Cervical ripening consists of radical changes in the shape and consistency of the cervix before labor. The cervix is essentially a connective tissue structure comprising greater than 85% fibrous connective tissue.1 Prelabor cervical ripening was associated with a reduction in collagen content and a rise in the activity of enzymes such as collagenases and neutrophil elastase and involves a process in which cellular number decreases.2,3 The exact mechanism of cervical ripening is currently unknown. Failure of the cervix to ripen may result in prolonged pregnancy; conversely, premature ripening may lead to early labor or incompetent cervix and preterm birth.
Apoptosis, together with mitosis, controls the number of cells in a given tissue. Apoptosis has been described extensively and occurs in a number of physiologic processes.4,5 Examples include the involution of the Wolffian and müllerian ducts in embryogenesis.6 Apoptosis differs from necrosis in that it is an active form of cell death dependent on the internal machinery of the cell. Necrosis is an accidental or unplanned death caused by factors external to the cell.4,5
Tissue changes of the uterine cervix before labor was postulated to be the result of the proliferation of fibroblasts and smooth muscle cells followed by active or programmed cell death.7,8 Leppert and Yu showed that the numbers of dying smooth muscle cells in the cervix increased with cervical softening in pregnant rats between days 12 and 21 of gestation. The morphologic characteristics of the chromatin cleavage in apoptosis were identified.8 In a subsequent study, Leppert found that the incidence of apoptosis cells in rat cervix increased throughout gestation.9 Onapristone, an antiprogesterone, was found to significantly inhibit apoptosis. That study suggested that active cell death played a role in the initiation of parturition. Leppert hypothesized that apoptosis of cervical stroma cells initiated biochemical pathways in the cervix resulting in a perturbation of collagen structure and changes in proteoglycan composition. Apoptosis throughout gestation and its relation to cervical ripening has not been described in the human cervix. The purpose of this study was to determine whether remodeling of the lower uterine segment in the latter half of human pregnancy and labor is associated with changes in the percentage of cervical stroma cells undergoing apoptosis compared with non-pregnant control women.
Materials and Methods
Between July 1999 and March 2000, we collected lower uterine segment tissue samples immediately after cesarean delivery from a convenience sample of 11 women in spontaneous active labor, 13 women before labor, and ten nonpregnant premenopausal women undergoing hysterectomy for benign disease. Tissue collection was approved by the committee for the rights of human subjects at the University of North Carolina at Chapel Hill. Spontaneous active labor was defined as 2 cm or greater cervical dilation with at least 80% effacement if less than 4 cm dilated and at least three regular uterine contractions in 10 minutes. No labor was defined as no cervical dilation with less than 50% cervical effacement and no regular uterine contractions. Pregnant subjects were excluded from eligibility if evidence of clinical chorioamnionitis existed. Nonpregnant subjects were excluded if clinical or pathologic evidence of pelvic infection or malignancy was noted. After informed consent and immediately after infant delivery or hysterectomy, a 1-cm3 sample of lower uterine segment was obtained from the anterior aspect of the uterus and immediately snap-frozen in liquid nitrogen and stored at −80C. Two modalities were used to identify apoptotic cells within the cervical stroma: terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate marker nick-end labeling staining, which identifies the fragmented DNA associated with apoptosis, and ligase-mediated polymerase chain reaction (PCR), which identifies apoptosis by amplifying specific kilobase nuclear DNA fragments.
Terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate marker nick end-labeling staining was performed on 6-μm cryosections using the Apop Tag In-Situ Apoptosis Detection kit (Intergen Co., Gaithersburg, MD). Briefly, tissue cryosections were fixed in formalin. Endogenous peroxidases were quenched with 3% hydrogen peroxide in 100% methanol. Residues of digoxigenin nucleotide were catalytically added to the DNA by terminal deoxynucleotidyl transferase. Antidigoxigenin antibody peroxidase conjugate was then applied. Filtered 0.05% diaminobenzidine (Sigma Chemical Co., St. Louis, MO) with 0.02% hydrogen peroxide was then applied to the sample. The tissue was then counterstained with toluidine blue and the slides were examined by light microscopy. Cryosections of postweaning mouse mammary gland were used as positive controls. A negative control for each section was made by substituting distilled water for terminal deoxynucleotidyl transferase. Digital images of ten randomly selected high-power fields (approximately 770× final magnification) in each sample were obtained using a Nikon Microphot FXA microscope (Nikon, Tokyo, Japan) and Scion Image software (Scion, Frederick, MD). Apoptotic nuclei, differentiated from nonapoptotic nuclei by their brown labeling, were counted. The total number of nuclei and the number of apoptotic nuclei in ten high-power fields for each sample were determined manually using Scion Image software. The percentage of apoptotic nuclei ([number apoptotic nuclei/total number nuclei] × 100) was calculated for each sample. Only those nuclei with brown staining and morphologic criteria of apoptosis were considered apoptotic. Apoptotic nuclei can be identified by chromatin condensation resulting in a nuclear appearance of single or multiple dark bodies. Areas of necrosis and inflammation were avoided in the analysis. Investigators were blinded to study group during analysis.
To confirm the presence of apoptosis, nucleosomal ladders in apoptotic cells were detected using the ApopAlert LM-PCR Ladder Assay Kit (Clontech, Palo Alto, CA). Genomic DNA was isolated from cryopreserved tissue using the protocol described by Ausubel et al.10 Ligation-mediated PCR was used to detect nucleosomal ladders in aptotic cells as described by Staley et al.11 Dephosphorylated adaptors were ligated to the DNA fragments generated during apoptosis. The adapter oligonucleotides annealed to the 5′-phosphorylated blunt ends, creating a primer for PCR. Twenty-five and 28 cycles of PCR were performed for each sample. Electrophoresis of each reaction mixture was then performed on 1.2% agarose and ethidium bromide. Purified genomic DNA from calf thymus was used as a positive control for each reaction.
To detect a 10% difference in the percentage of apoptotic cells per subject between study groups assuming a power of 0.90, an alpha of .05, and a standard deviation (SD) of 5%, approximately ten subjects per group were needed (assuming a similar percentage of apoptotic cells and SD seen in prelabor rat cervices).9 Statistical analysis was done using the Kruskal–Wallis test or rank-sum test for continuous data and the Fisher exact test for categorical data. A P value of .05 was considered significant except when multiple pairwise comparisons were made between the three study groups using the Krukal–Wallis test. To control for multiple comparisons, a P value of .02 was considered statistically significant (P = .05/k, where k = the number of pairwise comparisons).12
We collected lower uterine segment samples from 11 women in labor, 13 women not in labor, and ten nonpregnant women. Demographic and pregnancy characteristics of each group are presented in Table 1. A mean of 1805 cells was counted in each sample. Apoptotic nulcei were identified in the cervical stroma tissue by their intense brown labeling (Figure 1). The median percentage of apoptotic nuclei was 0.7 (inter-quartile range 0.4, 1.4) for the nonpregnant group, 7.5 (interquartile range 6.6, 11.2) for the pregnant nonlaboring group, and 11.6 (interquartile range 8.3, 16.7) for the pregnant laboring group (P < .001). The percentage of apoptotic nuclei differed significantly across the three study groups: nonpregnant compared with pregnant laboring, P < .001; nonpregnant compared with pregnant nonlaboring, P < .001; and pregnant laboring compared with pregnant nonlaboring, P = .006.
Electrophoresis of lower uterine segment extracts showed evidence of nucleosomal ladders of approximately 200-bp fragments, confirming the presence of apoptosis in the lower uterine segment during the latter half of pregnancy. At 25 PCR cycles, all 11 samples from laboring pregnant women, five of 13 samples from nonlaboring pregnant women, and none of ten samples from nonpregnant women showed DNA fragmentation characteristic of apoptosis (P < .001). Figure 2 shows the ligation-mediated PCR assay of cervical stroma tissue from nonpregnant, nonlaboring pregnant, and laboring pregnant women (25 PCR cycles). Bands forming nucleosomal ladders at approximately 200-bp intervals can be seen in lanes 1 and 3–6. At 28 PCR cycles, seven of 13 samples from nonlaboring pregnant women and none of ten samples from nonpregnant women showed nucleosomal ladders (P = .007).
We found that human pregnancy is associated with an increase in lower uterine segment stroma cell apoptosis compared with nonpregnant control women. Laboring pregnant women were also found to have an increased percentage of stroma cells undergoing apoptosis compared with nonlaboring women. Our findings support studies in rodents suggesting that cervical stroma cell apoptosis increases with gestation. Although the molecular basis of this association is unknown, we hypothesize that maternal or fetal factors that signal the onset of labor stimulate the cervical stroma to undergo apoptosis. Other studies showed that cellular apoptosis might be hormonally influenced. The antiprogesterone onapristone increased apoptosis in the rat.9
Our study is limited by several factors. Specimens from pregnant women were obtained during cesarean delivery, and the biopsy sites may not have represented the functional internal cervical os. Although we attempted biopsy at the point of the anatomic cervical os in pregnant and nonpregnant uteri, different areas of the lower uterus might have been sampled. We did not differentiate stroma cells into fibroblast and smooth muscle cells during histologic analysis. Cervical stroma cells consist of more than 80% fibroblast cells, and Leppert9 showed a similar percentage of apoptotic nuclei in smooth muscle and fibroblast cells, so we did not examine the proportion. The use of terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate marker nick-end labeling staining was criticized because of the presence of false-positive staining with necrosis. To avoid this possible error, we specifically avoided areas of necrosis, if present. We confirmed our findings by the presence of DNA laddering after ligase-mediated PCR seen in specimens from pregnant compared with control women. These DNA fragments seen on gel electrophoresis are widely recognized hallmarks of apoptosis.11 The primary indication for cesarean delivery in our laboring patients was arrest of active phase of labor. Abnormal rather than normal progress of labor or associated subclinical chorioamnionitis might be associated with an increased incidence of cervical stroma cell apoptosis. However, there is a progression of the incidence of cervical stroma apoptosis between nonpregnant, pregnant unlabored, and pregnant labored women, suggesting that pregnancy affects cellular apoptosis, which is then magnified by the process of labor. Understanding the relation between dysfunctional labor and cervical stroma apoptosis requires a study of the cervix in women who labor and deliver vaginally.
Our data suggest that cervical stroma cell apoptosis is a cause or an effect of cervical remodeling that occurs during dilatation and effacement in pregnancy. Further knowledge of the significance of stromal cell apoptosis might be exploited to treat disorders of premature cervical dilatation and effacement or encourage cervical ripening when indicated.
1. Danforth DN. The fibrous nature of the human cervix, and its relation to the isthmic segment in gravid and nongravid uteri. Am J Obstet Gynecol 1947;53:541–60.
2. Danforth DN, Veis A, Breen M, Weinstein HG, Buckingham JC, Manalo P. The effect of pregnancy and labor on the human cervix: Changes in collagen, glycoproteins, and glycosaminglycans. Am J Obstet Gynecol 1974;1:641–51.
3. Uldbjerg N, Ulmsten U, Ekman G. The ripening of the human cervix in terms of connective tissue biochemistry. Clin Obstet Gynecol 1982;26:14–26.
4. Kerr IFR, Harmon BV. Definition and incidence of apoptosis: An historical perspective. In: Tomei LD, Cope FO, eds. Apoptosis: The molecular basis of programmed cell death. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1991:5–29.
5. Sayill J. Review: Apoptosis in disease. Eur J Clin Invest 1994;24:715–23.
6. Wylie AH, Duvall E. Cell injury and death. In: McGee JD, Isaacson PG, Wroghjt NA, eds. Oxford textbook of pathology. Volume 1. Principles of pathology. Oxford: Oxford University Press, 1992:141–7.
7. Yu SY, Tozzi CA, Babiarz J, Leppert PC. Collagen changes in rat cervix in pregnancy—polarized light microscopic and electron microscopic studies. Proc Soc Exp Biol Med 1995;209:360–8.
8. Leppert PC, Yu SY. Apoptosis in the cervix of pregnant rats in association with cervical softening. Gynecol Obstet Invest 1994;3:150–4.
9. Leppert PC. Proliferation and apoptosis of fibroblasts and smooth muscle cells in rat uterine cervix throughout gestation and the effect of the antiprogesterone onapristone. Am J Obstet Gynecol 1998;178:713–25.
10. Ausubel FM, Brent R, Kingston RE, Moore DD, Sidman JG, Smith JA, et al. Current protocols in molecular biology. New York: Greene Publishing Associates and John Wiley & Sons, 1994.
11. Staley K, Blaschke AJ, Chun J. Apoptotic DNA fragmentation is detected by a semiquantitative ligation-mediated PCR of blunt DNA ends. Cell Death Differ 1997:66–75.
© 2001 The American College of Obstetricians and Gynecologists
12. Moore DS, McCabe GP. Introduction to the practice of statistics. 3rd ed. New York: WH Freeman and Company, 1999.