Naturally occurring steroids in Xenopus oocyte during meiotic maturation. Unexpected presence and role of steroid sulfates

Abstract

In the ovary, oocytes are surrounded by follicle cells and arrested in prophase of meiosis I. Although steroidogenic activity of follicle cells is involved in oogenesis regulation, clear qualitative and quantitative data about the steroid content of follicles are missing. We measured steroid levels of Xenopus oocytes and follicles by gas chromatography-mass spectrometry. We show that dehydroepiandrosterone sulfate is the main steroid present in oocytes. Lower levels of free steroids are also detected, e.g., androgens, whereas progesterone is almost undetectable. We propose that sulfatation is a protective mechanism against local variations of active steroids that could be deleterious for follicle-enclosed oocytes. Steroid levels were measured after LH stimulation, responsible for the release by follicle cells of a steroid signal triggering oocyte meiosis resumption. Oocyte levels of androgens rise slowly during meiosis re-entry whereas progesterone increases abruptly to micromolar concentration, therefore representing the main physiological mediator of meiosis resumption in Xenopus oocyte.

Introduction

In amphibian ovary like in all vertebrates, the prophase-arrested oocyte inside its follicle is surrounded by steroidogenic cells that are both endocrine and paracrine. They release steroid hormones that, after dilution and transport by the bloodstream, are recognized by classical steroid receptors in peripheral target cells where they primarily mediate the transcription of multiple genes. During the whole period of oogenesis, the growing oocyte is arrested in prophase of the first meiotic division inside its follicle envelope. How locally, inside the ovary, the steroidogenic signal released by follicle cells is received and regulated by the oocyte is unknown.

Following an extensive growth period, the follicle-enclosed oocyte reaches its full size (Dumont, 1972). A surge of the hypophyseal gonadotropin LH triggers the release of the oocyte into the oviduct, i.e., ovulation, and simultaneously promotes the transition from prophase I to metaphase II of the ovulated full-grown oocyte, known as meiotic maturation (Ryan and Grant, 1940, Wasserman and Masui, 1974). In amphibians, it is clearly established that LH acts on steroidogenic cells of the follicle envelope to stimulate both a systemic and a local release of steroids. This local steroid signal induces meiotic maturation of the full-grown oocyte (Masui, 1967, Smith and Ecker, 1969, Smith and Ecker, 1971). Hence, the Xenopus oocyte was established as a model system to study the signaling pathway induced by a steroid, but also to elucidate mechanisms controlling the interphase to M-phase transition that led to the identification of the M-phase Promoting Factor, MPF, initially known as Maturation Promoting Factor (Dunphy et al., 1988, Gautier et al., 1988, Masui and Markert, 1971). Steroid action on amphibian fully-grown oocyte is non-genomic as it does not require transcription (Masui and Markert, 1971, Wasserman and Masui, 1974, Wasserman and Smith, 1979). This model is therefore convenient to study post-transcriptional action of steroid hormones that does not require a nuclear step. Numerous studies focused on the transduction steps leading to MPF activation in Xenopus oocytes and provided a comprehensive view of this molecular cascade. It starts by a decrease of the cAMP concentration and PKA activity, allowing de novo synthesis of cyclin B1 and Mos kinase from stored mRNAs, and it ends with MPF activation (Haccard and Jessus, 2006). However, the identity of the steroid physiologically responsible for the release from the prophase arrest as well as its initial targets are still unknown.

More than 40 years ago, it was reported in Rana pipiens that progesterone is able to induce oocyte meiotic maturation (Masui, 1967, Schuetz, 1967, Smith et al., 1968). In Xenopus, the isolated prophase oocyte, released from its follicle envelopes, undergoes in vitro meiotic maturation after progesterone addition to the bathing medium (Jacobelli et al., 1974, Reynhout et al., 1975). This steroid is therefore widely used to induce oocyte maturation and is thought to be the physiological steroid responsible for triggering this process (Jacobelli et al., 1974, Reynhout et al., 1975). Moreover, the production of three main steroids, progesterone, testosterone and oestradiol, measured by radioimmunoassay, was reported to increase following the in vitro stimulation of Xenopus follicles by gonadotropin hormones (El-Zein et al., 1988, Fortune, 1983, Fortune et al., 1975, Fortune and Tsang, 1981, Lutz et al., 2001, Redshaw and Nicholls, 1971). Neither oestradiol nor other estrogens induce meiotic maturation and their physiological roles, if any, are unknown. It was thereafter reported that the main steroid produced quantitatively in response to LH is testosterone (El-Zein et al., 1988, Fortune and Tsang, 1981, Lutz et al., 2001) which induces in vitro meiotic maturation as efficiently as progesterone, as do many other C₁₉ and C₂₁ steroids (Baulieu et al., 1978). It is important to notice that all of these qualitative and quantitative results were obtained by radioimmunological assays. Despite the use of specific antibodies supposed to avoid cross-reactivity, analysis of steroid by radioimmunoassay, especially in lipid-rich oocyte extracts, must be considered as tentative and requires further molecular confirmation. Another inherent drawback of these radioimmunological approaches is that they do not allow the detection of new unexpected steroids.

The identity of steroid(s) liberated by follicle cells and how they reach the oocyte, either by simple diffusion or through gap junctions that connect follicle cells to the oocyte, remain unknown. These questions are of prime importance to guide future research devoted to the characterization of the steroid receptor(s) responsible for the induction of oocyte maturation. Direct biochemical approaches through the study of saturable progesterone binding sites in oocyte extracts, initiated more than 30 years ago (Ozon and Belle, 1973), were unsuccessful essentially due to the large stores of lipoproteins and lipids accumulated during growth of this giant cell. In contrast, the presence of mRNA coding different steroid receptors was discovered in Xenopus oocyte (Bayaa et al., 2000, Lutz et al., 2001, Tian et al., 2000). Although the mRNA of the classical nuclear progesterone receptor (nPR) is present in oocyte, the expression and the subcellular localization of its protein product were not convincingly demonstrated (Bagowski et al., 2001, Liu et al., 2005). More recently, a membrane bound progesterone receptor (mPR) has been cloned in zebrafish and the mRNA encoding this seven transmembrane domains receptor was subsequently identified in Xenopus oocyte (Josefsberg Ben-Yehoshua et al., 2007, Tokumoto et al., 2006, Zhu et al., 2003). Xenopus mPR binds progesterone with the affinity expected for oocyte maturation (Josefsberg Ben-Yehoshua et al., 2007), indicating that progesterone could be the physiological ligand of mPR. However, testosterone that is also produced by follicle cells and able to trigger oocyte maturation, does not bind mPR (Josefsberg Ben-Yehoshua et al., 2007). Numerous papers have suggested the existence of membrane androgen receptors in various systems (Rahman and Christian, 2007). Homologs of these have not yet been reported in Xenopus but may well be present. The Xenopus oocyte also contains mRNA encoding the classical nuclear testosterone receptor (nTR). Antisense oligonucleotides directed against this mRNA were reported to inhibit testosterone-induced maturation, indicative of a signaling function for this steroid in oocyte maturation (Lutz et al., 2001, Lutz et al., 2003, Sen et al., 2011). However, before concluding that non-genomic effects of testosterone signal through nTR in Xenopus oocyte, it will be necessary to further characterize this receptor at the protein level.

All in all, we know that steroidogenic follicle cells release both progesterone and testosterone. Both steroids are in vitro inducers of meiotic maturation of denuded Xenopus oocytes. We also know that the oocyte contains at least three mRNA coding for putative steroid receptors (nPR, mPR and nTR) that are potentially recruited to initiate the molecular cascade leading to MPF activation. A pending question is the nature of the true physiologically active steroid involved. Does it bind to a membrane-bound receptor localized on the external leaflet of the oocyte plasma membrane or does it enter into the oocyte to reach intracellular receptor(s), either associated with intracellular membranes or in the diffusible ooplasm? To clarify which steroids are released by follicle cells at the time of ovulation and could be implicated in oocyte meiosis re-entry, we quantified by gas chromatography coupled to mass spectrometry more than 40 steroids. A first result is that inside its follicle, the fully-grown prophase-arrested oocyte contains several free unconjugated steroids. Among them are testosterone, pregnenolone and oestradiol. A second observation is that unexpectedly the oocyte also contains two major steroid sulfates: pregnenolone sulfate (PREG-S) and dehydroepiandrosterone sulfate (DHEA-S). After follicle stimulation by LH, both progesterone and androgens levels are increased inside the oocyte. Interestingly, the production of steroid sulfates is also stimulated, albeit transiently. These analytical results prompted us to study the biological efficiency of the different free steroids found in the oocyte on meiotic maturation and to search for a possible function of oocyte hydroxysteroid sulfates in this process.