The hormonal control of parturition

Abstract

Parturition is a complex physiological process that must occur in a reliable manner and at an appropriate gestation stage to ensure a healthy newborn and mother. To this end, hormones that affect the function of the gravid uterus, especially progesterone (P4), 17β-estradiol (E₂), oxytocin (OT), and prostaglandins (PGs), play pivotal roles. P4 via the nuclear P4 receptor (PR) promotes uterine quiescence and for most of pregnancy exerts a dominant block to labor. Loss of the P4 block to parturition in association with a gain in prolabor actions of E₂ are key transitions in the hormonal cascade leading to parturition. P4 withdrawal can occur through various mechanisms depending on species and physiological context. Parturition in most species involves inflammation within the uterine tissues and especially at the maternal-fetal interface. Local PGs and other inflammatory mediators may initiate parturition by inducing P4 withdrawal. Withdrawal of the P4 block is coordinated with increased E₂ actions to enhance uterotonic signals mediated by OT and PGs to promote uterine contractions, cervix softening, and membrane rupture, i.e., labor. This review examines recent advances in research to understand the hormonal control of parturition, with focus on the roles of P4, E₂, PGs, OT, inflammatory cytokines, and placental peptide hormones together with evolutionary biology of and implications for clinical management of human parturition.

CLINICAL HIGHLIGHTS.

Parturition is controlled by hormones acting in an endocrine and/or paracrine manner to either maintain pregnancy (e.g., progesterone) or transform the uterine myometrium and cervix into the labor state (e.g., inflammatory and stress-associated signals). Understanding of the cell and molecular biology underlying the action and regulation of specific progestation and prolabor hormones involved in human parturition is revealing novel therapeutic strategies (e.g., boosting the progesterone block and/or preventing its withdrawal) to optimize the clinical control of birth timing (e.g., prevent pre- and postterm birth).

1. INTRODUCTION

Birth of a viable offspring through the process of parturition is a defining event for reproduction in viviparous species. Parturition is fundamentally a function of uterine activity. As the site for internal gestation, also a defining trait of viviparity, the uterus accommodates the conceptus (products of conception): the developing embryo/fetus with its placenta, extraembryonic membrane, and amniotic fluid. For most of pregnancy the muscular uterine wall, the myometrium, is composed of hypertrophied myometrial smooth muscle cells that are generally relaxed, producing minimal tone with only sporadic and weak asynchronous contractions. The myometrium is also compliant, distending to accommodate the increasing volume of the developing conceptus. At the same time, the uterine outlet, the cervix, is composed of stroma with a dense extracellular matrix (ECM) and secretory endocervical mucosa that forms a rigid and closed barrier to contain the conceptus. The inner endometrial lining of the uterus, referred to as the decidua, interfaces directly with trophoblast cells of the conceptus and establishes a local microenvironment that is immunologically quiescent to tolerate the allogeneic fetal cells.

At parturition, the myometrium transforms to become the engine for birth. Myometrial cells become electrically coupled and produce forceful, synchronized, and rhythmic contractions that force the conceptus against the cervical outlet. At the same time, collagen fibrils in the cervix ECM remodel to soften the stroma, allowing the uterine outlet to dilate in response to increased intrauterine pressure produced by myometrial contractions. In the decidua, immune/inflammatory quiescence transforms to tissue-level inflammation with the infiltration of activated maternal immune cells that produce factors, especially prostaglandins (PGs), that weaken the adjacent amnion membrane, leading to its rupture, and induce myometrium contraction and softening of cervical stroma. Myometrial contractions coupled with cervix dilation and membrane rupture eventually force the fetus, and then the placenta with the attached amnion-chorion membrane and decidua membrane, through the cervix, bony pelvic outlet, and vaginal canal. These uterine tissue-specific paracrine and endocrine events are reinforced by the neuroendocrine release of oxytocin (OT), which further propagates uterine contractions to facilitate labor and delivery (for review see Refs. 1, 2) (FIGURE 1).

FIGURE 1.

Endocrine, paracrine, and neuroendocrine interactions controlling parturition. Multiple physiological factors (some examples shown) contribute signals impacting uterine tissues that are integrated into a putative inflammatory load. In a threshold-limited manner, inflammatory load induces progesterone (P4)/P4 receptor (PR) withdrawal leading to loss of the P4/PR block to labor, which includes increased local inflammation due to loss of P4/PR anti-inflammatory activity, leading to increased local production of inflammatory cytokines, some of which [e.g., prostaglandin (PG)F_2α] are potent uterotonins. P4/PR withdrawal also increases expression of ESR1 [encodes estrogen receptor-α (ERα)], allowing circulating estrogens (mainly estradiol) to increase response of uterine tissues to uterotonins [especially PGF_2α and oxytocin (OT)] and increase extracellular matrix (ECM) breakdown in the cervix and fetal membrane, leading to cervix softening and membrane weakening, respectively. A neuroendocrine positive feedback loop involving OT amplifies contractions at the time of active labor: phasic myometrium contractions, cervix softening and dilation, and membrane rupture.

Transformation of the gravid uterus from quiescence to active labor is controlled primarily by hormonal interactions involving two classes of hormones, uterotonins, typified by PGs and OT, that directly affect myometrium contraction and uterotropins, primarily the steroid hormones progesterone (P4) and estradiol (E₂), that promote uterine growth and produce a phenotype conducive to pregnancy maintenance or parturition (FIGURE 2). Here we review recent advances in the understanding of how these hormones control the process and timing of parturition.

FIGURE 2.

The majority of pregnancy is a quiescent state, with low myometrial contractile activity, maintained by progesterone (P4), which is the dominant hormonal effector of ongoing gestation. The progression from quiescent pregnancy to latent (early) labor is facilitated by a functional progesterone withdrawal and increased estradiol (E₂) effect. Uterotonins [prostaglandins (PGs) and oxytocin (OT)] are the predominant effectors of active labor and postpartum uterine involution. Image is an adaptation of material from Ref. 3.

2. PROGESTERONE

The hormonal control of pregnancy and parturition is dominated by P4. In all species, P4 supplied by luteinized granulosa cells in the ovarian corpus luteum (CL) modifies the endometrium to create a site conducive to establishing a uterine pregnancy. In some species (e.g., mouse, goat), the maternal CL persists throughout pregnancy and is the sole source of P4 to maintain pregnancy. In other species (e.g., sheep, human), the placenta also produces P4 (4, 5). In women, the placenta is the principal source of P4 during the second and third trimesters (4, 6, 7).

In all species, actions of P4 to maintain pregnancy are mediated by the nuclear P4 receptor (PR) isoforms, PR-A and PR-B, that function primarily as ligand-activated transcription factors. PR-A and PR-B share amino acid sequence except that PR-B has an extra 164 NH₂-terminal residues (8). Both isoforms are detectable by immunostaining in myometrial, cervical, and decidual cells in the gravid uterus of all species studied so far, and in each cell type the PRs localize exclusively to the nucleus, consistent with their function as ligand-activated transcription factors. The importance of P4/PR signaling for the maintenance of pregnancy is demonstrated by numerous clinical and animal studies showing that disruption of P4/PR signaling with PR antagonists, such as RU486 (aka mifepristone) and onapristone, induces labor and the full parturition cascade at all stages of pregnancy (9, 10).

In the 1950s Arpad Csapo (11) proposed the “progesterone block” hypothesis, which posits that for most of pregnancy P4 actively blocks labor. Consistent with the genomic mode of steroid hormone/nuclear receptor action, P4/PR is thought to maintain pregnancy and block parturition by affecting expression of specific genes in uterine cells during pregnancy (12). P4/PR is considered to affect expression of genes encoding factors collectively referred to as contraction-associated proteins (CAPs). Studies in vitro and in vivo (mainly in mice and rats) suggest that P4/PR inhibits expression of the genes that encode procontraction CAPs (e.g., GJA1 that encodes connexin-43, which electrically couples myometrial cells to synchronize contractions; PTGS2 that encodes the rate-limiting enzyme for the synthesis of bioactive PGs that promote myometrial cell contraction; OXTR that encodes the OT receptor to increase responsiveness to OT) and increases expression of genes encoding prorelaxation factors (e.g., adrenergic receptor subtypes that mediate relaxation in response to adrenaline) (13–15). P4/PR also decreases responsiveness to E₂ in uterine cells by inhibiting expression of ESR1 that encodes estrogen receptor-α (ERα) (16, 17). In myometrial cells E₂/ERα increases expression of prolabor CAP genes and in cervix stromal cells expression of genes encoding metalloproteinases that break down collagen, leading to softening and dilation of the cervix (18–22). These effects are consistent with the “seesaw hypothesis” proposed by Arpad Csapo in the 1970s, which posits that the labor state of the gravid uterus is controlled by the balance between the P4 block and prolabor stimuli, especially E₂, OT, and PGs (23). For most of pregnancy the P4/PR block prevails.

Studies in transgenic mice have revealed specific roles of PR-A and PR-B in murine pregnancy and parturition. Mice lacking both PRs (24–27) develop normally to adulthood; however, female mice are infertile because of neuroendocrine disorders and multiple defects in the reproductive tissues including endometrial hyperplasia and inflammation, defective mammary gland development, and anovulation. Female mice lacking only PR-A (i.e., having only PR-B) are also infertile, with a reproductive phenotype similar to that of the mice lacking both PRs. However, PR-B mediates a normal response to P4 in the mammary epithelium in the PR-A-null mice, suggesting that PR-B is sufficient for mammary development and lactation whereas PR-A is critical for uterine function and neuroendocrine control of reproduction (28, 29). In contrast, mice lacking only PR-B (i.e., having only PR-A) have normal pregnancy and parturition but defective mammary gland morphogenesis and lactation (28). The data suggest that PR-A alone is sufficient to mediate P4 actions to establish and maintain murine pregnancy.

With regard to cell type-specific roles for P4/PR in pregnancy maintenance, recent studies suggest that P4/PR signaling in myometrial cells is not necessary for pregnancy and parturition (30). Mice lacking PR expression in smooth muscle cells, which included uterine myometrial cells, had markedly decreased fertility, primarily due to failure of embryos to transit from the oviducts to the uterine lumen. Even though myometrial cell PR deletion impaired adaptation of the myometrium to pregnancy, in the few pregnancies established by embryos that managed to enter the uterus, implantation, pregnancy, and parturition were normal. This suggests that, at least in mice, P4/PR signaling in myometrial cells is not necessary for pregnancy establishment and maintenance. Interestingly, in decidual stromal cells P4/PR signaling was not affected by myometrial cell PR deletion, suggesting that the decidua in mice is the key uterine tissue mediating the P4/PR actions necessary for pregnancy. Taken together, data from PR isoform ablation experiments in mice provide novel insights into the cell/tissue-specific roles of PR-A and PR-B in pregnancy and parturition, with PR-A appearing to play a major role. Whether this applies to pregnancy in other species, especially human, is unclear.

Siiteri and colleagues (31, 32) in the 1970s proposed that P4 maintains pregnancy by exerting anti-inflammatory actions. It is now generally accepted, on the basis of clinical and animal data, that parturition is an inflammatory event associated with increased expression in uterine tissue of genes encoding proinflammatory cytokines and activation and increased infiltration of neutrophils and macrophages into each of the uterine compartments (33–39). Animal and clinical studies also show that parturition is induced by inflammatory stimuli such as intrauterine infection and chorioamnionitis (40–45). Multiple lines of evidence suggest that P4/PR blocks labor, at least in part, by inhibiting the transcriptional response of uterine cells and tissues to prolabor/proinflammatory stimuli (27, 46–53). Thus, data in the last 20–30 years justify updating Csapo’s seesaw model for the hormonal control of parturition to include inflammation and proinflammatory hormones as prolabor factors that are counterbalanced by anti-inflammatory actions of P4/PR.

The NF-κB transcription factor complex is a key mediator of inflammatory stimuli and is activated in the uterine tissues in association with the onset of labor (54–56). Several lines of evidence in a variety of cell types indicate that P4/PR exerts anti-inflammatory activity by inhibiting NF-κB. In kidney fibroblast cells P4/PR inhibits NF-κB activity by physically interacting (albeit weakly) with the P65 subunit of NF-κB (57). In myometrial cells, P4/PR increases expression and decreases the degradation of IκBα that inhibits NF-κB by binding to and sequestering the NF-κB complex to the cytoplasm (47, 58) and by inducing expression of dual-specificity phosphatase 1, which decreases NF-κB translocation to the nucleus (57, 58). Some studies suggest that P4/PR indirectly decreases NF-κB activity by affecting other transcription factors and/or coregulators (59–61). Studies of DNA promoter occupancy showed that in human myometrial cells P4/PR-B decreases interleukin (IL)-1β-induced interaction of NF-κB with cognate response elements in the PTGS2 promoter (47). Thus, the anti-inflammatory actions of P4/PR appear to be promoter specific. Recent transcriptome analyses in myometrial cells suggest that P4/PR inhibition of NF-κB involves the activator protein-1 (AP-1) transcription factor complex. In a human myometrial cell line and in human term myometrium, P4/PR anti-inflammatory activity does not affect all IL-1β/NF-κB-responsive genes, but instead promoter motif analysis shows that genes whose promoters contain response elements to both NF-κB and AP-1 were enriched in the cohort of IL-1β-responsive genes inhibited by P4/PR (51). Studies in other cell types show that AP-1 and NF-κB function in synergy to mediate gene transcription in response to proinflammatory stimuli (62, 63). In myometrial cells, AP-1 is thought to be a master transcription factor for the expression of genes encoding key CAPs, especially GJAI (64, 65). Importantly, PR-A and PR-B each physically interact with AP-1 subunits in myometrial cells in an isoform- and ligand-dependent manner (61). PR-A interacts with the FOS subunits, whereas PR-B has preference for JUN homodimers. PR-A-FOS in a ligand-independent manner induces expression of GJA1, whereas P4/PR-B-JUN inhibits GJA1 expression. According to this model, for most of pregnancy P4/PR-B via interaction with JUN blocks parturition by inhibiting CAP gene expression. Whether this applies to the expression of NF-κB-dependent inflammatory genes is not known.

P4/PR may promote uterine quiescence indirectly by modulating expression of specific transcription factors that affect the expression of CAP genes. This mechanism is exemplified by the observation that in myometrial cells P4/PR increases expression of the zinc finger E-box-binding transcriptional repressor ZEB1 that decreases expression of key contractile CAP genes (66). In addition, P4/PR may increase expression, either directly or via ZEB1, of specific miRNAs, especially miR-199a-3p and miR-214, that repress CAP and inflammatory gene expression (67). At parturition, P4/PR withdrawal (discussed below) would decrease ZEB1, leading to increased miR-200 that further decreases ZEB1 and ZEB2 to remove ZEB inhibition of CAP genes. Loss of P4/PR would also decrease levels of the anti-inflammatory miR-199a-3p and miR-214. This regulatory network may be critical in establishing the labor state of the myometrium.

Regulation of the inflammatory state in the gravid uterus by P4/PR is not limited to myometrium. In decidual stromal cells P4/PR induces expression of IL15, which is thought to promote the differentiation of resident uterine natural killer cells to produce an immune quiescent cellular microenvironment that is tolerant of the allogeneic conceptus (68–70). In decidual stromal cells IL-1β inhibits expression of IL-15, and this is prevented by P4/PR (70).

Thus, taken together current data demonstrate that P4/PR blocks labor via multiple pathways leading to the inhibition of CAP gene expression and the prevention of tissue-level inflammation (FIGURE 3).

FIGURE 3.

Signaling pathways by which progesterone (P4)/P4 receptor (PR) blocks labor. P4/PR via decidual cell expression of IL-15 promotes immune quiescence at the maternal-fetal interface. P4/PR directly inhibits expression of ESR1, which decreases levels of estrogen receptor-α (ERα), leading to refractoriness to estradiol (E₂)/ERα-induced expression of contraction-associated protein (CAP) genes. P4/PR increases expression of ZEB1, whose downstream effects inhibit CAP and inflammatory cytokine gene expression. P4/PR interacts physically with the activator protein-1 (AP-1) transcription factor complex to inhibit activity of AP-1 and NF-κB, leading to a broad anti-inflammatory effect and inhibition of AP-1-driven CAP genes. For most of pregnancy P4/PR effects (green) dominate. STAT5b, signal transducer and activator of transcription 5b; 20αHSD, 20α-hydroxysteroid dehydrogenase.

3. PROGESTERONE WITHDRAWAL

Just as the P4 block to labor dominates the hormonal control of pregnancy maintenance, P4 withdrawal is a major trigger for parturition. Loss of the P4/PR block alone is sufficient to induce the full parturition cascade; and this appears to be a conserved trait among viviparous species as evidenced by the induction of labor by RU486 in essentially all viviparous studied to date. This implies that the uterus would be poised for parturition at all stages of pregnancy were it not for the P4/PR block. In this system the mechanism for P4/PR withdrawal and how it is induced is central to the hormonal regulation of parturition.

P4 withdrawal can occur systemically, locally, or functionally (FIGURE 4). Although the mechanism of P4 withdrawal that triggers term parturition in a particular species is considered that species’ primary mechanism for the hormonal control of parturition, it is now clear that the various P4 withdrawal modalities may coexist and function synergistically or in response to specific physiological states. For example, although systemic P4 withdrawal triggers parturition in mice, a local and/or functional P4 withdrawal appears to operate in response to parturition induced by inflammation (71).

FIGURE 4.

Modes of progesterone (P4) withdrawal. Systemic P4 withdrawal occurs by P4 metabolism in the source tissue [e.g., corpus luteum (CL) via 20α-hydroxysteroid dehydrogenase (20αHSD) or placenta via P450c17]. Similarly, local withdrawal occurs by P4 metabolism by 20αHSD in uterine target cells to prevent P4 from interacting with P4 receptors (PRs). Functional P4 withdrawal occurs by changes in PR transcriptional activity caused by interaction with specific repressors, transrepression by phosphorylated PR-A, and/or direct activation of contraction-associated protein (CAP) gene expression by liganded PR-A. A4, androstenedione; ERα, estrogen receptor-α; 20αOHP, 20α-hydroxyprogesterone.

3.1. Systemic P4 Withdrawal

Systemic P4 withdrawal is caused by decreased production of P4 from the CL or placenta. In both tissues reduced P4 release is caused by increased P4 metabolism. In mice and rats parturition is preceded by increased activity in CL cells of the 20α-hydroxysteroid dehydrogenase (20αHSD) enzyme, a member of the aldoketo reductase family that converts P4 to 20α-hydroxyprogesterone (20αOHP) which lacks progestin activity (72–74). Loss of P4 removes negative feedback on the hypothalamus-pituitary, which increases follicle-stimulating hormone that increases E₂ production by developing follicles, which leads to the postpartum estrus typical of rodents. In sheep increased expression in placental trophoblast cells of Cyp17a1 that encodes the 17α-hydroxylase, 17/20 lyase (P450c17) enzyme, decreases P4 release by converting it to androstenedione (A4) (75). A4 is then converted to E₂ in placental cells, leading to a coordinated increase in maternal E₂ levels. Thus, induction of 20αHSD in the CL (mice) and P450c17 in the placenta (sheep) each cause systemic P4 withdrawal coupled with an increase in E₂. The net effect of P4 withdrawal and increased E₂ drives the uterine tissues to the labor state.

In human pregnancy, P4 is supplied initially by the CL and subsequently (during the later two-thirds of pregnancy) by the placenta. Unlike the sheep, human placental cells do not express CYP17A1 and continue to produce P4 before and during parturition, with maternal P4 levels decreasing only after delivery of the placenta. Thus, human parturition occurs without systemic P4 withdrawal.

3.2. Local P4 Withdrawal

Although parturition in most species is preceded by a systemic P4 withdrawal, the level of circulating total P4 is not completely abolished, and in some species such as mice the levels are within the range that establishes pregnancy (76). This suggests the existence of other P4 withdrawal mechanisms. One of those mechanisms is local P4 withdrawal mediated by P4 metabolism in PR-expressing target cells.

The core principle of local P4 withdrawal is that P4 entering target cells is prevented from interacting with PRs by conversion to a PR-inactive form. This may be mediated by the 20αHSD enzyme that also functions in CL cells to decrease the release of P4 (61, 71, 77). In mice, systemic P4 withdrawal is caused by increased 20αHSD in CL cells, and this is induced by PGF_2α (78, 79). Although parturition in mice lacking 20αHSD is generally delayed, because of persistent CL P4 production, in some 20αHSD-deficient mice parturition was prolonged even though circulating P4 levels declined, suggesting that systemic P4 withdrawal is not sufficient to precipitate parturition (80, 81). Local P4 withdrawal via 20αHSD-mediated P4 metabolism in the uterine tissues may be a mechanism to compensate for insufficient systemic P4 withdrawal. Thus, 20αHSD acts as a gatekeeper to metabolize P4 to 20αOHP before it accesses the PRs. The 5α-reductase enzymes also may contribute to P4 metabolism in cervical stromal cells (82). Recent studies show that expression of AKR1C1, the gene that encodes 20αHSD, increases in human term myometrium in association with the onset of labor (83). This finding is consistent with earlier studies showing that in term myometrium the level of 20αOHP relative to P4 increases with labor (84, 85). These data support the hypothesis that 20αHSD-mediated local P4 withdrawal occurs in the human parturition process. Thus, local P4 metabolism in conjunction with a decline in circulating P4 levels and possibly altered PR activity (see below) represents redundancy to ensure a P4/PR withdrawal that is sufficient to initiate parturition.

The regulation of AKR1C1 expression leading to altered 20αHSD activity is a critical element in the hormonal control of parturition. Interestingly, this is linked to the P4/PR-ZEB1 signaling network. Expression of AKR1C1 is inhibited by the signal transducer and activator of transcription (STAT)5b transcription factor (81, 86). STAT5b-deficient mice have increased CL 20αHSD activity leading to lower than normal circulating P4 levels and increased incidence of midgestation pregnancy loss (81). Interestingly, the STAT5B promoter is a target for repression by miR-200 (77), whose levels are inhibited by P4/PR via ZEB1 (see above). Thus, a plausible model is that for most of pregnancy the P4/PR-ZEB1 interaction suppresses miR-200, allowing expression of STAT5B, leading to production of STAT5b that then suppresses AKR1C1, leading to low levels of 20αHSD allowing P4 to interact with PRs in target cells, leading to sustained suppression of AKR1C1. At parturition, loss of P4/PR removes the ZEB1-mediated inhibition of miR-200, leading to miR-200-mediated inhibition of STAT5B and loss of STAT5b repression of AKR1C1, leading to increased 20αHSD that prevents P4 from accessing PRs. This regulatory network contains positive feedback loops such that a small decrease in P4/PR activity is amplified to cause a significant local loss of P4 via 20αHSD metabolism. This local reduction in P4 increases miR-200-mediated repression of ZEB1 and ZEB2 that releases ZEB-mediated inhibition of CAP gene expression and response to inflammatory stimuli to transform the uterus to the labor state.

3.3. Functional P4 Withdrawal

Functional P4 withdrawal may occur by several mechanisms, each involving modulation of PR transcriptional activity. In response to P4 binding PR-A and PR-B each affect the expression of specific and overlapping gene sets in a cell type-specific manner, with the full-length PR-B having more robust P4-induced transcriptional activity than PR-A. In most cell types examined so far, including human myometrial cells, PR-A decreases P4/PR-B transcriptional activity in a concentration (based on immunoblot assay)-dependent manner (59, 87, 88). This transrepressive action of PR-A is thought to be a mechanism to regulate response to P4 especially when PR-A exceeds PR-B. This effect led to the PR-A/PR-B hypothesis for functional P4 withdrawal, which posits that an increase in the PR-A-to-PR-B ratio in uterine cells causes functional loss of the P4/PR block to labor. The hypothesis is supported by studies in myometrial cell lines, in term myometrium tissue, and in mice genetically modified to alter the PR-A-to-PR-B ratio in uterine cells (49, 50, 52, 89–93). Taken together the data suggest that the P4/PR block to labor is mediated by PR-B and that this is antagonized by increased abundance and transrepressive activity of PR-A. Some studies suggest that functional P4/PR withdrawal is also mediated by inhibition of PR-B transcriptional activity via decreased levels and activity of critical PR transcriptional coactivators (94, 95) and by reduced interaction of PR-B with DNA elements in uterine cells (96).

A core tenet of the PR-A/PR-B hypothesis is increased abundance and transrepressive activity of PR-A in uterine cells. Recent studies of epigenetic marks in the distal PGR promoter (leads to full-length mRNA encoding PR-B) and the proximal PGR promoter (leads to a truncated mRNA encoding PR-A) show that term human myometrium levels of histone modifications are higher in the proximal PR-A promoter compared with the distal PR-B promoter (89, 97, 98). This may explain why PR-A levels increase in myometrium with advancing gestation and in association with term labor whereas PR-B levels remain relatively constant (90).

PR-A abundance and activity may also be affected by posttranslational modifications. The PR isoforms can be phosphorylated at multiple serine residues depending on cell type and physiological context that affect PR stability and transcriptional activity (99). Studies in endometrial and mammary cancer cells show that that IL-1β via the p38 MAPK-induced pSer increases the stability of PR-A (100). Likewise, in human myometrial cells IL-1β increases the transrepressive activity and stability of PR-A (49, 50). These findings suggest that inflammatory stimuli induce PR-A-mediated functional P4/PR withdrawal in myometrial cells. This concept was expanded by the finding that phosphorylation of PR-A at the serine-344/345 locus (pSer^344/345-PRA) is induced by IL-1β, is necessary for PR-A transrepression of PR-B-mediated anti-inflammatory actions, and occurs in human term myometrium in association with the onset of labor (46, 49, 50).

PR-A may also promote parturition by directly increasing expression of CAP genes in myometrial cells. Studies of local P4 withdrawal suggest that P4 withdrawal via local 20αHSD-mediated metabolism produces unliganded PR-A that directly induces expression of genes encoding stimulatory CAPs (101). Loss of P4 leading to unliganded PR-A shifts the balance from P4/PR-B/JUN to PR-A/FOS activity, leading to increased CAP gene expression and unfettered response to proinflammatory stimuli producing a prolabor transcriptome (61, 65). Thus, as with local P4 withdrawal, functional P4/PR withdrawal appears to be redundant, involving various mechanisms including site-specific serine phosphorylation of PR-A, ligand-independent transcriptional activity of PR-A, and/or reduced direct transcriptional activity of PR-B.

3.4. Hormonal Control of Progesterone Withdrawal

A key implication of the P4/PR block hypothesis is that withdrawal of the block is a central event in the hormonal control of parturition and that this is induced by the convergence of endocrine and/or paracrine signals depending on physiological context (FIGURE 5). This concept is supported by studies in mice and sheep, in which the hormonal control of parturition is relatively well characterized.

FIGURE 5.

Comparison of endocrine and paracrine pathways to progesterone (P4)/P4 receptor (PR) withdrawal in human (solid line), ovine (dashed line), and murine (dotted line) pregnancy. A comparative approach highlights the unique roles of the murine corpus luteum and ovine placental P4 metabolism in P4/PR withdrawal versus the suspected mechanism of functional P4 withdrawal in human myometrium/decidua. CRH, corticotropin-releasing hormone; DHEA, dehydroepiandrosterone; E₂, estradiol; PGF_2α, prostaglandin F_2α. Image is an adaptation of material from Ref. 2.

In mice, PGF_2α produced by epithelial cells of the maternal endometrium increases 20αHSD in CL cells, leading to a decrease in P4 release and a decline in systemic P4 (79, 102–104). Loss of P4 removes negative feedback to the maternal hypothalamus-pituitary, leading to increased production of follicle-stimulating hormone and increased E₂, thus synchronizing P4 withdrawal and estrogen activation. The control of PGF_2α production by endometrial cells in mouse pregnancy is not clearly defined but is thought to involve some form of gestation clock mechanism (105).

In sheep, P4 withdrawal leading to term parturition is triggered by increased activity of the fetal hypothalamic-pituitary-adrenal (HPA) axis leading to a surge in cortisol level. Cortisol increases expression in ovine trophoblast cells of Cyp17a1 that encodes P450c17, which converts P4 to A4. The diversion of P4 from release to metabolism decreases systemic P4 levels in the ewe, leading to parturition (75). Ovine parturition does not occur if the fetal adrenal cortisol surge is prevented (e.g., by fetal adrenalectomy or fetal hypophysectomy) and can be induced prematurely by administration of glucocorticoid to the fetus (106). Importantly, A4 produced from P450c17-mediated metabolism of P4 is converted by sheep placental cells to E₂, leading to a coordinated increase of E₂ that acts on the uterine cells to promote labor. Thus, as with the mouse, in sheep parturition the systemic P4 withdrawal induced by specific enzymatic activity is coordinated with systemic estrogen activation.

Examination of the mechanism that increases activity of the fetal sheep HPA axis has shown that this is due to increased PGE₂ released by the placenta that sensitizes the fetal adrenal cortex to adrenocorticotropin (ACTH) (107). Thus, the ovine placenta, and specifically its production of PGE₂, is a trigger for parturition that via fetal adrenal cortisol induces systemic P4 withdrawal.

In contrast to the sheep, human parturition is not controlled by activity of the fetal HPA axis. Even in pregnancies in which the fetal HPA axis fails to produce cortisol (i.e., congenital adrenal hyperplasia and anencephaly), parturition occurs normally at term (108, 109). Moreover, administration of synthetic glucocorticoid to women presenting with threatened preterm labor does not appear to augment parturition or increase the incidence of preterm birth. Nonetheless, fetal organ system maturation is promoted by glucocorticoid. Also in contrast to the sheep, the human placenta lacks CYP17A1 and as such P4 production does not decline at any stage of pregnancy. Only after delivery of the placenta does systemic maternal P4 decline (110) (FIGURE 6).

FIGURE 6.

Comparison of relative serum levels of progesterone (P4) and estradiol (E₂) throughout gestation in human, sheep, and mouse. The level of P4 drops precipitously immediately before parturition in sheep and mouse but in human remains persistently elevated until after parturition. A functional withdrawal of P4/P4 receptor (PR), therefore, is not reflected in serum hormone levels in the context of human pregnancy. Image is an adaptation of material from Ref. 2.

Functional and/or local P4 withdrawal to initiate human parturition may be triggered by multiple redundant mechanisms, including a response to inflammatory stress. Studies in murine and nonhuman primate models show that parturition is induced by proinflammatory stimuli administered to the mother or fetus (38, 41–43, 111, 112). In humans, PG production by the fetal membranes and PG levels in amniotic fluid increase before active labor (113–117) and administration of PGs at any time in pregnancy can initiate parturition (118, 119). Taken together, the data suggest that inflammatory cytokines including locally produced uterotonic PGs, especially PGF_2α, are important hormonal triggers of parturition. In human myometrial cell lines, PGF_2α increased mRNA encoding PR-A (120) and IL-1β increased PR-A stability and transrepressive activity (50). This suggests, as described above, that inflammation induces PR-A-mediated functional P4 withdrawal (121).

4. INFLAMMATION

Clinical studies show that human parturition is affected by intrinsic physiological and extrinsic pathological inflammatory stressors. Normal term parturition is associated with tissue-level inflammation in the myometrium and chorion-decidua interface (37, 39), and risk for preterm labor leading to preterm birth is elevated in pregnancy affected by intrauterine infection, clinically silent upper genital tract infection, and bacterial vaginosis (122–125). Contributors to inflammatory triggers of parturition could include intrinsic physiology related to inflammatory signals from the fetus related to normal fetal development, distension of the uterus as the conceptus grows (126–130), and senescence of placental trophoblast cells (105, 131–136). Extrinsic factors could include maternal signals derived from environmental and psychosocial stressors (137) mediated by placental corticotropin-releasing hormone (CRH; discussed below) (138), intrauterine infection (139), and the maternal microbiome (140, 141). Recent studies in mice suggest that parturition is triggered by uterus-intrinsic pathways mediated by IL-33 produced by uterine interstitial fibroblasts and resident innate lymphoid cells, and that this mechanism for parturition initiation occurs without systemic P4 withdrawal, demonstrating a role for the immune system in the control of parturition (142).

In mice, fetal lung maturation is coordinated with parturition via pulmonary surfactant protein-A (SP-A), a major component of lung function necessary for air breathing. Intra-amniotic administration of SP-A to mice at midgestation induces labor via the infiltration of fetal macrophages into the maternal myometrium and activation of myometrial cell IL-1β and NF-κB. In contrast, inhibition of SP-A by administration of a blocking antibody delays normal parturition (143). Whether this system links fetal lung maturation with inflammation in the myometrium or intrauterine tissues in other species is not known. In humans, fetal macrophages have not been detected in the myometrium, and fetal SP-A levels are elevated several weeks before the onset of term labor and do not correlate in time with parturition (144).

Recent studies of the microbiome colonizing the upper vaginal tract suggest that signals from specific microbiota affect parturition (145, 146). Detailed multiomic analyses of the vaginal microbiota showed that risk for preterm birth was inversely related to the levels of Lactobacillus crispatus and was associated with taxa whose presence is correlated with elevated levels of proinflammatory cytokines in cervicovaginal fluid (140, 147–149). Although the mechanistic link between microbiome and parturition remains elusive, a longitudinal study integrating vaginal microbiota and cervicovaginal fluid immunophenotype in a cohort of women at high risk for preterm birth indicates involvement of maternal immune function, especially the complement system (150). Future studies may reveal microbiome-based biomarkers and therapeutics for preterm birth risk detection and prevention, respectively. Ongoing analyses linking vaginal microbiome community state and maternal immune function with risk for preterm birth support the concept that external factors contribute to the inflammatory load of pregnancy, affecting the trajectory toward the threshold for inflammation-induced P4/PR withdrawal.

As described above, proinflammatory stimuli may trigger parturition by inducing P4 withdrawal via systemic (e.g., PGF_2α-induced AKR1C1/20αHSD in CL cells), local (e.g., inflammatory cytokine induction of AKR1C1/20αHSD in uterine target cells), and/or functional (e.g., inflammatory cytokine modulation of PR-A transcriptional activity) mechanisms. It is possible, therefore, that the hormonal control of parturition is governed by the net inflammatory load on the uterine tissues derived from the maternal and/or fetal compartments converging to induce P4/PR withdrawal. It is also possible that the various modes for P4/PR withdrawal have specific sensitivity to inflammatory stimuli and that this is species dependent; this would account for the variation that has been observed in the hormonal control of normal term parturition seen across viviparous species. In this threshold-limited model, loss of P4/PR-B anti-inflammatory activity coupled with tissue-level inflammation would lead to a local positive-feedback inflammatory state that increases production of PGs to promote myometrial contraction, remodeling of the cervix ECM, and weakening of the amnion membrane. Loss of P4/PR-B would also remove inhibition of ESR1, leading to increased ERα in myometrial and cervical stromal cells. This increase in ERα would allow E₂ to increase OTR expression and sensitivity to OT and increase expression of metalloproteinases to remodel the ECM in the cervix. Moreover, loss of P4/PR signaling would deactivate the ZEB transcription factors via the miRNA-200s, leading to derepression of genes encoding CAPs and inflammatory mediators and 20αHSD (via loss of STAT5b). The net effect would be to produce a phenotype in uterine cells that allows uterotonic hormones to induce active labor (FIGURE 7).

FIGURE 7.

The inflammatory threshold hypothesis for the control of parturition timing. During pregnancy, inflammatory stimuli derived from multiple intrinsic (e.g., fetal development, placenta senescence) and extrinsic (e.g., maternal stress, intrauterine infection) factors contribute to a net inflammatory load impacting the uterine tissues. For most of pregnancy the progesterone (P4)/P4 receptor (PR) anti-inflammatory block to labor prevents uterine tissue-level inflammation. Above a threshold level, inflammatory stimuli induce P4 withdrawal via mechanism outlined in FIGURE 4. Loss of the P4/PR block to labor allows a positive-feedback proinflammatory state in the uterine tissues that induces active labor via increased contraction-associated protein (CAP) gene expression and increased production of and sensitivity to prolabor uterotonins. The timing of parturition is determined by the inflammatory load trajectory and the inflammatory load threshold for inflammation-induced P4/PR withdrawal. The threshold is normally reached at term but can be reached earlier in pathological states (e.g., intrauterine infection).

5. UTEROTONIC HORMONES

The parturition process involves uterotonic hormones that directly affect contraction of the myometrium, dilation of the cervix, and weakening of the amnion membrane. Two key uterotonins are PGs and OT. In addition, E₂/ERα signaling in myometrial and cervical stromal cells increases responsiveness to PGs and OT and promotes cervix softening. It is notable that active labor in response to induced P4/PR withdrawal (e.g., ovariectomy or RU486 treatment) has a latency, which likely is the time required for gene expression changes and tissue level processes that alter uterotonin production and responsiveness to produce the labor phenotype.

5.1. Estrogen

Animal studies show that E₂ increases expression of genes in myometrial, cervical, and decidual cells whose products promote tissue-level changes associated with parturition (151). This includes proteolytic enzymes that remodel the cervix and amnion ECM to cause cervix softening needed for dilation and membrane weakening required for rupture (152, 153). E₂ also increases responsiveness to uterotonic hormones, especially OT and PGs, by increasing expression of genes encoding OT and PG receptors in uterine cells (154–156) and enhances contraction efficiency by increasing expression of GJA1 to promote formation of gap junctions between myometrial cells (157–160) necessary for synchronized contractions.

During the latter half of human pregnancy placental trophoblast cells convert C19 androgen produced by the fetal adrenal cortex to estrogens (E₂, estrone, estriol, and estetrol) (161–163). Consequently, the uterine tissues are exposed to a high-estrogen milieu for most of human pregnancy (110, 164). Yet for much of this time the uterus is not responsive to prolabor estrogen actions and remains quiescent. This is likely because of dominance of the P4/PR block and P4/PR inhibition of ESR1, thus desensitizing uterine cells to prolabor E₂/ERα-mediated actions.

The role of estrogens in the hormonal control of human parturition timing is unclear. Pregnancies affected by abnormally high placental estrogens due to congenital fetal adrenal hyperplasia have normal term parturition (109). The timing and process of parturition are also normal in cases of placental aromatase deficiency (165–167), anencephaly (168, 169), placenta sulfatase deficiency (170–172), and congenital adrenal lipoid hyperplasia (173, 174), in which maternal estrogen levels are significantly lower than normal. Notably, in such pregnancies E₂ production by maternal tissues persists, and therefore the uterine tissues are exposed to some, albeit lowered, estrogenic drive that may be sufficient to promote parturition upon P4/PR withdrawal. A plausible explanation is that for most of human pregnancy inhibition of ESR1 expression by P4/PR desensitizes uterine cells to proparturition effects of estrogens. This is reversed upon P4/PR withdrawal, allowing even lower than normal levels of estrogens to alter the uterine tissues in favor of parturition. Thus, as with other species, in human parturition P4/PR withdrawal and E₂/ERα activation appear to be synchronized.

5.2. Oxytocin

OT, a small (9 amino acid) neuropeptide, is produced in the supraoptic and paraventricular nuclei of the hypothalamus and by decidual cells in the gravid uterus and plays a key role in the hormonal control of parturition (for review see Refs. 2, 175). OT production by the decidua and OXTR expressed by myometrial cells are each upregulated by estrogen (176–178). OT via the OXTR is a potent uterotonin that induces uterine contractions during the process of parturition. These effects occur mainly during the expulsive phase of active labor. Expression of OXTR in myometrial cells is increased by E₂/ERα and P4/PR withdrawal (175, 179–181), leading to increased responsiveness to OT (182). Locally produced OT may be a key component of the estrogen proparturition effects on the uterus. Interestingly, in decidual stromal cells OT increases PG production, which could augment contraction of adjacent myometrium (178).

Interaction of OT with receptors on myometrial cells is controlled by a balance between its production by the brain and decidual cells and its inactivation by oxytocinase, a cystine aminopeptidase produced by the decidua, chorion, and placenta (183, 184). For most of pregnancy, metabolism of OT prevails. Secretion of OT from the posterior pituitary occurs as part of the lactation letdown reflex and during labor in response to cervical dilation and vaginal distension (175, 185), referred to as the Ferguson reflex, a neuroendocrine loop that creates positive feedback to augment uterine contractions in response to cervix dilation and vaginal distension. OT secretion from the posterior pituitary may also be controlled in a circadian manner and in response to environmental stressors (186), which could be a mechanism via which psychosocial or other external influences are transduced to influence the process of parturition.

Whether pituitary OT is essential for parturition is questionable, since pregnancy and parturition are normal in women with a dysfunctional posterior pituitary (187), and parturition is delayed but not prevented in clinical studies and animal models of inhibition of OT with OXTR antagonists (188–190). Moreover, mice lacking OT have normal term parturition but dysfunctional lactation (191, 192). Overall, current data suggest that the role of OT in parturition is complex and redundant. Although it is a potent stimulatory uterotonin and contributes to the expulsive phase of labor and produces the postpartum tonic contraction of the uterus to facilitate hemostasis, it does not appear to be necessary for the initiation of parturition, and the full parturition cascade can proceed in the absence of OT action. This likely reflects redundancy in the hormonal control of parturition.

5.3. Prostaglandins

PGs, especially PGF_2α and PGE₂, play major paracrine roles in the hormonal control of the timing and process of parturition (193–196). PGs are synthesized by various cell types including resident immune cells, especially activated macrophages and neutrophils in the placenta, amnion, and chorion-decidua and myometrium. PGs are formed from arachidonic acid by the actions of phospholipase A₂ and the prostaglandin-endoperoxide synthase-1 and -2 (PTGS1 and PTGS2) enzymes (197). In the human gestational tissues PTGS2 (aka COX2) is induced by proinflammatory stimuli and is the key rate-limiting enzyme that provides substrate, PGH₂, for the PG synthetic pathway. PGH₂ is then converted to PGF_2a, PGE₂ and PGI₂, PGD₂, and thomboxane A₂ by specific isomerases (197).

Tissue levels of PGs are decreased by the 15-hydroxyprostaglandin dehydrogenase (HPGD) enzyme that converts PGs to inactive forms (198). Interestingly, HPGD levels are high in the chorion and decrease during active labor at term and in preterm labor (spontaneous or infection associated) (199, 200). The data suggest that regionalization of HPGD serves to regulate PG access to the myometrium and cervix. In this model, loss of the HPGD chorion barrier may provide a mechanism by which the human fetus triggers parturition, allowing amnion PGs (possibly produced in response to intrauterine infection) to access the uterine tissues.

Clinical and animal studies show that pharmacological inhibition of PG synthesis inhibits labor (194). In humans, PG levels in the amniotic fluid increase during the prelude to labor (201), suggesting that PGs contribute to labor induction. This is applied clinically by exogenous administration of PGF_2α and PGE₂ to promote uterine contraction and cervical dilation. In mice, parturition at term is triggered by increased expression and activity of Ptgs1 in the uterine epithelium, which increases PGF_2α release to induce CL regression and 20αHSD-mediated and systemic P4 withdrawal (188). In sheep, increased PGE₂ production by placental trophoblast cells increases sensitivity of fetal adrenal cortical cells to ACTH and activity of the fetal HPA axis to produce the cortisol surge that triggers parturition via systemic P4 withdrawal (107). Human parturition involves increased expression of PTGS2 in the gestational tissues; this leads to increased local and amniotic fluid levels of PGF_2α and PGE₂ (202–204) that may induce functional P4 withdrawal by inhibiting PR transcriptional activity and/or local P4 withdrawal by inducing 20αHSD-mediated P4 metabolism.

Specific actions of PGs on the gravid uterus are controlled by the type of PG synthesized, the extent to which they are locally inactivated by HPGD, and the temporal, spatial, and cell type-specific pattern of PG receptors on target cells. PGF_2α induces myometrial cell contraction via the FP receptor. PGE₂ induces myometrial contraction via the EP1 and EP3 receptors and relaxation via the EP2 and EP4 receptors (205, 206). FP and EP3 receptors are present in low quantity in the myometrium for most of pregnancy, likely a consequence of P4/PR inhibition. At parturition, FP receptors increase in the uterine fundus but decline in the lower uterine segment, whereas EP4 expression increases in the lower segment (205, 206). This is consistent with the concept that most of the expulsive force during labor is generated by the fundus, whereas the lower uterine segment must be less contracted to allow movement of the fetus through the dilated cervix.

6. PLACENTAL PEPTIDE HORMONES

The placenta is a remarkable endocrine organ producing multiple hormones whose function is generally to modify maternal physiology in favor of maintaining pregnancy and providing the fetus with oxygen and nutrients. The role of placental hormones in the control of parturition varies among species. As discussed above, the human placenta provides P4 to maintain pregnancy. Whether the placental trophoblast cells and the syncytiotrophoblast participate in the hormonal control of parturition is not clearly apparent. A plethora of hormones are produced by the placenta and are released into the maternal compartment, and most of these affect maternal physiology to favor pregnancy (207). One hypothesis for the contribution of placental signals in parturition is that senescence of placental cells and tissues occurs with advancing gestation, producing local senescence-associated inflammatory factors that contribute to the inflammatory load on the adjacent uterine tissues (135, 208).

Corticotropin-releasing hormone (CRH) produced by cells in the human placenta (syncytiotrophoblast and trophoblast) is thought to play an intermediary role to integrate fetal signals to initiate parturition. In human pregnancy, production of placental CRH, which is identical to CRH produced by the hypothalamus, can be detected in the maternal circulation beginning in the second trimester. Levels increase with advancing gestation, and that increase becomes exponential during the third trimester. The rate of maternal CRH increase can serve as a biomarker to predict proximity to parturition. When compared to the trajectory of maternal CRH in women with a normal term delivery, a higher rate of increase beginning early in the second trimester is predictive of preterm birth and a lower rate of increase is predictive of postterm birth (209, 210). An important concept to emerge from these studies is that events occurring early in pregnancy (e.g., factors affecting placental CRH production) set a trajectory for pregnancy outcome (210, 211). This could reflect a temporal and developmental link between the conceptus and parturition in human pregnancy whereby placental CRH reflects the developmental trajectory of the placenta and/or the fetus (210).

In the human fetus, placental CRH may affect the HPA axis by increasing production of pituitary ACTH and directly affecting fetal adrenal cortical cells by increasing production of dehydroepiandrosterone (DHEA). As DHEA is the principal C19 substrate for estrogen production by the human placenta, this would increase maternal estrogen levels (212). Importantly, unlike the adult HPA axis, in which cortisol inhibits CRH production, cortisol stimulates CRH production by human placental trophoblast cells (213, 214). This may produce a positive feedback loop whereby fetal cortisol increases CRH, which then increases maternal E₂ levels (via increased DHEA). E₂ could then participate in the hormonal control of parturition. Although plausible, this model must be considered from the perspective of data described above suggesting that the fetal HPA axis and maternal E₂ levels do not participate in the hormonal control of human parturition (FIGURE 8).

FIGURE 8.

Proposed model in which maternal and fetal cortisol stimulate corticotropin-releasing hormone (CRH) production by the trophoblast. This would generate positive feedback loops to result in increased estradiol (E₂) to promote maternal expression of contraction-associated proteins (CAPs) and contribute to the onset of parturition. Cortisol-induced fetal organ maturation is also positively affected by placental-fetal feedback in this system. DHEA, dehydroepiandrosterone. Image is an adaptation of material from Ref. 215.

7. EVOLUTIONARY PERSPECTIVE

As with all aspects of reproductive biology, the hormonal control of parturition is likely subjected to strong natural selection. Parturition is a major event in the viviparous reproduction process and imparts a significant risk for maternal and neonate survival, and consequently reproductive fitness. As reproductive fitness drives natural selection, parturition traits that optimize reproduction in the context of other traits, some of which are species specific, in the reproduction process, such as seasonality, cyclical or continuous fertility, litter size, and neonate maturity, were likely subjected to strong selection pressure. This explains, at least in part, the diversity among viviparous species in the hormonal control of parturition.

The hormonal control of human parturition is thought to have been influenced by the evolution of encephalization, a trait unique to the hominid lineage that increased the size and complexity of the brain (216–218). This concept posits that benefits of encephalization in the hominid lineage produced strong positive selection pressure leading to a progressive increase in the rate of fetal brain growth and the size of the fetal head (FIGURE 9). The obstetric dilemma hypothesis predicts that the benefits of encephalization would be lost when the size of the fetal head exceeds the size of the maternal pelvic outlet (cephalopelvic disproportion), leading to reproduction failure (217, 219, 220). Coupled with the mechanical problems, the metabolic crossover hypothesis (221) suggests that energy required for rapid fetal brain growth is a limiting factor for pregnancy especially when it exceeds the capacity for maternal energy supply. One biological solution to the encephalization-related problems is to alter the hormonal control of parturition to shorten gestation such that the fetus is born before its brain/head size becomes larger than the pelvic outlet and the energy requirements for continued in utero fetal brain growth exceed the capacity for maternal energy supply. As discussed above, the timing of parturition is controlled by multiple drivers mediated by hormonal interactions that converge to induce P4/PR withdrawal. One intriguing possibility is that changes in PGR gene sequence altered the PR to increase sensitivity to PR-mediated functional P4/PR withdrawal. Recent studies of PGR gene sequence variation between human populations indicate that changes around the serine 344/345 locus affect risk for preterm birth (222). The Neanderthal genome suggests differences in the PGR gene compared with Homo sapiens that could have affected reproductive fitness (223). Interestingly, in the Neanderthal PGR has a threonine at position 344 instead of a serine at 344 in humans. Whether these changes affect sensitivity to inflammation-induced phosphorylation remains unknown and the subject of ongoing studies. It is tempting to speculate, however, that in Homo sapiens the presence of a serine at the 344 position increased sensitivity to inflammation-induced parturition via pSer^344/345-PRA-mediated functional P4/PR withdrawal that favored a solution to the obstetric dilemma, giving Homo sapiens a reproductive advantage. It is also possible that the change increased the risk for preterm birth, which is counterintuitive in terms of reproductive fitness, since preterm birth increases risk for neonate death and reproduction failure. However, as natural selection operates at the species level and over multiple generations, it is plausible that the benefits of encephalization outweighed the cost of reproduction failure due to preterm birth. Another evolutionary consideration is that although preterm birth imparts cost on the fetus in the current pregnancy, it preserves the female’s capacity for future pregnancies and/or to care for existing offspring. Clearly, pregnancy is a state of physiological stress for the female, especially late in gestation when, in the case of hominids, parturition involves birthing a large-headed fetus whose brain has a high energy demand. From a species-level evolutionary perspective natural selection would favor hormonal signals that trigger parturition to favor survival of the female and neonate, but with preference for the female based on her established capacity for future pregnancy and the possibility that she is responsible for the survival of current offspring.

FIGURE 9.

Visual comparison of the relative proportions of the fetal skull to the maternal pelvis in primates shows a maximal cephalopelvic ratio in humans. The obstetric dilemma hypothesis proposes that the biomechanical limitations of the pelvic outlet (blue outline) and the size of the human head (red outline) necessitate that parturition occur at a time that results in relative altriciality of the human neonate. Image is an adaptation of material from Ref. 2.

8. CLINICAL PERSPECTIVE

Attempts to clinically manipulate the timing of parturition are inherent in the practice of obstetrics. Numerous scenarios exist in which delivery before the onset of spontaneous labor at term, and slowing or halting preterm labor, is beneficial for the parturient and the fetus (224, 225). To this end, leveraging the hormonal mediators of parturition has been a primary focus in strategies to both induce and prevent parturition (FIGURE 10).

FIGURE 10.

The hormonal regulation of parturition provides an opportunity for therapeutic intervention, both for prevention of preterm labor (left) and for induction of labor (right). Interventions that directly target hormonal pathways are listed for the prevention of preterm labor (left) and induction of labor (right). Attempts to prevent preterm labor have targeted progesterone (P4) receptor (PR), oxytocin receptor (OTR), prostaglandin (PG) synthesis, and estrogen receptor (ER). These interventions include some with therapeutic benefit in a limited population (e.g., vaginal P4 for short cervix), some with unclear or unproven benefit (e.g., atosiban), and some harmful and contraindicated in pregnancy [e.g., diethylstilbestrol (DES)]. Medically indicated labor and management of miscarriage, in contrast, utilize all of the listed interventions routinely as the standard of care in common clinical scenarios.

8.1. Induction of Parturition

Medically indicated induction of labor, clinically defined as rhythmic, forceful, and painful contractions of the myometrium, softening and dilation (referred to as ripening) of the cervix, and weakening and possibly rupture of the amnion-chorion membrane, utilizes pharmacological methods that promote cervical ripening and uterine contractions via endogenous uterotonin signaling pathways. The principal agents used are synthetic analogs of OT (e.g., Pitocin) and PGE₂ (e.g., dinoprostone) and PGE₁ (e.g., misoprostol) (226–230). The PG analogs are used for cervical ripening and also promote contractions. In contrast, although OT is effective for inducing myometrial contractions, it is often used as an adjunct for cervical ripening (231). OT and PG analogs are presumed to function via their endogenous uterotonic pathways and possibly override or circumvent the P4/PR block to parturition (120).

RU486 and misoprostol (PGE₁) are used for cervical ripening in preparation for medical and surgical abortion and for the management of miscarriage (232, 233). As a PR antagonist, RU486 is a potent abortifacient and in animal models alone is sufficient to induce the full parturition cascade (234), which supports the P4/PR block hypothesis. Clinically, RU486 is used for cervical ripening in term pregnancy in clinical trials (235, 236); RU486 is not broadly used for induction of labor in viable pregnancies. This is likely due, at least in part, to regulatory barriers to its use (237).

The nonpharmacological, mechanical methods of cervical ripening (intracervical balloon catheters and, for pregnancy termination, osmotic dilators) and other obstetric interventions (membrane sweeping and amniotomy) are suspected to act at least in part via stimulation of endogenous PGs (238, 239). These methods of inducing parturition are often used in combination or sequentially in a single patient, and, overall, induction of labor has been demonstrated to decrease the rate of cesarean delivery compared to expectant management (240).

In a minority of patients who undergo cervical ripening and induction of labor, active labor is not achieved and an unplanned cesarean delivery is required (241). The risk factors and clinical scenarios that contribute to failed induction of labor in an individual patient are numerous, and so it follows that the contributing mechanisms of failure are likely complex and heterogeneous. Based on the seesaw and threshold models of parturition initiation, one can surmise that insufficient cervical ripening and failed induction of labor may result from a failure to overcome the P4/PR block. Whether a similar mechanism contributes to dysfunctional spontaneous labor (when parturition is initiated but does not progress normally; also a contributor to unplanned cesarean delivery) is also unknown.

8.2. Prevention of Preterm Parturition

Preterm birth is a substantial cause of neonatal morbidity and mortality. A contemporary epidemiological study has demonstrated that 65% of preterm births from 20 weeks to 36 weeks 6 days of gestation result from spontaneous preterm labor (versus iatrogenic preterm delivery) (242). The clinical presentations of preterm labor are heterogeneous; therefore, it is not surprising that translational and clinical research have failed to identify a universal prophylaxis or therapy for preterm labor. Several therapies have been tested that each attempt to restore uterine/cervical quiescence by exploiting hormonal mediators of parturition.

Because of the central role of the P4/PR block to parturition, progestins have been a major focus over several decades for both early pregnancy maintenance (243, 244) and prevention of preterm labor (for recent reviews on progestin-based prophylaxis for preterm birth prevention see Refs. 245–247). Boosting the P4/PR block and/or preventing its withdrawal is an attractive strategy in theory. Indeed, despite early (1950–1960) studies showing that P4 treatment has no tocolytic benefit for women in active labor (248), subsequent, albeit small, trials found that prophylaxis with the caproate salt of 17-hydroxyprogesterone (17HPC) may reduce preterm birth risk (249). Those findings motivated larger clinical trials of 17HPC prophylaxis (250) and vaginal P4 (251) for preterm birth prevention in women with elevated risk due to a history of preterm birth. Both trials produced beneficial outcomes, with a reduced incidence of preterm birth in women receiving progestin prophylaxis. Subsequent trials of 17HPC (252) and vaginal P4 (253), however, showed that neither therapy reduces the risk for preterm birth. Consequently, 17HPC is no longer recommended for preterm birth prevention. Vaginal P4 for women with increased preterm birth risk due to short cervix remains the only progestin-based prophylaxis in clinical use (254, 255). The general failure of progestin supplementation to prevent preterm birth is expected, since levels of P4 in the maternal circulation are persistently high even in pregnancies that end preterm and there is no evidence that preterm birth risk is related to endogenous P4 deficiency that can be corrected clinically (247). Intriguingly, promegestone (aka R5020), a P4 analog that binds the PR with high affinity and is not metabolized by 20αHSD/AKR1C1, delays parturition in a mouse model of inflammation-induced preterm birth (256). Clearly, improved understanding of the mechanism(s) for, and hormonal control of, P4/PR withdrawal in the physiology of human parturition will provide more effective therapeutic strategies to clinically control the timing of birth.

Additional tocolytic strategies to intervene on preterm contractions include atosiban, which functions as an OT receptor antagonist and also has anti-inflammatory effects in myometrial cells (257). This medication has demonstrated safety and is approved for clinical use in Europe; however, its efficacy in improving neonatal outcomes is uncertain, and a clinical trial to address this is ongoing (258). Indomethacin, an inhibitor of prostaglandin synthesis, is also utilized as a tocolytic in limited clinical scenarios for brief durations at early gestational ages; its overall effectiveness for improvement in neonatal morbidity has not been sufficiently demonstrated (259, 260). The use of indomethacin will continue to remain limited because of the risk of premature closure of the fetal ductus arteriosus, a well-known consequence of PG synthesis inhibition in the late fetal period.

The potential for fetal effects of hormonally directed therapies is not limited to PG inhibition. Diethylstilbestrol (DES) is an ER agonist that was used in the 1940s–1960s for, among other indications, prevention of miscarriage and preterm birth. It was ultimately found to cause gynecologic cancers in females who were exposed in utero, congenital anomalies of the reproductive tract in both females and males following in utero exposure, and concerns for third-generation effects (261). This sobering history underscores the importance of appreciation of the holistic maternal, placental, and embryonic/fetal effects of any pregnancy intervention, particularly when manipulating hormonal pathways that have distinct and essential roles in the mother and fetus.

9. CONCLUSIONS

The hormonal control of parturition is complex and redundant, integrating multiple physiological signals. Some of these signals are overt stressors, whereas others allow maternal adaptation to the conceptus and receipt of the fetus’ indication of its readiness for neonatal life. These endocrine and paracrine signals occur consequent to maternal and fetal-placental inputs and collectively result in uterine contractions, cervix softening/dilation, and membrane rupture and eventually in the completion of parturition. Upon this framework, unraveling the hormonal control of parturition must acknowledge that different viviparous species possess diverse controls for parturition timing and process. This diversity is shaped by species-specific natural selection that favors a parturition control system that increases reproductive fitness in the context of the extant constellation of reproductive traits. Despite variability, the P4 block to parturition mediated by the nuclear PRs and the induction of parturition by withdrawal of P4/PR signaling appears to be a conserved trait. Upstream signals that induce P4/PR withdrawal initiate downstream transducers to induce anatomic and physiological changes mediated by uterotonins (e.g., OT, PGs) that complete the parturition process. To better understand the physiology of parturition, analysis of existing data and those from ongoing and future studies must acknowledge and reconcile the diverse and common pathways across species in the context of natural history and evolutionary biology. This approach is yielding novel concepts to advance understanding of the hormonal control of parturition that will translate clinically to address adverse birth timing conditions, especially preterm labor leading to birth of a premature neonate.

GRANTS

This work was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (HD097279; HD103676) and the March of Dimes Prematurity Research Center, Collaborative of Ohio.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.M. conceived and designed research; S.M. prepared figures; E.H.-S. and S.M. drafted manuscript; E.H.-S. and S.M. edited and revised manuscript; E.H.-S. and S.M. approved final version of manuscript.