Independent regulation of prostaglandins and monocyte chemoattractant protein-1 by interleukin-1{beta} and hCG in human endometrial cells

Jihong Kang1, Ali Akoum3,4, Pierre Chapdelaine1, Philippe Laberge1,4, Patrice E. Poubelle2 and Michel A. Fortier1,4,5

Unité de Recherche en 1Ontogénie et Reproduction et 2 Centre de Recherche en Rhumatologie et Immunologie, Centre de Recherche du CHUL, 2705 Boul. Laurier, Ste-Foy, Québec, G1V 4G2, 3 Endocrinologie de la Reproduction Centre de Recherche de l'Hôpital St-François d'Assise and 4 Département d'Obstétrique et Gynécologie, Université Laval, Ste-Foy, Québec, G1K 7P4, Canada

5 To whom correspondence should be addressed at: Unité de Recherche en Ontogénie et Reproduction, Centre Hospitalier Universitaire de Québec (CHUL), Université Laval, 2705, Boul. Laurier, Sainte-Foy, Québec, G1V 4G2, Canada. Email: mafortier{at}crchul.ulaval.ca


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: Inflammatory mediators such as prostaglandins (PG), chemokines, cytokines and their interactions regulate reproductive functions. The relationship between PG and monocyte chemoattractant protein-1 (MCP-1) has not been elucidated in human endometrium. The presence of hCG receptors in the human endometrium suggests that this embryonic signal may exert a local function in this tissue. Our objectives were to investigate the possible association between PG and MCP-1 and to examine the role of hCG in interleukin-1{beta} (IL-1{beta})-regulated PG and MCP-1 production in human endometrium. METHODS: Primary cell cultures isolated from endometrial biopsies were used as an in vitro model. PG and MCP-1 levels were measured in the culture medium. RESULTS: IL-1{beta} stimulates the production of both PG and MCP-1. Neither COX inhibitors nor direct addition of PG affects MCP-1 production. By contrast, MCP-1 is able to induce PGE2 and PGF2{alpha} in a concentration-dependent manner but it does not appear to contribute to the increase in PG accumulation following IL-1{beta} stimulation. hCG inhibits IL-1{beta}-induced PG level. However, hCG has no effect on either basal or IL-1{beta}-mediated MCP-1 level. CONCLUSIONS: PG are not involved in the regulation of MCP-1 production in endometrial cells. hCG appears to play a local function in the endometrium.

Key words: hCG/human endometrium/IL-1{beta}/MCP-1/prostaglandins


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Menstruation and implantation are two critical reproductive processes involving inflammatory responses in the human endometrium (for a review, see Kelly et al., 2001Go). Inflammatory mediators such as prostaglandins (PG), chemokines and cytokines have been associated with these events (Jones et al., 1997Go). Although the precise mechanisms regulating menstruation and implantation are still unknown, increasing evidence suggests the involvement of the interaction between these mediators.

It is now widely acknowledged that PG are important regulators of female reproductive function (for a review, see Poyser, 1995Go). PG are produced from arachidonic acid (AA) stored in cell membrane phospholipids and liberated by phospholipases (PL). The conversion of AA to PG is catalysed by the rate-limiting enzyme cyclooxygenase (COX). Two isozymes have been identified, a constitutively expressed form COX-1 and an inducible form COX-2 (for a review, see Otto and Smith, 1995Go). PGE2 and PGF2{alpha} are the two most studied prostanoids in human endometrium. The aberrant ratio of PGE2:PGF2{alpha} is associated with menstrual disorders such as dysmenorrhoea and menorrhagia (for a review, see Poyser, 1995Go). In some species, an increase in endometrial prostaglandin production was observed around the peri-implantation period or at the implantation site, suggesting an important role for PG in the implantation process (for a review, see Psychoyos et al., 1995Go).

Chemokines are increasingly recognized as important regulators of uterine function (for a review, see Kayisli et al., 2002Go). Monocyte chemoattractant protein-1 (MCP-1), a member of the beta-chemokine family, displays chemotactic activity for monocytes/macrophages, T lymphocytes, basophils and natural killer cells and recruits them into sites of inflammation (for a review, see Leonard and Yoshimura, 1990Go). MCP-1 can be released by endometrial epithelial and stromal cells in culture (Arici et al., 1995Go). In human endometrium, its expression was detected in a cyclical pattern during the menstrual cycle (Jones et al., 1997Go). The highest level of MCP-1 is detected peri-menstrually (Arici et al., 1999Go).

It has been demonstrated that in some cell types, interleukin-1 (IL-1) is able to induce both PG and MCP-1 production (Efsen et al., 2001Go). IL-1{beta} is a pro-inflammatory cytokine expressed in the endometrium throughout the menstrual cycle, particularly around the time of implantation (Simon et al., 1995Go). It is postulated to mediate the dialogue between the blastocyst and the endometrium in implantation (Ibid, Simon et al., 1995Go). IL-1{beta} can induce PG production in human endometrial stromal cells and an epithelial cell line (Kniss et al., 1997Go; Huang et al., 1998Go; Tamura et al., 2002Go). In endometriotic cells, the production of MCP-1 is up-regulated by IL-1{beta} (Akoum et al., 1995Go). However, no information is available regarding the possible link between the regulation of PG and MCP-1 by IL-1{beta} in human endometrium. Previous studies have indicated that COX metabolites up-regulated MCP-1 in hepatic stellate cells (Efsen et al., 2001Go), but repressed MCP-1 formation in experimental glomerulonephritis (Schneider et al., 1999Go). These studies infer some cell specificity in the determination of the interaction between PG and MCP-1. It is therefore of interest to study the relationship between PG and MCP-1 in human endometrial cells.

The presence of LH/hCG receptors has been reported for many years in the human endometrium, but their physiological relevance is still questioned (Reshef et al., 1990Go; Rao, 1999Go). hCG, the embryonic signal for the maternal recognition of pregnancy in the human, appears to play an important role in blastocyst implantation (for a review, see Srisuparp et al., 2001Go). It is primarily produced by the trophoblast of the placenta. Besides its well-known luteotrophic function, hCG is able to inhibit apoptosis, promote decidualization (Han et al., 1999Go) and exert anticarcinogenic and immunomodulatory effects (Russo and Russo, 1994Go). Up-regulation of COX-2 gene expression in response to hCG has been demonstrated in both human endometrial stromal and epithelial cells (Han et al., 1996Go; Zhou et al., 1999Go). In addition, Arici et al. (1997)Go reported the induction of higher MCP-1 mRNA expression and protein production by LH/hCG in human ovarian follicles.

Thus, the aims of this study are to investigate the possible relationship between PG and MCP-1 and to study the effect of hCG on IL-1{beta}-mediated PG and MCP-1 production. Separated and non-separated human endometrial cell cultures are compared and non-separated human endometrial cells are used in the present study as the primary in vitro model.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Tissue collection and culture of endometrial cells
Endometrial biopsies were taken from women aged between 31 and 45 years with regular menstrual cycles undergoing gynaecological investigation for benign conditions. Biopsies were dated according to the stated last menstrual period (LMP). A total of 11 biopsies was used, five from proliferative and six from secretory phase. Informed consent for donation of anonymous endometrial samples was obtained before tissue collection. The tissue samples were placed in sterile Hanks' balanced salt solution (HBSS) containing 100 IU/ml penicillin, 100 µg/ml streptomycin and 0.25 mg/ml amphotericin B at 4°C and transported to the laboratory immediately. Endometrial tissues were placed in a 50 ml conical tube containing trypsin (0.02%; Boehringer Mannheim Corp., USA), collagenase (type II, 0.02%; Sigma, USA) and deoxyribonuclease I (DNase-I, 0.015%; Sigma) in HBSS and incubated at 37°C for 20 min in a shaking water bath. Cells obtained from the digestion were recovered by centrifugation. The cell pellets were resuspended in 15 ml HBSS and deposited on 12.5 ml Ficoll gradient (Amersham Biosciences, Sweden). The gradients were centrifuged at 800 g for 20 min. Endometrial cells were recovered at the interface and washed with HBSS. The cell pellets were resuspended in Roswell Park Memorial Institute (RPMI) 1640 (ICN Biomedicals, Inc., USA) containing 50 ng/ml gentamicin (Sigma), 10 mg/ml insulin (Sigma), 5 mg/ml transferrin (Sigma) and 10% fetal bovine serum (Wisent, Inc., Quebec, Canada) depleted of steroids by dextran–charcoal extraction. Endometrial cells were seeded in 24-well plates and incubated at 37°C in a humidified atmosphere of 5% CO2 in air. It is difficult to use standard haemocytometer or coulter counter techniques to evaluate the total number or the relative proportion of epithelial and stromal cells in the non-separated cell preparation. However, the relative proportion of cells is quite reproducible between experiments as evaluated subjectively by observation under the phase contrast microscope. We have demonstrated previously, using the bovine endometrium, that our method of isolation and purification as well as estimation of cell distribution was highly accurate and reproducible (Asselin et al., 1996Go). In the present study, when endometrial cells are fully attached and after culture medium replacement, we estimate that the relative proportion of epithelial cells was ~50% but could vary from 40 to 60% between experiments. As an additional measure, we have evaluated the expression of cytokeratin, an epithelial cell marker, and vimentin, a stromal (mesenchyma) cell marker, in a different set of experiments. Dual labelling with a fluorescent secondary antibody confirmed our previous results claiming a proportion of 50% epithelial (cytokeratin positive) cells (Figure 1F). The presence of contaminating immune cells was evaluated in three parallel cultures. Lymphocytes do not adhere to plastic and those eventually present in suspension were eliminated at the time of medium change. The presence of macrophages was evaluated by Diff-Quick and non-specific esterase staining (Roberge et al., 1994Go).



View larger version (122K):
[in this window]
[in a new window]
 
Figure 1. Characteristics of primary human endometrial cells in culture. Photomicrographs of separated epithelial cells (A), stromal cells (B) and non-separated cells (C) under phase-contrast microscope. Indirect immunofluorescent staining of the same section is shown on the right. Epithelial cells were stained for cytokeratin (D), stromal cells stained for vimentin (E). Antibody against cytokeratin was used for non-separated endometrial cells (F). Magnification:x100.

 
Isolation of separated epithelial and stromal cells
The separation of epithelial and stromal cells was performed as described previously (Akoum et al., 1995Go). Briefly, the pellets of non-separated endometrial cells obtained above were resuspended in 10 ml of culture medium. Unit gravity sedimentation of the cell suspension (Freshney, 2000, p. 222)Go was conducted at room temperature for 1 h. The top 8 ml of medium was defined as the stromal-enriched fraction and the bottom 2 ml as the epithelial-enriched fraction. The stromal fraction was transferred to 24-well plates and incubated at 37°C in an atmosphere of 5% CO2 in air for 30 min. The medium containing floating cells and debris was replaced and the adherent stromal cells were maintained in culture. The epithelial fraction was diluted with culture medium and placed into 75 cm2 culture dishes for a preincubation process, after which the contaminating stromal cells were eliminated by selective adhesion (Freshney, 2000, p. 211)Go. Epithelial cells were recovered and seeded into 24-well plates and incubated at 37°C in a humidified atmosphere of 5% CO2:95% air.

Treatments
The medium was changed every 2 days. After confluence, cells were treated with fresh, serum-free RPMI 1640 medium containing different concentrations of IL-1{beta} (0–1 ng/ml) (Sigma), IL-1{beta} (1 ng/ml) plus indomethacin (10 µmol/l; Sigma), IL-1{beta} (1 ng/ml) plus NS-398 (10 µmol/l; Cayman, USA), hCG (0.1 µg/ml, 10 000 IU/mg; Sigma), IL-1{beta} (1 ng/ml) plus hCG (0.1 µg/ml), PGE2 (0–10 µmol/l), PGF2{alpha} (0–10 µmol/l) or MCP-1 (0–50 ng/ml) in triplicate at 37°C in an atmosphere of 5% CO2:95% air. After 24 h, the culture medium was collected and kept at –20°C for prostaglandin and MCP-1 measurement. Cells were recovered for protein assay.

Immunofluorescence microscopy analysis
Primary separated and non-separated human endometrial cells were grown on Lab Tek 4-well chamber slides (Nalge Nunc International, USA). Cells were fixed with cold methanol for 6 min at room temperature and then rinsed with phosphate-buffered saline (PBS). Slides were then blocked in PBS containing 10% goat serum for 1 h at room temperature. After washing in PBS (three times, 5 min each), slides were incubated with mouse monoclonal antibodies against cytokeratin (recognizing 4, 5, 6, 8, 10, 13 and 18 isoforms) or vimentin (Sigma) at a dilution of 1:200 in PBS containing 0.5% goat serum for 1 h at room temperature. Immunofluorescence was obtained by reaction with a fluorescent secondary antibody (Alexa Fluor® 488 goat anti-mouse IgG; Molecular Probes, Inc., USA) at a dilution of 1:300 in PBS containing 0.5% goat serum for 1 h at room temperature. Slides were then rinsed three times with PBS. Fluorescence was visualized using a Zeiss Axiovert 100-Inverted microscope (Zeiss, Germany) and images captured and integrated using the Northern Exposure program (Empix Imaging Inc., Canada).

Enzyme immunoassays (EIA) for PG
An EIA technique was used for PGE2 and PGF2{alpha} measurement as described previously (Asselin et al., 1996Go), which utilized acetylcholinesterase-linked PG tracers (Cayman Chemical Company, USA). The inter- and intra-assay coefficients of variation (n=12) were 16 and 10% respectively.

MCP-1 ELISA
MCP-1 in the culture medium was measured using a previously reported sandwich enzyme-linked immunosorbent assay (Akoum et al., 1996Go). Briefly, 250 ng/well of mouse monoclonal anti-human MCP-1 antibody (R & D systems, USA) and a 1:1000 dilution of rabbit polyclonal anti-human MCP-1 antibody (a generous gift from Dr Paul H.Naccache) were used for detection. TMB (3,3',5,5',-tetramethylbenzidine; Bio-Rad Laboratories Ltd, Canada) was used as the substrate. The enzymatic reaction was stopped by the addition of sulphuric acid. The optical density was determined at 450 nm. MCP-1 concentration was interpolated from the standard curve using recombinant human MCP-1 (R & D systems). The sensitivity for MCP-1 was 50 pg/ml. The intra- and inter-assay coefficients of variation were <6%.

Western blot analysis
Proteins from cultured human non-separated endometrial cells were extracted and quantified as described recently (Chapdelaine et al., 2001Go). Aliquots of 10 µg protein of each sample were separated on 10% sodium dodecyl sulphate–polyacrylamide gel electrophoresis and then transferred onto nitrocellulose membranes. The membranes were blocked overnight at 4°C in PBS containing 5% fat-free dry milk and 0.05% Tween 20. After 1 h of incubation at room temperature with rabbit anti-sheep COX-1 polyclonal antibody or rabbit anti-sheep COX-2 polyclonal antibody (Merck–Froost, Canada) at a dilution of 1:6000 in 1% milk solution, the blots were washed with PBS containing 0.05% Tween 20. A secondary antibody (goat anti-rabbit IgG conjugated with horseradish peroxidase; Jackson Immunoresearch Laboratories, USA) was used for another 1 h at room temperature at a dilution of 1:10 000 in 3% milk solution. Finally, the blots were washed and immunoreactive proteins were visualized with an enhanced chemiluminescent system (Renaissance; NEN, Life Science Products, USA). The density of the signals was determined by computer-assisted densitometry (Biolmage, Visage 110s; Genomic Solutions Inc., USA).

Statistical analysis
Data are presented as the mean±SEM. Statistical analysis was performed using analysis of variance (ANOVA) followed by Fisher's protected least significant difference, Duncan's new multiple range and Student–Newman–Keuls multiple comparison tests (Super ANOVA; Abacus Concepts, USA). The interaction between the effect of IL-1{beta} and cell types and the interaction between hCG and IL-1{beta} were tested using two-way ANOVA. P<0.05 was considered statistically significant.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Characteristics of primary human endometrial cell cultures
Separated human endometrial cells were prepared from the biopsies and morphological characteristics of epithelial, stromal and non-separated cells were examined under a phase-contrast microscope. Epithelial cells exhibited a typical cuboidal morphology (Figure 1A), stromal cells an elongated fibroblast morphology (Figure 1B) and non-separated cell cultures a combination of both morphologies (Figure 1C). There was no difference in gross morphology between cells obtained from the proliferative or the secretory phase of the cycle. The presence of contaminating monocytes was found to be <0.02%, thus ruling out their contribution to the present findings on PG and MCP-1 production.

Epithelial and stromal cells in separated and non-separated cells were identified by cell-specific staining of intermediate filament proteins. Epithelial cells were positively labelled with the epithelium-specific cytokeratin antibody (Figure 1D and F), whereas stromal cells did not react with this antibody. By contrast, stromal cells exhibited a strong vimentin staining (Figure 1E) and epithelial cells weak background staining (data not shown).

Prostaglandin accumulation in the culture medium of human endometrial epithelial, stromal and non-separated cells
PG accumulation in the culture medium of epithelial, stromal and non-separated cells following treatment with IL-1{beta} (1 ng/ml) for 24 h was illustrated in Figure 2. IL-1{beta} significantly increased both PGF2{alpha} and PGE2 in stromal cells. In epithelial cells, IL-1{beta} induced a significant up-regulation of PGF2{alpha}, but not PGE2. Basal PG levels were comparable in stromal and epithelial cells. However, the effect of IL-1{beta} in stromal cells was significantly higher than in epithelial cells for both PGF2{alpha} and PGE2. In non-separated cells, IL-1{beta} stimulated significantly PGE2 and PGF2{alpha} (P<0.05).



View larger version (16K):
[in this window]
[in a new window]
 
Figure 2. Prostaglandin (PG) accumulation by human endometrial stromal, epithelial and non-separated cells. Separated or non-separated cells were seeded in 24-well plates. After reaching confluence, cells were incubated with interleukin-1{beta} (IL-1{beta}) (0 and 1 ng/ml) for 24 h. The culture medium was collected for PG assay by enzyme immunoassay. The results are presented as mean±SEM of four different experiments for stromal and epithelial cells from the same subjects, 11 for non-separated cells. *Significantly different compared with the control in the same cell type (P<0.05).

 
Time-course of the effect of IL-1{beta} on COX proteins, prostaglandin and MCP-1 levels
The effect of IL-1{beta} (1 ng/ml) on COX-1 and COX-2 protein expression from 0 to 24 h was illustrated in Figure 3. The increase in COX-2 protein level was already noticeable at 2 h (3.7-fold as time 0) and proceeded progressively up to 24 h (4.8-fold as time 0) (Figure 3A). There was no detectable change in COX-1 protein expression. A significant stimulation of PG accumulation was detectable after 2 h, whereas the up-regulation of MCP-1 was evident after 4 h (data not shown).



View larger version (31K):
[in this window]
[in a new window]
 
Figure 3. Effects of interleukin-1{beta} (IL-1{beta}) and IL-1{beta} plus cyclooxygenase (COX) inhibitors on COX protein expression, prostaglandin (PG) and monocyte chemoattractant protein-1 (MCP-1) levels in human non-separated endometrial cells. (A) Time-course of the effect of IL-1{beta} on COX-1 and COX-2 protein expression. Confluent cells were incubated with IL-1{beta} (1 ng/ml) for different time-periods. Proteins were extracted for COX-1 and COX-2 assay. (B, C and D) After reaching confluence, cells were treated with IL-1{beta} (0, 0.01, 0.1 and 1 ng/ml), IL-1{beta} (1 ng/ml) plus indomethacin (10 µmol/l) and IL-1{beta} (1 ng/ml) plus NS-398 (10 µmol/l) for 24 h. The culture medium was collected for PG and MCP-1 assay by enzyme immunoassay and enzyme-linked immunosorbent assay respectively. Cell pellets were recovered for COX protein measurement. The results are presented as mean±SEM of six experiments. (B) Bands with different superscripts and (C and D) bars with asterisks are significantly different compared with the other groups (P<0.05).

 
Effect of IL-1{beta} and COX inhibitors on prostaglandin level
The effects of IL-1{beta} and the combination of IL-1{beta} and COX inhibitors on COX protein and prostaglandin levels were determined in non-separated endometrial cells. Confluent cells were incubated with different doses of IL-1{beta} and IL-1{beta} plus COX inhibitors for 24 h. Western blot analysis showed that IL-1{beta} increased COX-2 protein level in a concentration-dependent manner (1.4-, 2.7-, 6.3-fold versus control respectively) (Figure 3B). There was a significant increase in COX-2 protein level by 1 ng/ml IL-1{beta} compared with control. No difference was observed between IL-1{beta} alone and the combination of IL-1{beta} and COX inhibitors on either COX-1 or COX-2 protein levels.

Both PGE2 and PGF2{alpha} levels were increased in a concentration-dependent manner by IL-1{beta} (Figure 3C). Co-treatment of cells with indomethacin (10 µmol/l), a non-specific COX inhibitor, or NS-398 (10 µmol/l), a COX-2 specific inhibitor, significantly reduced IL-1{beta}-induced prostaglandin secretion.

Effects of IL-1{beta} and COX inhibitors on MCP-1 secretion by human endometrial cells
MCP-1 concentration was measured in the culture medium from the same experiments used for determination of PG production. MCP-1 was detectable in the medium of all cultures but in variable levels. IL-1{beta} enhanced MCP-1 production in a concentration-dependent manner (Figure 3D). Co-treatment of COX inhibitors with IL-1{beta} did not have any effect on MCP-1 levels compared with IL-1{beta} alone.

Effect of prostaglandins on MCP-1 level
Non-separated endometrial cells in culture were incubated with different concentrations of PGE2 (0–10 µmol/l) or PGF2{alpha} (0–10 µmol/l) for 24 h. Table I showed that PGF2{alpha} and PGE2 did not have any significant effect on MCP-1 levels (P>0.05).


View this table:
[in this window]
[in a new window]
 
Table I. Effects of prostaglandin (PG) on monocyte chemoattractant protein-1 (MCP-1) production by human non-separated endometrial cells; data are expressed as mean±SEM (n=3 for PGE2 and n=2 for PGF2{alpha})

 
Effect of MCP-1 on prostaglandin secretion
After reaching confluence, cells were treated with different concentrations of MCP-1 (0, 0.5, 5 and 50 ng/ml) for 24 h. As illustrated in Figure 4, MCP-1 induced a significant increase of PGE2 compared with control (P<0.05).



View larger version (13K):
[in this window]
[in a new window]
 
Figure 4. Effect of monocyte chemoattractant protein-1 (MCP-1) on prostaglandin (PG) secretion by human non-separated endometrial cells. Cells were incubated with different doses of MCP-1 (0, 0.5, 5 and 50 ng/ml) for 24 h. The culture media were collected for PG measurement by enzyme immunoassay. The results are presented as mean±SEM of three experiments. *Significantly different compared with control (P<0.05).

 
Effect of hCG on IL-1{beta}-induced prostaglandin levels
We investigated the effect of hCG on IL-1{beta}-induced prostaglandin levels. Cells were exposed to hCG (0 and 0.1 µg/ml), IL-1{beta} (1 ng/ml) and IL-1{beta} (1 ng/ml) plus hCG (0.1 µg/ml) for 24 h. Western blot analysis did not detect any difference of either COX-2 or COX-1 protein between IL-1{beta} and IL-1{beta} plus hCG (Figure 5A).



View larger version (24K):
[in this window]
[in a new window]
 
Figure 5. Effect of hCG on interleukin-1{beta} (IL-1{beta})-induced prostaglandin (PG) and monocyte chemoattractant protein-1 (MCP-1) production by human non-separated endometrial cells. After reaching confluence, cells were incubated with hCG (0 and 0.1 mg/ml), IL-1{beta} (1 ng/ml) and IL-1{beta} (1 ng/ml) plus hCG (0.1 µg/ml) for 24 h. Proteins were extracted for COX-1 and COX-2 assay (A). The culture medium was collected for PG (B) and MCP-1 measurement (C). The results are presented as mean±SEM of four experiments. *Significantly different compared with the other groups (P<0.05).

 
As illustrated in Figure 5B, 0.1 µg/ml hCG did not have any appreciable effect on either PGE2 or PGF2{alpha} levels compared with control. However, the addition of 0.1 µg/ml hCG to 1 ng/ml IL-1{beta} inhibited IL-1{beta}-induced PG accumulation. A significant reduction was observed only for PGF2{alpha} based on two-way ANOVA (P<0.05).

Effects of hCG on IL-1{beta}-induced MCP-1 production
As illustrated in Figure 5C, no change was observed in either basal or IL-1{beta}-induced MCP-1 levels in response to 0.1 µg/ml hCG.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
In the present study, we have used an in vitro system based on human primary endometrial cell cultures to study the interactions between cytokines and PG. First we have compared the characteristics of separated and non-separated cell cultures. In agreement with other studies (Matthews et al., 1992Go), epithelial cells, but not stromal cells, were positive for cytokeratin in our culture system. By contrast, expression of vimentin is much higher and consistent in stromal cells than in epithelial cells. Using these characteristics, the relative proportion of epithelial cells could easily be identified with cytokeratin in the non-separated cell cultures (Figure 1F). We next describe the pattern of PG synthesis in separated epithelial and stromal cultures and compare them with non-separated cell cultures. Most studies on human endometrial cells in vitro are based on the use of stromal cells. This phenomenon may result from the tendency of epithelial cells to be less viable, more difficult to isolate and grow slower. The presence of IL-1 receptor type I (IL-1R tI) mRNA has been reported in the human endometrium (Simon et al., 1993Go), especially in the epithelium. However, in the present study stromal cells are more responsive to IL-1{beta} than epithelial cells (Figure 2). One reason may be that epithelial cells, when losing the three-dimensional structure occurring in vivo, also lose functional characteristics (for a review, see Fleming, 1999Go). It has been shown by others that the interactions between endometrial epithelial cells (EEC) and stromal cells (ESC) do influence their specific characteristics in vitro (Osteen et al., 1994Go; Pierro et al., 2001Go). In this study, epithelial and stromal cells were seeded together after enzymatic digestion. Several factors have motivated the choice of this model. First, the co-culture approach is much more efficient in terms of yield of viable cells and allows an optimal use of the precious human endometrial samples. Second, although the three-dimensional structure is probably not maintained with the non-separated cell model, it does allow physical and paracrine interactions between stromal and epithelial cells. This is strongly suggested by the pattern of PG accumulation found in non-separated cells (Figure 2). Indeed, the level of PG accumulation and the response to IL-1{beta} is lower or equal to that observed for epithelial cells which represent ~50% of the cell population in confluent non-separated cultures. Although less complete than the use of separated and non-separated endometrial cell cultures in parallel, we believe that co-culture represent an interesting compromise for in vitro study of endometrial function. No significant difference in basal PG accumulation was observed between cultures prepared from biopsies taken at different stages of the cycle.

The first indication of the induction of PG by IL-1{beta} in human endometrium dates back to 1990 (Mitchell et al., 1990Go). In stromal cells, it was shown that IL-1{beta} can elevate both PGE2 and PGF2{alpha} levels by increasing COX-2 mRNA and protein levels (Tamura et al., 2002Go). This effect involves protein kinase A (PKA), nuclear factor kappa B (NF-KB), and/or the extracellular signal-regulated kinase 1 and 2 (ERK1/2) signalling pathways (Tamura et al., 2002Go). In our study, similar induction of PG by IL-1{beta} is observed in non-separated endometrial cells and correlated with a concentration-dependent increase of COX-2 but not COX-1 protein levels (Figure 3). The time-course of COX-2 induction with a significant increase at 2 h shows a much faster expression than reported before in stromal cells (Huang et al., 1998Go). The difference may be due to the presence of epithelial cells.

MCP-1 can be released by many reproductive tissues. A variety of growth factors and cytokines such as tumor necrosis factor-{alpha} (TNF-{alpha}), IL-1 and interferon-{gamma} (IFN-{gamma}) stimulates MCP-1 secretion. In human endometrium, MCP-1 level is the lowest around the time of ovulation when estrogen is high (Arici et al., 1999Go). In glandular epithelial cells, MCP-1 mRNA level is higher than in stromal cells, but not regulated (Arici et al., 1995Go). No increase in MCP-1 mRNA levels was found in glandular epithelium treated with IL-1 (Arici et al., 1995Go). The mechanisms by which MCP-1 gene transcription is regulated are agonist specific. NF-KB pathway has been suggested for the regulation of MCP-1 by mediators such as IL-1, glucocorticoid, lipopolysaccharide (LPS) and TNF-{alpha} (King et al., 2001Go; Ueda et al., 1997Go). Our data show that in human non-separated endometrial cells, IL-1{beta} stimulation of MCP-1 production is concentration-dependent and highly reproducible, regardless of the possible difference in the proportion of epithelial and stromal cells in each confluent culture (Figure 3D). This indicates that the proportion of epithelial and stromal cells was fairly similar between experiments or that interactions between these cell types make them equivalent in terms of MCP-1 production in response to IL-1{beta}. The increase of MCP-1 in response to IL-1{beta} in the endometrium may be correlated with the activated NF-KB observed during menstruation.

Previous studies have shown that PG can modulate other inflammatory mediators including MCP-1. In ovine and bovine corpus luteum (CL), administration of PGF2{alpha} induces MCP-1 mRNA expression (Tsai et al., 1997Go). PGE2, an important immunomodulator and mediator of inflammation within the uterus, stimulates MCP-1 release from placenta (Denison et al., 1998Go). Additionally, COX inhibitors modulate MCP-1 levels in different ways depending on cell types (Kyrkanides et al., 2002Go). In human endometrium, high expression of MCP-1 and COX-2 is observed before the initiation of menstruation, coinciding with leukocyte accumulation in the uterus (Jones et al., 1997Go). Collectively, this led us to hypothesize that there could be a relationship between PG and MCP-1 in the endometrium. We show that treatment with indomethacin, a non-specific COX inhibitor, or NS-398, a COX-2-specific inhibitor, markedly reduces PG levels stimulated by IL-1{beta} (Figure 3C). COX inhibitors alone, especially indomethacin, also decreased basal PG accumulation in the culture media (data not shown). However, neither indomethacin nor NS-398 has any effect on either IL-1{beta}-induced (Figure 3D) or basal MCP-1 level (data not shown). Comparing our data with those obtained in hepatic stellate cells and glomerular mesangial cells (Schneider et al., 1999Go; Efsen et al., 2001Go) confirms that the interaction between PG and MCP-1 varies depending on tissues or cell types. We next investigated MCP-1 production in response to direct addition of PG and found that MCP-1 secretion is not sensitive to PGF2{alpha} or PGE2 (Table I). Recent studies demonstrated the presence of PGF2{alpha} receptor as well as PGE2 receptors in human endometrium (Milne et al., 2001Go; Milne and Jabbour, 2003Go). Thus, the lack of response is likely due to the inability of PG to modulate MCP-1 level in endometrial cells. Globally, neither COX inhibitors nor direct addition of PG had any effect on MCP-1 level, suggesting that PG do not play a role in the regulation of MCP-1 synthesis in human endometrial cells. Interestingly, the incubation of cells with MCP-1 gave rise to a significant concentration-dependent increase in PGE2, but not PGF2{alpha} production (Figure 4). A time-course study was then conducted to test whether MCP-1 could contribute to PG production induced by IL-1{beta}. The results revealed that a significant increase in PG production was already detectable 2 h following exposure of cells to IL-1{beta} (1 ng/ml), while the up-regulation of MCP-1 was evident only after 4 h (data not shown). Moreover, IL-1{beta} stimulates both PGF2{alpha} and PGE2 whereas MCP-1 preferentially stimulates PGE2. These observations suggest that PG production in response to IL-1{beta} stimulation is independent of MCP-1.

Despite its important role in implantation (Simon et al., 1994Go), IL-1{beta} can disrupt blastocyst implantation by inhibiting decidualization of endometrial stromal cells (Kariya et al., 1991Go; Frank et al., 1995Go). hCG, a heterodimeric glycoprotein hormone required to maintain pregnancy, exerts its effects through LH/hCG receptors by stimulating progesterone secretion from the corpus luteum in the first trimester. Moreover, hCG is suggested to have important roles within the uterus through an effect on cytokine production (Uzumcu et al., 1998Go). IL-6 and TNF-{alpha} are stimulated by hCG in human endometrial cells (Uzumcu et al., 1998Go). We have tested the possibility that hCG could have an effect on IL-1{beta} via modulation of PG and/or MCP-1 production. We found that at the concentration of 0.1 µg/ml, hCG did not have any effect on basal PG secretion (Figure 5B). The co-treatment of hCG (0.1 µg/ml) with IL-1{beta} (1 ng/ml), however, reduced PG production induced by IL-1{beta}. A significant reduction was observed only for PGF2{alpha} (P<0.05). The mechanism by which hCG reduced IL-1{beta}-induced PG is unknown. In gonads, LH/CG receptors are coupled to cAMP generation and act through PKA. This would, however, decrease NF-KB and inhibit MCP-1 production, which is not the case here. In primary baboon endometrial epithelial cells and a human endometrial cell line, hCG does not induce cAMP generation, but rather a PKA-independent ERK1/2 cascade that would also decrease MCP-1 formation (Srisuparp et al., 2001Go). Therefore, according to the literature, all mechanisms of hCG action that could decrease PG production would also lead to a reduction of MCP-1 (Tamura et al., 2002Go). Therefore, further investigation of this particular action must be conducted to elucidate the mechanisms involved. As indicated in previous studies, suppression of PG synthesis is required for successful implantation (Smith and Kelly, 1988Go). This finding in our study supports the local role of hCG in the endometrium. hCG may indirectly modulate the receptive endometrium to facilitate the implantation process. In contrast, hCG did not have any effect on either basal or IL-1{beta}-induced MCP-1 level (Figure 5C). Human CG receptors have been well characterized on human endometrial epithelial and stromal cells (Reshef et al., 1990Go). The failure of hCG to suppress IL-1{beta}-induced MCP-1 is surprising because hCG can inhibit various cellular responses mediated by NF-KB. Inhibition of NF-KB activation by hCG is not cell type specific (Manna et al., 2000Go). Given the range of hCG concentrations tested (from 0.1 to 10 µg/ml), it seems unlikely that the effective dose was missed. Therefore, MCP-1 levels do not appear to be directly modulated by hCG in human endometrial cells.

In conclusion, IL-1{beta} stimulates the production of both PG and MCP-1 in non-separated human endometrial cells by apparently distinct and independent mechanisms. hCG, added in combination with IL-1{beta}, is able to block the induction of PG synthesis but not that of MCP-1. These results support the role for cytokines in the regulation of endometrial cell function and of a local action of hCG in the uterus.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
This study was supported by a grant from Canadian Institute of Health Research (CIHR) to M.A.F.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Akoum A, Lemay A, Brunet C and Hebert J (1995) Secretion of monocyte chemotactic protein-1 by cytokine-stimulated endometrial cells of women with endometriosis. Le groupe d'investigation en gynecologie. Fertil Steril 63, 322–328.[ISI][Medline]

Akoum A, Lemay A, McColl S, Turcot-Lemay L and Maheux R (1996) Elevated concentration and biologic activity of monocyte chemotactic protein-1 in the peritoneal fluid of patients with endometriosis. Fertil Steril 66, 17–23.[ISI][Medline]

Arici A, MacDonald PC and Casey ML (1995) Regulation of monocyte chemotactic protein-1 gene expression in human endometrial cells in cultures. Mol Cell Endocrinol 107, 189–197.[CrossRef][ISI][Medline]

Arici A, Oral E, Bukulmez O, Buradagunta S, Bahtiyar O and Jones EE (1997) Monocyte chemotactic protein-1 expression in human preovulatory follicles and ovarian cells. J Reprod Immunol 32, 201–219.[CrossRef][ISI][Medline]

Arici A, Senturk LM, Seli E, Bahtiyar MO and Kim G (1999) Regulation of monocyte chemotactic protein-1 expression in human endometrial stromal cells by estrogen and progesterone. Biol Reprod 61, 85–90.[Abstract/Free Full Text]

Asselin E, Goff AK, Bergeron H and Fortier MA (1996) Influence of sex steroids on the production of prostaglandins F2 alpha and E2 and response to oxytocin in cultured epithelial and stromal cells of the bovine endometrium. Biol Reprod 54, 371–379.[Abstract]

Chapdelaine P, Vignola K and Fortier MA (2001) Protein estimation directly from SDS-PAGE loading buffer for standardization of samples from cell lysates or tissue homogenates before Western blot analysis. Biotechniques 31, 478–482.[ISI][Medline]

Denison FC, Kelly RW, Calder AA and Riley SC (1998) Cytokine secretion by human fetal membranes, decidua and placenta at term. Hum Reprod 13, 3560–3565.[Abstract]

Efsen E, Bonacchi A, Pastacaldi S, Valente AJ, Wenzel UO, Tosti-Guerra C, Pinzani M, Laffi G, Abboud HE, Gentilini P et al. (2001) Agonist-specific regulation of monocyte chemoattractant protein-1 expression by cyclooxygenase metabolites in hepatic stellate cells. Hepatology 33, 713–721.[CrossRef][ISI][Medline]

Fleming H (1999) Structure and function of cultured endometrial epithelial cells. Semin Reprod Endocrinol 17, 93–106.[ISI][Medline]

Frank GR, Brar AK, Jikihara H, Cedars MI and Handwerger S (1995) Interleukin-1 beta and the endometrium: an inhibitor of stromal cell differentiation and possible autoregulator of decidualization in humans. Biol Reprod 52, 184–191.[Abstract]

Freshney RI (2000) Culture of Animal Cells: A Manual of Basic Techniques, 4th edn. John Wiley & Sons, Inc, New York, USA.

Han SW, Lei ZM and Rao CV (1996) Up-regulation of cyclooxygenase-2 gene expression by chorionic gonadotropin in mucosal cells from human fallopian tubes. Endocrinology 137, 2929–2937.[Abstract]

Han SW, Lei ZM and Rao CV (1999) Treatment of human endometrial stromal cells with chorionic gonadotropin promotes their morphological and functional differentiation into decidua. Mol Cell Endocrinol 147, 7–16.[CrossRef][ISI][Medline]

Huang JC, Liu DY, Yadollahi S, Wu KK and Dawood MY (1998) Interleukin-1 beta induces cyclooxygenase-2 gene expression in cultured endometrial stromal cells. J Clin Endocrinol Metab 83, 538–541.[Abstract/Free Full Text]

Jones RL, Kelly RW and Critchley HO (1997) Chemokine and cyclooxygenase-2 expression in human endometrium coincides with leukocyte accumulation. Hum Reprod 12, 1300–1306.[ISI][Medline]

Kariya M, Kanzaki H, Takakura K, Imai K, Okamoto N, Emi N, Kariya Y and Mori T (1991) Interleukin-1 inhibits in vitro decidualization of human endometrial stromal cells. J Clin Endocrinol Metab 73, 1170–1174.[Abstract]

Kayisli UA, Mahutte NG and Arici A (2002) Uterine chemokines in reproductive physiology and pathology. Am J Reprod Immunol 47, 213–221.[CrossRef][ISI][Medline]

Kelly RW, King AE and Critchley HO (2001) Cytokine control in human endometrium. Reproduction 121, 3–19.[Abstract/Free Full Text]

King AE, Critchley HO and Kelly RW (2001) The NF-kappaB pathway in human endometrium and first trimester decidua. Mol Hum Reprod 7, 175–183.[Abstract/Free Full Text]

Kniss DA, Zimmerman PD, Garver CL and Fertel RH (1997) Interleukin-1 receptor antagonist blocks interleukin-1-induced expression of cyclooxygenase-2 in endometrium. Am J Obstet Gynecol 177, 559–567.[ISI][Medline]

Kyrkanides S, Moore AH, Olschowka JA, Daeschner JC, Williams JP, Hansen JT and Kerry O'Banion M (2002) Cyclooxygenase-2 modulates brain inflammation-related gene expression in central nervous system radiation injury. Brain Res Mol Brain Res 104, 159–169.[ISI][Medline]

Leonard EJ and Yoshimura T (1990) Human monocyte chemoattractant protein-1 (MCP-1). Immunol Today 11, 97–101.[CrossRef][ISI][Medline]

Manna SK, Mukhopadhyay A and Aggarwal BB (2000) Human chorionic gonadotropin suppresses activation of nuclear transcription factor-kappa B and activator protein-1 induced by tumor necrosis factor. J Biol Chem 275, 13307–13314.[Abstract/Free Full Text]

Matthews CJ, Redfern CPF, Hirst BH and Thomas EJ (1992) Characterization of human purified epithelial and stromal cells from endometrium and endometriosis in tissue culture. Fertil Steril 57, 990–997.[ISI][Medline]

Milne SA and Jabbour HN (2003) Prostaglandin (PG) F(2alpha) receptor expression and signaling in human endometrium: role of PGF(2alpha) in epithelial cell proliferation. J Clin Endocrinol Metab 88, 1825–1832.[Abstract/Free Full Text]

Milne SA, Perchick GB, Boddy SC and Jabbour HN (2001) Expression, localization, and signaling of PGE(2) and EP2/EP4 receptors in human nonpregnant endometrium across the menstrual cycle. J Clin Endocrinol Metab 86, 4453–4459.[Abstract/Free Full Text]

Mitchell MD, Edwin S and Romero RJ (1990) Prostaglandin biosynthesis by human decidual cells: effects of inflammatory mediators. Prostagl Leukot Essent Fatty Acids 41, 35–38.[ISI][Medline]

Osteen KG, Rodgers WH, Gaire M, Hargrove JT, Gorstein F and Matrisian LM (1994) Stromal-epithelial interaction mediates steroidal regulation of metalloproteinase expression in human endometrium. Proc Natl Acad Sci USA 91, 10129–10133.[Abstract/Free Full Text]

Otto JC and Smith WL (1995) Prostaglandin endoperoxide synthases-1 and -2. J Lipid Mediat Cell Signal 12, 139–156.[CrossRef][ISI][Medline]

Pierro E, Minici F, Alesiani O, Miceli F, Proto C, Screpanti I, Mancuso S and Lanzone A (2001) Stromal-epithelial interactions modulate estrogen responsiveness in normal human endometrium. Biol Reprod 64, 831–838.[Abstract/Free Full Text]

Poyser NL (1995) The control of prostaglandin production by the endometrium in relation to luteolysis and menstruation. Prostagl Leukot Essent Fatty Acids 53, 147–195.[ISI][Medline]

Psychoyos A, Nikas G and Gravanis A (1995) The role of prostaglandins in blastocyst implantation. Hum Reprod 10 (Suppl 2), 30–42.[ISI][Medline]

Rao CV (1999) No mRNAs for the LH/hCG receptor in human endometrium? Fertil Steril 72, 374–375.[CrossRef][ISI][Medline]

Reshef E, Lei ZM, Rao CV, Pridham DD, Chegini N and Luborsky JL (1990) The presence of gonadotropin receptors in nonpregnant human uterus, human placenta, fetal membranes, decidua. J Clin Endocrinol Metab 70, 421–430.[Abstract]

Roberge CJ, de Medicis R, Dayer JM, Rola-Pleszczynski M, Naccache PH and Poubelle PE (1994) Crystal-induced neutrophil activation. V. Differential production of biologically active IL-1 and IL-1 receptor antagonist. J Immunol 152, 5485–5494.[Abstract/Free Full Text]

Russo J and Russo IH (1994) Toward a physiological approach to breast cancer prevention. Cancer Epidemiol Biomarkers Prev 3, 353–364.[Abstract]

Schneider A, Harendza S, Zahner G, Jocks T, Wenzel U, Wolf G, Thaiss F, Helmchen U and Stahl RA (1999) Cyclooxygenase metabolites mediate glomerular monocyte chemoattractant protein-1 formation and monocyte recruitment in experimental glomerulonephritis. Kidney Int 55, 430–441.[CrossRef][ISI][Medline]

Simon C, Piquette GN, Frances A and Polan ML (1993) Localization of interleukin-1 type I receptor and interleukin-1 beta in human endometrium throughout the menstrual cycle. J Clin Endocrinol Metab 77, 549–555.[Abstract]

Simon C, Frances A, Piquette GN, el Danasouri I, Zurawski G, Dang W and Polan ML (1994) Embryonic implantation in mice is blocked by interleukin-1 receptor antagonist. Endocrinology 134, 521–528.[Abstract]

Simon C, Pellicer A and Polan ML (1995) Interleukin-1 system crosstalk between embryo and endometrium in implantation. Hum Reprod 10 (Suppl 2), 43–54.[ISI][Medline]

Smith SK and Kelly RW (1988) The release of PGF2 alpha and PGE2 from separated cells of human endometrium and decidua. Prostagl Leukot Essent Fatty Acids 33, 91–96.[ISI][Medline]

Srisuparp S, Strakova Z and Fazleabas AT (2001) The role of chorionic gonadotropin (CG) in blastocyst implantation. Arch Med Res 32, 627–634.[CrossRef][ISI][Medline]

Tamura M, Sebastian S, Yang S, Gurates B, Fang Z and Bulun SE (2002) Interleukin-1beta elevates cyclooxygenase-2 protein level and enzyme activity via increasing its mRNA stability in human endometrial stromal cells: an effect mediated by extracellularly regulated kinases 1 and 2. J Clin Endocrinol Metab 87, 3263–3273.[Abstract/Free Full Text]

Tsai SJ, Juengel JL and Wiltbank MC (1997) Hormonal regulation of monocyte chemoattractant protein-1 messenger ribonucleic acid expression in corpora lutea. Endocrinology 138, 4517–4520.[Abstract/Free Full Text]

Ueda A, Ishigatsubo Y, Okubo T and Yoshimura T (1997) Transcriptional regulation of the human monocyte chemoattractant protein-1 gene. Cooperation of two NF-kappaB sites and NF-kappaB/Rel subunit specificity. J Biol Chem 272, 31092–31099.[Abstract/Free Full Text]

Uzumcu M, Coskun S, Jaroudi K and Hollanders JM (1998) Effect of human chorionic gonadotropin on cytokine production from human endometrial cells in vitro. Am J Reprod Immunol 40, 83–88.[ISI][Medline]

Zhou XL, Lei ZM and Rao CV (1999) Treatment of human endometrial gland epithelial cells with chorionic gonadotropin/luteinizing hormone increases the expression of the cyclooxygenase-2 gene. J Clin Endocrinol Metab 84, 3364–3377.[Abstract/Free Full Text]

Submitted on June 23, 2003; resubmitted on March 19, 2004; accepted on July 20, 2004.