Functional Hierarchy of Uterotonics Required for Successful Parturition in Mice

Our experiments showed that oxytocin receptor signaling is necessary for induction of COX-2 and connexin 43 expression during parturition and that both oxytocin/oxytocin receptor signaling and PGF_2α receptor signaling are major components for successful parturition. In the current study, it was found that expression of the oxytocin receptor, PGF_2α receptor, COX-2, and connexin 43 in uteri increased during parturition, whereas the expression of COX-1 and PGE₂ receptors 1 to 4 decreased or did not change during parturition compared with that in WT on GD 17.0. Alternatively, an increase in the expression of COX-2 and connexin 43 during parturition was not observed in the uteri of Oxtr^−/−. The importance of COX-2 and connexin 43 in parturition has been suggested in previous reports. It has been shown that inflammation induced COX-2 expression in the uterus and that inhibition of the enzymatic activity of COX-2 prevented inflammation-mediated preterm labor (12). Connexin 43 contributes to the induction of physical and biochemical connectivity among myometrial cells and to the formation of extensive waves of depolarization and contraction over large areas of the uterus (23). Decline of blood progesterone concentration is critical for the onset of parturition in rodents. There is evidence that the progesterone receptor directly suppresses the expression of COX-2 and connexin 43. Previous studies showed that the progesterone receptor inhibited induction of COX-2 expression via both direct interaction with nuclear factor κB (NF-κB) (24) and induction of expression of the NF-κB inhibitor IκBα in myometrial cells (25). The progesterone receptor also binds to the promoter region of connexin 43 through p54nrb (non-POU domain–containing, octamer binding protein) and directly suppresses connexin 43 expression (26). Alternatively, there is evidence that the progesterone receptor indirectly suppresses oxytocin receptor expression. It was reported that ZEB1 and ZEB2, zinc finger E-box binding homeobox proteins, suppress oxytocin receptor transcription in the uterus until parturition. A decline of progesterone concentration induces downregulation of ZEB1 and ZEB2 at term. This downregulation induces oxytocin receptor transcription at term (27). There is no evidence that the progesterone receptor directly or indirectly suppresses expression of Ptgfr. In the current study, upregulation of COX-2 and connexin 43 transcription was impaired in the uterus of Oxtr^−/− during parturition, whereas PGF_2α receptor expression was not. Because Oxtr^−/− showed a progesterone concentration similar to that in WT on GD 19.0, the lower induction of COX-2 and connexin 43 expression is not caused by insufficient progesterone decline on term. The oxytocin receptor, COX-2, and connexin 43 are all expressed in the rodent myometrium in late pregnancy (17, 28, 29). It was reported that the oxytocin receptor activated the MAPK pathway and NF-κB via the protein kinase A pathway. Oxytocin increased COX-2 expression and prostaglandin synthesis via MAPK in vitro (30). A study using ovariectomized Ptgfr^−/− showed that COX-2 expression increased after induction of oxytocin receptor expression in the uterus (28). Oxytocin receptor signaling also enhanced connexin 43 protein expression via activated NF-κB in vitro (31). It is likely that the MAPK pathway and the protein kinase A–NF-κB pathway via oxytocin receptor signaling enhance COX-2 and connexin 43 expression during parturition after the decline of progesterone concentration. To our knowledge, our findings provide the first evidence that oxytocin receptor signaling is essential for enhancement of COX-2 and connexin 43 expression during parturition.

Deletion of the PGF_2α receptor gene or COX-1 gene (a major gene responsible for the synthesis of PGF_2α in luteolysis) in mice resulted in a deficit of luteolysis at term, and progesterone concentration did not decline during late pregnancy in these deficient mice (9, 32). PGF_2α receptor signaling induces downregulation of blood progesterone concentrations by inhibition of progesterone biosynthesis in the corpora lutea (33). Alternatively, it was reported that mice with double knockout of the genes for COX-1/oxytocin showed normal onset of parturition. Oxtr expression in the corpora lutea declined in late pregnancy, and exogenous oxytocin infusion delayed the decline of progesterone concentration and onset of parturition (32). Those studies suggest that oxytocin has a luteotrophic effect in pregnancy, although Oxt^−/− and Oxtr^−/− showed normal parturition and a serum progesterone decline similar to that in WT mice. In the current study, Oxtr^−/−;Ptgfr^−/− and Oxt^−/−;Ptgfr^−/− showed a plasma progesterone concentration similar to that in Ptgfr^−/− on GD 19.0. Considering the previous study, oxytocin receptor signaling appears to have the opposite action of PGF_2α in luteolysis in late pregnancy. However, oxytocin and oxytocin receptor gene deficiencies were not able to restore the decline of progesterone concentration and normal onset of parturition in complete blocking of PGF_2α receptor signaling by gene deficiency, unlike COX-1 gene deficiency. These findings suggested that a luteotrophic action of oxytocin/oxytocin receptor signaling in the corpora lutea in late pregnancy is negligible.

In the current study, Oxtr^−/−;Ptgfr^−/− and Oxt^−/−;Ptgfr^−/− administered RU486 were capable of starting parturition. Uteri in both Oxtr^−/− and Ptgfr^−/− at late pregnancy lost contractile responses to oxytocin and PGF_2α, respectively (3, 9). These results suggest that there are certainly other uterine contractile mechanisms for the onset of parturition other than oxytocin/oxytocin receptor signaling and PGF_2α receptor signaling. Inhibition of the enzymatic activity of COX-2 postponed the onset of parturition in mice (28, 34). PGE₂ has been reported to have potent uterotonic activity in the periparturient uterus. PGE₂ receptors 1 and 3 were expressed in the myometrium at term of parturition, and PGE₂ receptor 3 induced contractility in the human pregnant myometrium (13). Alternatively, it is known that PGE₂ plays an essential role in cervical ripening during term. Downregulation of 15-prostaglandin dehydrogenase, which inactivates PGE₂, is critical for cervix ripening at term (35). COX-2 expression increased after induction of oxytocin receptor expression in the uterus during parturition (28) and COX-2 expression was not induced in the uterus of Oxtr^−/− during parturition in this study. These results suggest that induction of COX-2 expression in the uterus at term is not essential to induce contractility for parturition onset. Further investigation is needed to find contractile factors responsible for the onset of parturition. Cervical ripening is important for normal onset and smooth progress of parturition. Impairment of cervical ripening caused failure of parturition onset (22). Alternatively, early cervical ripening caused preterm birth (35). In the current study, Oxtr^−/−;Ptgfr^−/− administered RU486 showed cervical histology similar to that in WT, Oxtr^−/−, and RU486-injected Ptgfr^−/− at term, suggesting that oxytocin/oxytocin receptor signaling and PGF_2α receptor signaling are not likely to contribute to cervical ripening.

Oxtr^−/−;Ptgfr^−/− and Oxt^−/−;Ptgfr^−/− administered RU486 showed greatly prolonged parturition and their pups remained in the uterus. Oxtr^−/−;Ptgf^-+/− showed prolonged parturition as did Oxtr^−/−;Ptgfr^−/− and Oxt^−/−;Ptgfr^−/−, although the percentage of Oxtr^−/−;Ptgf^-+/− showing prolonged parturition (22.2%) was lower than the percentages of Oxtr^−/−;Ptgfr^−/− and Oxt^−/−;Ptgfr^−/− (71.4% and 75.0%, respectively). These findings suggested that both oxytocin/oxytocin receptor signaling and PGF_2α receptor signaling play an important role in the progress of parturition. Sensitivity of uterine contraction to vasopressin increased at late pregnancy (7). Vasopressin induced uterine contraction not via the vasopressin receptor but via the oxytocin receptor in the mouse uterus in late pregnancy (3, 6). In the current study, all of the phenotypes in Oxt^−/−;Ptgfr^−/− administered RU486 were similar to those of Oxtr^−/−;Ptgfr^−/− administered RU486. These findings suggested that vasopressin did not compensate for oxytocin function via the oxytocin receptor and did not have a major contribution to the progress of parturition in mice.

In humans, maternal serum progesterone concentrations do not vary significantly in the late third trimester of pregnancy, and there is no evidence of a fall in maternal plasma or uterine tissue progesterone at labor, unlike those of rodents (36). However, administration of RU486 can induce human labor at term (37). Additionally, some reports suggest the importance of “functional progesterone withdrawal” in initiation of human labor. Progesterone receptors are expressed as two protein isoforms. Progesterone receptor-A and progesterone receptor-B are structurally identical with the exception of N-terminal extension that is specific to progesterone receptor-B. This extended region possesses a transactivation activity that contributes to distinct cell- and promoter-specific transactivation properties of the two isoforms (38). At term in human labor, the progesterone receptor-A/progesterone receptor-B ratio increases due to increased progesterone receptor-A expression in the myometrium (39). Progesterone receptor-A represses the transcriptional activity of progesterone receptor-B in human myometrial cells (40). Additionally, the expression of progesterone receptor coactivators is decreased in the myometrium at term (41). In rodents, Ptgfr^−/− studies clearly indicated that a decline of maternal blood progesterone concentration via luteolysis in late pregnancy is essential for initiation of labor (9, 11). However, there is no genetic evidence that a decline of progesterone receptor activity is necessary for initiation of labor, because progesterone receptor-A–deficient female mice were infertile and progesterone receptor-B–deficient female mice had sustained pregnancy and gave birth to viable offspring (38, 42). RU486 binds to both progesterone receptors and the glucocorticoid receptor as an antagonist. There is no evidence that inhibition of the glucocorticoid receptor induces labor in humans (37, 43). In rodents, uterine-specific glucocorticoid receptor–deficient female mice had sustained pregnancy and were able to deliver their pups (44). All of these findings suggest that a decrease of progesterone receptor activity is a common event in humans and rodents for initiation of labor. A schematic representation of the hierarchy for uterotonics in the progress of parturition is shown in Fig. 5. Our results indicated that the expression of oxytocin receptor, PGF_2α receptor, COX-2, and connexin 43 is induced during parturition in mice. In humans, oxytocin and PGF_2α have been used to induce or augment labor in clinical practice (1, 8). The expression of oxytocin receptor, PGF_2α receptor, COX-2, and connexin 43 was upregulated at term in labor in the human myometrium (17, 45–47). Additionally, oxytocin upregulated the expression of COX-2 in human myometrial cells in vitro (30). These findings suggest that the hierarchy for uterotonics obtained from the current study in mice has similarity with the mechanism in humans. Our study can help to further understand the complex mechanisms of human parturition.

Preterm birth is the main cause of neonatal morbidity and mortality: it causes 25% to 50% of long-term neurologic impairment in children and ∼70% of neonatal deaths and 36% of infant deaths (48). It is recognized that an increase in uterine contractions and shortening and dilatation of the cervix are features of active parturition, although the processes of human parturition are not fully understood. Recent reports suggest that some tocolytic agents might be effective in women with preterm labor. A promising nonpeptide oxytocin receptor antagonist, GSK221149A (retosiban), is currently being used in a phase 3 clinical trial for treatment of spontaneous preterm labor (49). Selective PGF_2α receptor antagonists OBE022 and OBE002 are currently being used in a phase 2 clinical trial for pregnant women with preterm labor (50). Administration of retosiban to women with spontaneous preterm labor was associated with an increase of 8.2 days in time to delivery compared with that in women administered a placebo and significant reduction in preterm births (51). Perinatal survival rate for preterm infants dramatically increased week by week between 22 weeks and 31 weeks of gestation, indicating that a longer period of uterine quiescence with tocolytic reagents can contribute to a decrease in the risk (52). We showed that the oxytocin receptor and PGF_2α receptor have a complementary relationship in the process of parturition. Our results suggest that administration of an oxytocin receptor antagonist in combination with a PGF_2α receptor antagonist would be more effective for spontaneous preterm labor.

In conclusion, the oxytocin receptor is an upstream regulator of COX-2 and connexin 43 expression in the uterus during parturition, and both oxytocin/oxytocin receptor signaling and PGF_2α receptor signaling are major components for successful parturition. Our results thus indicate the transcriptional and functional hierarchy of uterotonics required for successful parturition.