Cyclooxygenase-2-derived Prostaglandin E2 Directs Oocyte ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 48, pp. 37117–37129, December 1, 2006 © 2006 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

Cyclooxygenase-2-derived Prostaglandin E2 Directs Oocyte Maturation by Differentially Influencing Multiple Signaling Pathways*□ S

Received for publication, August 25, 2006, and in revised form, September 27, 2006 Published, JBC Papers in Press, October 4, 2006, DOI 10.1074/jbc.M608202200

Toshifumi Takahashi‡1, Jason D. Morrow§, Haibin Wang‡2, and Sudhansu K. Dey‡§¶3 From the Departments of ‡Pediatrics, §Pharmacology, and ¶Cell and Developmental Biology, Division of Reproductive and Developmental Biology, Vanderbilt University Medical Center, Nashville, Tennessee 37232 The process of oocyte maturation, which impacts ovulation and fertilization, is complex and requires an integration of the endocrine, paracrine, juxtacrine, and autocrine signaling pathways. This process involves an intimate interaction between the oocyte and encircling cumulus cells within a follicle, a unique venue for somatic and germ cell communication. Cumulus cell expansion and resumption of meiosis with germinal vesicle breakdown are major events in oocyte maturation. Cyclooxygenase-2 (COX-2)derived prostaglandin E2 (PGE2) is a known critical mediator of oocyte maturation, but the diverse function of this lipid mediator in oocyte maturation, ovulation, and fertilization has not been fully appreciated. We show here that gonadotropins in coordination with PGE2 signaling via its cell surface G-protein-coupled EP2 and EP4 receptor subtypes direct cumulus cell expansion and survival and oocyte meiotic maturation by differentially impacting cAMPdependent protein kinase, MAPK, NF-␬B, and phosphatidylinositol 3-kinase/Akt pathways. This study is unique in the sense that it provides evidence for new site- and event-specific involvement of these signaling pathways under the influence of COX-2-derived PGE2 during the critical stages of this somatic-germ cell interaction, an absolute requirement for oocyte maturation.

Prostaglandins (PGs)4 are involved in various female reproductive functions, including ovulation, fertilization, and implantation (1, 2). The cyclooxygenase (COX) isozymes COX-1 and COX-2

* This work was supported in part by National Institutes of Health Grants, HD12304, HD33994, HD050315, and P01-CA-77839. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. 1 and 2. 1 Present address: Dept. of Obstetrics and Gynecology, Yamagata University School of Medicine, 2-2-2 Iidanishi, Yamagata 990-9585, Japan. 2 Recipient of the Solvay-Mortola research award from the Society for Gynecologic Investigation. 3 Recipient of Method to Extend Research in Time (MERIT) award from the NICHD, National Institutes of Health. To whom correspondence should be addressed. E-mail: [email protected]. 4 The abbreviations used are: PG, prostaglandin; GVBD, germinal vesicle breakdown; COX-2, cyclooxygenase-2; WT, wild type; MI, metaphase I; MII, metaphase II; HX, hypoxanthine; PI3K, phosphatidylinositol 3-kinase; COC, cumulus-oocyte complex; TRITC, tetramethylrhodamine isothiocyanate; PMSG, pregnant mare serum gonadotropin; hCG, human chorionic gonadotropin; PKA, cAMP-dependent protein kinase; MAPK, mitogen-activated protein kinase; ERK, extracellular signal-regulated kinase; TUNEL, terminal dUTP nick-end labeling; FSH, follicle-stimulating hormone; hFSH, human FSH; JNK, c-Jun N-terminal kinase; RT, reverse transcription.

DECEMBER 1, 2006 • VOLUME 281 • NUMBER 48

are the primary producers of PGs. Although COX-1 is considered to be constitutive, COX-2 is induced by inflammatory stimuli, including cytokines and growth factors (3). Gene targeting experiments in mice have revealed distinct functions of these isozymes. COX-1-deficient females are mostly fertile but show parturition defects. In contrast, Cox-2⫺/⫺ females are mostly infertile because of severely impaired ovulation, fertilization, and implantation (4). Although ovulation failure in Cox2⫺/⫺ mice is rescued by PGE2 supplementation (5), implantation defects are improved by prostacyclin (PGI2) working via peroxisome proliferator-activated receptor-␦ (6). Effects of PGE2 are normally mediated through G protein-coupled receptors EP1, EP2, EP3, and EP4 (7). Gene targeting studies showed that although EP2-deficient females have ovulation and fertilization defects similar to those in Cox-2⫺/⫺ females (8 –10), EP1- and EP3-deficient mice are apparently fertile (11). In contrast, most EP4⫺/⫺ pups die shortly after birth because of patent ductus arteriosus, precluding studies on reproductive phenotypes (12, 13). Oocytes are arrested at the germinal vesicle stage before the surge of pituitary luteinizing hormone. This surge induces meiotic resumption of oocytes acting via its receptors on theca or granulosa cells, because luteinizing hormone receptors are scant in cumulus cells (14). Although luteinizing hormone signaling in theca or granulosa cells is not clearly understood, recent findings show that epidermal growth factor-like growth factors, amphiregulin, epiregulin, and betacellulin, are induced by hCG, and these growth factors promote meiotic resumption and cumulus expansion by depositing cell matrix through expression of hyaluronan synthase-2, COX-2, and tumor necrosis factor-induced protein-6 (TSG-6) (15). Moreover, oocyte meiotic maturation is associated with cumulus expansion induced by gonadotropins (16). Cumulus expansion is important for ovulation, fertilization, and subsequent embryonic development (4, 5, 8). PGE2 induces oocyte meiotic maturation in cultured mouse ovaries (17), and indomethacin, a nonspecific COX inhibitor, attenuates gonadotropin-induced cumulus expansion and germinal vesicle breakdown (GVBD) of oocytes (18 –20). However, indomethacin, which blocked the release of PGs, failed to inhibit GVBD induced by gonadotropin-releasing hormone and angiotensin II (21, 22), suggesting complex roles of PGs in oocyte maturation. We have shown that ovulated oocytes from Cox-2⫺/⫺ mice on the C57BL/6J/129 background had very few first polar bodJOURNAL OF BIOLOGICAL CHEMISTRY

37117

Differential Signaling by PGE2 in Oocyte Maturation ies (4). Our more recent work, however, shows that Cox-2⫺/⫺ oocytes have normal development compared with wild-type oocytes when matured spontaneously in vitro (10). Moreover, normal GVBD is observed in Cox-2⫺/⫺ mice after superovulation (5). In contrast, oocytes with signs of fragmentation are frequently observed in Cox-2⫺/⫺ mice on CD1 background on day 2 of pregnancy, suggesting defective oocyte maturation, resulting in fertilization failure. In the absence of definitive roles of PGs in oocyte maturation, we initiated an in-depth investigation on the roles of PGs in oocyte maturation in vivo and in vitro using Cox-2⫺/⫺ mice on the CD1 background and the signaling pathways involved in this process.

EXPERIMENTAL PROCEDURES Reagents—Minimum essential medium and fetal bovine serum were purchased from Invitrogen. Human follicle-stimulating hormone (hFSH) was a gift from National Hormone and Peptide Program (NIDDK, National Institutes of Health, Bethesda, MD). PGE2, butaprost, sulprostone, 11-deoxy-prostaglandin E1 (11-deoxy-PGE1), prostaglandin A1 (PGA1), and AH6809 were purchased from Cayman Chemicals (Ann Arbor, MI). H-89, U0126, SB203580, and SP60012 were purchased from Calbiochem. LY294002 and antibodies for ERK1/2, phosphoERK1/2, Akt, and phospho-Akt were purchased from Cell Signaling Technologies (Danvers, MA). Antibodies for COX-2, EP1, EP2, EP3, and EP4 were purchased from Cayman Chemicals. Antibodies for actin were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibody for cleaved caspase-3 was purchased from Promega (Madison, WI). Antibody for Bcl-2 was purchased from GeneTex (San Antonio, TX). Fluorescein isothiocyanate-, rhodamine (TRITC)-, or peroxidase-conjugated secondary antibody was purchased from Jackson ImmunoResearch (West Grove, PA). Other reagents were obtained from Sigma unless otherwise indicated. Animals and Treatments—The disruption of the Cox-2 gene in mice was described previously (23). PCR of genomic DNA determined the genotypes (4). Adult female mice on the CD1 background (2– 6 months old) were used for all experiments. Generation of CD1 Cox-2⫺/⫺ mice was described previously (24). All mice used were housed in the Institutional Animal Care Facility according to National Institutes of Health and institutional guidelines for laboratory animals. To induce superovulation, mice received intraperitoneal injections of 5 IU of pregnant mare serum gonadotropin (PMSG) followed by injections of 5 IU of hCG 48 h later as described previously (10, 24). To examine whether PGE2 can rescue normal oocyte maturation in CD1 Cox-2⫺/⫺ mice, wild-type or Cox-2 mutant females received subcutaneous injections of vehicle (3% ethanol in saline) or PGE2 (40 ␮g/mouse) concurrent with hCG administration and followed by a second injection 4 h later (5). Ovulated cumulus-oocyte complexes were recovered from the oviduct 14 –15 h after hCG injection. Cumulus cells were removed by a brief hyaluronidase treatment, and the meiotic status was evaluated by nuclear staining. Culture of Cumulus-Oocyte Complex—Cumulus-oocyte complexes (COCs) were isolated by puncturing large antral follicles with a 26-gauge needle. In some cases, ovulated COCs were collected from the oviduct. For experiments with in vitro

37118 JOURNAL OF BIOLOGICAL CHEMISTRY

oocyte maturation, COCs were collected 46 – 47 h after PMSG injection. COCs with uniform compacted cumulus cells were used for in vitro culture. COCs were transferred into the medium containing 4 mM hypoxanthine (HX) to prevent spontaneous GVBD. The culture medium was bicarbonate-buffered minimum essential medium supplemented with 75 mg/liter penicillin G, 50 mg/liter streptomycin sulfate, and 3 mg/ml crystallized lyophilized bovine serum albumin (basal medium). To examine FSH-induced oocyte meiotic maturation, groups of 20 –30 COCs were cultured in 100-␮l droplets of basal medium containing 4 mM HX and 100 IU/liter hFSH covered with mineral oil in a humidified atmosphere of 5% CO2 in air at 37 °C. For cumulus expansion experiments, 1% fetal bovine serum instead of bovine serum albumin was added to the basal media. For assessment of oocyte meiotic status, cumulus cells were removed by hyaluronidase digestion, and denuded oocytes were fixed in 4% buffered formalin, mounted on a glass slide with glycerol/phosphate-buffered saline (1:1) solution containing 2 ␮g/ml Hoechst 33258, and examined under a fluorescence microscope. Oocyte meiotic status was classified as follows: germinal vesicle, MI, or MII. To assess cumulus expansion, images of COCs were captured under an inverted microscope, and the projected cumulus areas of COCs were calculated using the Image J Imaging System software version 1.3 (National Institutes of Health, Bethesda, MD). In some experiments, the degree of cumulus expansion was determined as 0 (no expansion) to 4 (complete expansion) (25). In Vitro Fertilization—To assess whether PGE2 can restore normal fertilizing capacity of Cox-2 null oocytes, we performed in vitro fertilization experiments in both wild-type and Cox-2 mutant mice receiving vehicle or PGE2 injections (5, 10). PGE2 was given subcutaneously at a dose of 40 ␮g/mouse concurrently with hCG injection followed by a second injection 4 h later. Ovulated COCs were collected from the oviduct after 14 –16 h of hCG injection, placed into droplets of 200 ␮l of HTF medium, and cultured in 5% CO2 in air at 37 °C until sperm insemination. Sperm were collected from the cauda epididymidis of mature wild-type mice and were preincubated for 2 h in 400 ␮l of HTF medium to allow capacitation. After capacitation, sperm were introduced into 200 ␮l droplets containing COCs. Six hours after insemination, COCs were transferred into droplets of 100 ␮l of HTF medium, and cumulus cells were removed by repeated pipetting. Fertilization was assessed by the appearance of two pronuclei in oocytes that were further cultured in KSOM medium for 5 days under 5% CO2 in air at 37 °C. The number of blastocysts developed was recorded at the end of culture. Prostaglandin Assays—Twenty COCs were cultured in 500-␮l droplets of the HX media in the presence or absence of hFSH, and culture media were collected at 3, 6, and 18 h posttreatment for PG analysis by gas chromatography/negative ion chemical ionization mass spectrometric assay (6). Immunofluorescence—Immunofluorescence was performed as described (26). Antibodies specific to COX-2 (1:300), TSG-6 (1:100), EP1 (1:100), EP2 (1:100), EP3 (1:100), EP4 (1:100), Akt (1:80), phospho-Akt (1:100), or cleaved caspase-3 (1:100) were VOLUME 281 • NUMBER 48 • DECEMBER 1, 2006

Differential Signaling by PGE2 in Oocyte Maturation Ptgerep4 (389 bp), 5⬘-CCTGGATTTACATCCTTCTCCG-3⬘ and 5⬘-ACTAACCTCATCCACCAACAGG3⬘. The reaction conditions for PCR were as follows: Ptgerep1, denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s, and extension at 72 °C for 30 s; Ptgerep2, denaturation at 95 °C for 30 s, annealing at 65 °C for 60 s, and extension at 72 °C for 60 s; Ptgerep3, denaturation at 95 °C for 30 s, annealing at 62 °C for 30 s, and extension at 72 °C for 30 s; Ptgerep4, denaturation at 95 °C for 30 s, annealing at 62 °C for 60 s, and extension at 72 °C for 60 s. PCRs were run for 35 cycles. Amplified products were separated by electrophoresis on 2% agarose gels and visualized by ethidium bromide staining. Apoptosis Detection—DNA fragmentation was detected by terminal deoxynucleotide transferase-mediated dUTP nick-end labeling (TUNEL) technique using DeadEndTM colorimetric or Fluorometric TUNEL Systems (Promega, Madison, WI; catalog numbers G7130 and G3250) according to the manufacturer’s instructions. Statistical Analysis—Each oocyte FIGURE 1. COX-2 deficiency compromises gonadotropin-induced oocyte maturation and cumulus maturation experiment was con⫺/⫺ expansion in vivo and in vitro. A, deferred meiotic maturation of Cox-2 oocytes in response to hCG in vivo. GVBD was scored for COCs isolated at various times after hCG injection. Numbers inside bars indicate the ducted at least three times independnumbers of oocytes with GVBD/total numbers of oocytes examined. B, impaired cumulus expansion in Cox- ently with 20 –30 oocytes per treat2⫺/⫺ mice in response to hCG in vivo. Projected areas of COCs were recorded at the indicated times after hCG injection. C, deferred meiotic maturation of Cox-2⫺/⫺ oocytes by FSH in culture. COCs collected 46 – 47 h after ment group. Differences between PMSG injection were cultured in HX medium containing 100 IU/liter FSH for 18 h. D, impaired cumulus expan- means were assessed by Student’s t sion of Cox-2⫺/⫺ COCs in response to FSH in culture. *, significantly different from WT mice (p ⬍ 0.01). test or were calculated by one-way analysis of variance followed by Fisher used. The images were captured using the Image J Imaging LSD post hoc test using StatView software (Abacus Concepts, System software (version 1.3) under an inverted fluorescence Berkeley, CA). Each experiment was repeated at least three times unless stated otherwise. Values are shown as mean ⫾ S.E. Data microscope. Immunoblotting—Thirty COCs or cumulus cells from 30 expressed as percentages were analyzed by using ␹2 tests. SignifiCOCs were lysed, and lysates were analyzed by immunoblot- cant differences are defined as p ⬍ 0.05. ting for TSG-6, total Akt, phospho-Akt (1:1000), Bcl-2, or actin as described (26). Protein bands were visualized using chemilu- RESULTS minescent detection reagents (Pierce) and quantitated using a Gonadotropin-induced Oocyte Maturation and Cumulus densitometer. Expansion Are Compromised in Cox-2⫺/⫺ Mice—Previous RNA Isolation and RT-PCR—Total RNA was extracted from observations of ovulation and fertilization defects in Cox-2⫺/⫺ cumulus cells using TRIzol reagents (Invitrogen) according to mice on the C57BL/6J/129 background suggested that COX-2the manufacturer’s instructions. Reverse transcription with oli- derived ovarian PGs are critical to these events (4, 10). Furthergo(dT) primers generated cDNAs from 5 ␮g of total RNA using more, significantly improved ovulation with frequently Superscript II. DNA amplification was carried out with a Taq observed fragmented eggs in CD1 Cox-2⫺/⫺ females indicated polymerase (Invitrogen) using the following primers: Ptgerep1 that the processes of oocyte maturation and follicular rupture (148 bp), 5⬘-GGGGGCTGGAACTCTAACTC-3⬘ and 5⬘-TGAC- during ovulation have differential requirements for COX-deCCTCAGGGGTAGGA-3⬘; Ptgerep2 (401 bp), 5⬘-AGGACTTC- rived PGs (24). In fact, we observed in CD1 Cox-2⫺/⫺ females GATGGCAGAGGAGAC-3⬘ and 5⬘-CAGCCCCTTACAC- that an early induction of COX-1 by gonadotropin in mural TTCTCCAATG-3⬘; Ptgerep3 (262 bp), 5⬘-TACCTGTTTCCC- granulosa and theca cells partially offsets the loss of COX-2 for TGGGTCTG-3⬘ and 5⬘-CAAAGGTTCTGAGGCTGGAG-3⬘; completion of follicular rupture, whereas cumulus cells surroundDECEMBER 1, 2006 • VOLUME 281 • NUMBER 48

JOURNAL OF BIOLOGICAL CHEMISTRY

37119

Differential Signaling by PGE2 in Oocyte Maturation Because oocyte meiosis resumption and progression, including GVBD and MI to MII transition, are temporally regulated dynamic events, we next examined the kinetics of in vivo oocyte maturation induced by hCG. As illustrated in Fig. 1A, although all wild-type oocytes underwent GVBD (the first visible sign of meiosis resumption) at 6 h post-hCG stimulation, Cox-2⫺/⫺ oocytes showed delayed resumption of meiosis, most of which exhibited GVBD at 15 h post-hCG administration. Under normal conditions, in parallel to meiotic oocyte maturation, the surrounding cumulus matrices underwent physical and physiological changes termed cumulus expansion preparing for “soonto-occur” sperm-oocyte fertilization events. Thus, we compared the cumulus expansion status between wild-type and Cox-2⫺/⫺ mice in vivo. Again, we observed deferred and compromised cumulus expansion in Cox-2⫺/⫺ mice (Fig. 1B). For example, wild-type COCs showed visible signs of expansion at 3 h after hCG injection and reached full expansion at 12 h. In contrast, the FIGURE 2. FSH induces COX-2 expression in cumulus cells (A) and PGE2 production (B) in culture. COCs expansion of Cox-2⫺/⫺ COCs was were cultured in media containing HX or HX ⫹ 100 IU/liter hFSH for 18 h. At indicated times, COCs and culture media were collected and subjected to immunofluorescence analysis of COX-2 and PG assays, respectively. first noted at 9 h post-hCG adminPropidium iodide (PI) was used as nuclear staining. BF, bright field. Bar, 100 ␮m. PGI2 was measured as 6-keto- istration with a largely reduced PGF1␣. *, significantly different from Cox-2⫺/⫺ mice (p ⬍ 0.05). expanded area compared with that of wild-type COCs. Collectively, the ing the oocyte lacking COX-1 expression fail to correct the COX-2 results suggest that oocyte maturation is lessened in response to deficiency-associated aberrant oocyte maturation and cumulus ovulation stimulus in Cox-2⫺/⫺ mice. This finding led us to expansion (24). Thus, an in-depth understanding of compromised examine whether gonadotropin-induced oocyte maturation is oocyte maturation events in the absence of COX-2 remains elusive impaired in vitro in the absence of COX-2. and instigated us to further explore the role of PGs in oocyte matWe have shown previously that in vitro maturation of uration in vivo and in vitro using Cox-2⫺/⫺ mice and the potential C57BL/6J/129 Cox-2⫺/⫺ oocytes in the presence of fetal bovine signaling pathways involved in this process. serum and FSH is comparable with wild-type oocytes (10). To examine in vivo oocyte maturation status in CD1 Cox- Under the same culture condition, we further observed that 2⫺/⫺ mice, we employed a standard superovulation regimen there were no differences between the wild-type and CD1 Coxusing a combination of PMSG and hCG protocol as described 2⫺/⫺ oocytes in their ability to undergo GVBD and progression previously (10, 24). All wild-type oocytes retrieved from super- to the MII (⬃80%) stage (data not shown). This observation ovulated mice 15 h after an hCG injection underwent GVBD, indicates that serum-derived factors compensate for the loss of and a large number of them were arrested at MII stage (supple- COX-2 or, alternatively, COX-2-derived PGs are not always mental Fig. 1, A and B). In contrast, deferred resumption and necessary for oocyte maturation in vitro. To address this issue, progression of meiosis were observed in Cox-2⫺/⫺ oocytes in we next used a serum-free medium containing 4 mM HX, a response to hCG stimulation. For example, the majority of null nonspecific phosphodiesterase inhibitor that prevents spontaoocytes was arrested in MI stage, followed by those that pro- neous maturation of oocytes. Interestingly, a compromised gressed to MII stages. Most strikingly, a small percentage of null oocyte meiotic maturation and cumulus expansion evaluated oocytes still exhibited intact GV, indicating their failure to by GVBD ratio and projected area of cumulus expansion, resume meiosis (supplemental Fig. 1, A and B). These results respectively, were observed when Cox-2⫺/⫺ COCs were culconfirm our initial observation of compromised oocyte matu- tured under serum-free conditions and challenged with FSH ration in the absence of COX-2. (Fig. 1, C and D, and supplemental Fig. 1C). In addition, the


VOLUME 281 • NUMBER 48 • DECEMBER 1, 2006

Differential Signaling by PGE2 in Oocyte Maturation 4 h with sustained high levels until termination of the culture at 18 h. This temporal induction of COX-2 in COCs is well correlated with PG accumulation in the culture medium. We noted that PGE2 is the major PG secreted upon induction of COX-2 by FSH in wild-type COCs with low to undetectable levels in Cox-2⫺/⫺ COCs similarly cultured (Fig. 2B). These results further support the previous view that COX-2-derived PGE2, but not COX-1, is a key player in regulating oocyte maturation and cumulus expansion (8, 24, 27). Because PGE2 is the primary PG produced by cumulus cells in response to gonadotropin both in vivo and in vitro, we subsequently tested whether PGE2 could rescue defective oocyte maturation and cumulus expansion in Cox-2⫺/⫺ mice. With respect to meiotic maturation, although FSH was fully functional in stimulating wild-type oocyte GVBD and progression into the MII stage, PGE2 alone was also able to induce meiotic resumption but to a lesser extent. A combined treatment with FSH and PGE2 exerted a similar effect as FSH alone (Fig. 3, A and B). This observation is FIGURE 3. PGE2 differentially regulates oocyte maturation and cumulus expansion in vitro. A and B, PGE2 with our findings modestly rescues COX-2 deficiency-induced oocyte meiotic maturation defects. WT and Cox-2⫺/⫺ COCs were consistent cultured in media containing HX, HX ⫹ 100 IU/liter FSH, HX ⫹ 1 ␮M PGE2, or HX ⫹ FSH ⫹ PGE2 for 18 h. The described above showing FSH-inpercentage of oocytes with GVBD (A) or at metaphase II (MII) (B) was scored at termination of culture. C and duced COX-2 expression and PGE 2 D, PGE2 restores normal cumulus expansion in Cox-2⫺/⫺ mice. Representative photographs of COCs after various treatments in culture (C). Bar, 100 ␮m. Projected areas of COCs were calculated. Bars with different secretion in cumulus cells. Howletters are significantly different (p ⬍ 0.01). ever, our observation of less efficient meiotic resumption by PGE2 is percentage of Cox-2⫺/⫺ oocytes progressing to the MII stage in suggestive of alternative pathways under FSH regulation, response to FSH stimulation was significantly lower in compar- namely the epidermal growth factor family of growth factors, ison to wild-type mice (supplemental Fig. 1D). These results ovarian steroids, and sterols (15, 29 –33). We thus speculated faithfully phenocopied the in vivo observation of defective that FSH and PGE2 have additive effects in inducing maturation oocyte maturation in Cox-2⫺/⫺ females and provoked us to of Cox-2⫺/⫺ oocytes in culture. Indeed, it was interesting to further explore the underlying molecular mechanisms and sig- note that FSH or PGE2 alone had a modest stimulatory role in naling pathways associated with gonadotrophins and COX-2- inducing GVBD and MI to MII transition in null oocytes, derived PGs during oocyte maturation using this simply whereas FSH and PGE2 cotreatment additively improved defined culture system. oocyte meiotic maturation in the absence of COX-2 (Fig. 3, A PGE2 Differentially Regulates Oocyte Maturation and Cumu- and B). In the same context, we also examined cumulus expanlus Expansion in Vitro—Previous studies have demonstrated a sion in culture. It is interesting to note that PGE2 alone was spatiotemporal expression of COX-2 in mural granulosa and sufficient to induce nearly normal cumulus expansion in both cumulus cells in response to preovulatory gonadotropin stim- wild-type and Cox-2⫺/⫺ COCs (Fig. 3, C and D). No additive ulation in mice (24, 27, 28). To further elucidate the physiolog- effects were observed when COCs of either genotype were ical significance of COX-2-derived PGs in oocyte maturation in cotreated with FSH and PGE2, although FSH did exert some vitro, we first assessed COX-2 expression by immunofluores- stimulatory effects on mutant COC cumulus expansion. It is cence in wild-type COCs in culture. As shown in Fig. 2A, conceivable that gonadotropin-induced alternative signaling COX-2 proteins are progressively induced in cumulus cells by molecules in addition to PGE2 are involved in regulating oocyte FSH stimulation; first detectable expression was seen as early as meiotic maturation as well as cumulus expansion. Nonetheless, DECEMBER 1, 2006 • VOLUME 281 • NUMBER 48


37121

Differential Signaling by PGE2 in Oocyte Maturation These findings provoked us to elucidate further the mechanism by which COX-2-derived PGE2 induces cumulus expansion. PGE2 Restores Normal Cumulus Expansion via Up-regulating TSG-6 Expression in Cox-2⫺/⫺ COCs in Culture—Cumulus expansion involves matrix remodeling, and TSG-6 is considered as an important player in this event (34). Previous investigation has shown that in vivo and in vitro expression of TSG-6 in Cox-2⫺/⫺ COCs is lower as compared with wild-type mice (35). We wanted to confirm this earlier observation in our experiments that used mice on the CD1 background by immunofluorescence and immunoblotting. Immunofluorescence of TSG-6 was strongly detected in the matrix of COCs collected from superovulated wildtype mice after 15 h of hCG injection (Fig. 4A). In contrast, only a faint immunofluorescence of TSG-6 was detected in cumulus cells, but not in the matrix, of Cox2⫺/⫺ COCs (Fig. 4A). This downregulated TSG-6 expression in Cox2⫺/⫺ COCs was further confirmed by immunoblotting. Antibodies against TSG-6 recognized two bands, a 36-kDa band, corresponding to free TSG-6, and a 120-kDa band, representing inter-␣-trypsin inhibitor-TSG-6 complex. We found that Cox-2⫺/⫺ COCs recovered from superovulated females had less TSG-6 proteins of either type when compared with wild-type COCs receiving the same treatment (Fig. 4B). We speculated that this remarkable reduction of TSG-6 FIGURE 4. PGE2 restores cumulus cell expansion and survival by up-regulating TSG-6 expression in Cox- expression and incorporation into 2ⴚ/ⴚ COCs in culture. COX-2 deficiency leads to reduced TSG-6 expression and incorporation into cumulus matrices (A and B), and increased cumulus cell apoptosis (C–F). PGE2 or PGE2 plus FSH, but not FSH alone, the cumulus matrix compromises restores normal TSG-6 expression (G), cumulus cell expansion (H), and survival (I) in culture. Ovulated COCs the structural integrity of the were collected 15 h after hCG injection from WT and Cox-2⫺/⫺ mice. TSG-6 expression was analyzed by immu- matrix, leading to cumulus cell nofluorescence (A) and immunoblotting (B). Cumulus cell apoptosis was examined by TUNEL assay (green), immunostaining of cleaved caspase-3 (red), and immunoblotting of Bcl-2 (C–F). TUNEL-positive or cleaved death. Thus, we examined the apocaspase-3-positive cumulus cells from WT (n ⫽ 54) and Cox-2⫺/⫺ (n ⫽ 44) mice were calculated. Band intensi- ptosis status of superovulated wildties were quantified by densitometry from three replicate experiments. *, significantly different from WT mice type and Cox-2⫺/⫺ COCs. We (p ⬍ 0.01). Upon in vitro maturation, COCs were cultured in HX, HX ⫹ 100 IU/liter FSH, HX ⫹ FSH ⫹ 1 ␮M PGE2 for 18 h. TSG-6 expression and cell apoptosis were examined (G–I). At least 30 COCs from each group were noted that a substantial number of examined and repeated three times. Bars with different letters are significantly different (p ⬍ 0.01). Propidium cumulus cells undergo apoptosis iodide (PI) or Hoechst 33258 was used as nuclear staining. BF, bright field. Bar, 100 ␮m. (TUNEL-positive cells) upon the loss of COX-2 (Fig. 4C). This the results reinforce the concept that de novo synthesis of PGE2 increased cumulus cell death in Cox-2⫺/⫺ mice was associated via COX-2 in cumulus cells is essential for the functional integ- with down-regulation of Bcl-2, an anti-apoptotic protein (36, rity of COCs during oocyte maturation and cumulus expansion. 37), but an up-regulation of caspase-3, a key mediator of apo-



Differential Signaling by PGE2 in Oocyte Maturation receptors in this process remained unexplored. In this study, we addressed this question using RT-PCR and immunofluorescence analysis. COCs were collected after 15 h of hCG injection and were subjected to immunofluorescence analysis for EP receptors. For RT-PCR analysis, cumulus cells dispersed from oocytes by hyaluronidase digestion were used. As shown in Fig. 5, A and B, cumulus cells, but not oocytes, expressed EP2, EP3, and EP4. To characterize the role of these receptors in cumulus expansion in vitro, wild-type COCs were cultured with butaprost (an EP2 agonist), sulprostone (an EP1/EP3 mixed agonist), or 11-deoxy-PGE1 (an EP2/EP4 mixed FIGURE 5. EP2 and EP4 receptors are important for cumulus expansion. A and B, PGE2 receptor expression agonist) for 18 h. As shown in Fig. 5, in COCs. EP1, EP2, EP3, and EP4 mRNAs and proteins were analyzed by RT-PCR and immunofluorescence, respectively. Mouse kidney RNA was used as a positive control for the EP1. ⫹, ⫺, presence or absence of RT C and D, 11-deoxy-PGE1, but not reaction. M, marker. C and D, activation of EP2 and EP4 induces cumulus expansion in culture. COCs were butaprost or sulprostone, mimicked cultured in hypoxanthine medium containing butaprost, sulprostone, or 11-deoxy-PGE1 for 18 h. Numbers inside bars indicate numbers of expanded COCs/total number of COCs examined. *, significantly different from the role of native PGE2 in inducing other groups (p ⬍ 0.01). PI, propidium iodide. Bar, 100 ␮m. cumulus expansion. The results suggest that both EP2 and EP4 subptosis (38, 39) (Fig. 4, C–F). The results collectively suggest that types are required for cumulus expansion in vitro. To provide the defective cumulus expansion in Cox-2 mutant mice is further evidence on the involvement of EP2 and EP4 receptors because of down-regulated TSG-6 function and enhanced in oocyte meiotic resumption and cumulus expansion, we used cumulus cell death. To explore further insights to this defect, EP receptor antagonists in culture, such as AH6809 (an EP1/ we examined the regulatory effects of FSH and PGE2 on TSG-6 EP2 mixed antagonist) and AH23848 (an EP4 antagonist). expression and cell apoptosis status in cultured COCs. As Wild-type COCs were cultured in the presence or absence of shown in Fig. 4, G–I, FSH or PGE2 alone was sufficient to AH6809, AH23848, or a combination of AH6809 and AH23848 induce TSG-6 expression and facilitate cell survival in wild-type for 1 h before FSH treatment. It was interesting to observe that COCs in culture, whereas PGE2, but not FSH, was capable of blockade of either EP2 or EP4 activity did not substantially up-regulating TSG-6 expression and restoring normal cumulus influence FSH-induced oocyte meiotic maturation (Fig. 6, A expansion in the absence of COX-2, although a modest role of and B), but significantly inhibited FSH- or PGE2-triggered FSH in restoring cell survival was observed in mutant COCs. It cumulus expansion (Fig. 6, C and D). The results are consistent is noteworthy that cotreatment with FSH and PGE2 not only with our above observations that cumulus cells, but not exhibits strong TSG-6 expression, but also its incorporation oocytes, are the primary target of PGE2-EP2/4 signaling. Moreinto the cumulus matrix of COCs in both genotypes (Fig. 4G), over, this impaired cumulus expansion in response to FSH by giving a smeared staining pattern. This suggests that although EP2/EP4 antagonists is associated with significantly less inducPGE2 is an upstream inducer of TSG-6 expression in cumulus tion of TSG-6 (Fig. 6E), reinforcing that TSG-6 is a downstream cells, other signaling molecules under FSH stimulation in col- target of PGE2. However, a modest inhibitory effect on oocyte laboration with PGE2 further stimulate TSG-6 expression and meiotic maturation was observed when COCs were cotreated incorporation into the COC matrix. It was also interesting to with AH6809 and AH23848. We speculate that this is a secondnote that even the basal level of cumulus cell apoptotic activity ary effect that resulted from severely restrained cumulus was higher in Cox-2⫺/⫺ mice in comparison with that of wild- expansion under the cotreatment condition, suggesting that type mice. Nonetheless, the results suggest that TSG-6 reciprocal oocyte-somatic cell interactions are critical for norexpression is under the regulation of COX-2-PGE2 signaling mal oocyte maturation. More importantly, the results add new in response to gonadotropin during oocyte maturation and evidence that PGE2 signaling is primarily mediated by both EP2 cumulus expansion. Our next objective was to see which of and EP4 subtypes. the four PGE2 receptor subtypes, EP1, EP2, EP3, and EP4, PGE2 Directs Cumulus Cell Expansion and Survival during mediate PGE2 signaling in these events. Oocyte Maturation by Using Different Signaling Cascades—We PGE2 Functions through EP2 and EP4 Receptors during next asked which PGE2 downstream signaling pathways are Oocyte Maturation and Cumulus Expansion in Vitro—Al- involved in oocyte maturation and cumulus expansion. though previous pharmacological and genetic evidence has Increasing evidence points toward the involvement of the shown the importance of EP2 in mediating PGE2 function dur- PI3K-Akt signaling cascade in gonadotropin-induced oocyte ing oocyte maturation (8, 10, 35), the participation of other EP meiotic maturation (40, 41). Thus, we first examined the DECEMBER 1, 2006 • VOLUME 281 • NUMBER 48


37123

Differential Signaling by PGE2 in Oocyte Maturation sponding to germinal vesicle breakdown. Interestingly, barrel-shaped signals of phospho-Akt were observed after 12 h in culture. The fluorescence intensity of phospho-Akt in Cox-2⫺/⫺ oocytes was less than those of wild-type oocytes and so was the percentage of oocytes positive for phospho-Akt (Fig. 7B). Again, it was interesting to note that a combined treatment of FSH and PGE2 significantly increased the percentage of phospho-Akt positive oocytes in Cox-2⫺/⫺ mice compared with FSH treatment alone, indicating that substantially restored cumulus expansion by PGE2 in null COCs improves oocyte meiotic maturation. To provide further evidence on the participation of the Akt pathway during oocyte maturation, we employed a pharmacological approach using LY294002 to selectively block PI3K and thus Akt signaling in culture. As shown in Fig. 8, A–C, LY294002 treatment in a dose-dependent manner inhibited FSH-induced oocyte Akt activation as well as meiosis resumption and progression, which was not corrected by PGE2 cotreatment. This is consistent with our observation of a lack of EP receptor expression in oocytes, suggesting that oocyte Akt FIGURE 6. Pharmacological silencing of EP2 and EP4 attenuates FSH-induced oocyte meiotic maturation phosphorylation is activated by an (A and B), cumulus expansion (C and D), and down-regulation of TSG-6 expression (E) in vitro. COCs were cultured in media containing HX, HX ⫹ 100 IU/liter FSH, HX ⫹ FSH ⫹ 1 ␮M PGE2 for 18 h. COCs were pretreated alternative stimulus independent of with AH6809 (50 ␮M), AH23848 (50 ␮M), or AH6809 (50 ␮M) plus AH23848 (50 ␮M) for 1 h before addition of FSH the COX-2-PGE2 signaling axis. or PGE2. BF, bright field, PI, propidium iodide, Bar, 100 ␮m. The bars with different letters are significantly Because gonadotropin also induces different (p ⬍ 0.01). Akt phosphorylation in cumulus cells expression of phospho-Akt in COCs in wild-type and Cox-2⫺/⫺ during in vivo oocyte maturation (supplemental Fig. 2), the physimice during in vivo oocyte maturation. Phospho-Akt was ological significance of this signaling cascade in these cells remains detected as early as 3 h after hCG injection in both oocytes and unexplored. Therefore, we examined cumulus cell expansion in cumulus cells of wild-type mice, but not in those from Cox-2 response to LY294002. As shown in Fig. 8D, we observed a signifmutant mice as assessed by immunofluorescence and immuno- icant decrease in FSH-induced cumulus expansion in the presence blotting (supplemental Fig. 2, A–C). Interestingly, at 6 h post- of LY294002, which was not restored by PGE2 cotreatment, indihCG injection, phospho-Akt signals were detected in all cating that PI3K-Akt signaling in regulating cumulus cell funcoocytes examined in wild-type mice as opposed to only 50% of tion lies downstream of COX-2-derived PGE2. This is consistCox-2⫺/⫺ oocytes showing phospho-Akt signals (supplemental ent with our subsequent observation of more prominent Akt Fig. 2, A and B). This timely activation of Akt signaling in phosphorylation by FSH in wild-type cumulus cells as commaturing oocytes is well correlated with the timing of the pared with Cox-2⫺/⫺ cells (Fig. 8E) and of Akt activation by GVBD event as illustrated in Fig. 1 and suggests that Akt sig- PGE2 alone in cumulus cells in culture (Fig. 8F). In addition, naling is an important pathway for ensuring normal oocyte alteration of TSG-6, but not COX-2 expression, in response to meiotic maturation. In contrast, its deferred activation in the FSH by LY294002 (Fig. 8, G and H) further places the PI3K-Akt absence of COX-2 leads to defective oocyte maturation in vivo. signaling pathway downstream of the COX-2-PGE2 signaling This finding led us to examine Akt activation in oocytes during axis during cumulus expansion. in vitro maturation. Because PGE2 restores normal cumulus cell expansion As shown in Fig. 7A, we observed peri-nuclear signals for (Fig. 3C) as well as cumulus cell survival (Fig. 4H) in Coxphospho-Akt after 8 h in culture. Signals were initially scattered 2⫺/⫺ females, we further asked whether these processes but concentrated at the center of oocytes after 10 h, corre- under the influence of PGE2 utilize differential signaling cas-



Differential Signaling by PGE2 in Oocyte Maturation

FIGURE 7. COX-2 deficiency causes deferred Akt activation in oocytes during meiotic maturation in vitro. COCs were cultured in media containing HX, HX ⫹ 100 IU/liter FSH, HX ⫹ FSH ⫹ 1 ␮M PGE2 for 18 h. After removing cumulus cells, oocytes were stained with anti-phospho-Akt (p-Akt) antibody and propidium iodide (PI). BF, bright field. Bar, 50 ␮m. p-Akt immunostaining was classified as follows: no signals, perinuclear staining (type A), scattered staining (type B), round-shaped staining (type C), or barrel-shaped staining (type D). Numbers inside bars indicate total numbers of oocytes examined.

cades during oocyte maturation. Early studies have demonstrated that cumulus expansion in response to gonadotropin or other stimuli depends on multiple signaling pathways, including cAMP/PKA (42– 45), ERK1/2 (34, 46 – 49), p38 MAPK (34, 49), as well as PI3K-Akt signaling (40, 41). Thus, we next used selective antagonists of these pathways and tested their effects under PGE2 stimulation in cumulus cells. As depicted in Fig. 8I, we surprisingly observed that attenuation of ERK1/2, p38 MAPK, or nuclear factor ␬B (NF-␬B) by U0126, SB203580, or PGA1, respectively, failed to show any effect on PGE2-stimulated cumulus cell survival, but profoundly blocked PGE2-induced cumulus expansion. It is conceivable that PGE2 functions through alternative pathways in regulating cell survival. Indeed, we noted that blockade of the PI3K-Akt pathway by LY294002 substantially dampened cumulus cell survival, perhaps resulting in modest impairment of cumulus expansion seen in response to PGE2. Similar effects were also noted when PKA or c-Jun N-terminal kinase (c-JNK), which is known to participate in Xenopus oocyte meiotic maturation (50, 51), was selectively inhibited by H-89 or SP60012, respectively (Fig. 8I). It is interesting to note that PGE2 up-regulated the expression of its own synthesizing enzyme COX-2 in cumulus cells during oocyte maturation (Fig. 8J). We speculated that this positive feedback regulation of COX-2 contributes to de novo ampliDECEMBER 1, 2006 • VOLUME 281 • NUMBER 48

fication of PGE2 signaling and consequently to the regulation of cumulus expansion. To test this hypothesis, we analyzed the consequence of inhibition of differential signaling cascades on COX-2 expression. As illustrated in Fig. 8J, we observed that blockade of the ERK1/2, p38 MAPK, or NF-␬B signaling cascade that leads to defective cumulus expansion, but not PI3K-Akt, PKA, or c-JNK pathway, antagonized the PGE2-induced up-regulation of COX-2 expression. This finding reinforces the idea that PGE2 by a feed-forward mechanism regulates cumulus cell expansion via ERK1/2, p38 MAPK, or NF-␬B signaling, although its effects on cell survival are mediated through the PI3K-Akt, PKA, or c-JNK pathway. Nonetheless, the results highlight the diversity of PGE2 signaling in regulating cumulus cell expansion and survival via different downstream pathways. To further explore the physiological relevance of COX-2derived PGE2 in oocyte meiotic maturation and their fertilizing capacity, we examined in vivo oocyte maturation in Cox-2 null mice receiving PGE2 injections. As shown in Table 1, PGE2 largely restores normal oocyte meiotic maturation in vivo in the absence of COX-2. Furthermore, these null oocytes with PGE2 pretreatment in vivo during the preovulatory period exhibit normal fertilizing capacity in vitro (Table 2). The results reinforce that COX-2-derived PGE2 is at least one of the key players in regulating the resumption and progression of oocyte meiotic JOURNAL OF BIOLOGICAL CHEMISTRY

37125




Differential Signaling by PGE2 in Oocyte Maturation TABLE 1 PGE2 rescues oocyte meiotic maturation in Cox-2ⴚ/ⴚ mice in vivo

Superovulated oocytes were collected from WT and Cox-2⫺/⫺ mice 14 –15 h after hCG injection. Oocytes were examined and classified as follows: GV, germinal vesicle; MI, metaphase I; MII, metaphase II by nuclear staining. The percentage of oocytes classified for each stage of meiosis for WT and Cox-2⫺/⫺ mice was calculated by dividing the number of oocytes at each stage of meiosis by the total number of ovulated oocytes. Data are presented as mean ⫾ S.E. Values with different superscript letters within the same column are statistically significant (p ⬍ 0.05; analysis of variance). Treatment

No. of tested mice with ovulation

No. of ovulated oocytes

GV %

%

%

WT ⫹ vehicle Cox-2⫺/⫺ ⫹ vehicle Cox-2⫺/⫺ ⫹ PGE2

12 7 5

30 ⫾ 3a 16 ⫾ 3b 25 ⫾ 4a

0⫾0 2⫾2 0⫾0

31 ⫾ 4a 54 ⫾ 5b 34 ⫾ 4a

69 ⫾ 4a 44 ⫾ 5b 66 ⫾ 4a

MI

MII

TABLE 2 PGE2 restores normal fertilizing capacity of Cox-2ⴚ/ⴚ oocytes Superovulated COCs were subjected to in vitro fertilization. The appearance of two pronuclei (PN) indicated successful fertilization. The number of blastocysts formed from fertilized eggs was recorded after 5 days of culture.

a

Treatment

No. of tested mice with ovulation

No. of 2-PN oocytes/ total oocytes

No. of blastocysts/ total 2-PN oocytes

WT ⫹ vehicle Cox-2⫺/⫺ ⫹ vehicle Cox-2⫺/⫺ ⫹ PGE2

13 9 3

269/347 (78%) 73/172 (42%)a 44/66 (67%)

254/269 (94%) 56/73 (77%)a 39/44 (89%)

Statistical analysis was performed by ␹ 2 test; p ⬍ 0.01.

maturation during ovulation to further ensure their normal developmental potential.

DISCUSSION In this investigation, we studied in-depth the roles of PGs in oocyte maturation using genetic, molecular, physiological, and pharmacological approaches. The highlights of our investigation are that COX-2-derived PGE2 in response to gonadotropin coordinates oocyte meiotic maturation, cumulus cell expansion, and survival by utilizing different downstream signaling cascades. Furthermore, this study elucidates a novel mechanism of COX-2 expression in cumulus cells by PGE2 that facilitates cumulus expansion. The findings of decreased accumulation of TSG-6 in the cumulus matrix with enhanced cumulus cell apoptosis in the absence of COX-2 provide evidence for the underlying cause of compromised fertilization in Cox-2⫺/⫺ mice. In addition, here we provide pharmacological evidence that both EP2 and EP4 receptors are potentially important in mediating PGE2 bioactivity during oocyte maturation and cumulus expansion. Activation and expansion of the cumulus oophorum as well as timely resumption of oocyte meiosis are critical to successful ovulation and fertilization (52). The participation of COX-2derived PGE2 in these processes has been elucidated by a number of pharmacological and genetic studies. In this study, we demonstrate that FSH up-regulates cumulus COX-2 expression and thus PGE2 production in culture, which phenocopies the in vivo induction of COX-2 and PGE2 synthesis in cumulus

cells in response to the preovulatory gonadotropin surge, suggesting its role in cumulus expansion and oocyte maturation. The rescue of defective cumulus expansion in Cox-2⫺/⫺ mice in vivo or in vitro by PGE2 provides definitive evidence that this PG is critical to this process. TSG-6, which is expressed in cumulus and mural granulosa cells in preovulatory follicles (53–56), is known to modulate extracellular matrix remodeling and is a known marker of cumulus expansion, but not meiotic maturation of oocytes. Our observation of down-regulation of TSG-6 in cumulus cells in the absence of COX-2, but its up-regulation by PGE2, provides evidence that TSG-6 is a downstream target of PGE2. This is consistent with other studies showing down-regulation of hCG or FSH-induced TSG-6 in cumulus cells missing EP2 or COX-2 (34, 35). TSG-6 secreted into the cumulus cell matrix binds to hyaluronic acid and other hyaluronic binding proteins, such as PTX-3 or serum-derived inter-␣-trypsin inhibitor, to stabilize the matrix bed. The stabilization of the matrix by these molecules is essential for cumulus expansion and ovulation. Gene-targeting experiments in mice have shown multiple female reproductive defects that include reduced ovulation, defective cumulus expansion, and fertilization failure in the absence of these factors (57– 61). Early studies have provided genetic evidence that EP2 is a key player in ovulation and fertilization (8, 9, 62). Similar to Cox2⫺/⫺ mice, EP2-deficient mice also show defective cumulus expansion in addition to reduced ovulation and fertilization

FIGURE 8. PGE2 directs cumulus cell expansion and survival using different signaling cascades. A–D, inactivation of PI3K-Akt signaling inhibits FSHinduced oocyte meiotic maturation and cumulus expansion. WT COCs were cultured in medium containing HX, HX ⫹ 100 IU/liter FSH, HX ⫹ FSH ⫹ 1 ␮M PGE2 for 18 h. COCs were pretreated with various concentrations of LY294002 for 1 h before exposing to FSH or PGE2. Numbers inside bars indicate numbers of p-Akt-positive oocytes/total number of oocytes tested. E, FSH induces Akt phosphorylation in WT, but not in COX-2⫺/⫺, cumulus cells. COCs were cultured in HX medium with 100 IU/liter FSH for 4 h. Cumulus cells were dispersed from COCs for immunoblotting. F, PGE2 activates Akt in cumulus cells. WT COCs were cultured with 1 ␮M PGE2. These experiments were repeated twice with similar results. G and H, silencing of PI3K-Akt signaling attenuates FSH-induced TSG-6 but not COX-2, expression. I, PGE2 regulates cumulus cell expansion and survival through multiple signaling pathways. COCs were cultured in HX medium with 1 ␮M PGE2 with or without H-89, U0126, SB203580, SP60012, PGA1, or LY294002 for 18 h. J, PGE2 up-regulates COX-2 expression in cumulus cells. At least 30 COCs from each group were tested and repeated three times. Bars with different letters are significantly different (p ⬍ 0.05). PI, propidium iodide. BF, bright field. Bar, 100 ␮m.



37127

Differential Signaling by PGE2 in Oocyte Maturation cumulus cell survival with modest effects on cumulus expansion. These observations also suggest that PGE2 modulation of these signaling pathways lies downstream of gonadotropin-induced oocyte maturation. Another striking observation is the up-regulation of COX-2 by PGE2 involving the same signaling pathways that promote cumulus cell expansion. This would then suggest that COX-2-derived PGE2 occupies a pivotal place downstream of gonadotropins in amplifying its own signaling via a “feedforward-feedback” loop not only to initiate but also to sustain the oocyte maturation process after termination of the gonadotropin surge. Our study also reveals that direct targets of PGE2 signaling are the cumulus cells but not the oocytes that lack EP receptors. In conclusion, the present investigation underscores the diversity of PGE2 signaling during oocyte maturation. This study is clinically relevant to female fertility with special FIGURE 9. A diagram depicts the potential signaling network impacted by gonadotropins and PGE2 reference to women of reproductive during oocyte maturation and cumulus expansion. ages who consume COX-2-selective inhibitors or other nonsteroidal (10). However, our present observations provide pharmacolog- anti-inflammatory steroids that are nonspecific inhibitors of ical evidence that both EP2 and EP4 are important players in COX-1 and COX-2. cumulus expansion for the following reasons. First, 11-deoxyPGE1, a mixed agonist for EP2/EP4, but not butaprost, a selec- Acknowledgment—We thank Anthony J. Day (Medical Research tive EP2 agonist, induces cumulus expansion. Second, these Council Immunochemistry Unit, Department of Biochemistry, processes are attenuated by a mixed EP2/EP4 antagonist. Third, Oxford University, UK) for the generous gift of TSG-6 antibodies. EP2, EP3, and EP4 are expressed in cumulus cells, and hCG up-regulates EP2 and EP4 expression in these cells (28). How- REFERENCES ever, the role of EP3 in cumulus cell function does not appear to 1. Dey, S. K., Lim, H., Das, S. K., Reese, J., Paria, B. C., Daikoku, T., and Wang, be critical, because EP3 null females do not exhibit overt reproH. (2004) Endocr. Rev. 25, 341–373 2. Wang, H., and Dey, S. K. (2005) Prostaglandins Other Lipid Mediat. 77, ductive phenotypes (11). Because EP4 deficiency results in peri84 –102 natal lethality (12, 13), it remains unanswered whether EP2 and 3. Smith, W. L., and Dewitt, D. L. (1996) Adv. Immunol. 62, 167–215 EP4 have distinct and/or overlapping function in female 4. Lim, H., Paria, B. C., Das, S. K., Dinchuk, J. E., Langenbach, R., Trzaskos, reproduction. J. M., and Dey, S. K. (1997) Cell 91, 197–208 Cumulus cell expansion in response to gonadotropins has 5. Davis, B. J., Lennard, D. E., Lee, C. A., Tiano, H. F., Morham, S. G., Wetsel, been shown to be governed by multiple signaling pathways (52). W. C., and Langenbach, R. (1999) Endocrinology 140, 2685–2695 6. Lim, H., Gupta, R. A., Ma, W. G., Paria, B. C., Moller, D. E., Morrow, J. D., However, our observation of increased apoptosis in addition to DuBois, R. N., Trzaskos, J. M., and Dey, S. K. (1999) Genes Dev. 13, impaired expansion of cumulus cells missing COX-2 adds a 1561–1574 new complexity to this process. Because PGE2 can simulate 7. Negishi, M., Sugimoto, Y., and Ichikawa, A. (1995) Biochim. Biophys. Acta gonadotropins in cumulus expansion, we explored site- and 1259, 109 –119 event-specific participation of PKA, MAPK, NF-␬B, and PI3K8. Hizaki, H., Segi, E., Sugimoto, Y., Hirose, M., Saji, T., Ushikubi, F., MatAkt pathways during PGE2-induced oocyte maturation. suoka, T., Noda, Y., Tanaka, T., Yoshida, N., Narumiya, S., and Ichikawa, A. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 10501–10506 Indeed, a remarkable diversification of these pathways in 9. Kennedy, C. R., Zhang, Y., Brandon, S., Guan, Y., Coffee, K., Funk, C. D., cumulus cell expansion and survival has emerged from this Magnuson, M. A., Oates, J. A., Breyer, M. D., and Breyer, R. M. (1999) Nat. study (Fig. 9). For example, although ERK1/2, p38 MAPK, and Med. 5, 217–220 NF-␬B pathways are critical to PGE2-induced cumulus expan- 10. Matsumoto, H., Ma, W., Smalley, W., Trzaskos, J., Breyer, R. M., and Dey, sion, they are not essential for cumulus cell survival. In contrast, S. K. (2001) Biol. Reprod. 64, 1557–1565 PI3K-Akt, PKA, and c-JNK pathways are more involved in 11. Ushikubi, F., Segi, E., Sugimoto, Y., Murata, T., Matsuoka, T., Kobayashi,




12.

13.

14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

24. 25. 26. 27. 28. 29. 30. 31. 32.

33. 34. 35. 36. 37. 38.

T., Hizaki, H., Tuboi, K., Katsuyama, M., Ichikawa, A., Tanaka, T., Yoshida, N., and Narumiya, S. (1998) Nature 395, 281–284 Nguyen, M., Camenisch, T., Snouwaert, J. N., Hicks, E., Coffman, T. M., Anderson, P. A., Malouf, N. N., and Koller, B. H. (1997) Nature 390, 78 – 81 Segi, E., Sugimoto, Y., Yamasaki, A., Aze, Y., Oida, H., Nishimura, T., Murata, T., Matsuoka, T., Ushikubi, F., Hirose, M., Tanaka, T., Yoshida, N., Narumiya, S., and Ichikawa, A. (1998) Biochem. Biophys. Res. Commun. 246, 7–12 Peng, X. R., Hsueh, A. J., LaPolt, P. S., Bjersing, L., and Ny, T. (1991) Endocrinology 129, 3200 –3207 Park, J. Y., Su, Y. Q., Ariga, M., Law, E., Jin, S. L., and Conti, M. (2004) Science 303, 682– 684 Sato, E., and Yokoo, M. (2005) Ital. J. Anat. Embryol. 110, Suppl. 1, 205–217 Neal, P., Baker, T. G., McNatty, K. P., and Scaramuzzi, R. J. (1975) J. Endocrinol. 65, 19 –25 Downs, S. M., and Longo, F. J. (1982) Am. J. Anat. 164, 265–274 Downs, S. M., and Longo, F. J. (1983) J. Exp. Zool. 228, 99 –108 Downs, S. M., and Longo, F. J. (1983) Anat. Rec. 205, 159 –168 Yoshimura, Y., Karube, M., Oda, T., Koyama, N., Shiokawa, S., Akiba, M., Yoshinaga, A., and Nakamura, Y. (1993) Endocrinology 133, 1609 –1616 Yoshimura, Y., Nakamura, Y., Oda, T., Yamada, H., Nanno, T., Ando, M., Ubukata, Y., and Suzuki, M. (1990) Biol. Reprod. 43, 1012–1018 Dinchuk, J. E., Car, B. D., Focht, R. J., Johnston, J. J., Jaffee, B. D., Covington, M. B., Contel, N. R., Eng, V. M., Collins, R. J., and Czerniak, P. M. (1995) Nature 378, 406 – 409 Wang, H., Ma, W. G., Tejada, L., Zhang, H., Morrow, J. D., Das, S. K., and Dey, S. K. (2004) J. Biol. Chem. 279, 10649 –10658 Buccione, R., Vanderhyden, B. C., Caron, P. J., and Eppig, J. J. (1990) Dev. Biol. 138, 16 –25 Wang, H., Matsumoto, H., Guo, Y., Paria, B. C., Roberts, R. L., and Dey, S. K. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 14914 –14919 Joyce, I. M., Pendola, F. L., O’Brien, M., and Eppig, J. J. (2001) Endocrinology 142, 3187–3197 Segi, E., Haraguchi, K., Sugimoto, Y., Tsuji, M., Tsunekawa, H., Tamba, S., Tsuboi, K., Tanaka, S., and Ichikawa, A. (2003) Biol. Reprod. 68, 804 – 811 Ashkenazi, H., Cao, X., Motola, S., Popliker, M., Conti, M., and Tsafriri, A. (2005) Endocrinology 146, 77– 84 Conti, M., Hsieh, M., Park, J. Y., and Su, Y. Q. (2006) Mol. Endocrinol. 20, 715–723 Jamnongjit, M., Gill, A., and Hammes, S. R. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 16257–16262 Byskov, A. G., Andersen, C. Y., Nordholm, L., Thogersen, H., Xia, G., Wassmann, O., Andersen, J. V., Guddal, E., and Roed, T. (1995) Nature 374, 559 –562 Tsafriri, A., Cao, X., Ashkenazi, H., Motola, S., Popliker, M., and Pomerantz, S. H. (2005) Mol. Cell. Endocrinol. 234, 37– 45 Ochsner, S. A., Day, A. J., Rugg, M. S., Breyer, R. M., Gomer, R. H., and Richards, J. S. (2003) Endocrinology 144, 4376 – 4384 Ochsner, S. A., Russell, D. L., Day, A. J., Breyer, R. M., and Richards, J. S. (2003) Endocrinology 144, 1008 –1019 Vaux, D. L., Cory, S., and Adams, J. M. (1988) Nature 335, 440 – 442 Adams, J. M., and Cory, S. (1998) Science 281, 1322–1326 Kuida, K., Zheng, T. S., Na, S., Kuan, C., Yang, D., Karasuyama, H., Rakic,


P., and Flavell, R. A. (1996) Nature 384, 368 –372 39. Woo, M., Hakem, R., Soengas, M. S., Duncan, G. S., Shahinian, A., Kagi, D., Hakem, A., McCurrach, M., Khoo, W., Kaufman, S. A., Senaldi, G., Howard, T., Lowe, S. W., and Mak, T. W. (1998) Genes Dev. 12, 806 – 819 40. Hoshino, Y., Yokoo, M., Yoshida, N., Sasada, H., Matsumoto, H., and Sato, E. (2004) Mol. Reprod. Dev. 69, 77– 86 41. Kalous, J., Solc, P., Baran, V., Kubelka, M., Schultz, R. M., and Motlik, J. (2006) Biol. Cell 98, 111–123 42. Dekel, N., and Kraicer, P. F. (1978) Endocrinology 102, 1797–1802 43. Webb, R. J., Marshall, F., Swann, K., and Carroll, J. (2002) Dev. Biol. 246, 441– 454 44. Tsafriri, A., Chun, S. Y., Zhang, R., Hsueh, A. J., and Conti, M. (1996) Dev. Biol. 178, 393– 402 45. Conti, M. (2002) Biol. Reprod. 67, 1653–1661 46. Su, Y. Q., Denegre, J. M., Wigglesworth, K., Pendola, F. L., O’Brien, M. J., and Eppig, J. J. (2003) Dev. Biol. 263, 126 –138 47. Su, Y. Q., Wigglesworth, K., Pendola, F. L., O’Brien, M. J., and Eppig, J. J. (2002) Endocrinology 143, 2221–2232 48. Leonardsen, L., Wiersma, A., Baltsen, M., Byskov, A. G., and Andersen, C. Y. (2000) J. Reprod. Fertil. 120, 377–383 49. Shimada, M., Hernandez-Gonzalez, I., Gonzalez-Robayna, I., and Richards, J. S. (2006) Mol. Endocrinol. 20, 1352–1365 50. Chie, L., Amar, S., Kung, H. F., Lin, M. C., Chen, H., Chung, D. L., Adler, V., Ronai, Z., Friedman, F. K., Robinson, R. C., Kovac, C., Brandt-Rauf, P. W., Yamaizumi, Z., Michl, J., and Pincus, M. R. (2000) Cancer Chemother. Pharmacol. 45, 441– 449 51. Mood, K., Bong, Y. S., Lee, H. S., Ishimura, A., and Daar, I. O. (2004) Cell. Signal. 16, 631– 642 52. Richards, J. S. (2005) Mol. Cell. Endocrinol. 234, 75–79 53. Carrette, O., Nemade, R. V., Day, A. J., Brickner, A., and Larsen, W. J. (2001) Biol. Reprod. 65, 301–308 54. Fulop, C., Kamath, R. V., Li, Y., Otto, J. M., Salustri, A., Olsen, B. R., Glant, T. T., and Hascall, V. C. (1997) Gene (Amst.) 202, 95–102 55. Mukhopadhyay, D., Hascall, V. C., Day, A. J., Salustri, A., and Fulop, C. (2001) Arch. Biochem. Biophys. 394, 173–181 56. Yoshioka, S., Ochsner, S., Russell, D. L., Ujioka, T., Fujii, S., Richards, J. S., and Espey, L. L. (2000) Endocrinology 141, 4114 – 4119 57. Sato, H., Kajikawa, S., Kuroda, S., Horisawa, Y., Nakamura, N., Kaga, N., Kakinuma, C., Kato, K., Morishita, H., Niwa, H., and Miyazaki, J. (2001) Biochem. Biophys. Res. Commun. 281, 1154 –1160 58. Salustri, A., Garlanda, C., Hirsch, E., De Acetis, M., Maccagno, A., Bottazzi, B., Doni, A., Bastone, A., Mantovani, G., Beck Peccoz, P., Salvatori, G., Mahoney, D. J., Day, A. J., Siracusa, G., Romani, L., and Mantovani, A. (2004) Development (Camb.) 131, 1577–1586 59. Fulop, C., Szanto, S., Mukhopadhyay, D., Bardos, T., Kamath, R. V., Rugg, M. S., Day, A. J., Salustri, A., Hascall, V. C., Glant, T. T., and Mikecz, K. (2003) Development (Camb.) 130, 2253–2261 60. Varani, S., Elvin, J. A., Yan, C., DeMayo, J., DeMayo, F. J., Horton, H. F., Byrne, M. C., and Matzuk, M. M. (2002) Mol. Endocrinol. 16, 1154 –1167 61. Zhuo, L., Yoneda, M., Zhao, M., Yingsung, W., Yoshida, N., Kitagawa, Y., Kawamura, K., Suzuki, T., and Kimata, K. (2001) J. Biol. Chem. 276, 7693–7696 62. Tilley, S. L., Audoly, L. P., Hicks, E. H., Kim, H. S., Flannery, P. J., Coffman, T. M., and Koller, B. H. (1999) J. Clin. Investig. 103, 1539 –1545


37129