In mice, the catabolism of progesterone to the inactive C21-steroid 5α dihydroprogesterone by the enzyme steroid 5α-reductase facilitates cervical ripening, while mice deficient in this enzyme fail to deliver at term despite normal uterine contractions and progesterone withdrawal in blood [¹⁹]. In mice with a targeted disruption of 20α-hydroxysteroid dehydrogenase (20α-HSD) gene (which converts progesterone to the biologically inactive metabolite 20α-dihydroprogesterone), the mean duration of pregnancy is significantly prolonged compared to the controls [²⁰]. In cultured rat ovarian granulose cells IL-1β stimulates 20α-hydroxysteroid dehydrogenase activity [²¹]; however it is unknown if such an effect may also take place in myometrium during labor. In humans, experiments with endocervical cells of women in labor show that, during cervical ripening and parturition, 17β-hydproxysteroid dehydrogenase type 2 activity decreases, resulting in increased 20α-hydroxyprogesterone and estradiol levels because of decreased conversion of 20α hydroxyrogesterone to progesterone and decreased conversion of estradiol to estrone. These changes result in inactivation of progesterone responses [²²].

The role of progesterone receptors in human pregnancy has been recently extensively reviewed [²³]. In humans, there are two major isoforms of progesterone receptor, PR-A and PR-B, which belong to the nuclear receptor superfamily, as well as many other isoforms which have so far given evidence of being less significant. The expression of the various isoforms may contribute to the functional withdrawal of progesterone during labor. In human myometrial cells, the ratio of PR-A:PR-B mRNA increases 2- to 3-fold compared with the nonlaboring state, mainly due to overexpression of PR-A. This change induces a “functional estrogen activation” through increased estrogen receptor α (ERα) expression [²⁴]. PR-A may also suppress the transcriptional activity of PR-B, which is the main receptor for the nuclear signal transduction of progesterone [²⁵, ²⁶]. Apart from myometrial contractions, the functional progesterone withdrawal due to the altered expression of PR-A, PR-B isoforms may also contribute to the cervical changes during labor [²⁷, ²⁸].

The binding of progesterone to PR-C, which is a soluble form of the receptor, may sequester available progesterone away from PR-B and thereby diminish its biological effect [¹⁸]. In another study of women in labor, PR-C protein levels were increased in cytoplasmic fractions of fundal myometrial cells, but not in cells of the lower uterine segment, while PR-B protein levels increased only in the laboring fundal endometrium. In the same study PR-A could not be detected despite the presence of PR-A mRNA. This paradoxical increase in PR-B protein may be the result of the associated reduced progesterone activity. In human telomerase reverse transcriptase- (hTERT-) immortalized human myometrial cells, PR-C compromises PR-B transactivation. The upregulation of PR-B and PR-C expression was associated with activation of NF-κB, while upregulation of NF-κB was observed in pregnant mouse uterus, coinciding with PR isoform expression. In addition, treatment of myometrial cells with the cytokine IL-1β further induces NF-κB activation and concomitant increase in PR expression, which suggests that labor-associated changes in myometrium are associated with uterine inflammatory pathways [²⁹].

A number of other progesterone receptors (mPR-α, mPR-β, and mPR-γ ) have been identified which are believed to be structurally related to the G-protein complex. mPR-α and mPR-β receptors were found to be expressed in human pregnant myometrial cells, and progesterone activation of these receptors may induce nongenomic actions, resulting in inhibition of adenyl cyclase, phosphorylation of myosin light chains, and contractions of human myometrial cells in labor. It is suggested that mPR at the end of pregnancy downregulates steroid receptor coactivator-2 (SRC-2), which in combination with an altered PR-B:PR-A ratio may lead to decreased transcriptional activation of PR-B [³⁰]. Controversy exists regarding the function of mPRs on cell membrane, some authors supporting the view that the isoforms of these receptors are intracellular and reside primarily on the endoplasmic reticulum, while progesterone does not induce any G protein-dependent signaling of mPRs [³¹].

PRs interact with coregulators that either increase (coactivators) or decrease (corepressors) their transcriptional activity. These molecules include among others the CREB-binding protein (CBP) and the steroid receptor coactivators SRC-1, SRC-2, and SRC-3. In fundal myometrial cells of women in labor, the mRNA levels of SRC-2, SRC-3, and CBP are decreased compared to the nonlabor samples. Nuclear levels of CBP, SRC-2, and SRC-3 proteins were also reduced in fundal myometrium during labor. In the mouse uterus in labor, levels of SRC-1, SRC-3, and CBP were also reduced. The reduced levels of these coactivators were related to decreased acetylated histone H3 in both human and mouse, suggesting the closing of chromatin structure and the reduced expression of PR-responsive genes as a possible mechanism for the initiation of labor [³²]. Histone acetylation/deacetylation can also affect NF-κB activity and proinflammatory gene expression (e.g., IL-6, IL-8, COX-2, and RANTES), while histone deacetylase inhibitors exert anti-inflammatory effects in myometrium which may favor uterine quiescence [³³].

p54nrb (non-POU-domain-containing, octamer binding protein) is another molecule that acts as a PR transcriptional corepressor in vitro and which decreases in rat myometrium at term pregnancies. Its decreased expression at the time of labor may contribute to the increased expression of labor-associated genes [³⁴].

An antagonism between NF-κB and progesterone exists in reproductive tissues during labor. NF-κB belongs to the superclass 4 of transcriptional factors and is involved in the synthesis of many proinflammatory mediators such as cytokines and COX-2. Inflammation is a critical trigger for the initiation of term and preterm labor [³⁵], and NF-κB as a transcriptional factor is implicated in proinflammatory stimuli signaling. Proinflammatory cytokines such as IL-1β and TNF-α, as well as lipopolysaccharides (LPSs), can stimulate NF-κB activity in the uterus. Additionally, the promoter region of cytokines genes such as IL-1β, IL-8, and IL-6 contains NF-κB recognition elements, and NF-κB promotes the synthesis of cytokines [³⁶].

NF-κB also stimulates prostaglandin synthesis, regulates matrix metalloproteinase expression, connexin-43 (a gap junction protein), oxytocin receptor, and may express inhibitory PR isoforms or increase progesterone metabolism [¹⁸, ³⁶]. In lower segment fibroblasts and in amnion cells IL-1β induces the synthesis of NF-κB regulated genes, while progesterone attenuates these effects [³⁷]. Progesterone exhibits anti-inflammatory properties that maintain uterine quiescence. MCP-1 is a chemokine which is produced by decidual, endometrial, myometrial, and trophoblast cells. It is increased in laboring pregnant myometrium prior to the onset of labor and attracts leukocytes from the circulation to the myometrium. MCP-1 is also increased in cultured myometrial cells in response to IL-1β. Experimentally, in choriodecidual cells MCP-1 expression is suppressed by progesterone, while, by contrast, it is stimulated by NF-κB [³⁸–⁴⁰].

Antagonism between nuclear factors such as NF-κB and PR for nuclear cofactors is another proposed mechanism of functional progesterone withdrawal. During labor, the level of NF-κB changes in the intracellular environment, and its activity is regulated by interactions with the nuclear cofactors that increase gene expression, such as CBP/p300. The antagonism between NF-κB and PR for CBP may contribute to progesterone withdrawal and expression of labor-promoting genes [⁴¹].

Progesterone acts as an immunomodulator that interacts with the immune system and exerts anti-inflammatory effects throughout pregnancy. Progesterone inhibits the activity of dendritic cells (DCs) that generate proinflammatory responses and favor the induction of tolerogenic DCs. It also controls the activity of natural killer (NK) cells and the differentiation of T cells into T-helper cell type 2 (Th2) like clones. The Th2 phenotype induced by progesterone is a prerequisite for the maintenance of pregnancy [⁴², ⁴³].

Progesterone-induced blocking factor (PIBF) is a protein which is released from lymphocytes in response to progesterone, mediates the immunological effects, and induces the production of Th2 dominant cytokines like IL-3, IL-4, and IL-10 [⁴⁴]. The functional progesterone withdrawal and the attenuation of progesterone's effects may contribute, among other factors, to the swift from a Th2 to a Th1 dominant effect towards labor. Inflammatory response involves Toll-like receptors, and progesterone modulates the expression of these receptors in mouse cervix and placenta [⁴⁵].

Cytokines such as IL-1β and TNF-α increase PGHS-2 and therefore the synthesis of prostaglandins (PG), while they downregulate prostaglandin 15-hydroxy dehydrogenase (PGDH), the enzyme that converts PGs into inactive metabolites. On the contrary, progesterone promotes PGDH expression [⁶]. Progesterone also downregulates prostaglandin F2-alpha receptor, which is increased by IL-1β [⁴⁶]. Via the use of expression microarray and quantitative reverse transcriptase PCR it has been shown that in human myometrial cells medroxyprogesterone acetate, a synthetic progestogen, affects genes involved in inflammatory response and cytokine activity. Among the downregulated genes of the study were those of IL-1β, Il-6, Il-11, Il-24, COX-2, and connexin 43 [⁴⁷].

Clinically, women with threatened preterm labor have significantly lower serum concentrations of PIBF and IL-10 and significant higher serum concentrations of the proinflammatory cytokines, IL-6 and interferon-γ (IFN-γ), compared with healthy pregnant women of the control group [⁴⁸].

A novel pathway for progesterone withdrawal and increased expression of contraction-associated genes has recently been described based on experiments in mice and human endometrial cells. The levels of micro-RNAs (mi-RNAs), a class of posttranscriptional regulators that belong to the mi-RNA-200 family, increase with advanced gestation, as do the levels of connexin-43 and oxytocin receptor, while the levels of ZEB1 and ZEB2 (zinc finger E-box binding homeobox proteins 1 and 2) repressor transcriptional factor decline. ZEB1 and ZEB2 inhibit the expression of connexin-43 and oxytocin receptor and block the oxytocin-induced contraction of myometrial cells. ZEB1 and ZEB2 also inhibit the expression of the miRNA-200 family, which in turn downregulate ZEB1, ZEB2. Progesterone increases the levels of ZEB1, and thus progesterone withdrawal reverses the inhibitory effect of ZEB on mi-RNAs. As mi-RNAs increase, they cause repression of ZEB and thereby promote the synthesis of connexin-43 and oxytocin receptor, thus facilitating labor [⁴⁹].

In summary, all the above described mechanisms may collectively impair progesterone regulation of gene expression, and this functional progesterone withdrawal promotes uterine contractility, effecting labor.

Progesterone has clinical applications with regard to either the initiation of labor or the arrest of preterm labor. In women, treatment with antiprogestins induces labor and delivery, but the most significant applications concern prevention of preterm birth. Numerous studies and meta-analyses have been conducted to evaluate the efficacy of progesterone in reducing the incidence and complications of preterm labor. Progestational agents may reduce the number of preterm labor and their complications in certain cases, but the optimal route of administration, the optimal dose, and the frequency of administration have not been clearly defined [⁵⁰–⁵⁴].

The rise of CRH is associated with a concomitant fall in CRH binding protein (CRHBP) in late pregnancy. The binding of CRH to CRHBP makes it biologically inactive and removes it from circulation. The decrease in CRHBP during the final three weeks of pregnancy results in higher concentrations of active CRH in circulation that contribute to the onset of labor [⁵⁵, ⁵⁶]. In both term and preterm labor, the expression of CRH mRNA from the placenta is increased, with this increase being higher in preterm deliveries. However, placental CRHBP mRNA expression does not change despite the fact that CRHBP decreases in maternal circulation before labor. This leads to the hypothesis that another source of CRHBP exists, such as a fetal source that may be responsible for this decrease [⁵⁵, ⁵⁷].

However, despite the fact that the levels of CRH in maternal circulation rise by as much as 1000 times compared to the nonpregnant state, only a mild hypercortisolism occurs, mainly due to estrogen-induced production of cortisol-binding protein, and the maternal axis continues its function. Thus the functional target of placental CRH is not the maternal pituitary adrenal axis but the fetal pituitary adrenal axis [⁵⁶]. Placental CRH stimulates adrenocorticotropic hormone (ACTH) production from the fetal pituitary. ACTH stimulates fetal adrenals to produce dehydroepiandrosterone (DHEA), dehydroepiandrosterone-sulphate (DHEA-S), and cortisol. ACTH is also produced in the placenta through paracrine mechanisms. Placental CRH also exerts a direct effect by stimulating fetal adrenal zone cells (Figure 2). Fetal adrenal DHEA is metabolized to estrogens in the placenta that favor parturition [², ⁵⁸, ⁵⁹]. The produced cortisol exerts a stimulatory effect on the placenta to further produce CRH, thus a positive loop is established that causes placental CRH to rise exponentially as pregnancy advances. Estrogens, progesterone, and nitric oxide inhibit CRH production, while a number of neuropeptides exert a stimulatory effect [²]. Cytokines influence CRH production. IL-1β and TNF as well as lipopolysaccharides activate the hypothalamic-pituitary-adrenal axis and induce CRH production, while CRH regulates IL-1β, IL-6, and lipopolysaccharides production by immune cells [⁶⁰]. Cytokines also influence the metabolism of cortisol. The enzyme 11β-hydroxysteroid dehydrogenase (11β-HSD) converts cortisol to the inactive metabolite cortisone. Cytokines such as IL-1β, IL-6, and TNF-α inhibit human placenta 11β-HSD type 2 isoform, and this could result in increased cortisol levels and therefore increased CRH production by the placenta [⁶].

Cortisol produced by fetal adrenals acts on fetal lungs and produce surfactant protein A (SP-A) that activates inflammatory signals in the uterus, which consequently enhance myometrial contractility [¹⁸]. SP-A signals via Toll-like receptors. These are membrane-spanning receptors that act as key regulators in immune system responses [⁶¹]. SP-A exerts its action via activation of NF-κB in many cells. In the mouse, SP-A increases the production of IL-1β and NF-κB by amniotic fluid macrophages and causes migration of macrophages in the uterus, resulting in the activation of the inflammatory cascade [⁶²].

CRH may also facilitate the expression of contraction-associated genes. In myometrial smooth muscle cells from nonpregnant women, CRH enhances connexin-43 mRNA and protein expression through nuclear transcription factor activator protein 1 (AP-1) activation and upregulation of c-Fos expression, which is an AP-1 subunit. This is accomplished in a positive CRH-dose-dependent manner. As a result, in late pregnancy the high levels of CRH possibly cause an associated increased expression of connexin-43 that promote myometrial contractions [⁶³]. AP-1 also increases gene expression in response to many other stimuli, including cytokines [⁶⁴]. CRH may also increase the myometrial response to prostaglandin F2α and therefore modulate the onset of labor [⁶⁵].

CRH exerts its actions through activation of specific receptors, termed CRH-R1 and CRH-R2. These receptors belong to the family of seven transmembrane (7TMD) G protein coupled receptors (GPCRs). CRH receptors are expressed in several tissues in pregnancy. Both CRH-R1 and CRH-R2 and their variants have been detected in human myometrium, and it is believed that the expression of various isoforms plays a role in both myometrium quiescence and transition to a contractile phenotype during labor [⁶⁶]. Alternative splicing of CRH-R mRNA results in different receptor isoforms (such as CRH-R1α and CRH-R1d) with different activities that generate a more contractile phenotype [⁶⁷]. In myometrial strips from laboring and nonlaboring pregnant women, it was shown that CRH could inhibit spontaneous contractility of nonlaboring but not of laboring myometrium, acting via the CRH-R1 receptor. This effect further implies that CRH may exhibit a dual role at the myometrium [⁶⁸].

It is suggested that CRH-R1 promotes relaxation of smooth muscles via intracellular signals that inhibit phosphorylation of the contractile protein myosin light chain (MLC20). The decreased expression of proteins such as Gα_s-protein (a G protein subunit that activates the cAMP-dependent pathway by activating adenylate cyclase) towards the end of pregnancy turns the balance towards myometrial contractility. The increased expression of CRH-R2 protein toward onset of labor increases MLC20 phosphorylation and myometrial contractility [⁶⁷].

CRH-R1 and CRH-R2 have been identified both in the upper and lower uterine segment of laboring and nonlaboring human myometrium, but during labor CRH-R1 decreases significantly at the upper segment but not at the lower one. This decreased expression may be related to the fact that during labor the fundus of the uterus switches to a highly contractile state, while the lower segment remains relatively quiescent [⁶⁹]. Recently a new isoform of CRH-R has been identified in human pregnant but not in nonpregnant myometrium. This isoform is termed CRH-R1β/d and shows similarities with both CRH-R1β and -R1d, such as reduced affinity for CRH, negligible signaling, and mainly cytoplasmic localization. Its expression is maximum at the early third trimester, and it is increased by estradiol-17β [⁷⁰].

CRH receptors may also be activated by other agonists such as the urocortins. Urocortin 2 (UCN 2) is a CRH-R agonist that is expressed in human pregnant myometrium and interacts with CRH-R2 receptors and through a signal transduction pathway involving the protein kinase C (PKC), the extracellular signal-regulated kinases ERK1/2, and the RhoA/ROK. Urocortin 2 interaction stimulates MLC20 phosphorylation, which in turn induces myometrial contractility [⁷¹]. Although urocortins 2 and 3 are expressed in the human placenta, it is not known if they act by increasing myometrial contractility during labor [⁷²]. Recent data have reported that, in human placentas from women with PPROM and chorioamnionitis, inflammatory events are associated with changes to the CRH-related mechanism of labor, such as increased CRH, UCN 2, and CRH-R1 mRNA expression and decreased UCN 3 and CRH-R2 expression [⁷³].

Immune system responses are also associated with changes to the CRH-related mechanisms of labor. Experiments on both term and preterm, laboring and nonlaboring human myometrium show that active labor is associated with increase of CRH-R1α and CRH-R1β mRNA in term and preterm myometrium. IL-1β induces CRH-R1 _Tα mRNA and prostaglandin synthetase 2 (PTGS2) mRNA and protein levels, decreases CRH-R1β mRNA, and impairs CRH-induced cAMP production. It may therefore be hypothesized that IL-1β acts as a regulator of CRH-R1 expression and function that contributes to a preparatory contractile environment necessary for the initiation of labor [⁷⁴].

CRH may also act on both term and preterm cervical fibroblasts and increase IL-8 expression, which is an important mediator of cervical ripening [⁷⁵].

Several blood, amniotic fluid, or vaginal fluid markers have been proposed as predictors of preterm labor, including interleukins, cortisol, fibronectin, ADAM-8, ITAC, estriol, a-FP, b-HCG, and others [⁷⁶–⁸¹]. Many of these studies have evaluated the clinical usefulness of CRH measurements to predict preterm labor [⁵⁶, ⁸², ⁸³]. Urocortin may also serve as a biomarker of labor in women with threatened preterm birth before 34 weeks [⁸⁴]. However, these observations have mostly diagnostic and not therapeutic benefit, since no CRH/CRH receptor antagonist has so far been identified that can be used in humans to prevent preterm labor.