| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Cell Biology Unit, Christian de Duve Institute of Cellular Pathology (P.B.C., C.G., Y.E., P.J.C., E.M., P.H.), and Department of Pathology, Saint-Luc University Clinics (C.G., E.M.), Université Catholique de Louvain, B-1200 Brussels, Belgium
Address all correspondence and requests for reprints to: Etienne Marbaix, Cell Biology Unit, Christian de Duve Institute of Cellular Pathology, Université Catholique de Louvain, Avenue Hippocrate, 75, B-1200 Brussels, Belgium. E-mail: marbaix{at}cell.ucl.ac.be.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MMPs are a large family of zinc-dependent extracellular endoproteinases that process a variety of structural and regulatory molecules at neutral pH (4, 5). MMP-9, also referred to as gelatinase B, type IV collagenase or 92-kDa gelatinase, is usually secreted as a 92-kDa glycosylated prometalloenzyme. Activation of the proenzyme involves the proteolytic removal of the N-terminal prodomain conserved in all MMPs, resulting in an 82-kDa active enzyme. MMP-9 degrades the denatured collagens (gelatin) with high specific activity. It can also cleave native type IV, V, and XI collagens; elastin; and a variety of non-ECM molecules (6). MMP-9 is clearly linked to cancer and wound-repair. After grafting of cancer cells, mice devoid of MMP-9 show markedly reduced angiogenesis (7). Moreover, production of MMP-9 by infiltrating inflammatory cells was increased in carcinoma tissues (8). MMP-9 is also involved in cell migration upon activation by the urokinase/plasmin system (9).
Besides a semiquantitative analysis showing the presence of residual MMP-9 mRNA levels in nonperimenstrual endometria (10), reports of MMP-9 expression in the human endometrium along the menstrual cycle are essentially based on immunohistochemical detection or in situ hybridization. Unlike the expression of other endometrial MMPs that seems restricted to one specific cell type (stromal cells for MMP-1 and -3, epithelial cells for MMP-7), MMP-9 was detected in or around all cell types present in the endometrium. However, MMP-9 mRNA appeared restricted to foci of stromal cells at the perimenstrual phase in one study (11) and was not detected in another study that only explored the nonmenstrual phases of the cycle (12). The protein could be extracted from endometrial tissue collected throughout the cycle, exclusively in its latent form during the nonmenstrual phases and partially as an active form at menstruation, likely as a result of MMP-3 cleavage (13, 14). Activation of proMMP-9 was also increased during episodes of irregular dysfunctional bleeding on levonorgestrel treatment (15). MMP-9 was immunolocalized: 1) in neutrophils, eosinophils and macrophages throughout the cycle, with an increase at menstrual phase (16, 17); 2) in epithelial glandular cells during the late proliferative and/or the secretory phase (14, 16, 18), with a more intense staining at the apical portion of the cells (17); and 3) in a few stromal cells during the secretory phase (14), and at menstruation (17, 19). MMP-9 was also detected around, and even in, arteriolar walls during the midproliferative, secretory, and menstrual phases (17).
The detection of foci expressing MMP-9 further suggests the existence of local regulators such as cytokines. Among these, LEFTY-A, initially identified as an endometrial bleeding-associated factor (EBAF), appears as a good candidate. Human LEFTY-A is a recently discovered member of the TGF-ß family, secreted as a 42-kDa precursor susceptible to proteolytic cleavage (20, 21, 22). Active forms are able to induce MAPK activity and to inhibit TGF-ß signaling (23, 24). In the human endometrium in vivo, we recently quantified a dramatic (3-log) increase in LEFTY-A mRNA expression during the perimenstrual phase (25). Using explant culture from proliferative endometria, we demonstrated the ability of recombinant LEFTY-A to induce proMMP-3 and proMMP-7 expression. This effect, as well as endogenous LEFTY-A mRNA expression, was completely inhibited by P, combined or not with estradiol.
In this study, we first measured endometrial MMP-9 mRNA concentration along the normal menstrual cycle in vivo, in comparison with LEFTY-A mRNA expression (25). We next examined whether P, combined with estradiol, regulates MMP-9 in culture of endometrial explants, by measuring mRNA expression (again in comparison with LEFTY A mRNA) and protein production, release, and activation. MMP-9 was immunostained in the same samples, before and after culture. Finally, we directly tested the ability of recombinant LEFTY-A to modulate MMP-9 expression and/or activation in explant culture.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Normal endometrial tissue was sampled at various phases of the menstrual cycle from hysterectomy specimens (n = 51) or biopsies (n = 9). An ideal menstrual cycle of 28 d was divided into the following phases: early proliferative (d 6–8, n = 12), midproliferative (d 9–11, n = 1), late proliferative (d 12–14, n = 10), early secretory (d 15–18, n = 6), midsecretory (d 19–22, n = 6), late secretory (d 23–27, n = 10), and perimenstrual (d 28–5, n = 15). All patients were premenopausal (range, 25–55 yr old), were not treated with ovarian steroids and either were operated for conditions unrelated to the endometrium or had a biopsy for routine evaluation of infertility. Patients gave informed consent, and the study was approved by the Ethical Committee of the Université Catholique de Louvain, in accordance with the Declaration of Helsinki of the World Medical Association.
The collection and culture of endometrial explants were performed as previously described (25). Briefly, endometrium was gently scraped with a sterile surgical blade, from hysterectomy specimens, and put in ice-cold PBS; biopsies were immediately collected in ice-cold PBS. Part of the endometrial tissue was fixed in 4% formaldehyde for histological examination; dating was first estimated, following established microscopic criteria, and was finely adjusted according to clinical information of the last menstrual period whenever available. Noncultured samples were quickly frozen at –80 C in lysis buffer (SV Total RNA Isolation System, Promega, Madison, WI). Alternatively, tissue explants were cultured as described (2). Briefly, tissue samples were cut in pieces of about 1-mm side, with a sterile surgical blade, and placed in tissue culture inserts (Millipore, Bedford, MA; 24 explants/30-mm insert for RNA extraction; six explants/12-mm insert in triplicates for experiments addressing the effect of recombinant LEFTY-A). DMEM (Life Technologies, Inc., Merelbeke, Belgium), devoid of serum and phenol red, was placed in the lower chamber and renewed daily (300 µl or 1.2 ml in 12-mm and 30-mm inserts, respectively). Medium was without hormonal addition (–H), or supplemented with 1 nM water-soluble 17ß-estradiol (+E), 100 nM P (+P), or their combination (+E+P; Sigma-Aldrich, Bornem, Belgium) and, when applicable, with 5 ng/ml recombinant LEFTY-A (R&D Systems, Abingdon, UK), 4 µM cycloheximide (Sigma, St. Louis, MO), or 10 µM monensin (Calbiochem/Merck, Darmstadt, Germany). After collection, all media were supplemented with 0.05 vol of 1 M Tris-HCl buffer (pH 7.5), 1% (vol/vol) Triton-X-100, 0.1 M CaCl2, 60 mM NaN3, and kept frozen at –20 C until biochemical analysis. At the end of the culture, explants were frozen at –80 C in lysis buffer until RNA extraction.
RNA extraction and reverse transcription (RT)
Total RNA was extracted using the SV Total RNA Isolation System (Promega, Leiden, The Netherlands), and 200 ng total RNAs was reverse-transcribed by using the Thermoscript RT-PCR System (Life Technologies, Inc., Paisley, UK) as described (25).
Oligonucleotide primers used for PCR amplification
Specific oligonucleotide primer sequences used in competitive PCR (cPCR) were from Huang et al. (26) for the human MMP-9 and from Cornet et al. (25) for the human LEFTY-A and ß-ACTIN. The 3'-end primer used to prepare the MMP-9 competitor was redesigned (5'-CCCCACTTCTTGTCGCTGTCCTCCTTACCCAG-3') to limit the extent of the deletion and ensure an amplification efficiency similar to that of the target cDNA. For real-time PCR, the primer pair and Taqman probe for MMP-9 were from Van Trappen et al. (27), and the ß-ACTIN oligonucleotides were adapted from Kreuzer et al. (28) (sense: 5'-AGCCTCGCCTTTGCCGA-3'; Taqman probe: 5'-FAM-CCGCCGCCCGTCCACACCCGCC-TAMRA-3'; antisense: 5'-TCACGCCCTGGTGCCTG-3'). Customized primers were obtained from Life Technologies, Inc. or Biosource (BioSource International, Camarillo, CA), and their optimal hybridization temperatures were determined by comparing different temperatures in parallel PCR using a Biometra T-gradient thermocycler (Westburg, Leusden, The Netherlands).
Preparation of competitors for cPCR and standards for real-time PCR
The competitor for MMP-9 was constructed by deleting a 66-bp fragment from the corresponding target cDNA. Truncated fragments were synthesized by PCR using a 3'-end competitor primer (see above) as previously described (25). The competitors for LEFTY-A and ß-ACTIN were those used in our previous work (25).
MMP-9 and ß-ACTIN standards for real-time PCR were constructed in a similar way. Purified DNA products obtained after PCR amplification with appropriate primers were cloned into pCR II-TOPO vector (TOPO TA cloning kit; Life Technologies, Inc., Paisley, UK). Plasmids were transfected and amplified in Escherichia coli XL-1 blue competent bacteria (Stratagene, Amsterdam, The Netherlands), purified, and quantified by spectrophotometry. Known amounts (1 µg) of plasmids were linearized with Xho I or Xba I to obtain MMP-9 and ß-ACTIN standards, respectively, and were stored at –20 C as aliquots that were thawed only once and used immediately.
cPCR
cPCR for LEFTY-A, MMP-9, and ß-ACTIN was performed as described (25). The optimal hybridization temperature used for LEFTY-A, ß-ACTIN, and MMP-9 amplification was 65 C, 68 C, and 62 C, respectively.
Real-time PCR
Real-time PCR was performed with the LightCycler-FastStart DNA Master Hybridization probes (Roche/Boehringer, Mannheim, Germany) in the LightCycler apparatus (Roche/Boehringer) according to the manufacturer’s protocol. Concentration of probes and magnesium were optimized for each primer set. Aliquots of 3 µl of 3-fold-diluted RT products were subjected to PCR with 500 nM adequate paired primers, 200 nM Taqman probe, 7.25 mM and 3 mM MgCl2 for MMP-9 and ß-ACTIN, respectively, in a total vol of 20 µl. A control PCR without template DNA was performed in each experiment. A standard curve was generated from serial 10-fold dilutions of the purified linearized plasmid containing the appropriate cDNA (see above). The real-time PCR with Taqman probe was performed as follows. After an initial denaturation at 95 C for 10 min, the DNA was amplified through 40–45 cycles of 20 sec at 95 C, and 15 sec at 60 C or 20 sec at 67 C for MMP-9 and ß-ACTIN, respectively. The fluorescence signal (F1/F2) was measured after each amplification cycle. The PCR products were stored at 4 C until electrophoresis analysis. All PCR measurements were performed in duplicates. Data were analyzed by using the LightCycler Software (version 3.5, Roche/Boehringer). Regression analyses of the threshold cycle values of the standard dilution series were used to determine the amplification efficiency. The absolute amount of molecules in the experimental samples was determined by extrapolating the threshold cycle values from the standard curves. We considered that a PCR experiment was valid when: 1) no amplification was detected without DNA template as a control; 2) the slope of the logarithmic standard curve was around –3.3 (range, –3.1 to–3.5) to ensure full efficiency of the PCR (calculated as 10–1/slope); and 3) the amplification curve of the samples was parallel to those of the standards. The ß-ACTIN values were used for standardization as previously described (25), and the standardized ratio (MMP-9 mRNA/ß-ACTIN mRNA or LEFTY-A mRNA/ß-ACTIN mRNA) will hereafter be referred to as the relative amount of MMP-9 or LEFTY-A mRNA, respectively.
Western blotting
Western blotting for MMP-3 and MMP-7 was performed as previously described (25).
Zymography
MMP-9 was assayed in conditioned media or in explants by gelatin zymography as previously described (2). Explants were homogenized in 200 µl PBS (pH 7.4) with a Dounce potter. The band intensity was quantified by the ScnImage program (Scion Corporation, Frederick, MD) after scanning the zymogel. Data were normalized according to protein concentration in conditioned media or in explant homogenates determined using the Bradford method [(29); Bio-Rad, Richmond, CA].
Immunolocalization of MMP-9
Histological sections were immunolabeled for MMP-9 using a purified mouse monoclonal antibody (IgG1, clone 56–2A4; provided by K. Iwata, Fuji Chemical Industries Ltd, Toyama, Japan). Controls were obtained by omission of the primary antibody or its replacement by a nonrelevant antibody of the same isotype, raised against the immediate early antigen of cytomegalovirus (Argene, Varhiles, France).
After removal of paraffin and inactivation of endogenous peroxidases with 0.3% H2O2 for 30 min at room temperature, slides aimed at detecting MMP-9 were transferred in 0.01 M sodium citrate buffer (pH 5.8) and incubated into a water bath at 98 C for 75 min to unmask the antigenic sites. Nonspecific binding sites were blocked by incubating the slides for 1 h into 50 mM Tris-HCl (pH 7.4) containing 10% normal goat serum and 1% BSA, and the histological sections were incubated overnight at 4 C with 0.5 µg/ml of the primary antibody. After two washes in deionized water and two in 50 mM Tris-HCl (pH 7.4), slides were incubated for 1 h at room temperature with peroxidase-conjugated dextran molecules carrying antimouse secondary antibodies (Envision, DakoCytomation, Glostrup, Denmark). The peroxidase activity was revealed by incubation for 10 min at room temperature with 0.5 mg/ml diaminobenzidine in the Tris-HCl buffer (pH 7.4). Sections were then washed in tap water and slightly counterstained with hematoxylin.
Double-immunolabeling
Histological sections were immunolabeled for MMP-9 and for CD10 (clone 56C6; Novocastra, Newcastle, United Kingdom), CD15 (Leu-M1; Becton Dickinson, San Jose, CA), or CD68 (clone PGM-1; DakoCytomation). Slides were first immunolabeled for MMP-9, as described above, with 0.25 µg/ml of the primary mouse monoclonal antibody clone 56–2A4. After peroxidase staining, slides were washed and incubated in 0.1 N HCl for 1 h at room temperature to remove the antibodies. After washing in water, nonspecific binding sites were blocked for 30 min (see above), and the slides were incubated overnight at 4 C with 2 µg/ml or 5 µg/ml anti-CD15 or anti-CD68 antibodies, respectively, or with a 1:10 dilution of the anti-CD10 antibody. Controls for the second immunolabeling were obtained by omission of the anti-CD primary antibody. After two washes in deionized water and two washes in 50 mM Tris-HCl (pH 7.4), slides were incubated for 30 min at room temperature with 40 µg/ml of a secondary antimouse monoclonal antibody coupled to alkaline phosphatase (AP) (Biomakor, Rehovot, Israel). To further amplify the signal, slides were washed and incubated for 30 min at room temperature with 2 µg/ml of an AP-coupled anti-AP antibody (Dako-Cytomation). The AP activity was revealed by incubation for 1 h at room temperature, in the dark, with 1 mg/ml Fast Red, 0.4 mg/ml Naphthol AS-MX, and 0.15 mg/ml Levamisole in 0.1 M Tris-HCl buffer, pH 7.4 (Sigma). Sections were then washed in tap water and mounted in aqueous medium (Fluorescent Mounting medium; DakoCytomation).
Statistical analysis
Statistical significance was tested using the Wilcoxon matched-pairs or two-sample tests or the paired Student’s t test, as appropriate. Differences were interpreted as significant at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We recently demonstrated that radioactive cPCR is an accurate, sensitive, and reproducible method to precisely quantify specific mRNA molecules (25). To measure MMP-9 mRNA amounts in endometrial samples, we first compared cPCR with real-time PCR using Taqman probes (Fig. 1A
). The Taqman probe assay and cPCR gave similar results for the 13 samples tested, and duplicate experiments were highly reproducible. The linearity, dynamic range, sensitivity, and reproducibility of the Taqman probe assay was further compared with our previous cPCR results (25). ß-ACTIN mRNA was selected to standardize target mRNA amounts between patients (25). As shown in Fig. 1B, the Taqman probe assay allowed a precise quantification over a range of 5 logs [similar to cPCR (25)]; the detection level measured was 10 molecules, compared with 40 with cPCR (25). The reproducibility of the RT-PCR using the Taqman probe was examined by comparing data obtained over a large dynamic range (from 10–106) of target molecules from duplicate PCR measurements performed on one (Fig. 1C) or duplicate (Fig. 1D) set(s) of RT products. The PCR amplification proved extremely reproducible, with even better correlation factors (r 
0.997; Fig. 1C) than reported for cPCR [r = 0.970 (25)]. The RT step was also found to be highly reproducible for both MMP-9 and ß-ACTIN mRNAs (r 0.990; Fig. 1D). Based on these observations, we decided to use the Taqman probe technology for all measurements of MMP-9 and ß-ACTIN mRNA.
|
Using the Taqman probe assay, we quantified the amount of MMP-9 mRNA relative to ß-ACTIN mRNA in 57 endometria sampled throughout the normal menstrual cycle. The relative amount of MMP-9 mRNA varied considerably between patients, even within a given phase of the cycle (Fig. 2A). Despite these individual differences, the relative amount of MMP-9 mRNA was significantly (P < 0.001) increased during the perimenstrual phase (n = 14), compared with all other phases combined (n = 43). The median value was approximately 40-fold higher in the perimenstrual group (1.75 x 10–2) than in the nonperimenstrual group (4.48 x 10–4).
|
). In remarkable contrast with MMP-9 mRNA and because 14 new samples were added to our 43 previous values (25), we observed a significant increase of the relative amount of LEFTY-A mRNA during the late secretory phase (P < 0.001 when compared with all other nonperimenstrual phases), in addition to the perimenstrual phase (P < 0.001 compared with proliferative, early, and midsecretory phases; not significant when comparing late secretory with perimenstrual values). The median of the late secretory group (3.98 x 10–3) was similar to that in the perimenstrual group (3.72 x 10–3) and 140-fold higher than that of the other phases of the cycle (2.66 x 10–5).
Because of the large interpatient variations, relative amounts of MMP-9 and LEFTY-A mRNAs were compared with one another in individual samples (Fig. 2C). All samples with low relative amounts of both MMP-9 and LEFTY-A mRNAs were from nonperimenstrual and nonlate secretory phases. Most samples with high relative amounts of both MMP-9 and LEFTY-A mRNAs were from endometria showing histological signs of menstrual-like tissue breakdown (black symbols: 11 perimenstrual and two proliferative). Most samples with low relative amounts of MMP-9 mRNA, but high relative amounts of LEFTY-A mRNA were from the late secretory phase or from the perimenstrual phase (gray or black symbols). Finally, all samples with high relative amounts of MMP-9 mRNA, but low relative amounts of LEFTY-A mRNA, were from nonperimenstrual and nonlate secretory phases.
Altogether, the increase in LEFTY-A mRNA before that of MMP-9 mRNA at the end of the cycle: 1) indicates that both genes are under the control of ovarian steroids, but their response differ by specific adjustments in time and magnitude; and 2) suggests that LEFTY-A triggers the increased MMP-9 expression in perimenstrual samples. Both questions were addressed in subsequent experiments performed on cultured explants.
MMP-9 mRNA concentration increases in explant culture and is marginally controlled by combined estradiol and P
In the absence of added hormones, relative amounts of MMP-9 mRNA quickly increased in explant culture of nonperimenstrual endometrium, to reach maximal amounts after 4–8 h (Fig. 3A). Surprisingly, the addition of estradiol and P had no consistent effect on this early increase. Upon longer explant culture (for 16–48 h), high relative amounts of MMP-9 mRNA, similar to those observed in perimenstrual cycling endometria, were maintained in the absence of added hormones, reaching values up to 350-fold higher after 24 h of culture than in corresponding noncultured samples (P < 0.001), with a 21-fold median increase. Interestingly, the relative amounts of MMP-9 mRNA slightly decreased in the presence of estradiol and P after reaching the maximal value, suggesting a delayed effect of the ovarian steroids on MMP-9 mRNA transcription or stability. Indeed, after 24 h of culture, relative amounts of MMP-9 mRNA were marginally, but consistently, lower in the presence of estradiol and P (Fig. 3B) in explants from both 12 proliferative (2.5-fold in average) and 12 secretory endometria (1.6-fold).
|
Because estradiol and P weakly influenced MMP-9 mRNA concentration in explant culture, we next investigated their effect on MMP-9 protein expression and activation. The production of total MMP-9 (proMMP-9 and MMP-9) and its activation were quantified by gelatin zymography in media conditioned by explants cultured for up to 48 h in the absence or in the presence of the ovarian steroids. In their absence, MMP-9 became detectable in the medium after 8 or 16 h of culture (Fig. 4A), exclusively as the proform. The active form appeared with a delay that varied among samples. The earliest detection was after 24 h of culture; at that time, active MMP-9 was detected in 12 of 27 samples (Fig. 4B), alone or together with the proform. In contrast, both latent and active forms of MMP-2 appeared in the conditioned medium as soon as 2–4 h after the start of the culture (Fig. 4A). Addition of estradiol and P did not consistently affect total MMP-9 release (Fig. 4C) but dramatically reduced proMMP-9 activation at all times of culture (P < 0.001; Fig. 4B).
|
). Indeed, homogenates had to be concentrated about 6-fold to allow detection of zymographic activities similar to those in corresponding media. As in conditioned media, estradiol and P did not influence the amount of total MMP-9 in the tissues (n = 3). Instead, the translation inhibitor cycloheximide prevented appearance of MMP-9 in conditioned media, indicating that the released protein was synthesized during culture probably as a result of the increased mRNA levels (Fig. 5B). In all six paired comparisons tested, cycloheximide further enhanced these levels by approximately 3-fold (Fig. 5C), suggesting that the transcriptional control of MMP-9 depends on a repressor factor, which would be down-regulated at the start of culture.
|
We next investigated which cells were responsible for the increased expression of MMP-9 in the perimenstrual endometrium in vivo and in nonperimenstrual endometrium ex vivo. In noncultured endometria, MMP-9 was detected in epithelial, stromal, endothelial, and inflammatory cells throughout the menstrual cycle, without major difference between the phases (Fig. 6, A–D). Epithelial signal was prominent at all phases of the cycle. More interestingly, we observed that some clusters of stromal cells were intensively stained for MMP-9 in menstrual endometria (Fig. 6D, box at right), at foci corresponding to disruption of tissue architecture, as shown by the disappearance of argyrophilic fibers (Fig. 6F to be compared with Fig. 6E). To identify the cellular origin of MMP-9 in the stroma, sections of menstrual endometria were double-immunolabeled for MMP-9 and CD10 (fibroblasts), CD15 (leukocytes), or CD68 (macrophages). For a direct correlation, serial sections are presented in Fig. 6, H–J. Fibroblastic stromal cells were largely stained for both CD10 and MMP-9 throughout the tissue (Fig. 6H). No CD15-positive leukocytes were detected in the illustrated field (Fig. 6I); but, when present in other areas, the majority of these inflammatory cells expressed MMP-9 (inset in Fig. 6I). In contrast, CD68-positive macrophages were rarely stained for MMP-9 (Fig. 6J).
|
) and that was not appreciably modified by the addition of estradiol and P (Fig. 6K). However, some isolated stromal cells were more intensively stained for MMP-9 when explants were cultured for 24 h in the presence of the secretory pathway inhibitor, monensin. The increased intracellular signal secondary to the retention of secretory products was observed in such stromal cells, both in the absence (Fig. 6M) or presence (not shown) of ovarian steroids. In contrast, the epithelial and endothelial staining appeared unaffected by the treatment with monensin. Silver staining of adjacent sections showed that collagen fibers disappeared in explants cultured without ovarian steroids (Fig. 6O). In contrast, fibers were preserved not only in explants cultured with estradiol and P (Fig. 6N), which prevented activation of proMMP-9, but also in explants cultured with monensin (Fig. 6P), which completely prevented release of (pro)MMP-9 as assessed by zymography of conditioned medium (not shown). Recombinant LEFTY-A regulates MMP-9 expression in cultured explants
To test whether LEFTY-A can regulate MMP-9 expression, as suggested by their sequential mRNA up-regulation in vivo, recombinant LEFTY-A was added to explant culture (Fig. 7). Production and activation of MMP-9 were measured as in Fig. 4. The effect of LEFTY-A was heterogeneous between patients, leading us to discriminate two groups. In a first group (n = 9; referred to as responsive), MMP-9 release was increased by up to 3-fold upon addition of 5 ng/ml recombinant LEFTY-A (Fig. 7A, left panel, closed symbols), whereas in the other group (n = 5; referred to as nonresponsive), MMP-9 was not enhanced in response to exogenous LEFTY-A (open symbols). LEFTY-A had the same stimulatory effect on MMP-3 and -7 production in responsive endometria but not in nonresponsive ones (data not shown). The heterogeneity in responsiveness could be correlated neither to dating of the samples nor to batches of recombinant LEFTY-A. When tested in four responsive endometria, induction of MMP-9 expression was consistently measured at 5, 25, and 125 ng/ml LEFTY-A concentrations but not at lower concentrations (0.2 and 1 ng/ml; data not shown). The stimulation of MMP-9 release by LEFTY-A in explants of the responsive group was totally lost when estradiol and P were added in the culture medium (Fig. 7A, right panel). Taken together, these data suggest that LEFTY-A coordinately regulates MMP-3, -7 and -9 production and that ovarian steroids repress this response. In contrast, LEFTY-A had no effect on proMMP-9 activation, either in the absence or in the presence of ovarian steroids (Fig. 7B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
40-fold) at menstruation. These observations are consistent with previous reports based on cPCR (10, 30) and with a study by in situ hybridization, in which MMP-9 mRNA was only detected during menstruation (11), probably due to differences in the sensitivity of the techniques. Although stromal fibroblasts were shown to express MMP-9 mRNA at menstruation (11), all cell types could contribute to the menstrual mRNA increase, including neutrophils because their transcriptional activity can be stimulated after maturation (31, 32). Immunodetectable MMP-9 was found throughout the cycle in all cell types of the tissue, including stromal fibroblasts, epithelial cells, leukocytes, and macrophages, in agreement with some previous reports (14, 16). However, the predominant signal varied according to the cycle phase. Whereas MMP-9 was mostly epithelial during the proliferative and secretory phases, intense MMP-9 staining was localized in sporadic foci of stromal cells in menstrual endometrium, that is in the fragmenting tissue where collagen fibers were disrupted. MMP-9 mRNA or protein were detected by in situ hybridization or immunohistochemistry in similar foci (11, 33), most likely corresponding to areas of gelatinolytic activity identified by in situ zymography (18, 33). Whereas proMMP-9 preexisting in nonperimenstrual endometrium remained sequestered during culture, released (pro)MMP-9 was synthesized by stromal cells after the explants were put in culture as a result of a major early increase in MMP-9 mRNA. Altogether, the data strongly suggest that stromal cells are responsible for the increased expression of MMP-9 at menstruation, in foci where the endometrial ECM is degraded. Because such foci point to the existence of local regulators, we first focused on LEFTY-A, because this cytokine was initially identified as characteristic of bleeding endometria. Addition of 14 new samples to our previous analysis (25) clearly shows that the surge of LEFTY-A mRNA expression during the late secretory phase, when circulating concentrations of estradiol and P are falling, precedes the rise of MMP-9 mRNA by approximately 3 d. This delay suggests that cytokines, such as LEFTY-A, act on MMP-9 expression as local relays of the global control by ovarian steroids. The potential involvement of cytokines in the local in vivo control of MMP expression is directly supported by the demonstration that, in explants cultured without steroid hormones, recombinant LEFTY-A was sufficient to up-regulate MMP-9 expression, as well as that of other related MMPs such as MMP-3 and -7, in the majority of the samples tested. In contrast, LEFTY-A had no effect on MMP-9 activation. The amount and forms of endogenous LEFTY-A were not measured in the present study and may explain the unresponsiveness of a minority of samples to the addition of recombinant form. The mechanism by which LEFTY-A up-regulates expression of MMP-9, -3, and -7 remains speculative. LEFTY-A may act by activating the MAPK pathway (24) or by inhibiting the TGF-ß pathway (23).
Although we provide evidence that LEFTY-A can act as an important regulator of endometrial ECM breakdown, other cytokines can regulate MMPs expression. IL-1
is present in the human endometrium throughout the cycle (with cyclical modulations) and controls MMP expression in cultures of endometrial cells (34, 35). TNF- rapidly but transiently induces the expression of (pro)MMP-9 by purified stromal cells (36). TGF-ß was proposed to act as a relay of P in the inhibition of MMP-3 and -7 production by endometrial cells (37, 38). Because LEFTY-A can negatively interact with the TGF-ß pathway (23), the absence of responsiveness to recombinant LEFTY-A in some explants could also result from the presence of endogenous TGF-ß.
How are ovarian steroids and cytokine networks integrated to control MMP expression in the human endometrium? The control of local cytokines by ovarian steroids was best demonstrated in cultures of proliferative endometria that showed a strong (17-fold) inhibition of LEFTY-A mRNA expression by combined estradiol and P, compared with culture without hormones (25). The marginal nonsignificant effect of the addition of ovarian steroids to secretory endometria is attributed to residual endogenous P. A weaker inhibition was observed for MMP-9 mRNA (2-fold) at both phases of the cycle, starting at 24 h of culture, whereas ovarian steroids had no effect on the protein level during the first day of culture. However, we previously observed a slight inhibition of (pro)MMP-9 release by explants cultured for 2–3 d with combined estradiol and P (3), which is consistent with the delayed inhibition observed at the mRNA level.
P causes a dual inhibition on MMP-1 production in the human endometrium, both by preventing the release of IL-1 and by suppressing the stimulating effect of this cytokine on MMP-1 (34). Our observations (Ref. 25 and the present report) demonstrate that the control by LEFTY-A of MMP-3 and -7 (and to a limited extent, of MMP-9) represents another target for a dual inhibition by estradiol and P, upstream and downstream of the cytokine. In contrast, the major effect of ovarian steroids with respect to proMMP-9 is on its activation; it reflects inhibition by these hormones of the expression of the MMP-9 activator, MMP-3 (13).
To our surprise, the delay between LEFTY-A and MMP-9 mRNA increase was no longer observed in explant culture. The further increase of MMP-9 mRNA in the presence of cycloheximide indicates that early MMP-9 expression in cultured explants is not induced by a newly expressed protein but rather results, at least partially, from the decreased expression of a repressor whose continuous synthesis is required to control MMP-9 expression. Decreased expression of this repressor could result from: 1) the absence in explant culture of factors present in vivo (for instance, circulating factors); 2) the relaxation of mechanical tension in the dissected tissue, which was shown to induce expression of several MMPs (39); and/or 3) the premature (or artificial) activation or release of potential inducers, such as reactive oxygen species (40) or cytokines already present in the endometrium in an inactive state. There is increasing evidence that nuclear factor (NF)-
B is involved in the regulation of cytokines and MMPs, including MMP-9, in the human endometrium (reviewed in Ref. 41). NF-B has been recently proposed as a key regulator for the initiation of menstruation, based on the observation that withdrawal of the ovarian steroids from cultures of endometrial stromal cells stimulated the production of prostaglandin F2 through a reactive oxygen species-induced NF-B activation (40). Moreover P receptor can directly interact with the RelA (p65) subunit of NF-B (42), and progestins have been shown to inhibit cytokine expression through a NF-B-mediated mechanism (43).
Despite their limitations, experiments in explant culture mimic several in vivo observations and provide important clues to suggest that MMP-9 activity in the human endometrium in vivo is controlled overall by estradiol and P, finely tuned by local inducers, including cytokines such as LEFTY-A, and potentially repressors such as TGF-ß, and ultimately regulated by the control of activator(s). This hypothesis is entirely consistent with the emerging concept that regulators of MMP expression and activity in the human endometrium are integrated in a complex network. According to this view, up-regulation of menstrual MMPs, toward the end of the endometrial cycle, results from: 1) the local clearance of receptors for ovarian steroids; and/or 2) the increased production of MMP-inducing cytokines such as IL-1 and LEFTY-A in response to the fall in circulating ovarian steroids. Clearly, the net effect depends on the relative concentration between the different partners of this network. The exhaustive identification of these partners and their interactions will require considerable additional investigations.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: AP, Alkaline phosphatase; cPCR, competitive PCR; E, 17ß-estradiol; EBAF, endometrial bleeding-associated factor; ECM, extracellular matrix; MMP, matrix metalloproteinase; NF, nuclear factor; P, progesterone; RT, reverse transcription.
This work was supported by grants from the Belgian Fonds de la Recherche Scientifique Médicale (3.4555.02 to E.M.), from Interuniversity Attraction Poles and Concerted Research Actions (to P.J.C.), and from Organon (to E.M.). P.B.C. is a Research Fellow, and P.H. a Research Associate of the Belgian Fonds National de la Recherche Scientifique.
Received July 1, 2004.
Accepted October 28, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
and its dual block by ovarian steroids. Proc Natl Acad Sci USA 94:10341–10345 production through nuclear factor-B activation via oxygen radicals in human endometrial stromal cells: potential relevance to menstruation. J Reprod Dev 50:215–225[CrossRef][Medline]
B and the progesterone receptor. J Biol Chem 271:6217–6224 B and steroid receptor-signaling pathways. Endocr Rev 20:435–459This article has been cited by other articles:
 |
|
 |
|
  H. P. Gaide Chevronnay, C. Galant, P. Lemoine, P. J. Courtoy, E. Marbaix, and P. Henriet Spatiotemporal Coupling of Focal Extracellular Matrix Degradation and Reconstruction in the Menstrual Human Endometrium Endocrinology, November 1, 2009; 150(11): 5094 - 5105. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Selvais, H. P. Gaide Chevronnay, P. Lemoine, S. Dedieu, P. Henriet, P. J. Courtoy, E. Marbaix, and H. Emonard Metalloproteinase-Dependent Shedding of Low-Density Lipoprotein Receptor-Related Protein-1 Ectodomain Decreases Endocytic Clearance of Endometrial Matrix Metalloproteinase-2 and -9 at Menstruation Endocrinology, August 1, 2009; 150(8): 3792 - 3799. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Pretto, H. P. Gaide Chevronnay, P. B. Cornet, C. Galant, D. Delvaux, P. J. Courtoy, E. Marbaix, and P. Henriet Production of Interleukin-1{alpha} by Human Endometrial Stromal Cells Is Triggered during Menses and Dysfunctional Bleeding and Is Induced in Culture by Epithelial Interleukin-1{alpha} Released upon Ovarian Steroids Withdrawal J. Clin. Endocrinol. Metab., October 1, 2008; 93(10): 4126 - 4134. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. P. Gaide Chevronnay, P. B. Cornet, D. Delvaux, P. Lemoine, P. J. Courtoy, P. Henriet, and E. Marbaix Opposite Regulation of Transforming Growth Factors-{beta}2 and -{beta}3 Expression in the Human Endometrium Endocrinology, March 1, 2008; 149(3): 1015 - 1025. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. Cederroth, O. Schaad, P. Descombes, P. Chambon, J.-D. Vassalli, and S. Nef Estrogen Receptor {alpha} Is a Major Contributor to Estrogen-Mediated Fetal Testis Dysgenesis and Cryptorchidism Endocrinology, November 1, 2007; 148(11): 5507 - 5519. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Collette, R. Maheux, J. Mailloux, and A. Akoum Increased expression of matrix metalloproteinase-9 in the eutopic endometrial tissue of women with endometriosis Hum. Reprod., December 1, 2006; 21(12): 3059 - 3067. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Malik, K. Day, I. Perrault, D.S. Charnock-Jones, and S. K. Smith Reduced levels of VEGF-A and MMP-2 and MMP-9 activity and increased TNF-{alpha} in menstrual endometrium and effluent in women with menorrhagia Hum. Reprod., August 1, 2006; 21(8): 2158 - 2166. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Wahab, A.H. Taylor, J.H. Pringle, J. Thompson, and F. Al-Azzawi Trimegestone differentially modulates the expression of matrix metalloproteinases in the endometrial stromal cell Mol. Hum. Reprod., March 1, 2006; 12(3): 157 - 167. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Vassilev, C. M. Pretto, P. B. Cornet, D. Delvaux, Y. Eeckhout, P. J. Courtoy, E. Marbaix, and P. Henriet Response of Matrix Metalloproteinases and Tissue Inhibitors of Metalloproteinases Messenger Ribonucleic Acids to Ovarian Steroids in Human Endometrial Explants Mimics Their Gene- and Phase-Specific Differential Control in Vivo J. Clin. Endocrinol. Metab., October 1, 2005; 90(10): 5848 - 5857. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
