Sphingosine 1-phosphate in amniotic fluid modulates cyclooxygenase-2 expression in human amnion-derived WISH cells.

The metabolism of arachidonic acid, in particular the generation of prostaglandins (PGs), has been proposed to play a key role in the regulation of labor. Moreover, several extracellular proteins have been reported to modulate PG synthesis in amnion cells. In this study, we found that lipid components dissolved in the amniotic fluid modulate PG synthesis in WISH human amnion cells and identified one of these components as a sphingosine 1-phosphate (S1P). WISH cells express several S1P receptors including S1P1, S1P2, and S1P3. When WISH cells were stimulated with S1P, PGE2 synthesis increased in a concentration-dependent manner, showing maximal activity at around 100 nM. S1P treatment also caused the up-regulation of cyclooxygenase-2 (COX-2) mRNA and protein, which was apparent within 3-12 h of stimulation. In terms of the intracellular signaling pathway of S1P-induced WISH cell activation, we found that S1P stimulated two kinds of MAPK, ERK, and p38 kinase. We examined the roles of these two MAPKs in S1P-induced COX-2 expression. S1P-induced COX-2 expression was blocked completely by PD-98059 but not by SB-203580, suggesting that ERK has a critical role in the process. Transfection of S1P1 or S1P3 but not of S1P2 antisense oligonucleotide inhibited S1P-induced COX-2 expression and PGE2 production in WISH cells, indicating the involvements of S1P1 and S1P3 in the processes. This study demonstrates the physiological role of S1P in amniotic fluid and its effect on the modulation of COX-2 expression and PGs synthesis in WISH cells.

The metabolism of arachidonic acid, in particular the generation of prostaglandins (PGs), has been proposed to play a key role in the regulation of labor. Moreover, several extracellular proteins have been reported to modulate PG synthesis in amnion cells. In this study, we found that lipid components dissolved in the amniotic fluid modulate PG synthesis in WISH human amnion cells and identified one of these components as a sphingosine 1-phosphate (S1P). WISH cells express several S1P receptors including S1P 1, S1P 2 , and S1P 3 . When WISH cells were stimulated with S1P, PGE 2 synthesis increased in a concentration-dependent manner, showing maximal activity at around 100 nM. S1P treatment also caused the up-regulation of cyclooxygenase-2 (COX-2) mRNA and protein, which was apparent within 3-12 h of stimulation. In terms of the intracellular signaling pathway of S1P-induced WISH cell activation, we found that S1P stimulated two kinds of MAPK, ERK, and p38 kinase. We examined the roles of these two MAPKs in S1P-induced COX-2 expression. S1P-induced COX-2 expression was blocked completely by PD-98059 but not by SB-203580, suggesting that ERK has a critical role in the process. Transfection of S1P 1 or S1P 3 but not of S1P 2 antisense oligonucleotide inhibited S1P-induced COX-2 expression and PGE 2 production in WISH cells, indicating the involvements of S1P 1 and S1P 3 in the processes. This study demonstrates the physiological role of S1P in amniotic fluid and its effect on the modulation of COX-2 expression and PGs synthesis in WISH cells.
The arachidonic acid metabolism has been proposed to play a key role in the regulation of labor (1,2). In particular, generation of prostaglandins (PGs), 1 such as PGE 2 and PGF 2␣ , mod-ulate labor by inducing uterine contractions (1,2). PGE 2 is known to be the main prostanoid produced by amnion cells (1). Moreover, PGs can be synthesized from arachidonic acid by the activity of cyclooxygenase (COX) (3). Of the two known COXs, COX-2, which is induced by extracellular signals, plays a key role in the modulation of PG production and labor (3,4). Several cytokines, such as interleukin-1␤ (IL-1␤) and tumor necrosis factor ␣, have been shown previously (5,6) to stimulate the generation of PGE 2 in amnion cells. Moreover, some inflammatory cytokines, including IL-1␤, were found to enhance the expression level of COX-2 in amnion cells (5,6). Although some cytokines have been reported to regulate the expression of COX-2, further factors involved in COX-2 expression and PG synthesis should be considered.
Many reports have demonstrated the involvement of lipid factors in cellular responses. In particular, it is reasonable to assume that lipid ligands might be involved in the regulation of labor in combination with several kinds of protein factors. Sphingosine 1-phosphate (S1P) is an important lipid mediator that exerts a wide range of physiological activities (7)(8)(9)(10)(11)(12). For example, S1P has been reported to induce chemotactic migration and angiogenesis in human umbilical vein endothelial cells (8). S1P induces cellular chemotaxis and modulates cytokine release in mature human dendritic cells emerging of Th2 immune responses (11), and in human bronchial epithelial cells S1P has been reported to regulate interleukin-8 secretion (12). In terms of S1P cell surface receptors, a family of G proteincoupled receptors called the EDGs has been shown to contain specific S1P receptors (13)(14)(15)(16). This family includes EDG1/ S1P 1 , EDG3/S1P 3 , EDG5/S1P 2 , and EDG6/S1P 4 (13)(14)(15)(16). Although many previous reports have demonstrated the pivotal role of S1P in the modulation of several biological responses via its specific receptor, the role of S1P in amnion cells has not been studied previously.
In this study, we aimed to investigate whether amniotic fluid contains certain lipid ligands that modulate COX-2 expression and PGE 2 synthesis. We found that S1P exists in the amniotic fluid and acts as a potential regulator of labor response.

EXPERIMENTAL PROCEDURES
Reagents-The reverse transcription-polymerase chain reaction kit was purchased from Invitrogen, and fetal calf serum was purchased from Hyclone (Logan, UT). [ 3 H]S1P and [␥-32 P]dCTP were from PerkinElmer Life Sciences, enhanced chemiluminescence reagents and the PGE 2 enzyme immunoassay kit were from Amersham Biosciences, rabbit polyclonal COX-2 antibodies were from Cayman Chemical (Ann Arbor, MI), and phospho-ERK1/2, phospho-p38, and ERK2 antibodies were purchased from New England Biolabs (Beverly, MA). PD-98059 and SB-203580 were obtained from Biomol (Plymouth Meeting, PA) and were dissolved in dimethyl sulfoxide before being added to the cell culture. The final concentrations of dimethyl sulfoxide in culture were 0.1% or less.
Preparation of Lipid Extract of Amniotic Fluid-Amniotic fluid was prepared from healthy volunteers who had been pregnant for eight to nine months or during labor. Informed consent was obtained from all volunteers, and this study was approved by the local institutional review board at Dong-A University Hospital. Prepared amniotic fluid was extracted with chloroform as described previously (17). Briefly, 0.5 ml of amniotic fluid was extracted with ice-cold chloroform:methanol (1:2) and then with 2 ml of chloroform and 2 ml of 1 M KCl, 100 l of 7 N NH 4 OH. The acidic phase (the lower phase) and the alkaline phase (upper phase) were then separated, 3 ml of chloroform and 200 l of concentrated HCl were added to the alkaline phase, and the lower chloroform phase was retained. Each phase was evaporated under nitrogen and stored at Ϫ70°C until use. Just before assay, the dried samples were dissolved by sonication in phosphate-buffered saline (PBS) containing 0.4% BSA (w/v). PGE 2 Assay-PGE 2 levels were determined using an enzyme immunoassay kit, according to the manufacturer's instructions. Briefly, 50 l of standard or sample was pipetted into the wells of a 96-well plate. Aliquots of mouse polyclonal PGE 2 antibody and PGE 2 conjugated to alkaline phosphatase were then added to each well, and the plate was incubated at room temperature for 1 h. After incubation, the wells were washed six times with 200 l of PBS containing 0.05% Tween 20, and the 3,3Ј,5,5Ј-tetramethylbenzidine substrate was added. Wells were read at 670 nm with an enzyme-linked immunosorbent assay reader 30 min after adding substrate.
Quantitative Measurement of S1P-The amount of S1P in amniotic fluid was measured as described previously (18). In brief, acidified methanol containing C 17 -S1P, as an internal standard, was added to amniotic fluid. After partitioning the fluid with chloroform-sodium hydroxide solution, S1P was recovered from the aqueous phase. The release of monophosphate from S1P by alkaline phosphatase produced sphingosine, which was successively derivatized with o-phthalaldehyde and quantified using a high pressure liquid chromatography system equipped with a fluorescence detector.
Cell Culture-The WISH human amnion cell line was obtained from the American Type Culture Collection (Manassas, VA). WISH cells were cultured in RPMI 16040 medium supplemented with 2 mM Lglutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 10% fetal calf serum at 37°C in a humidified 5% CO 2 atmosphere. The cells were subcultured twice weekly by trypsinization and seeded in either 12 (2 ϫ 10 5 cells/well)-or 6-well plates (5 ϫ 10 5 cells/well). The cells were then stimulated with lipid extract or S1P for various lengths of time ranging from a few minutes to 12 h in the presence of S1P with or without inhibitors.
Ligand Binding Analysis-Ligand binding analysis was performed as described previously (19). Briefly, WISH cells were seeded at 1 ϫ 10 5 cells per well into a 24-well plate and cultured overnight. After blocking the cells with blocking buffer (33 mM HEPES, pH 7.5, 0.1% BSA in RPMI) for 2 h, several concentrations of 3 H-labeled S1P were added to the cells in binding buffer (PBS containing 0.1% BSA), in the absence or presence of unlabeled 10 M S1P, and incubation was continued for 3 h at 4°C with continuous shaking. The samples were then washed five times with ice-cold binding buffer, and 200 l of lysis buffer (20 mM Tris, pH 7.5, 1% Triton X-100) was added to each well. After 20 min at room temperature, the lysates were collected and counted using a ␥-ray counter (19).
Western Blot Analysis-WISH cells were plated in a 6-well plate and treated with the lipid extract of amniotic fluid or with S1P for different times. The cells were then washed with cold PBS, scraped off, and pelleted at 700 ϫ g at 4°C. The cell pellet obtained was resuspended in lysis buffer (50 mM Tris-HCl, pH 8.0, 5 mM EDTA, 150 mM NaCl, 0.5% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, and protease inhibitor mixture) and cleared by centrifugation, and the supernatant was saved as a whole-cell lysate. Proteins (30 g) were separated by 10% reducing SDS-PAGE and electroblotted in 20% methanol, 25 mM Tris, and 192 mM glycine onto a nitrocellulose membrane. The membrane was then blocked with 5% nonfat dry milk in Tris-buffered saline, Tween 20 (25 mM Tris-HCl, 150 mM NaCl, and 0.2% Tween 20), incubated with antibodies for 4 h, washed, and incubated for 1 h with secondary antibodies conjugated to horseradish peroxidase. Finally, the membrane was washed and developed using an enhanced chemiluminescence system.
RNA Isolation and Northern Blot Analysis-WISH cells were cultured for the indicated times at 37°C in 1 M S1P and then washed three times with PBS containing 2% bovine serum albumin. RNA was isolated using a Tri-Reagent kit (Molecular Research Center, Cincinnati, OH). Aliquots (2 g) of total RNA were denatured and fractionated by gel electrophoresis using a 1% agarose gel containing 2.2 M formaldehyde. The RNA was then transferred by capillary action in 20ϫ SSC (3 M NaCl, 0.3 M sodium citrate, pH 7.0) onto a nylon membrane. Blots were incubated with specific DNA probes for human COX-2, which had been labeled with [␣-32 P]dCTP by random priming using a Prime-␣-Gene kit (Promega, Madison, WI). Glyceraldehyde-3-phosphate dehydrogenase probe was used as an internal RNA loading control.
Transfection of Antisense Nucleotides of S1P Receptors-Antisense oligonucleotide corresponding to the region of translation initiation of S1P 1 , S1P 2 , or S1P 3 was used to inhibit the expression of S1P receptors, as described previously (20). Sense nucleotides for each receptor were used as controls. Finally, 20 M of antisense oligonucleotide for each S1P receptor were transfected using LipofectAMINE reagents (Invitrogen) according to the manufacturer's instructions. The cells were incubated for 48 h prior to each experiment. Statistics

Amniotic Fluid Lipid Extracts Stimulate PGE 2 Synthesis via COX-2 Expression in WISH Cells-Because
COX-2 has been reported to synthesize PGs in amnion cells, we investigated whether a lipid in the amniotic fluid modulates PGE 2 synthesis and COX-2 expression in amnion-derived cells. For this purpose, we extracted lipids from amniotic fluid, as described under "Experimental Procedures." The three dried types of lipid extracts were dissolved by sonication in a suspension buffer (0.4% BSA in PBS). WISH cells were incubated in serum-free medium in the absence or presence of the three lipid extracts of amniotic fluid for various lengths of time. We found that the addition of the chloroform extract with acidic solvent caused a significant increase in PGE 2 synthesis in WISH cells (Fig. 1A). When WISH cells were incubated with the lipid extract, PGE 2 synthesis was enhanced in a time-dependent manner, showing maximal activity 6 h after stimulation (Fig.  1A). PGE 2 synthesis was increased 2-fold as compared with the untreated cells at 6 h (Fig. 1A). We also performed Western blot analysis using antibodies recognizing COX-2. Stimulation of WISH cells with the lipid extract of the amniotic fluid induced 72-kDa COX-2 expression in a time-dependent manner (Fig.  1B), and this was maximally up-regulated 6 -12 h after stimulation (Fig. 1B). The results obtained suggest the existence of certain lipid components in the amniotic fluid that stimulate PGE 2 synthesis via COX-2 expression. A previous report demonstrated that activated charcoal adsorbs short lengths of lipid components such as S1P (21). When we stimulated WISH cells with charcoal-stripped amniotic fluid, as described previously (21), we did not observe any significant increase in the PGE 2 or COX-2 expression of the cells (Fig. 1). This result supports our notion that S1P may be involved in the amniotic fluid-induced expression of COX-2 in WISH cells. S1P Is Dissolved in Amniotic Fluid-To check whether amniotic fluid contains S1P, we performed quantitative analysis for S1P in amniotic fluid that was prepared from volunteers who had been pregnant for eight to nine months using an analytical method described previously (18). It was found to contain 19.7 Ϯ 2.38 nM (Table I). To investigate the role of S1P in the regulation of labor, we checked whether S1P levels during labor (37.5 Ϯ 4.27 nM) were found to be higher than the level prior to labor (19.7 Ϯ 2.38 nM) (  Fig. 2A),  . S1P-treated WISH cells were lysed, and the extracted protein was immunodetected with COX-2-specific antibody. The COX-2 protein levels shown are representative of three independent experiments (B). Western blotting analysis was also performed with antiactin antibody to confirm equal loadings (B). which was quantified after subtracting nonspecific binding. The addition of a 1000-fold molar excess of unlabeled S1P prior to the addition of [ 3 H]S1P reduced this binding. The specific binding of S1P in WISH cells was proven by adding 25 nM [ 3 H]S1P, which achieved saturation at ϳ200 nM [ 3 H]S1P for 5 ϫ 10 4 cells ( Fig. 2A). Moreover, the specific binding curve was found to be biphasic, showing an initial saturation at 50 nM per 5 ϫ 10 4 cells and next saturation at 200 nM per 5 ϫ 10 4 cells ( Fig. 2A). This indicates that WISH cells express at least two different receptors. To determine which isoforms of S1P receptor are expressed on WISH cells, we analyzed the mRNA expressions of different S1P by semi-quantitative RT-PCR. As shown in Fig. 2B, WISH cells expressed several forms of S1P receptors, namely S1P 1 , S1P 2 , and S1P 3 , and the expression levels of S1P 1 and S1P 3 were higher than that of S1P 2 (Fig. 2B). However, we were unable to detect S1P 4 expression (Fig. 2B). We confirmed that the RT-PCR product obtained without the addition of reverse transcriptase did not contain a DNA band in the gel (Fig. 2B). This result correlates with that of the ligand binding analysis, which showed the existence of multiple S1P receptors in WISH cells. S1P Stimulates PGE 2 Synthesis via COX-2 Expression in WISH Cells-To reveal the physiological role of S1P in amnion cells, we investigated the effect of S1P on PGE 2 synthesis. As shown in Fig. 3A, the stimulation of WISH cells with various concentrations of S1P enhanced PGE 2 release into the medium by 2-fold versus the unstimulated control, which showed maximal activity at a S1P concentration of 100 nM. Previous reports (3,4) have shown that COX-2 is responsible for the synthesis of PGE 2 when induced by extracellular stimuli. COX-2 is also known to be induced by extracellular stimuli (5, 6). The expression level of COX-2 protein was examined by Western blotting using anti-COX-2 antibodies. When WISH cells were stimulated with various concentrations of S1P, the 72-kDa COX-2 protein was increased in a concentration-dependent manner (Fig. 3B). This effect was maximal at an S1P concentration of 100 nM (Fig. 3B). Moreover, the concentration dependence of S1P-induced COX-2 expression was similar to that of S1Pinduced PGE 2 release, indicating that COX-2 is responsible for PGE 2 synthesis. The time dependence of S1P-induced COX-2 up-regulation was also tested. When WISH cells were stimulated with 100 nM S1P for various lengths of time, S1P enhanced the expression level of COX-2 within 3 to 12 h (Fig. 4A). Western blotting analysis was also performed using anti-actin antibody to confirm equal loadings (Fig. 4A). We then examined the effect of S1P on the transcriptional level of COX-2 by Northern blotting with a COX-2 probe. As shown in Fig. 4B, the stimulation of WISH cells with 100 nM S1P induced an accumulation of COX-2 mRNA transcripts 0.5-1 h after treating with 100 nM S1P (Fig. 4B).
ERK Is Involved in the S1P-induced Expression of COX-2 in WISH Cells-Mitogen-activated protein kinase (MAPK) has been reported to mediate extracellular signals to the nucleus in various cell types (22). In this study, we examined whether S1P stimulates MAPKs by using Western blot analysis with antiphospho-specific antibodies against each enzyme. When WISH cells were stimulated with 2 M S1P for different times, the phosphorylation level of ERK was transiently increased, showing maximal activity within 2-5 min of stimulation (Fig. 5A), and returned to baseline 10 min after stimulation (Fig. 5A). Another important MAPK, p38 kinase, was also transiently activated by S1P stimulation with kinetics that resembled those of ERK activation (Fig. 5A). We also examined the concentration dependence of S1P-induced ERK and p38 kinase activation. When WISH cells were stimulated with various concentrations of S1P, ERK and p38 kinase were activated in a concentration-dependent manner (Fig. 5B). In terms of ERK activation, S1P caused significant activation at 10 nM and maximal activation at 100 nM (Fig. 5B). However, p38 kinase activation was induced by ϳ0.1 nM, and S1P was maximal at 100 nM S1P (Fig. 5B). To determine the role of each MAPK on S1P-induced COX-2 expression, we used two different MAPK inhibitors. The preincubation of WISH cells with PD-98059, a specific MAPK/ERK kinase inhibitor, prior to S1P treatment completely inhibited COX-2 expression by S1P (Fig. 5C). Moreover, SB-203580, a specific p38 kinase inhibitor, did not affect S1P-induced COX-2 expression. These results indicate that MAPK/ERK kinase-dependent ERK activity but not p38 kinase is essentially involved in S1P-induced COX-2 expression. S1P-induced COX-2 Expression Is Mediated by a Partially Pertussis Toxin (PTX)-sensitive Pathway-We investigated the role of PTX-sensitive G protein on S1P-induced COX-2 expression. Cultured WISH cells were preincubated with 100 ng/ml of PTX prior to stimulation with 100 nM S1P, which up-regulated COX-2. We found that pretreatment with PTX partially inhibited COX-2 expression by S1P (Fig. 6A). This result shows that S1P induces COX-2 expression in a partially PTX-sensitive manner. Fig. 5C shows that S1P-stimulated COX-2 expression was completely inhibited by PD-98059, thus indicating the criticality of the role played by ERK in the process. We also checked the effect of PTX on S1P-induced ERK activation. When WISH cells were preincubated with 100 ng/ml of PTX prior to being stimulated with 2 M S1P, S1P-induced ERK activation was also found to be partially inhibited (Fig. 6B). This result supports our notion that S1P stimulates COX-2 expression via both PTX-sensitive and -insensitive pathways and that ERK is critically involved in the process. S1P 1 and S1P 3 but Not S1P 2 Are Involved in S1P-induced COX-2 Expression-In Fig. 2B, we showed that WISH cells express multiple isoforms of S1P receptor. To reveal the identity of the S1P receptor involved in S1P-induced COX-2 expression, we used S1P receptor antisense oligonucleotide, as described previously (20). Transfection with the antisense oligonucleotides to S1P 1 or S1P 3 significantly suppressed S1Pmediated COX-2 expression and PGE 2 production (Fig. 7, A  and B). As shown in Fig. 7A, in WISH cells transfected with S1P 1 -or S1P 3 -specific antisense oligonucleotide, S1P-induced PGE 2 production was reduced by ϳ51% (n ϭ 4) or 42% (n ϭ 4), respectively. In contrast, transfection with antisense oligonucleotide to S1P 2 did not affect the S1P-induced responses (Fig.  7, A and B). As control experiments, we confirmed that transfection with sense oligonucleotide to any S1P receptor had no effect on S1P-stimulated COX-2 expression and PGE 2 production (Fig. 7, A and B). These results suggest that S1P requires both S1P 1 and S1P 3 for full activation of the intracellular signal needed for COX-2 induction and PGE 2 production. The results also coincide with our data, namely that S1P-induced COX-2 expression is mediated by a partially PTX-sensitive pathway.

DISCUSSION
In the present study, we found that S1P, a lysophospholipid, is dissolved in human amniotic fluid and that it modulates the activity of human amnion-derived WISH cells. We then focused on the effect of S1P on PGE 2 synthesis and the expression of COX-2, which is critically involved in the regulation of labor.
The modulation of labor response remains an important is-  sue in the biology of reproduction. Many groups have reported that several kinds of cytokines are involved in the regulation of labor (3)(4)(5)(6)23). However, the role of lipid mediators on the modulation of labor response has not been studied. In this study, we demonstrate for the first time that human amniotic fluid contains lipid components involved in PGE 2 synthesis and COX-2 expression. We further characterized the lipid mediators and identified one of them as S1P. In a previous study (18), we developed a quantitative analytical method for S1P in biological samples, and then using this developed method, we found that human amniotic fluid contains significant amounts of S1P. The level of S1P in amniotic fluid prepared from volunteers who had been pregnant for eight to nine months was around 19 pmol/ml (7.1 pmol/mg of protein) (Table I). In addition, S1P levels in amniotic fluid increase by ϳ2-fold during labor. These data suggest that S1P play a role in labor. To determine the role of S1P in amniotic fluid, we examined the effect of purified S1P on amnion-derived WISH cells. The stimulation of human amnion-derived WISH cells with purified S1P induced COX-2 expression at the mRNA and protein level (Fig.  4). Because the lipid extract of amniotic fluid-induced PGE 2 secretion, mediated by COX-2 expression, we suggest that S1P dissolved in the amniotic fluid modulates COX-2 expression in amnion cells. The role of S1P on the modulation of COX-2 expression has already been reported (24). Davaille et al. (24) demonstrated that S1P stimulates COX-2 expression leading to anti-proliferative activity in human hepatic myofibroblasts. Our study indicates that S1P also regulates COX-2 expression in amnion-derived cells, suggesting that it has a role in the modulation of labor. Moreover, this finding strongly suggests not only protein cytokines, such as IL-1␤, but also lipid mediators like S1P might be involved in the regulation of labor.
Previous reports (13)(14)(15)(16) have demonstrated that S1P binds to several isoforms of the cell surface receptors, namely S1P 1 , S1P 2 , S1P 3 , and S1P 4 . When we performed RT-PCR to check the expression pattern of S1P receptors on WISH cells, we found that WISH cells express S1P 1 , S1P 2 , and S1P 3 (Fig. 2B). According to previous reports (25)(26)(27), S1P 1 exclusive couples to G proteins of the G i /G o family, whereas S1P 2 and S1P 3 are coupled to G i /G o and also to G q , G s , and G 13 . In our study, we investigated the effect of PTX (which specifically inactivates G i /G o -mediated signaling pathways) on S1P-induced signaling. When WISH cells were pretreated with 100 ng/ml PTX for 24 h prior to S1P stimulation, S1P-induced COX-2 expression was partially inhibited (Fig. 6A). S1P-stimulated ERK activation was also partially inhibited by PTX pretreatment, as shown in Fig. 6B. These results suggest that S1P modulates ERK activation leading to COX-2 expression and that not only PTXsensitive G protein-coupled receptors but also PTX-insensitive G protein-coupled receptors are involved in the process in WISH cells. We also determined the identities of the S1P receptors that are involved in the regulation of COX-2 expression and PGE 2 synthesis in WISH cells by S1P receptor antisense oligonucleotide transfection. As shown in Fig. 7, S1P-induced COX-2 expression and PGE 2 synthesis were inhibited by the down-regulation of S1P 1 and S1P 3 but not S1P 2 , which suggests that S1P 1 and S1P 3 , but not S1P 2 , are involved in S1Pinduced COX-2 expression in WISH cells. The results correlate with the finding that S1P-induced COX-2 expression is mediated by a partially PTX-sensitive signaling pathway.
To investigate the signal pathway of COX-2 expression by S1P in WISH cells, we checked the role of two different kinds of MAPKs. We observed that S1P stimulates both ERK and p38 kinase activity (Fig. 5, A and B). To determine the role of ERK or p38 kinase on S1P-induced COX-2 expression, we pretreated WISH cells with two different MAPK-specific inhibitors, PD-98059 and SB-203580 (specific ERK and p38 kinase inhibitors, respectively). Preincubation of WISH cells with PD-98059 prior to S1P stimulation completely inhibited COX-2 expression by S1P (Fig. 5C); however, SB-203580 did not affect S1P-induced COX-2 expression (Fig. 5C). This indicates that ERK, but not p38 kinase, plays a key role in S1P-induced COX-2 expression in amnion cells. Previously, we characterized the signaling pathways involved in IL-1␤-induced COX-2 expression in WISH cells and showed that p38 kinase, but not ERK, is involved in the expression of COX-2 (23). We also demonstrated that the addition of 1-butanol completely blocked IL-1␤-induced COX-2 expression in WISH cells and suggested the critical role played by phospholipase D in the process (23). In the present study, when we stimulated S1P in the presence of 1-butanol, COX-2 expression was unaffected (data not shown). These results suggest that the intracellular signaling pathways involved in the modulation of COX-2 by S1P differ from those involved in the modulation of COX-2 by IL-1␤ in WISH cells.
In conclusion, the present study has two important findings, first, that the lipid mediator S1P is present in human amniotic fluid, and second, that this acts on the S1P-specific receptors S1P 1 , S1P 2 , and S1P 3 expressed on human amnion-derived WISH cells. Because this study is the only report on the existence of lipid mediators in human amniotic fluid and on the potential role of S1P in labor response, further studies on the pathophysiological and physiological roles of S1P in labor response are required.