SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Proposed Mechanisms of Cervical Ripening and Dilatation
  5. Discussion
  6. Conclusion
  7. References

In spite of impressive advances in biochemistry and molecular biology, it has not yet been possible to fit the individual biochemical components of cervical ripening and dilatation to a uniform clinical moiety or to uncover any regulatory mechanisms. The production of interleukin-8 by activated fibroblasts and macrophages plays a key role in cervical ripening, since this cytokine induces chemotaxis, activation, and degranulation of neutrophilic granulocytes with the consequent release of various proteases, including collagenase. In addition, the extravasation of neutrophilic granulocytes is mediated—as in the early stage of an acute inflammatory reaction—by a brief increase in adhesiveness of vascular endothelium. This is known to be modulated by the cytokine-induced increase in the expression of endothelial adhesion molecules. Furthermore, an increase in pro-inflammatory cytokine and proteinase concentrations in preterm delivery seems to occur at earlier stages of cervical dilatation than in term delivery. It is also well known that in patients with chorioamnionitis, the levels of pro-inflammatory cytokines are elevated in amniotic fluid, maternal serum, cervical secretion, placenta, and other compartments of the placento-maternal unit, and are associated with preterm uterine contractions. We have demonstrated for the first time that cytokine concentrations in the lower uterine segment in patients with chorioamnionitis are strongly elevated. We conclude from our data that increased concentrations of pro-inflammatory cytokines may also play a pivotal role in cervical softening and dilatation during chorioamniotic infection. Our data agree with the hypothesis of Liggins who stated nearly 20 years ago that cervical ripening may be an inflammatory reaction, which leads to increased prostaglandin synthesis, preterm labour and finally to preterm delivery.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Proposed Mechanisms of Cervical Ripening and Dilatation
  5. Discussion
  6. Conclusion
  7. References

On the one hand, the firm and closed uterine cervix is an essential prerequisite for an unimpaired pregnancy through to maturation of the fetus. On the other hand, physiological cervical ripening processes are essential for the timely onset of labour and a rapid and coordinated cervical dilatation is the key process for normal delivery. The fascinating metamorphosis of the cervix from an organ closing the uterus to a passage of more than 10 cm diameter for delivery of the baby is not the result of a passive, sphincter-like function of the uterine cervix, but rather of active and dynamic reorganisation processes within the extracellular matrix.

In spite of the impressive advances in biochemistry and molecular biology it has, as yet, not been possible to fit the individual biochemical components of cervical ripening and dilatation to a uniform clinical moiety or to uncover any regulatory mechanisms.

Proposed Mechanisms of Cervical Ripening and Dilatation

  1. Top of page
  2. Abstract
  3. Introduction
  4. Proposed Mechanisms of Cervical Ripening and Dilatation
  5. Discussion
  6. Conclusion
  7. References

In contrast to cervical ripening, which needs, at least from a clinical point of view, a few weeks, the rapid dilatation of the cervix at birth, which takes place within hours, requires the rapid formation and action of catabolic enzyme systems.

First indications for collagenolysis in cervical tissue during delivery were provided by the morphological investigations of Junqueira et al, who observed a fibre-free halo surrounding neutrophilic granulocytes, which migrated into the cervical stroma during parturition with subsequent degranulation1.

Later serial analyses of collagenase and elastase activities in cervical tissue and maternal serum at different stages of cervical dilatation revealed an increasing collagenase activity up to a dilatation of 6–8 cm, corresponding to an immunohistochemically detectable progressive invasion of the cervical stroma by neutrophilic granulocytes2.

At <2 cm dilatation, only a small number of neutrophils are present especially inside the capillaries. In contrast, a large number of neutrophils are detectable in the stroma when cervical dilatation progresses.

In immunohistochemical investigations with a quantitative determination of various leukocyte populations in the lower segment, we have found a significant quantitative increase in stroma invasion by granulocytes with progressive cervical dilatation3.

The results of experiments to identify fibroblast collagenase messenger ribonucleic acid (mRNA) by means of in situ hybridisation (ISH) indicated that a decisive participation of fibroblasts to the increased availability of collagenases during parturition is unlikely4.

In contrast to other investigations5, we merely found isolated eosinophilic granulocytes in the stroma of the lower uterine segment, which suggests that this leukocyte population is of minor significance for term delivery.

Interleukins

The results quoted above pose the question as to which triggering mechanisms may be responsible for these morphological findings. Without doubt, the production of interleukin-8 (IL-8) by activated fibroblasts and macrophages plays a key role since this cytokine induces chemotaxis, activation, and degranulation of neutrophilic granulocytes with the consequent release of various proteases. Further indications for the importance of IL-8 in parturition are supported by the detection of IL-8 production in human cervical tissue6, our recent observation of the IL-1β induced increase in the production of IL-8 by human lower uterine segment fibroblasts7, and the achievable cervical ripening and dilatation demonstrated in animal experiments upon local administration of IL-88. The latter process is accompanied by an invasion of neutrophilic granulocytes into the cervical stroma and an increased collagenase activity in cervical tissue9. Furthermore, investigations in humans have shown a significantly increased IL-8 concentration in the myometrium of the lower uterine segment in women with cervical dilatation of more than 3 cm and labour as compared with women with an unripe cervix and no labour. In addition, a clear correlation was found between the rise in IL-8 concentration and the increases in concentration of the granulocytic matrix metalloproteinases (MMP-8 and MMP-9). The latter analyses were performed without a differential grading of cervical dilatation10. Hence we carried out the first serial studies on the cytokine concentrations in the lower uterine segment at various stages of cervical dilatation during parturition at term and obtained the results shown in Table 1.

Table 1.  Median concentrations (25th–75th percentile) of cytokines, cell adhesion molecules, matrix metalloproteinases, and tissue inhibitor of MMP-1 in the lower uterine segment at term parturition.
Cervical dilatation (cm)<22–<44–6>6
 n= 8–12n= 8–12n= 9–12n= 7–12
  1. aP < 0.05 versus <2 cm.

  2. bP < 0.01 versus <2 cm.

  3. cP < 0.05 versus 2 to <4 cm.

  4. dP < 0.01 versus 2 to <4 cm.

Cytokines (pg/mg total protein)
Tumour necrosis factor-α8.0 (1.9–17)9.8 (0.2–679)22 (0.4–142)6.7 (1.1–194)
Interleukin-1β1.3 (0.3–4.9)2.6 (0.2–679)22a (0.4–142)13b (1.1–194)
Interleukin-613 (3.3–491)91 (4.6–7295)1226a (22–8760)1171a (170–4222)
Interleukin-817 (2.0–649)28 (2.6–11677)2081a (4.4–12450)1627b (218–6336)
Cell adhesion molecules (ng/mg total protein)
Endothelial leukocyte adhesion molecule-10.2 (0.12–0.32)0.21 (0.12–0.27)0.36 (0.29–0.44)0.22 (0.17–0.28)
Intercellular adhesion molecule-12.24 (1.49–3.04)4.49 (2.7–6.47)6.73 (3.66–7.73)4.4a (2.38–5.9)
Vascular cell adhesion molecule-15.5 (4.18–6.54)8.26 (3.66–11.1)4.39 (2.67–5.46)4.36 (3.32–6.39)
Platelet endothelial cell adhesion molecule-12.41 (1.56–3.96)3.00 (1.90–3.65)2.19 (1.42–2.43)1.49 (1.15–1.86)
Metalloproteinases and their inhibitor (ng/mg total protein)
MMP-832 (26–57)64 (41–95)97 (24–112)114a (68–269)
MMP-915 (11–25)25 (19–46)70b,c (55–245)102b,c (55–165)
Tissue inhibitor of MMP-153 (45–101)69 (60–78)251a (110–520)309b,d (106–835)

Up to a cervical dilatation of 4–6 cm, there was a significant increase in the concentrations of tumour necrosis factor-α (TNFα), IL-1β, IL-6, and IL-8; on further dilatation no further increases in cytokine tissue levels were observed.

Thus, the increasing concentration of IL-8 in the lower uterine segment at a dilatation of up to 4–6 cm also coincides, not only with an increase in granulocytic stroma invasion, but also with an increase in MMP-8 and MMP-9 concentrations up to complete cervical dilatation.

This is paralleled by an increase in tissue inhibitor of MMP-1 concentration (Table 1). Furthermore, we studied the mRNA expression for MMP-9, and their counter regulatory peptides (tissue inhibitor of MMP-1 and tissue inhibitor of MMP-2) with use of reverse transcriptase-polymerase chain reaction (TPCR)11.

In term delivery, at <2 cm dilatation, one of five specimens showed MMP-9 mRNA, at 2–<4 cm, three of five specimens, at 4–6 cm dilatation two of five specimens, and at >6 cm dilatation one of five specimens (Fig. 1). Tissue inhibitor of MMP-1 mRNA was detectable in all specimens.

image

Figure 1. Detection of MMP-9 mRNA in the lower uterine segment at term delivery: at <2 cm dilatation one of five specimens showed MMP-9 mRNA, at 2–<4 cm three of five specimens, at 4–6 cm dilatation two of five specimens, and at >6 cm dilatation one of five specimens.

Download figure to PowerPoint

We also localised mRNA by ISH. These data indicated that mRNA for MMP-9 were mainly produced by glandular epithelial cells (Fig. 2), but also by lower uterine segment fibroblasts and by macrophages seen in the stroma.

image

Figure 2. Localisation of MMP-9 mRNA in the lower uterine segment at term delivery by in situ hybridisation: MMP-9 mRNA was mainly produced by glandular epithelial cells.

Download figure to PowerPoint

The coincidence in time of the increase of the IL-1β concentration with the rise of the MMP-8 and MMP-9 concentrations is possibly of considerable significance since, in vitro, IL-1β stimulates the release of arachidonic acid from human myometrium cells12 as well as the rate of prostaglandin E2 (PGE2) synthesis by various intrauterine cell species (amnion13, chorion14, decidua15), induces collagenase and elastase production, which is down regulated by tissue inhibitors of proteases16, and thus inhibits matrix synthesis. Hence the stimulation of other, non-granulocytic MMPs, such as those from fibroblasts, by increasing IL-1β concentrations, would also be feasible.

A significant correlation between cytokine concentrations in the lower uterine segment and the duration of labour was only found for IL-1β, IL-6, and IL-8. In addition, there were significant correlations between the number of neutrophils in the stroma and MMP-9 concentrations and duration of labour, but not for MMP-8 concentrations.

One possible source of IL-8 in the lower uterine segment may be fibroblasts. As found in our investigations17, IL-1β significantly stimulates IL-8 production by fibroblasts of the lower uterine segment in humans 35-fold. This effect was time and dose dependent. The stimulation of fibroblasts with combinations of IL-1β with transforming growth factor-β and platelet derived growth factor augmented the IL-1β effect on IL-8 production dramatically. When progesterone was added, IL-8 secretion decreased by 7–34%.

Cell Adhesion Molecules

The prerequisite for extravasation of neutrophilic granulocytes is—as in the early stage of an acute inflammatory reaction—a brief increase in adhesiveness of vascular endothelium, which is known to be modulated by the cytokine induced increase in the expression of endothelial adhesion molecules18. We have examined immunohistochemically the expression of some of these adhesion molecules in tissues sample from the lower uterine segment.

Endothelial leukocyte adhesion molecule-1 (ELAM-1) (CD62E), which modulates, among other processes, the adhesion of neutrophilic granulocytes, monocytes, and various T-cell populations to activated endothelium, could not be detected on the endothelium of cervical vessels by means of a monoclonal antibody when the cervix was closed. In contrast, at cervical dilatation of more than 6 cm on average, about 80% of all endothelial cells expressed endothelial leukocyte adhesion molecule-1.

Similar results were found for vascular cell adhesion molecule-1 (VCAM-1) (CD 106), which modulates the adhesion of those leukocytes that carry integrin α4β1 (VLA-4) (e.g. monocytes, lymphocytes), as well as basophilic and eosinophilic, but not neutrophilic granulocytes18.

Intercellular adhesion molecule-1 (ICAM-1) (CD 54), which also modulates the adhesion of lymphocytes and neutrophilic granulocytes to human umbilical vein endothelial cells18, is expressed by almost all endothelial cells irrespective of the stage of cervical dilatation19. We have tried to quantify these immunohistochemical findings and the results can be summarised as follows:

  • the expression of ELAM-1 increases significantly from a cervical dilatation of 2–3 cm and reaches a maximum at dilatation of >6 cm
  • a lower, but still significant increase, in the expression of VCAM-1 is observed at a dilatation >6 cm in comparison to one of <2 cm
  • the expression of ICAM-1 is already at a maximum with an unripe cervix and does not increase further during parturition.

Quantification of adhesion molecule concentrations by enzyme immunoassay revealed that the median ICAM-1 concentrations at 2–<4 cm cervical dilatation, 4–6 cm dilatation, and >6 cm dilatation were significantly higher than at <2 cm dilatation. The maximum concentration was found at 4–6 cm (Table 1).

Furthermore, the concentration of ICAM-1 was also found to be dependent on the duration of labour: we observed a peak in the concentration at >12–24 hours duration of labour. The median concentration at >6–12 hours (4.8 ng/mg TP) and at >12–24 hours (5.9 ng/mg TP) was significantly higher than at ≤6 hours (2.5 ng/mg TP). The concentrations of ELAM-1, VCAM-1, and platelet endothelial cell adhesion molecule-1 were not found to be related to cervical dilatation or duration of labour.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Proposed Mechanisms of Cervical Ripening and Dilatation
  5. Discussion
  6. Conclusion
  7. References

We would suggest that, in the lower uterine segment, increasing concentrations of pro-inflammatory cytokines could induce a higher adhesiveness of the capillary endothelium that may cause extravasation of neutrophils.

Our biochemical results can be summarised in the model shown in Fig. 3 (which of course must remain hypothetical with respect to the individual regulatory cycles).

image

Figure 3. Hypothetical model for cervical dilatation at term.

Download figure to PowerPoint

Cervical dilatation requires the rapid activation of catabolic processes involving mainly collagen, but also other elemental matrix proteins. Above all, the triggering mechanisms are still unclear.

One promising hypothesis proposes that a decrease in the progesterone receptor concentration and thus reduced progesterone activity in the lower uterine segment may cause an increase in IL-8 production by cervical fibroblasts. It is a matter of speculation whether or not, at a cervical dilatation of 2–3 cm, a subclinical infection after ascension of pathogens or macrophages from the vagina may be a causal factor. The fact that the active phase of birth begins at this point in time provides support for this assumption. It is fairly certain that, on the one hand, the increased production of TNF-α and IL-1β, e.g. from activated macrophages, induces an increased expression of endothelial adhesion molecules with subsequent extravasation of granulocytes in the cervical stroma. While it is also possible, that the chemotaxis and degranulation of these cells in the stroma is triggered by an increase in concentration of IL-8. An increase in the concentration of hyaluronic acid at this time can be considered as a potent inducer of IL-1 and TNF-α synthesis by various leukocyte populations. At least in vitro, TNF-α and IL-1β promote the production of IL-8 from human cervical fibroblasts.

Furthermore, the increased synthesis of IL-6 by these cells stimulates PG and leukotriene synthesis. This can lead to a dilatation of cervical vessels and accordingly further promote the extravasation of leukocytes.

The proteases released after degranulation of neutrophilic granulocytes encounter an already destabilised collagenous fibre network. One of the main causes of this destabilisation, which occurs during cervical ripening and labour, is an increase in the decorin/collagen ratio.

We believe that the biochemical changes within the extracellular matrix may be similar. This hypothesis is supported by some of our results in patients with preterm delivery.

Determinations of cytokine concentrations in the lower uterine segment showed a significant increase in IL-1β, IL-6 and IL-8 concentrations: median concentrations of IL-1β, IL-6 and IL-8 at a cervical dilatation of 2–<4 cm were significantly higher compared with a dilatation of <2 cm. The concentrations of IL-6 and IL-8 were higher still at ≥4 cm cervical dilatation, whereas IL-1β concentration dropped significantly. Tumour necrosis factor-α concentration did not change.

The concentration of ICAM-1 was found to be significantly higher at a cervical dilatation of ≥4 cm compared with 2–<4 cm. Concentrations of VCAM-1 and ELAM-1 did not change.

Investigations concerning MMPs had the following results: median concentrations of MMP-8 and MMP-9 at 2–<4 cm cervical dilatation were significantly higher compared with <2 cm. The concentration of tissue inhibitor of MMP-1 was not associated with the stage of cervical dilatation. In addition, the concentration of MMPs and its inhibitors were not influenced by the duration of labour.

Furthermore, we studied mRNA expression for MMP-9, and their counter regulatory peptides (tissue inhibitor of metalloproteinases-1 and tissue inhibitor of MMP-2) using reverse TPCR. Messenger ribonucleic acid for MMP-9 was found in two of five specimens from women who delivered preterm with a cervical dilatation <2 cm, but could not be detected in women with a cervical dilatation of between 2 and 3 cm yet was present in all the samples (4/4) tested at ≥4 cm cervical dilatation.

The presence of mRNA for tissue inhibitors of MMP-1 was detectable in all preterm and term specimens irrespective of the stage of cervical dilatation.

We also localised mRNA by ISH in preterm parturition and found that, like term specimens, mRNA for MMP-9 is mainly produced by glandular epithelial cells.

Surprisingly, there was no remarkable increase in the number of neutrophils migrating into the cervical stroma with increasing cervical dilatation in preterm delivery; neutrophils were predominantly localised in the vessels. This finding is not consistent with the above mentioned increase in IL-8 and MMP-8 concentrations and needs further explanation.

Although no definite conclusions can be drawn from our preterm delivery studies, because of the small number of samples investigated, similarities are obvious between term and preterm delivery. This especially applies to the corresponding increase in cytokines, MMP-8, MMP-9 and tissue inhibitor of MMP-1 concentrations in the lower uterine segment. However, the increase in pro-inflammatory cytokine and proteinase concentrations in preterm delivery seems to occur at earlier stages of cervical dilatation than in term delivery.

In a more recent study, we compared IL-6 and IL-8 concentrations in the lower uterine segment in patients with (n= 33) and without (n= 33) chorioamnionitis. In patients with chorioamnionitis, the median IL-6 concentrations were higher (61.5 and 19.4 pg/mg protein, respectively [P < 0.01]). The same applied to the median IL-8 concentrations (162.3 and 13.4 pg/mg protein, respectively [P < 0.001]).

It is well known that in patients with chorioamnionitis, the levels of pro-inflammatory cytokines are elevated in amniotic fluid, maternal serum, cervical secretion, placenta, and other compartments of the placento-maternal unit, and are associated with preterm uterine contractions20. However, we demonstrated for the first time that cytokine concentrations in the lower uterine segment in patients with chorioamnionitis are strongly elevated.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Proposed Mechanisms of Cervical Ripening and Dilatation
  5. Discussion
  6. Conclusion
  7. References

We conclude from our data that increased concentrations of pro-inflammatory cytokines may also play a pivotal role in cervical softening and dilatation during chorioamniotic infection.

Our data agree with the hypothesis of Liggins21, who stated nearly 20 years ago that cervical ripening may be an inflammatory reaction, which leads to increased PG synthesis, preterm labour and finally to preterm delivery. It is a great challenge to the obstetrician not only to elucidate fundamental mechanisms of preterm labour, but also to develop more biochemically substantiated concepts for the prevention of preterm birth. In this context, basic studies in physiological parturition may prepare the way for a better understanding of processes associated with preterm parturition with the aim of developing more effective drugs to stop preterm labour and to prolong pregnancy until term.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Proposed Mechanisms of Cervical Ripening and Dilatation
  5. Discussion
  6. Conclusion
  7. References
  • 1
    Junqueira LC, Zugaib M, Montes GS, Toledo OM, Krisztan S, Shigihara KH. Morphologic and histochemical evidence for the occurrence of collagenolysis and for the role of neutrophilic polymorphonuclear leukocytes during cervical dilatation. Am J Obstet Gynecol 1980;138: 272281.
  • 2
    Osmers R, Rath W, Adelmann-Grill BC, et al. Origin of cervical collagenase during parturition. Am J Obstet Gynecol 1992;166: 14551460.
  • 3
    Winkler M, Fischer D-C, Ruck P, Marx T, Kaiserling E, Rath W. Interleukin-8—assoziierte Zunahme der Zahl neutrophiler Granulozyten im zervikalen Stroma als zentraler Mechanismus der Zervixeröffnung bei der Termingeburt. Geburtsh Frauenheilk 1997;57: 300302.
  • 4
    Osmers R, Rath W, Adelmann-Grill BC, et al. Origin of cervical collagenase during parturition. Am J Obstet Gynecol 1992;166: 14551460.
  • 5
    Knudsen UB. Cervical ripening. A rat model for investigation of contractile and passive biomechanical properties, with focus on antigestagens, eosinophil granulocytes and mast cells. Acta Obstet Gynecol Scand 1996;75: 8889.
  • 6
    Barclay CG, Brennand JE, Kelly RW, Calder AA. Interleukin-8 production by the human cervix. Am J Obstet Gynecol 1993;169: 625632.
  • 7
    Winkler M, Rath W, Fischer DC, et al. Regulation of interleukin-8 synthesis in human lower uterine segment fibroblasts by cytokines and growth factors. Obstet Gynecol 2000;95: 584588.
  • 8
    El Maradny E, Kanayama N, Halim A, Maehara K, Sumimoto K, Terao T. Biochemical changes in the cervical tissue of rabbit induced by interleukin-8, interleukin-1β, dehydroepiandrosterone sulphate and prostaglandin E2: a comparative study. Hum Reprod 1996;11: 10991104.
  • 9
    Rechberger T, Woessner F. Collagenase, its inhibitors, and decorin in the lower uterine segment in pregnant women. Am J Obstet Gynecol 1993;168: 15981603.
  • 10
    Osmers RGW, Bläser J, Kuhn W, Tschesche H. Interleukin-8 synthesis and the onset of labor. Obstet Gynecol 1995;86: 223229.
  • 11
    Winkler M, Menon R, Tschesche H, Kemp B, Rath W, Fortunato SJ. Gelatinases and their inhibitors in human lower uterine segment during term and preterm parturition. J Soc Gynecol Investig 2002;9(Suppl 1):210A.
  • 12
    Molnár M, Romero R, Hertelendy F. Interleukin-1 and tumor necrosis factor stimulate arachidonic acid release and phospholipid metabolism in human myometrial cells. Am J Obstet Gynecol 1993; 169,825169,829.
  • 13
    Romero R, Durum S, Dinarello CA, Oyarzun E, Hobbins JC, Mitchell MD. Interleukin-1 stimulates prostaglandin biosynthesis by human amnion. Prostaglandins 1989;37: 1322.
  • 14
    Lundin-Schiller S, Mitchell MD. Prostaglandin production by human chorion laeve cells in response to inflammatory mediators. Placenta 1991;12: 353363.
  • 15
    Mitchell MD, Edwin S, Romero RJ. Prostaglandin biosynthesis by human decidual cells: effects of inflammatory mediators. Prostaglandins Leukot Essent Fatty Acids 1990;41: 3538.
  • 16
    Ito A, Leppert PC, Mori Y. Human recombinant interleukin-1α increases elastase-like enzyme in human uterine cervical fibroblasts. Gynecol Obstet Invest 1990;30: 239241.
  • 17
    Winkler M, Rath W, Fischer DC, et al. Regulation of interleukin-8 synthesis in human lower uterine segment fibroblasts by cytokines and growth factors. Obstet Gynecol 2000;95: 584588.
  • 18
    Carlos TM, Harlan JM. Membrane proteins involved in phagocyte adherence to endothelium. Immunol Reviews 1990;114: 528.
  • 19
    Winkler M, Ruck P, Horny H-P, et al. Expression of cell adhesion molecules by endothelium in the human lower uterine segment during parturition at term. Am J Obstet Gynecol 1998;178: 557561.
  • 20
    Winkler M, Rath W. Zur rolle von zytokinen bei weheninduktion, zervixreifung und blasensprung. Z Geburtsh Neonatol 1996;200(Suppl 1):112.
  • 21
    Liggins GC. Cervical ripening as an inflammatory reaction. In: EllwoodDA, AndersonABM, editors. The Cervix in Pregnancy and Labor: Clinical and Biochemical Investigations. Edinburgh : Churchill-Livingstone, 1981: 19.