoto, 1992). On the other hand, the gradual increase in cyclin B1 levels and CDK1 activity acts as a timing mechanism to modulate the formation of stable K-MT interaction in oocytes (Davydenko, Schultz, & Lampson, 2013). Herein, we found that the acetylation-mimetic mutant tubulin-K40Q induces erroneous K-MT attachments in oocytes ( Figure 5). In combination with the differential meiotic progression between young and old oocytes (Eichenlaub-Ritter & Boll, 1989), we propose that advanced maternal age induces the loss of HDAC3 in oocytes, likely through elevating tubulin acetylation, results in kinetochore-microtubule misattachments, thereby contributing to the meiotic defects and aneuploidy generation ( Figure 5g). In addition, SIRT2 reduction has recently been reported to contribute to oocyte aging by influencing acetylation status of histone H4K16 (Zhang et al., 2014). Therefore, the present study cannot rule out that other pathways might be regulated by HDAC3 to affect meiotic activity in oocytes.

| MATERIAL S AND ME THODS
All chemicals and culture media were obtained from Sigma unless otherwise specified.

| Animals
42-week-old female ICR mice were used as a model of reproductive aging, 3-week-old female mice were used as a control. All animal protocols were approved by the Animal Care and Use Committee of Nanjing Agricultural University, and all experiments were conducted in accordance with the guidelines of the local animal ethical committee and the Animal Care and Use Committee of Nanjing Agricultural University.

| Antibodies
Rabbit polyclonal anti-HDAC3 antibodies were purchased from Santa

| cRNA synthesis and overexpression experiments
Total RNA was extracted from 50 denuded oocytes using the Arcturus PicoPure RNA Isolation Kit (Applied Biosystems), and cDNA was generated using Quantitect Reverse Transcription Kit (Qiagen). The related primers are listed in Table S1. Plasmid construction and mRNA synthesis were conducted as we reported previously (Gao et al., 2018). In brief, PCR products were cloned into the pCS2 + vector with Myc tags, and cRNAs were made using in vitro transcription with SP6 mMESSAGE mMACHINE (Ambion) according to the manufacturer's instruction. Tubulin mutants were generated by site-directed mutagenesis according to the published protocol .
The same amount of RNase-free PBS was injected as a control. After injection, oocytes were cultured in medium containing 2.5 µM milrinone for 20 hr to ensure the translation of cRNA. Following multiple washes, oocytes were cultured in milrinone-free M16 medium for further experiments.

| Immunofluorescence
Oocytes were fixed in 4% paraformaldehyde for 30 min and permeabilized with 0.5% Triton X-100 for 20 min at room temperature.
After blocking in 1% BSA in PBS for 1 hr, oocytes were incubated with primary antibodies overnight at 4°C. Following three washes, oocytes were labeled with secondary FITC-or Cy5-conjugated antibody for 1 hr at room temperature. Samples were mounted on glass slides in a drop of antifade medium (Vectashield) and then examined under a laser scanning confocal microscope (LSM 710). Fluorescent intensity was measured by placing a small circle around the signals using ImageJ software (U.S. National Institutes of Health).
The average cytoplasmic fluorescence intensity was subtracted as background.

| Chromosome spread
Chromosome spreading was conducted as we previously described (Li et al., 2017). Zona pellucidae was removed prior to fixation by brief exposure to the Tyrode buffer (pH 2.5). Oocytes were washed in M2 medium and then fixed in a drop of 1% paraformaldehyde with 0.15% Triton X-100 on glass slide. After air drying, a standard immunofluorescent staining procedure was performed as mentioned above. Kinetochores were labeled with CREST and chromosomes were stained with Hoechst 33,342. System. β-actin was used as a loading control.

| Statistical analysis
Experiments were performed at least in triplicate. Data are presented as means ± SD, unless otherwise stated. Differences between two groups were analyzed by Student's t test. Multiple comparisons between more than two groups were analyzed by one-way ANOVA test using Prism 5.0. p < .05 was considered to be significant.

ACK N OWLED G M ENTS
This work was supported by National Natural Science Foundation of China (NO. 31970789 and NO. 31771660).

S U PP O RTI N G I N FO R M ATI O N
Additional supporting information may be found online in the Supporting Information section at the end of the article.