The effect of mifepristone administration on leukocyte populations ...

Molecular Human Reproduction vol.6 no.6 pp. 541–548, 2000

The effect of mifepristone administration on leukocyte populations, matrix metalloproteinases and inflammatory mediators in the first trimester cervix Fiona C.Denison1,4, Simon C.Riley1, Catherine L. Elliott1,4, Rodney W.Kelly2, Andrew A.Calder1 and Hilary O.D.Critchley1 1Obstetrics

and Gynaecology and 2Medical Research Council Reproductive Biology Unit, University of Edinburgh, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh, EH3 9ET, UK 3Department of Maternal and Fetal Medicine, Division of Paediatrics, Obstetrics and Gynaecology, Imperial College School of Medicine, Queen Charlotte’s Hospital, London, W6 OXG 4To

whom correspondence should be addressed at: Obstetrics and Gynaecology, Centre for Reproductive Biology, University of Edinburgh, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh, EH3 9ET, UK. E-mail: [email protected]

Cervical ripening is analogous to an inflammatory reaction characterized by an influx of inflammatory cells and an increase in inflammatory mediators. The anti-gestogen mifepristone is highly effective in inducing cervical ripening in women throughout gestation. However, its mechanism of action is largely unknown. The aim of the study was to investigate the effect of in-vivo administration of mifepristone on inflammatory cells and mediators in the cervix. Cervical biopsies were taken from women undergoing a first trimester termination of pregnancy at 0, 6, 12, 24 and 36 h (n ⍧ 6 per group) after mifepristone administration. Biopsies were fixed for immunohistochemistry and also cultured for subsequent analysis of culture media by radioimmunoassay or enzyme-linked immunosorbent assay. After administration of mifepristone (6–24 h), there was an increase in immunostaining for leukocyte common antigen (CD45), neutrophil elastase, monocytes (CD68), and matrix metalloproteinases (MMP)-1, -8 and -9. Immunostaining for MMP-2 and tissue inhibitor of metalloproteinases (TIMP)-1, -2 and -4 were unaffected by mifepristone treatment. Secretion of monocyte chemotactic protein (MCP-1) was significantly (P < 0.05) increased from biopsies taken 6–24 h after mifepristone administration. Cervical biopsies also released interleukin-8 (IL-8), prostaglandin (PG) E2, PGF2α and prostaglandin metabolites (PGEM and PGFM) although their secretion was unaffected by mifepristone treatment. This study suggests that mifepristone may, in part, effect cervical ripening by modulating the influx of inflammatory cells into the cervix, up-regulating MMP expression and inducing chemokine secretion by cervical tissue. Key words: cervical ripening/inflammation/leukocyte/mifepristone/pregnancy

Introduction Progesterone is essential for the maintenance of pregnancy with its withdrawal initiating cervical ripening and parturition throughout gestation in women. The anti-gestogen, antiglucocorticoid mifepristone (RU486) (Baird, 1993), is a highly effective abortifacient, particularly in combination with prostaglandin (PG), it induces cervical ripening and increases myometrial sensitivity (Bygdeman et al., 1994; Carbonne et al., 1995; Bugalho et al., 1996; Elliott et al., 1998). Mifepristone acts as a progesterone antagonist or a partial agonist, depending on whether progesterone is present or absent (Spitz and Bardin, 1993). Its mechanism of action at the cellular level is highly complex and a variety of hypotheses have been proposed. These include interfering with the dissociation of receptor and heat shock proteins and inducing a conformational change in the receptor, rendering it able to bind to but unable to activate the progesterone response element of responsive genes (Elashry et al., 1989; Clemm et al., 1995). The exact mechanisms by which mifepristone increases cervical compliance, reduces cervical resistance and effects cervical ripening are not well understood. Studies have failed © European Society of Human Reproduction and Embryology

to demonstrate any alterations in cervical morphology, collagenolytic activity, collagen content, plasminogen activator values, muscle contractility, 12-hydroxyeicosatetraenoic acid production or subsequent in-vitro bioconversion of radiolabelled arachidonic acid to thromboxane, PGE2 or PGF2α after in-vivo administration of mifepristone (Radestad et al., 1990; Heidvall et al., 1992; Bokstrom et al., 1994, 1998; Bokstrom and Norstrom, 1995). However, a decrease in the ratio of α-2 to β-adrenoceptors (Falkay, 1990), an increase in mast cells and signs of new capillary formation in the cervical stroma (Radestad et al., 1993) have been found after mifepristone administration. In addition, in both animal models and the human uterus, mifepristone has been shown to induce cellular infiltration, modulate inflammatory mediator release and inhibit the activity of prostaglandin dehydrogenase (PGDH) (Marbaix et al., 1992; Cheng et al., 1993; Jones et al., 1997; Patel et al., 1999). Its action on inflammatory cells and mediators in the cervix is not known. Cervical ripening has been likened to an inflammatory reaction (Liggins, 1981) characterized by an influx of inflammatory cells, particularly neutrophils and monocytes (Junqueira 541

F.C.Denison et al.

et al., 1980), into the cervical stroma. The chemokines interleukin-8 (IL-8) and monocyte chemotactic protein-1 (MCP-1), prostaglandins (PGE2, PGEM, PGF2α, and PGFM), matrix metalloproteinases (MMPs) and their inhibitors (TIMPs) have all been proposed as having a role in regulating this influx and in effecting tissue remodelling (Barclay et al., 1993; Ledingham et al., 1999). The aims of this study were, therefore, to investigate the effect of in-vivo administration of mifepristone to women undergoing a first trimester termination of pregnancy on the influx of inflammatory cells and the presence of the MMPs. The MMPs examined were MMP-1 and -8, important for breakdown of fibrillar collagens, the major structural collagens of cervix, and MMP-9, which degrades collagen IV of basement membranes (Leppert et al., 1995; Osmers et al., 1995). The MMP regulatory TIMP family was also examined. In addition, the secretion of IL-8, MCP-1 and prostaglandins by cervical explants after mifepristone administration was also examined to investigate their role in induction of cervical ripening.

Materials and methods Sample collection A total of 30 nulliparous women of ⬍9 weeks amenorrhoea who were to have a therapeutic suction termination of pregnancy under general anaesthesia were recruited. The women were randomized into five treatment groups, each consisting of six women to receive either no treatment (0 h) or 200 mg of oral mifepristone at 6, 12, 24 or 36 h prior to termination. After this period, some patients began to bleed and abort. This was the standard dosage regimen for termination used in this centre for several years and had the same efficacy as higher (600 mg) doses (McKinley et al., 1993). The demographic details of these women have been already described in detail in a previous publication reporting on leukocyte traffic in decidua in early pregnancy (Critchley et al., 1996). Exclusion criteria included: age ⬍16 years, serious medical condition, previous cervical surgery or inability to give informed consent. Immediately prior to termination, a cervical biopsy (4 mm3; ecto- and endo-cervix) was taken using Shumaker punch biopsy forceps consistently from the anterior lip of the cervix just to the side of the midline. Surgeons performing the terminations were blinded to the treatment allocation of the women. Biopsies were divided into two pieces on a sterile surgical swab. Tissues for subsequent immunohistochemical studies were placed in 10% neutral buffered formalin to fix overnight at 4°C, then washed in 70% ethanol prior to routine paraffin embedding. Biopsies destined for culture were placed in Roswell Park Memorial Institute (RPMI) 1640 culture medium at 4°C, for transport. These studies were approved by the Lothian Research Ethics Committee and informed and written consent was obtained from women prior to entry into the study. Immunohistochemistry Immunohistochemistry was performed as detailed previously (Ledingham et al., 1999; Riley et al., 1999). Paraffin embedded sections (5 µm) of cervix were dewaxed, rehydrated and endogenous peroxidase activity blocked in H2O2 (2% v/v in H2O) for 30 min at 23°C. Slides were washed in 0.01 mol/l phosphate buffered saline (PBS, 10 min) and blocked in diluted normal horse serum or goat serum (both 1:100; Vectastain, Vector Laboratories, Peterborough, UK; 30 min) as appropriate for the detection antibody. Excess blocking solution was removed and slides were incubated for 18 h

542

at 4°C under humid conditions with a primary antibody. The antibodies used were: MMP-1 affinity-purified rabbit polyclonal (Triple Point Biologics, Forest Grove, OR, USA, 1:10000), MMP-2 mouse monoclonal (Calbiochem, Nottingham, UK, 1:50), MMP-8 affinity-purified rabbit polyclonal (Triple Point Biologics, 1:2000), MMP-9 mouse monoclonal (Triple Point Biologics, 1:50), TIMP-1 affinity-purified rabbit polyclonal (Triple Point Biologics, 1:250), TIMP-2 affinitypurified rabbit polyclonal (Triple Point Biologics, 1:500), TIMP-3 affinity-purified rabbit polyclonal (Triple Point Biologics, 1:500), and TIMP-4 affinity-purified rabbit polyclonal (Chemicon, Harrow, UK, 1:1000). Our laboratory has confirmed the specificities of these antibodies previously (Ledingham et al., 1999; Riley et al., 1999a,b). Primary antibodies were detected using horse anti-mouse and goat anti-rabbit biotinylated secondary antibodies (Vectastain). Immunohistochemistry for leukocyte common antigen (CD45), monocyte marker (CD68) and neutrophil elastase was performed as previously described in detail (Critchley et al., 1996); all primary antibodies were diluted 1:50. Avidin–peroxidase complex (Vectastain) was added according to the manufacturer’s instructions. Specific immunoreactivity was identified by the application of the chromagen 3,3⬘-diaminobenzidine (Vectastain) that produces a brown colour. Sections were counterstained with haematoxylin prior to mounting and examination by light microscopy. All immunostaining experiments contained both parallel negative (where the primary antibody was omitted or replaced with an appropriate non-immune serum; see Figure 2C later) and positive controls (where a section of tonsil or human placenta was examined, as appropriate for the antibody, data not shown). Explant culture Reagents were from Sigma, Poole, UK unless otherwise stated. Explants were washed in PBS, dissected into small pieces 1–2 mm3 and cultured in quadruplicate (one explant/well) in 1 ml RPMI 1640 supplemented with 10% fetal calf serum (preliminary testing demonstrated that the serum used did not contain detectable amounts of cytokines that were subsequently measured), gentamicin (20 µg/ml), penicillin (50 IU/ml), streptomycin (50 µg/ml, Gibco, Paisley, UK), and L-glutamine (2 mmol/l) in a 24-well plate (Costar, High Wycombe, UK) for 24 h. Cytokine assays IL-8 was measured by a specific radioimmunoassay, with sensitivities and cross-reactivities as described previously in detail (Kelly et al., 1994). Rabbit anti-sera (raised against 72 amino acid variant by peptide synthesis using fluorenyl methoxy carbonyl) was used at 1:20 000. Radiolabelled human recombinant IL-8 (Becton-Dickinson, Oxford, UK) was prepared by iodination using chloramine-T. The intra- and inter-assay variations were 7 and 12% respectively. MCP-1 was measured by a specific radioimmunoassay as described for IL-8 (Kelly et al., 1994). Cross reactivities with MCP-2 and MCP-3 were ⬍0.1%. Goat anti-sera against MCP-1 (R&D Systems, Oxford, UK) was used at 1:12 500 and radiolabelled human recombinant MCP-1 (R&D Systems) was prepared by iodination using chloramine-T and purified by chromatography on a carboxy methyl silica column. The intra- and inter-assay coefficients of variation showed 6.3 and 8.6% relative standard deviations (R.S.D.) respectively. Prostaglandin assays Concentrations of PGE2, PGEM, PGF2α and PGFM were measured in the samples using specific enzyme-linked immunosorbent assays (ELISAs). The PGE2 and PGEM ELISAs measured a stable oximated derivative, using antibodies whose cross-reactivities and specificities have been reported elsewhere (Kelly et al., 1986; Kelly and Smith,

Mifepristone actions on cervix 1987). All dilutions were made using ELISA buffer: 150 mmol/l NaCl; 100 mmol/l Tris; 50 mmol/l Phenol Red solution; 2 mmol/l EDTA; 1 mmol/l 2-methylisothiazolone, Boehringer Mannheim, Lewes, UK; 1 mmol/l bromonitrodioxane, Boehringer Mannheim; 2 mg/ml bovine serum albumin (BSA); 0.05% Tween-20; pH 7.2. Plates (Costar, DNA-bind) were coated by covalent binding of affinitypurified donkey anti-rabbit serum and blocked with 0.1% BSA. For the PGE2, PGEM, PGF2α and PGFM assays the inter-assay coefficients of variation were 15.0, 14.7, 18.3 and 14.6% R.S.D. and the intraassay coefficients of variation were 7.8, 4.1, 5.2 and 6.8% R.S.D. respectively. Data and statistical analysis Due to the heterogeneity of the small samples obtained, semiquantitative analysis was not considered possible or meaningful. All sections were assessed qualitatively and blindly by two observers independently, whose findings displayed a high degree of comparability. Statistical analysis of the tissue explant data was performed using a Student’s unpaired t-test (StatView 4.1, Abacus Inc, Berkeley, CA, USA). The data were normally distributed and are expressed as ng/mg (mean ⫾ SEM) and P ⬍ 0.05 was considered to be statistically significant.

Results Immunolocalization of leukocyte markers in cervix CD45 (leukocyte common antigen) In biopsies from untreated women (Figure 1A) a few isolated cells, which were located mainly in cervical stroma, stained immunopositive for CD45. No clear differences were observed in biopsies obtained from women after 6 h treatment compared with untreated controls. However, 12–24 h after mifepristone administration there was a marked increase in discrete, immunopositive cells as cellular infiltrates, located predominately within the cervical connective tissue stroma (Figure 1B and 1C). Neutrophil elastase In biopsies from control women, only a few cells within the surface epithelium and stroma stained immunopositive for neutrophil elastase (Figure 1D). More than 12 h after mifepristone administration (Figure 1E and 1F), there was a marked increase in immunopositive cells, as cellular infiltrates, located mainly within cervical stroma and blood vessels. CD68 (monocyte/macrophage cell marker) Biopsies taken from women who had not received mifepristone were largely immunonegative for CD68 (Figure 1G) in both epithelial and stromal compartments. After 12 h treatment with mifepristone, a population of cells within cervical stroma stained immunopositive for CD68 (Figure 1H and 1I, some of these cells are highlighted by arrows). Immunolocalization of MMPs and TIMPs MMP-1 (interstitial collagenase) In tissue from control women, there were a few scattered cells in the cervical stroma which stained specifically for MMP-1 (Figure 2A). There was also weak, patchy positive immunoreactivity for MMP-1 in stroma in connective tissue and also in cervical surface epithelium which may also be associated with the connective tissues. In addition, in biopsies taken more than

6 h post-mifepristone administration (Figure 2B and 2C), there was a marked increase in cells in the stroma staining specifically for MMP-1. The patchy positive immunostaining in the epithelium connective tissue was comparable with untreated women (data not shown). A representative negative control section (Figure 2D) shows no non-specific immunostaining. MMP-8 (neutrophil collagenase) In untreated women there was patchy positive immunostaining for MMP-8 in cervical epithelium and stroma (Figure 2E). In addition, there were a few isolated cells within the cervical stroma which stained specifically for MMP-8. After 24 h treatment with mifepristone, there was a marked increase in the population of stromal cells, which stained specifically for MMP-8. There was also some patchy positive immunostaining of epithelium and connective tissue stroma, which was comparable with control sections (Figure 2F and 2G). MMP-9 In untreated women, MMP-9 was immunolocalized weakly to some cervical epithelial cells (Figure 2H). In biopsies taken more than 24 h after mifepristone administration, there was a marked increase in positive immunoreactivity for MMP-9 throughout the cervical epithelium, basement membrane and stromal connective tissue (Figure 2I). MMP-2 In untreated women, diffuse positive immunoreactivity for MMP-2 was visualized in surface and glandular epithelium and in the stromal compartment particularly associated with connective tissue and some stromal cells with minimal staining in cervical smooth muscle (Figure 2J). In-vivo treatment with mifepristone had no effect on MMP-2 immunolocalization (data not shown). TIMPs-1, -2, -3 and -4 In untreated women, TIMPs-1, -2 and -4 were immunolocalized to cervical surface and glandular epithelium, stromal connective tissue and cells (data not shown). Mifepristone administration had no effect on the pattern of immunostaining for TIMP-1, -2 or -4 at any timepoint (data not shown). Cervical explants were immunonegative for TIMP-3. Secretion of chemokines and prostaglandins Cervical biopsies released MCP-1, IL-8, PGE2, PGEM, PGF2α and PGFM. Significantly (P ⬍ 0.05) more MCP-1 was released by biopsies taken 6–24 h post-mifepristone administration than from control biopsies (Figure 3). Mifepristone treatment had no effect on the secretion of IL-8 or any prostaglandin of the mediators studied (Table I).

Discussion This study demonstrates that after administration of mifepristone there was an influx of leukocytes manifested as an increase in immunostaining for leukocytes (CD45), specifically neutrophils (neutrophil elastase) and monocytes (CD68). Similarly, there was an increase in localization of MMP-1, -8 and -9 after mifepristone treatment. MMP-2, TIMP-1, -2 and -4 were also present in cervical biopsies, although mifepristone had no effect on immunostaining. In addition, cervical biopsies 543

F.C.Denison et al.

Figure 1. Immunohistochemical localization in representative sections of cervical biopsies of CD45 (A, B and C), neutrophil elastase (D, E and F) and CD68 (G, H and I) taken (A, D and G) before and (B, C, E, F, H and I) after mifepristone administration. In untreated women, a few isolated cells stained immunopositive for (A, 0 h) CD45 and (D 0 h) neutrophil elastase. Control biopsies were also immunonegative for (G, 0 h) CD68. After mifepristone treatment, there was increased staining as stromal cellular infiltrates for CD45 (B, 12 h and C, 24 h) and neutrophil elastase (E, 12 h and F 24 h) and as discrete immunopositive stromal cells for CD68 (H 12 h and I 24 h). se ⫽ surface epithelium; st ⫽ stroma; ge ⫽ glandular epithelium; large arrows ⫽ cellular infiltrate; small arrows ⫽ specific cellular staining. Scale bars ⫽ 100 µm.

secreted MCP-1, IL-8, PGE2, PGEM, PGF2α and PGFM; however, only the release of MCP-1 by cervical biopsies was significantly increased 6–24 h after mifepristone administration. It is acknowledged that the findings reported in this study have their limitations. The technique of immunohistochemistry is non-quantitative and given the heterogeneity of staining, semi-quantitative or quantitative analysis was not appropriate or possible. In addition, due to the size, small numbers and heterogeneity of biopsies due to the differences in ecto- and 544

endo- cervix obtained, further studies such as dual localization were not possible. However, given these provisos, the reported patterns were consistent within groups (at least five out of six biopsies per group), between two independent observers and both positive (CD45, neutrophil elastase, CD68, MMP-1, -8 and -9) and negative (MMP-2, TIMP-1, -2 and -4) findings were demonstrated post-mifepristone administration. Mifepristone administration increased immunostaining for neutrophils and monocytes, which is analogous with physiological cervical ripening (Junqueira et al., 1980) and similar

Mifepristone actions on cervix

Figure 2. Immunohistochemical localization in representative sections of cervical biopsies taken of matrix metalloproteinase (MMP)-1 (A, B and C), negative control (D), MMP-8 (E, F and G), MMP-9 (H and I) and MMP-2 (J) before mifepristone (A, D, E, H and J) and after mifepristone (B, C, F, G and I) administration. In control women, a few discrete stromal cells stained immunopositive for (A) MMP-1. After mifepristone administration, there was a marked increase in discrete immunopositive stromal cells (B ⫽ 6 h and C ⫽ 24 h). (D) Representative negative control section with primary antibody omitted. In (E, 0 h) control and (F, 24 h) mifepristone-treated women there was patchy positive epithelial and stromal staining for MMP-8. In addition, (F and G) after mifepristone administration there was an increase in discrete stromal immunopositive cells. In (H, 0 h) control women, MMP-9 was weakly immunolocalized to surface epithelium. After treatment, there was a marked increase in positive immunoreactivity for MMP-9 in surface epithelium and stromal connective tissue (I, 24 h). MMP-2 was immunolocalized diffusely to surface epithelium and stromal connective tissue in both control (J, 0 h) and mifepristone-treated women (data not shown). se ⫽ surface epithelium; st ⫽ stroma; ge ⫽ glandular epithelium; small arrows ⫽ specific cellular staining. Scale bars ⫽ 100 µm.

to the pattern seen in the guinea pig cervix after administration of anti-gestogen (Hegele-Hartung et al., 1989). It has been previously demonstrated that there is a significant increase in mast cells within cervical stroma 24 h post-mifepristone administration (Radestad et al., 1993). Mast cells may, therefore, account for some of the leukocytes, which did not stain with the neutrophil elastase or monocyte marker. Mast cells

were not stained for in this study due to the scarcity of cervical biopsy tissue. Leukocyte and mast cell invasion of the connective tissue stroma and their subsequent degranulation releasing collagenolytic enzymes could be a mechanism by which mifepristone induces cervical softening. Chemokines play a pivotal role in leukocyte migration and activation. The significant increase in MCP-1 release by 545

F.C.Denison et al.

a prostaglandin-independent mechanism or indirectly affects prostanoid concentrations at the local level by modulation of PGDH expression (Patel et al., 1999). Specific stromal cellular immunostaining for both interstitial (MMP-1) and neutrophil collagenase (MMP-8) was increased in biopsies taken ⬎24 h after mifepristone administration. Up-regulation of MMP-1 release by mifepristone has been previously demonstrated in human endometrial stroma and fibroblasts (Singer et al., 1997; Lockwood et al., 1998), however stimulation of MMP-8 by progesterone antagonism is a novel finding. MMP-8 is unique in that it is expressed exclusively in inflammatory conditions (Balbin et al., 1998). Although originally described as being released specifically by neutrophils, it has recently been shown to be secreted by cytokine-activated fibroblasts (Halinen et al., 1996; Hanemaaijer et al., 1997). The majority of immunopositive stromal cells were of fibroblast morphology. This is therefore consistent with MMP-1 being a fibroblast product and release of MMP-8 by cytokine activated fibroblasts, such as would be found during the inflammatory process of cervical ripening. Minimal immunostaining for the gelatinase MMP-9 was observed in control biopsies with a marked increase in immunoreactivity in biopsies taken ⬎24 h after mifepristone administration. Physiological concentrations of progesterone have previously been demonstrated to suppress MMP-9 release by rabbit cervical fibroblasts (Imada et al., 1997) and human trophoblast cells with MMP-9 inhibition being antagonized by progesterone withdrawal (Shimonovitz et al., 1998). The high levels of progesterone during normal pregnancy may act as a protective mechanism to prevent premature cervical ripening. Up-regulation of MMP-9 post-mifepristone administration suggests that MMP-9 may play a role in mifepristone-induced cervical ripening. In contrast, immunolocalization of the gelatinase MMP-2 was unaffected by mifepristone administration. Whether MMP-2 is under progesterone regulation is controversial, with studies variously demonstrating inhibition of release by progesterone (Marbaix et al., 1992; Irwin et al., 1996) and no effect (Lockwood et al., 1998). It may be therefore that cervical MMP-2 is not progesterone regulated or that inhibition of its release was not detectable by immunohistochemistry. The immunolocalization of TIMPs-1, -2 and -4 in cervix was similar to the distribution previously demonstrated by this

explants post-mifepristone is consistent with the increase in CD68 immunostaining and may provide a mechanism for monocyte migration. There was a trend towards an increase in IL-8 release post-mifepristone administration but this failed to reach significance. This may be due to the lack of an effect or the low number of biopsies in each sample group. Alternatively, neutrophil chemotaxis may be the result of local neutrophil chemotactic factors released by monocytes (Wuyts et al., 1994). There was no difference in release of any prostanoids measured which is consistent with previous data which failed to demonstrate any alteration in the concentrations of PGE2 or PGF2α in first trimester cervical mucus postadministration of mifepristone (Bokstrom et al., 1995). This suggests that mifepristone either induces cervical ripening by

Figure 3. Release of monocyte chemotactic protein (MCP)-1 and interleukin (IL)-8 from 24 h cervical explant cultures taken from control patients (0 h) and patients treated with mifepristone 6–36 h before termination. Significantly more MCP-1 was released 6–24 h after mifepristone administration than from control biopsies (P ⬍ 0.05)*. Values are expressed as ng/mg ⫾ SEM.

Table I. Release of prostanoids from the first trimester human cervix after administration of mifepristone in vivo. Biopsies were taken 6, 12, 24 and 36 h post-mifepristone administration. Values are expressed as ng/mg wet weight tissue ⫾ SEM (in parentheses), with tissue collected from six women per time point and biopsies cultured in quadruplicate Prostanoid (ng/mg wet weight ⫾ SEM) Time (h)

PGF2α

PGFM

PGE2

PGEM

Control (0) 6 12 24 36

1.85 5.20 1.47 2.03 3.18

0.52 0.55 0.50 0.65 0.74

0.99 1.05 1.62 1.83 2.02

0.45 0.59 0.82 0.24 0.60

(0.71) (4.31) (0.60) (0.89) (1.27)

(0.17) (0.30) (0.18) (0.19) (0.29)

(0.39) (0.69) (0.56) (0.67) (0.68)

(0.17) (0.30) (0.50) (0.06) (0.20)

PGF2α ⫽ prostaglandin F2α; PGFM ⫽ prostaglandin F metabolite; PGE2 ⫽ prostaglandin E2; PGEM ⫽ prostaglandin E metabolite.

546

Mifepristone actions on cervix

laboratory (Ledingham et al., 1999), and was not affected by mifepristone administration. This conflicts with studies in rabbit cervical fibroblasts in which progesterone stimulated the release of TIMP-1 and -2 (Imada et al., 1994) but is consistent with those in human endometrial stromal cells where progesterone had no effect on TIMP-1 release (Lockwood et al., 1998). The increased immunostaining of MMP-1, -8 and -9 in conjunction with no alteration in TIMP localization would favour tissue remodelling and degradation. In summary, therefore, the following hypothesis is proposed to explain the mechanisms by which mifepristone could induce cervical ripening. A fall in the effect of local progesterone, effected by mifepristone administration, would induce new blood capillary formation (Radestad et al., 1993) and upregulate MMP-9 release, which would degrade vascular basement membrane. In conjunction with release of chemokines including MCP-1 and IL-8, this would favour accumulation of infiltrating leukocytes, specifically neutrophils, monocytes and mast cells (Radestad et al., 1993), in cervical vasculature and their subsequent emigration and degranulation within the connective tissue stroma. In addition, increased release of the collagenases MMP-1 and MMP-8 by stromal cells would further promote remodelling and loosening of the connective tissue stroma. Finally, the failure of mifepristone to effect changes in TIMP expression would alter the homeostatic balance between MMPs and TIMPs towards tissue remodelling and degeneration. Future studies, using larger numbers of biopsies per time point, could investigate this hypothesis by quantifying MMP and TIMP release by ELISA and measuring release of other inflammatory mediators, e.g. IL-1 and TNF-α.

Acknowledgements Dr Fiona C Denison was supported by a Clinical Research Training Fellowship from Action Research (S/F/0705). Technical support was provided by Miss Gail Baldie and Miss Deborah Mauchline. We also acknowledge the clinical support of Dr Joo Thong. This study was supported by funding from the Scottish Hospital Endowments Research Trust (Project grant no. 1389; SCR) and the Medical Research Council (Project grant no. G9406438PA; HODC).

References Baird, D.T. (1993) Antigestogens. Br. Med. Bull., 49, 73–87. Balbin, M., Fueyo, A., Knauper, V. et al. (1998) Collagenase 2 (MMP-8) expression in murine tissue-remodelling processes – Analysis of its potential role in postpartum involution of the uterus. J. Biol. Chem., 273, 23959– 23968. Barclay, C.G., Brennand, J.E., Kelly, R.W. et al. (1993) Interleukin-8 production by the human cervix. Am. J. Obstet. Gynecol., 169, 625–632. Bokstrom, H., Atterfelt, P., Alexandersson, M. et al. (1998) Preoperative cervical softening before first trimester legal abortion by mifepristone and misoprostol – A double-blind, randomized, clinical, biochemical, and immunohistochemical study. Contraception, 58, 157–163. Bokstrom, H. and Norstrom, A. (1995) Effects of mifepristone and progesterone on collagen-synthesis in the human uterine cervix. Contraception, 51, 249–254. Bokstrom, H., Norstrom, A. and Radestad, A. (1994) Cervical smooth-muscle contractile activity after treatment with mifepristone and progesterone. Contraception, 49, 115–123. Bokstrom, H., Norstrom, A. and Wiqvist, N. (1995) Cervical mucus concentration of prostaglandins E2 and F2α after pretreatment with mifepristone in the first trimester of pregnancy. Prostaglandins, 49, 41–48. Bugalho, A., Bique, C., Pereira, C. et al. (1996) Uterine evacuation by vaginal

misoprostol after second trimester pregnancy interruption. Acta Obstet. Gynecol. Scand., 75, 270–273. Bygdeman, M., Swahn, M.L., Gemzell-Danielsson, K. and Gottlieb, C. (1994) The use of progesterone antagonists in combination with prostaglandin for termination of pregnancy. Hum. Reprod., 9, (Suppl.1), 121–125. Carbonne, B., Brennand, J.E., Maria, B. et al. (1995) Effects of gemeprost and mifepristone on the mechanical-properties of the cervix prior to first trimester termination of pregnancy. Br. J. Obstet. Gynaecol., 102, 553–558. Cheng, L., Kelly, R.W., Thong, K.J. et al. (1993) The effect of mifepristone (RU486) on the immunohistochemical distribution of prostaglandin E and its metabolite in decidual and chorionic tissue in early pregnancy. J. Clin. Endocrinol. Metab., 77, 873–7. Clemm, D.L., Macy, B.L., Santiso-Mere, D. et al. (1995) Definition of the critical cellular components which distinguish between hormone and antihormone activated progesterone receptor. J. Ster. Biochem. Mol. Biol., 53, 483–495. Critchley, H.O.D., Kelly, R.W., Lea, R.G. et al. (1996) Sex steroid regulation of leukocyte traffic in human decidua. Hum. Reprod., 11, 2257–2262. Elashry, D., Onate, S. A., Nordeen, S. K. et al. (1989) Human progesteronereceptor complexed with the antagonist RU486 binds to hormone response elements in a structurally altered form. Mol. Endocrinol., 3, 1545–1558. Elliott, C.L., Brennand, J.E. and Calder, A A. (1998) The effects of mifepristone on cervical ripening and labor induction in primigravidae. Obstet. Gynecol., 92, 804–809. Falkay, G. (1990) Effects of antiprogestogen (RU486) treatment on myometrial and cervical alpha and beta-adrenoceptors in pregnant rabbits. Hum. Reprod., 5, 924–927. Halinen, S., Sorsa, T., Ding, Y.L. et al. (1996) Characterization of matrix metalloproteinase (MMP-8 and –9) activities in the saliva and in gingival crevicular fluid of children with Down’s syndrome. J. Periodontol., 67, 748–754. Hanemaaijer, R., Sorsa, T., Konttinen, Y.T. et al. (1997) Matrix metalloproteinase-8 is expressed in rheumatoid synovial fibroblasts and endothelial cells. J. Biol. Chem., 272, 31504–31509. Hegele-Hartung, C., Chwalisz, K., Beier, H.M. et al. (1989) Ripening of the uterine cervix of the guinea-pig after treatment with the progesterone antagonist onapristone (ZK 98,299): an electron microscopic study. Hum. Reprod., 4, 369–377. Heidvall, K., Radestad, A., Christensen, N.J. et al. (1992) Production of 12hydroxyeicosatetraenoic acid in early pregnant uterine cervix. Lack of correlation to mifepristone induced cervical ripening – a double-blind randomized biomechanical and biochemical- study. Prostaglandins, 43, 473–482. Imada, K., Ito, A., Itoh, Y. et al. (1994) Progesterone increases the production of tissue inhibitor of metalloproteinases-2 in rabbit uterine cervical fibroblasts. FEBS Lett., 341, 109–112. Imada, K., Ito, A., Sato, T. et al. (1997) Hormonal regulation of matrix metalloproteinase-9 gelatinase B gene expression in rabbit uterine cervical fibroblasts. Biol. Reprod., 56, 575–580. Irwin, J.C., Kirk, D., Gwatkin, R.B.L. et al. (1996) Human endometrial matrix metalloproteinase-2, a putative menstrual proteinase – Hormonal regulation in cultured stromal cells and messenger RNA expression during the menstrual cycle. J. Clin. Invest., 97, 438–447. Jones, R.L., Kelly, R.W. and Critchley, H.O.D. (1997) Chemokine and cyclooxygenase-2 expression in human endometrium coincides with leukocyte accumulation. Hum. Reprod., 12, 1300–1306. Junqueira, L.C.U., Zugaib, M., Montes, G.S. et al. (1980) Morphologic and histochemical evidence for the occurrence of collagenolysis and for the role of neutrophilic polymorphonuclear leukocytes during cervical dilation. Am. J. Obstet. Gynecol., 138, 273–281. Kelly, R.W. and Smith, S.K. (1987) Progesterone and antiprogestins, a comparison of their effect on prostaglandin production by human secretory phase endometrium and decidua. Prost. Leuk. Med., 29, 181–186 Kelly, R.W., Deam, S., Cameron, M.J. et al. (1986) Measurement by radioimmunoassay of prostaglandins as their methyl oximes. Prost. Leuk. Med., 24, 1–14. Kelly, R.W., Illingworth, P., Baldie, G. et al. (1994) Progesterone control of interleukin-8 production in endometrium and chorio-decidual cells underlines the role of the neutrophil in menstruation and parturition. Hum. Reprod., 9, 253–258. Ledingham, M.A., Denison, F.C., Riley, S.C. et al. (1999) Matrix metalloproteinases –2 and –9 and their inhibitors (TIMPs) are produced by the human uterine cervix but their secretion is not regulated by nitric oxide donors. Hum. Reprod., 14, 2089–2096.

547

F.C.Denison et al. Leppert, P.C. (1995) Anatomy and physiology of cervical ripening. Clin Obstet Gynecol., 2, 267–279. Liggins, G.C. (1981) Cervical ripening as an inflammatory reaction. In Ellwood, D.A. and Anderson, A.B.M. (eds), The Cervix in Pregnancy and Labour, Clinical and Biochemical Investigations. Churchill Livingstone, Edinburgh, Scotland, UK, pp. 1–9. Lockwood, C.J., Krikun, G., Hausknecht, V.A. et al. (1998) Matrix metalloproteinase and matrix metalloproteinase inhibitor expression in endometrial stromal cells during progestin-initiated decidualization and menstruation-related progestin withdrawal. Endocrinology, 139, 4607–4613. Marbaix, E., Donnez, J., Courtoy, P.J. et al. (1992) Progesterone regulates the activity of collagenase and related gelatinase-A and gelatinase-B in human endometrial explants. Proc. Natl Acad. Sci. USA, 89, 11789–11793. McKinley, C., Thong, K.J. and Baird, D.T. (1993) The effect of dose of mifepristone and gestation on the efficacy of medical abortion with mifepristone and misoprostol. Hum. Reprod., 8, 1502–1505. Osmers R.G.W., Adelmann-Grill, B.C., Rath W. et al. (1995) Biochemical events in cervical ripening dilatation during pregnancy and parturition. J. Obstet. Gynaecol., 21, 185–194. Patel, F.A., Clifton, V.L., Chwalisz, K. et al. (1999) Steroid regulation of prostaglandin dehydrogenase activity and expression in human term placenta and chorio-decidua in relation to labor. J. Clin. Endocrinol. Metab., 84, 291–9. Radestad, A., Bydgeman, M. and Green, K. (1990) Induced cervical ripening with mifepristone (RU486) and bioconversion of arachidonic acid in human pregnant uterine cervix in the first trimester. Contraception, 41, 283–292. Radestad, A., Thyberg, J. and Christensen, N.J. (1993) Cervical ripening with mifepristone (RU486) in 1st trimester abortion – an electron-microscope study. Hum. Reprod., 8, 1136–1142. Riley, S.C., Leask, R., Chard, T. et al. (1999a) Secretion of matrix metalloproteinase-2, matrix metalloproteinase-9 and tissue inhibitor of metalloproteinases into the intrauterine compartments during early pregnancy. Mol. Hum. Reprod., 5, 376–381. Riley, S.C., Leask, R., Denison, F.C. et al. (1999b) Secretion of tissue inhibitors of metalloproteinases by human fetal membranes, decidua and placenta at parturition. J. Endocrinol., 162, 351–359. Shimonovitz, S., Hurwitz, A., Hochner-Celnikier, D. et al. (1998) Expression of gelatinase B by trophoblast cells: Down-regulation by progesterone. Am. J. Obstet. Gynecol., 178, 457–461. Singer, C.F., Marbaix, E., Kokorine, I. et al. (1997) Paracrine stimulation of interstitial collagenase (MMP-1) in the human endometrium by interleukin 1 alpha and its dual block by ovarian steroids. Proc. Natl Acad. Sci. USA, 94, 10341–10345. Spitz, I.M. and Bardin, C.W. (1993) Clinical pharmacology of RU486- an antiprogestin and antiglucocorticoid. Contraception, 48, 403–444. Wuyts, A., Proost, P., Put, W. et al. (1994) Leukocyte recruitment by monocyte chemotactic proteins (mcps) secreted by human phagocytes. J. Immunol. Methods, 174, 237–247. Received on January 26, 2000; accepted on March 14, 2000

548