| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Minireview |
a MRC Group in Fetal and Neonatal Health and Development, Departments of Physiology and Obstetrics and Gynecology, University of Toronto, Toronto, Ontario, Canada M5A 1A8
b Department of Obstetrics and Gynecology, University of Ottawa, Ottawa, Canada K1H 8L6
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONOVINE PARTURITION STUDIESHUMAN PARTURITIONSTEROID HORMONE REGULATION OF...CONCLUSION: A COMPARATIVE...REFERENCES |
|---|
ACTH, adrenal cortex, anterior pituitary, cortisol, CRH, cytokines, decidua, estradiol, estradiol receptor, glucocorticoid receptor, hormone action, hypothalamus, NPY, oxytocin, parturition, pregnancy, progesterone, syncytiotrophoblast, trophoblast
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONOVINE PARTURITION STUDIESHUMAN PARTURITIONSTEROID HORMONE REGULATION OF...CONCLUSION: A COMPARATIVE...REFERENCES |
|---|
In most mammalian species studied toward the end of gestation and at the onset of labor, there is an increase in fetal plasma glucocorticoid concentration due to maturation and sustained activation of the fetal hypothalamic-pituitary-adrenal (HPA) axis [5]. Under normal conditions the fetal tissues are exposed to increasing levels of bioactive glucocorticoids in the final 10–15 days prior to delivery. These glucocorticoids induce maturational changes in the fetal organs such as the lungs, liver, kidneys, and gut in preparation for a successful transition from intra- to extrauterine life [5]. In addition, this surge in glucocorticoid has been suggested to be integral to the cascade of events leading to the onset of parturition [6]. Concurrent with the rise in fetal glucocorticoid concentration there is a progressive increase in plasma, amniotic fluid, and intrauterine tissue concentration of prostaglandin (PG), in particular PGE2 followed by an increase in PGF2
at the onset of labor [6]. These PGs have been identified as key mediators of the events of labor including cervical ripening, uterine contractility, membrane rupture, uteroplacental blood flow, and fetal adaptation to the process of labor [6]. Recent evidence has suggested that the rise in fetal glucocorticoid production directs the increase in intrauterine PG production and the onset of labor [6]. The purpose of this review is to examine the relationship between maturation of the fetal HPA axis, the increase in fetal glucocorticoid production, and intrauterine PG synthesis during late gestation and the onset of labor. We suggest that sustained activation of the fetal HPA axis leads to elevated fetal adrenal glucocorticoid production that both directly and indirectly (through a glucocorticoid stimulated intermediate) increases intrauterine PG production. We will review the current evidence in support of this hypothesis at the onset of ovine parturition and demonstrate that glucocorticoids both directly increase fetal placental PG production and indirectly increase PG production by maternal uterine tissues through the stimulation of placental estradiol synthesis. We will compare the events of ovine parturition with those of human parturition and provide evidence for the direct glucocorticoid regulation of PG production by the human fetal membranes and placenta. In the human, we will suggest a similar indirect effect of glucocorticoids on PG production mediated by glucocorticoid-stimulated intrauterine corticotropin releasing hormone (CRH) synthesis (Fig. 1). Finally, we will speculate on the direct molecular mechanism by which glucocorticoids could regulate PG production.
|
|
OVINE PARTURITION STUDIES |
|---|
|
TOPABSTRACTINTRODUCTIONOVINE PARTURITION STUDIESHUMAN PARTURITIONSTEROID HORMONE REGULATION OF...CONCLUSION: A COMPARATIVE...REFERENCES |
|---|
Fetal HPA Axis Activation
The critical role of the fetal HPA axis in the events of parturition was first established more than 30 yr ago, based on studies by Liggins and colleagues who recognized that pregnant ewes fed V. californicum at a particular time in pregnancy had prolonged gestation length and their fetuses had markedly hypoplastic pituitary and adrenal glands. Further in vivo studies found that fetal hypophysectomy or adrenalectomy, disruption of the fetal hypothalamic-pituitary stalk, and/or lesions of the paraventricular nucleus of the fetal hypothalamus similarly caused a prolongation of gestation length [8–10]. In addition, intrafetal administration of ACTH, cortisol, or the synthetic glucocorticoid dexamethasone was found to induce delivery of the ovine fetus prematurely [11–13]. These in vivo studies highlighted the critical role of the fetal HPA axis in the determination of gestation length and the timing of labor.
Recent studies have led to a greater understanding of the relationship between fetal HPA function and the onset of labor. Corticotropin-releasing hormone is produced by the parvocellular neurons of the hypothalamus; its secretion into the hypophyseal portal system drives the pituitary to synthesize proopiomelanocortin (POMC) and secrete ACTH. Fetal hypothalamic CRH mRNA and peptide expression is present by Day 60 of gestation (term 147–150 days); this expression increases gradually until Day 120 and then increases markedly over the final 20 days of gestation [14,15]. Corticotropin-releasing hormone acts at the pituitary via membrane bound receptors; CRH receptor subtype I (R1) has been identified within the ovine fetal pituitary (J.C. Rose, personal communication). The CRH-R1 receptor mRNA expression and CRH binding are present in the fetal pituitary by Day 100 of gestation, increase by Day 135, and then decrease toward term [16] (J.C. Rose, personal communication). These data suggest that hypothalamic CRH has its greatest effect on pituitary function from Days 120–135 of gestation, and the marked increase in CRH expression in late gestation may reflect a decreased pituitary sensitivity to CRH at that time. In addition, CRH and cortisol down-regulate CRH receptor expression and CRH binding; in late gestation CRH and cortisol levels are high and may contribute to the decreased sensitivity of the fetal pituitary to CRH [16]. In vivo, fetal pituitary responsiveness to intrafetal CRH infusion increases from Day 110 until Day 125 and then decreases until term; these findings are consistent with decreased pituitary CRH-R1 expression and CRH binding in late gestation [17]. Thus, it appears that although hypothalamic CRH expression continues to increase through late gestation and may initiate HPA activation, CRH action may be blunted at term and may not be the key to sustained function of the axis at this time.
Arginine vasopressin (AVP) is produced by the supraoptic and paraventricular nuclei (magnocellular and parvocellular neurons) of the hypothalamus; AVP production by the parvocellular neurons has been linked to the regulation of pituitary ACTH synthesis and secretion [18]. AVP mRNA and peptide are present in the parvocellular neurons of the fetal sheep pituitary gland by Day 60 of gestation; however, this expression does not alter with gestational age [18]. AVP acts via V1 receptors; the expression of these receptors in the fetal sheep pituitary throughout gestation has not been studied. In vitro, CRH and AVP increased POMC mRNA expression and ACTH output in a dose-dependent manner from cultured fetal sheep pituitary cells [18, 19]. AVP had a greater stimulatory effect than CRH in cells at Days 120–138 of gestation; at term the potency of CRH and AVP was similar. CRH and AVP together had an additive effect on fetal pituitary cell ACTH output [19]. Through the end of gestation, intrafetal administration of either AVP or CRH increased pituitary ACTH production; although responsiveness to CRH decreased and responsiveness to AVP increased as a function of gestational age [17, 20]. At Day 120 of gestation, AVP and CRH had a synergistic effect on ACTH production while at all other ages the effect was simply additive [20]. Therefore, the sustained increase in fetal pituitary ACTH synthesis and secretion through late gestation may be differentially regulated by both AVP and CRH. AVP and CRH may be involved in the early rise in ACTH synthesis and secretion. AVP may stimulate the rise in ACTH through the last 20 days of gestation. AVP and CRH effects, mediated by separate postreceptor mechanisms, may be additive and act to maintain the high levels of ACTH at term and with the onset of labor.
ACTH is produced by the corticotroph cells within the pars distalis and pars intermedia of the anterior pituitary. After cleavage of POMC by the action of prohormone convertase 1 (PC1), ACTH can then be cleaved by prohormone convertase 2 (PC2) to form smaller molecular weight products, corticotrophin-like intermediate lobe peptide (CLIP), and -melanocyte-stimulating hormone (-MSH). Within the ovine fetal pituitary, POMC mRNA is expressed by Day 60 of gestation; POMC expression in the pars intermedia increases progressively from Day 60 until Day 100 and then remains relatively constant through the end of gestation [15, 21]. Expression of POMC increases from Day 60 until Day 120 within the superior and inferior aspects of the pars distalis. After Day 120 of gestation, POMC expression is localized mainly to the inferior aspect of the pars distalis, and this expression is increased markedly by term, but no further increases are observed with the onset of labor [21, 22]. These increases in pars distalis POMC expression are reflected in increased immunoreactive ACTH expression within the corticotroph cells and increased circulating fetal plasma levels of ACTH(1–39) [17, 23]. The progressive rise in pituitary ACTH synthesis and secretion may be mediated by a change in the expression of PC1. The level of PC1 expression within the fetal pituitary increases toward the end of gestation suggesting that increased ACTH output could be attributable in part to increased cleavage of its precursor POMC [22]. In addition, the level of fetal pituitary PC2 does not change through gestation, suggesting that the ACTH produced is not increasingly cleaved to form -MSH and CLIP [22]. Thus, despite decreasing CRH responsiveness and maintained AVP action, POMC expression and ACTH secretion from the fetal pituitary continues to rise through the end of gestation and the onset of labor. The potential mechanism(s) of this sustained production and secretion will be discussed later.
In fetal sheep, plasma levels of ACTH(1–39) increase gradually from Day 110 of gestation and a further surge in output is observed with the progression of labor [17, 22]. The rise in plasma ACTH concentration is followed by sustained increases in fetal plasma cortisol levels approximately 10 days later [24]. Adrenocorticotropin hormone induces expression of the key adrenal steroidogenic enzymes involved in de novo cortisol synthesis, P450 side chain cleavage enzyme (P450scc) and P450 C17/C21 hydroxylase (P450C17). Therefore it has been suggested that increasing ACTH drives cortisol synthesis by adrenal cortical cells in late gestation [25, 26]. However, under conditions of low, constant plasma ACTH levels, adrenal cortisol production continued to increase through the end of gestation, suggesting that developmental changes occurring within the fetal adrenal also contribute to the rising cortisol levels in late gestation [27, 28]. Expression of ACTH receptor mRNA increases modestly after Day 130 of gestation and increases in response to stress such as hypoxemia [29]. Adrenocorticotropin hormone also increases its own receptor signal transduction by enhancing receptor coupling to adenylate cyclase, thereby facilitating increased adrenal sensitivity to ACTH through the end of gestation [30].
Several studies have examined the trigger(s) of fetal HPA activation and the mechanism(s) of sustaining this activation. Leptin, the adipocyte hormone, has been shown to reduce expression of neuropeptide Y (NPY) mRNA in the arcuate nucleus, and recent studies have suggested that NPY can stimulate HPA axis function by increasing CRH production. Thus, a decrease in leptin at the end of gestation would release inhibition of NPY and lead to HPA axis activation. Support for this possibility comes from a recent study in which intraventricular infusion of leptin to the late gestation ovine fetus caused a decrease in fetal plasma ACTH and cortisol concentrations. Presumably, reduced output of endogenous leptin would allow HPA activity to increase [31]. Oxytocin (OT) is expressed in the parvocellular and magnocellular neurons of the fetal hypothalamus, and OT expression in the hypothalamus increases with gestational age. Oxytocin increased ACTH output by cultured fetal pituitary cells, and the administration of CRH and OT together caused a greater ACTH output compared to either secretagogue alone. Thus, OT could play a role in HPA activation and exert synergistic effects with CRH in late gestation [32].
There are several mechanisms by which HPA activity could be sustained. Firstly, there appears to be an attenuation of the normal negative feedback regulation of HPA function [33]. Fetal plasma and pituitary levels of the high affinity cortisol binding protein, corticosteroid-binding globulin (CBG) increase through gestation. Corticosteroid-binding globulin may act to bind cortisol at the level of the pituitary and thereby limit the bioavailability of free cortisol to exert negative feedback [34–37]. The enzyme 11ßhydroxysteroid dehydrogenase (11ßHSD) that interconverts cortisol and its inactive metabolite cortisone is expressed within the fetal pituitary and preferentially metabolizes cortisol to cortisone, thereby reducing the effects of cortisol on pituitary ACTH production [38]. At present, however, the nature of fetal pituitary isoform(s) of 11ßHSD is unclear. Cortisol acts through glucocorticoid receptors (GR); fetal pituitary GR mRNA and cortisol binding can be detected from Day 60 of gestation and increases from Day 135 until term. Levels of GR mRNA in the anterior pituitary then decrease with the progression of labor [22, 39]. Fetal hypothalamic GR also declines in late pregnancy [32]. Collectively these data suggest that pituitary ACTH output may be incompletely inhibited by the increasing cortisol levels due to reduced hypothalamic and/or pituitary GR expression.
Estrogen may also influence activity of the fetal HPA axis. Intrafetal estradiol infusion increased hepatic CBG and 11ßHSD-I expression in late gestation [40]. Fetal plasma concentrations of estrone sulfate increase progressively during ovine pregnancy and in some studies physiologic increases in fetal plasma estrogen concentrations increase fetal plasma ACTH concentrations. It is suggested that this effect is mediated by estrogen-increasing hypothalamic CRH, and the CRH promoter contains an estrogen responsive element [41]. The fetal hypothalamus, hippocampus, and brainstem have been found to express high levels of estrogen sulfatase activity [42], suggesting that within the fetal brain, local estrogen production may increase and potentially influence HPA function. Furthermore, estrogen has been shown to decrease GR expression, suggesting that the changes in pituitary and/or hypothalamic GR expression in late pregnancy may reflect increased local estrogen production (unpublished results).
Prostaglandins, specifically PGE2, may also exert a positive feedforward effect on the fetal HPA axis that is not subject to negative feedback regulation. Plasma cortisol and PGE2 concentrations increase with similar time course in fetal sheep during late pregnancy [43, 44]. Intrafetal PGE2 infusion stimulated fetal plasma ACTH and cortisol production [45, 46]; inhibition of PGE2 production at both term labor and RU486-induced preterm labor decreased fetal plasma ACTH and cortisol levels [47–49]. Young et al. [49] showed that interruption of the hypothalamic-pituitary stalk blocked the effects of fetal PGE2 on the fetal HPA axis, suggesting that PGE2 exerted its effects at or above the level of the pituitary. In addition, PGE2 has been shown to increase adrenal cortical cell P450C17 expression and activity, suggesting that PGE2 could directly increase de novo adrenal cortisol synthesis [50]. Clearly, the mechanisms regulating the sustained activation of fetal HPA activity through the end of gestation are complex and multifactorial, reflecting a complex interplay between the developmental and endocrine changes occurring at this time (Fig. 2).
|
Intrauterine Events of Parturition
A current hypothesis of ovine parturition suggests that cortisol produced by the fetal adrenal acts to induce the expression of the enzyme P450C17 in fetal placental trophoblast tissue, which diverts pregnenolone away from the production of progesterone toward the production of C19 steroids required for placental estradiol synthesis. The ovine placenta also uses C19 precursors produced by the fetal adrenal for estradiol synthesis throughout gestation [33]. However, at the end of gestation, the increased expression and activity of P450C17 leads to additional local production of C19 precursors in the placenta, causing a surge in placental estradiol production and also a decline in placental progesterone secretion [51, 52].
In turn, it has been suggested that estradiol increases the expression and activity of PGHS-II within the intrauterine tissues leading to the production of two key PGs, PGE2 and PGF2. These PGs act to increase uterine contractility, induce cervical ripening, regulate uteroplacental blood flow, and modulate fetal adaptation to the events of labor. Increased estradiol output also triggers the expression of a cassette of CAPs such as connexin 43, OT-receptor, and PG receptors. These CAPs allow the myometrium to contract in synchronous, polar fashion in response to contractile stimulants such as PGs and OT. However, closer examination of the timing of these endocrine changes during late pregnancy suggests that this may not be the correct sequence of events that lead to the onset of labor and delivery of the ovine fetus.
Fetal plasma PGE2 levels gradually increase from Day 120 of gestation, and this rise in PGE2 concentration occurs with a similar time course to the rise in fetal plasma cortisol [53]. It clearly precedes the late-gestation sharp rise of maternal and fetal plasma estradiol levels [53]. The placental trophoblast expression of PGHS-II increases before the increase in placental P450C17 mRNA and protein and before plasma estradiol changes [54]. Moreover, intrafetal estradiol infusion did not increase expression of PGHS-II mRNA in the sheep placenta [6], although intrafetal administration of cortisol did increase placental PGHS-II expression and plasma PGE2 concentrations [11]. These observations suggest that cortisol, and not estradiol, might increase placental trophoblast PGHS-II expression. The pronounced increase in maternal plasma 13,14-dihydro-15-keto-prostaglandin F2 (PGFM) concentration occurs with the onset of labor, well after the progressive rise in fetal plasma PGE2 and correlating with increasing endometrial PGHS-II expression and uterine contractility [53]. Estradiol has been shown to increase PGHS-II expression significantly in nonpregnant ovine myometrium and in nonpregnant ovine endometrium after progesterone priming [55]. These observations suggested that placental estradiol may stimulate nontrophoblast intrauterine tissue PGHS-II expression/activity to produce PGF2, that in turn may contribute to uterine activity.
On the basis of this evidence, we hypothesized that there may be two separate pathways of intrauterine PG production: a cortisol-dependent/estradiol-independent pathway within placental trophoblast (fetal) tissue and an estradiol-dependent pathway within maternal intrauterine tissues. To test this hypothesis, we infused late gestation sheep fetuses with cortisol in the presence and absence of the aromatase inhibitor 4-hydroxyandrostenedione (4-OHA) and determined changes in placental and uterine PGHS-II expression and PG output. Intrafetal cortisol infusion increased expression of PGHS-II mRNA and protein in placental trophoblast and levels of PGE2 in fetal plasma. These changes occurred even in the presence of 4-OHA [56]. Thus, these events did not depend on increased placental estrogen output. However, maternal endometrial epithelial PGHS-II mRNA and protein increased only in the presence of increases in both fetal plasma cortisol concentration and placental estradiol production [56]. Recently, the presence of the glucocorticoid receptor within the fetal mononuclear trophoblast cells and the absence of the estrogen receptor (ER) from these cells has been reported. These observations preclude a direct estradiol effect on PGHS-II expression within the fetal tissue and are consistent with GR-mediated glucocorticoid regulation of PGHS-II within the fetal placenta [57, 58]. The ER has been identified within maternal endometrium in late gestation supporting a role for estrogen in the regulation of endometrial PGHS-II [58].
These data suggest that at the end of gestation there are two separate pathways of PG production. Fetal adrenal cortisol leads to the induction of fetal placental PGHS-II expression and PGE2 output and placental estradiol production induces maternal endometrial PGHS-II expression and PGF2 output contributing to uterine activity. To link these two pathways of PG production we have suggested that placental PGE2 acts in an autocrine/paracrine manner to upregulate the expression of placental P450C17, thereby driving the terminal surge in placental estradiol production. This suggestion is supported by studies showing that PGE2 increased P450C17 in cultured bovine adrenal cells and that PGF2 increased P450C17 expression in bovine preovulatory cells [50, 59]. Prostaglandin E2 acts through membrane-bound receptors linked to either adenylate cyclase leading to increased cAMP production or to inositol triphosphate and increased intracellular Ca2+ [60]. Expression of P450C17 is regulated by activation of a cAMP response element within the 5' promoter sequence suggesting that PGE2 could provoke an increase in intracellular cAMP and stimulate a cAMP response element (CRE)-mediated increase in P450C17 activity [61]. Thus, PGE2 could act as a glucocorticoid-stimulated intermediate that elicits an autocrine/paracrine upregulation of P450C17 within fetal trophoblast cells of the placenta leading to the surge in placental estradiol synthesis observed in the final days of gestation. Cortisol-stimulated PGE2 may also function as a positive mediator of fetal HPA activation, as previously discussed, thereby creating a positive feedforward loop between fetal HPA activity and intrauterine PG production [45, 62].
Myometrial Events at Parturition
At the onset of labor, the uterine activity pattern evolves from long duration, low amplitude, and low frequency contractures to high frequency, high amplitude, and short duration contractions associated with increased intrauterine pressure. Myometrial contractility has been defined as occurring in two distinct yet overlapping phases: uterine activation and uterine stimulation [7]. Uterine activation involves the upregulation of CAPs in myometrium. These proteins allow the myometrium to respond to contractile stimulants in a coordinated fashion; CAPs include the gap junction protein connexin 43 (Cx-43), the OT receptor, the PG receptors (EP isoforms 1–4 and FP), the ER, and protein constituents of ion channels, particularly Ca2+ channels [7]. Uterine stimulation refers to the synchronous, contractile response of the uterus to uterotonins including PG and OT [7]. Garfield et al. [63] first demonstrated that gap junctions form with increased size and number in the myometrium of pregnant sheep at term and with parturition. Since that time, Cx-43 has been identified as the subunit protein of the gap junction. The expression of Cx-43 in ovine myometrium increases with the onset and progression of ACTH-induced labor [7]. Oxytocin receptor and ER mRNA have been identified in the late gestation myometrium; their expression did not change with betamethasone-induced preterm labor [58]. However, with the onset of term labor, mRNA transcripts for the ER, its associated heat shock proteins (hsp70 and hsp90), and the OT receptor were increased [55].
Regulation of CAP expression and myometrial contractility has been linked with the increase in plasma estrogen:progesterone (E2:P4) ratio [7]. The effects of steroids on CAP gene expression are, in turn, dependent on changes in myometrial stretch [64]. Thus, in ovariectomized rats treated with maintenance levels of estrogen, uterine distention increases numbers of transcripts encoding Cx-43, OT receptors, and FP receptors, but this action is blocked by concurrent administration of progesterone [65–67]. In unilaterally pregnant rats, distention of the contralateral (nonpregnant) horn results in similar levels of Cx-43 mRNA as the control horn on Day 22, the time of labor, but not at Day 18 when maternal plasma progesterone concentrations are still high. These studies help explain the higher incidence of preterm labor in women with multiple pregnancies and greater uterine distention. It also relates CAP expression to fetal growth and may explain the more gradual rise in Cx-43 gene expression that occurs in the human uterus during late pregnancy. Our own recent measurements indicate that PGHS-II in intrauterine tissues can also be considered as a CAP gene. Administration of nimesulide, a PGHS-II inhibitor, not only decreases plasma PG concentrations but also leads to decreased expression of mRNA for hsp70 and OT receptor [55]. Thus, the final stage of the labor cascade involves the complex interaction of the fetal and placental endocrine signals leading to a maternal contractile response. The precise regulation of these events continues to be examined.
Summary
Based on the evidence presented, we propose the following model for the onset of parturition in sheep (Fig. 3). Toward the end of gestation there is a gradual and sustained increase in the placental trophoblast expression of PGHS-II expression and PGE2 production under the regulation of increasing fetal cortisol. Placental PGE2 in turn mediates an autocrine/paracrine increase in placental P450C17 expression/activity resulting in a surge in placental estrogen production that is superimposed on a gestation-dependent increase in estrogen output. Placental PGE2 is also secreted into the fetal compartment where it acts to sustain fetal HPA axis activation. Estrogen up-regulates the expression of maternal endometrial PGHS-II, leading to increased PGF2 output and induces the expression of CAPs. Consequently, myometrial contractility is stimulated by an orderly progression of events from fetus to fetal-placenta to maternal uterine tissues.
|
|
HUMAN PARTURITION |
|---|
|
TOPABSTRACTINTRODUCTIONOVINE PARTURITION STUDIESHUMAN PARTURITIONSTEROID HORMONE REGULATION OF...CONCLUSION: A COMPARATIVE...REFERENCES |
|---|
Prostaglandin output is reflected in increased PG synthetic capacity by intrauterine tissues and increased concentrations of PGs and/or metabolites in amniotic fluid, maternal plasma, and urine at labor, term and preterm. It is now clear that these changes precede clinical evidence of increased uterine activity [79] rather than occurring as a consequence of the labor process [80]. Administration of PG synthase inhibitors suppresses uterine activity and prolongs gestation, whereas exogenous PGs are used to stimulate uterine activity and enhance cervical ripening at any stage of gestation. Recent studies have suggested linkages between fetal HPA activation (or elevated glucocorticoid) and regulation of enzymes resulting in increased PG output in human fetal membranes, and this is discussed in more detail below.
Glucocorticoid-Mediated PG Synthesis
The intrauterine tissues, myometrium, placenta, decidua, and fetal membrane, express the enzymes to produce PGs from membrane phospholipids at term; however, it is the fetal membranes that have been identified as the main site of PG production at the end of gestation and the onset of labor [6]. The fetal membranes are composed of the amnion, the innermost layer, which is a single layer of epithelial cells overlying a thick layer of connective tissue embedded in which are fibroblasts (mesenchymal cells) and the chorion trophoblast layer that lies adjacent to the maternal decidua [81]. Both at term and preterm labor, PGHS-II expression within the amnion epithelium and mesenchyme and the chorion trophoblast layer increases leading to an increased output of PGE2 and PGF2 [82–85] (Fig. 4). Primary culture preparations have been used to study the regulation of amnion and chorion PG synthesis. Glucocorticoids have been identified as positive regulators of amnion PG synthesis, and preliminary studies suggest that glucocorticoids may similarly affect chorion PG synthesis [86–90]. This stimulatory effect of glucocorticoids in the amnion was found to be receptor dependent, linked to the activity of protein kinases, and involved increased PGHS-II mRNA and protein expression [87–90].
|
Amnion culture preparations may contain both epithelial and mesenchymal cells in varying proportions [87]. Hence the cell type responsible for glucocorticoid-mediated changes in PGHS-II has been sought. In a mixed amnion cell culture preparation, glucocorticoid exposure upregulated PGHS-II mRNA and immunoreactive protein in the amnion mesenchymal cells but not in the amnion epithelial cells [87]. Recently, purified amnion epithelial and mesenchymal cell preparations have been established [86, 91]. Both amnion epithelial and mesenchymal cells expressed immunoreactive PGHS-II and produced PGE2 in culture. Both amnion cell types expressed immunoreactive GR in intact fetal membrane preparations [86, 92]. However, results are conflicting regarding the basal and glucocorticoid-stimulated PGE2 production and PGHS-II expression [86, 91]. Prostaglandin E2 output by amnion epithelial cells but not mesenchymal cells was limited by substrate availability, suggesting that epithelial cell PG output may be more strictly regulated by phospholipase activity [86]. Although amnion epithelial cells produce less PGE2 per cell than mesenchymal cells, there are 7–10 times more epithelial cells than mesenchymal cells in the amnion at term, such that the relative contribution of each cell type to the increased pool of PGs is difficult to determine [81]. Whittle et al. [86] found that purified epithelial cells responded better to glucocorticoid with increased PGE2 output, whereas Blumenstein et al. [91] found that the mesenchymal cells responded to added glucocorticoid, as in mixed amnion cell cultures. The different responses of the epithelial cells in mixed versus purified cultures following glucocorticoid treatment suggests that there may be paracrine interactions between the amnion epithelial and mesenchymal cells involved in the regulation of PG production.
These reports of glucocorticoid-mediated amnion cell PG production are in conflict with studies examining PG production by WISH cells, an immortalized amnion cell line. Du Val et al. [93] reported that dexamethasone decreased PGHS-II expression and activity by amnion WISH cells. Other studies have reported that dexamethasone decreased cytokine-induced PGHS-II expression and activity by these cells through interference with transcription factor binding at the NF
B and/or CRE 5' promoter sequences of the PGHS-II gene [94–96]. The potential mechanism(s) of the glucocorticoid-mediated increase in PGHS-II expression and activity will be discussed later. These results may also reflect culture conditions. Glucocorticoids stimulated PGHS-II in cultured amnion cells that had been in culture for more than 3 days but decreased PGHS activity in freshly dispersed amnion cells [97]. Resolution of these different responses will be most helpful in understanding the mechanism of glucocorticoid effects on PGHS-II.
Regulation of PGHS-II expression and PGE2 output by human fetal membranes is clearly multifactorial, and an impressive list of potential regulatory agents including cytokines, growth factors, and activators of cAMP has been generated from in vitro experiments [98–103]. Extrapolation of these results to the in vivo situation should be made with caution. For example, cytokines such as interleukin (IL)-1ß and IL-6, mediators of the inflammatory response associated with infection-driven preterm labor, increase levels of mRNA encoding phospholipase A2 and PGHS-II in amnion cells and in decidua [104–106]. Interleukin-1 also stimulates the output from decidua of other cytokines including IL-6 and IL-8 [105, 106], thereby establishing a positive cytokine-PG cascade. Cytokines provoke release, in in vitro experiments, of other uterotonins including CRH from decidua, membranes, and placenta [107]. However, IL-10 an anti-inflammatory cytokine also produces a modest stimulation of PGE2 output and of PGHS-II mRNA levels in human amnion and chorion, but IL-10 antagonizes the upregulation of PGHS-II provoked by IL-1ß [108] Thus, the balance of locally generated cytokines in vivo may be critical in determining the tissue response. Interestingly, administration of IL-10 to pregnant rhesus monkeys prevented the induction of uterine contractions following IL-1ß intra-amniotic [109], suggesting that there may be a balance between these locally acting substances in regulating eicosanoid biosynthesis.
Glucocorticoid-Mediated PG Metabolism
NAD+-dependent prostaglandin dehydrogenase (PGDH), is the key enzyme responsible within the intrauterine tissues for the metabolism of E and F series PGs, producing biologically inactive PG metabolites [110]. Immunoreactive PGDH has been localized to the chorion trophoblast cells of the fetal membranes and to the syncytiotrophoblast of the villous placenta in tissue collected during the second half of human gestation [111]. The PGDH activity and mRNA was lower in cultured chorion trophoblast cells and chorion trophoblast tissue explants collected from patients in labor at term compared to tissue from women not in labor. However, there was no change in the expression/activity of PGDH within placental trophoblast cells with labor [112]. Chorionic PGDH mRNA, protein, and activity were decreased in approximately 15% of idiopathic preterm labor patients and was decreased further in patients with infection-associated preterm labor due to an inflammatory degradation of the chorion trophoblast layer [113, 114]. These data suggest that PGDH expression may limit the production of intrauterine PGs through gestation; however, at term and preterm labor, a decrease in chorionic PGDH expression/activity could contribute to the net increase in intrauterine PG production (Fig. 4).
Recent evidence suggests that progesterone produced by the placental and chorionic trophoblast cells acts in an autocrine/paracrine manner to maintain PGDH expression and activity in these cells and thereby provides a mechanism for sustained intrauterine PG metabolism through gestation [115, 116]. Cortisol and dexamethasone decreased the PGDH mRNA and activity of both chorion and placental trophoblast cells in vitro and therefore increased PG output by these cells [116]. The effect of cortisol was reversed by the addition of exogenous progesterone, suggesting that at term increasing cortisol concentrations may compete with progesterone in the regulation of PGDH; the resultant effect is a net decrease in PGDH expression leading to an overall increase in intrauterine PG production [116]. This interaction appears to occur at the GR rather than mineralocorticoid receptor or progesterone receptor (unpublished data). Glucocorticoids may also be generated locally in the fetal membranes (see below), raising the possibility of regional paracrine regulation of PGDH in a manner that would overcome stimulation of the enzyme by progesterone.
Role of 11ßHSD
We have argued above that cortisol plays a central role in the regulation of PG production by intrauterine tissues during late pregnancy and at term. Cortisol could be derived from sustained secretion by the fetal HPA axis or released acutely in response to fetal stress. It could be derived from the maternal adrenal, or it could be generated locally within the fetal membranes through the activity of the enzyme 11ßHSD. This enzyme belongs to the short-chain alcohol dehydrogenase class of enzymes capable of interconverting cortisol and its inactive metabolite cortisone. There are two major isoforms, 11ßHSD-I that is bidirectional, but generally operates as a reductase converting cortisone to cortisol, and 11ßHSD-II that acts predominantly as a dehydrogenase [117]. The 11ßHSD-II is expressed in the kidney, where it metabolizes cortisol and prevents saturation by glucocorticoids of mineralocorticoid receptors and in the placenta where it regulates transplacental transfer of cortisol from mother to fetus. This activity not only protects the fetus from the high concentrations of cortisol in the maternal compartment but also determines the level at which fetal HPA function is modulated by maternally derived cortisol. Regulation of syncytiotrophoblast 11ßHSD-II is multifactorial. Based on an elegant series of in vivo studies in the baboon, the 11ßHSD-II oxidative system has been established to be present within the placental syncytiotrophoblast cells, and its activity is increased by estrogen [118, 119]. At term an increase in placental 11ßHSD-II activity results in a greater conversion of maternal cortisol to cortisone and reduces the suppression of fetal HPA axis by maternal cortisol. Thus, there is an increase in fetal ACTH production and maturation of the fetal adrenal gland [118, 119]. In vitro studies with primary cultures of human placental trophoblast showed that 11ßHSD-II was also upregulated by activators of protein kinase (PK)C and PKA; forskolin, for example, increased the activity and mRNA of 11ßHSD-II severalfold. Conversely, 11ßHSD-II mRNA and protein was decreased by nitric oxide donors and by progesterone [120]. Recent studies have shown that placental 11ßHSD-II activity is decreased by both PGE2 and PGF2 in a Ca2+-dependent manner [121]. Thus, increased PG output, at term and with the onset of labor, would decrease cortisol inactivation in the placenta, potentially making available more cortisol to affect PG synthetic and metabolic enzymes. This relationship represents a possible positive feedforward regulatory loop involving placental cortisol production and intrauterine PG output (Fig. 5I).
|
The 11ßHSD-I mRNA and protein have been localized to the human fetal membranes, in particular to the chorion trophoblast cells, decidua, amnion epithelium, scattered amnion mesenchymal cells, and to the endothelium of placental blood vessels. The highest level of expression was found within the chorion [122]. The enzyme was found to function almost exclusively in the reductase direction to produce cortisol from cortisone [122]. Addition of cortisone to cultured chorion trophoblast cells decreased PGDH activity and PGDH mRNA; this effect was blocked in the presence of carbenoxolone, an inhibitor of 11ßHSD-I activity [123]. In contrast to the placental trophoblast 11ßHSD-II, cultured chorionic trophoblast cell 11ßHSD-I activity was increased by both PGE2 and PGF2 [122]. Thus, chorion trophoblast cells can utilize inactive cortisone to produce cortisol that may then act in an autocrine/paracrine fashion to decrease PGDH mediated PG metabolism within the chorion and to increase PGHS activity. This relationship represents a second possible feedforward regulatory loop involving intrauterine cortisol bioavailability and PG output (Fig. 5II). A similar inter-relationship with PGHS-II may occur in amnion. Here, locally formed cortisol could act in a paracrine fashion to increase amnion PGHS-II expression/activity and PGE2 synthesis. In turn, amnion PG production may influence chorionic 11ßHSD-I activity. The 11ßHSD-I enzyme expression is also present within the amnion epithelial and mesenchymal cells; amnion epithelial cells treated with cortisone had increased PGE2 output; this effect was blocked by coincubation with carbenoxolone [124]. Thus, these preliminary data suggest the possibility of a third positive feedforward loop within the amnion epithelium linking PG synthesis and cortisol metabolism (Fig. 5III).
Role of CRH
In addition to the direct effects of cortisol on PG production, cortisol may indirectly increase fetal membrane and placental PG production through the stimulation of intrauterine CRH production. Intrauterine CRH is homologous to hypothalamic CRH and is produced by the fetal membranes and placental syncytiotrophoblast cells; CRH expression can be detected by 10 wk of gestation and increases progressively through normal pregnancy [125]. Maternal plasma levels of CRH follow the pattern of increased placental CRH mRNA expression; however, CRH does not appear to affect the maternal HPA axis because it is bound in the maternal circulation to a specific CRH binding protein (CRH-BP). The expression of CRH-BP also increases through normal gestation [126, 127] but then decreases in the 3–4 wk prior to the onset of labor at term or preterm. Total CRH and free (active) CRH increases exponentially over that time consistent with CRH acting as a trigger for the events of parturition [126]. Cortisol increases CRH mRNA and CRH output in vitro [127, 128]; in vivo, the antenatal administration of maternal glucocorticoid increased maternal plasma concentration and intrauterine tissue expression of CRH [129, 130]. Placental CRH may affect myometrial activity directly through multiple receptor subtypes or indirectly through production of uterotonins such as PGs. In human myometrium CRH receptor subtypes 1, 1ß, 2, and variant c are expressed in both the smooth muscle cells and the connective tissue; however, CRH receptor subtypes 1 (CRH-R1) are the predominant forms; levels of mRNA encoding this protein are higher in myometrium from patients in labor either at term or preterm [131]. Corticotropin-releasing hormone acts through this receptor to 1) increase cAMP that inhibits myosin light chain kinase activity and 2) decrease myometrial PGE2 output. Therefore, CRH mediates uterine relaxation. In late gestation, uterine activity becomes regionalized; the lower segment relaxes to form the birth canal and the fundal region remains contractile providing the expulsive force required for labor. Hillhouse and coworkers [132, 133], in an elegant series of experiments, have suggested that at term increased output of OT stimulates PKC activity resulting in phosphorylation of the CRH receptor. This decreases the receptor affinity for CRH and the receptor coupling to the Gs regulatory protein resulting in a diminished relaxation effect of CRH. In contrast, Hillhouse and coworkers have also reported that CRH can increase the myometrial contractility in response to PGF2 and OT. Corticotropin-releasing hormone may contribute to stimulation of the uterine fundal region by increasing the output of PGE2 and PGF2 from the fetal membranes [128]. This activity appears to increase both increased expression of PGHS-II mRNA and decreased PG metabolism in chorion, through attenuation of PGDH activity, an effect that is presumably mediated by cAMP. It is also possible that at term CRH continues to mediate myometrial quiescence contributing to the relaxation of the lower uterine segment. Thus, CRH may play a dual role regulating both uterine quiescence and contractility; the effects of CRH on the uterus during pregnancy and at term are complex, depend upon regional variation in receptor subtype distribution and activity, and its interaction with uterotonins such as OT and PG.
Maternal CRH concentrations are elevated and CRH-BP levels are decreased in some patients in preterm labor, although this depends on the stage of pregnancy and the underlying cause of threatened preterm birth [134–138]. One cause may be stress, of either maternal or fetal origin. The former may be psychosocial stress. Hobel et al. [138] reported that increases in maternal CRH levels at 18–20 wk and 28–30 wk gestation were associated with maternal stress levels at 18–20 wk gestation. The fetus responds to a compromised intrauterine circumstance, such as hypoxemia with upregulation of the fetal HPA axis and increased output of cortisol from the transitional zone of the fetal adrenal gland [139]. Cortisol stimulates placental CRH output, through a mechanism that requires activation of an upstream cAMP response element. Corticotropin-releasing hormone may act locally in the placenta as a vasodilator to increase blood flow [140]. It also acts directly on the fetal zone of the fetal adrenal gland to increase output of DHEA, the C19 estrogen precursor, thereby promoting uterine activation [141]. We have suggested that sustained elevation of placental CRH output promotes PG production from the fetal membranes and effects stimulation of the activated myometrium, particularly in the fundal region [7]. Therefore, within the fetal membranes and placenta, glucocorticoids may indirectly mediate PG production through the stimulation of intrauterine CRH production. This relationship represents a fourth feedforward loop in a glucocorticoid-stimulated cascade of events within the intrauterine tissues facilitating PG production (Fig. 5IV). In addition, intrauterine CRH production at the onset of labor may feed back on the fetal HPA axis to promote cortisol production (Fig. 5).
Summary
A decrease in chorion trophoblast PGDH expression/activity and an increase in amnion and chorion PGHS-II expression/activity would allow a net increase in membrane PG production. These PGs may act at the membrane to provoke membrane rupture or pass through the membrane to act at the underlying cervix and myometrium to facilitate cervical ripening and myometrial contractility. A decrease in placental PG metabolism at term would permit an increase in net local PG output capable of maintaining uteroplacental blood flow during myometrial contractility and allow PGs to enter the fetal circulation to exert endocrine effects (Fig. 4). We suggest that these changes in intrauterine PG production are under direct and indirect glucocorticoid regulation. Intrauterine glucocorticoid bioavailability is a consequence of sustained fetal HPA activation and alterations in the tissue-specific level of intrauterine 11ßHSD activity.
|
STEROID HORMONE REGULATION OF PG PRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONOVINE PARTURITION STUDIESHUMAN PARTURITIONSTEROID HORMONE REGULATION OF...CONCLUSION: A COMPARATIVE...REFERENCES |
|---|
Glucocorticoids are lipophilic hormones that freely diffuse across the cell membrane into the cytoplasm and ultimately translocate to the nucleus where they exert genomic effects. There is some evidence to suggest that glucocorticoid cell entry may also be regulated by specific membrane-associated receptors that bind glucocorticoids and mediate rapid nongenomic glucocorticoid effects through a G-protein signaling system [145]. Once in the cytoplasm, glucocorticoids bind to their specific receptor (GR). Two isoforms of the receptor exist, GR and GRß [143]. However, only the GR isoform is capable of altering gene expression [143]. The ligand-receptor complex translocates to the nucleus, becomes phosphorylated, and dimerizes. The GR-ligand homodimer binds to a GRE to induce transcription of the target gene through interaction with the basal transcription apparatus and/or to induce a conformational change in the chromatin structure thereby increasing the accessibility of other regulatory factors [143, 145]. In addition to the classical mode of steroid hormone receptor action, protein-protein interactions between GR and other transcription factors can occur [145, 146]. This mode, referred to as crosstalk, involves the interaction of the GR-ligand complex with the AP-1 transcription factor family. The interaction between AP-1 factors and GR-ligand complex can be through direct protein-protein interaction, indirect through bridging proteins or through competition for a limiting cofactor [143, 145, 146]. The resultant interaction is either a non-DNA-bound transcription factor linked to a DNA-bound-transcription factor or the sequestration of two transcription factors into a non-DNA-binding complex. This transcription factor crosstalk can occur at either of the transcription factor's DNA binding sites, at an adjacent nonrelated transcription factor DNA binding site, or at a composite response element that is an overlap of the transcription factor's DNA binding sites [143, 146]. The net effect of crosstalk is either transcription factor synergism or negative interference of gene expression; the effect is dependent upon specific cell conditions including cell differentiation and stage of the cell cycle [143, 145, 146]. Recently, transcription factor crosstalk has been demonstrated to occur between nuclear hormone receptors and the NFB transcription factors, CRE, C/EBP, or the octamer binding factors in addition to GR and the AP-1 family. Thus, glucocorticoids can exert genomic and nongenomic influences through a variety of mechanisms; the absence of a GRE or nGRE on a given gene does not preclude glucocorticoid regulation of that gene [143, 145, 146].
The gene for PGHS-II is an acute response gene; its expression is undetectable in most cells but can be rapidly upregulated by cytokines, mitogens, hormones, and endotoxin [144]. Studies using an amnion cell line have described the induction of PGHS-II mRNA, protein, and PG synthesis within 30 min of cytokine exposure [147]. Traditionally, glucocorticoids have been classified as potent anti-inflammatory agents due to their effective inhibition of stimulated PGHS-II expression. The PGHS-II 5' promoter region does not contain an nGRE, and the mechanism of glucocorticoid inhibition has been postulated to be mediated through interference with an enhancer protein, inhibition by NFB, and/or a decrease in PGHS-II mRNA stability [142]. We have provided evidence for the upregulation of PGHS-II gene expression in response to glucocorticoids within fetal trophoblast-derived cells. Zakar et al. [88] demonstrated that the positive effect of glucocorticoids on PGHS activity within trophoblast-derived human amnion cells was receptor-dependent, did not require a cortisol-stimulated protein intermediate, and involved an increase in PGHS-II mRNA level that was not due to increased mRNA stability. A similar increased PG output and enhanced metabolism of arachidonic acid in response to glucocorticoid stimulation has been reported in fetal rat lungs, rat gastric mucosa, murine fibroblasts, bone marrow-derived myeloid leukemia cells, and rat renomedullary cells [148–152]. This stimulatory effect was characterized as dependent upon glucocorticoid concentration and cell differentiation. The 5' promoter of PGHS-II contains a full GRE at -726–717 base pairs that could mediate an increase in gene expression [153, 154]. Alternatively, given the recent evidence supporting the synergistic interaction of transcription factors, GR could interact with a member of the AP-I or C/EBP families at either of the respective transcription factor DNA binding sites or a composite response element. Alternatively, GR could interfere with a repressor transcription factor or synergize with a unique fetal trophoblast cell transcription factor to mediate a trophoblast cell-specific PGHS response to glucocorticoid. However, the trophoblast cells in vitro do retain an anti-inflammatory response to glucocorticoid administration. Several investigators have reported that glucocorticoids inhibit cytokine (IL-1ß, tumor necrosis factor , epidermal growth factor)-induced PGHS-II expression and this effect is mediated through negative interference with transcription factor binding at the two NFB promoter sites, the CRE and a third as yet unidentified enhancer site [154–156]. Glucocorticoids have also been shown to increase the expression of IB, the inhibitor protein of NFB, suggesting a possible dual mechanism of glucocorticoid inhibition of NFB activity [157]. We suggest that the proinflammatory effects of glucocorticoid occur by the upregulation of basal state PGHS-II expression, whereas the anti-inflammatory effects of glucocorticoids occur through the downregulation of the induced state of PGHS-II expression. Thus, glucocorticoid can exert differential effects depending on the state of the cell activity.
This hypothesis may explain in part the differential effects of glucocorticoids on freshly dispersed human amnion cells versus cultured cells [97]. Freshly dispersed amnion cells produce large quantities of proinflammatory cytokines as well as large quantities of PGE2. Administration of glucocorticoids at this time causes a decrease in PGE2 production [97]. As time in culture progresses, the production of the proinflammatory cytokines decreases as does the production of PGE2 [158]. At this time the administration of glucocorticoid causes an increase in PGHS-II expression and PGE2 output [87–90]. We suggest that freshly dispersed cells have an increased level of PGHS-II expression and activity induced by autocrine/paracrine action of the proinflammatory cytokines; glucocorticoids cause a characteristic inhibition of induced PGHS-II expression. As time in culture progresses, the cytokine induction of PGHS-II is attenuated, and PGHS-II expression decreases to a basal state capable of reinduction by glucocorticoids. Thus, depending upon the basal level of PGHS-II expression, glucocorticoids can exert a biphasic effect in cultured amnion cells. Clearly, glucocorticoid regulation of PGHS-II gene expression and enzyme activity is a complex interplay of cellular events unique to each cell type studied.
|
CONCLUSION: A COMPARATIVE ANALYSIS OF OVINE AND HUMAN PARTURITION |
|---|
|
TOPABSTRACTINTRODUCTIONOVINE PARTURITION STUDIESHUMAN PARTURITIONSTEROID HORMONE REGULATION OF...CONCLUSION: A COMPARATIVE...REFERENCES |
|---|
|
FOOTNOTES |
|---|
1 Correspondence: W.L. Whittle, Department of Physiology, University of Toronto, 1 Kings College Circle, Toronto, ON, Canada M5A 1A8. FAX: 416 978 4940; wendy.whittle{at}utoronto.ca 
Accepted: October 20, 2000.
Received: June 28, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONOVINE PARTURITION STUDIESHUMAN PARTURITIONSTEROID HORMONE REGULATION OF...CONCLUSION: A COMPARATIVE...REFERENCES |
|---|
