Involvement of the Serine Protease Inhibitor, SERPINE2, and the Urokinase Plasminogen Activator in Cumulus Expansion and Oocyte Maturation

The serpin peptidase inhibitor, clade E, member 2 (SERPINE2) inhibits urokinase-type plasminogen activator (PLAU) and tissue-type plasminogen activator. Higher SERPINE2 expression levels were detected in cumulus cells of human immature oocytes than in those of mature oocytes. The objective of this study was to evaluate whether high SERPINE2 levels in cumulus cells are associated with oocyte immaturity. Using the mouse cumulus–oocyte complex as an experimental model, the effects of elimination and overexpression of SERPINE2 in cumulus cells on cumulus expansion and oocyte maturation were assayed by in vitro maturation. Serpine2 and PLAU transcripts were the most highly expressed serpins and plasminogen activators, respectively. Their expression was coordinately regulated in cumulus cells during gonadotropin-induced oocyte maturation. Silencing of Serpine2 expression using small interfering RNAs or blockage of SERPINE2 protein using a specific antibody had no effect on oocyte maturation. However, overexpression of Serpine2 or exogenous supplementation with high levels of SERPINE2 impaired cumulus expansion and oocyte maturation, probably by decreasing hyaluronan synthase 2 (Has2) and versican (Vcan) mRNA expression. Amiloride, a specific PLAU inhibitor, also suppressed these processes. PLAU supplementation of the oocyte in vitro maturation medium caused earlier and more extensive expansion of cumulus cells and oocyte maturation that may be mediated by increased Has2 mRNA expression. However, these effects were neutralized by coincubation with SERPINE2 or amiloride and PLAU. In conclusion, SERPINE2 and PLAU are involved in cumulus expansion and oocyte maturation. High SERPINE2 levels impair these processes, probably by decreasing cumulus matrix gene expression as well as reducing cumulus hyaluronan contents and inhibiting PLAU activity. These findings may explain why cumulus cells surrounding immature human oocytes express high SERPINE2 levels.

PA expression levels upregulated in cumulus cells just before ovulation [13] suggests involvement in follicle wall rupture during ovulation [13][14][15][16][17]. However, Liu et al showed that plasmin, the best known target of PA, was not critical for ovulation [32]. PAs and their cognate serpin inhibitors have been detected in cumulus cells [27]; nevertheless, their involvement in oocyte maturation during pre-ovulation needs clarification. Higher SERPINE2 expression levels were detected in cumulus cells of human immature oocytes than in those of mature oocytes [33]. Therefore, we here assumed that high SERPINE2 levels were correlated with oocyte immaturity. To verify this, we used mouse cumulus-oocyte complexes (COCs) as a model for evaluating the association of SERPINE2 levels with cumulus expansion and subsequent oocyte maturation.

Ethics statement
This study was approved by the Mackay Memorial Hospital Institutional Review Board (reference number 09MMHIS024) with written consent for the use of human cumulus cells. Written consent for the use of cumulus cells was obtained from 20 patients undergoing intracytoplasmic sperm injection treatment. All animals contributed to this study were maintained in the Animal Center at the Department of Medical Research, Mackay Memorial Hospital. The animal use protocol has been reviewed and approved by the Mackay Memorial Hospital Institutional Animal Care and Use Committee with an approval number MMH-A-S-100-45. All efforts were made to minimize suffering.

Collection of human cumulus cells
Patients undergoing classical in vitro fertilization treatments at the Center of Reproductive Medicine, Mackay Memorial Hospital, Taiwan received controlled ovarian hyperstimulation by application of the gonadotropin-releasing hormone antagonist protocol. COCs from follicles .14 mm were collected using transvaginal ultrasound and a 16-gauge needle and were exposed to 80 IU hyaluronidase in Quinn's Advantage Fertilization medium (Sage BioPharma, Bedminster, NJ) for 20 s at 37uC to dissolve hyaluronan. Of the 46 COCs, 26 and 20 had mature and immature oocytes, respectively. The cumulus cells were individually separated from the COCs under an Olympus SZX7 stereomicroscope (Tokyo, Japan). They were mixed with 20 ml of extraction buffer from the Arcturus PicoPure RNA Isolation Kit (Applied Biosystems, Foster City, CA) for total RNA isolation and stored at 280uC until use. Cumulus cells individually collected from 10 other COCs were fixed on slides using 4% (v/v) paraformaldehyde for immunohistochemical staining.

Collection of mouse cumulus cells
The mice (age, 21-24 days) were injected with 5 IU of pregnant mare serum gonadotropin (PMSG; Sigma-Aldrich, St. Louis, MO) and sacrificed by cervical dislocation after 46 h. The ovaries were removed and briefly rinsed with PBS. COCs were isolated by puncturing antral follicles with a 30-gauge needle under an Olympus SZX7 stereomicroscope. To study the effect of luteinizing hormone on Serpine2 and Plau expression in cumulus cells during oocyte maturation, PMSG-primed mice were injected with 5 IU of human chorionic gonadotropin (hCG; Sigma-Aldrich). Ovaries were removed 3, 6, and 9 h after hCG treatment. COCs were isolated by puncturing antral follicles as described above. For post-ovulation COCs, the ovaries were removed 12 h after hCG injection, and the COCs were collected by flushing the oviducts with PBS. All COCs were treated with 150 IU hyaluronidase in PBS for 3 min at 37uC, the oocytes were removed, and cumulus cells were collected by centrifugation at 1000 6g for 3 min at room temperature.

Quantitative real-time RT-PCR (qRT-PCR)
Total RNA of cumulus cells was extracted using the Arcturus PicoPure RNA Isolation Kit and directly reverse transcribed into a 50 ml first-strand cDNA pool using a High Capacity cDNA Archive Kit (Applied Biosystems) according to the manufacturer's instructions. qRT-PCR was performed using primers (Table S1) [24]. The housekeeping genes, human ribosomal protein L19 and mouse hypoxanthine guanine phosphoribosyltransferase gene, were used as internal loading controls to normalize relative gene expression levels. PCR amplification efficiency for each tested gene was examined to ensure that it was equivalent to that of the housekeeping gene examined in a cDNA dilution series. PCR was performed in a total volume of 20 ml, containing 50 ng of tissue cDNA, 250 nM each of the forward and reverse primers, and 10 ml of 26 SYBR Green Master Mix (Applied Biosystems). All reactions were performed in triplicate and run on an ABI/PRISM 7500 Fast Sequence Detection System (Applied Biosystems) under the following conditions: 95uC for 20 s, and then 40 cycles at 95uC for 3 s and 60uC for 30 s. The threshold cycle (Ct) was defined as the fractional cycle number at which the reporter fluorescence, i.e., the number of amplified copies, reached a fixed threshold. Melting curve analysis was performed to verify that only a single product had formed in the reaction. The identity of the PCR products was confirmed by DNA sequencing. Relative quantification of mRNA expression was calculated by the 2 2DDCt method [34].

SERPINE2 and PLAU immunolocalization and hyaluronan status on COCs
COCs were transferred onto slides, air dried, and fixed in 4% paraformaldehyde for 15 min. The slides were incubated in blocking solution [10% (v/v) normal goat serum in PBS] for 1 h at room temperature and then incubated with anti-SERPINE2 or control antiserum (1:1000), with rabbit anti-PLAU antiserum (1:100; Santa Cruz Biotechnology, Santa Cruz, CA), or with preimmune rabbit serum (1:500; Jackson ImmunoResearch, West Grove, PA) in blocking solution at 4uC for 16 h. To assess the hyaluronan status in cumulus cells, slides were incubated with biotinylated hyaluronic acid binding protein (HABP; 1:200, Sigma-Aldrich, cat. no. H9910) in blocking solution at 4uC for 4 h. After washing three times in PBS with slight agitation for 5 min each, the slides were treated with fluorescein isothiocyanateconjugated goat anti-rabbit IgG (1:1000; Jackson ImmunoRe-

In vitro maturation (IVM)
To assess the extent of cumulus cell expansion, COCs isolated from PMSG-primed ovaries that had even diameters of approximately 100 mm and contained a nucleus (germinal vesicle, GV) were cultured in IVM medium as described previously [35,36] with some modifications. The IVM medium consisted of MEMa medium (Life Technologies, Grand Island, NY) supplemented with 10% fetal bovine serum (Sigma-Aldrich), 0.23 mM sodium pyruvate, 75 mU/ml follicle-stimulating hormone (FSH), 50 mg/l streptomycin, 60 mg/l penicillin, and 1 mg/l epidermal growth factor (EGF), pH 7.4. COCs were incubated in 150-ml microdrops of IVM medium supplemented with SERPINE2 (0.03, 0.06, or 0.12 mg/ml), anti-SERPINE2 antibody (1:1000), amiloride (300 mM), or PLAU (20 U; Millipore, Billerica, MA) and overlaid with mineral oil for approximately 16-20 h in a humidified 5% CO 2 atmosphere at 37uC. For control experiments, COCs were incubated in IVM medium without supplementation. After IVM, the diameters of expanded cumulus cells were scored. Next, the COCs were treated with 150 IU hyaluronidase in IVM medium for 3 min at 37uC, and cumulus cells were removed by repeated pipetting. The morphology of oocyte nuclei was observed, and the oocytes were classified as immature [GV or metaphase I (MI) stage] or mature (MII stage, with the extrusion of the first polar body). Oocyte maturation rate was determined after 16 h of culture by counting the number of MII oocytes among the total oocytes used in an assay.

Treatment of COCs with small interfering RNA (siRNA)
siRNA against mouse Serpine2 (catalog #20720-Serpine2; Dharmacon, Thermo Fisher Scientific, Lafayette, CO) and a non-targeting negative control siRNA (catalog #D-001206-05; Dharmacon, Thermo Fisher Scientific) dissolved in Accell siRNA delivery media were used according to the manufacturer's instructions. COCs isolated from PMSG-primed ovaries were incubated with 1, 2, or 3 mM siRNAs for 24 h in 150 ml MEMa medium supplemented with 10 mM milrinone (a phosphodiesterase inhibitor, Sigma-Aldrich, cat. no. M4659), 50 mg/l streptomycin, 60 mg/l penicillin, 0.23 mM pyruvate, and 3 mg/ml bovine serum albumin (Sigma-Aldrich). The optimal concentration for both siRNAs was 3 mM. After 24 h incubation, the COCs were transferred to IVM medium and cultured in a humidified 5% CO 2 atmosphere at 37uC for 16 h. Cumulus expansion and oocyte maturation were then assessed as described above. Serpine2 mRNA levels in cumulus cells were examined by qRT-PCR. To analyze whether SERPINE2 protein was knocked down, COCs were transferred onto slides and examined by immunohistochemistry as described above.

Construction of the mouse Serpine2 expression vector
The DNA fragment of the pIRES2-DsRed2 vector (Clontech Laboratories, Mountain View, CA) containing the multiple cloning site (MCS) and the red fluorescence protein coding region (DsRed2) was amplified by PCR using primer pairs bearing EcoRI sites (forward primer 59-TTCGAATTCTGCAGTCGACGG-TACC-39, reverse primer 59-TTTGAATTCATCTA-GAGTCGCGGCCGC-39; Figure S1a). Thirty-five PCR cycles were performed under the following conditions: denaturation at 95uC for 30 s, annealing at 55uC for 30 s, and extension at 72uC for 1 min. The PCR product was verified by agarose gel electrophoresis and DNA sequencing and ligated into an EcoRIdigested pCX-EGFP vector (Addgene, Cambridge, MA) to form the pCX-DsRed2 intermediate vector ( Figure S1b and c).
Since Serpine2 is predominantly expressed in mouse seminal vesicles [28], total RNA was extracted from that tissue using an RNeasy Mini Kit (Qiagen, Valencia, CA) and reverse transcribed into cDNA with a High-Capacity cDNA Archive Kit (Applied Biosystems) according to the manufacturer's instructions. The 1220-bp full-length mouse Serpine2 cDNA (NCBI Reference Sequence NM_009255.4) was amplified by RT-PCR from the cDNA pool using a Serpine2 primer pair (forward primer 59-GAAGGAACCATGAATTGGC-39, reverse primer 59-TTCCTTTGTCTGTCCTTCAGG-39). Thirty-five cycles of PCR were performed under the following conditions: denaturation at 95uC for 30 s, annealing at 55uC for 30 s, and extension at 72uC for 1 min. The PCR product was verified by agarose gel electrophoresis and DNA sequencing and cloned into the pGEM-T Easy vector (Promega, Madison, WI) by TA cloning.
The full-length Serpine2 cDNA was excised with XmaI and cloned into MCS of the pCX-DsRed2 vector to create the Serpine2 expression vector pCX-Serpine2-DsRed2 ( Figure S1d). The construct was sequenced to verify the sequence and orientation of the reading frame. This construct enabled the simultaneously translation of both SERPINE2 and DsRed2 for monitoring SERPINE2 protein expression by red fluorescence.

Serpine2 overexpression in COCs
COCs isolated from PMSG-primed ovaries were transfected with 500 ng of the Serpine2 expression vector pCX-Serpine2-DsRed2 or the vehicle vector pCX-DsRed2 using PolyJet DNA In Vitro Transfection Reagent (SignaGen Laboratories, Gaithersburg, MD) in 150 ml of MEMa medium supplemented with 10 mM milrinone (as mentioned above) but without FSH and EGF for 12 h. The COCs were washed three times using IVM medium, transferred to fresh medium, and cultured in a humidified 5% CO 2 atmosphere at 37uC for 16 h. Cumulus expansion and oocyte maturation were assessed as described above. Serpine2 mRNA levels in cumulus cells were examined by qRT-PCR.
(C) and Plau (D) mRNA levels in mouse cumulus cells surrounding developing oocytes (three mice for each group) during gonadotropin-induced oocyte maturation. Bars indicate means 6 SD of three independent experiments each. { P,0.05, # P,0.001, *P,0.0001 compared with the cumulus cells 46 h after PMSG injection alone (PMSG, 46 h). Immunofluorescent staining revealed SERPINE2 (E) and PLAU (F) protein levels in mouse COCs following gonadotropin treatments. COCs were fixed on slides and immunostained using anti-SERPINE2 antiserum and anti-PLAU antibody as described in the Materials and Methods: a, 46 h after PMSG; b, 3 h after hCG; c, 6 h after hCG; d, 9 h after hCG; e, 12 h after hCG administration; f, immunostaining with the control serum (con). Scale bars, 100 mm. doi:10.1371/journal.pone.0074602.g002

Statistical analysis
Data are presented as means 6 SD. Differences were analyzed by one-way analysis of variance followed by the Bonferroni post hoc test using GraphPad Prism 5.0 (GraphPad Software, San Diego, CA). P,0.05 was considered significant.

SERPINE2 was highly expressed in cumulus cells of immature human oocytes
We analyzed the expression levels of the four SERPINs of the PA system in cumulus cells of the mature human oocyte by qRT-PCR and found that SERPINE2 were the most highly expressed ( Figure  1A). Next, we compared SERPINE2 mRNA expression levels in cumulus cells collected from mature and immature human oocytes. Cumulus cells from immature oocytes expressed significantly (P,0.0001) higher SERPINE2 mRNA levels than those from mature oocytes ( Figure 1B). Similarly, considerably more SERPINE2 protein was detected in cumulus cells from immature human oocytes at the GV and MI stages ( Figure 1C, a and b, respectively) than in those from mature MII oocytes or with the control staining of MII oocytes ( Figure 1C, c and d, respectively). Other similar cases are shown in Figure S2.
Serpine2 and Plau were highly expressed in mouse cumulus cells during oocyte maturation We analyzed the expression profiles of the four Serpins of the PA system in cumulus cells surrounding mature mouse oocytes. Similar to the results with human cumulus cells, Serpine2 mRNA was the most highly expressed in mouse cumulus cells (Figure 2A). Next, we analyzed the gene expression patterns of SERPINE2targeted serine proteases in the cumulus cells of mature mouse oocytes using qRT-PCR. Plau mRNA was the most highly expressed, followed by Plat and Prss8 mRNAs. F2 mRNA was almost undetectable ( Figure 2B).
To examine the in vivo expression pattern of Serpine2 and Plau mRNAs in mouse cumulus cells during oocyte maturation, the cumulus cells were collected at various intervals during gonadotropin-induced oocyte maturation. Serpine2 mRNA was highly expressed 46 h after PMSG treatment and reached a maximum level 3 h after hCG administration, gradually decreasing to its lowest level 12 h after hCG administration ( Figure 2C). Plau mRNA was at a low level 46 h after PMSG treatment; however, it peaked 3 h and 6 h after hCG treatment and then gradually decreased to a low level after 12 h ( Figure 2D). The relative changes in Plau mRNA levels were much greater than the changes in Serpine2 mRNA levels ( Figure 2C and 2D). Expression of SERPINE2 and PLAU proteins was consistent with their mRNA expression in cumulus cells. SERPINE2 was at a relatively high level following PMSG administration and 3 h after hCG treatment ( Figure 2E, a and b). PLAU was at a relatively low level after PMSG treatment but peaked approximately 3-6 h after hCG treatment ( Figure 2F, a-c). From 6 h after hCG on, SERPINE2 protein levels were gradually decreased to a very lower level ( Figure 2E, c-f); on the contrary, PLAU protein levels were still at higher levels at 6 h after hCG ( Figure 2F, c) and then sharply decreased to a very low level thereafter ( Figure 2F, d-f).
Serpine2 silencing or SERPINE2 protein blockage had no effect on cumulus expansion and oocyte maturation in vitro IVM is often used to culture compact immature oocytes collected from PMSG-primed ovaries ( Figure 3A, a) for developing MII mature oocytes with fully expanded cumulus cells ( Figure 3A,  b). To examine the effect of Serpine2 silencing on oocyte maturation in cumulus cells, siRNA was used to knockdown Serpine2 mRNA expression during IVM. Cells were also treated with SERPINE2 antiserum to examine the effect of blocking SERPINE2 protein.
No detrimental effects on COC structure or morphology ( Figure  3A, c-e) or on the extent of cumulus expansion ( Figure 3B) were observed, although Serpine2 mRNA was significantly decreased (P,0.0001) and SERPINE2 protein was knocked down by the introduction of Serpine2 siRNA ( Figure 3C and Figure S3, respectively).
The treatments had no effect on oocyte maturation (Table S2). As shown in Figure 3D, more than 70% of oocytes reached the MII stage even when Serpine2 mRNA in cumulus cells was knocked down. Oocyte maturation was comparable in the media with and without control or Serpine2 siRNAs or specific anti-SERPINE2 antiserum. Taken together, these findings indicated that eliminating SERPINE2 in cumulus cells had no apparent effect on cumulus expansion and oocyte maturation.

SERPINE2 overexpression in cumulus cells impaired cumulus expansion and oocyte maturation in vitro
To test whether high SERPINE2 levels affected cumulus expansion and oocyte maturation, mouse COCs were transfected with a vector carrying Serpine2. The COC morphology was symmetrical with the outward expansion pattern of cumulus cells from the oocyte in both the untreated control and after transfection with control plasmid DNA ( Figure 4A, a and b); however, the cumulus cell was compact or in an unexpanded state after transfection with the Serpine2 plasmid ( Figure 4A, c). SERPINE2 protein was significantly overexpressed in cumulus cells after transfection and culturing for 16 h, although most of the protein expression was in the outer layer of cumulus cells ( Figure  4A, d). Similarly, exogenously added SERPINE2 resulted in compact, unexpanded cumulus cells that tightly encircled the oocyte ( Figure 4A, e-g). Serpine2 mRNA was significantly overexpressed in cumulus cells after transfection and culturing for 16 h ( Figure 4B). Serpine2 overexpression significantly reduced the extent of cumulus expansion compared to transfection with the control plasmid (P,0.0001; Figure 4C, gray bars). Exogenous SERPINE2 also significantly inhibited cumulus cell expansion compared to the control ( Figure 4C, black bars).
Serpine2 overexpression in cumulus cells or exogenously added SERPINE2 significantly reduced oocyte maturation, with most oocytes halting at the MI stage (Table S3). Introduction of the Serpine2 plasmid into cumulus cells significantly reduced oocyte maturation by approximately 45% compared with the control group ( Figure 4D, gray bars). SERPINE2 supplemented exoge- nously also significantly reduced oocyte maturation by approximately 26-42% ( Figure 4D, black bars).

PLAU protein was involved in cumulus expansion and oocyte maturation
Since PLAU was the most highly expressed serine protease in cumulus cells, we examined PLAU effects on cumulus expansion and oocyte maturation. COC expansion was visible at 6 h of culture ( Figure 5A, a) and had fully expanded cumuli with an average diameter of 236 mm after approximately 16-20 h of culture ( Figure 5A, b, and 5B, open bar). PLAU supplementation significantly expanded the COCs to an average diameter of 291 mm ( Figure 5A, d, and 5B, gray bar), and the expansion occurred earlier at 6 h during IVM compared with that in the control group ( Figure 5A, a and c). Furthermore, the PLAUsupplemented cumulus cells degraded earlier (at 20 h) than the control cells ( Figure 5A, b and e), which generally degraded at 24 h.
To determine whether the PLAU effect on oocyte maturation was specific, amiloride, a specific inhibitor of PLAU [37], was added to the IVM medium. As shown in Figure 5A, f, and 5B (black bar), cumulus expansion was significantly diminished, and cumulus cells remained encircling the GV oocyte at 20 h, similar to the effects of SERPINE2 addition ( Figure 4A, g, and 4C, black bars). To further demonstrate that the inhibition of cumulus expansion was due to PLAU suppression, amiloride or SERPINE2 was coincubated with PLAU during IVM. Intriguingly, the COC morphology appeared normal ( Figure 5A, g and h), and the extent of cumulus cell expansion was comparable to that in the control group ( Figure 5B, hatched bars).
PLAU significantly promoted oocyte maturation (P,0.05), whereas amiloride significantly reduced oocyte maturation (P,0.0001) ( Figure 5C and Table S4). Coincubation with PLAU and amiloride or SERPINE2 ( Figure 5C, hatched bars) reduced maturation to levels comparable with the control group (approximately 53% and 56%, respectively, vs. 67% for the control). Taken together, these data suggested that PLAU was involved in cumulus expansion and oocyte maturation and that its effects could be modulated by SERPINE2.

Excessive PLAU and SERPINE2 altered matrix gene expression and the hyaluronan status of cumulus cells during IVM
To examine the effect of excessive PLAU and SERPINE2 on the temporal gene expression pattern of the matrix genes in cumulus cells, cumulus cells at 3, 6, and 16 h, the critical time points, during IVM culture were collected from COCs and analyzed by qRT-PCR. PLAU significantly enhanced but SERPINE2 significantly down-regulated cumulus hyaluronan synthase 2 (Has2) mRNA levels at 3 and 6 h of IVM compared with that in the control IVM group (Figure 6A, a). Vcan mRNA levels in cumulus cells were significantly diminished by exogenous SERPINE2 at 3 and 6 h and by PLAU at 6 h of IVM, but with a PLAU-induced surge at 16 h of IVM (Fig. 6A, b). Cumulus Tnfaip6 mRNA levels were enhanced by exogenous PLAU after 3 h of IVM; however, this enhancement disappeared at 6 and 16 h of IVM culture. SERPINE2 showed no effect on Tnfaip6 mRNA expression at all the time points (Fig. 6A, c). Furthermore, exogenous PLAU and SERPINE2 had no effect on Ptx3 mRNA expression in cumulus cells (Fig. 6A, d).
HAS2 is a critical enzyme required for matrix hyaluronan synthesis [38]. Since Has2 mRNA expression is the most affected, we analyzed the hyaluronan status of cumulus cells during IVM. After 6 h of culturing, mouse COCs showed moderate expansion with relatively high hyaluronan contents around the cumulus matrix (Fig. 6B, a and e), in contrast to the negative staining control (Fig. 6B, b and f). Hyaluronan staining appeared to cover the entire cell; however, the staining was clearly outside the cumulus cell when ovarian tissue slides were stained. Thus, this staining pattern is probably caused by the steric stacking of cumulus cells (Fig. S4). Intriguingly, exogenous PLAU increased hyaluronan contents (Fig. 6B, c and g), while SERPINE2 supplementation decreased the contents on the cumulus matrix compared with that in the control group (Fig. 6B, d and h).

Discussion
SERPINE2 overexpression in mouse COCs or its exogenous addition to IVM medium remarkably impaired cumulus expansion, with the COCs exhibiting a compact morphology with a high proportion of GV and MI oocytes, leading to a significant reduction in oocyte maturation ( Figure 4 and Table S3). These findings from the mouse model may explain why cumulus cells surrounding immature human oocytes expressed more SER-PINE2 than those surrounding mature oocytes and support our hypothesis that aberrantly high SERPINE2 levels correlate with oocyte immaturity.
PLAU and SERPINE2 were the most abundant PA and PA inhibitor, respectively, in murine cumulus cells, and their gene expression levels were coordinately regulated during gonadotropin-induced oocyte maturation (Figure 2). SERPINE2 decreased rapidly 6 h after hCG administration, consistent with apparent cumulus expansion, whereas PLAU remained at a high level. Thus, the net proteolytic activity of PLAU may contribute to the initiation of cumulus expansion. This interplay appears to suggest that the net activity of PLAU may be crucial for cumulus expansion and subsequent oocyte maturation. Furthermore, we found that PLAU depletion via its specific inhibitor, amiloride, largely impaired these biological processes (Figure 4).
Hagglund et al. reported high Serpine2 mRNA levels and low Plau mRNA levels in mouse cumulus cells [27] and suggested that SERPINE2 may provide inhibitory activity for protecting the mucified COC matrix from proteolytic degradation. However, we found that cumulus cells expressed both Plau and Serpine2 mRNA and protein during gonadotropin-induced oocyte maturation ( Figure 2). Furthermore, in vivo, the granulosa cells, which are far more numerous, may also produce these proteins. Hence, we examined their relative expression in cumulus and granulosa cells after hCG 3, 6, and 9 h by immunohistochemistry. Granulosa cells Figure 4. Effects of Serpine2 overexpression and addition of exogenous SERPINE2 on cumulus expansion. A, The morphologies of untreated and treated COCs after 16 h of IVM culture are shown: a, IVM control; b, transfected with the control plasmid; c, transfected with the Serpine2-expressing plasmid. d, Cultured COCs transfected with the Serpine2-expressing plasmid were fixed on slides, and SERPINE2 protein expression was monitored by epifluorescence microscopy. The red fluorescence represents coexpression of SERPINE2 and the red fluorescent protein. e-g, COCs were incubated with 0.03, 0.06, and 0.12 mg/ml SERPINE2 protein, respectively. B, qRT-PCR revealed the relative Serpine2 mRNA levels in cumulus cells transfected with Serpine2-expressing or vehicle plasmids. C, The extent of cumulus expansions are shown as COC diameters after 16 h of IVM culture (n = 100 for each group). D, MII oocyte maturation rate after 16 h of IVM culture. B-D, Data represent means 6 SD for five independent experiments. # P,0.001, *P,0.0001 vs. untreated control or control plasmid. doi:10.1371/journal.pone.0074602.g004  expressed these proteins at levels similar to cumulus cells, especially at 3 and 6 h of hCG treatment ( Figure S5). PLAU and SERPINE2 detected in mural granulosa cells may also become associated with the COC matrix during cumulus expansion.
We compared cumulus SERPINE2 and PLAU levels in COCs treated with hCG in vivo or cultured in vitro. Cumulus SERPINE2 levels had no significant difference at 3 h; however, they were significantly higher at 6 and 9 h after culturing in vitro compared with those treated with hCG in vivo (Fig. S6). This could be coordinated by FSH and EGF because it was found that FSH could enhance SERPINE2 expression and FSH coupled with EGF (1 ng/ml in this study) could even further enhance SERPINE2 expression in bovine granulosa cells cultured in vitro [39]. Furthermore, cumulus PLAU levels were significantly higher at 3 and 6 h when treated with hCG in vivo than cultured in vitro, despite absence in changes thereafter (Fig. S6). This revealed that PLAU synthesis is repressed in FSH-stimulated COCs by an oocyte-soluble factor [13]. Two cumulus matrix genes, Adamts1 (a disintegrin and metalloproteinase with thrombospondin-like motifs) and Vcan, have higher expression levels in COCs that were treated with hCG in vivo than those cultured in vitro [40]. These data illustrate that the expression of cumulus matrix genes is altered when COCs are cultured in vitro.
Our results may suggest that SERPINE2 has only limited influence or that other inhibitors may be involved in regulating the activity of PLAU and other proteases. This is supported by our result that siRNA silencing of Serpine2 in cumulus cells had no effect on cumulus expansion or oocyte maturation (Figure 3). Because the siRNA efficiency is still not high enough, we cannot rule out the possibility that the residual SERPINE2 is enough for maintaining normal function. Furthermore, Murer et al. reported that Serpine2 knockout in mice did not result in female infertility [41].
These results appear to indicate that SERPINE2 can be depleted and compensated for by other factors, but that its overexpression hinders cumulus ECM remodeling by proteases and thus impairs cumulus expansion and subsequent oocyte maturation. Cumulus expansion includes limited ECM remodeling and requires sophisticated regulation of proteases and protease inhibitors. A convincingly characterized example is that ADAMTS1 cleavage of VCAM mediates essential remodeling of the COC matrix during ovulation [5].
PA expression in cumulus cells is species-specific. PLAT is the major PA expressed in bovine granulosa cells [42] and rat and human cumulus cells [15,30] (Figure S7), whereas PLAU is the major PA detected in murine granulosa and cumulus cells [27]. Ny et al. found that plasmin activity in the mouse ovary was increased 2-8 h after hCG treatment, and most of the activity was generated by PLAU [43]. D'Alessandris et al. found that mouse PLAT and PLAU activities dramatically increased between 16 h and 20 h of IVM and concluded that both PAs might function to destabilize the expanded COC matrix [13]. Several studies also demonstrated that PAs play a crucial role in follicle wall rupture during ovulation [13][14][15][16][17]; however, their function in oocyte maturation during preovulation remained unclear. Our study provides the first evidence that PLAU and its inhibitor SERPINE2 are involved in murine cumulus expansion and oocyte maturation.
Cumulus matrix genes, e.g., Has2 and Vcan, are induced and these genes normally peak around 4-6 h after treatment with hCG in vivo or FSH/EGF in vitro [40]. The temporal cumulus matrix Has2 and Vcan expression and hyaluronan status were altered by exogenous supplementation of SERPINE2 and PLAU during IVM ( Figure 6). Has2 expression is correlated with hyaluronan synthesis, which is necessary for cumulus expansion [44]. SERPINE2 down-regulated Has2 expression; thus, reducing the matrix hyaluronan contents. VCAM, an important cumulus matrix proteoglycan, is involved in cumulus expansion [5]. SERPINE2 also down-regulated Vcan expression at critical time points of cumulus expansion during IVM. These findings may explain why excessive SERPINE2 resulted in compact, unexpanded cumulus cells ( Figure 4A). On the other hand, PLAU upregulated Has2 expression and increased hyaluronan contents. These effects may cause earlier expansion and degradation than that seen in the control group ( Figure 5A).
PLAU supplemented exogenously led to early cumulus expansion and matrix degradation and enhanced oocyte maturation ( Figure 5). This is probably through up-regulation of Has2 expression and increased hyaluronan contents in the cumulus matrix ( Figure 6). Coincubation of PLAU with amiloride or SERPINE2 neutralized the PLAU effects on cumulus expansion and subsequent oocyte maturation. These findings indicate that SERPINE2 can modulate PLAU activity. Superovulation treatments sometimes retrieve immature oocytes from assisted reproductive technology (ART) patients [45,46]. Thus, PLAU supplemented exogenously may be helpful for maturation of immature oocytes during IVM.
In summary, the present results support the involvement of SERPINE2 and its cognate serine protease PLAU in cumulus expansion and subsequent oocyte maturation. Depletion or elimination of SERPINE2 expression has no effect on cumulus expansion and oocyte maturation; however, high SERPINE2 levels bound to the cumulus ECM could down-regulate Has2 and Vcan expression and decrease matrix hyaluronan contents, leading to suppressed cumulus expansion and oocyte maturation. On the other hand, PLAU supplementation to IVM culture medium upregulated Has2 expression and increased matrix hyaluronan contents could be a potential therapeutic strategy for rescuing immature human oocytes collected during ART procedures, although further study is required.  Figure S2 Immunofluorescence staining for SERPINE2 protein levels in human cumulus cells. COCs were collected from patients whose oocytes had all 3 nuclear stages (GV, MI, MII). Cumulus cells collected from immature oocytes at the GV (a) or MI stages (b) and from mature oocytes at the MII stages (c) were immunostained with anti-SERPINE2 antibody (green for case 2 and 3 or red for case 4), and cumulus cells of MII oocytes were immunostained with the control serum (d). For contrast, the slides were counterstained with Hoechst 33258 (blue, case 2 and 3) or hematoxylin (blue, case 4). Scale bars, 100 mm. (TIF) Figure S3 Immunohistochemistry of SERPINE2 protein in cumulus cells treated with Serpine2 siRNA. COCs treated without (control, a and e) or with control siRNA (b and f), or Serpine2 siRNA (c and g) were cultured in IVM medium for 16 h. After culturing, COCs were transferred and fixed onto slides and immunostained by anti-SERPINE2 (green) or control antiserum (d and h). For contrast, the slides were counterstained with Hoechst 33258 (blue, a-d) or photographed under differential interference contrast microscopy (e-h). Scale bars, 100 mm. (TIF) Figure S4 Hyaluronan matrix staining of the tissue section and COC. The hyaluronan on ovarian sections that were PMSG-primed and treated with hCG for 3 h (a) or COCs that were cultured for 6 h in vitro and then transferred onto slides (b) were stained with HABP (green) as described in Materials and Methods. For contrast, the slides were counterstained with Hoechst 33258 (H33258, blue, c and d) and the merged images are also shown (e and f). Scale bars, 100 mm. (TIF) Figure S5 Immunolocalization of SERPINE2 and PLAU in ovarian follicles during gonadotropin treatment. Ovarian sections from PMSG-primed and hCG administration for 3, 6, and 9 h were immunostained using anti-SERPINE2 antiserum and anti-PLAU antibody as described in Materials and Methods: a and e, 3 h after hCG; b and f, 6 h after hCG; c and g, 9 h after hCG; d and h, immunostaining with the control serum (con). Scale bars, 100 mm. (TIF) Figure S6 Cumulus SERPINE2 and PLAU protein levels in COCs treated with hCG in vivo or cultured in vitro. COCs isolated from PMSG-primed ovaries (a and f), treated with hCG for 3, 6, 9, and 12 h (b-e, respectively) or IVM culture for 3, 6, 9, and 16 h (g-j, respectively), were immunostained using anti-SERPINE2 antiserum and anti-PLAU antibody as described in Materials and Methods. The slides counterstained with Hoechst 33258 were photographed using a fluorescence microscope (Olympus BX 40) equipped with an Olympus DP-70 digital camera. The percentage of positively stained cells was determined using TissueQuest software (TissueGnostics, Vienna, Austria). A chi-square test was performed to independently compare the significance of difference in expression levels of SERPINE2 or PLAU in cumulus cells at different time points. *P,0.0001 compared with in vivo and in vitro samples at the same time point. Scale bars, 100 mm. (TIF) Figure S7 PLAT expression in cumulus cells of human oocytes. qRT-PCR revealed the relative levels of serine protease mRNAs in cumulus cells of mature (n = 16) human oocytes. Bars indicate means 6 SD of sixteen independent experiments each. *P,0.0001 compared with F2 mRNA. (TIF)