| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
research-article |
Equipe Endotoxines3 and Equipe Signalisation et Régulations Cellulaires,4 Institut de Biochimie et Biophysique Moléculaire et Cellulaire, Centre National de la Recherche Scientifique, UMR-8619, Université Paris-Sud, 91400 Orsay, France
ABSTRACT
Surfactant protein A (SFTPA1), a member of the collagenous lectin (collectin) family, was first described as a major constituent of lung surfactant, but has recently also been found in the female genital tract. Various microorganisms colonize this area and may cause intrauterine infection or trigger preterm labor. We found that SFTPA1 was not produced in the uterus. Instead, it was immunodetected transiently in rat myometrium at the end (Days 19 and 21) of gestation, but not postpartum, and in cultured myometrial cells. Fluorescence microscopy showed that Texas Red-labeled SFTPA1 bound to myometrial cells. This result was confirmed by biochemical approaches. [125I]-SFTPA1 bound to two myometrial cell proteins (55 and 210 kDa). This interaction was dependent on the integrity of the collagenlike domain of SFTPA1. SFTPA1 rapidly activated mitogen-activated protein kinase 1/3 (MAPK1/3) in myometrial cells. Bacterial lipopolysaccharide (LPS), an agent known to trigger uterine contractions and preterm birth, also activated MAPK1/3. The prolonged treatment of myometrial cells with LPS or SFTPA1 upregulated PTGS2 (COX2) protein levels. The addition of rough-type LPS to SFTPA1 blocked the interaction of SFTPA1 with its binding sites and the activation of MAPK1/3 and PTGS2 by SFTPA1. Our data provide the first demonstration of a direct effect of SFTPA1 on rat myometrial cells and inhibitory cross talk between SFTPA1 and LPS signals, providing new insight into the mechanisms of normal and preterm parturition.
female reproductive tract, immunology, parturition, signal transduction, uterus
Parturition is a sequence of integrated physiologic events: fetal membrane rupture, cervical dilatation, myometrial contractility, placental separation, and uterine involution [1, 2]. Gestation time is very precise in humans, but one in eight babies is born prematurely, with an associated high risk of disability. Recent investigations have focused on the sequence of molecular events occurring just before and during normal labor, with the hope of elucidating the mechanism underlying preterm births [3, 4]. Infection is one of the major environmental factors leading to premature birth, with at least 40% of preterm births thought to be attributable to intrauterine infection [5]. This major effect of infection results from the effects of bacterial molecules involved in inflammatory pathways. Bacterial lipopolysaccharide (LPS), a component of the cell wall of Gram-negative bacteria capable of inducing septic shock, has been implicated in the pathophysiology of preterm labor [6, 7]. Intrauterine LPS infusion in pregnant rats also has been shown to induce preterm delivery [8]. LPS is a potent proinflammatory molecule that binds to specific LPS receptors and initiates signal transduction, leading to activation of the mitogen-activated protein kinase (MAPK) and nuclear factor-kappa B (NF-kappa B) pathways [9]. The LPS receptor (Toll-like receptor 4) has been detected on cultured myometrial cells, which respond to LPS signaling by the activation of PKCzeta and NF-kappa B activation [10]. Kinase cascades also play a prominent role, culminating in the activation of MAPKs, including the p42 and p44 kinases (also called extracellular signal-regulated kinases 1 and 2; MAPK1/3), MAPK14, and MAPK8/9 (formerly known as p38 and JUNK, respectively). These three families of kinases may be activated independently or simultaneously, in response to various extracellular stimuli, such as physical stress, inflammatory cytokines, growth factors, and bacterial components. Several lines of evidence indicate that MAPKs regulate uterine contractions [11–13] and myometrial cell proliferation [14]. Furthermore, MAPK1/3 activation has been shown recently to be associated with uterine contractility and preterm labor in rats [15], via the production of prostaglandins [16]. Prostaglandins are synthesized from arachidonic acid by PTGS1 (also known as COX1), a weakly constitutively expressed cyclooxygenase, and the inducible PTGS2 enzyme (previously known as COX2). PTGS2 is absent from most normal tissues but is upregulated in response to proliferative and inflammatory stimuli. The release of endogenous prostaglandins is involved in the LPS-mediated increase in myometrial contractility in pregnant rats [17]. Furthermore, activated MAPK1/3 increases PTGS2 expression in human uterine myocytes, leading to the production of prostaglandins [18], which may induce uterine contraction [16].
Endogenous factors also may be involved in parturition, as seems to be the case for a protein secreted by the fetal lung, surfactant protein A (SFTPA1, also termed SP-A). Indeed, SFTPA1 secretion by the maturing fetal lung near term may provide the stimulus for amniotic fluid macrophage activation and migration to the maternal uterus, triggering an NF-kappa B signaling cascade within the uterus, leading to labor [19]. SFTPA1 may thus have an indirect effect on the uterus. SFTPA1 is a lung "collectin" associated with pulmonary surfactant phospholipids. Like other members of the collectin family, SFTPA1 has a globular C-terminal domain with calcium-dependent lectin activity, and an N-terminal collagenlike domain [20]. The globular domain of SFTPA1 recognizes carbohydrates and lipids on the surface of bacteria, viruses, and fungi, resulting in a role for SFTPA1 in the clearance and killing of several pathogens [21].
LPS, which also is a strong inducer of labor in pathologic conditions, is also one of the ligands of SFTPA1. SFTPA1 binds to the lipid A portion of LPS devoid of the O-antigenic chain (rough-type LPS) [22], and modulates LPS responses in vitro and in vivo [23, 24]. SFTPA1 also specifically recognizes receptors present on the membranes of leukocytes and alveolar epithelial type II cells and modulates several functions of these cells, such as phagocytosis, cytokine release, the production of reactive oxygen and nitrogen species, chemotaxis, and surfactant secretion.
SFTPA1 is synthesized mostly in alveolar epithelial type II cells and Clara cells, but this protein also is present at other sites, and Sftpa1 mRNA can be detected in a significant number of nonpulmonary tissues [25]. Akiyama et al. reported the immunohistologic detection of SFTPA1 in uterine duct lumina, but not in the uterus of Sftpa1-null mice [26]. SFTPA1 and SFTPD also have been detected in the genital tract of mares [27]. The location of SFTPA1 in the uterus may therefore be related to the role of this protein in parturition, which may involve its potency as a macrophage activator, as recently demonstrated by Condon et al. [19]. We therefore investigated the putative direct effects of SFTPA1 on the uterine smooth muscle: the myometrium.
Using biochemical, immunologic, and functional analyses, we detected SFTPA1 binding sites in myometrial cells and tissues and demonstrated that SFTPA1 binding led to the activation of MAPK1/3 and the induction of PTGS2 expression. These data provide the first evidence of a direct effect of SFTPA1 on uterine tissues. Cross talk between LPS and SFTPA1 also has been found in the control of uterine signaling.
Rough- and smooth-type LPS from Salmonnella minnesota (Re-LPS, S-LPS), bovine serum albumin (BSA), 1,3,4,6-tetrachloro-3
,6ß-diphenylglycouril (iodogen), 1,2-dipalmitoyl-sn-glycerol-3-phosphocholine (DPPC), collagenase type VII, and Saccharomyces cerevisiae mannan were obtained from Sigma Chemical Co. (St. Louis, MO). Texas Red succinimidyl ester was obtained from Molecular Probes (Eugene, OR). Na[125I] (0.78 MBq/µl) was purchased from ICN Biomedical Inc. (Irvine, CA). Polyclonal rabbit anti-SFTPA1 antibody was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Horseradish peroxidase (HRP)-conjugated anti-rabbit IgG was obtained from Dako (Trappes, France). Insta-Pure reagent was obtained from Eurogentec (Angers, France). The HRP-conjugated anti-mouse IgG and Fluoro Guard antifade reagent were obtained from Bio-Rad (Marne la Coquette, France). Moloney murine leukemia virus reverse transcriptase, dinucleotide triphosphates (dNTPs), random hexamer primers, Taq polymerase, the anti-active MAPK1/3 antibody, and the MEK inhibitor 1,4-diamino-2,3-dicyano-1,4-bis(2-aminophenylthio) butadiene (U0126), were obtained from Promega (Madison, WI). Sftpa1 primers were purchased from MWG-Biotech (Roissy, France). Anti-PTGS2 antibody was obtained from Cayman Chemical (Montigny le Bretonneux, France). The cell culture medium (CM) used was RPMI 1640 (GIBCO, Grand Island, NY) containing 2 mM L-glutamine, 100 IU/ml penicillin, and 100 µg/ml streptomycin, supplemented with 10% heat-inactivated (56°C, 30 min) fetal calf serum (FCS) from BioMedia (Boussens, France). All other cell culture reagents were obtained from Invitrogen (Cergy Pontoise, France).
For each experiment, ten to twelve 28-day-old prepubertal Wistar female rats (Janvier, France) were treated by intraperitoneal injection of 30 µg estradiol or vehicle alone (0.2 ml sunflower oil plus 3% ethanol) for 2 days. They were killed at the age of 30 days by carbon dioxide asphyxiation. Pregnant rats (2 to 3 rats) were used at different stages of gestation (Day 12: mid-gestation; Days 19–21: end of gestation; pp: 1 day postpartum). All treatments were performed in accordance with the principles and procedures outlined in the European guidelines for the care and use of experimental animals. The uterus was removed, and the myometrium was separated from the endometrium by stripping, as previously described [14, 28].
Primary cultures of myometrial cells were prepared from 10 to 12 nonpregnant rats by collagenase type II digestion, as previously described [29]. The endometrium was first removed from the uterus by stripping. Histologic sections showed that the samples consisted exclusively of smooth muscle fibers. This myometrial preparation was ground and incubated in dissociation medium (minimum essential medium [MEM] containing 0.05% collagenase, 0.02% DNase I, and 0.1% trypsin inhibitor) for 20 min at 37°C, with shaking. The cell suspension obtained from this first digestion was discarded to ensure that any residual endometrial cells were removed. The residual fragments of the myometrial preparation were then subjected to three successive 40-min digestions in the same dissociation medium. Cell suspensions were collected and cultured in MEM supplemented with 10% FCS in Petri dishes at 37°C under an atmosphere of 5% CO2/95% humidified air for 20 min to eliminate putative adherent cells, such as fibroblasts and macrophages. Myometrial cells in suspension then were cultured in the same conditions at a plating density of 15 x 103 cells/cm2. The medium was changed every 2 days, and the cells were kept in serum-free medium for 24 h before the experiment. The cells were positive for desmin and and 
actin, all of which are markers for smooth muscle cells (detected by immunofluorescence, data not shown).
The human lung epithelial cell line A549 and the human promonocytic cell line U937 used in some experiments were both obtained from the European Collection of Cell Cultures (ECCAC, Salisbury, UK). They were cultured in CM supplemented with 10% heat-inactivated and endotoxin-free FCS.
Myometrial and endometrial strips (25 mg of tissue) were allowed to equilibrate for 25 min in 5 ml of a pH 7.4 Krebs bicarbonate buffer consisting of 117 mM NaCl, 4.7 mM KCl, 1.1 mM MgSO4, 1.2 mM KH2PO4, 24 mM NaHCO3, 0.8 mM CaCl2, and 1 mM glucose (under a 95% O2/5% CO2 gas phase), with constant stirring. Tissues then were incubated for the appropriate time in the presence or absence of the agent to be tested. Reactions were stopped by immersing the uterine tissues in liquid nitrogen. Tissues were homogenized in 1 ml cold solubilization buffer A, consisting of 1% (v/v) Triton X-100 in 50 mM HEPES (pH 7.4), 150 mM NaCl, 100 mM NaF, 10% (v/v) glycerol, 10 mM sodium pyrophosphate, 200 µM orthovanadate, 10 mM EDTA, 10 µg/ml aprotinin and leupeptin, and 0.5 mM phenylmethylsulphonyl fluoride (PMSF) and centrifuged at 10 000 x g for 20 min at 4°C.
When used, starved primary myometrial cells were incubated in the presence of various agents, and reactions were stopped by aspiration of the incubation medium and the addition of 50 µl solubilization buffer A. Cells were detached by scraping on ice and centrifuged at 10 000 x g for 20 min at 4°C.
Preparation of Membrane and Cytosolic Fractions
Endometrial and myometrial strips (50 mg) were homogenized with an Ultra-Turax homogenizer (Janke and Kunkel, IMLAB, France) in 1 ml cold buffer B containing 10 mM Tris-HCl, 0.5 mM EDTA, 250 mM sucrose, 10 µg/ml aprotinin, 10 µg/ml leupeptin, and 1 mM PMSF, and were centrifuged for 5 min at 700 x g. Supernatants then were centrifuged for 20 min at 100 000 x g. The resulting supernatants represented the cytosolic fractions, and the corresponding pellets, suspended in buffer A, constituted the membrane fractions. Protein concentrations were determined using a Bradford assay.
Western Blot Analysis of Phosphorylated MAPK1/3
Equivalent amounts of protein (40 µg) from the 10 000 x g supernatants of detergent extracts were heated for 10 min at 95° C with Laemmli sample buffer and analyzed by 10% SDS-PAGE. The separated proteins were transferred to nitrocellulose sheets and probed with a polyclonal anti-active MAPK1/3 antibody (1:5000 v/v). The immunoreactive bands were visualized by enhanced chemiluminescence after incubation with horseradish peroxidase-conjugated anti-rabbit IgG. The intensities of immunoreactive bands were quantified with a densitometer (Molecular Dynamics, Sunnyvale, CA).
Serum-starved myometrial cells in 24-well plates were incubated in the presence or absence of Re-LPS (100 ng/ml) or SFTPA1 (20 µg/ml) for the indicated times. Detergent-extracted proteins (40 µg) were subjected to electrophoresis and Western blotting with polyclonal anti-PTGS2 antibodies (1:1000).
We analyzed the binding of fluorescent SFTPA1 by incubating myometrial cells grown on glass coverslips to 60% confluence with 20 µg/ml Texas red-labeled SFTPA1 in the presence or absence of unlabeled SFTPA1. Reactions were stopped by washing the coverslips in PBS. Control experiments were performed with fluorescent BSA replacing the fluorescent SFTPA1. Cells then were fixed by incubation for 15 min in 4% formaldehyde in PBS. A Carl Zeiss Axiophot 2 instrument (Le Pecq, France) was used for fluorescence microscopy.
Total cellular RNA was isolated from rat fetal lung (collected 1 day before parturition) and from myometrial strips of nonpregnant and pregnant (12 and 21 days of gestation) rats with Insta-Pure reagent, according to the manufacturer's protocol, and 5 µg total RNA was reverse transcribed to generate cDNA using 200 units Moloney murine leukemia virus reverse transcriptase, 0.2 mM dNTPs, and 10 µM random hexamer primers. The target cDNA was amplified using one tenth of the reverse transcribed cDNA preparation, 0.2 mM dNTPs, 2 mM MgCl2, 2.5 units Taq polymerase, and 100 pmol primers in PCR buffer. The following primers, based on the rat Sftpa1 mRNA published sequence (GenBank accession number: NM_017329), were used: Sftpa1 primer set 1 (expected size 591 bp): 5'-TTCACCCTCTTCTTGACTGT-3' (sense) and 5'-AGCCCCATCCAGGTAGTGGA-3' (antisense); and Sftpa1 primer set 2 (expected size 349 bp): 5'-TTGTCGCTGGTATCAAGTGC-3' (sense) and 5'-GATCCTTGCAAGCTGAGGAC-3' (antisense). Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) mRNA was used as an internal standard and was amplified with the following primer set (expected size 522 bp): 5'-CTGCACCACCAACTGCTTAG-3' (sense) and 5'-ACCACCCTGTTGCTGTAGCC-3' (antisense). The mixtures were amplified in a thermal cycler (iCycler; Bio-Rad) under the following conditions: 94°C (15 sec), 60°C (30 sec), and 72°C (30 sec) for 30 cycles. The resulting PCR products were analyzed by electrophoresis in a 1.5% agarose gel stained with ethidium bromide.
Isolation of Surfactant Protein A
Human SFTPA1 was purified from the bronchoalveolar lavage of patients with alveolar proteinosis. After centrifugation and sequential extractions with butanol and octyl glucoside [30], the purity of SFTPA1 was checked by SDS-PAGE in a 12% polyacrylamide gel under reducing conditions. The SFTPA1 samples used contained 0.12 pg endotoxin contaminant per microgram of protein, as determined with a Limulus Amebocyte Lysate (LAL) assay (BioWhittaker).
Preparation of Labeled SFTPA1 and its Collagenase-Resistant Fragment (CRF)
A fluorescent SFTPA1 derivative was prepared as previously described [31]. Briefly, SFTPA1 was incubated with Texas Red succinimidyl ester in 5 mM Tris-HCl buffer (pH 8.0). After dialysis, the covalent coupling of the fluorochrome was assessed by SDS-PAGE with viewing of the gel under ultraviolet light. The capacity of this fluorescent derivative of SFTPA1 to interact with DPPC vesicles was found to be identical to that of the native SFTPA1 [31].
Radioiodinated SFTPA1 was prepared by the iodogen method of Greenwood et al. [32]. The specific activity of the [125I]-labeled SFTPA1 was 2.25 x 106 cpm/µg. The purity of the preparation was assessed by SDS-PAGE and autoradiography. We then investigated whether the labeled SFTPA1 retained its capacity to bind DPPC and Re-LPS, two well-known ligands of SFTPA1 [21]. Briefly, solutions of DPPC or Re-LPS dissolved in organic solvents (2 µg in 100 µl) were subjected to evaporation under vacuum in the wells of solvent-resistant (polypropylene) microplates. Plates were saturated with 2% BSA, washed, and incubated overnight at 4°C or for 3 h at 37°C with solutions (100 µl) of radiolabeled SFTPA1 in a buffer (5 mM Tris, 150 mM NaCl, pH 7.4) supplemented with 0.2% BSA and 2 mM calcium chloride. Plates were thoroughly washed, and the residual bound radioactivity was recovered in 10% SDS and measured.
The CRF of SFTPA1 was obtained as previously described [33]. Unlabeled or radioiodinated samples of SFTPA1 were incubated with collagenase (0.5 units per µg of SFTPA1) in 5 mM Tris/5 mM CaCl2 at 37°C for 16 h. Controls were prepared in parallel by incubating samples in the absence of enzyme. The material recovered after incubation was stored frozen, for subsequent use at 4°C. The purity of the fragments was assessed by SDS-PAGE analysis. CRF prepared in these conditions retained its capacity to bind Re-LPS, as previously described [34].
Autoradiography of SFTPA1 Binding Proteins Using [125I]-SFTPA1
Tissues were homogenized in solubilization buffer B, consisting of 1% CHAPS in 50 mM Tris (pH 7.5) and 300 mM NaCl, supplemented with a cocktail of protease inhibitors (10 µg/ml aprotinin, 1 mM PMSF, 2 µg/ml pepstatin and leupeptin), 100 µM orthovanadate, and 2 mM iodoacetamide. The extracts were centrifuged at 12 000 x g for 20 min at 4°C. Detergent-extracted proteins were analyzed by SDS-PAGE in 8% polyacrylamide gels. Proteins were transferred onto nitrocellulose membranes. Membranes were blocked by incubation (overnight at 4°C) with 2% BSA in TS, which consisted of 5 mM Tris and 150 mM NaCl (pH 7.4). Blots were washed with 0.1% Tween-20 in TS and incubated (overnight at 4°C or for 2 h at room temperature) with 0.2% BSA in TS supplemented with [125I]-SFTPA1 (0.5 to 2 x 106 cpm), or [125I]-CRF (4 x 106 cpm), in the presence of 2 mM CaCl2. Blots were thoroughly washed with 0.1% Tween-20 in TS containing 2 mM CaCl2, dried, and placed against x-ray film for autoradiography (Hyperfilm; Amersham Biosciences).
All of the bands shown are from one representative experiment, and all densitometric data are the mean ± SD of three independent experiments.
Interaction of SFTPA1 with Uterine Proteins
SFTPA1 binding has been reported to involve specific SFTPA1 receptors in macrophages and alveolar type II cells [35].We assessed the ability of [125I]-SFTPA1 to bind to uterine proteins separated by SDS-PAGE and transferred onto nitrocellulose membranes. Two proteins (around 200 kDa and 55 kDa) interacting with [125I]-SFTPA1 were detected in myometrial and endometrial extracts (Fig. 1A). Equivalent amounts of protein were loaded into each well on SDS-PAGE. Our data, therefore, indicate that these putative SFTPA1 binding sites were much more abundant in myometrial than in endometrial extracts. The 200-kDa SFTPA1 binding protein also was detected in protein extracts from the lung epithelial type II cell line A549 and the macrophagelike cell line U937 obtained under the same experimental conditions (same amount of total protein). This 200-kDa SFTPA1 binding protein was first described by Chroneos et al. in U937 cells [35]. This protein was much more abundant in myometrial and endometrial extracts than in A549 and U937 extracts. The 55-kDa SFTPA1 binding protein also was much less abundant in A549 cells than in myometrial and endometrial extracts, and was not detectable in U937 cells, which can be considered to act as a negative control. These results provide the first evidence that SFTPA1 binding proteins are present in uterine tissues.
|
We used cytosol-enriched and membrane-enriched preparations from rat myometrial and endometrial tissues for the localization of SFTPA1 binding sites. An SDS-PAGE analysis of fractions containing equivalent amounts of total protein indicated that the 200-kDa SFTPA1 binding protein was present in similar amounts in membrane and cytosolic fractions (Fig. 1B). The 55-kDa SFTPA1 binding protein clearly was more abundant in the membrane fraction than in the cytosolic fraction. Both binding sites were more abundant in myometrial than in endometrial membranes.
As pregnancy-related changes occur in the levels of many proteins, including membrane receptors [36], we analyzed SFTPA1 binding protein levels in myometrial strips prepared from rats on Days 12 and 21 of gestation. SFTPA1 binding analysis demonstrated that two SFTPA1 binding proteins (55 and 210 kDa) were present in pregnant (midgestation and term gestation) rat myometrium (Fig. 1C). However, no quantitative differences were observed between the myometria of pregnant and nonpregnant rats.
Fluorescence Microscopy Analysis of SFTPA1 Binding to Myometrial Cells
The presence of SFTPA1 binding sites in the myometrium was confirmed by incubating living myometrial cells with a fluorescent SFTPA1 derivative. After 5 min at 37°C, fluorescent spots of SFTPA1 were observed on the periphery of cells (Fig. 2, A and B). Prior incubation of the cells with a 10-fold excess of unlabeled SFTPA1 abolished the binding of labeled SFTPA1 to myometrial cells (Fig. 2, E and F). The specificity of SFTPA1 binding was confirmed by the lack of binding of Texas Red-BSA to myometrial cells (Fig. 2, C and D).
|
Presence of SFTPA1 in Rat Myometrium at Late Gestation
SFTPA1 has been implicated in the initiation of labor in mice [19]. SFTPA1 and SFTPD also have recently been detected in the genital tract of mares [27]. The unexpected presence of SFTPA1 binding sites on myometrial cells led us to look for the possible presence of SFTPA1 in rat myometrium at various stages of gestation. Immunodetection of the SFTPA1 protein was carried out in detergent extracts prepared from myometrial tissues from pregnant and nonpregnant rats and rat primary myometrial cells in culture. Western blot analysis with an anti-SFTPA1 antibody (Fig. 3A) indicated that SFTPA1 was abundant in fetal lung (positive control obtained from fetus at Day 21 of gestation) but was not detectable in the myometrium of nonpregnant rats, in rats at midgestation (Day 12), and in primary cultures of myometrial cells. However, SFTPA1 was detected in myometrial extracts obtained from rats at Days 19 and 21 of gestation, and it disappeared at postpartum. The anti-SFTPA1 antibody recognized the same single polypeptide in rat fetal lung and myometrium from pregnant rats. The absence of detectable SFTPA1 in primary cultures of myometrial cells, and the transient presence of this protein during gestation, suggests that SFTPA1 is synthesized in the myometrium in response to a late gestation-specific stimulus, or that SFTPA1 is produced outside the myometrium. We analyzed Sftpa1 mRNA levels in rat fetal lung (positive control) and in the myometrium of nonpregnant or pregnant rats by RT-PCR. As expected, Sftpa1 mRNA levels were high in fetal lung. In contrast, this mRNA was undetectable in myometrium, even at Days 19 and 21 of gestation, when the corresponding protein was present. Thus, SFTPA1 was not synthesized in the myometrium, and its presence in this tissue may be due to a paracrine process.
|
Characterization of the Interaction Between SFTPA1 and SFTPA1 Binding Proteins in Myometrium
We characterized the interaction between SFTPA1 and its binding sites in the myometrium. The addition of unlabeled SFTPA1 abolished the binding of [125I]-SFTPA1 to both the 55-kDa and the 200-kDa SFTPA1 binding proteins of myometrial extracts (Fig. 4A). This result confirmed the specificity of the interaction of [125I]-SFTPA1 with the two myometrial proteins. SFTPA1 possesses a C-terminal globular domain with lectin activity and an N-terminal collagenlike domain. Mannan, a polysaccharide that efficiently inhibits the binding of SFTPA1 to immobilized sugars [37], did not block the binding of [125I]-SFTPA1 to the 55-kDa and 200-kDa SFTPA1 binding proteins. This suggests that the lectin activity of SFTPA1 is not involved in SFTPA1 binding to myometrial sites.
|
The C-terminal domain of SFTPA1 contains at least two metal binding sites [38]. The addition of calcium modifies the fluorescence and circular dichroism spectra of SFTPA1, reflecting a conformational change in the protein following calcium binding [39]. EDTA abolished the binding of [125I]-SFTPA1 to the 55- and 200-kDa proteins (Fig. 4A), suggesting that calcium is required for the interaction between SFTPA1 and myometrial binding sites.
We investigated whether a SFTPA1 with no collagen domain (removed by collagenase digestion) was able to interact with the SFTPA1 binding proteins of the myometrium. We isolated the CRF remaining after collagenase treatment of the radiolabeled SFTPA1 as previously described [33]. This fragment bound DPPC and Re-LPS, two well-known ligands of the globular domain of SFTPA1 (data not shown). [125I]-CRF did not interact with the 55- and 200-kDa proteins recognized by [125I]-SFTPA1 (Fig. 4B). Furthermore, unlabeled CRF did not inhibit the binding of [125I]-SFTPA1 to the 55- and 200-kDa proteins (data not shown). These results indicate that the globular domain of SFTPA1 alone was unable to interact with myometrial SFTPA1 binding sites, and that an intact collagenlike domain is required for the recognition of the two myometrial SFTPA1-binding proteins.
Influence of LPS on the Interaction of SFTPA1 with Myometrial Binding Sites
Some bacterial components, including Re-LPS, interact with SFTPA1. We therefore investigated whether Re-LPS affected the interaction of SFTPA1 with its myometrial binding sites. Blots of radiolabeled SFTPA1 (Fig. 5A, t – 1) showed that the prior incubation of [125I]-SFTPA1 for 1 h with Re-LPS from S. minnesota reduced the binding of [125I]-SFTPA1 to the 55- and 200-kDa SFTPA1 binding proteins. Binding also was inhibited when Re-LPS and [125I]-SFTPA1 were added simultaneously to the blotted proteins (Fig. 5A, t0). In contrast, when the blot was incubated for 1 h with [125I]-SFTPA1 before the addition of Re-LPS, no inhibition of SFTPA1 binding was observed (Fig. 5A, t + 1). Unlike Re-LPS, the complete form of LPS (smooth LPS) does not bind to SFTPA1 [22, 34, 40]. As expected, the presence of smooth LPS from S. minnesota did not inhibit [125I]-SFTPA1 binding in any of the conditions tested (Fig. 5B). Following the prior incubation of various concentrations of Re-LPS with [125I]-SFTPA1, we observed the dose-dependent inhibition of [125I]-SFTPA1 binding to the 200-kDa (Fig. 5C) and 55-kDa (Fig. 5D) myometrial cell proteins. Our data suggest that SFTPA1 was less efficiently recognized by the two myometrial SFTPA1 binding proteins when complexed with Re-LPS.
|
Effect of SFTPA1 on MAPK1/3 Activation in the Myometrium
Full-term labor in the rat is associated with an increase in basal MAPK1 activation [13, 41]. As SFTPA1 binding sites were present on rat myometrial cell membranes and SFTPA1 was present at late gestation, we evaluated the capacity of SFTPA1 to regulate MAPK1/3 activity. SFTPA1 induced MAPK1 activation (Fig. 6). The level of MAPK1 activation increased with the concentration of SFTPA1 with which the myometrial strips were incubated (Fig. 6A). The maximal response was obtained at a concentration of 20 µg/ml SFTPA1. This SFTPA1-mediated effect was rapid, reaching a maximum at about 2 to 5 min and then slowly declining (Fig. 6B). U0126, a selective inhibitor of MAPK1/3 kinase (MEK), markedly reduced the effect of SFTPA1, demonstrating the involvement of the MAPK1/3 cascade following the binding of SFTPA1. SFTPA1 also activated MAPK1 in myometrial cells in primary culture (Fig. 7). Consistent with the results of binding experiments (Fig. 4B) demonstrating that the CRF part of SFTPA1 did not bind to myometrial proteins, the CRF did not activate MAPK1 (Fig. 7). Furthermore, Re-LPS, which blocked the binding of SFTPA1 (Fig. 5A), also inhibited MAPK1 activation. LPS alone was able to stimulate MAPK1 activation (Fig. 7, A and B). Thus, SFTPA1 induced MAPK1 activation by interacting with myometrial binding sites.
|
|
Upregulation of PTGS2 Levels by SFTPA1
We investigated the effects of SFTPA1 on PTGS2 levels in the myometrium, using cells in primary culture. PTGS2, an enzyme upregulated by the MAPK1/3 pathway [18], plays an important role in parturition. Our data (Fig. 8, A and B) provide the first evidence that treatment with SFTPA1 increases PTGS2 levels in rat uterine smooth muscle cells. This effect was mainly detectable at 4 h and remained detectable at 8 h of treatment. In similar experimental conditions actin levels were unaffected. Our data also showed that LPS upregulated PTGS2 protein levels. SFTPA1 complexed with LPS, which failed to activate MAPK1, did not upregulate the expression of PTGS2 protein (Fig. 8, A and B).
|
This study provides the first evidence for the presence of SFTPA1 binding sites in rat myometrial and endometrial tissues and in primary culture of myometrial cells. In the myometrium, SFTPA1 induced the activation of MAPK1/3 and the upregulation of PTGS2, two enzymes playing a crucial role in parturition. We show that the uterine binding sites of SFTPA1 correspond to two proteins of around 55 and 200 kDa. Results obtained by fluorescence microscopy indicated that the binding of SFTPA1 to myometrial cells was rapid. The specificity of the myometrial SFTPA1 binding proteins was demonstrated by the complete inhibition of radiolabeled SFTPA1 binding in the presence of excess unlabeled ligand. Chroneos et al. [35] described a 210-kDa receptor for SFTPA1, SFTPA1R210, in macrophages, alveolar type II cells, and lymphocytes. This receptor was reported to be calcium dependent, mannose insensitive, and bound to the collagenlike domain of SFTPA1 [35, 42]. These features were identical to those of the larger of the two SFTPA1 binding proteins (about 200 kDa) that we observed in the myometrium and endometrium. This myometrial protein therefore probably corresponds to SFTPA1R210, which was recently identified as the unconventional cell surface myosin 18A [43].
The second SFTPA1 binding protein (about 55 kDa) was not detected in the promonocytic cell line U937, which was used as negative control, but was observed in endometrial and myometrial tissues and in the alveolar cell line A549. This protein is thus present in both lung and uterine cells. The interaction of SFTPA1 with the 55-kDa protein, like that with the 200-kDa protein, was not abolished or reduced by mannan, required the intact collagenlike domain of SFTPA1, and was calcium dependent. The addition of calcium induces conformational changes in the protein following calcium binding [39]. These conformational changes may thus be important for the recognition of SFTPA1 by its binding sites in myometrial cells. The 55-kDa myometrial SFTPA1 binding protein may correspond to a known SFTPA1 receptor: CD14 or calreticulin, both of which are 55-kDa proteins. Moreover, CD14 in the human female reproductive tract [44] interacts with SFTPA1 [34], and calreticulin complexed with CD91 [45] seems to be an active receptor of SFTPA1 [46]. However, the 55-kDa SFTPA1 binding protein of the myometrium is unlikely to be CD14 or calreticulin because U937 cells, which contain CD14 and calreticulin, do not contain the 55-kDa SFTPA1 binding protein. We found that SFTPA1 binding sites also were present in rat myometrium at different stages of gestation, without significant pregnancy-related changes. This suggests that the effects of SFTPA1 may occur in the uterus are independent of hormonal status.
We investigated the function of the SFTPA1 binding proteins detected on the surface of myometrial cells (Fig. 2). Previous studies demonstrated that MAPK pathway activation plays an important role in the regulation of both myometrial cell proliferation [14, 47] and the contraction associated with term labor and preterm labor [13, 41, 48, 49]. The end of rat gestation coincides with the upregulation of MAPK1/3 activity [49]. A recent study suggested that SFTPA1 secreted by the maturing fetal lung triggers the initiation of parturition [19]. We therefore investigated the potential link between SFTPA1 and MAPK1/3 activity in uterine tissues. Our results provide the first demonstration that the addition of SFTPA1 to the incubation medium stimulates MAPK1 in a dose- and time-dependent manner in myometrial strips and myometrial cells. Myometrial cells in primary culture are therefore an appropriate model for investigating the early and long-term effects of SFTPA1.
MAPK1/3 activated by SFTPA1 may operate at the level of caldesmon phosphorylation or by increasing calcium sensitivity, facilitating myometrial contraction [13]. Alternatively, MAPK1/3 activated by SFTPA1 may contribute to the PLA2 stimulation involved in prostaglandin production in rat myometrium [29] that leads to myometrial contraction [3]. Prostaglandin production is catalyzed by PTGS2, the expression of which is increased by MAPK1/3 activation in human uterine myocytes [18]. We showed in myometrial cells in culture that prolonged treatment with SFTPA1 resulted in a significant increase in PTGS2 protein levels. The presence of SFTPA1 in the myometrium may therefore be of physiologic relevance: it may enhance myometrial contraction or contribute to the multiple, coordinated processes associated with parturition.
It has been suggested that the SFTPA1 secreted by the maturing fetal lung may activate fetal macrophages, which migrate from the amniotic fluid to the maternal uterus, increasing the proinflammatory status of the mouse uterus and inducing labor [19]. However, the trafficking of fetal macrophages from the amniotic cavity or chorioamniotic membranes into the myometrium does not occur during human labor [50]. Our results show that, in addition to the indirect mechanism proposed for SFTPA1 [19], SFTPA1 binding sites are present at the end of gestation, together with high levels of SFTPA1. This protein does not seem to be synthesized in situ by the myometrial cells, as no Sftpa1 mRNA was detected. Our data therefore provide an additional mechanism by which SFTPA1, probably produced outside the uterus, could directly activate myometrial cells, thereby contributing to parturition. Human SFTPA1 levels in the amniotic fluid are known to increase from less than 3 µg/ml at 30 wk of gestation to more than 24 µg/ml at term [51]. The presence of immunoreactive SFTPA1 has already been reported in the lumen of mouse uterine gland ducts [26] and, more recently, in the genital tract of mares [27]. It remains to be determined why the myometrium is endowed with SFTPA1 binding sites, and how SFTPA1 gets to the myometrium. One plausible explanation is that following fetal membrane rupture at the end of gestation, SFTPA1 produced by the fetus and released into the amniotic fluid may come into contact with uterine tissues. This presence of SFTPA1 at the right time and in the right place might provide an additional mechanism for enhancing the myometrial contractions generally felt 12 to 24 h later in human parturition. However, even in the absence of disruption or leakage, the maternal-fetal barrier is not totally impenetrable, and substances can cross the placenta by various routes, including pinocytosis, endocytosis, and active transport. We therefore cannot exclude the possibility that SFTPA1 is transported by an as yet unknown mechanism, requiring further investigation. SFTPA1 has been detected in cord blood serum [52, 53]. This represents one possible route to the myometrium (a well-irrigated tissue), contributing to the regulation of delivery by SFTPA1.
Inflammation seems to play an important role in parturition [2, 54]. Infection has been shown to be tightly related to premature birth, and the administration of LPS, a proinflammatory bacterial component, has been reported to induce preterm labor. The cellular effects of LPS are mediated via its interaction with membrane Toll-like receptor 4 (TLR-4), which is expressed in a number of cell types, including human and rat myometrial cells [10]. Our study shows that in uterine tissues, LPS, like SFTPA1, can induce MAPK1 activation and PTGS2 expression.
Studies in Sftpa1-deficient mice have shown that SFTPA1 is involved in the clearance and killing of pathogens and in reducing the proinflammatory effects of LPS. This anti-LPS effect of SFTPA1 also was observed in our study, because complexing LPS with SFTPA1 markedly reduced the stimulatory effect of SFTPA1 (its capacity to activate MAPK1/3 and induce PTGS2). Further evidence for negative cross talk between LPS and SFTPA1 was provided by the observation that LPS impaired the recognition of SFTPA1 by its binding sites on the myometrium. Our data suggest that the negative cross talk between LPS and SFTPA1 may account for the protective effect of SFTPA1 at the end of gestation, in line with the lower percentage of severe infections for term deliveries than for preterm births [55]. The presence of SFTPA1 and its cognate receptors in reproductive tissues, including the uterus in particular, provides new insight into the possible roles of this multifunctional protein in parturition: contributing to both the defense of the fetus against infections of the maternal genital tract, and as a hormone contributing to labor.
ACKNOWLEDGMENTS
We thank Monique Synguelakis for help with radiolabeling experiments and Ginette Vilain for expert technical assistance.
FOOTNOTES
1Supported by the Fondation pour la Recherche Médicale (contract grant number ACE20040901625), the Centre National de la Recherche Scientifique, and Université Paris-Sud. 
Correspondence: 2Zahra Tanfin, UMR 8619, CNRS, Batiment 430, Universite Paris Sud, 91400 Orsay, France. FAX: 33 169 853 715; e-mail: zahra.tanfin{at}ibbmc.u-psud.fr
Received: 17 October 2006.
First decision: 14 November 2006.
Accepted: 29 December 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  I. Garcia-Verdugo, Z. Tanfin, E. Dallot, M.-J. Leroy, and M. Breuiller-Fouche Surfactant Protein A Signaling Pathways in Human Uterine Smooth Muscle Cells Biol Reprod, August 1, 2008; 79(2): 348 - 355. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
