| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
research-article |
CIHR Group in Fetal Development and Health, Departments of Physiology, Obstetrics and Gynecology, and Medicine, University of Toronto, Toronto, Ontario, Canada M5S 1A8
ABSTRACT
The abnormal degradation of the extracellular matrix by matrix metalloproteinases (MMPs) in the fetal membranes has been proposed as a central event in preterm premature rupture of the membranes (pPROM). Prostaglandins (PGs) are thought to increase the risk of preterm premature rupture of the fetal membranes by causing matrix degradation. The aim of this study was to assess the mediating role of PGs on lipopolysaccharide (LPS)-induced MMP9 secretion in vitro. ELISA, zymography, and Western blotting were performed on cells and medium from cultures of purified chorion trophoblasts (CTs) and syncytiotrophoblasts (STs) from the human placenta and fetal membranes treated with LPS, meloxicam, (a selective prostaglandin-endoperoxide synthase 2 [PTGS2, previously known as cyclooxygenase 2] inhibitor), or replacement PGE2 or PGF2alpha. LPS significantly (P < 0.01) increased proMMP9 secretion and prostaglandin E2 (PGE2) output by cultured CTs and STs, but there was no effect on tissue inhibitor of matrix metalloproteinase 1 (TIMP1) secretion. In these cells, meloxicam significantly blocked LPS-induced proMMP9 secretion and PGE2 output (P < 0.01). Exogenous PGE2 and PGF2alpha significantly reversed the reduction in proMMP9 secretion caused by meloxicam in a dose-dependent manner (P < 0.01). The expression of PTGS2 protein in CTs and STs was increased dramatically after LPS treatment, but there was no significant effect on the expression of PTGS1 (previously known as cyclooxygenase 1), membrane-associated prostaglandin E synthases (membrane-associated PTGES, previously known as mPGES) 1 and 2, or cytosolic prostaglandin E synthase (cytosolic PTGES, previously knows as cPGES) proteins. Our results suggest that PGs may mediate the selective increase in MMP9 after exposure of trophoblast cells to LPS. There was no effect of LPS on TIMP1. Understanding this relationship may help in developing strategies for the prevention and management of pPROM and preterm labor.
human fetal membrane, lipopolysaccharide, matrix metalloproteinase, mechanisms of hormone action, prostaglandin, syncytiotrophoblast, trophoblast
Prevention and treatment of preterm premature rupture of the membranes (pPROM) and infection-associated preterm birth remain considerable challenges in obstetric practice. Infection, especially intrauterine infection (with a combination of Gram-negative and Gram-positive bacteria in the majority of preterm labor [PTL] and pPROM cases), is one of the important mechanisms of pPROM and PTL. It has been estimated that at least 40% of all preterm births occur to mothers with intrauterine infection, largely subclinical in nature. Moreover, the earlier the gestational age at delivery, the higher the frequency of intrauterine infection [1]. Although broad-spectrum antibiotic therapy increased the likelihood of prolongation of the latency period of pPROM and improved neonatal health by reducing the risk of respiratory distress syndrome, early sepsis, severe intraventricular hemorrhage, and severe necrotizing enterocolitis, there is little benefit in terms of the reduction in rate of preterm birth from only antimicrobial administration [2, 3]. Presently, the underlying mechanisms explaining the development of rupture of the membranes in PTL and pPROM after infection are still poorly understood, although microorganism-derived proteases have a direct effect on the fetal membranes, and bacteria-induced proteolysis of the fetal membrane matrix can be inhibited by the addition of antibiotics in vitro [4, 5].
Several studies have focused on fetal membrane remodeling, the role of matrix metalloproteinases (MMPs) and their inhibitors, tissue inhibitors of metalloproteinase (TIMPs), in the rupture of membranes with PTL. Extensive changes in collagen metabolism and increased expression of MMPs with decreased expression of TIMPs are associated with pPROM [6–12]. Among these enzymes, MMP9 is the one of the best studied proteases involved in pPROM. In human fetal membranes, MMP9 expression increases during infection, pPROM, and active term labor. Abnormal degradation of the extracellular matrix by MMPs in fetal membranes has been proposed as the central event in pPROM. Therefore, to reduce extracellular matrix degradation and the risk of PROM, downregulation of MMP expression and production or neutralization of MMP activities by endogenous inhibitors have been proposed as a potential therapeutical targets. Previous studies have shown that production of MMPs is regulated by endocrine and paracrine effectors [13–18], but the interrelationship between bacterial products (lipopolysaccharide [LPS]) and prostaglandins (PGs) is relatively unexplored.
PGs are known to be involved in the cascade of events leading to cervical softening, myometrial contractions, and labor at term and preterm [19, 20]. The concentrations of PGs increase in the amniotic fluid of women with preterm labor and intra-amniotic infection. Bacterial products, including LPS, increase PG production by the fetal membranes through changing PG synthesis and metabolism [19, 21–23]. Previous studies showed that LPS augments PG production through releasing the PG precursor arachidonic acid from membrane phospholipids, increasing PG synthetic capacities and decreasing PG metabolic capacities in human fetal membranes [24–27]. However, the precise regulation of PG synthesis cascade by different synthetases in relation to bacterial infection is not well understood. Although prostaglandin E2 (PGE2) and PGF2
are required for MMP production, and LPS can increase MMP production in human fetal membranes, a direct link between LPS, PG production, and MMP production in the fetal membranes has not been established [9, 15, 16].
Intra-amniotic infections are frequently associated with Gram-negative organisms. LPS, a component of Gram-negative bacteria, effects stimulation of PGs and cytokines from host cells [17, 21], thereby causing preterm birth [28, 29]. LPS is detectable in the amniotic fluid of patients with preterm labor, premature rupture of membranes, or clinical suspicion of chorioamnionitis. In the present study we challenged cultured trophoblast cells from human placenta and chorion with LPS in vitro in order to determine the changes of MMP9 and TIMP1 and the role of PGE2, PGF2, and enzymes on their synthetic pathways in the regulation of MMP9 production.
Human Placental and Fetal Membrane Collection and Cell Cultures
Placentas with attached fetal membranes were collected from normal-term (>37-wk gestation) pregnancies after elective cesarean delivery (nonlabor; n = 16). Patient consent and ethical approval were obtained before tissue collection in accordance with the Canadian Tri-Council guidelines and the regulations of Mount Sinai Hospital (Toronto, ON, Canada) and the University of Toronto. None of the patients had received PGs, corticosteroids, or oxytocin. Syncytiotrophoblast (ST) and chorionic trophoblast cells were prepared using the methods described previously [30].
Cells were plated in 24-well plates (0.5 x 106 to 1 x 106 cells/well) for zymography and ELISA analysis or in 60-mm diameter dishes (6 x 106 to 10 x 106 cells) for Western blotting, and were cultured in Dulbecco modified Eagle medium (DMEM)/F-12 (Gibco Invitrogen Corp., Grand Island, NY) supplemented with 10% fetal calf serum (Invitrogen Life Technologies Inc., Gaithersburg, MD) and antibiotics (1000 U/ml penicillin, 0.1 mg/ml streptomycin, and 0.23 µg/ml amphotericin; Sigma-Aldrich Corp., St. Louis, MO) at 37°C under 5% CO2/95% O2 for 72 h. The serum-free DMEM/F-12 then was used to replace the culture medium. After 12 h of starvation and/or preincubation with treatment, the further treatments were added and incubated for 18 h or other time periods as described.
LPS (Escherichia coli O55:B5), the prostaglandin-endoperoxide synthase 2 (PTGS2) inhibitor meloxicam, PGE2, and PGF2 were purchased from Sigma (St. Louis, MO). Substances were used at final concentrations ranging from 10–12 to 10–6 M. Vehicle-treated wells (controls) were present in each experiment. After incubation the medium was harvested and stored at –20°C until MMP9 and TIMP1 assays were performed. Cells were used for protein extraction and determination by the Bradford assay.
MMP9 and TIMP1 levels in the media were determined with an ELISA system kit (Amersham Biosciences, Little Chalfont, UK). An aliquot (100 µl) of the standard mixture or the sample was transferred to each microplate well previously coated with the monoclonal antibody, and the plate was allowed to stand for 1 h at room temperature without shaking. It then was aspirated and washed four times with wash buffer (0.01 M phosphate buffer, pH 7.0, containing 0.05% Tween 20). Next, a 100-µl aliquot of biotinylated antibody was added to each well and incubated for 1 h at room temperature without shaking. After washing, a 100-µl aliquot of streptavidin-horseradish peroxide conjugate reagent was added to all of the wells and allowed to stand for 2 h at room temperature. Subsequently, 100 µl substrate equilibrated at room temperature was dispensed into all wells and shaken on a microplate plate shaker for exactly 20 min at room temperature. The reaction was stopped with 100 µl of 1 M H2SO4, and the absorbance at 450 nm was measured with a microplate reader.
In order to detect proteolytic activity in conditioned media, about 10–20 µl (10 µg cell protein) harvested culture medium was electrophoresed under nonreducing conditions in a 10% acrylamide gel containing 1 mg/ml gelatin (Sigma-Aldrich Corp.), according to the method described by Fisher and Werb [31]. After electrophoresis, the gels were washed at room temperature for 1 h in 2.5% Triton X-100 and 50 mM Tris-HCl (pH 7.5), then incubated at 37°C overnight in buffer containing 150 mM NaCl, 5 mM CaCl2, and 50 mM Tris-HCl (pH 7.6). Thereafter, gels were stained with 0.1% (wt/vol) Coomassie Brilliant Blue R-250 (INC Biomedicals Inc., Aurora, Ohio) in 30% (vol/vol) isopropyl alcohol/10% glacial acetic acid for 60 min and destained in 10% (vol/vol) methanol/5% (vol/vol) glacial acetic acid. Quantification of the bands corresponding to 92-kDa gelatinase was performed by densitometry using Scion Image software (Scion Corp., Frederick, MD).
Radioimmunoassay for PGE2 Concentration
Concentrations of PGE2 were measured in culture medium collected after different times of treatment. PGE2 concentrations were determined using a specific radioimmunoassay described previously [32]. [3H]-PGE2 (Amersham International, Aylesbury, UK) and PGE2 polyclonal antibody (at a final dilution of 1:4000, raised in rabbits; gift from Dr. Tom Kennedy, University of Western Ontario, London, ON, Canada) were used. The mean intraassay and interassay coefficients of variation were 6.7% for PGE2 radioimmunoassay. Assay sensitivity routinely was approximately 15 pg/tube.
About 5 µg (for membrane-associated prostaglandin E synthases [PTGES] 1 and 2 or cytosolic PTGES) or 20 µg (for PTGS1 and PTGS2 and ß-actin) total protein per lane was incubated in SDS-PAGE sample buffer, subjected to SDS-PAGE analysis with 10%–12% acrylamide gel, and electrotransferred onto a nitrocellulose membrane. The membrane was blocked in 5% skim milk powder in 0.1% Tris-buffered saline/Tween 20 overnight. The membrane then was incubated with one of the following antibodies: mouse monoclonal anti-human PTGS2, rabbit polyclonal anti-ovine PTGS1, rabbit polyclonal anti-human cytosolic PTGES, rabbit polyclonal anti-human membrane-associated PTGES 1, or rabbit polyclonal anti-human membrane-associated PTGES 2 (Cayman Chemical Co./Cedarlane Laboratories Ltd., Hornby, ON, Canada) and mouse monoclonal anti-ß-actin (Sigma-Aldrich Canada Ltd., Oakville, ON, Canada). The membrane was then incubated with the appropriate secondary antibodies of either peroxidase-conjugated sheep anti-mouse IgG or peroxidase-conjugated donkey anti-rabbit IgG (Amersham Biosciences, Arlington Heights, IL). Immunoreactive proteins were visualized using the enhanced chemiluminescence Western blotting detection system (PerkinElmer).
Statistical analysis was carried out by ANOVA and Tukey test. Results are expressed as the mean ± SEM for the number of different experiments (placentas studied). Control cultures were conducted in the absence or presence of dimethyl sulfoxide and/or alcohol (0.01%). Statistical significance was set at P < 0.05. Calculations were performed using SigmaStat (Jandel Scientific software, San Rafael, CA).
Figure 1 illustrates the effect of LPS on the secretion of proMMP9 and TIMP1 from human chorion trophoblasts (CTs) and placental STs by ELISA and zymography analysis. There was a significant dose-response effect of LPS on the MMP9 secretion by ELISA (P < 0.05 or P < 0.01; Fig. 1, C and D). In contrast, TIMP1 levels were not significantly changed in the medium of either of these cell types (Fig. 1, A and B).
|
Figure 2 shows the measurement of proMMP9 secretion from CTs and STs by zymography. Changes in proMMP9 are similar to the results of ELISA analysis in those cells (Fig. 2, A and B).
|
Figure 3 shows the change in endogenous PGE2 output after LPS stimulation and the inhibition by meloxicam of LPS-induced PGE2 production in CTs and STs. The output of PGE2 was significantly elevated compared with control at each time point (P < 0.01; Fig. 2, A [ST] and B [CT]). Meloxicam significantly blocked LPS-induced PGE2 production at the 8-h time point in a dose-dependent manner (P < 0.01; Fig. 2, C [ST] and D [CT]).
|
The results presented in Figure 4 show the effect of LPS in the absence or presence of meloxicam, a selective inhibitor of PTGS2 activity, on proMMP9 secretion from CTs and STs determined by gelatin zymography analysis. Co-incubation with meloxicam diminished the effect of LPS (P < 0.05 or P < 0.01). The effect was concentration dependent, with the higher concentrations of meloxicam exerting the greatest degree of inhibition.
|
The results presented above suggest a link between LPS, PGs, and proMMP9 secretion in cultured human fetal membrane and placental cells. Therefore, the effect of PGs on proMMP9 secretion was determined in CT and ST cells treated with LPS + meloxicam. The cells were preincubated with meloxicam (1 x 10–6 M) for 12 h and LPS (200 ng/ml) with or without PGE2/PGF2 (1 x 10–12 M to 1 x 10–7 M) was added for a further 18 h. PGE2 reversed the blockade of meloxicam on LPS-induced proMMP9 in a concentration-response pattern (P < 0.05 or P <0.01; Fig. 5, A [CT] and C [ST]). Higher concentrations of PGE2 (1 x 10–9 M and 1 x 10–7 M) induced a significant increase of proMMP9 secretion, reaching values comparable with those seen following LPS stimulation. The effect of PGF2 was similar to that of PGE2 on these same cells (Fig. 5, B [CT] and D [ST]).
|
The observations that meloxicam-reduced proMMP9 secretion was restored by PGE2 in LPS-treated CTs and STs and that endogenous PGE2 production was induced by LPS in CTs and STs indicate that endogenous PG synthesis may be involved in the regulation of MMP9 secretion stimulated by LPS. The enzymatic steps involved were examined by determining the protein expression of PG synthesis enzymes, including PTGS1, PTGS2, membrane-associated PTGES1 and PTGES2, and cytosolic PTGES. As shown in Figure 6, PTGS2 protein expression was rarely detectable in the absence of LPS, but after treatment there was significantly increased expression of PTGS2 protein in CTs and STs in a time- and dose-response manner. Unlike PTGS2, there were no significant changes in protein levels of PTGS1, membrane-associated PTGES1 and PTGES2, and cytosolic PTGES with LPS stimulation.
|
Bacterial infection and inflammation are believed to trigger pPROM and provoke cervical changes that contribute to preterm birth. Bacterial infection may predispose women to rupture of the fetal membranes through several possible mechanisms, each of which induces degradation of the extracellular matrix. It has been documented that bacteria-derived proteases directly weakened the strength and elasticity of fetal membranes. Bacterial infection via bacterial products (for instance, LPS) activates the host inflammatory response through release of proinflammatory cytokines (like TNF and IL-1ß) to induce MMP production in vitro, thereby degrading extracellular matrix [4, 14, 17]. High concentrations of PGs were detected in the amniotic fluid of women with preterm labor [21]. In animal models of intra-amniotic infection, higher expression of PTGS2 gene and a higher percentage of preterm birth was found after intraperitoneal LPS injection. Cyclooxygenase inhibitors reduced the percentage of LPS-induced preterm birth in pregnant mice [28, 29]. These observations all suggest that PGs play a critical role in infection-induced preterm birth. Although it is well documented that PGs upregulate MMPs in the fetal membranes and that expression of MMPs increased with PROM and PTL [15, 16, 21], it remained to be investigated whether PGs were involved in infection-induced MMP production in the fetal membranes.
We found that LPS increased the secretion of proMMP9 from human CT cells and STs, consistent with the effect of LPS on human fetal membranes in a tissue culture model [9]. Moreover, our data showed that LPS-stimulated PGE2 production and PTGS2 expression paralleled the changes in the secretion of proMMP9. The increased levels of proMMP9 and PGE2 were reduced significantly following inhibition of PG synthesis with the selective PTGS2 inhibitor meloxicam. In addition, exogenous PGE2 reversed the blockade of meloxicam on LPS-induced proMMP9 secretion. Exogenous PGF2 restored proMMP9 secretion after meloxicam in a manner similar to the effect of PGE2 on those same cells. Thus, our results suggest that endogenous PGs are involved in LPS-induced MMP9 secretion in human fetal membranes and placenta. This provides evidence to support the proportion of PG-mediated LPS induction of MMP9 production in an autocrine manner in chorion and placental cells. In contrast, we did not find significant changes in TIMP1 secretion with LPS treatment. TIMPs inhibit the activity of secreted MMPs by the formation of a 1:1 complex with MMP. This raises the possibility of a shift in the balance between proteolysis by MMPs and their inhibition by TIMP after LPS treatment to favor tissue degradation.
It has been documented that all MMPs (except membrane-type MMPs) are released as inactive precursor forms (proMMPs). Activation of these enzymes is therefore an important regulatory step in MMP activity. The most commonly recognized mechanism of MMP activation involves proteolytic removal of a propeptide domain of proMMPs. This can be achieved by a range of endogenous proteinases, such as trypsin, plasmin, and other active MMP family members, forming MMP-activating cascades in vivo [33]. At sites of infection, various types of proteinases are released by pathogenic bacteria. Recent studies demonstrated that bacterial proteinases, such as Pseudomonas aeruginosa elastase, Vibrio cholerae proteinase, and Streptococcus pyogenes, strongly activated proMMP9 and other proMMPs via limited proteolysis to generate active forms of MMPs [34, 35]. Moreover, the inflammatory oxidants such as HOCl and ONOO– also activate proMMPs [36, 37]. Importantly, TIMP1 can be readily inactivated by ONOO– [38]. These findings indicate that proMMP9 activation is complex. Although proMMP9 activation has not been shown under the in vitro conditions of our present study, it is reasonable to speculate that in vivo bacteria and bacteria-induced inflammation may convert proMMP9 into active MMP9 at sites of infection. Thus, we propose that bacteria, at least in part through their product LPS, induce PG production, and their proteinases may increase matrix degradation by upregulating MMP production and converting proMMPs into active MMPs. Eventually, these processes trigger premature rupture of the membranes, cervical ripening, and preterm birth. Furthermore, our results raise the possibility of exploiting direct inhibition of MMP activity combined with PTGS2 inhibitor and antibiotics as a therapeutic target in the prevention and treatment of pPROM. The contributions of other MMPs (including MMPs 1 and 3) potentially induced by LPS stimulation were not evaluated in the present studies.
It is well established that the level of PGs within the placenta and fetal membranes depends on the balance between synthesis and degradation. The biosynthesis of PGs involves a multienzyme pathway by three sequential enzymatic steps, namely, phospholipase A2 enzymes, cyclooxygenase enzymes (PTGS1 and PTGS2), and various terminal prostanoid synthases (membrane-associated PTGES 1 and 2 and cytosolic PTGES and PGF synthase). Previous research reported that LPS increases the activity of phospholipase A2 type II and reduces the activity of 15-hydroxyprostaglandin dehydrogenase in human fetal membranes, thereby resulting in an increase of PG production [24, 25]. In this study we found that the expression of PTGS2 protein in CTs and STs was dramatically increased with LPS treatment in a time- and dose-dependent manner, but it was rarely detectable in control (untreated) cells. PTGS1 expression was not significantly affected by LPS treatment. Previous studies showed that human fetal membranes and placenta expressed membrane-associated PTGES and cytosolic PTGES protein [39, 40]. However, little is known of the effects of LPS on these enzyme activities. In the present study we further demonstrated that membrane-associated PTGES1 and PTGES2 proteins were expressed in cultured CTs and STs, but we did not find changes in these enzymes or in cytosolic PTGES with LPS treatment. Our results suggest that the PTGS2 step is responsible principally for the LPS-induced PG output and for the changes in MMP9 output that we have observed.
In conclusion, our results, together with previous data [24, 25], might allow us to suggest that during infection LPS stimulates production of PGE/PGF within the fetal membranes and placenta by upregulation of PTGS2 expression. PG induces MMP9 secretion from fetal membranes and placenta and increases the ratio of MMP9 to TIMP1. These changes contribute to degradation of extracellular matrix that predisposes the fetal membranes to rupture prematurely.
Correspondence: 1Wei Li, 1 King's College Circle, Medical Sciences Building, Room 3344, Department of Physiology, University of Toronto, Toronto, ON, Canada M5S 1A8. FAX: 416 978 4940; e-mail: weisun.li{at}utoronto.ca
Received: 3 September 2006.
First decision: 16 October 2006.
Accepted: 12 December 2006.
REFERENCES
in intact term fetal membranes. Placenta 1998; 19:625–630[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  C. Rampon, S. Bouillot, A. Climescu-Haulica, M.-H. Prandini, F. Cand, Y. Vandenbrouck, and P. Huber Protocadherin 12 deficiency alters morphogenesis and transcriptional profile of the placenta Physiol Genomics, July 1, 2008; 34(2): 193 - 204. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-H. Yen, T. Khayrullina, and D. Ganea PGE2-induced metalloproteinase-9 is essential for dendritic cell migration Blood, January 1, 2008; 111(1): 260 - 270. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
