Differential Expression and Regulation of Microsomal Prostaglandin ...

0021-972X/03/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 88(12):6040 – 6047 Copyright © 2003 by The Endocrine Society doi: 10.1210/jc.2003-030618

Differential Expression and Regulation of Microsomal Prostaglandin E2 Synthase in Human Fetal Membranes and Placenta with Infection and in Cultured Trophoblast Cells MARINA PREMYSLOVA, WEI LI, NADIA ALFAIDY, ALAN D. BOCKING, KAREN CAMPBELL, WILLIAM GIBB, AND JOHN R. G. CHALLIS Canadian Institute of Health Research (M.P., W.L., N.A., J.R.G.C.), Departments of Physiology, Obstetrics, Gynecology and Medicine, University of Toronto, Toronto, Ontario, Canada M5S 1A8; Department of Obstetrics and Gynecology and Physiology (A.D.B., K.C.), University of Western Ontario, London, Ontario, Canada N6A 5B8; and Departments of Obstetrics and Gynecology (W.G.), University of Ottawa, Ottawa, Ontario, Canada K1N 6N5 We have evaluated the effect of chorioamnionitis on the protein expression of microsomal and cytosolic prostaglandin E2 synthases (mPGES and cPGES) in preterm human placentae (PL) and fetal membranes (FM), by Western blot and immunohistochemistry, as well as the regulatory effect of IL-1␤ and TNF-␣ on mPGES, cPGES, and cyclooxygenase (COX)-2 expression in villous trophoblast (VT) and chorion trophoblast (CT) cell cultures. mPGES localized to the syncytiotrophoblast and vascular endothelium in PL and to the amnion epithelium, CT, and decidual cells in FM. cPGES protein was localized only to the syncytiotrophoblast in PL and had the same profile of expression as mPGES in FM. With infection, there was an increase in mPGES expression in PL and a decrease in the expression in FM. cPGES protein did not change

P

ROSTAGLANDIN E2 SYNTHASE (PGES) is a terminal enzyme in the cyclooxygenase (COX)-mediated PGE2 biosynthetic pathway. PGE2 is involved in the onset of parturition and promotes uterine contractility. There are two isoforms of PGES, a microsomal (mPGES) (1) and cytosolic (cPGES) form (2). Believed to be involved in inflammatory processes, mPGES may also be linked to the inducible pathway for prostaglandin synthesis. Induction of mPGES expression by proinflammatory stimuli has been shown in A-549 cells (3); in rat macrophages, brain, lung, spleen, stomach, kidney, and testis tissues (4); in bovine endometrium (5); and in human vascular smooth muscle cells (6), orbital fibroblasts (7), and rheumatoid synovial cells (8). cPGES is believed to be constitutively expressed and is unaltered by proinflammatory stimuli in various cells and tissues, except in rat brain, where it is significantly increased (2). The concentrations of inflammatory cytokines such as TNF-␣ and IL-1␤ are increased in amniotic fluid of women in preterm labor in the presence of chorioamnionitis (an inflammatory process involving the chorion, its fetal blood Abbreviations: COX, Cyclooxygenase; cPGES, cytosolic PGES; CT, chorion trophoblast; FM, fetal membrane(s); IHC, immunohistochemistry; PGDH, prostaglandin dehydrogenase; PGES, prostaglandin E2 synthase; mPGES, microsomal PGES; PL, placenta(e); RT, room temperature; ST, syncytiotrophoblast; TBS, Tris-buffered saline; VE, vascular endothelium; VT, villous trophoblast.

in either PL or FM with infection. In VT cells in culture, IL-1␤ up-regulated COX-2 protein expression but did not affect mPGES. However, TNF-␣ increased both mPGES and COX-2 protein expression in these cells. In CT cells in culture, IL-1␤ and TNF-␣ increased both mPGES and COX-2 protein levels. Neither IL-1␤ nor TNF-␣ affected cPGES in either VT or CT cells. We conclude that protein levels of mPGES, as well as COX-2, can be stimulated by cytokines, potentially contributing to the increased prostaglandin production at the time of infection-driven preterm labor. However, multiple mechanisms, which apparently are inductor- and cell-type-specific, exist for the regulation of these enzymes. (J Clin Endocrinol Metab 88: 6040 – 6047, 2003)

vessels, the umbilical cord, and the amnion) (9 –12). These cytokines activate the PG biosynthetic pathway primarily via induction of cyclooxygenase-2 (COX-2) and increased synthesis of PGE2 by human amnion (13), chorion, and decidual (14) cells. Furthermore, IL-1␤ can stimulate PG output not only by increasing expression of COX-2 in human villous and chorion trophoblast (VT and CT) cells but also by decreasing expression of prostaglandin dehydrogenase (PGDH) in VTs (15, 16). However, the importance of PGES, which acts downstream of COX-2, in regulating PG synthesis during preterm labor of infectious etiology in human placenta (PL) and fetal membranes (FM) is unknown. Previously, no changes in mPGES mRNA distribution with term labor or with infection-driven preterm labor were observed in amnion, choriodecidual, or villous tissues (17). However, we found increased mPGES protein expression in chorion (18), although not in myometrium (19) with labor, and others have reported increased mPGES mRNA expression in myometrial cells after treatment with IL-1␤ (20). In the present study, we examined whether PGES protein expression changed with infection-driven preterm labor, whether PGES enzyme in human PL and FM was regulated by proinflammatory cytokines, and whether there was a coordinate regulation of mPGES, cPGES, and COX-2 by cytokines. We examined mPGES and cPGES immunohistochemical localization and protein expression in preterm hu-

6040

Premyslova et al. • mPGES Expression with Infection

man membranes and PL with chorioamnionitis, and we determined the regulatory effect of IL-1␤ and TNF-␣ on mPGES, cPGES, and COX-2 protein expression using purified cultures of VT and CT cells. Materials and Methods Tissue collection PL and FM from preterm (24.5–36 wk gestation) vaginal delivery with (n ⫽ 16) or without (n ⫽ 16) chorioamnionitis were collected at St. Josephs Health Centre (London, Ontario, Canada). FM were separated from PL. Rolls of FM and PL tissue were immediately snap frozen in liquid nitrogen for Western blotting analysis or fixed in 4% paraformaldehyde and paraffin embedded for immunohistochemistry (IHC). Diagnosis of chorioamnionitis was confirmed by histopathology. Leukocyte infiltration was used as a criterion for chorioamnionitis. Term PL (n ⫽ 8) and membranes were also collected from normal uncomplicated pregnancies, at term, after elective cesarean section in the absence of labor (Mount Sinai Hospital, Toronto, Canada). Patient consent and ethical approval were obtained in accordance with the guidelines of Research Ethics Committees of Mount Sinai Hospital and the University of Toronto, as well as the Review Board for Research Involving Human Subjects of the University of Western Ontario (London, Ontario, Canada). Term tissues were used as a source of PL and CT cells, not affected by inflammation, labor or any complication of pregnancy that could change the expression of prostaglandin synthesizing enzymes.

IHC for mPGES and cPGES The PL and FM tissues with and without chorioamnionitis (10-␮m sections) were deparaffinized with xylene (EMD Chemicals, Inc., Gibbstown, NJ) and rehydrated in graded ethanol washes (six sections for each of four tissue groups were processed). Endogenous peroxidase activity was inhibited by treatment with 0.3% hydrogen peroxide in PBS for 30 min. The sections were then washed in 0.01 m PBS and incubated for 2 h at room temperature (RT) with normal goat serum to block nonspecific binding. Sections were incubated overnight, at ⫹4 C, with primary polyclonal antibodies against a peptide corresponding to amino acids 59 –75 (CRSDPDVERSLRAHRND) of human mPGES (Cayman Chemical, Ann Arbor, MI) or with primary polyclonal antibodies against a peptide corresponding to amino acids 58 – 67 (CIDPNDSKHK) of human cPGES (Cayman Chemical), both used at a dilution of 1:200 in antibody dilution buffer (1% BSA in 0.01 m PBS). Primary antibody binding was visualized using the Vectastain ABC Kit (Vector Laboratories, Burlingame, CA): sections were incubated for 2 h with antirabbit biotinylated secondary antibodies, at RT, followed by a PBS wash and 2-h incubation, at RT, with avidin biotin-peroxidase complex. 3,3⬘ diaminobenzidine (Sigma St. Louis, MO) was used as chromagen. The sections were counterstained with Carazzi’s hematoxylin, dehydrated in graded ethanol, and cover slips were applied. For negative controls, the primary mPGES and cPGES antibodies were pre-absorbed overnight with 10 ␮g/ml mPGES or cPGES blocking peptides (Cayman Chemical), respectively.

Western blot analysis of mPGES and cPGES protein expression in human preterm PL and FM with and without chorioamnionitis Levels of mPGES and cPGES protein in PL and FM tissues were assessed by Western blotting. Proteins were extracted from frozen tissues by homogenization on ice in T-PER protein extraction buffer (Pierce, Rockford, IL) containing Mini EDTA-free protease inhibitors (Boehringer Mannheim Biochemicals, Mannheim, Germany) and 100 ␮m sodium orthovanadate (Na2VO2, Sigma). Insoluble material was pelleted by centrifugation at 14,000 ⫻ g for 10 min at ⫹4 C, and the clarified lysates were collected. Protein concentrations in the supernatants were determined by the Bradford assay (21). Proteins samples (10 ␮g/lane for mPGES and 20 ␮g/lane for cPGES) were diluted and boiled for 3 min in sodium dodecyl sulfate-containing sample buffer (22) and then resolved on a 14% (vol/vol) sodium dodecyl sulfate-polyacrylam-

J Clin Endocrinol Metab, December 2003, 88(12):6040 – 6047 6041

ide gel. The proteins were transferred to a nitrocellulose membrane (23). Transfer was confirmed by protein visualization with Ponceau S (Sigma). Membranes were blocked for 2 h at 25 C in 5% skimmed milk powder diluted in Tris-buffered saline (TBS)-Tween, pH 7.6 [50 mm Tris-HCl, 150 mm NaCl, and 0.1% (vol/vol) Tween 20]. After blocking, membranes were incubated for 1 h with mPGES (1:1000 in blocking solution) or cPGES (1:1000) antibodies. Blots were washed 5 times (10 min each) with TBS-Tween and subsequently incubated for 1 h with horseradish peroxidase-conjugated goat antirabbit IgG (1:2000 in blocking solution) (Amersham Life Science, Quebec, Canada). Then, membranes were washed 5 times (10 min each) with TBS-Tween, and proteins bands were visualized with enhanced chemiluminescence Western blotting detection reagents (Amersham Biosciences, Buckinghamshire, UK). The membranes were exposed to X-OMAT blue film (Kodak Scientific Imaging Product, Rochester, NY). The relative intensities of immunoreactive bands were measured by scanning and analysis with Scion Image software (Scion Image Corporation, Frederick, MD). Protein bands were digitized, and the mean pixel density for each band was analyzed to obtain relative optical density units for each protein. Sample loading was standardized to the expression of ␤-globulin using specific antibodies (Santa Cruz Biotechnology, Inc., Santa Cruz, CA; 1:2000) (24). To compare measurements between different blots, a reference PL or FM sample was included in each gel.

VT and CT cell culture and treatment VT and CT cells were isolated and cultured using a modification of the technique described by Kliman et al. (25, 26). Briefly, approximately 60 g of villous tissue was digested with 0.125% trypsin (Life Technologies Inc., Grand Island, NY) and 0.02% deoxyribonuclease I (Sigma) in DMEM, three times for 30 min. The chorion, after the removal of decidua and blood clots, was digested three times for 60 min with the same digestion medium supplemented with 0.2% collagenase (Sigma). The dispersed villous or chorio-decidual cells were filtered through a 200␮m-pore-size nylon gauze and loaded onto a 42-ml preformed discontinuous Percoll (Sigma) gradient of 5–70%, in 5% steps of 3 ml each, then centrifuged (1,200 ⫻ g) at RT for 20 min. Cells between densities of 1.049 and 1.062 g/ml were collected and plated in 100-mm-diameter Petri dishes at 2 ⫻ 106 cells/ml in 15 ml DMEM containing 10% fetal bovine serum (Life Technology, Inc.) and 1% (vol/vol) of antibiotic-antimycotic solution (10,000 U/ml penicillin G, 10 mg/ml streptomycin sulfate, and 25 mg/ml amphotericin B in 0.9% sodium chloride) (Sigma). Cells were then cultured for 3 d at 37 C in 5% CO2-95% air. After 3 d, PL trophoblast cells aggregated to form a syncytium, and CT cells formed clumps. Chorion and PL cells were then washed twice with Hanks’ solution, equilibrated at 37 C, and cultured for 16 h in DMEM free of FCS and antibiotics. Cells viability was assessed by using trypan blue exclusion and was higher then 80% in all experiments.

Effect of cytokines After 16 h incubation in DMEM free of fetal bovine serum and antibiotics, the medium was changed, and the cells were treated with: 1) medium alone; and 2) IL-1␤ (recombinant human IL-1␤, 0.1 ng/ml; 1 ng/ml and 10 ng/ml) (Sigma) for 4 h, or with TNF-␣ (recombinant human TNF-␣, 0.1 ng/ml; 1 ng/ml and 10 ng/ml) (Sigma) for different periods of time (2, 4, 6, and 24 h). CT cells were treated with IL-1␤ (1 ng/ml) or TNF-␣ (10 ng/ml) for 4 h. We chose 4 h for IL-1␤ treatment, based on our preliminary experiments when we failed to observe any effect of IL-1␤ on mPGES expression after 24 h incubation. Cells were characterized by cytokeratin (an epithelial cell marker) or vimentin (a fibroblast cell marker) immunostaining as published previously (15). About 90% of PL and chorionic cells were cytokeratin positive. All experiments were performed in duplicate for PL cells and in singleton or duplicate for CTs (depending on the cell yield after extraction).

Western blotting for mPGES, cPGES, and COX-2 expression in cultured VT and CT cells Cultured cells were collected, proteins extracted, and Western blotting carried out as described for tissue samples. The membranes were

6042

J Clin Endocrinol Metab, December 2003, 88(12):6040 – 6047


FIG. 1. Immunohistochemical localization of mPGES and cPGES protein in preterm human PL and FM without and with chorioamnionitis. Brown color indicates positive staining. A–D, Immunostaining for mPGES (A and B in control and infected PL, respectively; C and D in control and infected FM, respectively). E–H, Immunostaining for cPGES (E and F in control and infected PL, respectively; G and H, in control and infected FM, respectively). For negative controls, primary mPGES and cPGES antibodies were pre-absorbed with mPGES (I and J for FM and PL, respectively) or cPGES (K and L for FM and PL, respectively) blocking peptides. AE, Amnion epithelium; D, decidua. mPGES (A), rather then cPGES (E), localizes to VE of human PL. There is a loss of CT layer in FM with CA and diminished mPGES (D) and cPGES (H) immunostaining, compared with controls (C and G). Original magnification for all panels, except D and H, ⫻400; for D and H, ⫻200.



sequentially probed with mPGES, cPGES, and COX-2 antibodies (1:1000 in blocking solution, Cayman Chemical).

preabsorption of the antibodies with the corresponding blocking peptides (Fig. 1; I, J, K, and L).

Data analysis

mPGES and cPGES protein expression in human preterm FM and PL tissues with and without chorioamnionitis

Student’s paired and unpaired t tests were used to compare the results for infected and noninfected tissue samples and to determine the effects of IL-1␤ and TNF-␣ on CT cells. One-way ANOVA, followed by Student-Newman-Keuls test, was used to compare dose effects of the cytokines on VT cells. Data are expressed as the mean ⫾ sem, with P ⫽ 0.05 as the limit of significance (Sigma Stat; Jandel Scientific Software, San Rafael, CA).

Results Immunohistochemical localization of mPGES and cPGES in human preterm PL and FM with chorioamnionitis

In PL, mPGES protein was localized to the syncytiotrophoblast (ST) and vascular endothelium (VE) in noninfected (Fig. 1A) as well as in infected with CA tissues (Fig. 1B). In noninfected FM, mPGES was localized to amnion epithelium (AE), CT, and decidual cells (Fig. 1C). With infection, there was a loss of normal structure of the membrane, loss of the chorion layer, and diminished mPGES immunostaining (Fig. 1D). cPGES was localized to ST in normal (Fig. 1E) and in infected (Fig. 1F) PL tissues. In FM, cPGES had the same pattern of expression as mPGES (Fig. 1, G and H, show cPGES localization in noninfected and infected FM, respectively). mPGES and cPGES immunostaining in FM and PL was entirely abolished after

FIG. 2. Western blot analysis of mPGES protein expression in noninfected with chorioamnionitis (⫺CA) and infected (⫹CA) human preterm PL and FM. A and B, Representative Western blots. A, mPGES expression in noninfected (⫺CA) and infected (⫹CA) PL tissues; B, mPGES expression in noninfected (⫺CA) and infected (⫹CA) FM. C, Histogram depicting the relative levels of mPGES expression in PL and FM without chorioamnionitis (open bars) and with chorioamnionitis (filled bars). Each bar represents the mean ⫾ SEM of 16 samples. *, P ⬍ 0.05, PL: ⫺CA vs. ⫹CA; **, P ⬍ 0.05, FM: ⫺CA vs. ⫹CA. a, Reference PL sample; b, reference FM sample.

By Western blotting, immunoreactive mPGES protein was present as a single band with the expected molecular weight of 16 kDa; immunoreactive cPGES was present as a 23-kDa band. In PL samples, the level of mPGES was increased (P ⬍ 0.05) with infection (Fig. 2, A and C). However, there was a decrease in the expression of mPGES within FM with infection (P ⬍ 0.05, Fig. 2, B and C), which is likely attributable to the loss of cells from the CT layer (see Fig. 1). There was no significant effect of chorioamnionitis on cPGES expression within these tissues (Fig. 3, A–C). Effects of IL-1␤ and TNF-␣ on mPGES, COX-2, and cPGES expression in primary VT and CT cell cultures

The increase in mPGES expression in infected PL tissue might be attributable to the elevated level of proinflammatory cytokines, mainly IL-1 ␤ and TNF-␣, known to be secondary to chorioamnionitis. To test this hypothesis and to show whether the changes in mPGES expression correlated with any changes in COX-2 and cPGES expression, we determined the effect of IL-1␤ and TNF-␣ on mPGES, COX-2, and cPGES protein expression in primary cultures of VT and CT cells, isolated from PL, at term, after elective cesarean section. In cultured VT cells, IL-1␤ (0 –10 ng/ml) did not affect

FIG. 3. Western blot analysis of cPGES protein expression in noninfected-with-chorioamnionitis (⫺CA) and infected (⫹CA) human preterm PL and FM. A and B, Representative Western blots. A, cPGES expression in noninfected (⫺CA) and in infected (⫹CA) PL tissues; B, cPGES expression in noninfected (⫺CA) and infected (⫹CA) FM. C, Histogram depicting the relative levels of cPGES expression in PL and FM without chorioamnionitis (open bars) and with chorioamnionitis (filled bars). Each bar represents the mean ⫾ SEM of 16 samples. a, Reference PL sample; b, reference FM sample.

6044


FIG. 4. Effect of IL-1␤ on mPGES, COX-2, and cPGES protein expression in VT cells. Cultured for 3 d, VT cells were maintained in serum and antibiotic [free medium for 16 h and then treated with an increasing dose of IL-1␤ (0 –10 ng/ml) for 4 h]. Proteins were extracted and subjected to Western blotting with mPGES-, COX-2-, or cPGESspecific antibodies. Histograms of relative levels of mPGES (A), COX-2 (B), and cPGES (C) expression are given under respective representative blots. Each bar represents the mean ⫾ SEM from at least three experiments. *, P ⬍ 0.05 vs. control group.

mPGES or cPGES expression during 4 h of incubation (Fig. 4, A and C). In contrast, COX-2 protein expression was upregulated by IL-1␤ at 10 ng/ml (Fig. 4B). TNF-␣ caused a 2-fold increase in mPGES expression at 10 ng/ml (Fig. 5A). COX-2 was also up-regulated by TNF-␣ (Fig. 5B). The increase was dose-dependent for both enzymes, with maximum effect at the concentration of 10 ng/ml TNF-␣ for mPGES and 1.0 ng/ml for COX-2. cPGES was unaffected by the range of TNF-␣ concentrations (Fig. 5C). The timedependent character of TNF-␣ stimulation of mPGES and COX-2 is shown in Fig. 6, A and B. The increase in mPGES


FIG. 5. The effect of TNF-␣ (0 –10 ng/ml) on mPGES, COX-2, and cPGES protein expression in cultured VT cells. Cells were cultured for 3 d, then maintained in serum and antibiotic-free medium for 16 h, and treated with increasing concentrations of TNF-␣ (0 –10 ng) for 4 h. Proteins were extracted and subjected to Western blotting with mPGES-, COX-2-, or cPGES-specific antibodies. Representative Western blots are shown. Histograms of relative levels of mPGES (A), COX-2 (B), and cPGES (C) expression are given under correspondent blots. The data are expressed as the mean ⫾ SEM from five to eight experiments. *, P ⬍ 0.05 vs. control group.

and COX-2 expression occurred within the first 2– 4 h of treatment. After 4 h, there was a decline in mPGES expression, whereas up-regulation of COX-2 lasted at least 24 h. cPGES remained unchanged under the same conditions of TNF-␣ treatment (Fig. 6C). In CT cells, IL-1␤ (1 ng/ml) and TNF-␣ (10 ng/ml) up-regulated both mPGES (Fig. 7, A and B) and COX-2 enzymes (Fig. 7, C and D) but did not affect cPGES expression (Fig. 7, E and F).


FIG. 6. The effect of TNF-␣ on mPGES, COX-2, and cPGES protein expression in cultured VT cells with time (0 –24 h). Cells were cultured for 3 d, then maintained in serum and antibiotic-free medium for 16 h, and treated with TNF-␣ (10 ng/ml) for different periods of time (0, 2, 4, 6, and 24 h). Proteins were extracted and subjected to Western blotting. Membranes were probed with mPGES-, COX-2-, and cPGESspecific antibodies. Representative Western blots are shown. Histograms of relative levels of mPGES (A), COX-2 (B), and cPGES (C) expression are given under correspondent blots. The data are expressed as the means ⫾ SEM from three experiments.

Discussion

We have demonstrated immunohistochemical localization of mPGES and cPGES proteins in human preterm FM and PL and shown that there is increased mPGES protein expression in human preterm PL with chorioamnionitis. In addition, we have shown a stimulatory effect of TNF-␣ on mPGES and COX-2 protein expression in cultured VT and CT cells prepared from term PL without infection. IL-1␤, though not affecting mPGES expression in VT cells, up-regulated COX-2 in these cells and caused an increase in mPGES as well as in COX-2 protein expression in CTs. Regulation of cPGES was clearly different from that of mPGES or COX-2 and was


unaltered by the amount of proinflammatory cytokines used in the present experiments. IHC showed a loss of the normal structure of the amniochorion and especially of the chorion layer in tissues from preterm labor patients with infection. There was diminished mPGES as well as cPGES immunostaining in these tissues. We have reported previously similar observations showing that reduction of mRNA, protein, and enzyme activity of 15-OH PGDH in chorion was accounted for by the loss of the trophoblast cells that accompany the infective process (27–29). Our finding of expression of mPGES, rather then cPGES, in VE implies that mPGES may be important for regulation of PL vascularity (and thus, PL blood flow) and that it may be acutely regulated. In the FM, cPGES had the same pattern of staining as mPGES, localizing to the amnion epithelial, CT, and decidual cell. Based on our culture experiments, its appearance is more constitutive and not amenable to short-term regulation at least by two proinflammatory cytokines. By Western blotting, we have shown increased mPGES expression in PL and decreased mPGES expression in FM with infection. The decrease may reflect the mechanical loss of CT cells in infected FM. However, there was no change in cPGES expression in these tissues, despite the same profile of immunolocalization for both isoforms. Our use of fullthickness membranes for these experiments might explain this difference. We suggest that mPGES, but not cPGES, has a relatively higher protein expression in chorion, compared with amnion epithelium and decidua. Thus, we measured a detectable decrease in mPGES, but not in cPGES protein expression, in the infected full-thickness FM, which exhibits loss of chorion layer with infection. The increase in mPGES in PL could result from stimulatory effects of proinflammatory cytokines. Indeed, we speculated that cytokines might also increase initially mPGES in chorion at an earlier time point in the infective process. The decrease seen in intact tissues would reflect then a later loss of chorion cells. Levels of proinflammatory cytokines are known to be elevated with chorioamnionitis (9 –12). Previous studies have shown the induction of COX-2 enzyme by proinflammatory cytokines (30 –33), and coordinate induction of COX-2 and mPGES (4, 5, 7). Therefore, we investigated whether mPGES was coordinately regulated with COX-2 in VT and CT cells. In VT cells, IL-1␤ had no effect on mPGES expression, but it induced the expression of COX-2 in a dose-dependent manner. In CTs, also cultured for 4 h, IL-1␤ up-regulated both mPGES and COX-2 enzymes. Although we did not measure PGE2 output in the present study, these results are consistent with increased PGE2 output from VTs (starting already after 4 h of incubation) and from CTs with Il-1 ␤ treatment, reported from our laboratory previously (15). It is not clear why VT and CT cells exhibited different responses to IL-1␤, although this could be related to the tissue-specific balance of pro- and anti-inflammatory stimuli. The up-regulation of COX-2, accompanied with unchanged mPGES expression in VTs under IL-1␤ treatment, may suggest also that there are different mechanisms involved in the regulation of the two enzymes. In contrast, TNF-␣ stimulated mPGES as well as COX-2 protein in VTs in a dose- and

6046



FIG. 7. The effect of IL-1␤ and TNF-␣ on mPGES, COX-2, and cPGES protein expression in cultured CT cells. CT cells were cultured for 3 d, then maintained in serum and antibioticfree medium for 16 h, and treated with medium alone, with 1 ng/ml IL-1␤, or 10 ng/ml TNF-␣ for 4 h. Proteins were extracted and subjected to Western blotting with mPGES-, COX-2-, or cPGES-specific antibodies. Representative Western blots are shown. Histograms of relative levels of mPGES (A and B), COX-2 (C and D), and cPGES (E and F) protein expression are given under correspondent blots. Open bars, Expression in control cells; hatched bars, in cells treated with IL-1␤; solid bars, in cells treated with TNF. Data are expressed as the mean ⫾ SEM from five to eight experiments. *, P ⬍ 0.05.

time-dependent manner. The increase in COX-2 expression at 0 –1 ng/ml TNF-␣ was followed by a decrease at 10 ng/ml (Fig. 5B). We speculate that this decrease might be the result of COX-2 negative self-regulation (34). CT cells also responded to TNF-␣ treatment by an increased level of mPGES and COX-2 proteins. These data suggest that the mechanisms of mPGES and COX-2 regulation are inductor specific. The shapes of dose- and time-response plots, which are not the same for mPGES and COX-2, support the possibility of the existence of different mechanisms responsible for their regulation. Further studies are required to examine whether there are differences in cellular and/or subcellular colocalization of these enzymes, and to determine changes that occur with acute stimulation. cPGES was unchanged in both cell types under either IL-1␤ or TNF-␣ action. These data are in agreement with studies using other cell and tissue types (2). The current findings regarding the stimulatory effect of

inflammatory cytokines on mPGES are consistent with previous reports on the ability of cytokines to up-regulate mPGES in other cell systems and tissues (1, 3– 8). Consistent with this, in human gestational tissues, there is increased expression of mPGES mRNA after IL-1␤ treatment in myometrium (20), and there is increased mPGES protein expression in chorion with labor (18). However, others did not find differences in mPGES mRNA expression with term labor in amnion, chorio-decidua, or villous tissues (17) or in human myometrium with labor (19). The reasons for these conflicting results are not apparent at the present time. In conclusion, our results suggest that chorioamnionitis up-regulates mPGES protein expression in human preterm PL and that proinflammatory cytokines, produced in response to chorioamnionitis, might underlie the effect. Thus, mPGES seems to be involved in the process of inflammation in the human PL and FM, and may contribute to an overall increase in PGE2



production at the time of preterm labor with infectious etiology. However, the mechanisms of mPGES and COX-2 induction do not seem to be coordinate in the trophoblast cells and may be both cell-type- and inductor-specific.

15.

Acknowledgments

16.

Received April 9, 2003. Accepted September 8, 2003. Address all correspondence and requests for reprints to: Marina Premyslova, Department of Physiology, Obstetrics, and Gynecology, Medical Science Building, University of Toronto, 1 King’s College Circle, Toronto Ontario, Canada M5S 1A8. E-mail: [email protected].

References 1. Jakobsson PJ, Thoren S, Morgenster R, Samuelsson B 1999 Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci USA 96:7220 –7225 2. Tanioka T, Nakatani Y, Semmyo N, Kojima F, Ikeda T, Fueki M, Ueno A, Oh S, Kudo I 2000 Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled with cyclooxygenase-1 in immediate prostaglandin E2 biosynthesis. J Biol Chem 275:32775–32782 3. Forsberg L, Leeb L, Thoren S, Morgenstern R, Jakobsson PJ 2000 Human glutathione dependent prostaglandin E synthase: gene structure and regulation. FEBS Lett 471:78 – 82 4. Murakami M, Naraba H, Tanioka T, Semmyo N, Nakatani Y, Kojima F, Ikeda T, Fueki M, Ueno A, Oh-ishi S, Kudo I 2000 Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol Chem 275:32783–32792 5. Parent J, Chapdelaine P, Sirois J, Fortier M 2002 Expression of microsomal prostaglandin E synthase in bovine endometrium: coexpression with cyclooxygenase type 2 and regulation by interferon-␶. Placenta 143:2936 –2943 6. Soler M, Camacho M, Escudero JR, Iniguez MA, Vila L 2000 Human vascular smooth muscle cells but not endothelial cells express prostaglandin E synthase. Circ Res 87:504 7. Han R, Tsue S, Smith T 2001 Up-regulation of prostaglandin E2 synthesis by interleukin-1␤ in human orbital fibroblasts involves coordinate induction of prostaglandin-endoperoxide H synthase and glutathione-dependent prostaglandin E2 synthase expression. J Biol Chem 277:16355–16364 8. Stichtenoth D, Thoren S, Bian H, Peters-Golden M, Jacobsson PJ, Crofford LJ 2001 Microsomal prostaglandin E synthase is regulated by proinflammatory cytokines and glucocorticoids in primary rheumatoid synovial cells. J Immunol 167:469 – 474 9. Dollner H, Vatten L, Halgunset J, Halgunset J, Rahimipoor S, Austgulen R 2002 Histological chorioamnionitis and umbilical serum levels of pro-inflammatory cytokines and cytokines inhibitors. Br J Obstet Gynaecol 109:534 –539 10. Romero R, Brody DT, Oyarzun E, Mazor M, Wu YK, Hobbins JC, Mitchell MD 1989 Infection and labor. III: A signal for the onset of parturition. Am J Obstet Gynecol 160:1117–1123 11. Romero R, Wu YK, Sirtori M, Oyarzun E, Mazor M, Hobbins JC, Mitchell MD 1989 Amniotic fluid concentrations of prostaglandin F2␣ 13, 14-dihydro15-keto-prostaglandin F2␣ (PGFM) and 11-deoxy-13, 14-dihydro-15-keto-11, 16-cyclo-prostaglandin E2 (PGEM-LL) in preterm labor. Prostaglandins 37: 149 –161 12. Tsunoda H, Tamatani T, Oomoto Y, Hirai Y, Kasahara T, Iwasaki H, Onozaki K 1990 Changes in interleukin-1 level in human amniotic fluid with gestational ages and delivery. Microbiol Immunol 34:377–385 13. Pollard JK, Mitchell MD 1993 Tumor necrosis factor alpha stimulates amnion prostaglandin biosynthesis primarily via an action on fatty acid cyclooxygenase. Prostaglandins 46:499 –510 14. Pollard JK, Thai D, Mitchell MD 1993 Evidence for a common mechanism of action of interleukin-1 beta, tumor necrosis factor-␣, and epidermal growth

17.

18. 19.

20. 21. 22. 23. 24. 25. 26.

27.

28. 29.

30. 31. 32. 33. 34.

factor on prostaglandin production in human chorion cells. Am J Reprod Immunol 30:146 –153 Pomini F, Caruso A, Challis JRG 1999 Interleukin-10 modifies the effect of interleukin-1␤ and tumor necrosis factor-␣ on the activity and expression of prostaglandin H synthase-2 and the NAD-dependent 15-hydroxyprostaglandin dehydrogenase in cultured term human villous trophoblast and chorion trophoblast cells. J Clin Endocrinol Metab 84:4665– 4651 Keelan JA, Mesnage S, Goodwin V, Mitchel MD 1998 Inhibition of 15hydroxyprostaglandin dehydrogenase expression and activity by cytokines in human placental trophoblasts. J Soc Gynecol Invest 5:48A Marvin KW, Parks B, Keelan JA, Sato A, Mitchel MD 2002 Expression of microsomal prostaglandin E synthase in human gestational tissues with labor at term and with infection and labor preterm. J Soc Gynecol Investig 9(Suppl): 452 (Abstract) Alfaidy N, Sun M, Challis JRG, Gibb W 2003 Expression of membrane prostaglandin E synthase (mPGES) in the human placenta and fetal membranes and the effect of labor. Endocrine 3:219 –226 Giannoulias D, Alfaidy N, Holloway A, Gibb W, Sun M, Lye SJ, Challis JRG Expression of prostaglandin I synthase, but not prostaglandin E synthase, changes in myometrium of women at term pregnancy. J Clin Endocrinol Metab 87:5274 –5282 Slater DM, Astle S, Thornton T 2002 Expression of inducible prostaglandin E synthase in human gestational tissues. J Soc Gynecol Investig 9(Suppl N1): 451 (Abstract) Bradford MM 1976 A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248 –254 Laemly UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680 – 685 Towbin H, Staehelin T, Gordon J 1979 Electrophoretic transfer of protein from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350 – 4354 Europe-Finner GN, Phaneuf S, Tolkovsky AM, Watson SP, Lopes-Bernal A 1994 Down-regulation of G ␣ in human myometrium in term and preterm labor: a mechanism for parturition. J Clin Endocrinol Metab 79:1835–1839 Kliman HJ, Nestler JE, Sermasi E, Sanger JM, Strauss JF 1986 Purification, characterization and in vitro differentiation of cytotrophoblasts from human term placenta. Endocrinology 118:1567–1582 Sun K, Yang K, Challis JRC 1997 Differential regulation of 11-␤ hydroxysteroid dehydrogenase type I and II by nitric oxide in cultured human placental syncytiotrophoblast and chorionic cell preparation. Endocrinology 138:4912– 4920 Pomini F, Patel F, Mancuso S, Challis JRG 2000 Activity and expression of NAD dependent 15 hydroxyprostaglandin dehydrogenase in cultured chorion trophoblast and villous trophoblast cells and chorion explants, before and with spontaneous labor. Am J Obstet Gynecol 182(1 Pt 1):221–226 Van Meir CA, Sangha RK, Walton JC, Matthew SG, Keirse MJ, Challis JRG 1997 15-Hydroxiprostaglandin dehydrogenase: implications in preterm labor with and without infection. J Clin Endocrinol Metab 83:969 –976 Van Meir CA, Sangha RK, Walton JC, Matthew SG, Keirse MJ, Challis JRG 1996 Immunoreactive 15-hydroxyprostaglandin dehydrogenase (PGDH) is reduced in fetal membranes of patients at preterm delivery in the presence of infection. Placenta 17:291–297 Dinarello CA, Sheldon MW 1993 The role of IL-1 in disease. N Engl J Med 328:106 –113 Mitchell MD, Romero RJ, Edwin SS, Trautman MS, Mitchell MD 1995 Prostaglandins and parturition. Reprod Fertil Dev 7:623– 632 Kent ASH, Sun MY, Sullivan MH, Elder MG 1993 The effect of interleukins 1␣ and 1␤ on prostaglandin production by human trophoblast cells from first and third trimesters. J Clin Endocrinol Metab 80:1641–1646 Imseis HM, Zimmerman PD, Samuels P, Kniss DA 1997 Tumor necrosis factor-␣ induces cyclooxygenase-2 gene expression in first trimester trophoblasts: suppression by glucocorticoids and NSAIDs. Placenta 18:521–526 Poligone B, Baldwin A 2001 Positive and negative regulation of NF-kB by COX-2. J Biol Chem 276:38658 –38664