| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Division of Animal Sciences,3 University of Missouri-Columbia, Columbia, Missouri 65211
Research Center for Transgenic Cloned Pigs,4 Chungnam National University, Daejeon, Korea
Institute of Animal Physiology & Genetics,5 Czech Academy of Sciences, 277 21 Lib
chov, Czech Republic
Department of Obstetrics & Gynecology,6 University of Missouri-Columbia, Columbia, Missouri 65211
ABSTRACT
The resumption of oocyte meiosis in mammals encompasses the landmark event of oocyte germinal vesicle (GV) breakdown (GVBD), accompanied by the modification of cell-to-cell communication and adhesion between the oocyte and surrounding cumulus cells. The concomitant cumulus expansion relies on microfilament-cytoskeletal remodeling and extracellular matrix (ECM) deposition. We hypothesized that this multifaceted remodeling event requires substrate-specific proteolysis by the ubiquitin-proteasome pathway (UPP). We evaluated meiotic progression, cytoskeletal dynamics, and the production of cumulus ECM in porcine cumulus-oocyte complexes (COCs) cultured with or without 10–200 µM MG132, a specific proteasomal inhibitor, for the first 22 h of in vitro maturation, followed by 22 h of culture with or without MG132. Treatment with 10 µM MG132 arrested 28.4% of oocytes in GV stage (vs. 1.3% in control), 43.1% in prometaphase I, and 16.2% in metaphase I, whereas 83.7% of control ova reached metaphase II (0% of MG132 reached metaphase II). The proportion of GV-stage ova increased progressively to >90% with increased concentration of MG132 (20–200 µM). Furthermore, MG132 blocked the extrusion of the first polar body and degradation of F-actin-rich transzonal projections (TZP) interconnecting cumulus cells with the oocyte. The microfilament disruptor cytochalasin E (CE) prevented cumulus expansion but accelerated the breakdown of TZPs. Ova treated with a combination of 10 µM MG132 and 10 µM CE underwent GVBD, despite the inhibition of proteasomal activity. However, 90.0% of cumulus-free ova treated with 10 µM MG132 remained in GV stage, compared with 16.7% GV ova in control. Cumulus expansion, retention of hyaluronic acid, and the deposition of cumulus ECM relying on the covalent transfer of heavy chains of inter-alpha trypsin inhibitor (I
I) were also inhibited by MG132. Cumulus expansion in control COCs was accompanied by the degradation of ubiquitin-C-terminal hydrolase L3, an important regulator of UPP. RAC1, a UPP-controlled regulator of actin polymerization was maintained at steady levels throughout cumulus expansion. We conclude that proteasomal proteolysis has multiple functions in the progression of oocyte meiosis beyond GV and metaphase I stage, polar body extrusion, and cumulus expansion.
cumulus expansion, FSH, germinal vesicle breakdown, LH, MG132, oocyte, oocyte maturation, porcine, proteasome, ubiquitin
Before the preovulatory gonadotropin surge, the mammalian oocytes arrested at the dictyate stage of meiosis I are surrounded by the cumulus oophorus, a specialized population of cumulus-granulosa cells in the antral ovarian follicle. Cumulus oophorus plays important roles in oogenesis, such as keeping the oocyte under meiotic arrest, inducting meiotic resumption, and supporting nuclear and cytoplasmic maturation of the oocyte [1]. The cumulus cells synthesize cAMP, a meiosis inhibitory factor that is the major survival factor in the antral follicle [2]. The resumption of oocyte meiosis is accompanied by the disappearance of the zona pellucida-spanning, actin microfilament-rich transzonal projections (TZPs) connecting the cumulus cells with the oocyte. This event is followed by the alteration of gap junctional communication in the cumulus oophorus, extensive rearrangement of actin microfilaments in the cumulus cells, and deposition of the extracellular matrix (ECM) parallel to cumulus expansion [3, 4]. Consequently, the communication between cumulus cells and the oocyte is altered in an expanded cumulus [5–7]. Metabolic uncoupling within the expanding cumulus-oocyte complexes (COCs) is believed to cause the decline of ooplasmic cAMP levels and lead to activation of CDC2/CDK1 kinase and mitogen-activated protein (MAP) kinase in the oocyte [8]. Subsequently, the surrounding mass of cumulus oophorus starts to expand because of massive hyaluronan synthesis [9] and extensive cytoskeletal rearrangement within the cumulus oophorus [4, 10–12]. The actin-rich TZPs are thought to be retracted and maintain fewer connections with the oocyte after cumulus expansion [13], but the actual mechanism altering the TZPs upon gonadotropin surge is not known.
Cumulus expansion in rodents [14–16] occurs after germinal vesicle breakdown (GVBD), but in the porcine model, partially expanded cumuli are observed prior to GVBD onset [17, 18]. In an in vitro culture system, cumulus expansion and metabolic uncoupling can be induced by FSH [19, 20] as well as by the adenyl-cyclase-activating drug forskolin [18], causing an increase in cAMP concentration in the cumulus cells [6, 15, 20]. Normal cumulus expansion in vitro is conducive to increased maturation rates [21] and higher rates of male pronuclear formation after sperm penetration [22]. Cumulus expansion is the result of interplay between microfilament-cytoskeletal rearrangement and increased synthesis of ECM composed mainly of hyaluronic acid (HA), glycosaminoglycans, and glycoproteins [5, 15, 23–25]. The FSH-induced synthesis of HA can be measured by the incorporation of [3H]-glucosamine into HA either in complexes alone (retained HA) or in medium plus complexes (total HA) [26]. The heavy chains (HCs) of serum derived from inter-alpha trypsin inhibitor (II) molecules become covalently linked to HA during mouse cumulus expansion in vivo [27, 28]. The covalent transfer of II to HA is catalyzed by the factors synthesized within the ovarian follicle [29–31]. The molecules of II can freely cross the blood-follicle barrier. Consequently, follicular fluid collected at any stage of folliculogenesis can be used to replace serum during porcine oocyte maturation in vitro [32].
Ubiquitin is a highly conserved protein of 76 amino acid residues that binds covalently in a tandem fashion to substrate proteins through a multistep enzymatic pathway [33]. The resultant polyubiquitin chain-tagged substrates are recognized by the 26S proteasome, a multicatalytic, substrate-specific protease composed of the barrel-shaped 20S proteasomal proteolytic core capped with one or two 19S regulatory complexes [34]. After the 19S regulatory complex removes the polyubiquitin chain, the substrate protein is translocated to 20S proteasomal core, where it is degraded into small peptides [35]. The rate of proteasomal proteolysis is controlled by deubiquitinating enzymes, including the ubiquitin-C-terminal hydrolases (UCHs) and ubiquitin-specific proteases [36]. Several groups reported the role of the ubiquitin-proteasome pathway (UPP) in the invertebrate oocyte maturation. Cyclins, CDK-inhibitor proteins, and c-Mos gene product are ubiquitinated and degraded by the 26S proteasome during the M phase of the cell cycle [37]. Takagi Sawada et al. [38] blocked starfish oocyte maturation by the addition of MG115, a specific proteasomal inhibitor, and proposed that the proteasomal activity triggers the activation of pre-MPF through the dephosphorylation of CDC2 kinase in response to the hormonal stimulus. Recently, the metaphase-anaphase transition was blocked effectively by MG132, a highly specific and reversible proteasomal inhibitor, during oocyte maturation in several mammalian species [39–42]. The inhibition of metaphase-anaphase transition by MG132 is consistent with the requirement of proteasomal activity for cyclin degradation at this stage of the cell cycle [43]. The present study has examined the effect of MG132 on early (GVBD) and late events (cytokinesis/polar body extrusion) of oocyte meiotic maturation and cumulus expansion of porcine COCs, and it investigated the interplay between proteasomal activities, microfilament dynamics, and ECM synthesis during in vitro maturation (IVM).
Collection and IVM of Porcine Oocytes
Ovaries were collected from prepubertal gilts at a local slaughterhouse and transported to the laboratory in a warm box (25°C–30°C). Cumulus-oocyte complexes were aspirated from antral follicles (3–6 mm in diameter) using an 18-gauge needle attached to a 10-ml disposable syringe. The COCs were washed three times in HEPES-buffered Tyrode lactate (TL-HEPES-PVA) medium containing 0.1% (w/v) polyvinyl alcohol (PVA) and three times with the maturation medium [44]. A total of 50 COCs were transferred to 500 µl of the maturation medium that had been covered with mineral oil in a four-well multidish (Nunc, Roskilde, Denmark) and equilibrated at 38.5°C, 5% CO2 in air. The medium used for oocyte maturation was tissue culture medium 199 (TCM199; Gibco, Grand Island, NY) supplemented with 0.1% PVA, 3.05 mM D-glucose, 0.91 mM sodium pyruvate, 0.57 mM cysteine, 0.5 µg/ml LH (L 5269; Sigma), 0.5 µg/ml FSH (F 2293; Sigma), 10 ng/ml epidermal growth factor (E 4127; Sigma), 10% porcine follicular fluid, 75 µg/ml penicillin G, and 50 µg/ml streptomycin sulfate. After 22 h of culture, oocytes were cultured without LH and FSH for 22 h at 38.5°C, 5% CO2 in the air. Reversible proteasomal inhibitor MG132 (Biomol Research Laboratories Inc., Plymouth Meeting, PA) was dissolved in ethanol to prepare the stock solution. All chemicals, including MG132 and cytochalasin E (CE), were added to IVM medium at the time of oocyte addition immediately prior to IVM.
Unless otherwise mentioned, all chemicals used in this study were purchased from Sigma Chemical Co. (St. Louis, MO).
Immunofluorescence and Epifluorescence Microscopy
Intact cumulus cell-enclosed oocytes and cumulus cell-free oocytes were fixed before IVM or after 44 h of IVM. To remove cumulus cells, oocytes were incubated with 0.1% hyaluronidase in TL-HEPES-PVA medium and were washed three times with TL-HEPES-PVA medium. Oocytes were transferred into the first well of a nine-well Pyrex brand glass plate (Fisher Scientific, Brightwaters, NY) filled with 400 µl of 37°C warm PBS, and 100 µl of 10% formaldehyde was slowly added to bring the final concentration of formaldehyde to 2% [45]. After 40 min of fixation at room temperature, oocytes were washed in two wells of PBS, fixative was removed, and the plates were wrapped in Handi-Wrap plastic wrap and stored for 1–7 days at 4°C. Prior to fluorescence processing, oocytes were permeabilized in PBS with 0.1% Triton X-100 at room temperature for 40 min. Samples were incubated with Alexa Fluor 488-phalloidin (1:50 dilution; Molecular Probes, Eugene, OR) and 2.5 µg/ml 4',6'-diamidino-2-phenylindole (DAPI; Molecular Probes) or 10 µg/ml propidium iodide (PI; DNA stains; Sigma) for 40 min to observe microfilament organizations.
To examine configurations of nuclear envelope and proteasomes at the germinal vesicle (GV) and GVBD stages of oocyte maturation, the formaldehyde-fixed oocytes were blocked for 25 min in 0.1 M PBS containing 5% normal goat serum and 0.1% Triton X-100 after permeabilization. Oocytes were incubated for 40 min with mAb 414, a mouse immunoglobulin G (IgG) against nuclear pore complex (1:200 dilution; BabCo/Covance, Berkeley, CA) and a rabbit polyclonal antibody against 20S proteasomal core recognizing multiple -type and β-type core subunits (1:200 dilution; Biomol). After washing, oocytes were incubated for 40 min with goat-anti-mouse (GAM)-IgG-fluorescein isothiocyanate to detect mAb 414, and goat-anti-rabbit (GAR)-TRITC to detect anti-20S proteasome /β subunit antibody (1:80 dilutions, respectively; Zymed Inc., San Francisco, CA). DNA was stained with DAPI (2.5 µg/ml). Image acquisition was performed using a Nikon Eclipse 800 microscope (Nikon Instruments Inc., Melville, NY) with a Cool Snap camera (Roper Scientific, Tucson, AZ) and MetaMorph software (Universal Imaging Corp., Downingtown, PA) and an Olympus Fluoview confocal laser scanning microscope. Data were archived on CD-R compact discs and edited using Adobe Photoshop 7.0 (Adobe Systems, Mountain View, CA).
Groups of 30 COCs isolated before (0 h) or after (22 h) in vitro culture in the presence of FSH/LH in medium with or without 100 µM MG132 were prepared for Western blot analysis by a procedure described previously [32]. Control in vivo COCs were obtained as described [32]. Shortly, COCs were washed in PBS containing a cocktail of protease inhibitors (PBS-PI) (Complete Mini; Boehringer Mannheim, Mannheim, Germany) and digested with 1 IU Streptomyces hyaluronidase (Merck-Calbiochem, Prague, Czech Republic) at 37°C for 1 h. The digested samples were centrifuged at 300 x g for 7 min to separate the matrix extract from the cell pellet. Reducing Laemmli buffer then was added to each fraction, and all samples were boiled at 99°C for 4 min, separated in 7.5% acrylamide/SDS gels, and transferred to Hypobond-P membranes (Amersham-Pharmacia, Prague, Czech Republic). Membranes were blocked with 5% nonfat dry milk in Tris-buffer solution (TBS) for 2 h at room temperature. Primary and secondary antibodies were diluted in TBS containing 5% BSA and 0.05% Tween 20. After electrophoresis, the proteins were detected with rabbit anti-human II (1:2000 dilution; DAKO, Carpenteria, CA) and anti-actin/ACTB (1:500 dilution; Sigma-Aldrich, Prague, Czech Republic) antibodies, followed by a horseradish peroxidase-linked F(ab')2 fragment of the anti-rabbit IgG, as a secondary antibody (1:10 000 dilution; Amersham-Biosciences) for 1 h at room temperature. RAC1 GTPase was detected using the mouse IgG generated against 21-kDa human RAC1 (Becton Dickinson, BD Transduction Laboratories). Ubiquitin C-terminal hydrolase L3 (UCHL3) was detected by the affinity-purified rabbit serum raised against a synthetic peptide corresponding to a C-terminal domain of human UCHL3 (MBL, Woburn, MA). The protein bands were detected by enhanced chemiluminiscence (Amersham-Pharmacia).
Hyaluronic Acid Synthesis Assay
Groups of 10 COCs were cultured for 23 h in 100 µl culture medium in the presence of FSH/LH and 2.5 µCi of [3H]-glucosamine hydrochloride (MGP, Zlín, Czech Republic) with or without 100 µM MG132. Synthesis of hyaluronic acid (HA) was measured using a procedure described previously [26]. Briefly, the cultures were terminated by adding 10 µl of a solution containing 50 mg/ml pronase (Sigma-Aldrich) and Triton X-100 in 0.2 M Tris buffer pH 7.8. The samples were incubated for 1 h at 38.5°C and then transferred onto Whatman 3MM filter paper circles. The circles were air dried and then washed through three changes of solution containing 0.5% cetylpyridinium chloride and 10 mM nonradioactive glucosamine hydrochloride (Sigma-Aldrich). The circles were dried once again, and radioactivity was measured using a liquid scintillation counter. Synthesis of HA was measured either in medium plus complexes (total HA) or within the complexes alone (retained HA).
Analysis of variance (ANOVA) was performed using the general linear model (GLM) procedure of the Statistical Analysis System (Cary, NC). The Duncan multiple-range test was used to compare mean value of individual treatments. The F-value was considered significant at P < 0.05.
MG132 Prevents Gonadotropin-Induced Degradation of TZPs and Cytoskeletal Rearrangements Within Cumulus Oophorus
Confocal microscopy was used to evaluate the TZPs and microfilament distribution at different stages of oocyte maturation with and without proteasomal inhibitors. Immature porcine oocytes were surrounded by a compact mass of round cumulus cells in the freshly isolated COCs. The innermost layer of cumulus/corona radiate cells was connected to oocyte surface through numerous F-actin-rich TZPs (Fig. 1A). Few TZPs remained on the ZP after IVM (Fig. 1B). Addition of 10 µM MG132 in the IVM medium for 44 h prevented cumulus expansion and breakdown of TZPs (Fig. 1C). When the COCs were cultured in the presence of MG132 for 22 h, and then cultured without MG132 for an additional 22 h, TZPs remained lodged in the ZP and connected to oocyte surface (Fig. 1D). However, the extracellular spaces between cumulus cells became enlarged at the periphery of COCs and bundles of filamentous actin (F-actin) typical of early stages of cumulus expansion [11, 12] emerged (Fig. 1D). In the COCs cultured in the absence of MG132 for 22 h and subsequently cultured with MG132 for 22 h, cumulus masses already expanded under the influence of FSH/LH stimulation within the first 22 h of culture, and few TZPs remained embedded in the ZP (Fig. 1E).
|
MG132 Prevents Cumulus Expansion and Decelerates Meiotic Maturation
Progression of oocyte meiosis was evaluated by conventional light microscopy of oocytes stained with DNA stain DAPI. GV retention was higher in the presence of MG132 for 44 h than it was after MG132 addition into the first or second 22 h of IVM only. Most oocytes became arrested in prometaphase I or metaphase I when COCs were cultured with MG132 for the whole 44 h. In the absence of MG132, 83.7% of ova reached metaphase II after 44 h of IVM (Fig. 2). The extent of cumulus expansion was evaluated under a dissecting microscope at 0, 22, and 44 h of IVM with or without MG132 and showed complete inhibition in the presence of 10 µM MG132 present in the IVM medium for the whole 44-h culture period (Supplemental Data Figure 1 available online at www.biolreprod.org).
|
Concentration-Dependent Inhibition of GVBD and Metaphase-Anaphase Transition by MG132
Exposure to progressively increasing concentrations of MG132 caused a dose-dependent delay of GVBD and meiosis in the ova from in vitro-matured COCs (Fig. 3A). The rates of ova with intact GV after 44 h of IVM were significantly (P < 0.05) higher in the presence of 150 and 200 µM MG132 than in the presence of 20–100 µM MG132. The rates of prometaphase I, metaphase I, anaphase I, telophase I, and metaphase II decreased proportionally with increased concentrations of MG132 (Fig. 3A). The arrest of oocyte meiosis at metaphase-anaphase transition was expected because of the known requirement of proteasomal activity for the degradation of cyclin B at this stage of meiosis. In a concomitantly cultured control group, porcine COCs were cultured with different concentrations of ethanol (a vehicle for MG132; Fig. 3B).
|
To examine the process of GVBD and accompanying redistribution of proteasomes in the maturing oocytes, COCs cultured in the presence or absence of 10 µM MG132 were fixed and incubated with a mixture of mAb 414, a mouse IgG against nuclear pore complex (NPC; marker of intact nuclear envelope signifying intact GV), and a rabbit polyclonal antibody to 20S proteasomal core recognizing multiple -type and β-type proteasomal core subunits. Gradual disappearance of NPCs was observed prior to chromosome condensation and prior to the disappearance of proteasomes from the GV during GVBD process in the control oocyte not exposed to MG132 (Fig. 3C, a–d). Both the proteasomes and NPCs were completely absent from metaphase I stage oocytes (Fig. 3C, e). The rate of GV retention in the presence of 100 µM MG132 for 44 h (Fig. 4A) was comparable to that in experiments with 150–200 MG132, and higher than when MG132 was added only during the first or second 22 h of IVM. The progression of meiosis was arrested in prometaphase I or metaphase I when COCs were cultured for 44 h with MG132 or when the COCs were cultured for the first 22 h with MG132 (Fig. 4A). Interestingly, a majority of the ova treated with 100 µM MG132 failed to extrude first polar body and reach metaphase II, even if MG132 was present only during the second half of the IVM interval (Fig. 4A). Most oocytes cultured with 100 µM MG132 for 44 h (Fig. 4B) displayed intact nuclear envelopes with abundant NPCs and a high content of proteasomes similar to control GV-stage oocytes.
|
To examine whether the treatment with MG132 treatment could be used to extend the period of IVM, COCs were cultured in the presence of 10 µM MG132 for 44 h in TCM199 medium and then for 44 h under control IVM conditions (22 h in IVM medium with hormones followed by 22 h in medium without hormones). Despite doubling the time of in vitro culture, 50.1% of metaphase II oocytes were observed after 88 h of IVM (Supplemental Data Figure 2, available online at www.biolreprod.org).
Inhibition of Oocyte Meiosis by MG132 in Relation to Structural Integrity of TZP and Presence of Cumulus Cells
The above experiments did not address whether the inhibition of proteasomal activity by MG132 prevents GVBD directly by preventing the proteolysis of cell cycle-controlling proteins in the oocyte, or indirectly by blocking the destruction of TZP and cumulus-oocyte uncoupling. We devised a combined treatment with 10 µM MG132 and 10 µM CE, a microfilament disruptor that could induce actin microfilament depolymerization in the TZPs, leading to TZP breakdown independent of hormonal stimulation of COCs. First, we cultured the COCs in the presence or absence of CE the first 22 h of maturation in FSH/LH-supplemented medium, followed by 22 h of culture without hormones and with CE. These treatments did not significantly alter GVBD and the progression to metaphase I, whereas few ova reached metaphase II due to the cytochalasin-induced block of karyokinesis and cytokinesis (first polar body extrusion; Fig. 5A). While cumulus expansion was blocked by CE after 44 h of culture, the treatment coincided with the disruption of TZPs from ZP and oocyte surface (Fig. 5B, a–a''). When the COCs were cultured in the presence of CE in IVM medium for 22 h and then cultured without CE for 22 h, the cumulus cells were connected loosely and few TZPs were present in the ZP (Fig. 5B, b–b''). In COCs cultured in the absence of CE for 22 h and then cultured with CE for next 22 h, the cumulus cells had already expanded within the first 22 h of culture, TZPs were disrupted, and weak staining of microfilaments was observed on the oocyte surface (Fig. 5B, c–c'') compared with the microfilament staining of TZPs and oocyte cortex in the control GV-stage oocytes (Fig. 5B, d–d'').
|
The combined treatment with 10 µM MG132 and 10 µM CE for 22 h did not inhibit the progression of meiosis up to metaphase I, a stage at which MG132 prevents the proteasomal degradation of cyclin B, a prelude to metaphase-anaphase transition [39]. Consequently, high rates of metaphase I (64.9%) ova were recorded, but the rates of GV stage ova (Fig. 5C) were similar to control, untreated, expanded COCs and COCs treated with CE alone. While cumulus expansion was blocked by the combination of MG132 and CE, the treatment with CE caused microfilament depolymerization and disruption of TZPs (Fig. 5D). Even though a low concentration of MG132 (10 µM) was not sufficient to maintain intact GV in the cumulus-enclosed ova, 90% of ova remained in GV stage when cumulus-free ova were subjected to 10 µM MG132 treatment for first 22 h of oocyte maturation (Fig. 5E). Thus, even a low concentration of MG132 prevented GVBD in the absence of cumulus cells.
MG132 Treatment Prevents the Synthesis and Deposition of Cumulus ECM
More than 90% of porcine COCs isolated from follicles 3 mm or larger undergo expansion during 24 h of culture in the presence of FSH [12]. In the present study, the presence of FSH/LH induced cumulus cells to expand within 24 h and synthesize HA (Fig. 6A). In the control FSH/LH-stimulated expanded complexes, the amount of HA retained within the complexes represented approximately 55% of the total amount of HA accumulated in the culture well (combined HA content of COCs and culture medium). Nevertheless, in the presence of FSH/LH and 100 µM MG132, all COCs remained compact, and total HA accumulation was reduced by 48% (P < 0.001; Fig. 6A). In addition, the retention of HA within the complexes decreased to <23% of the amount retained within the complexes cultured in the absence of MG132 (P < 0.05; Fig. 6A).
|
We next used immunoblotting with anti-II antibody to evaluate whether the heavy chains of serum-derived II-related molecules are covalently linked to HA in the presence of MG132, the covalent binding being required for ECM deposition. The nonexpanded control COCs (0 h; Fig. 6B) and COCs cultured for 22 h with MG132 (both digested with hyaluronidase) did not contain proteins reactive with II antibody. In contrast, matrix extracts of expanded COCs after culture without MG132 (22 h; Fig. 6B) or matrix extracts of in vivo-expanded COCs isolated from antral follicles of gilts treated sequentially with eCG and hCG (Fig. 6B), contained immunoreactive bands of approximately 120 kDa, 130 kDa, and 220 kDa. These respective bands were demonstrated recently in SDS-PAGE of porcine serum [32], and they correspond to II (bikunin plus HC1 and HC2), PI (bikunin plus HC3), and ILI (bikunin plus HC2), respectively [46, 47]. Two additional bands of approximately 75–85 kDa were detected in the matrix extracts of both types of COCs after digestion with hyaluronidase. Two other minor bands in the hyaluronidase matrix extract migrating at about 160–170 kDa likely represent dimers of the different types of heavy chains linked to the same HA strand at so close a proximity that the enzymatic digestion of the interposed HA fragment is prevented, as previously suggested by Chen et al. [29]. These observations confirm our previous results [32] that HCs from II-related molecules identified in the porcine serum were transferred and covalently linked to HA during porcine cumulus expansion in vivo and in vitro. However, cumulus expansion and the covalent transfer of HCs from II to HA were not observed in the presence of MG132 (Fig. 6). Immunoblotting with anti-actin antibody was used for normalization of II protein load.
MG132 Treatment Does Not Alter Microfilament Dynamics in Cumulus Cell Cytoplasm
The above data demonstrated that MG132 treatment prevented cumulus expansion and deposition of cumulus ECM. Apart from maintenance of F-actin-rich TZPs, MG132 treatment did not prevent microfilament dynamics within cumulus cell cytoplasm. We examined whether MG132 could affect the cumulus expansion by reducing the content of RAC1 GTPase, a major regulator of F-actin polymerization known to be turned over by UPP [48]. However, we observed neither major changes in RAC1 content in the course of cumulus expansion nor RAC1 accumulation in the presence of MG132 (Fig. 7). By epifluorescence microscopy, RAC1 colocalized with F-actin inside cumulus cells before and after cumulus expansion (Supplemental Data Figure 3, available online at www.biolreprod.org). Total content of β-actin (ACTB) was stable during cumulus expansion but increased in COCs matured in the presence of MG132 (Fig. 7B). In the same set of experiments, we also evaluated the cellular content of UCHL3. The UCHL3 and related deubiquitinating enzymes are believed to regulate the rate of proteasomal proteolysis [36]. Relative to ACTB expression, the content of UCHL3 was greatly diminished after 44 h of COC culture (Fig. 7A). This reduction of UCHL3 content seemed to occur during the last 12 h of culture and was not prevented by MG132 (Fig. 7B). It is possible that a feedback mechanism that signals the increased need for proteasomal degradation of proteins accumulating in the presence of a proteasomal core inhibitor also suppresses the expression of deubiquitinating enzymes that could be limiting factors of proteasomal degradation.
|
The involvement of the UPP in oocyte maturation has been suggested previously in various species and taxa. However, the present study offers novel insight into mechanisms by which proteasomal proteolysis contributes to both the regulation of the oocyte cell cycle and to remodeling of its surrounding cumulus mass. In the present study, 10 µM MG132 maintained the percentage of GV-intact ova near 30% and the percentage of metaphase I oocytes above 40% in cumulus-enclosed porcine ova. However, when 10 µM MG132 was used in conjunction with 10 µM CE, an agent capable of disrupting TZPs of cumulus cells, the percentage of GV oocytes was reduced to 10%, and the percentage of metaphase I oocytes was increased proportionally, reaching 
70%. The arrest at metaphase I in both cases can be explained by the inhibitory effect of MG132 on proteasome-dependent cyclin degradation, preventing metaphase-anaphase transition [39, 43]. Our finding of low incidence of GV-intact oocytes in the presence of MG132 in conjunction with CE treatment suggests that the depolymerization of actin-rich TZPs by CE overruled the inhibitory effect of 10 µM MG132 on GVBD in the cumulus-enclosed ova. Metaphase-anaphase transition thus appears to be more sensitive to proteasomal inhibitors than GVBD. While high concentrations of MG132 (20–200 µM) prevented GVBD in a concentration-dependent manner, almost complete GV retention (>90%) was observed when cumulus-free ova were exposed to 10 µM MG132. Therefore, a higher level of proteasomal activity may be necessary to induce GVBD process in the absence of cumulus cell-produced, meiosis-stimulating factors. Possible substrates of UPP during GVBD could include cell cycle regulators, nuclear lamins, and histones. The inhibitory action of 10 µM MG132 on GVBD could be counteracted by the cumulus cell-produced, freely diffusible, meiosis-stimulating agents that may not require TZPs for their transport into ooplasm. At the same time, breakdown of TZPs with CE could stop the influx of meiosis-inhibiting substances through cumulus-to-oocyte gap junctions (GJs), explaining why MG132 at low concentrations was not sufficient to maintain intact GVs in presence of cumulus cells and CE. These observations suggest that the maintenance of TZPs supports oocyte meiotic arrest, possibly by the means of sustained high levels of cAMP. According to this hypothesis, cAMP is produced by the cumulus cells and transported to a meiotically quiescent oocyte via TZPs terminated by cumulus cell-to-oocyte GJs, permeant to cAMP [49]. In a recent study, carbonoxolone, a selective inhibitor of gap junctional communication, promoted meiotic maturation of follicle-enclosed rat oocytes in vitro and caused a significant decrease in intraoocyte concentration of cAMP [50]. Our thesis that the gonadotropin-induced breakdown of TZPs depends upon proteasomal degradation is further supported by the observation that MG132 alone sufficiently protected TZPs from breakdown in the presence of gonadotropins. Consequently, the maintenance of TZPs in the presence of MG132 reduced the rate of GVBD, and this MG132-induced inhibition was cancelled by microfilament depolymerization with CE. Traditionally, the uncoupling of cumulus-to-oocyte communication at the onset of oocyte maturation has been interpreted as the withdrawal of TZPs from zona pellucida [51, 52]. Our data suggest that this process may involve regulated, substrate-specific proteasomal proteolysis of proteins associated with actin microfilaments within TZPs. For example, ubiquitination and proteasomal degradation of actin-associated RAC1 GTPase occurs during early stages of epithelial cell scattering, an event that requires microfilament depolymerization in focal adhesions [48]. In our studies, the levels of RAC1 appeared relatively stable during cumulus expansion, and F-actin was observed with and without MG132. This observation alone does not rule out that a small subset of RAC1 molecules associated with TZPs could be degraded locally at the time of meiotic resumption. While MG132 did not cause accumulation of RAC1, it did induce a substantial increase in content of total ACTB (monomeric and F-actin). It is possible that actin in the TZPs is directly ubiquitinated and degraded by proteasomes during meiotic maturation. Ubiquitination and proteasomal degradation of actin and actin-associated proteins has been suggested as a possible cause of skeletal muscle atrophy [53]. While high levels of RAC1 may be important for the sustenance of cumulus mass before and during expansion, degradation of UCHL3, which is an important rate-limiting factor of proteasomal proteolysis, may reflect an increased need for proteasomal proteolysis during cumulus expansion.
The exact function of the 26S proteasome in cumulus expansion remains to be elucidated. In the present study, proteasomal inhibitor MG132 prevented cumulus expansion, and cumulus cells remained packed around the oocytes after 44 h of culture in a gonadotropin-supplemented medium. During preovulatory cumulus expansion, cumulus cells synthesize a large amount of HA that is organized in a highly hydrated, mucoelastic matrix. Figure 6 shows that 100 µM MG132 affected the organization of cumulus ECM and prevented HC translocation to HA in the COCs stimulated with gonadotropins. Furthermore, the FSH/LH-stimulated synthesis of HA was significantly reduced by MG132. Lower concentrations of MG132 (e.g., 10 µM; data not shown) did not reduce HA synthesis measured as total synthesis, but they significantly reduced the retention of synthesized HA within the COCs and still prevented cumulus expansion. In addition to HA synthesis, serum components belonging to the II family play a key role in porcine cumulus ECM deposition [32]. The formation of HA-enriched cumulus ECM is not supported by II-immunodepleted serum while it occurs in the presence of purified II molecules [27]. In the present study, the effect of MG132 on covalent linkage of heavy chains of II to HA, critical for cross-linking of HA strands and stabilization of cumulus matrix, was studied. Our Western blotting results clearly demonstrate that heavy chains of serum-derived II-related molecules become covalently linked to HA in pig COCs expanded in vitro only in the absence of MG132. This observation indicates that proteasomal proteolysis is essential for cumulus expansion and assembly of cumulus ECM. To our knowledge, this is the first demonstration of the requirement of proteasomal proteolysis during the process of cumulus expansion.
The results of a combined treatment with MG132 and CE suggest that cytoskeletal reorganization precedes ECM deposition in the signaling cascade of cumulus expansion. A similar phenomenon was observed when COCs were cultured with progesterone receptor (PR) antagonist RU486. Cumulus cells of COCs treated with RU486 remained tightly packed around the oocytes, with little visible expansion [54]. Other proteases besides 26S proteasome could also be involved in the reorganization of cumulus ECM during cumulus expansion. ADAMTS-1, a member of the disintegrin and metaloproteinase family of proteases, has been shown to degrade members of the lectican family of proteoglycans [55–57]. Production of ADAMTS-1 protein is induced by LH in granulosa cells via progesterone- and PR-dependent pathways [58–60] and cleaves versican (VI variant) in human aorta in vivo, releasing a 70-kDa N-terminal fragment [61]. In the expanded COCs, a 70-kDa N-terminal VI fragment appears to be generated by an ADAMTS-1 proteaselike activity [60]. Since the cleaved N-terminal domain of versican binds the HA-rich matrix, the cleaved versican may serve to stabilize cumulus expansion of COC. The ADAMTS-1 mRNA in cumulus cells of COCs is dramatically upregulated in response to gonadotropins [54] and downregulated by RU486. While the cumulus expansion is compromised in the RU486-treated COCs, the treated cumulus cells still show patterns of HAS-2, TNFAIP6, versican, and ADAMTS-4 expression similar to those of control, expanded COCs [54]. This result is reminiscent of apparently normal cumulus HA synthesis in the presence of expansion-blocking 10 µM MG132 in our study.
Proteasomes may also participate in protein turnover inside the cells of expanding COCs. One example is the ubiquitin-proteasome dependent turnover of steroidogenic acute regulatory protein (STAR) in granulosa cells. STAR is synthesized as a 37-kDa preprotein in the cytoplasm of steroidogenic cells and is then rapidly transported into mitochondria, where it is cleaved, generating 32-kDa and 30-kDa proteins [62, 63]. STAR plays a key role in rapid modulation of steroidogenesis in the ovary [64] and testis [65]. Rat granulosa cells and immortalized human granulosa cells treated with MG132 accumulate STAR [66]. The MG132-induced STAR accumulation is further augmented in the presence of forskolin, LH, or FSH [66]. The cytosolic 37-kDa STAR preprotein, which is responsible for the steroidogenic activity of STAR, is the primary proteasomal substrate, and the inhibition of its degradation by MG132 causes the upregulation of progesterone production [66]. Proteasomal activity may also be important for quality control during the synthesis and secretion of ECM glycoprotein by cumulus cells. Ubiquitination and proteasomal degradation of improperly glycosylated, unassembled, and misfolded glycoproteins is a central element of the endoplasmic reticulum-associated ERAD system for protein quality control in the secretory pathway [67]. Accumulation of aberrant, misfolded ECM proteins in the presence of MG132 could prevent their interactions with HA after secretion.
Together, our data suggest that proteasomal activity has multiple functions in oocyte meiosis, in cumulus expansion, in the synthesis and processing of cumulus ECM, and in the modification of cumulus-oocyte communication. Maintenance of TZPs of cumulus cells may be required for the maintenance of meiotic arrest in the cumulus-enclosed oocytes, and the gonadotropin-induced proteasomal degradation of cytoskeletal proteins in these projections may contribute to resumption of oocyte meiosis. In the ooplasm, proteasomal activity is required for metaphase-anaphase transition, as shown previously in mammals, but also for the process of GVBD.
ACKNOWLEDGMENTS
We thank Nicole Leitman for technical support, the laboratory of Dr. Randall Prather for collecting porcine ovaries, and Kathryn Craighead for clerical support and manuscript editing.
FOOTNOTES
1Supported by National Research Initiative (NRI) Competitive Grants #2002-35203-12237 and 2007-35203-18274 from the US Department of Agriculture (USDA) Cooperative State Research, Education, and Extension Service, USDA-NRI Discovery Award 2005, and seed funding from the Food for the 21st Century Program of the University of Missouri-Columbia. Y.-J.Y. and C.-S.P. were supported by grant R11-2002-100-00000-0 from ERC Program of Korea Science & Engineering Foundation (KOSEF). E.N. was supported by award 305/05/0960 from the Grant Agency of the Czech Republic. 
Correspondence: 2Peter Sutovsky, Associate Professor, University of Missouri-Columbia, S141 ASRC, 920 East Campus Dr., Columbia, MO 65211-5300. FAX: 573 884 5540; e-mail: SutovskyP{at}missouri.edu
Received: 8 March 2007.
First decision: 26 May 2007.
Accepted: 8 October 2007.
REFERENCES
| ||||||||||||
