The E3 Ubiquitin Ligase SYVN1 Plays an Antiapoptotic Role in Polycystic Ovary Syndrome by Regulating Mitochondrial Fission

Polycystic ovary syndrome (PCOS) is one of the most common hormonal disorders among premenopausal women. PCOS is accompanied by many other reproductive, endocrinal, and metabolic disorders thus amassing the difficulties encountered by the women affected. However, there is limited information on its molecular etiology. Synoviolin (SYVN1) is an E3 ubiquitin ligase that is thought to participate in the pathology of PCOS. However, the expression and function of SYVN1 in PCOS are unknown. In this study, we found that downregulation of SYVN1 expression was followed by increased apoptosis in the granulosa cells (GCs) of patients with PCOS. Subsequent in vitro experiments indicated that the overexpression of SYVN1 inhibited apoptosis and mitochondrial fission. Furthermore, using immunoprecipitation and western blotting, we identified that SYVN1 promoted the degradation of Drp1 via the proteasome-dependent pathway. Additionally, we generated a PCOS model in female Sprague Dawley rats and treated them with an SYVN1 inhibitor, LS-102. We observed that the inhibition of SYVN1 increased Drp1 levels and exacerbated the degeneration of GCs in the PCOS rat model. Finally, in vitro and in vivo experiments showed that SYVN1 inhibits apoptosis and mitochondrial fission by promoting Drp1 degradation in GCs. These results highlight the function of SYVN1 in PCOS and provide a potential target for the clinical treatment of PCOS.


Introduction
Polycystic ovary syndrome (PCOS) is a common disorder in women with a prevalence of 8 to 13% among those of premenopausal age [1]. PCOS can have long-term consequences such as an increased risk for cardiovascular, metabolic, and other physiological conditions that may result in a lifetime of complications [2]. PCOS also contributes to around 75% of anovulatory infertility and is involved in impaired follicular maturation, anovulation, and biochemical pregnancy [3]. Therefore, the complex mecha-nisms behind PCOS need further characterization and understanding. For a cellular system to function efficiently, the proper regulation of mitochondrial dynamics is critical. Mitochondria are characterized as semiautonomous organelles that constantly undergo fusion and fission [4]. Dysregulation of fusion and fission processes could lead to many complications such as neurodegenerative, cardiovascular, and renal diseases [5,6]. It has also been widely accepted that mitochondrial complications may contribute to other diseases, especially in situations where the molecular etiology of the  2 Oxidative Medicine and Cellular Longevity disease has not been well explored. Recently, studies have implicated that a reduction in mitochondrial mass and its subsequent dysfunction contributes to the development of PCOS [7]. Granulosa cells (GCs) are the principal cells in the ovary that provide the support and environment for the developing oocyte. These cells further associate with female gametes and continue their interaction until primary and secondary oocyte maturation. Cumulus GCs transport the oocyte to the oviduct, and the mural GCs remain in the ovary to form luteal cells. These luteal cells are essential for implantation and zygote development [8]. Hence, GCs are necessary for appropriate follicular development, ovulation, steroidogenesis, and atresia [9,10]. Interestingly, previous studies have indicated that the apoptosis of GCs contributes to abnormal folliculogenesis, anovulation, and other complications in PCOS [11]. However, it is important to note that one important source of this apoptosis in GCs is the dysregulation of mitochondrial fission. Specifically, a study on PCOS in an androgenized rat model displayed an upregulation of the key mitochondrial fission marker, dynamin-related protein 1 (Drp1) [12]. Furthermore, this upregulation was associated with increased mitochondrial fission that contributed to apoptosis and autophagy in the GCs at the antral stage of development in the PCOS model. However, this response could be partially rescued in rats with the use of exogenous gonadotropin [12].
Protein regulation and recycling is an important process carried out by the ubiquitination-proteasomal pathway. Previously, a study using protein-protein interaction network analysis showed that synoviolin (SYVN1) may be associated with PCOS [13]. SYVN1 is an E3-ubiquitin ligase protein predominantly studied in the endoplasmic reticulum-(ER-) associated degradation of cells and accumulated misfolded proteins [14]. A study by Ge et al. observed that upregulated SYVN1 could significantly inhibit apoptosis in osteoarthritis [15]. However, the role that SYVN1 plays in PCOS is not well elucidated.
In this study, we initially observed that SYVN1 is highly downregulated in the GCs of PCOS patients. Additionally, using a human ovarian granulosa-like tumor (KGN) cell line, we also confirmed that downregulated SYVN1 is associated with increased apoptosis and increased mitochondrial fission. Using both in vitro and in vivo rat models, we have found that SYVN1 modulates apoptosis in GCs through the degradation of Drp1 and the downregulation of mitochondrial fragmentation and apoptosis. This is the first study, to our knowledge, that clearly illustrates the mechanistic role of SYVN1 in PCOS. Additionally, our study provides valid therapeutic targets for the development of potential treatment strategies for PCOS.

Methods and Materials
2.1. Patients. All samples for this study were collected from women who were referred to the Reproductive Medicine Center, Shanghai East Hospital. Informed consent was obtained from each study participant. Follicular fluid was collected from women (n = 60) either undergoing in vitro fertilization (IVF) or intracytoplasmic sperm injection. Participants were divided into control (n = 30) and PCOS (n = 30) groups. PCOS patients were enrolled based on the Rotterdam criteria as mentioned in a previous study [16]. In the PCOS group, patients exhibited two of three (control: patients undergoing fertility treatment due to male infertility or tubal infertility disorders; PCOS, patients with polycystic ovary syndrome). The data are expressed as mean ± standard deviation. (n = 4) * P < 0:05, * * P < 0:01, and * * * P < 0:001. Two-tailed unpaired Student's t-test. Cleaved-caspase-3: P < 0:0001, t = 16:64, and df = 6; Bax: P = 0:0011, t = 5:836, and df = 6; Bcl-2: P = 0:0100, t = 3:705, and df = 6.   Oxidative Medicine and Cellular Longevity characteristic pathologies: (a) oligomenorrhea or anovulation, (b) clinical or biological hyperandrogenism, or (c) polycystic ovary. The control group comprised women who displayed typical menstrual cycles and were undergoing IVF not due to the abovementioned reasons but due to male infertility or tubal infertility disorders. The study was performed in accordance with the Declaration of Helsinki, and ethical clearance was obtained from the ethics committee of the Shanghai East Hospital.

Oxidative Medicine and Cellular Longevity
(DHEA, 6 mg per 100 g body weight in 0.2 ml sesame oil) every day for a total of 21 days to induce hyperandrogenic PCOS [17,18]. The rats were then intraperitoneally injected with 1 ml/kg saline for a subsequent 28 days; (3) the PCOS model-SYVN1 inhibitor (LS-102) group, where the rats were induced with DHEA to have the PCOS phenotype as mentioned in the previous group. Then, the rats were intraperitoneally injected with LS-102 (50 mg/kg) every day for a total of 28 days.
All the animal experiments were approved by the Shanghai East Hospital and were carried out in accordance with the regulations set by the Shanghai East Hospital. Ethical clearance was obtained from the ethics committee of the Shanghai East Hospital.
2.6. Quantitative Real-Time PCR. TRIzol reagent was used to isolate the total RNA from the granulosa cell samples (Invitrogen). Using HiScript reverse transcriptase, the total RNA was used in reverse transcription (RT) PCR to obtain cDNA (RNase H; Genecopoeia, Rockville, MD, USA). Finally, relative SYVN1 gene expression was assessed using the step one plus real-time PCR system (Applied Biosystem, Waltham, MA, USA). The primer sequences used are listed as follows: forward, 5 ′ -CTCCTCCATCCACCAGTGC-3 ′ and reverse,

11
Oxidative Medicine and Cellular Longevity antibody (1 : 500, Abcam) for 1 h. Finally, slides were mounted in mounting media (Vector Laboratories), and images were captured with a fluorescence microscope (Olympus, Tokyo, Japan).

TUNEL Assay. Human granulosa cells and KGN cells
were cultured on glass coverslips and fixed using 4% PFA. Apoptotic cell analysis was performed using the terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) staining kit (Roche, Mannheim, Germany) according to the manufacturer's instructions. DAPI was used to counterstain the nuclei. Imaging of the TUNEL-positive cells, which have a pyknotic nucleus with dark red fluorescent staining, was performed using an inverted microscope.  2.13. Histological Assay and Follicle Count. Rat ovaries and vital organs from respective groups were initially fixed in 10% neutral-buffered formalin solution for 24 h at 4°C. The organs were dehydrated, embedded in paraffin, and sectioned serially at 5 μm thickness. The sections were carefully mounted onto a frosted glass slide for further staining with hematoxylin and eosin (HE). Imagining of the sections was performed with a bright field microscope. Follicle number was evaluated under a microscope in every tenth HEstained section to create a distance of approximately 50-60 μm between counts. The follicles were categorized as primordial, primary, antral, Graafian, and atretic follicles according to the definitions described by Furat Rencber et al. [19].
2.14. Transmission Electron Microscopy (TEM). Rat ovarian tissues were initially fixed for 2 h with 2.5% glutaraldehyde (Sigma-Aldrich). Tissues were washed three times in 0.1 M PBS, dehydrated in graded alcohol, and stained with OsO 4 and uranyl acetate. The tissues were embedded into Durcupan Fluka resin and polymerized for 48 h at 56°C. Blocks were further sectioned into 55 nm ultrathin-section and collected on Formvar-coated single-slot grids. Sections were subsequently stained by incubating in lead citrate. The cell ultrastructure was observed using a transmission electron microscope (FEI, Hillsboro, OR, USA).

Statistical Analysis.
Statistical analysis was performed using SPSS software version 17.0 (SPSS, Inc., Chicago, IL, USA). Data are presented as the mean ± standard deviation ðSDÞ. For normally distributed data, group differences were assessed using the Student t-test (comparison of two independent groups), one-way analysis of variance (ANOVA) with Tukey's multiple comparison test, or two-way ANOVA with Sidak's multiple comparison test (comparison of several independent groups). P value < 0.05 was considered to indicate a statistically significant difference.

Downregulated SYVN1 Is Potentially Associated with
Massive Apoptosis in Granulosa Cells of Polycystic Ovary Syndrome Patients. Initially, we assessed the mRNA expression levels of SYVN1 in primary GCs from patients with PCOS and identified that SYVN1 levels were significantly downregulated in PCOS patient cells (n = 30) when compared with control samples (Figure 1(a)). We also observed a significant decrease in SYVN1 protein levels in PCOS patient GCs (Figure 1(b)). We further visualized this decrease in PCOS GCs using immunofluorescence staining (Figure 1(c)). We observed an increase in the number of TUNEL-positive apoptotic cells among the PCOS GCs (Figure 1(d)). To further explore the role of apoptosis in PCOS, we checked the levels of key apoptotic markers such as cleaved caspase-3, Bax, and Bcl-2 using western blotting. We found a significant increase in the levels of cleaved caspase-3 and Bax, whereas a clear decrease in the Bcl-2 levels among PCOS GC samples (Figure 1(e)). This evidence indicates the potential involvement of SYVN1 in the apoptosis of GCs in PCOS patients.

Overexpression of SYVN1 Inhibited Apoptosis in KGN
Cells. To understand the role of SYVN1 in the apoptosis of GCs in PCOS, we used the human ovarian granulosalike tumor (KGN) cell line. We overexpressed SYVN1 using expression plasmids, silenced SYVN1 using siRNA in the KGN line, and confirmed its expression levels using RT-qPCR and western blotting (Figures 2(a) and 2(b)). Apoptosis levels were assessed using flow cytometry. SYVN1 overexpression decreased the percentage of apoptotic cells (Empty: 3.83% and SYVN1: 2.37%), whereas silencing of SYVN1 increased the percent of apoptotic cells (siSYVN1: 12.94%) (Figure 2(c)). TUNEL assay indicated that SYVN1 overexpression decreased apoptotic TUNEL + cells whereas silencing SYVN1 increased TUNEL + cells (Figure 2(d)). We also confirmed the levels of apoptotic markers using western blotting. Overexpression of SYVN1 levels significantly decreased cleaved caspase and Bax levels along with an increase in the Bcl-2 levels. Alternatively, silencing of SYVN1 significantly increased cleaved caspase and Bax levels with a decrease in Bcl-2 levels (Figure 2(e)). These results indicate that SYVN1 regulates apoptosis in ovarian GCs.

Overexpression of SYVN1 Inhibited Mitochondrial
Fission in KGN Cells. As mitochondrial fission is the initial event during apoptosis, we wanted to explore the role of SYVN1 in mitochondria. Using MitoTracker, we assessed KGN cells either overexpressing or silenced for SYVN1. Imaging results indicated that the silencing of SYVN1 significantly decreased the mitochondrial levels as indicated through decreased surface area exhibiting a positive signal for MitoTracker. Alternatively, SYVN1 overexpression clearly increased the mitochondrial signal (Figure 3(a)). Overexpression of SYVN1 showed highly interconnected and elongated mitochondrial structures, whereas silencing of SYVN1 resulted in stratified, disconnected, and circular mitochondria. These results indicate that SYVN1 could play a role in apoptosis by inhibiting mitochondrial fission. Next, we assessed the levels of key the mitochondrial fissionassociated protein Drp1 in KGN cells. Initially, it was clear that overexpression of SYVN1 significantly decreased Drp1 levels, whereas silencing of SYVN1 increased Drp1 (Figure 3(b)). We also checked the levels of mitofusin 1 (Mfn1) and observed that overexpression of SYVN1 significantly increased Mfn1 levels, whereas silencing SYVN1 decreased Mfn1 levels (Figure 3(c)). This further confirmed that SYVN1 inhibited mitochondrial fission.

SYVN1 Regulates Ubiquitination and Proteasomal
Degradation of Drp1 in KGN Cells. To understand the role of SYVN1 in mitochondrial fission, we overexpressed SYVN1 and assessed Drp1 levels using western blotting. Our data clearly indicate that the overexpression of SYVN1 caused a significant decrease in Drp1 protein levels (Figure 4(a)). Additionally, silencing of SYVN1 significantly increased Drp1 protein levels (Figure 4(b)). We wanted to assess whether SYVN1 regulates Drp1 protein through ubiquitination. Therefore, we examined whether SYVN1 was able to bind Drp1 in Co-IP experiments. In KGN cells, SYVN1 was able to bind to Drp1 (Figure 4(c)), and the ubiquitination of Drp1 was detected using an ubiquitination antibody in KGN cells overexpressing SYVN1 (Figure 4(d)). To confirm that SYVN1 was able to regulate the protein level of Drp1 through the proteasome pathway, we treated SYVN1 overexpressed cells with the proteasome inhibitor MG132, and the results showed that MG132 could indeed inhibit the downregulation of Drp1 level by SYVN1 (Figure 4(e)). Meanwhile, treatment of KGN cells with CHX revealed that the overexpression of SYVN1 was able to significantly reduce Drp1 stability (Figure 4(f)). These results showed that in KGN cells, the binding of SYVN1 to Drp1 promoted its ubiquitination and degradation.

SYVN1 Inhibited Apoptosis and Mitochondrial Fission by
Promoting Drp1 Degradation in KGN Cells. The role that SYVN1 played in the regulation of apoptosis was characterized by silencing SYVN1 and Drp1. Flow cytometry showed that the silencing of SYVN1 increased apoptosis in KGN cells, whereas the silencing of Drp1 gave the opposite result and decreased apoptosis in KGN cells. In addition, silencing Drp1 reversed apoptosis induced by SYVN1 knockout, increasing the level of apoptosis (Figure 5(a)). We confirmed these results using a TUNEL assay and observed that SYVN1 is necessary for the decrease in Drp1-mediated apoptosis ( Figure 5(b)). Western blotting also indicated that the silencing of SYVN1 caused a significant increase in apoptotic markers (Bax and cleaved caspase-3). However, silencing of Drp1 significantly decreased Bax and cleaved caspase levels. In parallel, we observed a decrease in Bcl-2, when SYVN1 was silenced, and an increase in Bcl-2, when Drp1 was silenced ( Figure 5(c)). We also assessed mitochondrial fission levels using MitoTracker, and it was clear that silencing of SYVN1 increased mitochondrial fission. In contrast, silencing of Drp1 significantly decreased mitochondrial fission as indicated by an increase in the mitochondrial surface area along with highly interconnected and elongated mitochondrial structures. However, the silencing of Drp1 along with SYVN1 reversed this effect and resulted in rarefied and disconnected mitochondrial structures ( Figure 5(d)). These results confirmed that SYVN1 decreased apoptosis and mitochondrial fission by inducing the degradation of Drp1.

SYVN1 Inhibited Apoptosis and Mitochondrial Fission in
Granulosa Cells in Rat with Polycystic Ovary Syndrome. Finally, we assessed the role of SYVN1 in inhibiting apoptosis and mitochondrial fission in a rat model of PCOS. We separated the female SD rats (N = 18) into two groups, a control (N = 6) and PCOS model groups (N = 12). To achieve a PCOS model, we injected rats intraperitoneally with DHEA for 21 days. These rats were again divided into two groups, and one group was assigned as the control and was injected with saline for 28 days (PCOS, N = 6). The other group was injected with LS-102, an SYVN1 inhibitor, for 28 days (LS-102, N = 6). After 28 days, the rats were euthanized, organs were extracted, and total protein was isolated. Using HE staining, we confirmed the proper establishment of PCOS as indicated by the degeneration of oocyte and GC layer, along with the presence of large follicles in the PCOS rat ovaries. We observed that the LS-102 aggravated the PCOS phenotype, with an increased number of large fluid-filled follicles and increased degeneration of oocytes posttreatment (Figure 6(a)). We further investigated the consequences of PCOS on follicle morphology by assessing the different follicles in each ovary (Figure 6(b)). The number of primary, secondary, antral, and Graafian follicles was significantly reduced in the PCOS group, whereas the number of atretic and cystic follicles was significantly increased. Compared to the PCOS group, the number of primordial, primary, secondary, antral, and Graafian follicles in the LS-102-treated group essentially remained unchanged (P < 0:05). However, the number of atretic and cystic follicles was significantly increased. This indicates that SYVN1 may play potential roles in follicular development. The samples were assessed for SYVN1 and Drp1 protein levels using western blotting. We observed that the DHEA-induced PCOS groups (saline and LS-102 treated) had significantly lower levels of SYVN1 when compared to the control group ( Figure 6(c)). A higher number of fragmented mitochondria were observed in the PCOS and LS-102 group when compared to the control group (Figure 6(d)). Western blotting indicated that the PCOS group displayed significantly high levels of cleaved caspase and Bax, but a significantly low level of Bcl-2 when compared to the control group. LS-102 treatment elevated the levels of apoptosis as indicated by increased cleaved caspase-3 and Bax levels, along with decreased Bcl-2 levels, compared to both control and PCOS groups (Figure 6(e)). These results indicated that SYVN1 plays an important role in inhibiting apoptosis and mitochondrial fission in the GC cells of rats with PCOS.

Discussion
Women with PCOS are considered at greater risk of many physiological and psychological long-term complications [2]. Genetics, oxidative stress, insulin resistance, lipid imbalance, and hyperinsulinemia have been identified as some of the contributing factors to PCOS [1,20,21]. However, the etiology of this disease is still unclear. The most common strategy used to alleviate PCOS is the use of drugs, such as metformin, that target metabolic and hormonal complications [22]. Alternative strategies use selenium to act as an antioxidant and to reduce the devastating consequences of ROS in PCOS [23]. A study by Olaniyan et al. suggested that ROS can contribute to the apoptosis of GCs, and the use of 3-nitropropionic acid could decrease ROS production and associated apoptosis [24]. In this study, we further examine the molecular pathway through which this apoptosis is caused in the GCs. Initially, using protein-protein interaction network analysis, we identified that SYVN1, the E3 ubiquitin ligase, could be associated with PCOS. We further confirmed that SYVN1 is significantly downregulated and is associated with increased apoptosis in PCOS granulosa and KGN cells. The increase in apoptosis was further characterized by cleaved caspase-3 and the proapoptotic marker Bax. Indeed, we observed an upregulation of these markers in PCOS cells and a clear decrease in antiapoptotic markers such as Bcl-2. Several reports on BCL-2 expression in PCOS have found decreases in the level of GCs in PCOS patients and in the ovarian tissues of PCOS animal models [25,26], which correspond with the results found in our study. Overall, our results imply that ovarian GC cell apoptosis is essential for the progression of PCOS.
In this study, we also observed that apoptosis in PCOS and the downregulation of SYVN1 are highly correlated with increased mitochondrial fission. Mitochondrial dynamics are complex, and the regulation of the fusion and fission process is vital [4]. Mitochondria fuse and divide as they migrate along the cytoskeleton, fusion is followed by fission, and the absence or deregulation of either could be deleterious to the cell [4]. In this study, we were specifically interested in the role of mitochondrial fission in PCOS. Drp1 is considered a vital regulator of mitochondrial fission, and cells lacking Drp1 are known to have highly interwoven mitochondrial nest-like structures [27,28]. Drp1, alternatively known as Dnm1 in yeast, functions through its interaction with its two partners, mitochondrial fission 1 (Fis1) and mitochondrial division protein 1 (Mdv1) [29,30]. The process of mitochondrial fission is probably initiated by Fis1 which recruits Mdv1 to the mitochondrial membrane [4,31]. Mdv1 nucleates the Dnm1/Drp1-GTP oligomer, which forms spirals around the mitochondrial membrane and serves the mitochondria into fragments [4]. Previous, studies have indicated that the upregulation of Fis1 leads to increased fragmentation of mitochondria, whereas downregulation of Fis1 leads to fused mitochondria [32].
In PCOS GCs, we observed that SYVN1 downregulation is associated with the upregulation of Drp1. Indeed, we could clearly observe an associated increase in mitochondrial fission and apoptosis in the PCOS models. Understanding the role of Drp1 in apoptosis would be essential to understanding its role in PCOS. Under apoptotic stimulus, Drp1 migrates to the mitochondrial membrane and recruits Mfn1 and Bax to initiate mitochondrial fission. Studies similar to ours [33,34] have indicated that the lack of Drp1 could decrease mitochondrial fission, cleaved caspase-3 production, and apoptosis. Inhibition of Drp1 appears to inhibit the apoptotic process just before Bax translocation and cytochrome C release [35]. Hence, the regulation of Drp1 is a vital step in the regulation of apoptosis in PCOS. SYVN1 is an E3 ubiquitin ligase protein identified to play a vital role in ER-associated degradation [14]. Defects in ER-associated degradation have been associated with diseases where the disposal of accumulating toxic protein is necessary, such as Alzheimer's and Parkinson's [36].
Interestingly, SYVN1 has been associated with the ubiquitination of these toxic proteins and degrading them through the proteasomal pathway, thus maintaining the balance in the ER and decreasing ER-associated stress [37,38]. SYVN1 is also known to mediate ubiquitination and degradation of proteins to regulate apoptosis in other diseases such as ER proteotoxic stress and liver injury [39] and lens epithelial cells [40]. Both of these diseases involve the misfolding and aggregation of proteins through SYVN1-mediated degradation via the ubiquitin-proteasome pathway. In this study, we identified that Drp1 coprecipitates with SYVN1 and that the upregulation of SYVN1 further degrades Drp1. Moreover, Drp1 degrades quickly in the presence of SYVN1. We also observed that the use of proteasomal inhibitor upregulated Drp1 levels even in the presence of SYVN1. Overall, our results indicate that SYVN1 regulates the levels of Drp1 using the proteasome/ubiquitin pathway.
In a DHEA-mediated PCOS rat model, SYVN1 is highly downregulated in PCOS and an SYVN1 inhibitor exacerbated the PCOS by significantly increasing Drp1 and apoptosis. In addition to increased apoptosis, we also found an elevated level of atretic follicles in the PCOS rat model and the number increased even further when Drp1 was upregulated by the use of protease inhibitor, implying that SYVN1 can influence folliculogenesis. In support of our results, Salehi et al. [12] observed that Drp1 is associated with excessive mitochondrial fission and apoptosis in GCs and found that follicular growth arrest is important in the pathogenesis of PCOS. They found that GC apoptosis and follicular growth arrest can disrupt folliculogenesis and anovulation, and that upregulated Drp1 can increase levels of mitochondrial fission. Overall, we obtained similar results, SYVN1 binds and degrades Drp1, and the loss of Drp1 thus modulates mitochondrial fission and apoptosis in ovarian GCs.

Conclusions
This study indicates that SYVN1 binds and degrades Drp1 and thus modulates mitochondrial fission and apoptosis in ovarian GCs. Moreover, the current study provides a novel target for the development of therapeutic strategies for PCOS.

Data Availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.