|
|
||||||||
a Department of Obstetrics and Gynecology, Feinberg School of Medicine, Northwestern University, Evanston, Illinois 60201
b Department of Obstetrics and Gynecology, Columbia University College of Physicians & Surgeons, New York, New York 10032
| ABSTRACT |
|---|
|
|
|---|
parturition, pregnancy, progesterone
| INTRODUCTION |
|---|
|
|
|---|
In mice, normal parturition at term is believed to result from luteolysis and subsequent progesterone (P4) withdrawal. P4 levels fall precipitously before delivery, paralleling the weight of corpora lutea [3]. Ovariectomy results in pregnancy termination within approximately 24 h in mice, and P4 replacement can prevent ovariectomy-induced delivery or prolong normal pregnancy beyond term [4].
In the setting of infection in the preterm period, it is unclear to what extent the signals for labor resemble those for spontaneous parturition at term. In both humans and mice, increased expression of prostaglandins and inflammatory cytokines (e.g., interleukin [IL]-1, IL-6, and tumor necrosis factor
) occurs during infection-induced preterm labor [512]. However, whether these factors participate in the signaling process by which bacterial organisms cause labor is not known [13]. A recent study in a mouse model demonstrated a rapid fall in serum P4 of 46% within 1 h, 68% within 4 h, and 87% within 10 h of i.p. administration of endotoxin (lipopolysaccharide) [14]. These data raise the possibility that endotoxin exposure induces labor in mice secondarily, via P4 withdrawal. We tested this hypothesis in an existing murine model by using an intrauterine rather than a systemic stimulus to model infection-induced labor [1012].
| MATERIALS AND METHODS |
|---|
|
|
|---|
All procedures involving animals were approved by the Columbia University and Evanston Northwestern Healthcare Institutional Animal Care and Use Committees and conformed to the Guide for Care and Use of Laboratory Animals (National Academy of Sciences, 1996). Pregnant CD-1 mice underwent timed matings and were anesthetized on Day 14.5 of a 19- to 20-day gestation, as previously described [1012].
Study Procedures
For studies to determine the rate of decline in serum P4 levels in response to a variety of interventions, mice were blindly allocated to receive one of the following four treatments: 1) intrauterine injection of sterile medium (previously shown to result in delivery at term) [1012]; 2) intrauterine injection of 106 heat-killed Escherichia coli bacteria (previously shown to result in delivery at term) [12, 13]; 3) intrauterine injection of 109 heat-killed E. coli (previously shown to result in preterm delivery in an average of 20 h) [12, 13]; or 4) ovariectomy (known to cause delivery in approximately 24 h).
All procedures were performed via midline laparotomy in the lower abdomen. In mice, the uterus is a bicornuate structure in which the fetuses are arranged in a pattern that resembles beads on a string. For intrauterine injections, the right uterine horn was identified and injected at a point between two adjacent fetuses, taking care not to inject individual fetal sacs. For ovariectomies, the vascular supply to each ovary was clamped with a hemostat for 1 min and the ovary was excised with a scalpel. The abdomen was closed in two layers with 4-0 polyglactin suture at the peritoneum and wound clips at the skin. Animals were allowed to recover individually in clean cages and were killed for serum collection at fixed times after surgery (0.75, 3.5, 5.5, or 8 h following intrauterine inoculation). Ovariectomized animals were harvested in a proportionally delayed schedule to account for the 4-h difference in average time to delivery compared with intrauterine high-dose injection of bacteria (i.e., 1, 5, 7.5, and 11 h after surgery).
To test the effect of exogenous steroid administration, a separate group of pregnant mice were pretreated with varying quantities of P4 diluted in 100 µl of sterile corn oil (01.5 mg s.c. per mouse; Sigma-Aldrich, St. Louis, MO). Two hours later, they underwent anesthesia and either ovariectomy or intrauterine inoculation with 109 killed E. coli. These animals were then either monitored to record time to delivery or they were killed at 8, 16, or 24 h after surgery for serum and tissue collection.
Serum P4 and estradiol (E2) levels were determined with the Immulite Immunoassay system (Diagnostic Products, Los Angeles, CA). Averages are reported as means ± SD. Statistical significance was determined using a t-test. P < 0.05 was considered statistically significant.
| RESULTS |
|---|
|
|
|---|
|
|
In order to test the ability of exogenous P4 to modulate labor induced by killed bacteria, mice were pretreated with 1.5 mg of P4 in oil s.c. or with oil alone 2 h before induction of anesthesia. They were then subjected to either ovariectomy or intrauterine inoculation with 109 killed E. coli (Fig. 3). Compared to placebo treatment, 1.5 mg of s.c. P4 significantly prolonged the average interval to delivery following both ovariectomy (from 24.4 ± 5.6 h to 44.8 ± 7.7 h; P < 0.001) and intrauterine bacterial inoculation (from 18.6 ± 3.6 h to 35.0 ± 17.3 h; P < 0.01). However, such treatment resulted in pharmacologically elevated P4 levels greater than 3-fold above physiological values as late as 16 h after surgery (Fig. 4).
|
|
In order to determine the effects of more physiologic P4 supplementation, we characterized the dose-response relationship between preoperative P4 (01.5 mg/mouse) and pregnancy prolongation after ovariectomy (Fig. 5). The lowest tested effective dose of P4 was 0.375 mg/mouse, which resulted in an average pregnancy prolongation of 14.6 h (P = 0.01). We next injected pregnant mice either with this dose of P4 or with vehicle 2 h before intrauterine inoculation with 109 E. coli. There was no difference in the interval to delivery between animals treated with this level of P4 and concurrent controls (Fig. 6; note that the average time to deliveryapproximately 21 his slightly longer than that previously determined after intrauterine bacterial inoculation19 hdue to the need to prepare a fresh lot of killed E. coli for this and the following series of experiments).
|
|
Finally, in order to determine whether this lower dose of s.c. P4 (0.375 mg/mouse) resulted in physiologic rather than pharmacologic replacement, serum P4 was measured 24 h after surgery in mice pretreated with either P4 or vehicle and subjected 2 h later to either ovariectomy or intrauterine bacterial inoculation (Fig. 7). The only statistically significant difference among the above regimens was between the ovariectomy group treated with oil and the bacteria-inoculated group treated with P4, with significantly higher serum P4 levels in the latter, presumably due to the contribution of continued endogenous production.
|
| DISCUSSION |
|---|
|
|
|---|
These observations are extremely important for interpreting the results of rodent models of infection-induced labor. We and others have demonstrated in the past that such models are associated with biochemical changes that mimic the condition in women (e.g., up-regulation of IL-1, IL-6, cyclooxygenase, and prostaglandins [10, 12, 1519]). Further support for the murine model is provided in the present paper, which demonstrates that induction of labor does not result from a process that is apparently irrelevant to human infection-induced parturition; namely, P4 withdrawal.
We have also shown that although supraphysiologic P4 treatment can result in delayed delivery following high-dose intrauterine bacterial administration, this effect is not present with a lower dose of P4. This observation reflects the well-known smooth muscle relaxant effects of P4, and serves to underscore the need for caution in attributing to physiologic roles of P4 what may in fact be pharmacologic phenomena.
Our model differs from that of Fidel et al. [14] in that we administered killed bacteria to the uterus, as opposed to lipopolysaccharide to the peritoneum. It can be assumed that at least a portion of the toxicity of killed E. coli is caused by the lipopolysaccharide present in their cell walls. However, systemic lipopolysaccharide administration results in a profound inflammatory insult akin to septic shock. Our use of intrauterine bacterial injections produces a more localized process with minimal to no systemic symptoms. We observed a drop in serum P4 after ovariectomy (60% at 1 h and 81% at 7.5 h) that was similar to the one reported in the Fidel study after i.p. injection of lipopolysaccharide (46% at 1 h and 87% at 10 h) [14]. Preterm delivery occurred in both cases. However, ovariectomy produces labor within approximately 24 h (a relatively constant interval in the literature), compared with 1112 h after endotoxin injection in the Fidel study. Although the time intervals from separate studies cannot be compared directly, the difference lends support to the conclusion that endotoxin exposure activates different pathways to labor than does P4 withdrawal alone.
As proposed by Skarnes and Harper [4], the abortifacient effect of endotoxin might be due to sensitization of the uterus by inflammatory byproducts such as prostaglandin F2
, followed by increasing uterine contractile activity as P4 levels fall. Thus, it remains possible that the level of circulating P4 is an important component of the process by which infection causes labor in mice. In this regard, it is notable that serum P4 falls to very low levels by 24 h after intrauterine bacterial exposure. Whether this is a direct result of the inflammatory process (e.g., prostaglandin-induced luteolysis [20]) or a secondary outcome related to a failing pregnancy is not known.
Although the signals that lead to term parturition in humans are not understood, they do not seem to depend on maintenance of circulating P4 levels, which do not begin to decline until after delivery. Nonetheless, it has been suggested that other changes affecting P4 bioavailability or activity may account for initiation of parturition in humans, including the E2:P4 ratio, concentrations of soluble receptors, binding of the P4 receptor to its nuclear response element, and other factors [21, 22]. Lending support for a critical role for P4 in the maintenance of human pregnancy are the observations that administration of the P4 antagonist mifepristone (RU-486) initiates labor in women and reduces the requirement for oxytocin during induction of labor [23]. In guinea pigs, a species that, like humans, does not experience a decline in serum P4 before the onset of labor, P4 antagonism with onapristone enhanced sensitivity to oxytocin by a factor of up to 30-fold, decreased the electrical input resistance of myometrium, and increased myometrial gap junctions [24]. A similar enhancement of oxytocin effect has been observed in nonhuman primates [25]. In human placental and chorionic trophoblast, P4 modulates the activity of prostaglandin dehydrogenase (PGDH), a key enzyme that metabolizes prostaglandins to inactive products [26]. PGDH activity is high during gestation and decreases in labor, which supports a "removal of inhibitor" hypothesis for the initiation of parturition. Thus, there may be more similarity between mice and humans in the control of normal parturition than was previously believed.
The degree of overlap in the mechanisms controlling term and infection-induced preterm parturition is not known either in humans or in mice, but the existing literature demonstrate remarkable similarity between the two species in the mechanisms of both normal and abnormal labor. The mouse affords several advantages when used as a pregnancy model. These include its relative low cost compared with that of using other species, its brief gestational period and high fecundity, its utility in gene expression modulation experiments, and the availability of the mouse genome sequence and a large collection of expressed sequence tag clones. Therefore, we believe that continued use of murine models holds great promise for studies of the physiology and pathophysiology of parturition.
| FOOTNOTES |
|---|
2 Correspondence: Emmet Hirsch, Department of Obstetrics and Gynecology, Evanston Northwestern Healthcare, 2650 Ridge Ave., Evanston, IL 60201. FAX: 847 773 5083; e-hirsch{at}northwestern.edu ![]()
Received: 27 September 2001.
First decision: 24 October 2001.
Accepted: 29 May 2002.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. Bizargity, R. Del Rio, M. Phillippe, C. Teuscher, and E. A. Bonney Resistance to Lipopolysaccharide-Induced Preterm Delivery Mediated by Regulatory T Cell Function in Mice Biol Reprod, May 1, 2009; 80(5): 874 - 881. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Schmitz, E. Souil, R. Herve, C. Nicco, F. Batteux, G. Germain, D. Cabrol, D. Evain-Brion, M.-J. Leroy, and C. Mehats PDE4 Inhibition Prevents Preterm Delivery Induced by an Intrauterine Inflammation J. Immunol., January 15, 2007; 178(2): 1115 - 1121. [Abstract] [Full Text] [PDF] |
||||
![]() |
X. Wang, H. Hagberg, C. Mallard, C. Zhu, M. Hedtjarn, C.-F. Tiger, K. Eriksson, A. Rosen, and B. Jacobsson Disruption of Interleukin-18, but Not Interleukin-1, Increases Vulnerability to Preterm Delivery and Fetal Mortality after Intrauterine Inflammation Am. J. Pathol., September 1, 2006; 169(3): 967 - 976. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Anderson, N. Brown, M. S. Mahendroo, and J. Reese Utilization of Different Aquaporin Water Channels in the Mouse Cervix during Pregnancy and Parturition and in Models of Preterm and Delayed Cervical Ripening Endocrinology, January 1, 2006; 147(1): 130 - 140. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Hirsch and H. Wang The Molecular Pathophysiology of Bacterially Induced Preterm Labor: Insights From the Murine Model Reproductive Sciences, April 1, 2005; 12(3): 145 - 155. [Abstract] [PDF] |
||||
![]() |
K. Yoshimura and E. Hirsch Interleukin-6 Is Neither Necessary Nor Sufficient for Preterm Labor in a Murine Infection Model Reproductive Sciences, October 1, 2003; 10(7): 423 - 427. [Abstract] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |