Molecular Human Reproduction vol.5 no.3 pp. 240–245, 1999

Interleukin-1 receptor-ligand interactions modulate interstitial collagenase-1 production by human endometrial fibroblasts

C.F.Singer1,3, E.Marbaix1,2, P.Lemoine1, Y.Eeckhout1 and P.J.Courtoy1,4 1Cell Biology Unit, Christian de Duve Institute of Cellular Pathology, Avenue Hippocrate 75 and 2Department of Pathology, Saint-Luc University Clinics, Louvain University Medical School, Avenue Hippocrate 10, B-1200 Brussels, Belgium 3 Present address: Division of Special Gynaecology, Department of Obstetrics and Gynaecology, University of Vienna Downloaded from https://academic.oup.com/molehr/article/5/3/240/1404548 by guest on 01 October 2021 Medical Center, Waehringer Guertel 18–20, A-1090 Vienna, Austria 4To whom correspondence should be addressed The expression of interstitial collagenase-1 in the cycling human endometrium is restricted to the perimenstrual phase and is a key event for matrix degradation that initiates menstruation. In the absence of ovarian steroids, collagenase production by endometrial fibroblasts is induced by epithelial cell-derived -1α. Media conditioned by endometrial epithelial cells were found to contain interleukin-1α but not interleukin-1β, and their capacity to induce collagenase production by endometrial fibroblasts correlated with interleukin-1α concentration in a saturable manner. Collagenase induction by recombinant interleukin-1α was severely inhibited by interleukin-1 receptor antagonist alone and abolished by its combination with soluble interleukin- 1 type-II receptor. By contrast, the association of the receptor antagonist with soluble type-I receptor was less effective than each factor alone. Induction of collagenase by epithelial cell-conditioned media was severely inhibited by neutralizing interleukin-1α antibodies, whereas the combination of receptor antagonist with soluble type-II receptor proved less effective. We conclude that the collagenase response of endometrial fibroblasts to epithelial cell-derived interleukin-1α is effectively blocked in vitro by soluble members of the interleukin-1 family and can thus be modulated in vivo by these or other local factors. Key words: endometrium/IL-1 receptor antagonist/interleukin-1/interstitial collagenase-1/soluble IL-1 receptor

Introduction neutralizing IL-1α, IL-1β, or IL-1Ra (Symons et al., 1993, -1 (IL-1) are classical ‘multifunctional’ : 1994; Arend et al., 1994). This complex array of interacting they are produced by, and affect, a wide variety of cell types. factors could finely modulate IL-1α responses in vivo (Burger In vivo, IL-1 exert a wide range of systemic effects such as et al., 1995). hypotension, fever, weight loss, and neutrophilia, but also a We have recently identified IL-1α as a major paracrine variety of local responses such as the controlled expression of stimulator of the production of interstitial collagenase (matrix growth factors, extracellular matrix components, and matrix metalloproteinase-1; MMP-1) by human endometrial metalloproteinases (for a review, see Dinarello, 1996). fibroblasts (Singer et al., 1997). Generally secreted as inactive Although IL-1 production in humans is generally considered proenzymes, MMPs are activated by limited proteolysis and to result from inflammation or tissue injury, evidence points degrade virtually all extracellular matrix proteins (for reviews, to a transient upregulation in embryogenesis (Sheth et al., see Woessner, 1991; Murphy and Reynolds, 1993; Hulboy 1991) and during the menstrual cycle (Cannon and Dinarello, et al., 1997). A number of observations suggest that among 1985; Polan et al., 1994). MMPs, interstitial collagenase (MMP-1) is a key enzyme in The IL-1 family comprises IL-1α, IL-1β, as well as the IL- the initiation of menstrual tissue degradation. In the normal 1 receptor antagonist (IL-1Ra). Two types of IL-1 receptor eutopic endometrium, MMP-1 expression is restricted to the have been identified in humans. Although IL-1α and IL-1β perimenstrual phase and confined to foci of fibroblast-like share little homology in primary sequence, the two cytokines stromal cells in the functionalis layer of the human endomet- exert their agonist effects through binding to the same type I rium, that is destined to be shed (Marbaix et al., 1995, 1996a; receptor (IL-1RI). Type II receptor (IL-1RII) also binds IL-1 Kokorine et al., 1996). Furthermore, inhibition of MMP-1 and but does not transduce a signal (Stylianou et al., 1992; Sims related matrix metalloproteinases prevents tissue degradation et al., 1993). The receptor antagonist IL-1Ra binds to IL-1RI in endometrial tissue explants (Marbaix, et al., 1996b). with an affinity comparable to that of IL-1α and IL-1β, but Considering the pivotal role of MMP-1 in initiating menstru- does not elicit biological responses (Arend and Dayer, 1990; ation, interference with its IL-1α-mediated stimulation should Hannum et al., 1990). Furthermore, extracellular fragments of have profound effects on endometrial function and morphology both receptor types (IL-1sRI and IL-1sRII) circulate in the in vivo. To investigate whether biological agents could modu- blood and function as natural ‘scavengers’ by binding and late IL-1 receptor-mediated MMP-1 production in endometrial

240 © European Society of Human Reproduction and Embryology Interleukin-1 regulates endometrial collagenase stromal cells, we analysed cultures of epithelial and of fibroblast cells isolated from human endometrial biopsies.

Materials and methods Cell cultures The study was approved by the Ethical Committee of the University of Louvain, Belgium. Endometrial tissue samples, derived from patients who had undergone hysterectomies or biopsies for non- endometrial pathologies, were generally obtained during the secretory phase. Epithelial cells and fibroblasts were then isolated as previously described (Zhang et al., 1995). In brief, fresh endometrial biopsies were washed with phosphate-buffered saline (PBS) and digested with Downloaded from https://academic.oup.com/molehr/article/5/3/240/1404548 by guest on 01 October 2021 400 IU/ml of collagenase IA (Sigma-Aldrich, Bornem, Belgium) at 37°C for 45 min. Fibroblasts were then separated from endometrial Figure 1. Inhibition by interleukin-1 receptor antagonist (IL-1Ra) glands by sequential filtration through nylon meshes with pore sizes of IL-1α- and IL-1β-induced matrix metalloproteinase-1 (MMP-1) of 300 and 35 µm respectively. Epithelial glands, collected on the production by endometrial fibroblasts. Confluent monolayers of second filter, were sedimented twice, then digested with trypsin. fibroblasts were cultured with 6 pM nominal concentration of recombinant human (rh)IL-1α or rh-IL-1β in the absence (–Ra) or Immunolabelling for vimentin and cytokeratin of isolated cells shortly in the presence (ϩ Ra) of a 1000-fold molar excess of rhIL-1Ra. after plating confirmed that fibroblasts were essentially pure and that Supernatants were collected after 24 h and collagenase activity was epithelial cells were highly enriched. measured. Values shown are means Ϯ SD (n ϭ 3) of samples from Fibroblasts were plated into culture flasks and propagated for 2–6 one representative patient. Whereas in absolute values the basal weeks (i.e. 3–5 passages) in Phenol Red-free Dulbecco’s minimal MMP-1 production and its response to IL-1α and β varied among essential medium (DMEM)/Ham’s F12 medium containing 100 IU/ patients, the level of inhibition by Ra was comparable in identical ml penicillin, 100 IU/ml streptomycin, 0.25 mg/ml amphotericin B, experiments performed on endometrial fibroblasts from two other and 10% fetal calf serum (FCS; all from Gibco Europe, Merelbeke, patients. Belgium). All experiments were carried out on confluent fibroblastic monolayers after three washes with PBS and replacement by medium devoid of FCS. Freshly isolated epithelial cells were cultured in IL-1α and IL-1β ELISA medium with FCS for 2–7 days, washed three times with PBS and IL-1α and IL-1β concentrations in media conditioned by epithelial further cultured for 24–48 h in medium without FCS. The latter cells for 24 or 48 h were determined in microtitre plates by medium, referred to as ‘conditioned by epithelial cells’, was then added an immunoassay (R&D Systems) according to the manufacturer’s to fibroblast monolayers in the absence or presence of recombinant protocol. Optical densities were estimated using a Titertek Multiscan proteins, as indicated. Alternatively, conditioned media were incubated MCC 340 spectrophotometer (Flow Laboratories, Doornveld, with monoclonal anti-IL-1α antibodies for 1 h at 20°C before addition Belgium). to fibroblasts. Curvilinear fitting Recombinant proteins and antibodies The relationship between IL-1α concentration and MMP-1 production Recombinant human IL-1α, IL-1β, IL-1Ra, IL-1sRI, and IL-1sRII was adjusted by least-square curvilinear fitting based on equation 1: were purchased from R&D Systems (Abingdon, UK) and added at MMP-1maxϫIL-1αi the indicated nominal concentrations (n.c.; due to adsorption on MMP-1 ϭ i α plastic at very low concentrations, effective values may fall very Ks ϩ IL-1 i much below nominal values; for example, a 6 pM n.c. of rhIL-1α where MMP-1i and MMP-1max are actual and maximal total colla- corresponded to a ~0.6 pM effective concentration, as measured by genase activities, expressed in IU/ml respectively; IL-1αi is the actual enzyme-linked immunosorbent assay (ELISA); this is close to max- concentration measured in conditioned medium applied to fibroblasts, imal stimulating concentration in our assays, (Singer et al., 1997). expressed in fM; Ks is the calculated half-maximal stimulatory IL- The monoclonal mouse anti-human IL-1α antibody (clone IC12.1) 1α concentration, expressed in fM. was purchased from Calbiochem (Bierges, Belgium).

Collagenase assays Results Total collagenase activity was determined at 25°C with [3H]-acetylated Recombinant IL-1α- and IL-1β-induced MMP-1 produc- collagen in solution (Eeckhout et al., 1986). One unit of collagenase tion by endometrial fibroblasts is suppressed by is defined as the amount of enzyme which degrades 1 µg of soluble IL-1Ra collagen/min. Full collagenase activation was achieved by treating To determine the respective ability of IL-1α and IL-1β to the conditioned media with 2 mM 4-aminophenylmercuric acetate stimulate MMP-1 production in the human endometrium, the (APMA; Sigma-Aldrich) for 2 h at 37°C. Antibodies directed against MMP-1 (a kind gift of Dr H.Nagase, Kansas University, Kansas City, recombinant cytokines were added at 6 pM n.c. to confluent KS, USA) abolished the collagenase activity in media conditioned fibroblast monolayers and MMP-1 production after 24 h of by endometrial fibroblasts, indicating that other collagenases such as culture was measured by collagenase assays (Fig 1). When MMP-8 and MMP-13 did not contribute to the collagenolytic activity added at equimolar concentrations, both cytokines augmented in this assay (data not shown). MMP-1 production with comparable potency. Addition of 241 C.F.Singer et al. Downloaded from https://academic.oup.com/molehr/article/5/3/240/1404548 by guest on 01 October 2021

Figure 3. Inhibition by interleukin-1 receptor antagonist (IL-1Ra) and soluble receptors of IL-1α-induced matrix metalloproteinase-1 Figure 2. Correlation between interleukin-1α (IL-1α) release from (MMP-1) production by endometrial fibroblasts. Confluent endometrial epithelial cells and resulting matrix metalloproteinase-1 fibroblast monolayers were incubated for 24 h with 30 pM nominal α (MMP-1) production by stromal cells. Epithelial cells were isolated concentration of recombinant human (rh) IL-1 , in the absence or from biopsies obtained during the secretory phase in eight patients. presence of 30 nM rhIL-1Ra (Ra), 150 nM rhIL-1sRI (SRI), and 30 After culture for various intervals (2–7 days) in the presence of nM rhIL-1sRII (SRII) alone, or in the indicated combinations. The fetal calf serum (3–5 intervals per patient), cells from each patient resulting MMP-1 induction was measured by collagenase assays. were transferred to serum-free media and cultured further for 24 h Values shown are mean Ϯ SD (n ϭ 3) of samples from one (closed circles, 1–4 media per patient, n ϭ 19) or 48 h (open representative patient. Whereas in absolute values the basal MMP-1 α circles, 1–3 media per patient, n ϭ 9). Conditioned media were production and its response to IL-1 varied among patients, the collected, assayed for IL-1α by enzyme-linked immunosorbent level of inhibition was comparable in an identical experiment assay and added to endometrial fibroblasts isolated from a single performed on endometrial fibroblasts from another patient. donor, for comparison. The resulting MMP-1 production by the fibroblasts was measured by collagenase assays after 24 h. All values are means of two measurements per medium with Ͻ10% IL-1α-induced MMP-1 production by endometrial variation. The correlation curve was drawn by curvilinear best- fibroblasts is inhibited by IL-1Ra and IL-1-soluble square fitting as explained in the text. receptors To further explore the possibility of inhibiting endometrial MMP-1 expression by interfering with the binding of IL-1α ~1000-fold molar excess of the IL-1 receptor antagonist (rhIL- to its receptor, the effects of a 1000-fold molar excess of rhIL- 1Ra) severely inhibited the stimulatory effect of either . 1Ra, rhIL-1sRI, and rhIL-1sRII, alone or in combination, on rhIL-1α-induced MMP-1 production by endometrial fibroblasts Induction of fibroblast-derived MMP-1 by epithelial was investigated (Figure 3). As shown in Figure 1, endometrial cell-conditioned medium correlates with its IL-1 fibroblasts produced only marginal amounts of MMP-1 in α α concentration in a saturable manner the absence of added IL-1 and this production increased considerably upon the addition of rhIL-1α. When rhIL-1Ra α β To analyse the natural release of IL-1 and IL-1 by human was added together with rhIL-1α, MMP-1 induction was endometrial cells, the concentration of both cytokines was greatly reduced, whereas rhIL-1sRI or rhIL-1sRII were weak measured in media conditioned by epithelial cells derived from inhibitors by themselves. Surprisingly, a combination of rhIL- endometrial biopsies obtained during the secretory phase, and 1Ra with rhIL-1sRI resulted in a less potent inhibition of cultured in the absence of ovarian steroids (Figure 2). Whereas rhIL-1α-induced MMP-1 expression than rhIL-1Ra alone. By β IL-1 in conditioned media was always below the level of contrast, the combination of rhIL-1Ra with rhIL-1sRII almost α detection (Ͻ60 fM; data not shown), IL-1 was found in all abolished MMP-1 induction. Again, the addition of rhIL-1sRI media but one, its maximum concentration being 1.8 pM. to this combination weakened its inhibition of rhIL-1α-induced As previously reported (Singer et al., 1997), epithelial- MMP-1 production. conditioned media strongly induced MMP-1 production. The capacity of these media to induce collagenase (MMP-1) MMP-1 induction in fibroblasts by epithelial cell- production by fibroblasts showed an excellent curvilinear conditioned media is partially repressed by a com- correlation (r ϭ 0.97) with the IL-1α concentration, present bination of IL-1Ra and IL-1sRII in the media. Curvilinear least-square fitting indicates that the The effectiveness of the combination of rhIL-1Ra with rhIL- estimated half-stimulatory concentration was as low as 53 Ϯ 1sRII to inhibit MMP-1 production by fibroblasts upon stimula- 35 fM (mean with 95% confidence intervals). In contrast to tion by epithelial cell-conditioned medium was next examined their inhibitory effect on explants (Singer et al., 1997), ovarian (Figure 4). Conditioned medium did not contain detectable steroids failed to decrease IL-1α release by cultured epithelial MMP-1, and only little collagenase activity was spontaneously cells in our experimental conditions (data not shown). produced by endometrial fibroblasts cultured without condi- 242 Interleukin-1 regulates endometrial collagenase

IL-1β immunostaining were not consistent (Tabibzadeh and Sun, 1992; Simo´n et al., 1993). Whereas explants from endometrial biopsies obtained during the proliferative or the secretory phases of the menstrual cycle and cultured for 24 h in the absence of ovarian steroids do not release measurable IL-1α into the culture medium, explants from biopsies obtained during the perimenstrual phase release considerable amounts of this cytokine over the same period (Singer et al., 1997). Upon longer culture times of non- menstrual explants in the absence of ovarian steroids, an appreciable release of IL-1α becomes apparent, a phenomenon that can be partially suppressed by the addition of physiological concentrations of 17β-oestradiol and progesterone (Singer Downloaded from https://academic.oup.com/molehr/article/5/3/240/1404548 by guest on 01 October 2021 et al., 1997). In contrast, purified epithelial cells cultured in Figure 4. Interference by the interleukin-1 (IL-1) family on the the absence of ovarian steroids also released IL-1α, but this induction by epithelial-derived conditioned medium of matrix release was not inhibited by the addition of 17β-oestradiol and metalloproteinase-1 (MMP-1) production by endometrial fibroblasts. Collagenase activity was measured in supernatants of confluent progesterone (data not shown). Further studies are clearly fibroblasts stimulated by epithelial cell-conditioned medium in the needed to determine whether the unresponsiveness of isolated absence or presence of 6 nM monoclonal anti-human IL-1α epithelial cells to ovarian steroids is due to the absence of antibody (IL-1α MAb), or a combination of 60 nM rhIL-1Ra (Ra) stromal–epithelial interactions (Cooke et al., 1997) and/or to with 150 nM rhIL-1sRII (SRII). Values shown are mean Ϯ SD the loss of ovarian steroid receptors upon culture. (n ϭ 3) of samples from one representative patient. Whereas in α β absolute values, the basal MMP-1 production and its response to The interaction of IL-1 and IL-1 with IL-1RI on target conditioned medium varied among patients, the level of inhibition cells can be modulated by receptor expression on the cell was comparable in identical experiments performed on endometrial surface and counterbalanced by the soluble inhibitory factors fibroblasts from two other patients. IL-1Ra, IL-1sRI, and IL-1sRII. Although IL-1Ra has nearly the same binding affinity to IL-1RI as IL-1α and IL-1β at equilibrium, its rate of association with IL-1RI is slower, so tioned medium. As shown in Figure 2, addition of epithelial that, when IL-1Ra dissociates from its cell surface receptor, cell-conditioned medium to endometrial fibroblasts greatly IL-1 will preferentially bind to the empty receptor (Arend, increased MMP-1 production. Whereas this induction could be 1993). This kinetic peculiarity might explain why IL-1α almost abolished following incubation of conditioned medium activity is inhibited only by a high molar excess of IL-1Ra α with a monoclonal anti-human IL-1 , incubation of conditioned (Arend et al., 1990). In our experiments, the presence of a medium with a precombination of rhIL-1Ra and rhIL-1sRII 1000-fold molar excess of IL-1Ra largely inhibited the stimula- was much less effective. tory effects of both IL-1α and IL-1β on MMP-1 production in human endometrial fibroblasts. Various combinations of IL-1Ra, IL-1sRI, and/or IL-1sRII, Discussion had distinct inhibitory effects on IL-1-stimulated fibroblasts. Both IL-1α and IL-1β are strong stimulators of MMP-1 Whereas the addition of IL-1α strongly induced MMP-1 production by fibroblasts of various tissues (Burger et al., production, this response was largely reversed by a 1000-fold 1995). Probably due to the presence of an IL-1 receptor molar excess of IL-1Ra and its combination with IL-1sRII accessory protein (IL-1R-AcP), which increases IL-1 binding proved most effective. By contrast, the combination of IL-1Ra affinity to IL-1RI (Greenfeder et al., 1995), and especially with IL-1sRI produced a considerably less potent inhibition. It because of powerful post-receptor amplification through might seem paradoxical that the combination of two inhibitory multiple phosphorylations by protein kinases (Croston et al., proteins, one neutralizing the ligand and the other competing 1995), occupancy of only a few receptors per cell may suffice with the receptor, is less powerful than each factor alone. to elicit maximal biological responses (O’Neill, 1995). In However, since IL-1Ra binds to IL-1sRI with an ~200-fold agreement with this concept, we found a half-maximal stimula- higher affinity than to IL-1RI (McMahan et al., 1991), addition tion of MMP-1 production at a concentration of ~50 fM IL- of IL-1sRI will neutralize the majority of the IL-1Ra (which 1α, i.e. several orders of magnitude below the KD of the rather binds to the soluble than to the membrane-attached ligand-receptor interaction. receptor), and therefore decrease its capacity to compete with When the concentrations of IL-1α and IL-1β were deter- IL-1α for binding to IL-1RI (Burger et al., 1995). mined in media conditioned by endometrial epithelial cells, The ability of the combination of IL-1Ra with IL-1sRII to only IL-1α was measurable in concentrations capable of inhibit MMP-1 production by fibroblasts was almost complete stimulating MMP-1 production by endometrial fibroblasts when induction was by recombinant IL-1α, but proved much (Singer et al., 1997); the concentration of IL-1β was always less able to antagonize the induction caused by the more below the level of detection (Ͻ60 fM). These findings are in complex epithelial cell-conditioned media. This weaker inhibi- agreement with previous studies which detected IL-1α in the tion cannot be ascribed to a high concentration of IL-1α epithelial cells of the human endometrium, while reports on in conditioned media (which was never Ͼ2 pM). Since a 243 C.F.Singer et al. neutralizing monoclonal anti-IL-1α antibody almost abolished and IL-1 receptor antagonist by soluble IL-1 receptors and levels of soluble IL-1 receptors in synovial fluids. J. Immunol., 153, 4766–4774. MMP-1 induction by conditioned media, the observation rather Bendtzen, K., Svenson, M., Jonsson, V. et al. (1990) Autoantibodies to suggests that additional factors are released by endometrial cytokines – friends or foes? Immunol. Today, 11, 167–169. cells and interfere with the inhibitory effect of IL-1Ra and IL- Burger, D., Chicheportiche, R., Giri, J.G. et al. (1995) The inhibitory 1sRII. One such factor might be IL-1sRI, known to circulate activity of human interleukin-1 receptor antagonist is enhanced by type II interleukin-1 receptor and hindered by type I interleukin-1 soluble receptor. in the blood of healthy humans at concentrations of 10–20 pM J. Clin. Invest., 96, 38–41. (Dinarello, 1996). However, the degree of IL-1sRI production Burrows, T.D., King, A. and Loke, Y.W. (1996) Trophoblast migration during and its local concentration in the human endometrium are human placental implantation. Hum. Reprod. Update, 2, 307–321. α Cannon, J.G. and Dinarello, C.A. (1985) Increased plasma interleukin-1 unknown. Alternatively, neutralizing anti-human IL-1 autoan- activity in women after ovulation. Science, 227, 1247–1249. tibodies, which function as natural buffers of IL-1α, have been Cooke, P.S., Buchanan, D.L., Young, P. et al. (1997) Stromal estrogen receptors found in up to 25% of healthy subjects. It is thus conceivable mediate mitogenic effects of estradiol on uterine epithelium. Proc. Natl. that autoantibodies similarly exist for IL-1Ra and IL-1sRII Acad. Sci. USA, 94, 6535–6540. Croston, G.E., Cao, Z. and Goeddel, D.V. (1995) NFkB activation by Downloaded from https://academic.oup.com/molehr/article/5/3/240/1404548 by guest on 01 October 2021 (Bendtzen et al., 1990; Svenson et al., 1990). interleukin-1 requires an IL-1 receptor-associated protein kinase activity. To what extent interactions such as those analysed in J. Biol. Chem., 270, 16514–16517. this report actually modulate IL-1α signalling in the human Dinarello, C.A. (1996) Biological basis for interleukin-1 in disease. Blood, 87, 2095–2147. endometrium, so as to affect production of other cytokines, Dower, S.K., Fanslow, W., Jacobs, C. et al. (1994) Interleukin-1 antagonists. e.g. tumour necrosis factor α (Laird et al., 1996), to finely Therapeutic Immunol., 1, 113–122. tune tissue differentiation in the functionalis layer, and possibly Eeckhout, Y., Delaisse´, J.M. and Vaes, G. (1986) Direct extraction and assay to regulate implantation efficiency (Simo´n et al., 1996) and of bone tissue collagenase and its relation to parathyroid-hormone-induced bone resorption. Biochem. J., 239, 793–796. trophoblast invasion (Burrows et al., 1996) will await further Granowitz, E.V., Porat, R., Mier, J.W. et al. (1992) Pharmacokinetics, safety, investigations. From a practical point of view, however, inhibi- and immunomodulatory effects of human recombinant interleukin-1 receptor tion of MMP-1-induced matrix breakdown may emerge as an antagonist in healthy humans. Cytokine, 4, 353–360. Greenfeder, S.A., Nunes, P., Kwee, L. et al. (1995) Molecular cloning and interesting therapeutical possibility to control dysfunctional characterization of a second subunit of the interleukin-1 receptor complex. endometrial bleeding (C.Galant et al., unpublished results), J. Biol. Chem., 270, 13757–13765. or to prevent endometriotic tissue remodelling and scarring Hannum, C.H., Wilcox, C.J., Arend, W.P. et al. (1990) Interleukin-1 receptor (Kokorine et al., 1997). Indeed, prevention of MMP-1-induced antagonist activity of a human interleukin-1 inhibitor. Nature, 343, 336–340. Hulboy, D.L., Rudolph, L.A. and Matrisian, L.M. (1997) Matrix breakdown can be achieved at the level of ovarian steroids, metalloproteinases as mediators of reproductive function. Mol. Hum. their IL-1α relay, or by direct inhibition of MMP-1 by Reprod., 3, 27–45. pharmacological inhibitors (Marbaix et al., 1996b). Since high Kokorine, I., Marbaix, E., Henriet, P. et al. (1996) Focal cellular origin and regulation of interstitial collagenase (matrix metalloproteinase-1) are related levels of IL-1Ra and IL-1sRII can be administered to humans to menstrual breakdown in the human endometrium. J. Cell Sci., 109, without noticeable side-effects (Granowitz et al., 1992; Dower 2151–2160. et al., 1994), they appear to be interesting candidates for Kokorine, I., Nisolle, M., Donnez, J. et al. (1997) Expression of interstitial collagenase (matrix metalloproteinase-1) is related to the activity of human MMP-1 suppression. If this intervention is envisaged, attempts endometriotic lesions. Fertil. Steril., 68, 246–251. to interfere with IL-1α-induced MMP-1 production in the Laird, S.M., Tuckerman, E.M., Saravelos, H. et al. (1996) The production of human endometrium should however consider the multiple tumour necrosis factor α (TNF-α) by human endometrial cells in culture. and complex interactions that may modulate the powerful Hum. Reprod., 11, 1318–1323. α Marbaix, E., Kokorine, I., Henriet, P. et al. (1995) The expression of interstitial effects of IL-1 on this tissue. collagenase in human endometrium is controlled by progesterone and by oestradiol and is related to menstruation. Biochem. J., 305, 1027–1030. Marbaix, E., Kokorine, I., Donnez, J. et al. (1996a) Regulation and restricted expression of interstitial collagenase suggest a pivotal role in the initiation Acknowledgements of menstruation. Hum. Reprod., 11 (Suppl. 2), 134–143. The authors wish to thank Professor J.Donnez and his co-workers Marbaix, E., Kokorine, I., Moulin, P. et al. (1996b) Menstrual breakdown of for providing endometrial tissues and Dr H.Nagase for the kind gift human endometrium can be mimicked in vitro and is selectively and of anti-MMP-1 antiserum. C.F.S was a recipient of the Haas-Teichen reversibly blocked by inhibitors of matrix metalloproteinases. Proc. Natl. Postdoctoral Fellowship of the Christian de Duve Institute of Cellular Acad. Sci. USA, 93, 9120–9125. Pathology, Brussels, Belgium. This work was supported by the McMahan, C.J., Slack, J., Cosman, D. et al. (1991) A novel IL-1 receptor, Belgian Fonds de la Recherche Scientifique Me´dicale, and by the cloned from B cells by mammalian expression, is expressed in many cell Fonds de De´veloppement Scientifique of the University of Louvain. types. EMBO J., 10, 2821–2832. This paper represents research results of the Belgian Programme on Murphy, G. and Reynolds, J.J. (1993) Extracellular matrix degradation. In Royce, P.M. and Steinmann, B. (eds), Connective Tissue and its Heritable Interuniversity Poles of Attraction and of Concerted Research Actions Disorders. Wiley-Liss, New York, p. 287. of the ‘Communaute´ Franc¸aise de Belgique’. O’Neill, L.A. (1995) Towards an understanding of the signal transduction pathways for interleukin-1. Biochim. Biophys. Acta, 1266, 31–44. Polan, M.L., Loukides, J.A. and Honig, J. (1994) Interleukin-1 in human References ovarian cells and in peripheral blood monocytes increases during the luteal phase: Evidence of a midcycle surge in the human. Am. J. Obstet. Gynaecol., Arend, W.P. and Dayer, J.-M. (1990) Cytokines and cytokine inhibitors or 170, 1000–1006. antagonists in rheumatoid arthritis. Arthritis Rheum., 33, 305–315. Sheth, K.V., Roca, G.L., Al-Sedairy, S.T. et al. (1991) Prediction of successful Arend, W.P., Welgus, H.G., Thompson, R.C et al. (1990) Biological properties embryo implantation by measuring interleukin-1-alpha and of recombinant human monocyte-derived interleukin-1 receptor antagonist. immunosuppressive factor(s) in preimplantation embryo culture fluid. Fertil. J. Clin. Invest., 85, 1694–1697. Steril., 55, 952–957. Arend, W.P. (1993) Interleukin receptor antagonist. Adv. Immunol., 54, Simo´n, C., Piquette, G.N., Frances, A. et al. (1993) Localization of interleukin- 167–227. 1 type I receptor and interleukin-1β in human endometrium throughout the Arend, W.P., Malyak, M., Smith, M.F. et al. (1994) Binding of IL-1α, IL-1β, menstrual cycle. J. Clin. Endocrinol. Metabol., 77, 549–555. 244 Interleukin-1 regulates endometrial collagenase

Simo´n, C., Gimeno, M.J., Mercader, A. et al. (1996) Cytokines-adhesion molecules-invasive proteinases. The missing paracrine/autocrine link in embryonic implantation ? Mol. Hum. Reprod., 2, 405–424. Sims, J.E., Gayle, M.A., Slack, J.L. et al. (1993) Interleukin-1 signaling occurs exclusively via the type I receptor. Proc. Natl. Acad. Sci. USA, 90, 6155–6159. Singer, C.F., Marbaix, E., Kokorine, I. et al. (1997) Paracrine stimulation of interstitial collagenase (MMP-1) in the human endometrium by interleukin- 1α and its dual block by ovarian steroids. Proc. Natl. Acad. Sci. USA, 94, 10341–10345. Stylianou, E., O’Neill, L.A.J., Rawlinson, L. et al. (1992) Interleukin-1 induces NFkB through its type I but not type II receptor in lymphocytes. J. Biol. Chem., 267, 15836–15841. Svenson, M., Hensen, M.B. and Bendtzen, K. (1990) Distribution and characterization of autoantibodies to interleukin-1α in normal human sera. Scand. J. Immunol., 32, 695–701. Downloaded from https://academic.oup.com/molehr/article/5/3/240/1404548 by guest on 01 October 2021 Symons, J.A., Young, P.R. and Duff, G.W. (1993) The soluble interleukin-1 receptor: ligand binding properties and mechanisms of release. Cytokine Res., 12, 381. Symons, J.A., Young, P.R. and Duff, G.W. (1994) Differential release and ligand binding of type II IL-1 receptors. Cytokine, 6, 555 (abstract 100). Tabibzadeh, S. and Sun, X.Z. (1992) Cytokine expression in human endometrium throughout the menstrual cycle. Hum. Reprod., 7, 1214–1221. Woessner, J.F. Jr. (1991) Matrix metalloproteinases and their inhibitors in connective tissue remodeling. FASEB J., 5, 75–82. Zhang, L., Rees, M.C. and Bicknell, R. (1995) The isolation and long-term culture of normal human endometrial epithelium and stroma. Expression of mRNAs for angiogenic polypeptides basally and on oestrogen and progesterone challenges. J. Cell Sci., 108, 323–331.

Received on June 15, 1998; accepted on November 25, 1998

245