Endocrinology of Parturition1

To insure optimal survival of the neonate, the timing of parturition must be tightly controlled. Parturition before term dramatically increases the risk of the birth of a nonviable premature infant. At a time after the baby is mature, the intrauterine environment becomes hostile. Postdate fetuses have an increased incidence of intrauterine demise. To maximize infant survival, mechanisms have evolved to allow for uterine distention and reduced contractions during pregnancy and the timely initiation of labor. In women this process is only partially understood. The control of the timing of labor is complex and involves interactions of the mother, the fetus, and the placenta plus membranes. In women, invasive study of the intrauterine environment is limited for ethical reasons. However, the control of pregnancy and parturition is highly species specific. Although much more is known about the process of parturition in rodents and sheep than in humans, there are no good nonprimate models of human parturition. Even subhuman primates have significantly different endocrine environments and controls than humans. There is a complex schema of the initiation of labor in women, and there is no simple chain of events as there are in other species. For instance, there cannot be shown an acute rise in circulating estrogen, nor an abrupt decline in circulating progesterone levels to signal the onset of labor in all pregnancies. Evidence suggests that there are multiple paracrine/autocrine events, fetal hormonal changes, and overlapping maternal/fetal control mechanisms for the triggering of parturition in women. For parturition to occur, two changes must take place in a woman’s reproductive tract. First, the uterus must be converted from a quiescent structure with dyssynchronous contractions to an active coordinately contracting organ with complex interlaced muscular components. This requires the formation of gap junctions between myometrial cells to allow for transmission of the contractile signal. The second change is that the cervical connective tissue and smooth muscle must be capable of dilatation to allow the passage of the fetus from the uterus. In pregnancy there is a dynamic balance between the forces that cause uterine quiescence and the forces that produce coordinated uterine contractility. There is also a balance between the forces that keep the cervix closed to prevent uterine emptying and the forces that soften the cervix and allow it to dilate. For delivery to occur, both balances must be tipped in favor of active uterine emptying. Whereas there may be a final common pathway for the initiation of labor, which involves alterations in prostaglandin and calcium metabolism, there are multiple, sometimes complementary, initiating factors involved in the onset of labor (1). These are endocrine, paracrine, and autocrine. The “final common pathway” to delivery is likely to be multiple, parallel, interactive paths that tip the balance in favor of coordinated uterine contractility and cervical dilation. These mechanisms involve a shift from progesterone to estrogen dominance, increased sensitivity to oxytocin, gap junction formation, and increased prostaglandin activity. Decreased nitric oxide (NO) activity and increased influx of calcium into myocytes are both required for uterine contractibility (2, 3). Complementary changes in the cervix involving a decrease in progesterone dominance and the actions of prostaglandins and relaxin, via connective tissue alterations, collagenolysis, and a decrease in collagen stabilization through metalloproteinase inhibitors, leading to cervical softening and dilation (4). It is clear that the above mentioned pathways are not all inclusive. Other factors, such as endothelin, are involved in uterine changes conducive to increased blood flow and myometrial activity (5). It is also clear that with many overlapping mechanisms, decrease or absence of a single component can be compensated by changes in other paths. By way of analogy, in mice with specific gene knockouts, CRH and oxytocin are not necessary for normal delivery. Although knockouts of cyclooxygenase (COX)-1, COX-2, phospholipase A2, and relaxin do alter the timing of labor, they do not preclude uterus emptying and cervical dilation in all animals (6). Thus, complementary actions for the system must come into play. It is also quite likely that many of the causes of preterm parturition differ from the initiators at term. In fact, there seem to be multiple causes of preterm parturition. For example, infection may overwhelm one or more control mechanisms of normal labor. Or, more specifically, a deficiency of a choriodecidual enzyme (such as 15-hydroxy-PG-dehydrogenase) may alter the metabolism of prostaglandin E2 (PGE2) resulting in excess and, ultimately, onset of labor (7, 8). On the other hand, in a novel and as yet undetermined way, prematurity in patients whose ovulation induction included human menopausal gonadotropins is associated with hyperrelaxinemia (9). Many initiating mechanisms may be involved in prematurity due to other causes, such as premature membrane rupture or the prematurity associated with uterine anomalies. Thus, whereas the onset of term labor seems to be a fairly predictable chain of events, preterm labor is as if the normal chain of events was entered at a site dependent on the etiology. Estrogen is essential for uterine development and function. Estrogens are responsible for the synthesis of the contractile proteins and the regulatory enzymes necessary for uterine contractility. In prepubescent girls, the uterus is small. It is only after estrogen is secreted that the uterus increases in size and develops the ability to respond to stimulants and inhibitors of contractions. Estrogen increases the concentration of receptors for oxytocin andα -adrenergic agents, which modulate membrane calcium channels (10). Estrogens are critical for intracellular communication. Estrogens increase connexin 43 synthesis and gap junction formation in the myometrium (11). This allows for coordinated uterine contractions. Estrogen also stimulates the production of prostaglandins F2α and E2, which stimulate uterine contractions (12). In women, although estradiol continues to remain high, it does not have a sharp predelivery increase as in sheep, where it is responsible for the onset of labor. Estrogens control cervical ripening. This may be associated with the down-regulation of the estrogen receptor. The control of the softening of the cervix, which involves rearrangement and realignment of collagen, elastin, and glycosaminoglycans such as decorin, is not well studied and is poorly understood (4). The metabolism of estrogens during pregnancy in humans and in other higher primates differs from that of all other species. The human placenta lacks significant amounts of 17-hydroxylase/17–20 lyase. This enzyme is needed in the synthetic pathway of estradiol from progesterone. Progesterone is synthesized from acetate and cholesterol in the placenta. Human pregnancy estrogen production is complex. The fetal zone of the adrenal gland produces dehydroepiandrosterone sulfate (DHEAS), which may be hydroxylated to 16-OH-DHEAS in the fetal liver. The 16-OH-DHEAS may be aromatized by the placenta to produce estriol, the major circulating estrogen of human pregnancy. In contrast to the nonpregnant state, during late human pregnancy the ovary is a minor source of circulating estrogens. Estradiol and estrone are synthesized by placental aromatization of DHEAS from both maternal and fetal sources; however, more than 90% of estriol is derived from fetal 16-OH DHA (13). Estriol concentrations in serum and saliva increase during the last 4–6 weeks of pregnancy. Throughout the last two fifths of pregnancy, levels of salivary estriol in women destined to have preterm deliveries are higher than in control women having term deliveries. There seems to be a 4-week advancement in the higher levels in women who will deliver preterm. Salivary estrogen has been suggested as a screen for the potential of preterm labor risk (14). During late pregnancy in women the placenta is the major source of circulating progesterone; the ovarian contribution is slight. In species in which pregnancy is luteal dependent, such as the rodent, progesterone withdrawal initiates the onset of labor. In women there is no confirmed prelabor fall of circulating progesterone. Serum progesterone levels do not vary significantly between women in labor and those not in labor. In early pregnancy, removal of the corpus luteum, the major source of progesterone at that stage of pregnancy, results in pregnancy loss (15). Progesterone receptor blockers such as RU486 result in the initiation of labor (16). This may be because RU486 stimulates CRH messenger RNA, suggesting that a progesterone decrease will result in an increased CRH effect (17). However, the actual amount of circulating progesterone throughout pregnancy is in excess of the concentration needed for uterine inhibition. Women with aβ -lipoproteinemia, who have circulating progesterone levels of less than 10 ng/mL throughout pregnancy, maintain their pregnancies normally and deliver normally at term (18). In pregnancy, progesterone is in dynamic balance with estrogen in the control of uterine activity. Progesterone in vitro decreases myometrial contractility and inhibits myometrial gap junction formation (2). Progesterone activity stimulates the uterine NO synthetase, which is a major factor in uterine quiescence. Progesterone down-regulates prostaglandin production, as well as the development of calcium channels and oxytocin receptors both involved in myometrial contraction (2). Calcium is necessary for the activation of smooth muscle contraction. In the cervix, progesterone increases tissue inhibitor of matrix metalloproteinase 1 (TIMP-1) (19). TIMP-1 inhibits collagenolysis. Thus, it is clear that progesterone is a major factor in uterine quiescence and cervical integrity. The factors that result in parturition must overcome the progesterone effect that predominates during the early pregnancy period of uterine quiescence. The activity of 17,20 hydroxysteroid dehydrogenase in fetal membranes increases around the time of parturition, leading to an increase in net 17β-estradiol and 20-dihydroprogesterone (20). This is a factor in altering the estrogen/progesterone balance. There may be decreased progesterone receptor levels at term resulting in a diminished progesterone effect. In summary, estrogen and progesterone activities are critical determinants of the balance between uterine quiescence and the factors that produce labor. Labor stimulation or inhibition will generally be produced by agents that alter this critical balance. Available evidence suggests that oxytocin performs an important, but not critical, role in the initiation of labor. Patients will deliver after hypophysectomy. Circulating oxytocin does not increase in late pregnancy. It does not increase in labor until after full cervical dilatation (21). However, the concentration of uterine oxytocin receptors increases toward the end of pregnancy (22). This results in increased efficiency of oxytocin action as pregnancy progresses. Estrogen increases oxytocin receptor expression and progesterone suppresses such estrogen-induced increase in cultured human myometrial cells (23). Oxytocin induces uterine contractions in two ways. Oxytocin stimulates the release of PGE2 and prostaglandin F2α in fetal membranes by activation of phospholipase C. The prostaglandins stimulate uterine contractility (24). Oxytocin can also directly induce myometrial contractions through PLC, which in turn activates calcium channels and the release of calcium from intracellular stores (25, 26). Oxytocin is locally produced in the uterus (27). The role of this local endogenous oxytocin is unknown. Nor is the direct effect of oxytocin on cervical dilatation well understood. Oxytocin infusion is used clinically to induce uterine contractions and labor. Oxytocin may accelerate cervical ripening at term, but it does not effectively or efficiently ripen an unripe cervix taking a long time at a low dose. Oxytocin is also less effective in causing uterine contractions in midpregnancy than at term. There is good evidence that prostaglandins are involved in the final pathway of uterine contractility and parturition. Prostacyclins, inhibitory prostaglandins present throughout early pregnancy, are also responsible for uterine quiescence during pregnancy. Arachidonic acid is the obligatory precursor of PGE2 and F2α, the major stimulatory prostaglandins (28). Although prostaglandins may not be obligatory for labor, as has been shown in knockout mice, they are of major importance in women (6). Prostaglandins are produced in the placenta and fetal membranes. The membranes, consisting of amnion and chorion form the amniotic sac. On the maternal side, the chorion is adherent to the decidua. Prostaglandin levels are increased before and during labor in the uterus and membranes (29, 30). Many factors affect the production of prostaglandins. Levels are decreased by progesterone and increased by estrogens (31–34). Several interleukins result in an increase in prostaglandin production (35). This may be the mechanism by which inflammatory cytokines result in premature labor. A leukocyte influx at the time of infection releases inflammatory cytokines that increase production of stimulatory prostaglandins. A significant proportion of cases of prematurity are related to intrauterine or cervical infection. CRH also increases prostaglandin production (36, 37). An increase in the circulating concentration of prostaglandin metabolites is found at the onset of labor (28). Prostaglandin is formed from arachidonic acid that is converted to prostaglandin H2 by the enzyme prostaglandin H synthetase (PGHS). PGHS-2 is an inducible form of the enzyme. PGHS-2 and COX-2 are the same enzyme but both are referred to in different papers and so is included here under both names. Cytokines increase the concentration of this enzyme 80-fold. Prostaglandins are degraded by 15- hydroxy-prostaglandin dehydrogenase. COX-2, the cyclooxygenase isoform that is cytokine inducible, is increased by NO. This is another mechanism by which prostaglandin production increases during inflammation. Inflammation-induced increase in prostaglandins can result in stimulation of uterine activity and cervical ripening (28). Local application of PGE1 and PGE2 is used clinically to induce cervical ripening. Prostaglandin F2α increases total glycosaminoglycan activity. PGE2 dilates cervical small blood vessels. PGE2 leads to cervical ripening associated with collagen degradation (4). There is controversy as to whether the action of prostaglandin on the cervix is direct or indirect. The prostaglandin effects on myometrium are, at least in part, direct actions. CRH is the hypothalamic hormone responsible for control of ACTH. It, in turn, is modulated by a typical long loop feedback system involving adrenal cortisol. Thus, is the mother ACTH stimulates the adrenal that makes cortisol, which decreases the production of CRH (negative feedback), resulting in a decrease in ACTH secretion. In human pregnancy, from the onset of the second trimester, the placenta is a major source of CRH secretion (38). Paradoxically, cortisol increases CRH production by the placenta (positive feedback) (39). Positive and negative feedback may be controlled by different signal transduction mechanisms. CRH stimulates fetal ACTH release. The resultant fetal adrenal steroidogenesis results in increased production of DHEA and DHEAS. These compounds are placental precursors of estrogens, including estradiol. It is postulated that increased placental CRH results in an estrogen-dominant environment conducive to parturition (40). Stimulation of the fetal adrenal also increases glucocorticoid production. Glucocorticoids are responsible for fetal lung maturation. Thus, CRH synchronizes maturation of the fetal lung with forces that induce the onset of labor. CRH of hypothalamic origin is a stress-induced hormone. With stress, cortisol is secreted, which increases placenta CRH production. This may be a factor in premature birth (41). Actions of CRH include dilation of the uterine vessels and stimulation of smooth muscle contractions, dilation of the fetal placental vessels via NO synthetase activation; and stimulation of prostaglandins F2α and E2 production by fetal membranes and decidua (42, 36, 37). Prostanoids increase CRH production by decidua and membranes (43). These are all actions conducive to the initiation of labor. CRH is also stimulated by inflammatory cytokines (43). However, the administration of cortisol does not induce labor in women. CRH rises in maternal serum starting at approximately the 16th week of gestation (44). Some data demonstrate that CRH increases at a more acute rate in the last 6–8 weeks of pregnancy. Before delivery, CRH binding protein decreases, resulting in more effective unbound CRH in maternal circulation (45). Women destined to have premature delivery have higher midpregnancy CRH levels than those who deliver at term (46). This higher level of CRH may be used as a marker for women at risk for prematurity. This elevated CRH may accelerate the timing of the process of parturition (43). Relaxin is a peptide hormone that is a member of the insulin family. Relaxin consists of A and B peptide chains linked together by two disulfide bonds. In women, circulating relaxin is a product of the corpus luteum of pregnancy, which is present in the ovary for the duration of pregnancy. Circulating relaxin is secreted in a pattern similar to that of human CG. That circulating relaxin is not critical for pregnancy maintenance is demonstrated by pregnancies generated by egg donation, which have no maternal circulating relaxin and are able to go to term and deliver spontaneously. However, relaxin is also a product of the placenta and decidua. Relaxin from these sources, which may act locally, is not secreted into the peripheral circulation (47). Premature birth is associated with increased circulating relaxin levels (48). Women who have superovulation with human menopausal gonadotrophins for either ovulation induction or in vitro fertilization have a significantly higher risk of premature birth. These women, who have multiple corpora lutea, have significant levels of hyperrelaxinemia. Logistic regression analysis has demonstrated that the extent of hyperrelaxinemia in these women is associated with increased levels of prematurity (9). Additional support for a role of relaxin in prematurity is that in spontaneous pregnancies women destined to have premature delivery have higher levels of relaxin at 30 weeks gestation than women who deliver at term (48). A potential mechanism for this relationship may be the action of relaxin on the cervix. Relaxin has been associated with cervical softening. Relaxin receptors are present on the human cervix (49). Some of the effects of relaxin include stimulation of procollagenase and prostromelysin, as well as a decrease in TIMP-1 (50). Relaxin is also capable of inhibiting contractions of nonpregnant human myometrial strips (51). Paradoxically, relaxin does not inhibit contractions of pregnant human uterine tissue (52). This may be because of the competitive effects of progesterone. It is vividly clear that a variety of endocrine systems play a role in the maintenance of uterine quiescence and the onset of parturition, with its attendant increase in uterine contractility and cervical ripening (Fig. 1). However, there is little doubt regarding the primacy of the balance between estrogen and progesterone. There are many factors that can tip the balance in favor of delivery early, late, or on time. These factors, such as prostaglandins or inflammatory cytokines, may directly affect the contractile mechanisms (53, 54). Other factors, such as oxytocin, CRH, or relaxin, may indirectly alter the actions of complementary systems. Physiologically, there may be many parallel operative systems that can compensate for anomalies or imbalance in other systems. However, our information is incomplete at present. We have only fragmentary data regarding the interaction of these systems and how the organism compensates for imbalances. For instance, exogenous estrogen cannot induce labor and exogenous progesterone cannot stop prematurity, yet both steroids are critical for the control of uterine activities. Simplified scheme of the endocrinological control of pregnancy and parturition in women. The balance between the effects of estrogen and progesterone is critical to maintenance of pregnancy and the onset of labor. Other important hormonal factors modulate this balance as shown in the scheme. Not all factors are represented. Those shown are endocrine factors demonstrated to be significant. Ignorance of the control mechanisms carries significant costs. There are as yet no clinically effective and useful agents to delay the onset of preterm labor. Available methods of labor induction, while clearly effective, are neither completely dependable nor safe. The cost of our inability to prevent prematurity is heartbreaking pregnancy loss or prolonged infant occupancy of neonatal intensive care units. It is only with increased understanding of the processes of parturition that we will be able to further improve the safety of the birth process.