This Article Abstract Full Text (PDF) Supplementary Table Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Read, C. P Articles by Mahendroo, M. S Search for Related Content PubMed PubMed Citation Articles by Read, C. P Articles by Mahendroo, M. S

Cervical remodeling during pregnancy and parturition: molecular characterization of the softening phase in mice

Department of Obstetrics and Gynecology, The University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas 75390-9032, USA

Correspondence should be addressed to M S Mahendroo; Email: mala.mahendroo{at}utsouthwestern.edu

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Acknowledgements References

 
Cervical remodeling during pregnancy and parturition is a single progressive process that can be loosely divided into four overlapping phases termed softening, ripening, dilation/labor, and post partum repair. Elucidating the molecular mechanisms that facilitate all phases of cervical remodeling is critical for an understanding of parturition and for identifying processes that are misregulated in preterm labor, a significant cause of perinatal morbidity. In the present study, biomechanical measurements indicate that softening was initiated between gestation days 10 and 12 of mouse pregnancy, and in contrast to cervical ripening on day 18, the softened cervix maintains tissue strength. Although preceded by increased collagen solubility, cervical softening is not characterized by significant increases in cell proliferation, tissue hydration or changes in the distribution of inflammatory cells. Gene expression studies reveal a potentially important role of cervical epithelia during softening and ripening in maintenance of an immunomucosal barrier that protects the stromal compartment during matrix remodeling. Expression of two genes involved in repair and protection of the epithelial permeability barrier in the gut (trefoil factor 1) and skin (serine protease inhibitor Kazal type 5) were increased during softening and/or ripening. Another gene whose function remains to be elucidated, purkinje cell protein 4, declines in expression as remodeling progressed. Collectively, these results indicate that cervical softening during pregnancy is a unique phase of the tissue remodeling process characterized by increased collagen solubility, maintenance of tissue strength, and upregulation of genes involved in mucosal protection.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Acknowledgements References

 
Premature birth is one of the most significant causes of perinatal morbidity in developed countries, and its incidence over the last decade has been increasing (Branum & Schoendorf 2002, Ananth et al. 2005). Investigations to define the causes of preterm birth remain a challenge in the obstetrical field as phenotypes of preterm labor and delivery vary extensively. As an example, precocious remodeling of the cervix in the absence of uterine contractions, clinically known as cervical incompetence or insufficiency in humans, is associated with preterm birth (Papiernik et al. 1986).

The cervix is a metabolically active organ in pregnancy composed of extracellular matrix (ECM) components such as collagen, elastin, proteoglycans, and hyaluronan in addition to stromal, epithelial, and smooth muscle cells (Leppert 1995). The cervix is extensively remodeled throughout gestation in order to open sufficiently for safe passage of the fetus during childbirth. Cervical remodeling during pregnancy and parturition is a single continuous process that can be loosely divided into four overlapping phases termed softening (phase 1), ripening (phase 2), dilation/labor (phase 3), and post partum repair (phase 4; Liggins 1978, Word et al. 2007).

In 1895, Hegar first described ‘softening’ of the lower uterine segment in association with human pregnancy at 4–6 weeks. The Hegar sign refers to the characteristic softening of the uterus in cervix that is evident on physical examination during early pregnancy, and it was customarily used to clinically diagnose pregnancy until the discovery of human chorionic gonadotropin many years later (Hegar 1895). Cervical softening can be defined as a change in the biomechanical properties of the cervix when compared with the nonpregnant cervix and is characterized by a progressive decrease in tissue stiffness without loss of tensile strength (Leppert & Yu 1994). The ability of the cervix to soften yet remain resistant to forces exerted upon it requires dual mechanisms in which there is an increase in the geometry (size) of the cervix along with maintained stiffness of the cervical wall (Drzewiecki et al. 2005). Physiologic softening of the rat cervix was first recognized by Harkness & Harkness (1959) who noted that distensibility of cervical tissue increased dramatically starting midgestation between days 11 and 12 (Harkness & Harkness 1959). An increase in the diameter of the cervical canal between gestation days 10 and 12 was used initially as a crude measurement to define cervical softening in the mouse (Leppi 1964). The change in diameter, however, was <1 mm and statistical analysis was not performed. Changes in tissue compliance during softening are proposed to be facilitated by changes in the composition or structure of the ECM in humans and rats (Danforth et al. 1974, Liggins 1978, Leppert 1995). This is supported by the fact that cervical incompetence is increased in women with inherited defects in collagen and elastin synthesis or assembly (e.g. Ehlers-Danlos and Marfan syndromes; Paternoster et al. 1998, Rahman et al. 2003). In addition, mice deficient in the ECM protein, thrombospondin 2, have altered collagen fibril morphology and premature cervical softening (Kokenyesi et al. 2004).

Cervical softening begins early in mammalian pregnancies and overlaps at the end of gestation in a second phase of matrix remodeling commonly referred to as cervical ripening. While the endocrine events that herald this stage differ from species to species, the series of events that bring about ripening are very similar. Cervical ripening has been studied much more extensively than the process of cervical softening. Ripening is characterized by an increase in hyaluronan content, loosening of the collagen matrix, increased collagen solubility (Hillier & Wallis 1982, Granstrom et al. 1989), changes in the distribution of inflammatory cells, increased tissue growth and hydration, and loss of tensile strength (Uldbjerg et al. 1983, Buhimschi et al. 2004, Straach et al. 2005, Timmons & Mahendroo 2006). In contrast to cervical ripening, the biochemical and molecular changes that occur during cervical softening are not well-defined. We propose that processes regulating softening events may differ from cervical ripening and are crucial to our understanding of cervical remodeling for successful delivery of young. The objective of the current series of experiments is to understand the midgestation cervical softening process in mice and to identify genes that may be integral to one or all stages of the cervical remodeling process.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Acknowledgements References

 
Mice
Animals were housed under a 12L:12D photoperiod (lights-on, 0600–1800 h) at 22 °C. Mice used in the present studies were of mixed strain (C57BL/6x 129SvEv). The C57BL/6x129SvEv mice were generated and maintained as a breeder colony at the University of Texas Southwestern Medical Center (Dallas, TX, USA). In general, mice in these studies were 3–6 months old and nulliparous. Female mice were housed overnight with males and checked at midday for vaginal plugs in order to obtain accurately timed pregnant mice. The day of plug formation was counted as day 0, and birth occurred in the early morning hours of day 19. Cervices collected before noon on the day of birth were estimated to be 10 h post partum and those collected before noon on the day after birth were estimated to be 24 h post partum. All studies were conducted in accordance with the standards of humane animal care as described in the NIH Guide for the Care and Use of Laboratory Animals. The research protocols were approved by the institutional animal care and research advisory committee.

Cervical tissue collection
Animals were given anesthesia with 1.25% tribromo-ethanol (25 µl/g of body weight injected intraperitoneally). After euthanasia by cervical dislocation, the uterine horns, cervix, and vagina were removed. Using a dissecting microscope, the cervix was isolated by transection at the utero-cervical junction (caudal to the uterine bifurcation) and removal of all vaginal tissue from cervical tissue specimens. All specimens for RNA extraction were immediately snap frozen in liquid nitrogen.

Tissue biomechanics
Tensile distensibility and maximum stretch of isolated cervical tissues was conducted according to a method developed by Harkness & Harkness (1959) and adapted in our laboratory. The excised cervix was mounted by means of two pins inserted through the cervical canal. One pin was attached to a calibrated mechanical drive and the other pin to a force transducer. Tissues were incubated in a water-jacketed bath containing physiologic saline solution (NaCl (120.5 mM), KCl (4.8 mM), MgCl₂ (1.2 mM), CaCl₂ (1.6 mM), NaH₂PO₄ (1.2 mM), NaHCO₃ (20.4 mM), dextrose (10 mM), and pyruvate (1.0 mM), pH 7.4 at 37 °C) bubbled with 95% O₂/5% CO₂. Baseline cervical dilatation (i.e. resting diameter of cervical os) was quantified by determining the difference in cervical diameter at 0 (pins juxtaposed with no tension) and the inner cervical diameter at the initiation of tension as the pins were separated. Thereafter, the inner diameter of the cervix was increased isometrically in 0.5 mm increments at 2-min intervals. The amount of force required to distend the cervix and the tension exerted by the stretched tissue were recorded. The diameter was increased until either forces exerted by the tissue reached a plateau or the tissue tore. Force was plotted as a function of cervical diameter. Cross-sectional area and stress (force/cross-sectional area) were calculated. The slope of the linear portion of the force–strain curve was computed as an index of tissue stiffness and elasticity (Young’s modulus).

Water content
Cervix tissue was weighed immediately after tissue collection (wet weight) and again after freeze–drying overnight (dry weight). Water content was determined by the ratio of the difference between the wet and dry cervical weights to the dry cervical weight (wet weight–dry weight)/wet weight (Anderson et al. 2006).

Collagen solubility
Dry cervical samples were homogenized and incubated in 0.5 ml of 0.5 M acetic acid, 1 mg/ml pepsin (Sigma, catalog no. P7012-1G) at 4 °C for 24 h with gentle agitation. Samples were spun at 12 000 r.p.m. (13 400 g) and supernatant containing soluble collagen was separated from the insoluble pellet. Soluble collagen was hydrolyzed by the addition of an equal volume of 12 M HCl for a final concentration of 6 M HCl. The insoluble collagen was hydrolyzed by the addition of 6 M HCl. Both fractions were incubated at 100 °C for 20 h. Samples were allowed to cool uncapped and incubated at 100 °C until HCl was evaporated. The dried hydrolysates were dissolved in 250 µl of water and the hydroxyproline concentration in each fraction measured by a colorimetric method according to methods as described (Stegemann & Stalder 1967). Briefly, 20 µl sample or standard was diluted in 500 µl water. Two hundred and fifty microliters of chloramine-T solution (1.41 g chloramine-T (Sigma), 10 ml H₂O, 10 ml n-propanol, and 80 ml OH-pro buffer) were added to each tube and incubated at RT for 20 min. (OH-pro buffer:50 g citric acid 1 H₂O, 12 ml acetic acid, 120 g sodium acetate 3H₂O, 24 g NaOH. H₂O is added to bring volume up to 1200 ml with water. pH to 6.0 with 6 N NaOH and then add 300 ml n-propanol). Two hundred and fifty microliters of aldehyde/perchloric acid solution (1 g p-dimethyl-amino-benzaldehyde (JT Baker, Phillipsburg, NJ, USA), 4 ml n-propanol, and 1.73 ml 70% perchloric acid) was added to all samples and standards, vortexed, and incubated at 60 °C for 15 min. Two hundred and fifty microliters of each sample were aliquoted into a 96-well microtiter plate and absorbance was measured at 558 nm on a Tecan Saffire2 microplate reader (Tecan, San Jose, CA, USA). A trans-4-hydroxy-L-proline (Sigma catalog no. 56250) standard series (6, 3,2,1,0.5, and 0.2 µg) and blank (water) were included in the experiment. All measurements were carried out in duplicate, and samples that had an absorbance higher than that of the most concentrated standard were diluted further and measured again. The hydroxyproline concentration in the soluble and insoluble fraction was interpolated from the standard curve and expressed as a percentage of the total OH-Pro in the sample. Total OH-Pro in the sample was determined by adding OH-Pro in soluble and insoluble fractions. The amount of total collagen was calculated by assuming the hydroxyproline content of collagen was 14% (Horgan et al. 1990). The concentration of collagen was determined by dividing the total collagen (µg) by the wet weight (mg) of the tissue. Four to six cervices were measured individually for each time point and data represent an average±S.E.M.

Immunohistochemistry
Freshly excised cervices were embedded in OCT compound (Tissue Tek; Bayer Corp., Elkhart, IN, USA) and frozen immediately in liquid nitrogen. Air-dried tissue sections (thickness, 5 µm) were fixed for 10 min in acetone. Nonspecific binding was blocked using 1.5% normal donkey serum for 20 min. Sections were incubated for 30 min at 25 °C with the monoclonal antibody, neutrophil 7/4, a rat anti-mouse neutrophil, and monocyte-specific antibody at a working dilution of 0.01 mg/ml (Serotec, Raleigh, NC, USA) or rat anti-mouse BM8, a macrophage-specific antibody (1:800 of a 0.5 µg/ml stock; BACEM Biosciences Inc., King of Prussia, PA, USA). Neutrophil 7/4 recognizes myeloid lineage leukocytes, including neutrophils and monocytes, but not macrophages (Henderson et al. 2003, Taylor et al. 2003). BM8 is a pan-macrophage marker that recognizes the F4/80 antigen found on cell membranes and the cytosol of mononuclear phagocytes (Mackler et al. 1999). Biotinylated donkey anti-rat (1:200; Jackson Laboratories, Westgrove, PA, USA) and alkaline phosphatase-conjugated avidin–biotin complex (Vector Laboratories, Burlingame, CA, USA) were applied in sequence followed with Vector Red substrate (Vector Laboratories). Tissues were counter-stained in hematoxylin for 10 s. The primary antibody was replaced with rat IgG2a (Caltag Laboratories, Burlingame, CA, USA) as a negative control. Three animals were tested for each time point. Nonpregnant cervices were collected during metestrus.

RNA isolation and quantitative real-time PCR
Total RNA was extracted from frozen tissue using RNA Stat 60 (Tel-test B, Friendswood, TX, USA). Total RNA was treated with DNase I (DNA free; Ambion, Austin, TX, USA) to remove any genomic DNA contamination. cDNA was synthesized using a TaqMan cDNA synthesis kit (Applied Biosystems, Foster City, CA, USA). RT-PCR was performed using SYBR Green and a PRISM 7900HT Sequence Detection System (Applied Biosystems). Aliquots (20 ng) of total cDNA were used for each PCR and were performed in triplicate. Expression of each gene was expressed relative to that of the housekeeping gene cyclophilin (Ppib). Relative levels of gene expression were determined by the ddCt method (Applied Biosystems User Bulletin #2).

Microarray
RNA from three separate cervical samples were obtained on days 10 and 12 of gestation and subjected to microarray experiments. Gene Spring 5.0 (Silicon Genetics, Redwood City, CA, USA) software was used in data analysis and statistical calculations. Affymetrix mouse expression chip 430 2.0 data were exported to the gene analysis software. Per chip normalization to the 50th percentile expression level and per gene normalization to the median expression of all samples was performed. Only probes with two out of three samples scored as either present or marginal were used in the data analysis. Probe sets with average expression of 40 or less in both experimental and control groups were excluded. Fold change expression was calculated as the median expression of day 12 over the expression of day 10 in mouse cervix tissue. Annotations for the probe sets were derived from the Affymetrix web site (http://www.affymetrix.com/) and were updated with publicly accessible databases: National Center for Biotechnology Information (http://www.ncbi.nih.gov/), Information Hyperlinked over Proteins (http://www.ihop-net.org/UniPub/iHOP/), and Swiss Prot (http://www.ebi.ac.uk/swissprot/) to verify gene identity and update annotations from current literature references. After generating a comprehensive gene list, the function and ontology of each gene were examined and then scored to determine possible relevance to cervical remodeling. From this list, a number of genes were chosen for further validation and analysis.

Statistical tests
Means of data were compared with Student’s t-test where appropriate. Two-tailed P-value 0.05 was considered statistically significant. For multiple comparisons, a Kruskal–Wallis one way ANOVA on ranks was performed followed by the Dunnett or Tukey method. For the microarray analysis, a gene was considered significant if one of two conditions were satisfied: first, if there was a two-fold change in mean expression of the gene from days 10 to 12 of gestation as calculated by Gene Spring, or second, if there was a statistically significant change in expression as determined by parametric testing adjusted for multiple comparisons. These two methods were combined to create a comprehensive and inclusive gene list. Statistical significance was defined in this context using a Benjamini and Hochberg false discovery rate (FDR) of <40%. An FDR <40% has been considered appropriate for exploratory research and gene discovery (Welle et al. 2002, Reiner et al. 2003). Real-time PCR was used to confirm the differential expression of genes found to be significant by the microarray analysis to reduce the possibility of finding false positives from the less stringent microarray analysis.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Acknowledgements References

 
Biomechanical characterization of cervical softening
To better define the time point at which cervical softening begins in the mouse when compared with earlier studies (Leppi 1964), biomechanical studies were carried out in cervices from nonpregnant and time-mated pregnant mice on gestation days 10, 11, 12, and 18 using methods that have been described previously (Word et al. 2005). Data shown for gestation day 18 is previously published and is shown in Fig. 1 for comparison to days 10–12 (Word et al. 2005). Cervical stiffness on day 10 (day 10) was not statistically significant from that of nonpregnant animals. Cervical tissues from nonpregnant animals and pregnant mice on day 10 were not distensible as the cervix could not be stretched to more than 3.5 mm before failure (Fig. 1A and B). By day 12, cervical stiffness (g/mm) decreased from 20.2±2.6 g/mm in nonpregnant animals to 6.2± 0.9 g/mm (P0.001, Fig. 1A), and cervical distensibility increased from 3.2±0.2 mm in nonpregnant animals to 6.9±0.5 mm on day 12 (P0.001). Cervical stiffness declined further from days 12 to 18 (P<0.001), and distensibility also increased dramatically from days 12 to 18 (P<0.001; Fig. 1A and B). Despite decreased tissue stiffness and increased cervical distensibility on day 12, the maximal load before failure was similar in non-pregnant and pregnant animals on days 10 and 12 (99±19 mN/mm² when compared with 129± 17 mN/mm², p=NS, Tukey) suggesting that collagen integrity is maintained (Table 1). In contrast, the maximal load before failure decreased on late day 18 to 41± 8 mN/mm² (when compared with day 12, P<0.05, Tukey), suggesting that, in contrast to the softened cervix, the ripened cervix lacks cervical strength (Table 1). Collectively, these data indicate that, in the mouse, a significant change in biomechanical properties of the cervix (i.e. cervical softening) occurs by gestation day 12, and that certain biomechanical changes during the ripening phase are distinct from those during cervical softening.

View larger version (21K): [in this window] [in a new window]   Figure 1 Biomechanical properties of the mouse cervix during early pregnancy. (A) Tissue stiffness was determined from the slope of the stress–strain relationship in cervices from nonpregnant (NP, open bar, n=8) and pregnant mice on gestation days 10 (hatched bar, n=6), 11 (diagonal lines, n=7), 12 (solid bar, n=8), and 18 and 19 (gray bar, n=11). *P0.001 when compared with NP. (B) Cervices from nonpregnant (NP) and pregnant mice on gestation days 10, 11, 12, and 18 and 19 were distended in 0.5 mm increments every 2 min until failure. Data are expressed as maximal distension in mm and represent mean±S.E.M. of 6–11 cervices per time point. *P0.001 when compared with nonpregnant.

View larger version (14K): [in this window] [in a new window]   Figure 2 Regulation of cervical weight in pregnant mice during pregnancy. Cervical dry (gray square) and wet (black circle) weights were determined in nonpregnant (NP) and pregnant mice on gestation days 10, 11, 12, and 18.75. Both the mean cervical wet and dry weights were greater just prior to delivery at day 18.75 when compared with cervical tissues from nonpregnant or pregnant mice earlier in gestation (*P<0.001, Dunnett). (n=5–11 cervices at each time point).

View larger version (19K): [in this window] [in a new window]   Figure 3 Collagen solubility, total collagen, and collagen concentration in cervical tissues from nonpregnant (NP), pregnant (days 7, 10, 12, and 18) and post partum (PP) mice. Percentage of solubilized collagen was determined in nonpregnant metestrus (hatched bar), pregnant (black bars), and post partum (grey bars) (A). Total amount of collagen in the cervix and concentration of collagen normalized to cervical wet weight are shown in (B and C) respectively. *P<0.05; **P<0.001, ANOVA Dunn test comparison with NP (n=4–6 cervices at each time point).

View larger version (91K): [in this window] [in a new window]   Figure 4 Immunolocalization of leukocytes (neutrophils and monocytes) and macrophages in cervical tissues from nonpregnant (NP) and pregnant mice on gestation days 10 and 12. Tissue sections were oriented in the sagittal plane through the cervical os. The cervical lumen lined by epithelial cells is indicated by arrows and cervical stroma by open arrowheads. Immunohistochemistry was conducted using MAB 7/4 (upper panels) and BM8 (lower panels). Positive immunoreactivity is indicated by red staining. 7/4-positive cells (leukocytes and monocytes) were predominantly localized in the epithelium, whereas macrophages were primarily localized in the subepithelial and stromal compartments of the cervix. Sections are representative of three cervices from each gestation day. Bar represents 20 µm.

The first criterion for differential expression of genes during cervical softening (mean difference in expression from days 10 to 12 by two-fold) yielded a gene list with 64 entries and is represented in Table 2 (upregulated genes) and Table 3 (downregulated genes). Using parametric statistical tests adjusting for multiple comparisons generated an additional gene list of 247 entries. This list is available as supplemental material (Supplemental Table 1, which can be viewed online at www.reproduction-online.org/supplemental/). Candidate genes identified in microarray analysis based on a two-fold change (Cxcl5, Mcp7, Pcp4, Spink5, and Tff1; Tables 2 and 3) were more often confirmed by RT-PCR than genes selected based on parametric statistical analysis (Aspn and Sdc2; Supplement).

Two genes, asporin (Aspn) and syndecan 2 (Sdc2), were selected as they encode ECM proteins that could influence changes in the cervical ECM during softening. Other candidate genes included chemokine ligand 5 (Cxcl5, a neutrophil chemoattractant expressed by epithelial cells), mast cell protease 7 (Mcp7, a serine protease expressed in mast cells), trefoil factor 1 (Tff1, a secreted protein involved in repair and protection of gastrointestinal mucosal epithelia), purkinje cell protein 4 (Pcp4, a neuronal specific calmodulin regulatory protein that inhibits apoptosis), serine proteinase inhibitor lymphoepithelial Kazal-type inhibitor (Spink5, a protease inhibitor that prevents degradation of proteins required for maintenance of an intact epithelial barrier), and decay accelerating factor (Daf1, a protein that suppresses inflammatory responses via inhibition of activation of the complement cascade; Lin et al. 2001, Kanamori et al. 2003, Singh et al. 2004, Yang et al. 2004, Funaba et al. 2006, Masui et al. 2006).

Regulation of gene expression in the cervix during pregnancy and parturition
Relative mRNA levels of selected genes were determined in cervical tissues from mice on gestation days 10 and 12. Asporin (Aspn), syndecan 2 (Sdc2), mast cell protease 7 (Mcp7), and decay accelerating factor 1 (Daf1) were found to be differentially expressed on the microarray, but further analysis with RT-PCR did not substantiate these differences (Fig. 5). Four genes with confirmed changes are indicated (Fig. 5). The analysis of relative mRNA levels of these genes was subsequently expanded to include cervical samples that were obtained throughout gestation (from day 8 through day 1 post partum) using RT-PCR (Fig. 6). Chemokine C-X-C ligand 5 (Cxcl5) encodes a neutrophil chemoattractant and was found to be upregulated on day 12 (Fig. 5). Though Cxcl5 expression varied throughout gestation, Cxcl5 was consistently expressed at higher levels during pregnancy than nonpregnancy (Fig. 5), but there was no specific pattern of expression with respect to duration of pregnancy (data not shown).

View larger version (20K): [in this window] [in a new window]   Figure 5 RT-PCR confirmation of differentially regulated genes identified in microarray comparisons between days 10 and 12 of gestation. Chemokine ligand 5 (Cxcl5), Kazal-type serine protease inhibitor 5 (Spink5), and trefoil factor 1 (Tff1) were upregulated on day 12, while purkinje cell protein 4 (Pcp4) was downregulated on day 12. Significant changes in Asporin1 (Aspn), Syndecan 2 (Sdc2), mast cell protease (Mcp7), and decay-accelerating factor (Daf1) mRNA were not detected by QRT-PCR. Expression values are displayed in logarithmic scale and represent mean±S.E.M. of six cervices at each time point. *P<0.05 when compared with day 10.

View larger version (15K): [in this window] [in a new window]   Figure 6 Expression of purkinje cell protein 4 (Pcp4, upper panel), Trefoil factor 1 (Tff1, middle panel), and Kazal-type serine protease inhibitor 5 (Spink5, lower panel) in cervical tissues from nonpregnant (NP) and pregnant mice. Cervices were obtained from nonpregnant (NP) and pregnant mice throughout gestation and the morning after parturition (10hPP). Expression of Pcp4, Tff1, and Spink5 was quantified using RT-PCR and normalized to that of cyclophilin B (Ppib) as described in Materials and Methods. Data represent mean±S.E.M. of an average of three cervices per time point with the exception of days 10 and 12 in which data are estimated from nine cervices per time point. P<0.001 by ANOVA Kruskall–Wallis test day 8 through 10hPP for all three genes analyzed.

Both Tff1 and Kazal-type serine peptidase inhibitor 5 (Spink5) were found to be expressed in cervix and upregulated on day 12 of pregnancy (Fig. 5). In addition, both genes were found to be highly upregulated with advancing gestation (Fig. 6, middle and lower panel). Tff1 mRNA levels were low in cervical tissues from nonpregnant and pregnant animals on days 8 and 9. During cervical softening, expression of Tff1 increased 2.8-fold from days 10 to 12 (from 2.6±0.6 to 7.3±1.4 relative units, P=0.007). Elevations in Tff1 during cervical softening was followed by a more dramatic increase in Tff1 mRNA prior to cervical ripening at term, such that Tff1 mRNA levels on late-gestation day 18 and after birth were over 1000-fold that in cervical tissues from nonpregnant or early pregnant animals (Fig. 6, middle panel).

Expression of Spink5 was also regulated in the cervix during softening and ripening with a gestational age-dependent pattern of expression similar to Tff1. Cervical mRNA levels of Spink5 increased exponentially during cervical softening, throughout parturition and the post partum period. Expression of Spink5 during cervical softening was increased 2.1-fold on day 12 versus day 10 (0.7±0.02 vs 0.15±0.04 relative units, P=0.05) (Fig. 4). Compared with early gestation (day 8), expression of Spink5 increased to greater than 650-fold after birth (PP; Fig. 6, lower panel).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Acknowledgements References

 
Molecular events regulating cervical softening are distinct from cervical ripening or dilation at term
Studies outlined in this work suggest that the molecular processes of cervical softening differ significantly from ripening. Figure 7 illustrates some of these differences. Increases in tissue proliferation, hydration or distribution of inflammatory cells in cervical stromal matrix were not evident during cervical softening. Increases in collagen solubility appear to precede biomechanical evidence of softening, indicating that this is one of the earlier events in the remodeling process. Further studies are required to determine the steps in collagen synthesis, cross-linking and fibril assembly that may be altered in early pregnancy, thus explaining the observed changes in our collagen solubility data. Changes in solubility are not predicted to result from increased collagen degradation as the overall cervical collagen content did not decline significantly during softening or ripening and was actually increased after parturition (Fig. 3). Previous studies are inconsistent with our findings as they report small increases in collagen content in mouse cervices prior to parturition; however, no statistical analysis was carried out (Rimmer 1973). Similar to previous studies, collagen concentration was greatest early in pregnancy (day 7) and lowest during cervical ripening (day 18) and 10 h post partum (Rimmer 1973). This suggests that the processes leading to increased collagen solubility precede the molecular events that control cervical tissue compliance during cervical softening and ripening (Fig. 1).

View larger version (21K): [in this window] [in a new window]   Figure 7 Representative schematic of the phases of mouse pregnancy. The changes in cervical physiology during mouse gestation can be divided into distinct yet overlapping phases: softening, ripening/dilation, and post partum repair. The ripening and dilation phases are described together as it is still not clear what processes may be distinct to these two phases. Processes described under post partum repair are those that occur in the immediate post partum period (within 24 h post partum).

In contrast to our observations with cervical softening, cervical ripening at the end of pregnancy is characterized by increased cell proliferation as well as reduced apoptosis of both the epithelia and the stroma which result in an increased circumference of the cervical lumen (Lee & Sherwood 2005, Lee et al. 2005). Tissue water content is also increased presumably due to the increased expression of aquaporin water channel proteins (Anderson et al. 2006) and hydrophilic glycosaminoglycans (ElMaradny et al. 1997, Straach et al. 2005). Increases in the glycosaminoglycan, hyaluronan, result in disruption of the collagen matrix which, along with tissue growth, facilitates cervical ripening (ElMaradny et al. 1997, Straach et al. 2005). These differences between softening and ripening which are illustrated in Fig. 7, may explain why in the ripened cervix there is loss of structural integrity, tensile strength, as well as an increase in dilation in response to uterine contractions of labor, yet during softening, the cervix retains its structural integrity, tensile strength, and is resistant to forces exerted upon it.

Distribution of inflammatory cells during cervical softening
Inflammatory cells and cytokines are increased during cervical dilation of labor in all mammalian species studied (Bokstrom et al. 1997, Thomson et al. 1999, Sennstrom et al. 2000, Kelly 2002). The role of inflammatory cells in ripening before dilation; however, is not clear in women or mice (Sakamoto et al. 2004, Sakamoto et al. 2005, Timmons & Mahendroo 2006). In the mouse, cervical ripening (Phase 2) is characterized by a change in distribution of inflammatory cells with increased staining of monocytes and/or neutrophils within the cervical stroma (Timmons & Mahendroo 2006). Studies conducted in cervical tissues from women and mice before and after cervical ripening (but before labor) failed to find significant increases in the number of granulocytes in the ripened cervix before labor (Sakamoto et al. 2004, 2005, Timmons & Mahendroo 2006). Macrophage numbers are reported to increase in women and one study in mice during ripening but the antibodies used may also recognize tissue monocytes thus a functional role of these cells in ripening remains unclear (Mackler et al. 1999, Sakamoto et al. 2004).

In contrast to the later phases of cervical remodeling, midtrimester cervical softening phase was associated with little change in the distribution of macrophages or neutrophils. In addition, relatively few differentially expressed genes were identified that are involved with activation of the immune system by gene microarray comparisons. One exception was the expression of Cxcl5, a neutrophil chemoattractant made by epithelial cells (Walz et al. 1991, Imaizumi et al. 2003). Cxcl5 was expressed throughout gestation (data not shown) and is consistent with the fact that neutrophils are confined to the epithelial region at both gestation days 10 and 12 as well as later in pregnancy (Timmons & Mahendroo 2006). We hypothesize that the localization of neutrophils in the epithelial region (Fig. 4, upper panel) helps to ensure that numerous surveillance mechanisms are in place during pregnancy to prevent ascending infections and protect the vulnerable, remodeling stromal matrix. Future studies to determine activation of inflammatory cells in a functional assay or measure cytokine levels directly will be required to verify that major inflammatory events do not mediate the softening process.

Expression of epithelial-specific genes during cervical softening
We initially hypothesized that changes in expression of genes affecting structure of collagen or elastin fibers would be identified in cervical tissues before and after cervical softening. This hypothesis is supported by i) decreased activity of the cross-linking enzyme, lysyl oxidase, in the mouse cervix during pregnancy, ii) changes in proteoglycan composition in the human and rat cervix during softening, and iii) the finding that mutations in the ECM proteins fibrillin-1 and thrombospondin 2 lead to premature cervical softening in women and mice respectively (Ozasa et al. 1981, Osmers et al. 1993, Kyriakides et al. 1998, Rahman et al. 2003, Kokenyesi et al. 2004). Contrary to our expectations, however, we were unable to detect significant changes in expression of ECM genes in the cervix from days 10 to 12. There are several possible explanations for these results. First, regulation of ECM proteins may occur at the post-transcriptional level that would not be reflected by changes in transcript expression using microarrays or RT-PCR. This possibility is suggested by studies that report no decrease in transcripts for small proteoglycans such as decorin, biglycan, and fibromodulin in the cervix from women at term despite a 50% reduction of total proteoglycan concentration (Westergren-Thorsson et al. 1998). Second, cervical epithelial cells undergo hyperplasia with advancing gestation such that the ratio of epithelial to stromal cells may change and therefore bias the microarray results towards epithelial-specific genes (Burger & Sherwood 1998, Lee et al. 2005). Since, in these studies, whole cervix was used instead of fractionating the epithelial from the stromal compartment, the contribution of potentially important stromal genes may have been diluted. Finally, significant changes in mRNA levels of proteins affecting the ECM may have occurred prior to day 10. Although changes in biophysical properties of the cervix were not detected until gestation day 12, significant changes in mRNA transcripts may have preceded cervical softening by several days. The fact that increases in collagen solubility (day 7) preceded changes in tissue compliance (day 12) further supports this idea. Studies are in progress to determine the biochemical and molecular changes that bring about increases in collagen solubility during early pregnancy.

Nevertheless, although genes encoding ECM molecules were not found to be differentially regulated during cervical softening, increased expression of several epithelial-specific genes were identified in the softened cervix on gestation day 12. The genes we identified are important in maintaining the integrity and barrier function of epithelia in the gut or skin. Tff1 and Spink5 are two genes that were significantly upregulated during cervical softening, and their expression progressively increased to peak at term and post partum. Based on their functions in gut and skin respectively, we propose that these transcripts encode proteins that may protect the cervix from infection and maintain the integrity and barrier function of the epithelia. Both of these genes are reported to be expressed in the female reproductive tract but have not been carefully explored with respect to female reproductive tract physiology (Wiede et al. 2001, Masui et al. 2006). The Tff1 gene encodes the estrogen-responsive protein, pS2, which is expressed in the gastric mucosa of humans and mice and involved in protection and restitution of mucosal epithelia (Shaoul et al. 2004, Hoffmann 2005). Tff1-deficient mice (Tff1^–/–) develop multiple antral gastric tumors from the loss of mucosal protection afforded by the pS2 protein (Lefebvre et al. 1996).

Spink5 is expressed in the differentiated layers of stratified epithelial tissue, and mutations in this gene have been shown to be the cause of autosomal-recessive Netherton syndrome (Netherton, 1958). Patients with this condition suffer from frequent bacterial infections. Spink5^–/– mice die a few hours after birth from dehydration due to the lack of an intact permeability barrier of skin (Hewett et al. 2005). The gene product, lymphoepithelial Kazal-type related inhibitor, is a serine protease inhibitor that attenuates the functions of stratum corneum tryptic enzymes, which degrade desmoglein 1 as well as other proteins involved in barrier formation and maintenance (Descargues et al. 2005).

The transcriptional upregulation of Spink5 and Tff1 during cervical softening and continuing through parturition suggests that the cervical epithelia may play a greater role in the process of parturition than was previously recognized. Changes in stromal matrix must orchestrate, and are integral to, the change in biomechanical properties of the cervix. In addition, epithelial cells together with neutrophils lining the cervical os, may serve an immunomucosal protective role to prevent inappropriate access of invading microorganisms to the stromal matrix, which may be in a vulnerable, ‘softened’ state as well as prevent ascending infections into the upper reproductive tract. As the cervix loses structural integrity during ripening, parturition and post partum recovery of the cervix, the need for epithelial barrier protection potentially becomes even greater. Further studies on the role of Spink5 and Tff1 in cervical remodeling may enhance our understanding of molecular events that control parturition and determine if alterations in expression or function of these epithelial-specific genes could result in accelerated remodeling of the cervical ECM, thereby leading to the complications of cervical insufficiency and preterm birth.

	Acknowledgements

Top Abstract Introduction Materials and Methods Results Discussion Acknowledgements References

 
The authors wish to thank Jesús Acevedo and Joseph Davis for their excellent technical assistance. The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work. This work is supported in part by NIH R01 HD043154 (M S Mahendroo) and P01-HD111149 (R A Word).

	Footnotes

 
Received 18 January 2007
First decision 26 February 2007
Revised manuscript received 18 April 2007
Accepted 27 April 2007

	References

Top Abstract Introduction Materials and Methods Results Discussion Acknowledgements References

 

Ananth CV, Joseph KS, Oyelese Y, Demissie K & Vintzileos AM 2005 Trends in preterm birth and perinatal mortality among singletons: United States, 1989 through 2000. Obstetrics and Gynecology 105 1084–1091.[Abstract/Free Full Text]

Anderson J, Brown N, Mahendroo MS & Reese J 2006 Utilization of different aquaporin water channels in the mouse cervix during pregnancy and parturition and in models of preterm and delayed cervical ripening. Endocrinology 147 130–140.[Abstract/Free Full Text]

Bokstrom H, Brannstrom M, Alexandersson M & Norstrom A 1997 Leukocyte subpopulations in the human uterine cervical stroma at early and term pregnancy. Human Reproduction 12 586–590.[Abstract/Free Full Text]

Branum AM & Schoendorf KC 2002 Changing patterns of low birthweight and preterm birth in the United States, 1981–1998. Paediatric and Perinatal Epidemiology 16 8–15.[CrossRef][ISI][Medline]

Buhimschi IA, Dussably L, Buhimschi CS, Ahmed A & Weiner CP 2004 Physical and biomechanical characteristics of rat cervical ripening are not consistent with increased collagenase activity. American Journal of Obstetrics and Gynecology 191 1695–1704.[CrossRef][ISI][Medline]

Burger LL & Sherwood OD 1998 Relaxin increases the accumulation of new epithelial and stromal cells in the rat cervix during the second half of pregnancy. Endocrinology 139 3984–3995.[Abstract/Free Full Text]

Danforth DN, Veis A, Breen M, Weinstein HG, Buckingham JC & Manalo P 1974 The effect of pregnancy and labor on the human cervix: changes in collagen, glycoproteins, and glycosaminoglycans. American Journal of Obstetrics and Gynecology 120 641–651.[ISI][Medline]

Descargues P, Deraison C, Bonnart C, Kreft M, Kishibe M, Ishida-Yamamoto A, Elias P, Barrandon Y, Zambruno G, Sonnenberg A et al. 2005 Spink5-deficient mice mimic Netherton syndrome through degradation of desmoglein 1 by epidermal protease hyperactivity. Nature Genetics 37 56–65.[CrossRef][ISI][Medline]

Drzewiecki G, Tozzi C, Yu SY & Leppert PC 2005 A dual mechanism of biomechanical change in rat cervix in gestation and postpartum: applied vascular mechanics. Cardiovascular Engineering V5 187–193.[CrossRef]

ElMaradny E, Kanayama N, Kobayashi H, Hossain B, Khatun S, She LP, Kobayashi T & Terao T 1997 The role of hyaluronic acid as a mediator and regulator of cervical ripening. Human Reproduction 12 1080–1088.[Abstract/Free Full Text]

Erhardt JA, Legos JJ, Johanson RA, Slemmon JR & Wang X 2000 Expression of PEP-19 inhibits apoptosis in PC12 cells. Neuroreport 11 3719–3723.[ISI][Medline]

Funaba M, Ikeda T, Murakami M, Ogawa K, Nishino Y, Tsuchida K, Sugino H & Abe M 2006 Transcriptional regulation of mouse mast cell protease-7 by TGF-ß. Biochimica et Biophysica Acta 1759 166–170.[Medline]

Granstrom L, Ekman G, Ulmsten U & Malmstrom A 1989 Changes in the connective tissue of corpus and cervix uteri during ripening and labour in term pregnancy. British Journal of Obstetrics and Gynaecology 96 1198–1202.[ISI][Medline]

Harkness ML & Harkness RD 1959 Changes in the physical properties of the uterine cervix of the rat during pregnancy. Journal of Physiology 148 524–547.[Free Full Text]

Hegar AEL 1895 Diagnose der frühesten Schwangerschaftsperiode. Deutsche Medizinische Wochenschrift 21 565–567.

Henderson RB, Hobbs JA, Mathies M & Hogg N 2003 Rapid recruitment of inflammatory monocytes is independent of neutrophil migration. Blood 102 328–335.[Abstract/Free Full Text]

Hewett DR, Simons AL, Mangan NE, Jolin HE, Green SM, Fallon PG & McKenzie AN 2005 Lethal, neonatal ichthyosis with increased proteolytic processing of filaggrin in a mouse model of Netherton syndrome. Human Molecular Genetics 14 335–346.[Abstract/Free Full Text]

Hillier K & Wallis RM 1982 Collagen solubility and tensile properties of the rat uterine cervix in late pregnancy: effects of arachidonic acid and prostaglandin F 2 . Journal of Endocrinology 95 341–347.[Abstract]

Hoffmann W 2005 Trefoil factors TFF (trefoil factor family) peptide-triggered signals promoting mucosal restitution. Cellular and Molecular Life Sciences 62 2932–2938.[CrossRef][ISI][Medline]

Horgan DJ, King NL, Kurth LB & Kuypers R 1990 Collagen crosslinks and their relationship to the thermal properties of calf tendons. Archives of Biochemistry and Biophysics 281 21–26.[CrossRef][ISI][Medline]

Imaizumi T, Kumagai M, Hatakeyama M, Tamo W, Yamashita K, Yoshida H, Munakata H & Satoh K 2003 Effect of 15-deoxy-12,14-prostaglandin J2 on IL-1beta-induced expression of epithelial neutrophil-activating protein-78 in human endothelial cells. Prostaglandins Leukotrienes and Essential Fatty Acids 69 323–327.[CrossRef][ISI][Medline]

Kanamori T, Takakura K, Mandai M, Kariya M, Fukuhara K, Kusakari T, Momma C, Shime H, Yagi H, Konishi M et al. 2003 PEP-19 overexpression in human uterine leiomyoma. Molecular Human Reproduction 9 709–717.[Abstract/Free Full Text]

Kelly RW 2002 Inflammatory mediators and cervical ripening. Journal of Reproductive Immunology 57 217–224.[CrossRef][ISI][Medline]

Kokenyesi R, Armstrong LC, Agah A, Artal R & Bornstein P 2004 Thrombospondin 2 deficiency in pregnant mice results in premature softening of the uterine cervix. Biology of Reproduction 70 385–390.[Abstract/Free Full Text]

Kyriakides TR, Zhu YH, Smith LT, Bain SD, Yang Z, Lin MT, Danielson KG, Iozzo RV, LaMarca M, McKinney CE et al. 1998 Mice that lack thrombospondin 2 display connective tissue abnormalities that are associated with disordered collagen fibrillo-genesis, an increased vascular density, and a bleeding diathesis. Journal of Cell Biology 140 419–430.[Abstract/Free Full Text]

Lee HY & Sherwood OD 2005 The effects of blocking the actions of estrogen and progesterone on the rates of proliferation and apoptosis of cervical epithelial and stromal cells during the second half of pregnancy in rats. Biology of Reproduction 73 790–797.[Abstract/Free Full Text]

Lee HY, Zhao S, Fields PA & Sherwood OD 2005 The extent to which relaxin promotes proliferation and inhibits apoptosis of cervical epithelial and stromal cells is greatest during late pregnancy in rats. Endocrinology 146 511–518.[Abstract/Free Full Text]

Lefebvre O, Chenard MP, Masson R, Linares J, Dierich A, LeMeur M, Wendling C, Tomasetto C, Chambon P & Rio MC 1996 Gastric mucosa abnormalities and tumorigenesis in mice lacking the pS2 trefoil protein. Science 274 259–262.[Abstract/Free Full Text]

Leppert PC 1995 Anatomy and Physiology of Cervical Ripening. Clinical Obstetrics and Gynecology 38 267–279.[CrossRef][ISI][Medline]

Leppert PC & Yu SY 1994 Apoptosis in the cervix of pregnant rats in association with cervical softening. Gynecologic and Obstetric Investigation 37 150–154.[CrossRef][ISI][Medline]

Leppi TJ 1964 A study of the uterine cervix of the mouse. Anatomical Record 150 51–65.[CrossRef][Medline]

Liggins GC 1978 Ripening of the cervix. Seminars in Perinatology 2 261–271.[ISI][Medline]

Lin F, Fukuoka Y, Spicer A, Ohta R, Okada N, Harris CL, Emancipator SN & Medof ME 2001 Tissue distribution of products of the mouse decay-accelerating factor (DAF) genes. Exploitation of a Daf1 knock-out mouse and site-specific monoclonal antibodies. Immunology 104 215–225.[CrossRef][ISI][Medline]

Mackler AM, Iezza G, Akin MR, McMillan P & Yellon SM 1999 Macrophage trafficking in the uterus and cervix precedes parturition in the mouse. Biology of Reproduction 61 879–883.[Abstract/Free Full Text]

Masui F, Kurosaki K, Mori T & Matsuda M 2006 Persistent trefoil factor 1 expression imprinted on mouse vaginal epithelium by neonatal estrogenization. Cell and Tissue Research 323 167–175.[CrossRef][ISI][Medline]

Netherton EW 1958 A unique case of trichorrhexis nodosa; bamboo hairs. A. M. A. Archives of Dermatology 78 483–487.[Medline]

Osmers R, Rath W, Pflanz MA, Kuhn W, Stuhlsatz HW & Szeverenyi M 1993 Glycosaminoglycans in cervical connective tissue during pregnancy and parturition. Obstetrics and Gynecology 81 88–92.[Abstract/Free Full Text]

Ozasa H, Tominaga T, Nishimura T & Takeda T 1981 Lysyl oxidase activity in the mouse uterine cervix is physiologically regulated by estrogen. Endocrinology 109 618–621.[Abstract]

Papiernik E, Bouyer J, Collin D, Winisdoerffer G & Dreyfus J 1986 Precocious cervical ripening and preterm labor. Obstetrics and Gynecology 67 238–242.[ISI][Medline]

Paternoster DM, Santarossa C, Vettore N, Dalla Pria S & Grella P 1998 Obstetric complications in Marfan’s syndrome pregnancy. Minerva Ginecologica 50 441–443.[Medline]

Rahman J, Rahman FZ, Rahman W, al-Suleiman SA & Rahman MS 2003 Obstetric and gynecologic complications in women with Marfan syndrome. Journal of Reproductive Medicine 48 723–728.[ISI][Medline]

Ramos JG, Varayoud J, Bosquiazzo VL, Luque EH & Munoz-de-Toro M 2002 Cellular turnover in the rat uterine cervix and its relationship to estrogen and progesterone receptor dynamics. Biology of Reproduction 67 735–742.[Abstract/Free Full Text]

Reiner A, Yekutieli D & Benjamini Y 2003 Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics 19 368–375.[Abstract/Free Full Text]

Rimmer DM 1973 The effect of pregnancy on the collagen of the uterine cervix of the mouse. Journal of Endocrinology 57 413–418.[ISI][Medline]

Robins SP, Milne G, Duncan A, Davies C, Butt R, Greiling D & James IT 2003 Increased skin collagen extractability and proportions of collagen type III are not normalized after 6 months healing of human excisional wounds. Journal of Investigative Dermatology 121 267–272.[CrossRef][ISI][Medline]

Sakamoto Y, Moran P, Searle RF, Bulmer JN & Robson SC 2004 Interleukin-8 is involved in cervical dilatation but not in prelabour cervical ripening. Clinical and Experimental Immunology 138 151–157.[CrossRef][ISI][Medline]

Sakamoto Y, Moran P, Bulmer JN, Searle RF & Robson SC 2005 Macrophages and not granulocytes are involved in cervical ripening. Journal of Reproductive Immunology 66 161–173.[CrossRef][ISI][Medline]

Sennstrom MB, Ekman G, Westergren-Thorsson G, Malmstrom A, Bystrom B, Endresen U, Mlambo N, Norman M, Stabi B & Brauner A 2000 Human cervical ripening, an inflammatory process mediated by cytokines. Molecular Human Reproduction 6 375–381.[Abstract/Free Full Text]

Shaoul R, Okada Y, Cutz E & Marcon MA 2004 Colonic expression of MUC2, MUC5AC, and TFF1 in inflammatory bowel disease in children. Journal of Pediatric Gastroenterology and Nutrition 38 488–493.[ISI][Medline]

Singh UP, Singh S, Boyaka PN, McGhee JR & Lillard JW Jr 2004 Granulocyte chemotactic protein-2 mediates adaptive immunity in part through IL-8Rß interactions. Journal of Leukocyte Biology 76 1240–1247.[Abstract/Free Full Text]

Stegemann H & Stalder K 1967 Determination of hydroxyproline. Clinica Chimica Acta 18 267–273.[CrossRef][ISI][Medline]

Straach KJ, Shelton JM, Richardson JA, Hascall VC & Mahendroo MS 2005 Regulation of hyaluronan expression during cervical ripening. Glycobiology 15 55–65.[Abstract/Free Full Text]

Taylor PR, Brown GD, Geldhof AB, Martinez-Pomares L & Gordon S 2003 Pattern recognition receptors and differentiation antigens define murine myeloid cell heterogeneity ex vivo. European Journal of Immunology 33 2090–2097.[CrossRef][ISI][Medline]

Thomson AJ, Telfer JF, Young A, Campbell S, Stewart CJ, Cameron IT, Greer IA & Norman JE 1999 Leukocytes infiltrate the myometrium during human parturition: further evidence that labour is an inflammatory process. Human Reproduction 14 229–236.[Medline]

Timmons BC & Mahendroo MS 2006 Timing of neutrophil activation and expression of proinflammatory markers do not support a role for neutrophils in cervical ripening in the mouse. Biology of Reproduction 74 236–245.[Abstract/Free Full Text]

Uldbjerg N, Ekman G, Malmstrom A, Olsson K & Ulmsten U 1983 Ripening of the human uterine cervix related to changes in collagen, glycosaminoglycans, and collagenolytic activity. American Journal of Obstetrics and Gynecology 147