The second messenger cAMP has been implicated in the regulation of mammalian and amphibian oocyte maturation. Although a decrease in intraoocyte levels of cAMP precedes germinal vesicle breakdown (GVBD), the gonadotropin induction of ovulation and oocyte maturation is associated with major increases of cAMP in ovarian follicles. In the mammalian system, isolated oocytes undergo spontaneous maturation in vitro but this process is blocked by treatment with a phosphodiesterase (PDE) inhibitor, IBMX, which increases intraoocyte cAMP levels. In contrast, the same inhibitor, when added to cultured follicles for a brief time, increases follicle cAMP levels, followed by the induction of GVBD. To resolve the paradoxical actions of this PDE inhibitor on the maturation of isolated and follicle-enclosed oocytes, we hypothesized that meiotic maturation requires opposing fluctuations of cAMP levels in the somatic granulosa and germ cells. Such opposing fluctuations may result from selective expression and regulation of PDEs in the somatic and germ cell compartments of the follicle. To test this hypothesis, PDE activity was manipulated in different follicular cells using type-specific inhibitors. The impact of the ensuing changes in cAMP levels in the two compartments was monitored by the induction of GVBD. In isolated oocytes, spontaneous GVBD was blocked by two inhibitors of type 3 PDE (cGMP-inhibited: CGI-PDE), milrinone and cilostamide. In contrast, treatment with an inhibitor for type 4 PDE (cAMP-specific), rolipram, was ineffective. These findings suggest that the oocyte expresses type 3 but not type 4 PDE and that increases in intraoocyte cAMP suppress GVBD. This hypothesis was confirmed by in situ hybridization studies with PDE3 and PDE4 probes. PDE3B mRNA was concentrated in oocytes while PDE4D was mainly expressed in granulosa cells. In cultured follicles, LH treatment induced oocyte maturation but the gonadotropin action was blocked by inhibitors of type 3 but not the type 4 PDE inhibitors. Furthermore, treatment with the type 4, but not the type 3, PDE inhibitor mimics the action of LH and induces oocyte maturation, presumably by increasing cAMP levels in granulosa cells. Our findings indicate that PDE subtypes 4 and 3 are located in follicle somatic and germ cells, respectively. Preferential inhibition of PDE 3 in the oocyte may lead to a delay in oocyte maturation without affecting the cAMP-induced ovulatory process in the somatic cells. Conversely, selective suppression of granulosa cell cAMP-PDE may enhance the gonadotropin induction of ovulation and oocyte maturation. Thus, in addition to the well-recognized differential expression and regulation of adenylate cyclase in the somatic and germ cell compartments of the follicle, we suggest that selective regulation and expression of PDEs may be involved in the regulation of cAMP levels and control of oocyte maturation in the preovulatory mammalian follicle.

Previous studies showed that epidermal growth factor (EGF) and TGFalpha mimic the action of LH on the resumption of oocyte maturation. We tested whether EGF-like agents, such as amphiregulin (AR), epiregulin (ER), and betacellulin (BTC), also mediate the LH stimulation of the ovulatory response in the rat. LH induced transient follicular expression of AR, ER, and BTC mRNA, reaching a maximum after 3-h incubation. Furthermore, the addition of ER, AR, and BTC to the culture medium could mimic some of LH actions. AR and ER fully simulated LH-induced resumption of meiosis in vitro, whereas BTC was less effective. To study the putative involvement of EGF-like factors in mediation of LH signal, the effect of the EGF receptor kinase inhibitor AG1478 was tested. When added with LH, AG1478, but not its inactive analog AG43, reduced EGF receptor phosphorylation and oocyte maturation compared with follicles treated with LH only. In addition to the inhibition of resumption of meiosis, AG1478 administration into the bursa (3 microg/bursa) resulted in 51% (P < 0.0005) inhibition of ovulation in the treated ovaries, compared with the untreated contralateral ones, as well as to the vehicle-treated ovaries (P < 0.02). LH, as well as ER, induced the expression of genes associated with the ovulatory response like rat hyaluronan synthase-2, cyclooxygenase-2, and TNFalpha-stimulated gene 6 mRNA, whereas AG1478 inhibited this effect of LH. Release of EGF-like factors from the membrane is dependent on activated metalloproteases. Indeed, Galardin, a broad-spectrum metalloprotease inhibitor, but not a specific matrix metalloprotease 2 and 9 inhibitor, suppressed meiotic maturation induced by LH. Conversely, meiotic maturation induced by ER was not affected by Galardin, thus, supporting the notion that LH releases follicular membrane-bound EGF-like agents. In summary, EGF-like factors such as ER, AR, and BTC seem to mediate, at least partially, the LH stimulation of oocyte maturation, ovulatory enzyme expression, and ovulation.

No abstract

No abstract

For more than a half a century, biologists have upheld the theory that in most mammalian species, oocytes are formed before or shortly after birth, but never in adulthood. This foundation of reproductive science has survived the rapid growth of new technology and knowledge and has remained virtually unchallenged until two recent papers were published by the group headed by Jonathan Tilly. The first paper claims that mouse germline stem cells (GSCs) replace ovarian follicles that have been rapidly lost through follicle death (Johnson et al., 2004). The second paper, recently published in Cell, proposes continuous immigration into mouse ovaries of GSCs derived from bone marrow (Johnson et al., 2005). How could so many investigators have overlooked these basic mechanisms for so long, or have they? One of the major conclusions of their first paperthat GSCs reside in the surface epithelium of mouse ovaries and are active throughout life (Johnson et al., 2004)-has now been modified by the authors themselves in their second paper (Johnson et al., 2005) in response to several critical comments (Gosden, 2004; Albertini, 2004; Greenfeld and Flaws, 2004; Telfer, 2004). Indeed many of the experiments described in the first paper are open to alternate explanations, and independent corroboration of their conclusions has still to be obtained. Here we express concerns about their second paper regarding the putative role of bone marrow cells in reproduction and the experimental rigor needed to verify this revolutionary hypothesis. Johnson et al. (2005) claim that "adult mouse ovaries can produce hundreds of new oocytes within 24 hours" after follicle destruction by doxorubicin. Oogenesisthe process by which mitotic germ stem cells undergo meiosis to the diplotene stage followed by the formation of follicles-normally requires at least one week in the developing mouse ovary. It is astonishing if this process can be completed within 24 hr as Johnson et al. (2005) propose. To reconcile these observations, it is important to test whether pre-existing follicles can recover after doxorubicin treatment or whether new germ cells undergoing accelerated oogenesis can be identified and, in either case, whether the purported oocytes are capable of supporting fertilization and subsequent embryo development. Indeed it is central to their hypothesis that the rates of meiosis and of follicle formation are dramatically different and, therefore, it falls upon these authors to show this experimentally. The authors have also shown that molecular markers normally associated with germ cells (Oct4, Mvh, Dazl, Stella, and Fragilis) are expressed in blood and bone marrow and, even more surprising, that expression levels of these proteins vary during the mouse estrous cycle. All of these proteins have been shown to be expressed in other cells and organs including brain and

Isolated oocytes will resume meiosis spontaneously in vitro whereas follicle-enclosed oocytes will remain in the dictyate stage when cultured unless they have been exposed to gonadotropins in vivo or in vitro. To examine the source of the follicular inhibitory influence, porcine oocytes have been cultured alone, with hemisections of follicle wall, granulosa cells, or with follicular fluid. Oocytes isolated from medium-sized (3-5 min) follicles resumed meiosis when cultured; 77.5 plus or minus 3.4 percent matured beyond the dictyate stage. When oocytes were cultured in the presence of follicle wall hemisections of medium and large (6-12 mm) follicles, the percentage of maturing oocytes was significantly reduced. The maturation of oocytes cultured in a medium containing 50 percent follicular fluid from small or large follicles was significantly inhibited. Resumption of meiosis was completely inhibited by co-culture of isolated oocytes with 10-7 granulosa cells from small, medium or large follicles. Addition of serially diminishing amounts of granulosa cells from 10-7 to 10-4 cells reduced the inhibitory influence. It is concluded that the granulosa cells are responsible for the maintenance of the oocytes in the dictyate stage within the follicle. The granulosa cells appear to exert their inhibitory influence upon meiosis by secretion of a chemical message into follicular fluid.

It is generally accepted that cyclic nucleotides are key signaling molecules in the control of oocyte meiotic resumption. Given the role of phosphodiesterases (PDEs) in cyclic nucleotide degradation, this study was undertaken to investigate the properties and regulation of PDEs expressed in rat oocytes. Cilostamide-sensitive PDE3 was the major activity detected in denuded oocytes, whereas no PDE3 activity could be detected in cumulus cells. Moreover, comparable levels of PDE3 activity were measured in cumulus-oocyte complexes (COCs) and in denuded oocytes. The oocyte PDE was recovered in the soluble fraction of the homogenate and immunoprecipitated with a specific PDE3A antibody. A significant and transient increase (P < 0.05) in PDE3 activity was measured in the oocytes after 30 min of culture (70 min after isolation) compared with immediately after collection (10 min after isolation). Conversely, no changes in activity were observed when denuded oocytes or cumulus cells were incubated for up to 130 min. Evaluation of oocyte maturation indicated that only 10% of oocytes had resumed meiosis at the peak of the PDE3 activity. A significant increase (P < 0.05) in PDE3 activity was measured in COCs when follicle-enclosed oocytes were cultured in the presence of hCG. Again, this increase preceded oocyte maturation. In conclusion, these data demonstrate that PDE3A is the major PDE form expressed in mammalian oocytes. PDE3A activity increases prior to resumption of meiosis in both spontaneous and gonadotropin-stimulated maturation. These findings strongly support the hypothesis that an increase in oocyte PDE3A activity is one of the intraoocyte mechanisms controlling resumption of meiosis in rat oocytes, at least in vitro.

hCG is a member of a family of glycoprotein hormones which share a common alpha-subunit, but differ in their hormone-specific beta-subunits. The CG beta-subunit is unique in that it contains a hydrophilic carboxyl-terminal extension with four serine O-linked oligosaccharides. To examine the role of the O-linked oligosaccharides and the carboxyl-terminal extension of hCG beta on receptor binding, steroidogenesis in vitro, and ovulation induction in vivo, site-directed mutagenesis and gene transfer methods were used. Wild-type hCG alpha and hCG beta expression vectors were transfected into an O-glycosylation mutant Chinese hamster ovary cell line to produce intact dimer hCG lacking the beta-subunit O-linked oligosaccharide units. In addition, a mutant hCG beta gene (CG beta delta T) was generated which contained a premature termination signal at codon 115. This gene was cotransfected with the hCG alpha gene into Chinese hamster ovary cells to produce hCG dimer which lacked the carboxyl-terminal amino acids 115-145 of hCG beta (truncated hCG). The O-linked oligosaccharide deficient or truncated hCG derivatives were examined for their ability to bind to the mouse LH/hCG receptor and stimulate cAMP and steroidogenesis in vitro. These studies show that the O-linked oligosaccharides and carboxyl-terminal extension play a minor role in receptor binding and signal transduction. In contrast, comparison of the stimulatory effects of truncated and wild-type hCG in a rat ovulation assay in vivo via either intrabursal or iv injection revealed that the truncated derivative was approximately 3-fold less active than wild-type hCG. These findings indicate that the carboxyl-terminal extension of hCG beta and associated O-linked oligosaccharides are not important for receptor binding or in vitro signal transduction, but are critical for in vivo biological responses.