Main

Parturition is the culmination of mammalian reproduction, a task essential for survival of the species. After a period of uterine quiescence to allow fetal growth and development, changes occur in myometrial contractility that result in efficient expulsion of the fetus. These events have likely evolved to enhance survival of both the fetus and the mother. The nature of these pathways, however, is at best incompletely defined.

Increased knowledge of the cascade of events that occur at parturition may lead to advances in combating preterm labor (PTL) or optimization of protocols for medically induced labor. As preterm birth is associated with both increased risk of neonatal mortality and chronic sequelae such as respiratory illness, cerebral palsy, and vision and hearing impairment, it is a major public health concern (1). In the United States, 12.7% of births are preterm (2). Although approximately 50% of preterm births are idiopathic (3), genetic factors seem to be important. Women whose mothers or sisters delivered preterm are at increased risk for premature labor (4,5), indicating a genetic component in the timing of labor. In addition, a study of a Utah population indicates greater genetic relatedness between families with preterm deliveries than control families (6). Because the population studied was descended from the people who established Mormonism in Utah and Mormons have low rates of substance abuse and sexually transmitted diseases, the population used in this study may represent individuals with few environmental risk factors for PTL (6). These studies highlight a role for genetic, and not just environmental factors in predisposition to PTL and serves as a call to researchers to identify and study genes important for parturition.

The mouse is a useful research tool for dissecting genetic factors involved in developmental processes. Mice are tractable to genetic manipulation, resulting in an array of available and potential knockout and transgenic mice suitable for studying the roles of specific genes in complex processes. Important to the study of parturition, mice have a conveniently short duration of gestation. Although the mouse has many advantages as a research tool, its relevance for increasing the understanding of human parturition has been questioned because of some dissimilarity in the parturition cascades of mice and humans. At the onset of normal murine parturition, prostaglandins trigger luteolysis, the structural and functional degradation of the steroidogenic corpora lutea, and a withdrawal of progesterone (Fig. 1) (7). Humans exhibit no such decline of serum progesterone at term (8). However, a functional progesterone withdrawal, mediated by a decrease in the progesterone receptor's (PR's) transcriptional activity, may be a part of normal human parturition (9). Furthermore, genetically altered mice in which parturition occurs without a dependence on declining serum progesterone levels have been described (10,11). This emphasizes the importance of other mechanisms of the parturition cascade known to be present in both mice and humans. In each species, there is an up-regulation of contractile-associated proteins (CAPs), such as oxytocin receptor (OTR), prostaglandin F2αreceptor (FP), and connexin43 (Cx43), in the uterus at term (12–15). Parturition studies using genetically altered mice are reviewed here (Table 1).

Figure 1
figure 1

Schematic for the cascade of events culminating in murine labor. Increased synthesis of PGF2α in the uterine epithelium by the sequential actions of cPLA2, COX-1, and PGF synthase results in luteolytic action through FP receptors in the ovary. A drop in circulating progesterone with luteolysis induces expression of myometrial CAPs and labor. CAP, contractile-associated protein; cPLA2, cytoplasmic phospholipase A2; COX-1, cyclooxygenase-1; FP, PGF2α receptor; PGF2α, prostaglandin F2α.

Full size image
Table 1 Parturition phenotype of genetically altered mice
Full size table

PROGESTERONE WITHDRAWAL

Throughout gestation, progesterone maintains uterine quiescence. A fall in circulating progesterone occurs with the onset of murine (7), but not human (8) labor. In mice, ovariectomy (OVX) induces PTL, whereas administration of progesterone delays parturition (16). Genetically altered mice have been used to study mechanisms for declining serum progesterone levels in mice and possible alternative mechanisms of progesterone withdrawal in humans.

20α-Hydroxysteroid Dehydrogenase Impacts Serum Progesterone Levels

20α-HSD has an important role in coordinating progesterone withdrawal at term in mice (17). 20α-hydroxysteroid dehydrogenase (20α-HSD) converts progesterone to an inactive metabolite and is expressed in the ovary (18). Although expression of 20α-HSD is barely detectable during mid to late gestation, it is up-regulated considerably at term (17). 20α-HSD−/− mice fail to initiate labor at term (17). Instead, labor is delayed 2–3 d compared with 20α-HSD+/+ controls, and the pups die in utero (17). In term 20α-HSD−/− females, progesterone fails to fall to the same level that it does in term wild-type mice (17). Administration of the progesterone antagonist RU486 to 20α-HSD−/− females 1 d before term results in productive parturition the next day without a decrease in circulating progesterone levels (17). This result indicates that the singular role for 20α-HSD in gestation and parturition is to regulate the levels of serum progesterone (17).

Prostaglandins Mediate Luteolysis and Progesterone Withdrawal

Progesterone withdrawal in mice is concurrent with luteolysis, the structural and functional degradation of the corpora lutea. A role for prostaglandin F2α (PGF2α) in luteolysis and progesterone withdrawal in mice has been demonstrated by the study of FP knockout mice (7). FP is normally expressed in the corpus luteum (19). FP−/− females establish and maintain pregnancies normally, but they fail to initiate labor (7). Serum progesterone levels do not decline in these animals at term, indicating that their corpora lutea retain their steroidogenic function (7). OVX in these dams leads to progesterone withdrawal and successful parturition (7). Mice deficient for prostaglandin receptors other than FP have no parturition defects, indicating that PGF2α is the critical prostaglandin isoform for inducing luteolysis (20).

The importance of prostaglandins in precipitating luteolysis and progesterone withdrawal is further supported by similar phenotypes in cytoplasmic phoshospholipase A2 (cPLA2)−/− and cyclooxygenase-1 (COX-1)−/− mice (21–24). cPLA2 and COX-1 are enzymes responsible for the synthesis of prostaglandins. cPLA2 converts membrane phospholipids to arachidonic acid. Arachidonic acid is then converted to prostaglandin H2 (PGH2), the intermediate required by PGF synthase, by different isoforms of COX (Fig. 2). cPLA2−/− females have a delay in labor (21,24). Administration of PR antagonist to gravid cPLA2-deficient females precipitates labor, indicating that persistent progesterone levels are responsible for delayed labor in these mice (24). COX-1−/− females also exhibit delayed parturition (22,23) and persistent progesterone levels (22). Labor can be induced at term by administration of PGF2α(22). No parturition phenotype has been reported for COX-2−/− mice because defects in ovulation and implantation prevent progression of gestation to this stage (25). However, COX-2 is up-regulated in the myometrium during parturition suggesting a role for this COX isoform in labor progression rather than initiation (26).

Figure 2
figure 2

Biochemical pathway of prostaglandin synthesis and degradation. Arachidonic acid conversion to the metastable intermediate PGH2 occurs through the action of the COX enzymes. PGH2 is converted to the specific prostanoid isoforms for receptor action by their respective synthases/isomerases. PGE2 and PGF2α, the main prostaglandins involved in parturition, are degraded by 15-HPGD. 15-HPGD, 15-hydroxyprostaglandin dehydrogenase; cPLA2, cytoplasmic phospholipase A2; COX, cyclooxygenase; PG, prostaglandin; TXA2, thromboxane A2.

Full size image

Models for Functional Progesterone Withdrawal in Humans

Although serum progesterone does not fall at term in humans (8), questions remain as to whether or not a functional progesterone withdrawal is part of human parturition. In some cases, administration of the PR antagonist mifepristone augments cervical ripening and precipitates labor in term women (27), indicating that functional progesterone withdrawal may have a role in human labor. Because it has been observed that each of the PR isoforms has distinct properties as a transcription factor (28), differential expression of PR isoforms at term seems a likely mechanism for functional progesterone withdrawal in humans. A recent mouse study indicates a potential importance of PR regulation at term. Mice deficient for Krüppel-like factor, Klf9, an Sp6 family transcription factor, have a delay in parturition with corresponding aberrant PR isoform A expression in the myometrium and insensitivity to progesterone antagonist at late gestation (29). Studies have been conducted on the ratio of expression of the different PR isoforms in the laboring human myometrium (30,31), but it is difficult to draw a compelling argument for PR regulation as a means of functional withdrawal from the data. However, evidence supporting functional progesterone withdrawal in humans comes in the form of an observed decrease in coactivators with histone acetylase activity in both humans and mice at term (9). Both expression of coactivators with histone acetylase activity and acetylated histone H3 are decreased at term in humans and mice (9). When trichostatin A (TSA), a histone deacetylase inhibitor, was administered to pregnant mice daily during the late gestation, parturition was delayed 24–48 h (9). TSA administration in this fashion leads to 200 times the histone acetylation of an untreated mouse at term (9). It is proposed that a decrease in coactivators and acetylation at term interrupts the transcriptional activity of the progesterone-PR complex leading to a functional progesterone withdrawal (9).

MYOMETRIAL CONTRACTION

Preparation for the myometrium to contract as one powerful unit is an important step in the start of the parturition cascade. As term approaches, expression of the CAPs, including Cx43, OTR, and FP, increases in the myometrium of humans and mice (12–15). The myometrium is thus activated and therefore able to respond to stimulants such as prostaglandins, which induce myometrial contractions and sensitize the myometrium to other uterotonic agents. Uterine PGF2α also increases to its highest levels just before labor begins (12). Prostaglandins and CAPs are important players for transforming the uterus, quiescent for most of gestation, to the powerful contractile organ necessary for expelling the fetus.

The Uterotonic Agent Oxytocin is Not Necessary for Parturition

Oxytocin (OT) is a strong uterotonic agent and is frequently used to induce labor (32,33). Although OTR expression increases in the myometrium approximately 10-fold at term (34), OTR−/− and OT−/− females have a normal timing and duration of labor (35–37). Surprisingly, COX-1−/−/OT−/− mice have a prolonged labor that initiates at normal term (22). This is also in contrast to the COX-1−/− mouse, which has impaired luteolysis and delayed labor (22,23). COX-1−/−/OT−/− mice undergo luteolysis normally (22). Therefore, it seems that in addition to having a role as a contractile agonist at the level of the uterus, OT has a role at the level of the ovary for maintaining the corpus luteum (22). Further examination of the role of OT in gestation and parturition has shown that infusion of OT leads to a delay in labor at low doses, highlighting its role in maintaining the corpus luteum, and precipitation of labor at high doses indicating its role as a uterine contractile agonist (38).

Expression of the Gap Junction Protein Cx43 in the Myometrium is Necessary for Parturition

The necessity of the CAP Cx43, which synchronizes myometrial cells, for successful parturition has been demonstrated by study of a novel genetically altered mouse (39). Because conventional Cx43−/− mice die shortly after birth (40), Cre-LoxP technology (41) was used to generate mice deficient for Cx43 specifically in smooth muscle tissues including the myometrium (39). Whereas the vast majority (89%) of gravid control mice deliver their litters between 4 and 8 am on d 19.5 of gestation, few (18%) of the myometrial Cx43−/− females deliver during this time period (39). The majority of myometrial Cx43−/− mice instead deliver after 8 am on d 19.5, with approximately half of that group delivering dead pups on d 20–22 of gestation (39). Uterine OTR and FP are up-regulated and serum progesterone declines normally in the myometrial Cx43−/− mice at term (39). This study highlights the importance of Cx43 and the coupling of myometrial cells for productive uterine contractility as well as the utility of the conditional knockout system for studying parturition.

15-Hydroxyprostaglandin Dehydrogenase Decreases at Term to Allow Prostaglandin Up-regulation

Mutant mice have shown a role for the synthesis and action of prostaglandins in the initiation of parturition (7,21–24). However, less is known about how the degradation of prostaglandins may be regulated to achieve increased prostaglandin levels and parturition at term. 15-Hydroxyprostaglandin dehydrogenase (15-HPGD) is the principal enzyme responsible for the breakdown of PGF2α as well as the less robustly expressed prostaglandin E2 (PGE2) (42,43). In humans, 15-HPGD mRNA decreases in chorion trophoblast cells in both term and preterm laboring women versus nonlaboring women (44,45). A similar decrease is noted in laboring mice (46). To examine the role of regulation of expression of this gene in the parturition cascade, mice hypomorphic for 15-HPGD were generated (11). These animals have a significant decrease in 15-HPGD activity in placenta, uterus, and fetal membranes compared with control animals (11). When mated to hypomorphic males, hypomorphic females labor approximately 1 d early (11). In the case of these pregnancies, PGF2α and PGE2 levels rise early (11). Interestingly, their labor proceeds without progesterone withdrawal, perhaps due to an ability of PGE2 to maintain the corpus luteum in the face of increasing PGF2α levels (11). These data indicate a role for decreased catabolism of prostaglandins, in addition to increased synthesis, in initiation of the parturition cascade (11). It also suggests that prostaglandins can mediate the induction of labor by mechanisms besides luteolysis and resulting progesterone withdrawal.

Small-Conductance, Calcium-Activated K+ Isoform 3 Channel Expression Decreases in Term Myometrium to Promote Contractility

The activity of several types of K+ channels in the myometrium is regulated throughout pregnancy to maintain uterine quiescence during gestation and generate forceful contractions at term (47–49). The role of small-conductance, calcium-activated K+ isoform 3 (SK3) channels in parturition has been examined using genetically altered mice. SK3-overexpressing females exhibit defective parturition, manifested by prolonged delivery and mortality of pups and dams (50). Weaker contractions are observed in uterus from term SK3-overexpressing versus wild-type females (51). Moreover, when administered stimuli precipitate PTL in wild-type mice, SK3-overexpressing mice do not complete labor even if serum progesterone levels decline (51). SK3 expression is decreased in both mice and humans at term (49,51), indicating an importance of regulation of this channel for parturition in both species.

CERVICAL RIPENING

Along with conversion of the quiescent uterus to a powerfully contracting unit, remodeling of the formerly rigid cervix is an important process for enabling successful parturition. During pregnancy, the cervix must be firm so that the fetus is not prematurely expelled. However, as term draws near, cervical softening and remodeling occurs, making the cervix more compliant and amenable to the birthing process. This process involves tissue growth, increase in cervical secretions, infiltration by inflammatory cells, and alterations in the extracellular matrix (52).

Relaxin Mediates Normal Cervical Ripening

Relaxin (RLX) is a hormone known to be important for connective tissue remodeling in sites including the cervix and pelvic ligaments (53,54). The primary source of rodent RLX is the corpus luteum (55). In rodents, levels of RLX increase during the second half of pregnancy (56). Studies in rats have indicated that in the absence of circulating RLX, labor is prolonged or prevented (53). The RLX knockout mouse is fertile, but has a low penetrance nonproductive labor phenotype (57). Duration of labor, measured as the time elapsed from delivery of the first pup until delivery of the last, was reported to be approximately 2 h in seven of seven RLX+/+ and six of eight RLX−/− mice (57). The remaining two RLX−/− females had abnormal labor. One of these females had a prolonged labor lasting approximately 15 h (57). The other birthed two dead pups and was killed 12 h later because of distress (57). Postmortem analysis revealed the presence of 10 additional dead pups in utero (57). Accordingly, a low penetrance phenotype of nonproductive parturition was also observed in dams deficient for LGR7, the highest affinity RLX receptor (58). The protracted labor phenotype associated with these two knockout models can perhaps by explained by a lack of normal pregnancy-related changes in the cervix, pubic symphysis, and vagina of RLX−/− females (57,59). Weight of pubic symphysis, vagina, and cervix in RLX−/− females does not increase by 1 d before term at the magnitude it does in RLX+/+ females (59). In addition, pregnant RLX−/− females have denser collagen staining in vagina, pubic symphysis, and cervix than do RLX+/+ females 1 d before term (59). Also, the pubic bones of RLX−/− females fail to widen to the extent that they do during the second half of pregnancy in wild-type gravid females (57). These data support a role for RLX in preparation of the birth canal for labor.

Progesterone Metabolism by Steroid 5α-Reductase Type 1 at the Cervix Allows Ripening

5α-Reductase type 1 (5α-R1)−/− mice also have a parturition defect attributable to poor cervical ripening (60,61). 5α-R1 metabolizes progesterone, and its expression has been noted in many tissues of the female reproductive tract (60). The majority of 5α-R1−/− dams have prolonged labor at 2–3 d postterm (60). A quarter of these females die during labor (60). If the pups are delivered via caesarean section at term, they are normal and can be nursed by foster mothers (60). 5α-R1−/− females delivering on time were noted to have decreased litter size (60). During the second half of gestation, circulating steroid hormone levels are comparable between 5α-R1+/+ and 5α-R1−/− dams (60). Additionally, gravid 5α-R1−/− females exhibit contractions that are sufficient for labor (61). However, at term, 5α-R1−/− females have a matrix of collagen fibers that is denser and more compact than that observed in 5α-R1+/+ females at term (61). Further analysis revealed that progesterone is elevated in the cervix and uterus of 5α-R1−/− females at term, despite the fact that a serum progesterone decline is observed (61). The prolonged labor phenotype observed can be reversed by administration of PR antagonists or OVX, as well as by administration of some 5α-reduced steroids, RLX, or OT (60,61). This novel report suggests the importance of local progesterone metabolism to the process of cervical ripening. Local progesterone concentrations also seem to be regulated for cervical ripening in humans, although by a different enzyme. 17β-Hydroxysteroid dehydrogenase, which converts 20α-hydroxyprogesterone to progesterone, is down-regulated in the endocervical cells in term women, leading to decreased local progesterone levels and favorable conditions for cervical ripening (62).

Additional studies using 5α-R1−/− female mice have indicated some potentially important participants in the process of cervical ripening. Hyaluronan (HA) increases during late pregnancy in a variety of mammals studied including the mouse and human and is the primary glucosaminoglycan present in the cervix at labor (63–66). HA is speculated to be important in cervical ripening because it has a high water affinity and increased tissue hydration may prevent the aggregation of collagen fibrils and reduce the tensile strength of this tissue (65,67). In 5α-R1+/+ females, hyaluronan synthase 2 (HAS2) mRNA is the HAS isoform found in greatest abundance in cervix 1 d before parturition, when cervical HA content is also increased (68). In the 5α-R1−/− female, HAS2 expression is reduced by 70% compared with 5α-R1+/+ females 1 d before term, correlating with a 68% reduction in HA compared with 5α-R1+/+ females (68). HAS2 up-regulation is also observed in women in labor (68). These data indicate a potential importance for the regulation of this enzyme in remodeling of the cervical extracellular matrix near term (68). In addition, 5α-R1−/− females do not display the late pregnancy-associated changes in expression of aquaporins and tight junction proteins seen in 5α-R1+/+ females, indicating a role for these types of proteins in mediating changes in cervical water content associated with cervical ripening (69,70).

CIRCADIAN INFLUENCE ON LABOR

A newly emerging area of interest is the relationship between circadian rhythmicity and parturition. Circadian rhythmicity is manifested by daily oscillations in locomotor activity, hormone levels, metabolism, and other parameters (71). A molecular clock comprised of several genes, among them the heterodimer partners Clock and Bmal1, underlies this phenomenon (71). The onset of parturition seems to be under circadian influences. Epidemiologic data from two separate studies indicate that humans tend to labor in the early morning hours (72,73). Time of day seems to factor into murine labor as well, as mice tend to labor during the dark phase of their daily cycle. This raises the question of how the molecular clock interacts with molecular events of the parturition cascade. The Clock mutant mouse, which has a deletion in Clock's transcriptional activation domain, has been noted to have a high incidence of extended but nonproductive labor indicating a role for genes of the molecular clock in the parturition process (74). In addition, we recently reported a possible role for OT in maintaining the normal circadian gating of labor in mice (75). A 6-h phase advance or delay in wild-type dams starting on d 4.5 of gestation precipitated no overall change in birth timing, whereas the same 6-h phase advance or delay in OT−/− dams leads to a random scattering of the time of birth throughout the day (75). OT is released in a circadian fashion. This rhythmicity of secretion is not altered in wild-type gravid females upon change in light:dark cycle although both wild-type and OT−/− dams shift their locomotor activity patterns in response to the disturbance (75). Therefore, it seems OT secretion may maintain the onset of labor with original circadian entrainment of pregnancy (75).

BACTERIALLY INDUCED PTL

Infection seems to contribute to or precipitate PTL in many instances, with 30% of PTL cases exhibiting intrauterine microbial colonization or histologic features of chorioamnionitis (76). Bacterially induced PTL is associated with an activation of the inflammatory response, including up-regulation of cytokines, chemokines, and prostaglandins (Fig. 3) (77). Even normal spontaneous labor has elements of an inflammatory response (78–80). Much study has been conducted in mice to better define the pathogenesis of bacterially induced PTL. Most mouse studies relevant to this phenomenon use injections of heat-killed Escherichia coli bacteria (HKE) (77) or lipopolysaccharide (LPS) (81), a component of the outer membrane of Gram-negative bacteria, to induce PTL.

Figure 3
figure 3

Bacterial infection-associated preterm birth pathway. Bacterial products, for example, endotoxin (LPS), activate toll-like receptors to produce proinflammatory products such as cytokines and prostaglandins that accelerate the time for parturition. NF-κB, nuclear factor-κB.

Full size image

No Progesterone Withdrawal in Bacterially Induced Labor?

Reports on whether or not bacterially induced labor is accompanied by progesterone withdrawal in mice are conflicting. Fidel et al. (81) observed a significant decrease in serum progesterone in mice injected peritoneally with LPS to precipitate PTL. However, Hirsch and Muhle (10) found that intrauterine injection of a high dose (109) of HKE did not result in a decrease of circulating progesterone of the magnitude observed resulting from OVX. However, the high dose of HKE precipitated labor more quickly than OVX (10). Whereas administration of progesterone at pharmacological levels lengthened the time to delivery in both the high-dose HKE and OVX groups, administration of progesterone at physiologic levels lengthens time to delivery in the OVX group only (10). The results of this study indicate that progesterone withdrawal is not the chief mechanism for bacterially induced PTL (10). The conflicting nature of these two reports may be because of the different routes of infection, with the peritoneal LPS injections in the first study leading to a systemic infection and the intrauterine HKE injection of the latter study leading to a more localized infection (10).

Cytokine Roles in Bacterially Induced PTL

A host of cytokines, such as IL-1, IL-6, and TNF, are found to be up-regulated in cases of infection-associated PTL in humans and mice (77). In animal models, IL-1 and TNF can induce increased prostaglandin levels, CAP up-regulation, uterine contractility, and PTL (82–86). Genetically altered mice have been helpful in determining the importance of specific cytokines in bacterially induced PTL. No individual cytokine has been found necessary for the pathogenesis of bacterially induced PTL. Mice deficient for IL-1 receptor (IL-1RI), which are completely devoid of IL-1 signaling, are not protected from PTL precipitated by HKE injections at multiple sites (87). Accordingly, IL-1β−/− females inoculated with HKE or LPS were as likely to undergo PTL as wild-type females undergoing the same treatment (88). Similarly, IL-6 seems to be dispensable for the pathogenesis of bacterially induced PTL as IL-6+/+ and IL-6−/− females are equally susceptible to PTL triggered by uterine HKE inoculation (89). To test the hypothesis that cytokines have redundant roles in the pathogenesis of bacterially induced PTL, mice lacking signaling from multiple cytokine receptors were examined. Mice treated with the anticytokines IL-1 receptor antagonist (IL-1ra) and soluble TNF receptor Fc fusion protein in combination with LPS undergo PTL at similar frequency as wild-type females treated with LPS alone (90). However, another report using IL-1 receptor (IL-1R1)−/−/TNF receptor (Tnfsrsf1a)−/− mice shows a decreased susceptibility to PTL induced by inoculation with HKE in the double knockouts (91). Perhaps, the discrepancy is due to complete eradication of receptor signaling in the knockout paradigm in contrast to residual receptor signaling in the case of anticytokine treatment. If this is the case, higher doses of these anticytokines may be more effective in inhibiting the observed PTL. The IL-1R1−/−/Tnfsrsf1a−/− study seems to indicate that having either IL-1 or TNF is necessary for the pathogenesis of bacterially induced PTL. However, there has been no report on the phenotype of Tnfsrsf1a single knockout mice. Such a study is warranted to establish whether the necessity for TNF signaling is responsible for the double knockout phenotype (91).

The role of cytokines in the pathogenesis of bacterially induced PTL was further explored in the IL-1R1−/−/Tnfsrsf1a−/− system (91). The double knockout mice showed significantly lower myometrial expression of COX-2 upon HKE inoculation compared with wild-type females indicating the importance of COX-2 in the pathogenesis of bacterially induced PTL (91). They did not show differential regulation of genes associated with cervical ripening compared with treated wild-types, indicating that the process of cervical ripening in bacterially induced PTL differs from that in term laboring females (91).

Toll-like Receptor Dependence of Bacterially Induced PTL

Bacteria are thought to initiate the inflammatory process and thereby precipitate PTL by interacting with toll-like receptors (TLRs), which orchestrate an immediate immune response. TLR stimulation leads to the activation of nuclear factor-κB (NF-κB), which induces cytokine and chemokine transcription (77). To examine whether TLRs are necessary for bacterially induced PTL, mice with normal TLR4 and mice with mutant TLR4, leading to LPS hyporesponsiveness, were used (46). Intrauterine injection of HKE resulted in PTL in the normal TLR4 but not mutant TLR4 strain at most doses (46). At the highest HKE dose (1010), 28% of mutant TLR4 females and 100% of normal TLR4 females underwent PTL (46). At 5 × 109 HKE, a dose sufficient to cause PTL in 100% of normal mice but none of the TLR4 mutant mice, cytokine and COX-2 up-regulation was observed in both strains (46). However, HPGD expression decreased in normal but not mutant females in response to HKE (46). This indicates that TLR4 may facilitate PTL by lowering prostaglandin catabolism (46).

Nonsteroidal Anti-inflammatory Drugs as Therapeutics

Because bacterially induced PTL is associated with an increase in prostaglandin production (77), inhibiting the synthesis of prostaglandins is a logical therapy in cases of potential infection-related PTL. Administration of nonsteroidal anti-inflammatory drugs (NSAIDs), which inhibit prostaglandin synthesis, have been shown to curtail the progression of both term labor and PTL (16,92–94). Currently used NSAIDs, such as indomethacin, block COX-1 and COX-2. Treatment with these drugs is associated with fetal and maternal side effects that have precluded their use (92,95). The efficacy of NSAIDs that inhibit a single COX isoform and their associated side effects were studied in mice as a potential remedy to these problems (96). Inoculation with LPS resulted in PTL within 24 h in all gravid females, whereas administration of LPS and indomethacin led to PTL in only 20% of dams (96). Most of the nonlaboring females demonstrated in utero fetal death (96). Coadministration of LPS and the COX-1 inhibitor SC-560 resulted in PTL in 58% of females, with 40% of nonlaboring females exhibiting in utero fetal demise (96). More successful was the coadministration of LPS and the COX-2 inhibitor SC-236 (96). This treatment resulted in PTL in only 8% of treated females, with 73% of the remaining females delivering viable fetuses (96). These data, in concert with the observation that 88% of gravid COX1−/− females treated with LPS deliver within 24 h, indicate that COX-2 derived prostaglandins are important for bacterially induced PTL and that COX-2 is a potentially important target for stopping PTL (96).

SURFACTANT PROTEIN-A AS PARTURITION SIGNAL

Pulmonary surfactant may have a role in regulating the inflammatory response associated with even normal term labor (97). Surfactant production begins in the fetal lung during the final third of gestation in humans and mice (97,98). Inadequate surfactant production results in respiratory distress in infants (98). A recent study suggests that in addition to its role in pulmonary function, secretion of surfactant, particularly surfactant protein-A (SP-A), from fetal lung has a role in signaling for parturition to begin in mice (97). It is proposed to play this role by activating amniotic fluid macrophages which travel to the maternal uterus and incite the inflammatory response (97). SP-A is found to up-regulate IL-1β and NF-κB in cultured amniotic fluid macrophages (97). Experiments using wild-type dams carrying fetuses heterozygous for β-galactosidase revealed that macrophages of fetal origin increase in the maternal uterus during the timeframe at which SP-A is secreted from fetal lung (97). Intraamniotic injection of SP-A caused PTL in gravid female mice within 6–24 h (97). Conversely, the injection of an SP-A antibody or NF-κB inhibitor led to a delay of labor more than 1 d postterm (97). In laboring humans, macrophages of fetal origin are not present in the myometrium (99,100). Although SP-A appears not to promote migration of fetal macrophages to the myometrium in women, it may have a different role in precipitating human labor (100).

CONCLUSION

The mouse has proved to be an important resource for increasing our knowledge of the processes related to parturition. Although the importance of CAPs and prostaglandins is evident in both humans and mice, the differences in progesterone regulation between the two species is more perplexing. Research into potential functional progesterone withdrawal in the human may lead us to an increased appreciation of similarity between human and murine labor. Additionally, the established murine models recapitulating infection-associated PTL are promising for increasing the understanding of inflammatory processes related to labor. Further knowledge on cervical ripening gained from the RLX and 5α-R1 knockout mice may lead us to new therapeutic targets for stimulating or delaying labor. The expanding circadian field provides us with a new group of genes whose roles in parturition merit further exploration. The ability of 15-HPGD hypomorphic mice, and potentially mice experiencing bacterial infection, to labor without a dependence on progesterone withdrawal indicates the importance of other mechanisms that are a part of labor initiation. Because these other mechanisms crucial for labor initiation are likely shared between mice and humans, continuing analysis of rodent models should yield further understanding of the physiology of pregnancy across species.