| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Obstetrics and Gynecology3
Molecular Genetics,4 University of Texas Southwestern Medical Center, Dallas, Texas 75390
Department of Medicine,5 University of California-Los Angeles, Los Angeles, California 90095
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
cervix, female reproductive tract, parturition, pregnancy, uterus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although endocrine events that initiate parturition differ from species to species, pregnancy in most mammals is maintained by the hormone progesterone. Parturition is initiated by a sharp reduction in the rate of progesterone formation and an increase in progesterone catabolism [2]. In the mouse, the decline in formation and increase in the breakdown of progesterone by the ovary (a process termed luteolysis) is initiated by uterine-derived prostaglandin F2
on Day 17 of a 19-day gestational period [3]. Animals with null mutations in the genes encoding phospholipase A2 (Pla2g4a) [4], the prostaglandin F2 receptor (Ptgfr) [5], cyclooxygenase (COX)-1 (Ptgs1) [6], or 20-hydroxysteroid dehydrogenase (Akr1c18) [7] fail to undergo normal parturition at term because of failed luteolysis. In these parturition-defective animal models, ovariectomy or antiprogestin treatment rescues the parturition defect. In steroid 5-reductase type 1 (SRD5A1)-deficient mice, luteolysis occurs, but parturition fails, primarily because cervical progesterone concentrations remain elevated as a result of attenuated catabolism of progesterone by SRD5A1. Ovariectomy, antiprogestin treatment, or relaxin rescues the parturition defect in SRD5A1-null mice [8, 9].
In women, serum progesterone levels do not decline before labor and delivery. The notion that a decline in progesterone function may occur at term in human pregnancy is supported by the observation that physiologic changes taking place during the preparatory and active phases of human parturition are indistinguishable from those in other species in which uterine and cervical changes seem to be induced by physiologic progesterone withdrawal. Some evidence suggests that functional withdrawal of progesterone action may occur before parturition in women via changes in the expression of progesterone-receptor coactivators [10] or differential expression of progesterone-receptor isoforms [11, 12]. In contrast, other data indicate that although progesterone-receptor antagonists act as abortifacients during the first trimester, concomitant administration of uterine contractile agents (e.g., oxytocin or prostaglandin analogs) are required thereafter for successful termination of pregnancy and delivery of the fetus [13, 14]. A significant number of women do not deliver in response to antiprogestins; several days are required before clinical effectiveness is observed [15, 16]. Labor is prolonged in women with poor cervical ripening, often leading to cesarean section for failure to progress in labor. The molecular mechanisms resulting in uterine contractions of labor in the absence of timely cervical ripening at term are unknown. Furthermore, significant changes in progesterone-responsive genes in myometrial tissues from women before and after the onset of labor have not been detected [17, 18]. Thus, it appears that both progesterone-independent and progesterone-dependent events mediate the initiation of parturition in women.
In the present study, we describe a mouse model in which the insertion of a transgene on mouse chromosome 6 results in failure to complete parturition even after withdrawal of progesterone. Transgene insertion in these mice (TgN(hApoB)1102SY) apparently has interrupted one or more genes that are crucial for normal parturition. Although the interrupted molecular pathways have not been identified, herein we used histologic, biomechanical, and functional assays to determine the pathophysiology of failed parturition in these animals.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ZK98299 (onapristone) was a gift from Schering Corp. (Kenilworth, NJ). Oxytocin was obtained from Burns Butler Vet Supply (Farmers Branch, TX; catalog no. 3015866).
Mice
Animals were housed under a 12L:12D photoperiod (lights-on, 0600– 1800 h) at 22°C. All mice used in the present studies were of mixed strain (C57BL/6J x SJL or C57BL/6 x 129SvEv). The C57BL/6J x SJL and C57BL/6 x 129SvEv mice were generated and maintained as a breeder colony at the University of Texas Southwestern Medical Center (Dallas, TX). Homozygous (Tg/Tg) mice were propagated by matings between heterozygous females (Tg/Wt) and homozygous males (Tg/Tg). This mating produced 50% of offspring bearing two copies of the transgene. Genotypes were determined using genomic DNA in a multiplex polymerase chain reaction (PCR) assay. To obtain timed-pregnant animals, females housed overnight with males were checked at midday for vaginal plugs. Plug day was counted as Day 0, and birth occurred during the early morning hours (0200–0300 h) of Day 19 in Wt/Wt and Tg/Wt animals. In the case of the Tg/Tg mice, birth did not occur; thus, animals were still pregnant on Day 19. For all experiments in which Day-19 animals were used, tissue was collected between 0700 and 0900 h. Thus, Wt/Wt animals were postpartum, but Tg/Tg animals were not. All studies were conducted in accordance with the standards of humane animal care described in the NIH Guide for the Care and Use of Laboratory Animals using protocols approved by an institutional animal care and research advisory committee.
Serum and Tissue Hormone Measurements
Blood was drawn from the inferior vena cava of pregnant mothers between Days 17 and 19 of gestation. Blood samples were collected from six or seven wild-type (Wt/Wt) or Tg/Tg mice for each time point. Cells were removed by centrifugation and serum stored at –20°C until steroid analyses were performed.
Progesterone, androstenedione, 20-hydroxyprogesterone, and 17ß-estradiol were quantified in cervical tissues and serum by RIA after chromatographic separation of the steroids on Sephadex LH-20 columns (Pharmacia, Inc., Piscataway, NJ). Steroid measurements were performed at the Oregon Regional Primate Research Center (Beaverton, OR). For tissue measurements, cervices were dissected from seven or eight animals per group and stored at –80°C until analyzed. For extraction, the tissues were thawed, dissected into two roughly equal portions, weighed on a microbalance, and homogenized in 1 ml of PBS at 4°C using a Polytron (Brinkmann Instruments, Inc., Westbury, NY). Steroids from both halves of the cervix were extracted into ether from the homogenates and subsequently measured as described for serum hormone measurements. For each sample, an average of the steroid concentration was calculated based on measurements from each half.
The intra-assay coefficient of variation was 6.5%, 6.4%, and 7.4% for estradiol, progesterone, and androstenedione, respectively. The average blank values were 0.65, 6, and 1.4 pg for estradiol, progesterone, and androstenedione, respectively. Coefficient of variation was not determined for 20-hydroxyprogesterone, because a purified 20-hydroxyprogesterone standard was not available.
Quantification of mRNA
Total RNA was extracted from frozen tissue using RNA Stat60 (Tel-Test "B," Inc., Friendswood, TX). Individual cervices from three or four mice were used to isolate total RNA for each genotype and time point. Samples were subsequently treated with DNA-free (Ambion, Inc., Austin, TX). Reverse transcription was performed with a TaqMan cDNA synthesis kit, and real-time PCR was done using SYBR Green and a PRISM7900HT Sequence Detection System (Applied Biosystems, Foster City, CA). Aliquots (20 ng) of total cDNA were used for each PCR reaction conducted in triplicate. Expression of each gene was normalized to the expression of the housekeeping-gene cyclophilin and expressed relative to one control cervix that was used as an external calibrator using the ddCt method (Applied Biosystems, Foster City, CA). Values were then expressed according to the relative abundance of each transcript in the cervix (estimated using the ddCt method for each gene expressed in the external calibrator).
Hormonal Manipulation
ZK98299 (1 mg; onapristone; Schering Corp., Kenilworth, NJ) was administered by subcutaneous injection. The steroid was dissolved in ethanol at a concentration of 20 mg/ml. Fifty microliters were added to triolene (Sigma, St. Louis, MO) and injected in a final volume of 200 µl. ZK98299 was administered once on the morning of Day 17 in Wt/Wt females or Day 18 in the Tg/Tg females. In Wt/Wt animals, delivery occurred 18–24 h after injection.
Oxytocin was administered by subcutaneous injections of 250–500 mU dissolved in 100 µl of PBS on Day 18.5. Animals were then monitored for delivery, which generally occurred within 1 h after injection. Animals were given up to three doses of oxytocin at an interval of 30 min between each dose if no initial response was observed.
Steroid Hormone Metabolism
Cervices and uteri were dissected from Wt/Wt and Tg/Tg mice on Day 18. Tissues were homogenized in 10 mM potassium phosphate, 150 mM potassium chloride, 0.3 M sucrose, and 1 mM EDTA. Protein was determined using the bicinchonic acid protein assay (Pierce, Rockford, IL). Progesterone metabolism in the uterus and cervix was assessed by incubating tissue homogenates (150 µg of protein) in 0.1 M Tris-citrate buffer (pH 7.0) and containing 5 µM [14C]progesterone (NEN, Boston, MA) and 5 mM NADPH (Sigma) in a total volume of 0.5 ml for 1 h at 37°C. Steroids were extracted into 5 ml of methylene chloride and taken to dryness under a stream of nitrogen. Steroids were dissolved in 20 µl of chloroform-methanol (2:1, v/v), spotted onto Silica Gel 150 TLC plates (4855-821; Whatman, Clifton, NJ), and resolved by development in chloroform-ethyl acetate (3:1, v/v). Radiolabeled steroids were visualized by exposing the plates to Kodak XAR-5 film (Eastman Kodak Co., Rochester, NY) for 12–16 h. Steroid metabolites were identified based on migration of standards as described previously [9].
Uterine Contractility
Mice were anesthetized with avertin (0.02 ml/g i.p.). Thereafter, a 21-gauge catheter was inserted into the uterus through a small abdominal incision. The catheter was threaded between the uterine wall and fetal membranes and connected to a Grass Model PT300 pressure transducer (Grass-Telefactor, West Warwick, RI). Contractile activity was recorded on a Grass 7C polygraph for 30 min to 1 h, and the animal was placed on a warming plate to maintain temperature. The transducer was calibrated before and after each recording session.
Assessment of Biomechanical Properties of the Cervix
Tensile properties of isolated cervical tissues were evaluated using modifications of the method developed by Harkness and Harkness [19]. The excised cervix was mounted by two pins inserted through the cervical canal. One pin was attached to a calibrated mechanical drive, and the other pin was attached to a force transducer. Tissues were incubated in a water-jacketed bath containing PSS at 37°C bubbled with 95% O2/5% CO2. Baseline cervical dilatation was quantified by determining the difference in micrometer readings at zero (pins juxtaposed), and the initiation of tension was recorded by the physiograph. Thereafter, the inner diameter of the cervix was increased isometrically in 1-mm increments at 2-min intervals to affect cervical distention. The amount of force required to distend the cervix was recorded. The diameter was increased until the forces exerted by the tissue reached a plateau or the tissue tore. Force was plotted as a function of cervical diameter. The slope of the linear portion of the force-strain curve was computed as tissue stiffness. A steeper slope indicates increased resistance to stretch and, therefore, decreased distensibility/elasticity (Young modulus). Because cervical size was dramatically different between Wt/Wt and Tg/Tg cervices, normalization of forces exerted by the stretched cervix to cross-sectional area would exaggerate the differences between the two genotypes. Thus, tension was expressed in grams. Data obtained from Tg/Tg and Tg/Wt animals at various stages of gestation were compared with that from Wt/Wt animals on Days 18 and 19 as reported previously [9].
Histology and Immunohistochemistry
Sagittal sections of cervix for histological analysis were fixed in 10% formalin or 4% paraformaldehyde. Sections were stained with Masson trichrome stain, which stains collagen fibrils blue.
For immunohistochemistry, freshly excised cervices were embedded in Tissue-Tek OCT compound and frozen immediately in liquid nitrogen. Frozen sections (thickness, 5 µm)were fixed in acetone for 10 min. Nonspecific binding was blocked using 1.5% normal donkey serum for 20 min. Sections were incubated for 30 min at 25°C with monoclonal rat anti-mouse neutrophil antibody 7/4 (working dilution, 0.01 mg/ml; Serotec, Raleigh, NC). Biotinylated donkey anti-rat antibody (1:200; Jackson Laboratories, Westgrove, PA) and alkaline phosphatase-conjugated avidin-biotin complex (Vector Laboratories, Burlingame, CA) were applied in sequence followed by 25-min incubation with Vector Red substrate (Vector Laboratories). Tissues were counterstained in hematoxylin for 10 sec. The primary antibody was replaced with rat immunoglobulin G2 (Caltag Laboratories, Burlingame, CA) as a negative control.
Affymetrix Gene Arrays
Total RNA was obtained from pooled cervices from Wt/Wt and Tg/ Tg animals on Gestational Day 18.5 and used to generate cRNA probes that were hybridized to Affymetrix murine genome U74A V2 chips. Signal intensity of each gene was analyzed using Microarray Suite 5.0 (Affymetrix) and then imported into Gene Spring 5.1 (Silicon Genetics, Redwood City, CA) for further analysis. Gene expression was then normalized to the average signal intensity of each chip, and individual gene expression values were normalized to the median value of the signal intensity of 6– 12 oligonucleotides present for each gene. Filters were applied to analyze genes with a minimum raw signal intensity of 100 in at least one of the chips and flagged as present or marginal.
Statistical Analysis
Differences in gene expression between specimens from Wt/Wt and mutant animals were determined using the Student t-test for normally distributed data or the Mann-Whitney U-test for data that were not normally distributed. Bonferroni adjustment was conducted if multiple independent comparisons were performed. Differences among three or more groups were determined using the Kruskal-Wallis one-way analysis of variance on ranks using Dunn pairwise comparisons against Wt/Wt animals. Ranges are mean ± SEM throughout.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Transgenic mice expressing the human apolipoprotein B gene (APOB), which encodes a large protein involved in plasma cholesterol transport, were created by injecting fertilized oocytes with DNA from a bacteriophage P1 clone harboring a 80-kilobase (kb) genomic fragment containing the entire human APOB gene [20]. Ten founder lines were established that integrated the transgene in 4–42 copies [20]. Lines expressed various levels of human APOB transcripts in liver tissue, had similar elevations in plasma APOB protein, were fertile, and transmitted the transgene to F1 progeny in a mendelian fashion. Homozygous females derived from one founder line, 1102, the formal designation of which is TgN(hApoB)1102SY, failed to deliver their offspring.
Founder line 1102 contains 10 copies of the 80-kb transgene per haploid genome. The transgene insertional locus was cloned, and the position of the insertion was mapped through construction and screening of a cosmid library containing genomic fragments from line 1102 and through PCR using sequence information derived from the preinsertional locus [21]. The position of the 800-kb insertion locus was mapped to mouse chromosome 6. The transgene insertion is associated with a deletion of 390 base pairs (bp) of mouse genomic sequence with no other detectable rearrangements. The transgene insertion position was further confirmed by fluorescent-based in situ hybridization of mouse metaphase chromosome preparations from heterozygous animals using the human apolipoprotein B gene as a probe. No transcription units were identified in the 390-bp deleted mouse sequences.
Females from this line maintained pregnancy for several days after the expected time of parturition, when either maternal death or fetal resorption occurred. The parturition defect was not related to overexpression of human APOB protein, because reproductive function (fertility and timely onset of parturition) was normal in nine homozygous lines with plasma APOB levels equal to or greater than the affected founder line. Parturition frequency was virtually normal in heterozygous females (Table 1). All homozygous transgenic pregnant females in the line TgN(hApoB)1102SY (hereafter referred to as Tg/Tg) failed to deliver, irrespective of litter genotype (Table 1), suggesting that parturition failure in Tg/Tg females was a maternal defect not related to male or fetal factors.
|
Expression of Parturition-Associated Genes in Reproductive Tissues of Tg/Tg Mice
In mice, parturition is preceded by luteolysis (progesterone withdrawal) and expression of several genes that are normally suppressed by progesterone. Uterine and cervical tissues were collected from Wt/Wt and Tg/Tg pregnant dams on the evening before the expected time of parturition (Day 18), and expression of genes known to be temporally and hormonally regulated in the uterus and cervix during pregnancy was quantified by real-time PCR (Fig. 1, A and B). Transcript levels of two genes that are normally increased after progesterone withdrawal on Gestational Day 18 (oxytocin receptor and connexin 43) were evaluated, as were genes that modulate progesterone function (i.e., Srd5a1 and the progesterone receptor). Additionally, mRNA expression of COX1 (Ptgs1) and COX2 (Ptgs2), which are rate-limiting enzymes involved in uterine prostaglandin production, and connexin 26, a gap junction protein expressed abundantly in the cervix, were evaluated.
|
Expression of connexin 43, Srd5a1, and progesterone receptor was reduced in uterine tissue from Tg/Tg compared with that in tissues from Wt/Wt mice (Fig. 1A). Oxytocin-receptor mRNA also was reduced, but not to a statistically significant level. Transcripts of Ptgs1 (COX1) were more abundant in the uterus and cervix compared to transcripts of Ptgs2 (COX2). Expression of Ptgs1 (COX1) was not different in Wt/Wt and mutant animals on Day 18, but Ptgs2 (COX2) expression, paradoxically, was increased in the Tg/Tg uterus. Connexin 26 was expressed at low levels in the uterus and was unchanged between the two genotypes. In the cervix, with the exception of the COX genes, expression of the examined genes was significantly decreased in Tg/Tg animals compared with that in Wt/Wt animals on Day 18 (Fig. 1B). Although expression of Ptgs2 (COX2) was similar in both groups, transcript levels were low to undetectable in the cervix at this stage of gestation (average cycle threshold [Ct] values of 28.0).
To determine whether decreased expression of SRD5A1 mRNA in uterine and cervical tissue from Tg/Tg animals resulted in a decline of progesterone metabolism at term, we evaluated steroid metabolism on Day 18 in uterine and cervical tissue homogenates (Fig. 1C). Metabolism of progesterone by SRD5A1 and 20-hydroxysteroid dehydrogenase appeared to be similar in uterine tissues homogenates of Tg/Tg pregnant mice compared with those of Wt/ Wt animals, despite the decline in Srd5a1 transcripts. In contrast, although detectable, SRD5A1 activity was decreased in cervical tissue homogenates from pregnant Tg/ Tg animals on Day 18, consistent with the decrease in Srd5a1. These data indicate that progesterone metabolism was reduced, but not absent, in the Tg/Tg cervix on Day 18.
Overall, the results presented in Figure 1 indicate that gene products normally increased in the reproductive tissues of preparturient animals (oxytocin receptor, Srd5a1, and connexin 43) are suppressed in tissues from Tg/Tg animals. Decreased mRNA and enzyme activity of SRD5A1 in the cervix of Tg/Tg mice (Fig. 1C) may result in increased levels of bioactive progesterone in this tissue, thereby leading to suppression of oxytocin receptor and connexin 43 mRNA.
Cervical Tissue Levels of Progesterone and 20-Hydroxyprogesterone
Tissue levels of progesterone and 20-hydroxyprogesterone in the cervix of Wt/Wt and Tg/Tg animals were measured by RIA on Days 17–19 (Fig. 2). Progesterone levels in the cervix were increased significantly in Tg/Tg animals on Day 18 compared with Wt/Wt animals on Day 18. However, consistent with the delay in luteolysis, tissue levels of progesterone on Day 19 decreased significantly in the Tg/Tg cervix to levels that approached those of Wt/Wt animals on Day 19 (Fig. 2A). Tissue levels of the progesterone metabolite, 20-hydroxyprogesterone, were similar in cervical tissues from Wt/Wt and Tg/Tg mice (Fig. 2B).
|
Progesterone Withdrawal and Oxytocin Signaling in Tg/Tg Pregnant Mice
To address the role of progesterone and other steroid hormones during parturition failure in Tg/Tg animals, levels of progesterone, androstenedione, and 17ß-estradiol were determined in serum from pregnant Wt/Wt and Tg/ Tg animals on Days 17–19 (Fig. 3). In Wt/Wt animals, serum levels of progesterone decreased significantly on Day 18, and these levels remained suppressed on the morning of Day 19. In Tg/Tg animals, high levels of serum progesterone were maintained on Day 18, but on Day 19, these decreased significantly to levels comparable with those of Wt/Wt animals. Serum levels of androstenedione (a C19 steroid produced by the mouse placenta that preserves function of the corpus luteum during pregnancy [22]) also were significantly increased in Tg/Tg animals on Day 18 (Fig. 3B). Estradiol levels were similar between the two genotypes (Fig. 3C). These data suggest that luteolysis, a process that results in decreased serum levels of progesterone before parturition, is delayed in Tg/Tg animals. Nevertheless, Tg/Tg animals failed to deliver on Days 19 and 20, even after marked decreases in serum progesterone on Day 19 and declining levels of androstenedione.
|
In species that undergo progesterone withdrawal, administration of progesterone-receptor antagonist or removal of the source of progesterone by ovariectomy during the latter stage of pregnancy results in uterine contractility, increased expression of uterine oxytocin receptors, cervical ripening, and delivery of the fetus [14]. To examine the role of attenuated progesterone metabolism in pregnant Tg/Tg mice, the progesterone-receptor antagonist, ZK98299, was administered, or ovariectomy was performed, on Day 18. These treatments did not reverse the parturition defect in Tg/Tg mice, although Wt/Wt mice delivered within 18–24 h (Table 2).
|
The role of oxytocin signaling in parturition failure of Tg/Tg animals was assessed by administration of high-dose oxytocin on Day 18 to Tg/Tg mice (Table 2). Up to three doses were administered at 30-min intervals, and animals were observed for delivery. All Wt/Wt mice delivered pups after oxytocin treatment, whereas none of the Tg/Tg mice delivered pups. After administration of oxytocin, Tg/Tg animals exhibited maternal behavior characteristic of parturient Wt/Wt mice (i.e., decreased locomotion, grooming of lower abdominal and anogenital regions, and lordosis contractions) [23]. This observation suggests that uterine responses to oxytocin are intact in Tg/Tg animals and that oxytocin-receptor mRNA is expressed, although at suppressed levels on Day 18 relative to those in Wt/Wt animals, in uterine tissues from pregnant Tg/Tg animals and is responsive to high concentrations of oxytocin.
Assessment of Uterine Function
Next, changes in uterine and cervical function during pregnancy were analyzed in Wt/Wt and Tg/Tg animals. Myometrial contractions were monitored in vivo using intrauterine pressure catheters in anesthetized mice on Day 16, late on Day 18, and on the morning of Day 19 (Table 3). The uterus was quiescent in Wt/Wt and Tg/Tg animals on Day 16, with rare contractions of low amplitude (data not shown). This quiescent pattern changed late on Day 18 in Wt/Wt females. Frequent spontaneous increases in intrauterine pressure (3.7 ± 0.8 mm Hg every 1.3 ± 0.2 min) were superimposed on slower cycles of increased intrauterine pressure (four to six per hour). Treatment with oxytocin resulted in significant increases in intrauterine pressure (6.6 ± 0.3 mm Hg) in 6 of 10 Wt/Wt animals on Day 18. Two animals manifested modest increases in uterine contractions in response to oxytocin (2.5–2.7 mm Hg), whereas oxytocin responses were not evident in two animals in labor (as determined by pups in birth canal at end of the experiment). In contrast, in four of five Day-18 Tg/ Tg animals, both the amplitude and frequency of spontaneous contractions were suppressed, and oxytocin responses were either absent or diminished (increase of <1.5 mm Hg, P 
0.01 vs. Wt/Wt). On the morning of Day 19, peak amplitude, duration, and frequency of uterine contractions in nondelivering Tg/Tg animals were indistinguishable from those of Wt/Wt animals in active labor (Table 3). In addition, in vivo oxytocin responses were of similar magnitude (increase in intrauterine pressure of 4.7 ± 0.7 mm Hg).
|
In studies not shown, the frequency, duration, and force amplitude of spontaneous and oxytocin-induced myometrial contractions were determined in uterine smooth muscle strips from Wt/Wt (Days 18 and 19) and Tg/Tg (Day 19) mice. Consistent with the in vivo assessments of uterine contractility, the amplitude, frequency, and duration of contractions in vitro were similar in both genotypes.
Experimental results indicating diminished uterine contractility and oxytocin responsiveness in Tg/Tg animals on Day 18 were consistent with diminished expression of connexin 43 and oxytocin-receptor mRNA in uterine tissues from Tg/Tg mice on Day 18 (Fig. 1). Although suppressed on Day 18, uterine contractility and oxytocin responses on Day 19 were of similar magnitude and duration as those in delivering Wt/Wt animals, but delivery failed to occur.
Assessment of Cervical Function During Pregnancy
Biomechanical properties of cervical tissues from Day-18 and Day-19 mice were determined by distending the cervix isometrically in physiologic saline solution in water-jacketed tissue baths and quantifying tension generated in response to progressive increases in cervical diameter (Fig. 4A). Cervical stiffness was calculated as the slope of the stress-strain curves. Distensibility was decreased in cervices from Tg/Tg mice on late Day 18 to early Day 19 compared with tissues from Tg/Wt animals (Fig. 4A), indicating a stiff, noncompliant cervix in Tg/Tg mice. Unexpectedly, cervical stiffness (i.e., poor distensibility) also was increased in heterozygous animals relative to Day-18 and Day-19 Wt/Wt animals (Fig. 4B).
|
In rodents, substantial changes in cervical elasticity occur during early to midpregnancy (cervical softening) [24], and this increase in elasticity progresses until the last day of pregnancy, when a second and more rapid stage (cervical ripening) occurs [19, 25]. The marked decrease in cervical distensibility in Tg/Tg animals during late pregnancy led us to evaluate the biomechanical properties of the cervix earlier in gestation during cervical softening that occurs by Day 12. The data indicate that even on Day 12, distensibility of the Tg/Tg cervix was compromised compared to that of Wt/Wt mice (Fig. 4B). The cervix of Tg/Tg animals on Day 12 was almost as rigid as that of nonpregnant Wt/ Wt mice. In heterozygous animals, cervical stiffness on Day 12 was intermediate between that of Wt/Wt and Tg/ Tg animals, and these differences became statistically different by late gestation (P 0.05). Biomechanical properties of the cervix changed during pregnancy in Tg/Tg animals, indicating that although distensibility of the Tg/Tg cervix was compromised on Day 19 compared with the Wt/ Wt cervix, substantial remodeling of the connective tissue occurred from Days 12 to 19 in these animals.
Thus, results from these experiments indicate that the Tg/Tg cervix is rigid on Day 12 compared with tissues from pregnant Wt/Wt animals. Although the cervix of Tg/Tg animals undergoes significant increases in cervical compliance from Day 12 to Day 19, and although the magnitude of this change in compliance is similar to that of Tg/Wt and Wt/Wt animals during pregnancy (i.e., stiffness decreases by two- to threefold), this remodeling is not sufficient to overcome the inelastic properties of the Tg/Tg cervix.
Relaxin is a polypeptide hormone known to cause softening and lengthening of the pubic ligament as well as cervical softening [26]. To evaluate Tg/Tg mice for possible abnormalities in relaxin function, length of the pubic ligament was measured in nonpregnant mice and pregnant animals on Day 18. Pubic ligament length increased from less than 1 mm in nonpregnant mice to 3.5 mm on Day 18 in Tg/Tg animals, similar to that in Wt/Wt animals and as reported previously [26]. Furthermore, mRNA levels of the relaxin receptor Lgr7 [27] were similar in Wt/Wt and Tg/ Tg cervical tissues (data not shown). These data suggest normal relaxin function in Tg/Tg mice.
Histologic Analysis of Cervical Tissues from Wt/Wt and Tg/Tg Animals
Histologically, normal cervical ripening is characterized by disorganized collagen fibrils, increased secretion of mucus by epithelial cells, and an influx of inflammatory cells [28]. Histological and immunohistological assessment of sagittal sections of the cervix from late-gestation Tg/Tg mice was performed to further evaluate the failure of cervical ripening in this mouse model. Sections from Day-18.5 Wt/Wt and Tg/Tg mice were stained with Masson trichrome to assess collagen organization. Immune cell migration was evaluated immunohistochemically using the antibody 7/4, which recognizes mouse neutrophils and monocytes but not macrophages [29, 30].
In Wt/Wt cervices on late Day 18, trichrome staining revealed cervical fibroblasts suspended in a loose array of disordered collagen fibers (morphometric features characteristic of compliant tissue) (Fig. 5, A and C). On the morning of Day 19, the cervix of Tg/Tg animals exhibited a denser, compact, heavily stained matrix of collagen fibers (Fig. 5, B and D). In addition, the size and number of dilated venules in the lamina propria and in the cervical stroma were markedly reduced in the cervix of Tg/Tg animals. Trichrome staining also indicated that epithelial cells of Wt/ Wt, but not Tg/Tg, endocervices contained abundant secretory vacuoles laden with mucins (Fig. 5, B and C). The number of neutrophils and/or tissue monocytes in the Tg/ Tg cervix was increased significantly compared with the time-matched Wt/Wt cervix (Fig. 6). In Wt/Wt animals on Day 15 (before cervical ripening), neutrophils and tissue monocytes were localized and concentrated just beneath the basal lamina of cervical epithelium (unpublished observations). On late Day 18, neutrophils and tissue monocytes increased in number and were redistributed from the subepithelium to the cervical stroma in Wt/Wt mice (Fig. 6A). In Tg/Tg cervices on late Day 18, large aggregates of neutrophils and/or monocytes were found throughout the cervical stroma as well as the subepithelium (Fig. 6B). The leukocyte congregates were not distributed uniformly throughout the female reproductive tract but were highly localized to the cervix of Tg/Tg animals. In contrast to the cervix, vaginal subepithelial leukocytes did not migrate into the underlying muscularis, and leukocytes did not accumulate in the myometrium or endometrium.
|
|
The overall size of the cervix was decreased in the Tg/ Tg mice. This observation was quantified by measurement of cervical wet weight (Fig. 7). On both Day 2 and Day 18, the weight of the cervix from Tg/Tg animals was decreased significantly compared to that from Wt/Wt animals, with the cervical weight from Tg/Wt animals being intermediate between that of Wt/Wt and Tg/Tg animals (Fig. 7). Despite the attenuated size of the Tg/Tg cervix on Day 12, cervical wet weight increased proportionately in mutant pregnant animals at term (i.e., approximately twofold). Furthermore, the water content of cervical tissue from Tg/Tg animals on Gestational Day 18 was 81%, a value almost identical to that in the Wt/Wt cervix (79.9%; unpublished observations), suggesting that although the cervix is significantly smaller in Tg/Tg animals, some adaptations of the tissue during gestation, such as cervical edema, occur.
|
Transcriptional Profile of Cervix from Tg/Tg Animals
To understand the molecular basis for alterations in cervical ripening in Tg/Tg animals better, cDNA microarrays were used to compare expression levels of several thousand transcripts in cervical tissues from Wt/Wt and Tg/Tg mice on Day 18 (data not shown). The majority of upregulated genes were inflammatory response genes, consistent with the histologic evidence that neutrophil and tissue monocytes are abnormally increased in the Tg/Tg cervical stroma. Table 4 lists the inflammatory response genes identified in the microarray that were verified by quantitative real-time PCR. These included interferon-induced gene products, which have been described as glucocorticoid-attenuated response genes (e.g., Garg49 and Garg39), cytokine receptors (Ccr5), and several proteases (e.g., Adamdec1). Gene products associated with apoptosis also were increased in the Tg/Tg cervix. For example, mRNA for a G protein-coupled receptor involved in glucocorticoid-induced apoptosis (T cell death-associated gene 8 [Gpr65]) was increased 4.9-fold in the mutant cervix compared with the Wt/Wt cervix (Table 4).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Luteolysis Is Delayed in Homozygous Transgenic Mice
In many species, luteolysis is a prerequisite for normal parturition [32]. Luteolysis involves a loss of the capacity by the corpus luteum to synthesize and secrete progesterone, followed by a loss of cells that comprise the corpus luteum itself. Prostaglandin F2, synthesized in the uterus primarily by uterine PTGS1 (COX1), is required to initiate luteolysis [32, 33]. It also facilitates a reduction in blood flow to the corpus luteum, an increase in luteal cell apoptosis, reduced progesterone biosynthesis (by decreasing transport of cholesterol into the endoplasmic reticulum), and increased progesterone metabolism through enhanced expression of the catabolic enzyme 20-hydroxysteroid dehydrogenase [32].
In Tg/Tg mice, luteolysis is delayed. This conclusion is supported by several lines of evidence. First, two genes that are normally repressed by progesterone (oxytocin receptor and connexin 43) were decreased in uterine and cervical tissues from Tg/Tg mice on the day before birth (Day 18) compared with Wt/Wt mice that have undergone physiologic luteolysis and progesterone withdrawal by this time point [33]. Second, serum and tissue progesterone concentrations and serum androstenedione levels are increased on Day 18 in Tg/Tg animals compared with Wt/Wt animals. The latter steroid hormone antagonizes luteolysis by inhibiting luteal cell apoptosis [22]. Third, the force and frequency of uterine contractions and contractile responsiveness to oxytocin are diminished in Tg/Tg animals on Day 18. Although delayed, luteolysis is not blocked in Tg/Tg animals. On the day of expected birth, serum and tissue progesterone concentrations decrease, and serum androstenedione concentrations decline. Uterine contractility appears to be normal on Day 19, also suggesting that luteolysis takes place. Nevertheless, parturition does not occur. These results, together with the inability of ovariectomy or progesterone-receptor antagonism to bring about parturition in these animals, further supports the conclusion that the parturition defect is not the result solely of delayed luteolysis or compromised tissue metabolism of progesterone, as observed in mice that are deficient in SRD5A1 [9] or steroid 20-hydroxylase [7].
Cervical Distensibility Is Impaired in Tg/Tg Mice
The present results indicate that the primary cause of failed delivery in Tg/Tg animals is insufficient remodeling of the cervix to allow expulsion of the fetus by the contracting uterus. Several lines of evidence support this conclusion. On Day 19, the force and frequency of uterine contractions are similar in Tg/Tg animals and parturient Wt/Wt animals. Biomechanical, histological, and immunohistological assessments, however, indicate that cervical remodeling and loosening of the extracellular matrix is inadequate in pregnant Tg/Tg mice on Days 18 and 19.
Cervical remodeling is initiated in Tg/Tg mice, as evidenced by increased elasticity of the cervix on Day 19 compared with that on Day 12. In Tg/Tg mice, however, further changes in elasticity do not occur before the expected time of parturition, and cervical distensibility is compromised in Tg/Tg animals at all time points in gestation.
The Phenotype of Tg/Tg Animals Is Distinct from Srd5a1-Null Mice
Like Srd5a1-knockout mice, parturition failure is associated with insufficient cervical ripening in Tg/Tg animals. Although SRD5A1 mRNA and enzyme activity are decreased in Day-18 cervical tissues from Tg/Tg animals, parturition failure in Tg/Tg mice cannot be a result of only compromised SRD5A1 activity in the cervix for several reasons. First, the weight of the pregnant cervix in the SRD5A1-deficient mouse is similar to that in the Wt/Wt mouse, in contrast to that in the Tg/Tg mouse. Second, delivery fails to occur in 100% of pregnant Tg/Tg dams, compared with 70% of SRD5A1-deficient mice. Third, antiprogestin treatment or ovariectomy does not rescue the parturition defect in Tg/Tg animals, as they do in SRD5A1-deficient mice. Fourth, cervical rigidity is more pronounced in Tg/Tg mice [8, 9]. Cervices of heterozygous animals at term are less distensible than those of Srd5a1-knockout mice. The fact that Tg/Wt animals deliver on time may be a result of the forceful uterine contractions on Day 19. Previously, we found that uterine contractions in Srd5a1-knockout mice are decreased in vivo at term, but these differences were not statistically different in the animals studied [9]. In Tg/Tg animals, uterine contractions were not compromised at term. Thus, although compromised uterine contractility in SRD5A1-deficient mice may contribute to parturition failure, normal and/or prolonged uterine contractions may be sufficient to deliver the fetus in Tg/Wt animals despite decreased elasticity in the heterozygous cervix.
Another difference in the Tg/Tg cervix compared with the SRD5A1-deficient cervix is the large numbers of inflammatory cells in the cervical stroma of Tg/Tg animals. This increase is supported by findings showing that several inflammatory response genes that are upregulated at term in the Tg/Tg cervix (Table 4) are not increased in the cervix of Srd5a1-knockout mice (data not shown).
Although some histologic features of the Tg/Tg unripe cervix are distinct from those in SRD5A1-deficient mice, some features are similar. Thus, the cervical stroma of both animals contained a dense collagen matrix, and tissue from both mice at term exhibited a lack of mucin-laden epithelial cells in the cervical mucosa.
Lack of Relationship Between Inflammatory Cell Infiltrate and Cervical Ripening in Tg/Tg Mice
Infiltrating leukocytes were abundant in the cervix of Tg/ Tg animals, but not in the uterus or vagina, late on Day 18. Previous studies have described increased macrophage infiltration into the mouse cervix before parturition [34] and increased leukocytes in the human cervix immediately postpartum [35], but our findings in the Tg/Tg mice indicate a far greater leukocyte infiltration in the Day-18.5 Tg/Tg cervix compared with Wt/Wt controls (Fig. 6) as well as the Day-15 Tg/Tg cervix (data not shown). Consistent with the increased density of 7/4-positive neutrophils and tissue monocytes, a number of inflammatory genes were upregulated significantly in the Day-18.5 Tg/Tg cervix. Given the proposed role of leukocytes in facilitating remodeling of the extracellular matrix during cervical ripening and the suppressive effect of progesterone on immune function, this increase in inflammatory cells seems to be paradoxical in light of the dense stromal matrix and cervical rigidity of Tg/Tg mice. In a recent study by Sakamoto et al. [36], cytokine and cytokine-receptor levels were determined in women before labor both with and without a ripe cervix and in labor or immediately after delivery. Compared with late pregnancy, cytokine (interleukin-8 and interleukin-1ß) levels increased after labor and vaginal delivery but not with cervical ripening, leading to the conclusion that leukocytes may not be involved in initiation of cervical ripening but, rather, in cervical dilatation or postpartum repair. Our findings in cervical tissues from Tg/Tg mice also suggest that the role of infiltrating inflammatory cells in the initiation of cervical ripening should be investigated further using animal models in which tissue samples can be obtained from defined time points before, during, and after cervical ripening.
Cervical Size in Tg/Tg Mice
The small size of the Tg/Tg cervix could result from a decrease in cell proliferation or from an increase in apoptosis. Dramatic upregulation of Gpr65, a gene involved in glucocorticoid-mediated apoptosis, suggests that increased apoptosis may, in part, facilitate reduction in cervical mass in Tg/Tg animals. Further studies are required to evaluate apoptotic events in the Tg/Tg cervix, and the temporal relationship between apoptosis and failure of cervical ripening during pregnancy.
A fortuitous transgene insertional mutation has led to inactivation, dysregulation, or abnormal function of cellular gene(s) required for normal cervical function during parturition. We hypothesize that the genetic malfunction is not related to progesterone withdrawal and is crucial for ripening of the cervix during parturition. Sequence analysis of a 1.6-megabase region spanning the transgene insertion site has led to the identification of 16 genes, none of which has been associated with parturition. Investigations are currently underway to identify the gene and molecular pathways responsible for the parturition phenotype.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Mala Mahendroo, Department of Obstetrics and Gynecology, UT Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75390-9032. FAX: 214 648 9242; mala.mahendroo{at}utsouthwestern.edu
Received: 18 April 2005.
First decision: 6 June 2005.
Accepted: 15 July 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-hydroxysteroid dehydrogenase. Mol Endocrinol 2005 19:431-440-Reduced androgens play a key role in murine parturition. Mol Endocrinol 1996 10:380-392[Abstract]-reductase type 1 knockout mice is due to impaired cervical ripening. Mol Endocrinol 1999 13:981-992This article has been cited by other articles:
|
|
 |
|
  S. L. Pierce, J. D.K. Kresowik, K. G. Lamping, and S. K. England Overexpression of SK3 Channels Dampens Uterine Contractility to Prevent Preterm Labor in Mice Biol Reprod, June 1, 2008; 78(6): 1058 - 1063. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. C. Timmons and M. Mahendroo Processes Regulating Cervical Ripening Differ From Cervical Dilation and Postpartum Repair: Insights From Gene Expression Studies Reproductive Sciences, December 1, 2007; 14(8_suppl): 53 - 62. [Abstract] [PDF]
|
|
|
|
 |
|
 C. P Read, R A. Word, M. A Ruscheinsky, B. C Timmons, and M. S Mahendroo Cervical remodeling during pregnancy and parturition: molecular characterization of the softening phase in mice Reproduction, August 1, 2007; 134(2): 327 - 340. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. L. McFarlin, W. D. O'Brien Jr, M. L. Oelze, J. F. Zachary, and R. C. White-Traut Quantitative ultrasound assessment of the rat cervix. J. Ultrasound Med., August 1, 2006; 25(8): 1031 - 1040. [Abstract] [Full Text] [PDF]
|
|
