Expression and Function of Toll-Like Receptor 4 in the Endometrial Cells of the Uterus
http://www.100md.com
《内分泌学杂志》
Royal Veterinary College, Departments of Veterinary Clinical Sciences (S.H., D.P.F., E.J.W., S.T.L., I.M.S.) and Pathology and Infectious Diseases (D.W.), University of London, North Mymms, Hatfield, Hertfordshire AL9 7TA, United Kingdom
Department of Veterinary Clinical Science and Animal Husbandry (H.D.), Faculty of Veterinary Science, University of Liverpool, Leahurst, Neston CH64 7TE, United Kingdom
Centre for Veterinary Science (C.E.B.), Department of Clinical Veterinary Medicine, University of Cambridge, Cambridge CB3 0ES, United Kingdom
Abstract
Prostaglandins have a central role in many endocrine functions in mammals, including regulation of the life span of the corpus luteum by prostaglandin F2 (PGF) and prostaglandin E2 (PGE), which are secreted by the uterine endometrium. However, the uterus is readily infected with bacteria such as Escherichia coli, which disrupt luteolysis. Immune cells detect E. coli by Toll-like receptor 4 (TLR4) binding its pathogenic ligand, lipopolysaccharide (LPS), although signaling requires accessory molecules such as CD14. The objective of this study was to determine the effect of E. coli or LPS on the function of bovine endometrial cells, and whether purified populations of epithelial and stromal cells express the molecules involved in LPS recognition. In addition, because the female sex hormones estradiol and progesterone modify the risk of uterine infection, their effect on the LPS response was investigated. Endometrial explants produced prostaglandins in response to LPS, with an increased ratio of PGE to PGF. Addition of LPS or E. coli to stromal and epithelial cells stimulated production of PGE and PGF and increased their cyclooxygenase 2 mRNA expression. The production of prostaglandins was abrogated by an LPS antagonist. In addition, estradiol and progesterone inhibited the production of PGE and PGF in response to LPS, indicating a role for steroid hormones in the response to bacterial infection. For the first time, Toll-like receptor 4 mRNA and CD14 mRNA and protein were detected in bovine endometrial stromal and epithelial cells by RT-PCR and flow cytometry. In conclusion, epithelial and stromal cells detect and respond to bacteria, which modulate their endocrine function.
Introduction
PROSTAGLANDINS HAVE a central role in many reproductive functions in mammals (1). The duration of ovarian cycles and pregnancy depends on the presence of an active corpus luteum in the ovary, and the life span of the corpus luteum is regulated by the secretion of prostaglandin F2 (PGF) and prostaglandin E2 (PGE) from the uterine endometrium (1). Both PGF and PGE are synthesized from arachadonic acid (AA) under the regulation of cyclooxygenase 2 (COX-2) isoenzymes in the endometrium (2, 3). The uterine epithelial cells predominantly secrete PGF, whereas the stromal cells produce PGE (4). The secretion of PGF, in response to endometrial oxytocin receptor activation, induces regression of the corpus luteum (luteolysis) and initiates the follicular phase of the ovarian cycle (1). In contrast, PGE has a luteotrophic role to help maintain the corpus luteum, particularly during pregnancy. However, infection of the uterus with bacteria disrupts the process of luteolysis, causing either premature regression of the corpus luteum or failure of luteolysis and an extended luteal phase (5).
For most of the reproductive cycle in humans and animals, the uterus is thought to be sterile or at least clear of pathogenic bacteria, but it is readily contaminated with bacteria during sexual intercourse and around the time of parturition. Sexually transmitted infections are widespread in the human population with an estimated 350 million new, mainly bacterial, cases each year and a prevalence of 5% in US women of reproductive age (www.who.int/topics/sexually_transmitted_infections). The majority of these infections are initially asymptomatic, but the consequences range from subfertility to severe pelvic inflammatory disease (PID) (6, 7, 8). Mammalian fertility is also compromised by PID associated with bacterial infections after parturition. Bos taurus is a particularly extreme species, where bacterial contamination of the uterus is ubiquitous after parturition in dairy cattle (9, 10, 11). In 15% of these animals, uterine bacterial infections persist for more than 3 wk after parturition, causing clinical diseases ranging from acute PID to chronic endometritis (12). These infections perturb normal ovarian cycles by suppressing follicular growth and disrupting luteolysis (5, 10). The well-characterized uterine disease, the ready availability of tissues, and the similarity of uterine bacterial pathogens among mammals, make cattle a good model for studying the effects of infection on the endocrine and immune functions of the uterus.
Escherichia coli are the most commonly isolated pathogenic bacteria from cases of clinical uterine disease in cattle (9, 10, 11). In the uterine lumen, there are high concentrations of the main pathogenic ligand of E. coli, lipopolysaccharide (LPS) (9). The endometrium provides a barrier against infection and an opportunity to detect these bacteria by innate immune receptors recognizing conserved sequences on pathogens, known as pathogen-associated molecular patterns (13). A key group of receptors that recognize pathogen-associated molecular patterns are the Toll-like receptors (TLRs), which were first identified on immune cells, but have since been identified on other cell types (14, 15, 16, 17, 18). Engagement of these receptors initiates a signaling cascade stimulating the production of immune mediators such as TNF and nitric oxide (NO), which orchestrate the immune response to clear the infection (14, 19). The TLR pathways can also stimulate the production of prostaglandins by immune cells (20). In the human endometrium, nine TLRs are expressed at the mRNA level including TLR4 (16, 18, 21, 22). It is the principal role of TLR4 to detect LPS, although signaling through TLR4 also requires accessory molecules such as CD14, LPS binding protein, and MD2 (14, 15).
The endometrium is regulated by changing concentrations of the female sex hormones estradiol and progesterone during the ovarian cycle, and these steroids also have a profound effect on the establishment of infections (23, 24). For example, in humans, rodents, and cattle, progesterone suppresses uterine immune function by decreasing the proliferative capacity of lymphocytes, thereby increasing the susceptibility to bacterial infection (23, 24). Furthermore, increased concentrations of progesterone are associated with decreased production of PGF (4, 25). Conversely, estradiol may play a role in the recruitment of immune cells as more macrophages are present in the endometrium when estradiol concentrations are high in rodents (24). However, the effect of steroids on the immune response to infection in the bovine uterus is less clearly defined.
The objective of the present study was to determine the effect of E. coli or LPS on the endocrine function of bovine endometrial cells, and whether purified populations of epithelial and stromal cells express the mRNA for molecules involved in LPS recognition and binding. In addition, the effect of steroids on the stromal and epithelial cell response to LPS was investigated.
Materials and Methods
Tissue explants, cell isolation, and culture
Bovine uteri from postpubertal nonpregnant animals with no evidence of genital disease were collected at a local abattoir immediately after slaughter and kept on ice until further processing in the laboratory. The physiological stage of the reproductive cycle for each genital tract was determined by observation of the ovarian morphology (26). Genital tracts with an ovarian stage I corpus luteum were selected for endometrial tissue and cell culture, and only the horn ipsilateral to the corpus luteum was used.
For tissue explants, the endometrium was cut into strips and placed into serum-free RPMI 1640 (Sigma, Poole, UK) supplemented with 50 IU/ml penicillin, 50 μg/ml streptomycin, and 2.5 μg/ml amphotericin B (Sigma), working under sterile conditions. The strips were then chopped into 1-mm3 pieces using a mechanical tissue chopper (McIlwain Laboratory Engineering, Gomshall, UK) and placed into HBSS (Sigma) (27). For each experimental group within the study, 50 mg of tissue were weighed in triplicate and then transferred onto sterile tissue-lined metal grids in six-well plates (Nunc, Nottingham, UK) with 4.25 ml serum-free RPMI 1640 per well (28). The tissue explants were incubated at 37 C, 5% CO2 in air, in a humidified incubator overnight, and then supernatants were removed and replaced with fresh media.
For cell isolation, endometrial tissue was used as previously described (29, 30) with the following modifications. Briefly, tissue was digested in 25 ml sterile filtered digestive solution, which was made by dissolving 50 mg trypsin III (Roche, Lewes, UK), 50 mg collagenase II (Sigma), 100 mg BSA (Sigma), and 10 mg DNase I (Sigma) in 100 ml phenol-red-free HBSS. After a 1.5-h incubation in a shaking water bath at 37 C, the cell suspension was filtered through a 40-μm mesh (Fisher Scientific, Loughborough, UK) to remove undigested material, and the filtrate was resuspended in phenol-red-free HBSS containing 10% fetal bovine serum (FBS; Sigma) and 3 μg/ml trypsin inhibitor (Sigma) (washing medium). The suspension was centrifuged at 100 x g for 10 min and, after two further washes in washing medium, the cells were resuspended in RPMI 1640 containing 10% FBS, 50 IU/ml penicillin, 50 μg/ml streptomycin, and 2.5 μg/ml amphotericin B. The cells were plated at a density of 1 x 105 cells in 2 ml per well using 24-well plates (Nunc). To obtain separate stromal and epithelial cell populations, the cell suspension was removed 18 h after plating, which allowed selective attachment of stromal cells (29). The removed cell suspension was then replated and incubated allowing epithelial cells to adhere (31). Stromal and epithelial cell populations were distinguished by cell morphology as previously described (29) and the purity was greater than 95% as determined by microscopy and the production of prostaglandins (4). The culture media was changed every 48 h until the cells reached confluence. All cultures were maintained at 37 C, 5% CO2 in air, in a humidified incubator.
Tissue explant and cell culture challenge
Tissue explants were challenged after 24 h culture with different concentrations of oxytocin (OT, 3–300 nM; Bachem, St. Helens, UK), LPS (0.03–1 μg/ml; Sigma; E. coli serotype 055.B5) and polymyxin B (2 μg/ml; Sigma) individually or in combination as indicated in Results.
Stromal and epithelial cells were challenged once confluence had been reached with different concentrations of AA (10–300 μM; Sigma), OT (100 nM), LPS (0.1–3 μg/ml), heat-killed E. coli (102–105 CFU/ml, isolated from a case of clinical bovine endometritis associated with pyrexia) (10), polymyxin B (2 μg/ml), 17- estradiol (3 pg/ml; Sigma), or progesterone (5 ng/ml; Sigma) individually or in combination and for the period of time indicated in Results.
For both tissue explants and cell cultures, the supernatants were harvested and frozen at –20 C until used for cytokine and prostaglandin determination. Endometrial cells were collected immediately for RNA isolation or flow cytometry.
Prostaglandin RIAs
Culture supernatants were analyzed for PGE and PGF by RIA as previously described (32). Samples were diluted in 0.05 M Tris buffer containing 0.1% gelatin and 0.01% sodium azide. Standards and tritiated tracers for the PGs were purchased from Sigma and Amersham International PLC (Amersham, Little Chalfont, Buckinghamshire, UK), respectively. The antisera were a generous gift from Prof. N. L. Poyser (University of Edinburgh, Edinburgh, Scotland, UK) and their cross-reactivities were: PGF antiserum, 0.54% with PGE; PGE antiserum, 0.47% with PGF (33). The limits of detection for PGE and PGF were 2 pg/tube and 1 pg/tube, respectively. The intraassay and interassay coefficients of variation were 4.4 and 7.8% for PGE, and 5.1 and 9.7% for PGF, respectively.
RT-PCR
Total RNA was isolated from cell cultures using the RNeasy Mini kit (Qiagen, Crawley, UK) and first strand cDNA was prepared using SuperScript II RNase H– Reverse Transcriptase (Invitrogen Life Technologies, Paisley, Scotland, UK) according to the manufacturers’ protocols. Amplification of cDNA used the following conditions: denaturation for 5 min at 94 C; followed by 30 cycles of 94 C for 30 sec, 54–56 C (depending on primer Tm) for 30 sec, and 72 C for 30 sec; followed by a final extension of 5 min at 72 C. Primer combinations were designed using the MacVector software package and were purchased from MWG (https://ecom.mwgdna.com). Primer sequences are presented in Table 1. All PCR products were sent for sequencing (MRC Geneservice, Cambridge, UK) and products showed >98% homology to published sequences on the NCBI database (http://www.ncbi.nlm.nih.gov/blast).
Monoclonal antibodies and flow cytometry
The sources of mouse mAb and isotypes, secondary reagents, and methods for flow cytometry have been described in detail (34). The defined surface antigens assessed and mAb used to detect the molecules were: CD14 (CCG33; IgG1) and CD45 (CC171; IgG2a). Isotype matched controls were murine mAb against chicken surface proteins AV20 (Bu-1, IgG1) and AV37 (chicken spleen cell subset, IgG2a) (kindly provided by F. Davison, Institute for Animal Health, Compton, UK). Bound mAb were detected with FITC and PE-labeled mouse isotype-specific reagents (Southern Biotechnology Associates, Birmingham, AL). In each case, 10,000 cells were analyzed on a FACSAria (BD Biosciences, Oxford, UK) and immunofluorescent staining was analyzed using the FC Express software (DeNovo Software, Ontario, Canada).
Determination of TNF- and NO concentrations
Levels of bioactive TNF- were determined using L929 cells as previously described (35), with the following modifications. The L929 cells were cultured in DMEM (Sigma) supplemented with 12.5% FBS, 50 IU/ml penicillin, 50 μg/ml streptomycin, and 20 mM HEPES buffer (Sigma). Cells were plated at a density of 2.5 x 104 cells per well in a 96-well plate (Nunc) in 100 μl medium. Cytotoxicity was determined by the colorimetric 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide assay involving the addition of 0.1 μg/ml 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide dye (Sigma) to each well. The cells were then lysed using 100 μl DMSO (Sigma) per well, and color development was read at 560 nm on a Spectra Max 250 (Molecular Devices). The limit of detection was 10 pg/ml; standards were made using recombinant human TNF- (Sigma) and cross-reactivity was determined using recombinant bovine TNF- (kindly provided by Prof. C. Howard, Institute for Animal Health, Compton, UK).
The concentrations of NO were measured using the Greiss Reagent System (Promega) according to the manufacturer’s instructions. The limit of detection was 2.5 μM nitrite.
Statistical analysis
Data were analyzed using mixed model ANOVA in SAS version 9.1. Results are quoted as mean ± SEM, and significance attributed when P < 0.05.
Results
Endometrial tissues and cells responds to OT in vitro with the production of prostaglandins
To determine whether the endometrial tissue and cell cultures were functional in vitro, the production of PGE and PGF in response to OT challenge was measured. Bovine endometrial tissue explants produced both PGE and PGF in response to OT stimulation (data not shown). Furthermore, stromal and epithelial cells isolated from these explants by tissue digestion released PGE and PGF, respectively, in response to OT challenge (Fig. 1). However, the media required supplementation with AA, the substrate for prostaglandin synthesis. Stromal cells produced little PGF and epithelial cells produced no PGE (stromal cells, 0.5 ± 0.1 ng/ml PGF; and epithelial cells, PGE levels below limits of detection at 100 nM OT stimulation). As expected, more PGF was produced in response to OT than PGE (172 ± 4 vs. 13 + 1 ng/ml at 100 nM OT stimulation).
LPS stimulates prostaglandin response by endometrial tissue explants
After parturition, the bovine uterus becomes contaminated with microorganisms, with E. coli being the most commonly isolated pathogen (10, 11). Because LPS is the main pathogenic ligand of E. coli and a strong activator of the innate immune response, LPS was added to tissue explant cultures and the production of PGE and PGF measured. Endometrial explants produced increasing amounts of PGE and PGF in response to LPS in a dose-dependent manner (Fig. 2, A and C). Moreover, this response was dependent on LPS because the LPS-stimulated production of PGE or PGF could be abrogated by addition of polymyxin B (Fig. 2, B and D), which neutralizes LPS (36).
Endometrial cells produce prostaglandin in response to LPS independently of leukocytes
Prostaglandins can be produced by a variety of cells, including leukocytes, in response to pathogenic stimulation (37). Therefore, leukocytes present in endometrial tissues explants could contribute to the production of PGs. Thus, endometrial stromal and epithelial cells cultures were established and analyzed for the presence of leukocytes using the pan-leukocyte marker CD45. No CD45 mRNA was detected in samples derived from endometrial epithelial or stromal cell cultures (Fig. 3A). This observation was further confirmed using flow cytometric analysis, which showed an absence of CD45+ cells (Fig. 3B).
After verifying the absence of immune cells in the endometrial cell preparations, the capacity of the latter cells to produce PGs in response to pathogenic stimulation was investigated. Both stromal and epithelial cells challenged with heat-killed E. coli produced increased concentrations of PGE and PGF, respectively, in response to bacterial challenge (Fig. 4). The prostaglandin response was abrogated by the addition of polymyxin B to the cultures (Fig. 4). Because LPS is considered the active component of E. coli, the production of prostaglandin by stromal and epithelial cells was analyzed after incubation of the cells with LPS alone. Both stromal and epithelial cells produced increasing amounts of PGE and PGF, respectively, in a dose-dependent manner (Fig. 5). Stromal cells produced little PGF in response to LPS (2.31 + 0.33 vs. 1.09 + 0.27 ng/ml at 1 μg/ml LPS). Interestingly, epithelial cells did produce PGE (81.8 + 17.3 vs. 24.3 + 5.0 ng/ml at 1 μg/ml LPS). However, as our aim was to focus on the effect of LPS on the endocrine function of uterine cells, subsequent results focused on the polarized production of PGE and PGF from stromal and epithelial cells, respectively. The PGE and PGF response to LPS was significantly abrogated by neutralizing LPS with polymyxin B (Fig. 5). Additionally, endometrial epithelial and stromal cells stimulated with LPS, up-regulated COX-2 mRNA expression (see Fig. 7).
Bovine endometrial cells express members of the LPS-receptor complex
The ability of cells to respond to LPS is dependent on the expression of, and signaling through the TLR4-CD14 receptor complex (14, 19). Using primers designed for bovine TLR4 and CD14, mRNA transcripts for both were detected by RT-PCR in samples derived from endometrial stromal or epithelial cells (Fig. 6). Flow cytometric analysis confirmed the expression of CD14 on stromal and epithelial cells (Fig. 3B).
NO and TNF in endometrial cells after LPS
Endometrial epithelial and stromal cells stimulated with LPS up-regulated inducible NO synthase (iNOS) and TNF- mRNA (Fig. 7). However, the concentrations of NO or TNF- in culture supernatants were below the limits of detection of the assays.
Progesterone and estradiol partially inhibit LPS-induced prostaglandin production
Preincubation of stromal and epithelial cells with either estradiol or progesterone for 48 h resulted in a significantly reduced production of prostaglandins in response to LPS challenge (Fig. 8). Additionally, the perturbation of prostaglandin secretion was greater following progesterone than estradiol treatment (P < 0.05).
Discussion
The uterus is often contaminated with bacteria during sexual intercourse or after parturition, causing considerable infertility in humans and animals (8, 10). In cattle the uterus is always contaminated with bacteria after parturition and E. coli infections cause uterine disease, which disrupts luteolysis with either premature regression of the corpus luteum or failure of luteolysis (5). As well as being responsible for secretion of PGE and PGF, which regulate corpus luteum life span, the endometrium provides the first line of defense against infection of the uterus. In the present study, we examined the effect of E. coli or its pathogenic ligand, LPS, on the function of the endometrium. Pure populations of stromal or epithelial cells secreted PGE or PGF, in response to E. coli or LPS and the effect was abrogated by an LPS antagonist. Furthermore, it was shown for the first time that bovine epithelial and stromal cells express transcripts for TLR4-CD14 receptor complex, which is necessary for LPS recognition (14, 15). Thus, it appears that endometrial cells have an immune role in detecting and responding to bacteria, as well as fulfilling an endocrine role. Furthermore, the immune response disrupts the endocrine function of the endometrium. However, it was intriguing to note that the response to the bacterial challenge was modulated by physiological concentrations of ovarian steroids.
Both PGE and PGF are synthesized from AA in many cell types, under the regulation of COX-2 isoenzymes and these prostaglandins have numerous endocrine and immune functions (1, 2, 38). The ability of endometrial explants to produce both PGE and PGF after OT stimulation was confirmed in the present study. However, these explants consisted of a variety of cells, so the individual cell phenotypes were isolated and two methods used to confirm that the cultures were not contaminated with immune cells. First, RT-PCR determined the absence of mRNA transcripts for the pan-leukocyte marker CD45 in epithelial or stromal cells. Second, flow cytometric analysis using a bovine CD45 antibody confirmed the absence of contamination with CD45+ cells. In response to OT, the stromal cells produced predominantly PGE, whereas epithelial cells secreted PGF, confirming previous observations (4). Because research on the interactions of uterine cells with invading pathogens is often hampered by the lack of cell accessibility or established models, these bovine tissue cultures provide a biologically relevant model to study the immune function of the endometrium.
The main pathogenic ligand of E. coli, LPS, was added to tissue explants at concentrations spanning the range found in the uterus of infected animals (9). There was a dose-dependent secretion of PGE and PGF in response to LPS, confirming data where a single concentration of 0.1 μg/ml LPS was used to challenge endometrial explants (27). These observations were extended in the present study by demonstrating the specificity of the tissue response using an LPS antagonist, which abrogated the increased production of prostaglandins, even for the maximal LPS challenge. The explants in both studies secreted more PGE than PGF in response to LPS. This greater ratio of PGE to PGF production is the reverse of that when endometrial cells or explants are stimulated with OT. After uterine infection, the altered prostaglandin secretion toward PGE may disrupt luteolysis and ovarian cycles, and provide unfavorable conditions for the establishment of pregnancy. The increased prostaglandin secretion by the endometrial explants could reflect the presence of immune cells within the explants, as they also respond to LPS challenge by secreting PGE and, to a lesser extent, PGF (20, 38). So, to examine further the uterine response to bacterial challenge, it was necessary to isolate and challenge pure populations of epithelial and stromal cells that were not contaminated by leukocytes.
Endometrial stromal and epithelial cells challenged with heat killed E. coli secreted PGE and/or PGF. To ensure the bacteria were biologically relevant, E. coli was isolated from the uterus of a cow with persistent uterine disease. Although it is difficult to reconcile the numbers of bacteria added to the cultures with those in the uterine lumen, they are likely to be of a similar magnitude as swabs collected from the uterus of infected cows often contain several hundred E. coli (10). Interestingly, epithelial cells responded to the microbial challenge at lower concentrations of bacteria than the stromal cells, and this differential sensitivity may reflect their spatial roles in the endometrium. Epithelial cells are in contact with the uterine lumen, so that sensitivity to fewer bacteria than stromal cells should ensure a rapid immune response to remove the pathogen. Indeed, direct contact of epithelial cells with the uterine lumen plays an important role in rats, where E. coli stimulates secretion of an anti-bacterial product (39). The effect of E. coli on secretion of prostaglandins by the present cell cultures was, in part, mediated by LPS as polymyxin B abrogated the response. Therefore, the next step was to challenge the endometrial cells with purified LPS.
Addition of appropriate concentrations of E. coli LPS to stromal and epithelial cells stimulated a dose-dependent secretion of PGE and PGF, respectively, and there was increased expression of COX-2 mRNA. Similar to the explant cultures, there was more PGE than PGF produced as a consequence of LPS challenge. In addition, both epithelial and stromal cells produced PGE. This is in contrast to the greater PGF to PGE ratio after OT stimulation and further confirms that uterine infections disrupt ovarian cycles. The production of PGE by the epithelial cells may be of considerable significance, and this modulation of the prostaglandin pathways in these cells warrants further investigation.
Work in humans and rats has focused on the increased secretion of inflammatory molecules such as IL-8 and TNF after LPS challenge of endometrial cells, and prostaglandin production does not appear to have been investigated (21, 40). The specificity of the LPS response in the present study was demonstrated by addition of polymixin B, which abrogated the secretion of prostaglandins. However, as TLR4 is the principal receptor for the recognition of LPS, it was important to determine whether the bovine endometrium expresses this receptor (14, 15).
The present study is the first to report mRNA transcript expression of the TLR4-CD14 complex in pure populations of bovine endometrial epithelial and stromal cells. In the human, TLR4 mRNA has been identified in the endometrium and in endometrial epithelial cell lines (18, 22). Recently, TLR4 mRNA and protein expression were reported in primary cultures of epithelial and stromal cells from the human endometrium (21). However, the absence of a suitable antibody to bovine TLR4 hampers the detection of TLR4 protein on bovine endometrial cells. The TLR4 signaling pathway requires accessory molecules such as CD14 (14, 15). In the present study, a bovine CD14 antibody was used to demonstrate that epithelial and stromal cells express CD14 protein, in addition to expressing CD14 at the mRNA level. In the human, CD14 was detected by flow cytometry only on stromal cells and not on epithelial cells, which may reflect a species difference (21).
The innate immune response is governed by the ability of immune cells to recognize pathogens and produce immune mediators, such as TNF- and NO (14). Because the endometrial cells expressed TLR4 and responded to LPS with an enhanced production of prostaglandins, the capacity of these cells to produce TNF- and NO was determined. Transcripts for iNOS, the enzyme required for the production of NO, and TNF- were up-regulated by endometrial cells after LPS challenge, but NO and TNF- were not detected in culture supernatants. These observations are in contrast to findings in the rat, where endometrial epithelial cells secrete detectable concentrations of TNF in response to LPS (40). In contrast, NO and TNF in the bovine endometrium may act in an autocrine or paracrine manner or may produce concentrations that are below the limits of detection of the assays in the present study. Indeed, the concentrations of TNF- and NO that stimulate the production of prostaglandins by bovine stromal and epithelial cells are around the limits of detection of the assays in the present study (41, 42).
Progesterone suppresses uterine immune responses resulting in increased susceptibility of the uterus to infection, at least in rodents and cattle (23, 43, 44, 45). The role of estradiol in infection is dependent on the species, tissue, and concentration of the sex hormone, but is generally thought to confer protection against uterine infections (24). In the present study, basal secretion of PGE by stromal cells was unaffected by steroids, as previously described (4), although the suppression of PGF secretion by estradiol and stimulation by progesterone was less marked than in the earlier study. However, the endometrial cells both secreted less prostaglandin after LPS challenge, when they were preincubated with physiological concentrations of progesterone or estradiol. The effect of estradiol was not as marked as that after progesterone treatment and this observation concords with the more potent suppression of immunity in the endometrium by progesterone (23, 43, 44, 45).
Taken together, the results showed that bovine endometrial stromal and epithelial cells express TLR4 and CD14, and produced a functional response to Escherichia coli and its pathogenic ligand, LPS. There was increased secretion of PGE and PGF, and greater expression of mRNA for inflammatory mediators. The increased secretion of prostaglandins may explain why luteolysis is often disrupted during uterine infection. However, the response to LPS was dependent on the steroid hormone milieu, which has important implications for the establishment of uterine infections and treatment of disease. In conclusion, epithelial and stromal cells have an important immune function to detect and respond to bacteria, as well as fulfilling endocrine roles in the endometrium.
Footnotes
This work was supported by grants from the Biotechnology and Biological Sciences Research Council (Grant No. S19795) and The Wellcome Trust (Grant No. 064155).
S.H., D.P.F., D.W., E.J.W., S.T.L., H.D., C.E.B., and I.M.S. have nothing to declare.
First Published Online October 13, 2005
Abbreviations: AA, Arachadonic acid; COX, cyclooxygenase; FBS, fetal bovine serum; iNOS, inducible NO synthase; LPS, lipopolysaccharide; NO, nitric oxide; OT, oxytocin; PGE, prostaglandin E2; PGF, prostaglandin F2; PID, pelvic inflammatory disease; TLR, Toll-like receptor.
Accepted for publication September 30, 2005.
References
Poyser NL 1995 The control of prostaglandin production by the endometrium in relation to luteolysis and menstruation. Prostaglandins Leukot Essent Fatty Acids 53:147–195
Smith WL, Garavito RM, DeWitt DL 1996 Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem 271:33157–33160
Arosh JA, Parent J, Chapdelaine P, Sirois J, Fortier MA 2002 Expression of cyclooxygenases 1 and 2 and prostaglandin E synthase in bovine endometrial tissue during the estrous cycle. Biol Reprod 67:161–169
Asselin E, Goff AK, Bergeron H, Fortier MA 1996 Influence of sex steroids on the production of prostaglandins F2 and E2 and response to oxytocin in cultured epithelial and stromal cells of the bovine endometrium. Biol. Reprod 54:371–379
Opsomer G, Grohn YT, Hertl J, Coryn M, Deluyker H, de Kruif A 2000 Risk factors for post partum ovarian dysfunction in high producing dairy cows in Belgium: a field study. Theriogenology 53:841–857
Fathalla MF 1991 Reproductive health: a global overview. Ann NY Acad Sci 626:1–10
Quayle AJ 2002 The innate and early immune response to pathogen challenge in the female genital tract and the pivotal role of epithelial cells. J Reprod Immunol 57:61–79
Butler D 2004 The fertility riddle. Nature 432:38–39
Dohmen MJ, Joop K, Sturk A, Bols PE, Lohuis JA 2000 Relationship between intra-uterine bacterial contamination, endotoxin levels and the development of endometritis in postpartum cows with dystocia or retained placenta. Theriogenology 54:1019–1032
Sheldon IM, Noakes DE, Rycroft AN, Pfeiffer DU, Dobson H 2002 Influence of uterine bacterial contamination after parturition on ovarian dominant follicle selection and follicle growth and function in cattle. Reproduction 123:837–845
Williams EJ, Fischer DP, Pfeiffer DU, England GC, Noakes DE, Dobson H, Sheldon IM 2005 Clinical evaluation of postpartum vaginal mucus reflects uterine bacterial infection and the immune response in cattle. Theriogenology 63:102–117
Sheldon IM, Dobson H 2004 Postpartum uterine health in cattle. Anim Reprod Sci 82–83:295–306
Beutler B 2004 Innate immunity: an overview. Mol Immunol 40:845–859
Beutler B, Hoebe K, Du X, Ulevitch RJ 2003 How we detect microbes and repsond to them: the Toll-like receptors and their transducers. J Leukoc Biol 74:479–485
Akira S 2004 Toll receptor families: structure and function. Semin Immunol 16:1–2
Schaefer TM, Fahey JV, Wright JA, Wira CR 2005 Innate immunity in the human female reproductive tract: antiviral response of uterine epithelial cells to the TLR3 agonist poly(I:C). J Immunol 174:992–1002
Strober W 2004 Epithelial cells pay a Toll for protection. Nat Med 10:898–900
Young SL, Lyddon TD, Jorgenson RL, Misfeldt ML 2004 Expression of Toll-like receptors in human endometrial epithelial cells and cell lines. Am J Reprod Immunol 52:67–73
Takeda K, Akira S 2004 TLR signaling pathways. Semin Immunol 16:3–9
Uematsu S, Matsumoto M, Takeda K, Akira S 2002 Lipopolysaccharide-dependent prostaglandin E2 production is regulated by the glutathione-dependent prostaglandin E2 synthase gene induced by the Toll-like receptor 4/MyD88/NF-IL6 pathway. J Immunol 168:5811–5816
Hirata T, Osuga Y, Hirota Y, Koga K, Yoshino O, Harada M, Morimoto C, Yano T, Nishii O, Tsutsumi O, Taketani Y 2005 Evidence for the presence of Toll-like receptor 4 system in the human endometrium. J Clin Endocrinol Metab 90:548–556
Pioli PA, Amiel E, Schaefer TM, Connolly JE, Wira CR, Guyre PM 2004 Differential expression of Toll-like receptors 2 and 4 in tissues of the human female reproductive tract. Infect Immun 72:5799–5806
Lewis GS 2004 Steroidal regulation of uterine immune defenses. Anim Reprod Sci 82–83:281–294
Beagley KW, Gockel CM 2003 Regulation of innate and adaptive immunity by the female sex hormones oestradiol and progesterone. FEMS Immunol Med Microbiol 38:13–22
Del Vecchio RP, Matsas DJ, Inzana TJ, Sponenberg DP, Lewis GS 1992 Effect of intrauterine bacterial infusions and subsequent endometritis on prostaglandin F2 metabolite concentrations in postpartum beef cows. J Anim Sci 70:3158–3162
Ireland JJ, Murphee RL, Coulson PB 1980 Accuracy of predicting stages of bovine estrous cycle by gross appearance of the corpus luteum. J Dairy Sci 63:155–160
Leung ST, Cheng Z, Sheldrick EL, Derecka K, Flint AP, Wathes DC 2001 The effects of lipopolysaccharide and interleukins-1, -2 and -6 on oxytocin receptor expression and prostaglandin production in bovine endometrium. J Endocrinol 168:497–508
Mann GE 2001 Hormone control of prostaglandin F2 production and oxytocin receptor concentrations in bovine endometrium in explant culture. Domest Anim Endocrinol 20:217–226
Fortier MA, Guilbault LA, Grasso F 1988 Specific properties of epithelial and stromal cells from the endometrium of cows. J Reprod Fertil 83:239–248
Cheng Z, Elmes M, Abayasekera DRE, Wathes DC 2003 Effects of conjugated linoleic acid on prostaglandins produced by cells isolated from maternal intercotyledonary endometrium, fetal allantochorion and amnion in late pregnant ewes. Biochim Biophys Acta 1633:170–178
Kim JJ, Fortier MA 1995 Cell type specificity and protein kinase C dependency on the stimulation of prostaglandin E2 and prostaglandin F2 production by oxytocin and platelet-activating factor in bovine endometrial cells. J Reprod Fertil 103:239–247
Cheng Z, Robinson RS, Pushpakumara PG, Mansbridge RJ, Wathes DC 2001 Effect of dietary polyunsaturated fatty acids on uterine prostaglandin synthesis in the cow. J Endocrinol 171:463–473
Poyser NL 1987 Effects of various factors on prostaglandin synthesis by the guinea-pig uterus. J Reprod Fertil 81:269–276
Werling D, Hope JC, Chaplin P, Collins RA, Taylor G, Howard CJ 1999 Involvement of caveolae in the uptake of respiratory syncytial virus antigen by dendritic cells. J Leukoc Biol 66:50–58
Trost LC, Lemasters JJ 1994 A cytotoxicity assay for tumor necrosis factor employing a multiwell fluorescence scanner. Anal Biochem 220:149–153
Jacobs DM, Morrison DC 1977 Inhibition of the mitogenic response to lipopolysaccharide (LPS) in mouse spleen cells by polymyxin B. J Immunol 118:21–27
Harris SG, Padilla J, Koumas L, Ray D, Phipps RP 2002 Prostaglandins as modulators of immunity. Trends Immunol 23:144–150
Janeway C, Travers P 2001 Immunobiology. 5th ed. New York: Garland Publishing
Fahey JV, Rossoll RM, Wira CR 2005 Sex hormone regulation of anti-bacterial activity in rat uterine secretions and apical release of anti-bacterial factor(s) by uterine epithelial cells in culture. J Steroid Biochem Mol Biol 93:59–66
Crane-Godreau MA, Wira CR 2005 CCL20/macrophage inflammatory protein 3 and tumor necrosis factor production by primary uterine epithelial cells in response to treatment with lipopolysaccharide or Pam3Cys. Infect Immun 73:476–484
Woclawek-Potocka I, Deptula K, Bah MM, Lee HY, Okuda K, Skarzynski DJ 2004 Effects of nitric oxide and tumor necrosis factor- on production of prostaglandin F2 and E2 in bovine endometrial cells. J Reprod Dev 50:333–340
Skarzynski DJ, Miyamoto Y, Okuda K 2000 Production of prostaglandin F2 by cultured bovine endometrial cells in response to tumor necrosis factor : cell type specificity and intracellular mechanisms. Biol Reprod 62:1116–1120
Lewis GS 2003 Steroidal regulation of uterine resistance to bacterial infection in livestock. Reprod Biol Endocrinol 1:117
Kaushic C, Zhou F, Murdin AD, Wira CR 2000 Effects of estradiol and progesterone on susceptibility and early immune responses to Chlamydia trachomatis infection in the female reproductive tract. Infect Immun 68:4207–4216
Kaushic C, Murdin AD, Underdown BJ, Wira CR 1998 Chlamydia trachomatis infection in the female reproductive tract of the rat: influence of progesterone on infectivity and immune response. Infect Immun 66:893–898(Shan Herath, Deborah P. Fischer, Dirk We)
Department of Veterinary Clinical Science and Animal Husbandry (H.D.), Faculty of Veterinary Science, University of Liverpool, Leahurst, Neston CH64 7TE, United Kingdom
Centre for Veterinary Science (C.E.B.), Department of Clinical Veterinary Medicine, University of Cambridge, Cambridge CB3 0ES, United Kingdom
Abstract
Prostaglandins have a central role in many endocrine functions in mammals, including regulation of the life span of the corpus luteum by prostaglandin F2 (PGF) and prostaglandin E2 (PGE), which are secreted by the uterine endometrium. However, the uterus is readily infected with bacteria such as Escherichia coli, which disrupt luteolysis. Immune cells detect E. coli by Toll-like receptor 4 (TLR4) binding its pathogenic ligand, lipopolysaccharide (LPS), although signaling requires accessory molecules such as CD14. The objective of this study was to determine the effect of E. coli or LPS on the function of bovine endometrial cells, and whether purified populations of epithelial and stromal cells express the molecules involved in LPS recognition. In addition, because the female sex hormones estradiol and progesterone modify the risk of uterine infection, their effect on the LPS response was investigated. Endometrial explants produced prostaglandins in response to LPS, with an increased ratio of PGE to PGF. Addition of LPS or E. coli to stromal and epithelial cells stimulated production of PGE and PGF and increased their cyclooxygenase 2 mRNA expression. The production of prostaglandins was abrogated by an LPS antagonist. In addition, estradiol and progesterone inhibited the production of PGE and PGF in response to LPS, indicating a role for steroid hormones in the response to bacterial infection. For the first time, Toll-like receptor 4 mRNA and CD14 mRNA and protein were detected in bovine endometrial stromal and epithelial cells by RT-PCR and flow cytometry. In conclusion, epithelial and stromal cells detect and respond to bacteria, which modulate their endocrine function.
Introduction
PROSTAGLANDINS HAVE a central role in many reproductive functions in mammals (1). The duration of ovarian cycles and pregnancy depends on the presence of an active corpus luteum in the ovary, and the life span of the corpus luteum is regulated by the secretion of prostaglandin F2 (PGF) and prostaglandin E2 (PGE) from the uterine endometrium (1). Both PGF and PGE are synthesized from arachadonic acid (AA) under the regulation of cyclooxygenase 2 (COX-2) isoenzymes in the endometrium (2, 3). The uterine epithelial cells predominantly secrete PGF, whereas the stromal cells produce PGE (4). The secretion of PGF, in response to endometrial oxytocin receptor activation, induces regression of the corpus luteum (luteolysis) and initiates the follicular phase of the ovarian cycle (1). In contrast, PGE has a luteotrophic role to help maintain the corpus luteum, particularly during pregnancy. However, infection of the uterus with bacteria disrupts the process of luteolysis, causing either premature regression of the corpus luteum or failure of luteolysis and an extended luteal phase (5).
For most of the reproductive cycle in humans and animals, the uterus is thought to be sterile or at least clear of pathogenic bacteria, but it is readily contaminated with bacteria during sexual intercourse and around the time of parturition. Sexually transmitted infections are widespread in the human population with an estimated 350 million new, mainly bacterial, cases each year and a prevalence of 5% in US women of reproductive age (www.who.int/topics/sexually_transmitted_infections). The majority of these infections are initially asymptomatic, but the consequences range from subfertility to severe pelvic inflammatory disease (PID) (6, 7, 8). Mammalian fertility is also compromised by PID associated with bacterial infections after parturition. Bos taurus is a particularly extreme species, where bacterial contamination of the uterus is ubiquitous after parturition in dairy cattle (9, 10, 11). In 15% of these animals, uterine bacterial infections persist for more than 3 wk after parturition, causing clinical diseases ranging from acute PID to chronic endometritis (12). These infections perturb normal ovarian cycles by suppressing follicular growth and disrupting luteolysis (5, 10). The well-characterized uterine disease, the ready availability of tissues, and the similarity of uterine bacterial pathogens among mammals, make cattle a good model for studying the effects of infection on the endocrine and immune functions of the uterus.
Escherichia coli are the most commonly isolated pathogenic bacteria from cases of clinical uterine disease in cattle (9, 10, 11). In the uterine lumen, there are high concentrations of the main pathogenic ligand of E. coli, lipopolysaccharide (LPS) (9). The endometrium provides a barrier against infection and an opportunity to detect these bacteria by innate immune receptors recognizing conserved sequences on pathogens, known as pathogen-associated molecular patterns (13). A key group of receptors that recognize pathogen-associated molecular patterns are the Toll-like receptors (TLRs), which were first identified on immune cells, but have since been identified on other cell types (14, 15, 16, 17, 18). Engagement of these receptors initiates a signaling cascade stimulating the production of immune mediators such as TNF and nitric oxide (NO), which orchestrate the immune response to clear the infection (14, 19). The TLR pathways can also stimulate the production of prostaglandins by immune cells (20). In the human endometrium, nine TLRs are expressed at the mRNA level including TLR4 (16, 18, 21, 22). It is the principal role of TLR4 to detect LPS, although signaling through TLR4 also requires accessory molecules such as CD14, LPS binding protein, and MD2 (14, 15).
The endometrium is regulated by changing concentrations of the female sex hormones estradiol and progesterone during the ovarian cycle, and these steroids also have a profound effect on the establishment of infections (23, 24). For example, in humans, rodents, and cattle, progesterone suppresses uterine immune function by decreasing the proliferative capacity of lymphocytes, thereby increasing the susceptibility to bacterial infection (23, 24). Furthermore, increased concentrations of progesterone are associated with decreased production of PGF (4, 25). Conversely, estradiol may play a role in the recruitment of immune cells as more macrophages are present in the endometrium when estradiol concentrations are high in rodents (24). However, the effect of steroids on the immune response to infection in the bovine uterus is less clearly defined.
The objective of the present study was to determine the effect of E. coli or LPS on the endocrine function of bovine endometrial cells, and whether purified populations of epithelial and stromal cells express the mRNA for molecules involved in LPS recognition and binding. In addition, the effect of steroids on the stromal and epithelial cell response to LPS was investigated.
Materials and Methods
Tissue explants, cell isolation, and culture
Bovine uteri from postpubertal nonpregnant animals with no evidence of genital disease were collected at a local abattoir immediately after slaughter and kept on ice until further processing in the laboratory. The physiological stage of the reproductive cycle for each genital tract was determined by observation of the ovarian morphology (26). Genital tracts with an ovarian stage I corpus luteum were selected for endometrial tissue and cell culture, and only the horn ipsilateral to the corpus luteum was used.
For tissue explants, the endometrium was cut into strips and placed into serum-free RPMI 1640 (Sigma, Poole, UK) supplemented with 50 IU/ml penicillin, 50 μg/ml streptomycin, and 2.5 μg/ml amphotericin B (Sigma), working under sterile conditions. The strips were then chopped into 1-mm3 pieces using a mechanical tissue chopper (McIlwain Laboratory Engineering, Gomshall, UK) and placed into HBSS (Sigma) (27). For each experimental group within the study, 50 mg of tissue were weighed in triplicate and then transferred onto sterile tissue-lined metal grids in six-well plates (Nunc, Nottingham, UK) with 4.25 ml serum-free RPMI 1640 per well (28). The tissue explants were incubated at 37 C, 5% CO2 in air, in a humidified incubator overnight, and then supernatants were removed and replaced with fresh media.
For cell isolation, endometrial tissue was used as previously described (29, 30) with the following modifications. Briefly, tissue was digested in 25 ml sterile filtered digestive solution, which was made by dissolving 50 mg trypsin III (Roche, Lewes, UK), 50 mg collagenase II (Sigma), 100 mg BSA (Sigma), and 10 mg DNase I (Sigma) in 100 ml phenol-red-free HBSS. After a 1.5-h incubation in a shaking water bath at 37 C, the cell suspension was filtered through a 40-μm mesh (Fisher Scientific, Loughborough, UK) to remove undigested material, and the filtrate was resuspended in phenol-red-free HBSS containing 10% fetal bovine serum (FBS; Sigma) and 3 μg/ml trypsin inhibitor (Sigma) (washing medium). The suspension was centrifuged at 100 x g for 10 min and, after two further washes in washing medium, the cells were resuspended in RPMI 1640 containing 10% FBS, 50 IU/ml penicillin, 50 μg/ml streptomycin, and 2.5 μg/ml amphotericin B. The cells were plated at a density of 1 x 105 cells in 2 ml per well using 24-well plates (Nunc). To obtain separate stromal and epithelial cell populations, the cell suspension was removed 18 h after plating, which allowed selective attachment of stromal cells (29). The removed cell suspension was then replated and incubated allowing epithelial cells to adhere (31). Stromal and epithelial cell populations were distinguished by cell morphology as previously described (29) and the purity was greater than 95% as determined by microscopy and the production of prostaglandins (4). The culture media was changed every 48 h until the cells reached confluence. All cultures were maintained at 37 C, 5% CO2 in air, in a humidified incubator.
Tissue explant and cell culture challenge
Tissue explants were challenged after 24 h culture with different concentrations of oxytocin (OT, 3–300 nM; Bachem, St. Helens, UK), LPS (0.03–1 μg/ml; Sigma; E. coli serotype 055.B5) and polymyxin B (2 μg/ml; Sigma) individually or in combination as indicated in Results.
Stromal and epithelial cells were challenged once confluence had been reached with different concentrations of AA (10–300 μM; Sigma), OT (100 nM), LPS (0.1–3 μg/ml), heat-killed E. coli (102–105 CFU/ml, isolated from a case of clinical bovine endometritis associated with pyrexia) (10), polymyxin B (2 μg/ml), 17- estradiol (3 pg/ml; Sigma), or progesterone (5 ng/ml; Sigma) individually or in combination and for the period of time indicated in Results.
For both tissue explants and cell cultures, the supernatants were harvested and frozen at –20 C until used for cytokine and prostaglandin determination. Endometrial cells were collected immediately for RNA isolation or flow cytometry.
Prostaglandin RIAs
Culture supernatants were analyzed for PGE and PGF by RIA as previously described (32). Samples were diluted in 0.05 M Tris buffer containing 0.1% gelatin and 0.01% sodium azide. Standards and tritiated tracers for the PGs were purchased from Sigma and Amersham International PLC (Amersham, Little Chalfont, Buckinghamshire, UK), respectively. The antisera were a generous gift from Prof. N. L. Poyser (University of Edinburgh, Edinburgh, Scotland, UK) and their cross-reactivities were: PGF antiserum, 0.54% with PGE; PGE antiserum, 0.47% with PGF (33). The limits of detection for PGE and PGF were 2 pg/tube and 1 pg/tube, respectively. The intraassay and interassay coefficients of variation were 4.4 and 7.8% for PGE, and 5.1 and 9.7% for PGF, respectively.
RT-PCR
Total RNA was isolated from cell cultures using the RNeasy Mini kit (Qiagen, Crawley, UK) and first strand cDNA was prepared using SuperScript II RNase H– Reverse Transcriptase (Invitrogen Life Technologies, Paisley, Scotland, UK) according to the manufacturers’ protocols. Amplification of cDNA used the following conditions: denaturation for 5 min at 94 C; followed by 30 cycles of 94 C for 30 sec, 54–56 C (depending on primer Tm) for 30 sec, and 72 C for 30 sec; followed by a final extension of 5 min at 72 C. Primer combinations were designed using the MacVector software package and were purchased from MWG (https://ecom.mwgdna.com). Primer sequences are presented in Table 1. All PCR products were sent for sequencing (MRC Geneservice, Cambridge, UK) and products showed >98% homology to published sequences on the NCBI database (http://www.ncbi.nlm.nih.gov/blast).
Monoclonal antibodies and flow cytometry
The sources of mouse mAb and isotypes, secondary reagents, and methods for flow cytometry have been described in detail (34). The defined surface antigens assessed and mAb used to detect the molecules were: CD14 (CCG33; IgG1) and CD45 (CC171; IgG2a). Isotype matched controls were murine mAb against chicken surface proteins AV20 (Bu-1, IgG1) and AV37 (chicken spleen cell subset, IgG2a) (kindly provided by F. Davison, Institute for Animal Health, Compton, UK). Bound mAb were detected with FITC and PE-labeled mouse isotype-specific reagents (Southern Biotechnology Associates, Birmingham, AL). In each case, 10,000 cells were analyzed on a FACSAria (BD Biosciences, Oxford, UK) and immunofluorescent staining was analyzed using the FC Express software (DeNovo Software, Ontario, Canada).
Determination of TNF- and NO concentrations
Levels of bioactive TNF- were determined using L929 cells as previously described (35), with the following modifications. The L929 cells were cultured in DMEM (Sigma) supplemented with 12.5% FBS, 50 IU/ml penicillin, 50 μg/ml streptomycin, and 20 mM HEPES buffer (Sigma). Cells were plated at a density of 2.5 x 104 cells per well in a 96-well plate (Nunc) in 100 μl medium. Cytotoxicity was determined by the colorimetric 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide assay involving the addition of 0.1 μg/ml 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide dye (Sigma) to each well. The cells were then lysed using 100 μl DMSO (Sigma) per well, and color development was read at 560 nm on a Spectra Max 250 (Molecular Devices). The limit of detection was 10 pg/ml; standards were made using recombinant human TNF- (Sigma) and cross-reactivity was determined using recombinant bovine TNF- (kindly provided by Prof. C. Howard, Institute for Animal Health, Compton, UK).
The concentrations of NO were measured using the Greiss Reagent System (Promega) according to the manufacturer’s instructions. The limit of detection was 2.5 μM nitrite.
Statistical analysis
Data were analyzed using mixed model ANOVA in SAS version 9.1. Results are quoted as mean ± SEM, and significance attributed when P < 0.05.
Results
Endometrial tissues and cells responds to OT in vitro with the production of prostaglandins
To determine whether the endometrial tissue and cell cultures were functional in vitro, the production of PGE and PGF in response to OT challenge was measured. Bovine endometrial tissue explants produced both PGE and PGF in response to OT stimulation (data not shown). Furthermore, stromal and epithelial cells isolated from these explants by tissue digestion released PGE and PGF, respectively, in response to OT challenge (Fig. 1). However, the media required supplementation with AA, the substrate for prostaglandin synthesis. Stromal cells produced little PGF and epithelial cells produced no PGE (stromal cells, 0.5 ± 0.1 ng/ml PGF; and epithelial cells, PGE levels below limits of detection at 100 nM OT stimulation). As expected, more PGF was produced in response to OT than PGE (172 ± 4 vs. 13 + 1 ng/ml at 100 nM OT stimulation).
LPS stimulates prostaglandin response by endometrial tissue explants
After parturition, the bovine uterus becomes contaminated with microorganisms, with E. coli being the most commonly isolated pathogen (10, 11). Because LPS is the main pathogenic ligand of E. coli and a strong activator of the innate immune response, LPS was added to tissue explant cultures and the production of PGE and PGF measured. Endometrial explants produced increasing amounts of PGE and PGF in response to LPS in a dose-dependent manner (Fig. 2, A and C). Moreover, this response was dependent on LPS because the LPS-stimulated production of PGE or PGF could be abrogated by addition of polymyxin B (Fig. 2, B and D), which neutralizes LPS (36).
Endometrial cells produce prostaglandin in response to LPS independently of leukocytes
Prostaglandins can be produced by a variety of cells, including leukocytes, in response to pathogenic stimulation (37). Therefore, leukocytes present in endometrial tissues explants could contribute to the production of PGs. Thus, endometrial stromal and epithelial cells cultures were established and analyzed for the presence of leukocytes using the pan-leukocyte marker CD45. No CD45 mRNA was detected in samples derived from endometrial epithelial or stromal cell cultures (Fig. 3A). This observation was further confirmed using flow cytometric analysis, which showed an absence of CD45+ cells (Fig. 3B).
After verifying the absence of immune cells in the endometrial cell preparations, the capacity of the latter cells to produce PGs in response to pathogenic stimulation was investigated. Both stromal and epithelial cells challenged with heat-killed E. coli produced increased concentrations of PGE and PGF, respectively, in response to bacterial challenge (Fig. 4). The prostaglandin response was abrogated by the addition of polymyxin B to the cultures (Fig. 4). Because LPS is considered the active component of E. coli, the production of prostaglandin by stromal and epithelial cells was analyzed after incubation of the cells with LPS alone. Both stromal and epithelial cells produced increasing amounts of PGE and PGF, respectively, in a dose-dependent manner (Fig. 5). Stromal cells produced little PGF in response to LPS (2.31 + 0.33 vs. 1.09 + 0.27 ng/ml at 1 μg/ml LPS). Interestingly, epithelial cells did produce PGE (81.8 + 17.3 vs. 24.3 + 5.0 ng/ml at 1 μg/ml LPS). However, as our aim was to focus on the effect of LPS on the endocrine function of uterine cells, subsequent results focused on the polarized production of PGE and PGF from stromal and epithelial cells, respectively. The PGE and PGF response to LPS was significantly abrogated by neutralizing LPS with polymyxin B (Fig. 5). Additionally, endometrial epithelial and stromal cells stimulated with LPS, up-regulated COX-2 mRNA expression (see Fig. 7).
Bovine endometrial cells express members of the LPS-receptor complex
The ability of cells to respond to LPS is dependent on the expression of, and signaling through the TLR4-CD14 receptor complex (14, 19). Using primers designed for bovine TLR4 and CD14, mRNA transcripts for both were detected by RT-PCR in samples derived from endometrial stromal or epithelial cells (Fig. 6). Flow cytometric analysis confirmed the expression of CD14 on stromal and epithelial cells (Fig. 3B).
NO and TNF in endometrial cells after LPS
Endometrial epithelial and stromal cells stimulated with LPS up-regulated inducible NO synthase (iNOS) and TNF- mRNA (Fig. 7). However, the concentrations of NO or TNF- in culture supernatants were below the limits of detection of the assays.
Progesterone and estradiol partially inhibit LPS-induced prostaglandin production
Preincubation of stromal and epithelial cells with either estradiol or progesterone for 48 h resulted in a significantly reduced production of prostaglandins in response to LPS challenge (Fig. 8). Additionally, the perturbation of prostaglandin secretion was greater following progesterone than estradiol treatment (P < 0.05).
Discussion
The uterus is often contaminated with bacteria during sexual intercourse or after parturition, causing considerable infertility in humans and animals (8, 10). In cattle the uterus is always contaminated with bacteria after parturition and E. coli infections cause uterine disease, which disrupts luteolysis with either premature regression of the corpus luteum or failure of luteolysis (5). As well as being responsible for secretion of PGE and PGF, which regulate corpus luteum life span, the endometrium provides the first line of defense against infection of the uterus. In the present study, we examined the effect of E. coli or its pathogenic ligand, LPS, on the function of the endometrium. Pure populations of stromal or epithelial cells secreted PGE or PGF, in response to E. coli or LPS and the effect was abrogated by an LPS antagonist. Furthermore, it was shown for the first time that bovine epithelial and stromal cells express transcripts for TLR4-CD14 receptor complex, which is necessary for LPS recognition (14, 15). Thus, it appears that endometrial cells have an immune role in detecting and responding to bacteria, as well as fulfilling an endocrine role. Furthermore, the immune response disrupts the endocrine function of the endometrium. However, it was intriguing to note that the response to the bacterial challenge was modulated by physiological concentrations of ovarian steroids.
Both PGE and PGF are synthesized from AA in many cell types, under the regulation of COX-2 isoenzymes and these prostaglandins have numerous endocrine and immune functions (1, 2, 38). The ability of endometrial explants to produce both PGE and PGF after OT stimulation was confirmed in the present study. However, these explants consisted of a variety of cells, so the individual cell phenotypes were isolated and two methods used to confirm that the cultures were not contaminated with immune cells. First, RT-PCR determined the absence of mRNA transcripts for the pan-leukocyte marker CD45 in epithelial or stromal cells. Second, flow cytometric analysis using a bovine CD45 antibody confirmed the absence of contamination with CD45+ cells. In response to OT, the stromal cells produced predominantly PGE, whereas epithelial cells secreted PGF, confirming previous observations (4). Because research on the interactions of uterine cells with invading pathogens is often hampered by the lack of cell accessibility or established models, these bovine tissue cultures provide a biologically relevant model to study the immune function of the endometrium.
The main pathogenic ligand of E. coli, LPS, was added to tissue explants at concentrations spanning the range found in the uterus of infected animals (9). There was a dose-dependent secretion of PGE and PGF in response to LPS, confirming data where a single concentration of 0.1 μg/ml LPS was used to challenge endometrial explants (27). These observations were extended in the present study by demonstrating the specificity of the tissue response using an LPS antagonist, which abrogated the increased production of prostaglandins, even for the maximal LPS challenge. The explants in both studies secreted more PGE than PGF in response to LPS. This greater ratio of PGE to PGF production is the reverse of that when endometrial cells or explants are stimulated with OT. After uterine infection, the altered prostaglandin secretion toward PGE may disrupt luteolysis and ovarian cycles, and provide unfavorable conditions for the establishment of pregnancy. The increased prostaglandin secretion by the endometrial explants could reflect the presence of immune cells within the explants, as they also respond to LPS challenge by secreting PGE and, to a lesser extent, PGF (20, 38). So, to examine further the uterine response to bacterial challenge, it was necessary to isolate and challenge pure populations of epithelial and stromal cells that were not contaminated by leukocytes.
Endometrial stromal and epithelial cells challenged with heat killed E. coli secreted PGE and/or PGF. To ensure the bacteria were biologically relevant, E. coli was isolated from the uterus of a cow with persistent uterine disease. Although it is difficult to reconcile the numbers of bacteria added to the cultures with those in the uterine lumen, they are likely to be of a similar magnitude as swabs collected from the uterus of infected cows often contain several hundred E. coli (10). Interestingly, epithelial cells responded to the microbial challenge at lower concentrations of bacteria than the stromal cells, and this differential sensitivity may reflect their spatial roles in the endometrium. Epithelial cells are in contact with the uterine lumen, so that sensitivity to fewer bacteria than stromal cells should ensure a rapid immune response to remove the pathogen. Indeed, direct contact of epithelial cells with the uterine lumen plays an important role in rats, where E. coli stimulates secretion of an anti-bacterial product (39). The effect of E. coli on secretion of prostaglandins by the present cell cultures was, in part, mediated by LPS as polymyxin B abrogated the response. Therefore, the next step was to challenge the endometrial cells with purified LPS.
Addition of appropriate concentrations of E. coli LPS to stromal and epithelial cells stimulated a dose-dependent secretion of PGE and PGF, respectively, and there was increased expression of COX-2 mRNA. Similar to the explant cultures, there was more PGE than PGF produced as a consequence of LPS challenge. In addition, both epithelial and stromal cells produced PGE. This is in contrast to the greater PGF to PGE ratio after OT stimulation and further confirms that uterine infections disrupt ovarian cycles. The production of PGE by the epithelial cells may be of considerable significance, and this modulation of the prostaglandin pathways in these cells warrants further investigation.
Work in humans and rats has focused on the increased secretion of inflammatory molecules such as IL-8 and TNF after LPS challenge of endometrial cells, and prostaglandin production does not appear to have been investigated (21, 40). The specificity of the LPS response in the present study was demonstrated by addition of polymixin B, which abrogated the secretion of prostaglandins. However, as TLR4 is the principal receptor for the recognition of LPS, it was important to determine whether the bovine endometrium expresses this receptor (14, 15).
The present study is the first to report mRNA transcript expression of the TLR4-CD14 complex in pure populations of bovine endometrial epithelial and stromal cells. In the human, TLR4 mRNA has been identified in the endometrium and in endometrial epithelial cell lines (18, 22). Recently, TLR4 mRNA and protein expression were reported in primary cultures of epithelial and stromal cells from the human endometrium (21). However, the absence of a suitable antibody to bovine TLR4 hampers the detection of TLR4 protein on bovine endometrial cells. The TLR4 signaling pathway requires accessory molecules such as CD14 (14, 15). In the present study, a bovine CD14 antibody was used to demonstrate that epithelial and stromal cells express CD14 protein, in addition to expressing CD14 at the mRNA level. In the human, CD14 was detected by flow cytometry only on stromal cells and not on epithelial cells, which may reflect a species difference (21).
The innate immune response is governed by the ability of immune cells to recognize pathogens and produce immune mediators, such as TNF- and NO (14). Because the endometrial cells expressed TLR4 and responded to LPS with an enhanced production of prostaglandins, the capacity of these cells to produce TNF- and NO was determined. Transcripts for iNOS, the enzyme required for the production of NO, and TNF- were up-regulated by endometrial cells after LPS challenge, but NO and TNF- were not detected in culture supernatants. These observations are in contrast to findings in the rat, where endometrial epithelial cells secrete detectable concentrations of TNF in response to LPS (40). In contrast, NO and TNF in the bovine endometrium may act in an autocrine or paracrine manner or may produce concentrations that are below the limits of detection of the assays in the present study. Indeed, the concentrations of TNF- and NO that stimulate the production of prostaglandins by bovine stromal and epithelial cells are around the limits of detection of the assays in the present study (41, 42).
Progesterone suppresses uterine immune responses resulting in increased susceptibility of the uterus to infection, at least in rodents and cattle (23, 43, 44, 45). The role of estradiol in infection is dependent on the species, tissue, and concentration of the sex hormone, but is generally thought to confer protection against uterine infections (24). In the present study, basal secretion of PGE by stromal cells was unaffected by steroids, as previously described (4), although the suppression of PGF secretion by estradiol and stimulation by progesterone was less marked than in the earlier study. However, the endometrial cells both secreted less prostaglandin after LPS challenge, when they were preincubated with physiological concentrations of progesterone or estradiol. The effect of estradiol was not as marked as that after progesterone treatment and this observation concords with the more potent suppression of immunity in the endometrium by progesterone (23, 43, 44, 45).
Taken together, the results showed that bovine endometrial stromal and epithelial cells express TLR4 and CD14, and produced a functional response to Escherichia coli and its pathogenic ligand, LPS. There was increased secretion of PGE and PGF, and greater expression of mRNA for inflammatory mediators. The increased secretion of prostaglandins may explain why luteolysis is often disrupted during uterine infection. However, the response to LPS was dependent on the steroid hormone milieu, which has important implications for the establishment of uterine infections and treatment of disease. In conclusion, epithelial and stromal cells have an important immune function to detect and respond to bacteria, as well as fulfilling endocrine roles in the endometrium.
Footnotes
This work was supported by grants from the Biotechnology and Biological Sciences Research Council (Grant No. S19795) and The Wellcome Trust (Grant No. 064155).
S.H., D.P.F., D.W., E.J.W., S.T.L., H.D., C.E.B., and I.M.S. have nothing to declare.
First Published Online October 13, 2005
Abbreviations: AA, Arachadonic acid; COX, cyclooxygenase; FBS, fetal bovine serum; iNOS, inducible NO synthase; LPS, lipopolysaccharide; NO, nitric oxide; OT, oxytocin; PGE, prostaglandin E2; PGF, prostaglandin F2; PID, pelvic inflammatory disease; TLR, Toll-like receptor.
Accepted for publication September 30, 2005.
References
Poyser NL 1995 The control of prostaglandin production by the endometrium in relation to luteolysis and menstruation. Prostaglandins Leukot Essent Fatty Acids 53:147–195
Smith WL, Garavito RM, DeWitt DL 1996 Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem 271:33157–33160
Arosh JA, Parent J, Chapdelaine P, Sirois J, Fortier MA 2002 Expression of cyclooxygenases 1 and 2 and prostaglandin E synthase in bovine endometrial tissue during the estrous cycle. Biol Reprod 67:161–169
Asselin E, Goff AK, Bergeron H, Fortier MA 1996 Influence of sex steroids on the production of prostaglandins F2 and E2 and response to oxytocin in cultured epithelial and stromal cells of the bovine endometrium. Biol. Reprod 54:371–379
Opsomer G, Grohn YT, Hertl J, Coryn M, Deluyker H, de Kruif A 2000 Risk factors for post partum ovarian dysfunction in high producing dairy cows in Belgium: a field study. Theriogenology 53:841–857
Fathalla MF 1991 Reproductive health: a global overview. Ann NY Acad Sci 626:1–10
Quayle AJ 2002 The innate and early immune response to pathogen challenge in the female genital tract and the pivotal role of epithelial cells. J Reprod Immunol 57:61–79
Butler D 2004 The fertility riddle. Nature 432:38–39
Dohmen MJ, Joop K, Sturk A, Bols PE, Lohuis JA 2000 Relationship between intra-uterine bacterial contamination, endotoxin levels and the development of endometritis in postpartum cows with dystocia or retained placenta. Theriogenology 54:1019–1032
Sheldon IM, Noakes DE, Rycroft AN, Pfeiffer DU, Dobson H 2002 Influence of uterine bacterial contamination after parturition on ovarian dominant follicle selection and follicle growth and function in cattle. Reproduction 123:837–845
Williams EJ, Fischer DP, Pfeiffer DU, England GC, Noakes DE, Dobson H, Sheldon IM 2005 Clinical evaluation of postpartum vaginal mucus reflects uterine bacterial infection and the immune response in cattle. Theriogenology 63:102–117
Sheldon IM, Dobson H 2004 Postpartum uterine health in cattle. Anim Reprod Sci 82–83:295–306
Beutler B 2004 Innate immunity: an overview. Mol Immunol 40:845–859
Beutler B, Hoebe K, Du X, Ulevitch RJ 2003 How we detect microbes and repsond to them: the Toll-like receptors and their transducers. J Leukoc Biol 74:479–485
Akira S 2004 Toll receptor families: structure and function. Semin Immunol 16:1–2
Schaefer TM, Fahey JV, Wright JA, Wira CR 2005 Innate immunity in the human female reproductive tract: antiviral response of uterine epithelial cells to the TLR3 agonist poly(I:C). J Immunol 174:992–1002
Strober W 2004 Epithelial cells pay a Toll for protection. Nat Med 10:898–900
Young SL, Lyddon TD, Jorgenson RL, Misfeldt ML 2004 Expression of Toll-like receptors in human endometrial epithelial cells and cell lines. Am J Reprod Immunol 52:67–73
Takeda K, Akira S 2004 TLR signaling pathways. Semin Immunol 16:3–9
Uematsu S, Matsumoto M, Takeda K, Akira S 2002 Lipopolysaccharide-dependent prostaglandin E2 production is regulated by the glutathione-dependent prostaglandin E2 synthase gene induced by the Toll-like receptor 4/MyD88/NF-IL6 pathway. J Immunol 168:5811–5816
Hirata T, Osuga Y, Hirota Y, Koga K, Yoshino O, Harada M, Morimoto C, Yano T, Nishii O, Tsutsumi O, Taketani Y 2005 Evidence for the presence of Toll-like receptor 4 system in the human endometrium. J Clin Endocrinol Metab 90:548–556
Pioli PA, Amiel E, Schaefer TM, Connolly JE, Wira CR, Guyre PM 2004 Differential expression of Toll-like receptors 2 and 4 in tissues of the human female reproductive tract. Infect Immun 72:5799–5806
Lewis GS 2004 Steroidal regulation of uterine immune defenses. Anim Reprod Sci 82–83:281–294
Beagley KW, Gockel CM 2003 Regulation of innate and adaptive immunity by the female sex hormones oestradiol and progesterone. FEMS Immunol Med Microbiol 38:13–22
Del Vecchio RP, Matsas DJ, Inzana TJ, Sponenberg DP, Lewis GS 1992 Effect of intrauterine bacterial infusions and subsequent endometritis on prostaglandin F2 metabolite concentrations in postpartum beef cows. J Anim Sci 70:3158–3162
Ireland JJ, Murphee RL, Coulson PB 1980 Accuracy of predicting stages of bovine estrous cycle by gross appearance of the corpus luteum. J Dairy Sci 63:155–160
Leung ST, Cheng Z, Sheldrick EL, Derecka K, Flint AP, Wathes DC 2001 The effects of lipopolysaccharide and interleukins-1, -2 and -6 on oxytocin receptor expression and prostaglandin production in bovine endometrium. J Endocrinol 168:497–508
Mann GE 2001 Hormone control of prostaglandin F2 production and oxytocin receptor concentrations in bovine endometrium in explant culture. Domest Anim Endocrinol 20:217–226
Fortier MA, Guilbault LA, Grasso F 1988 Specific properties of epithelial and stromal cells from the endometrium of cows. J Reprod Fertil 83:239–248
Cheng Z, Elmes M, Abayasekera DRE, Wathes DC 2003 Effects of conjugated linoleic acid on prostaglandins produced by cells isolated from maternal intercotyledonary endometrium, fetal allantochorion and amnion in late pregnant ewes. Biochim Biophys Acta 1633:170–178
Kim JJ, Fortier MA 1995 Cell type specificity and protein kinase C dependency on the stimulation of prostaglandin E2 and prostaglandin F2 production by oxytocin and platelet-activating factor in bovine endometrial cells. J Reprod Fertil 103:239–247
Cheng Z, Robinson RS, Pushpakumara PG, Mansbridge RJ, Wathes DC 2001 Effect of dietary polyunsaturated fatty acids on uterine prostaglandin synthesis in the cow. J Endocrinol 171:463–473
Poyser NL 1987 Effects of various factors on prostaglandin synthesis by the guinea-pig uterus. J Reprod Fertil 81:269–276
Werling D, Hope JC, Chaplin P, Collins RA, Taylor G, Howard CJ 1999 Involvement of caveolae in the uptake of respiratory syncytial virus antigen by dendritic cells. J Leukoc Biol 66:50–58
Trost LC, Lemasters JJ 1994 A cytotoxicity assay for tumor necrosis factor employing a multiwell fluorescence scanner. Anal Biochem 220:149–153
Jacobs DM, Morrison DC 1977 Inhibition of the mitogenic response to lipopolysaccharide (LPS) in mouse spleen cells by polymyxin B. J Immunol 118:21–27
Harris SG, Padilla J, Koumas L, Ray D, Phipps RP 2002 Prostaglandins as modulators of immunity. Trends Immunol 23:144–150
Janeway C, Travers P 2001 Immunobiology. 5th ed. New York: Garland Publishing
Fahey JV, Rossoll RM, Wira CR 2005 Sex hormone regulation of anti-bacterial activity in rat uterine secretions and apical release of anti-bacterial factor(s) by uterine epithelial cells in culture. J Steroid Biochem Mol Biol 93:59–66
Crane-Godreau MA, Wira CR 2005 CCL20/macrophage inflammatory protein 3 and tumor necrosis factor production by primary uterine epithelial cells in response to treatment with lipopolysaccharide or Pam3Cys. Infect Immun 73:476–484
Woclawek-Potocka I, Deptula K, Bah MM, Lee HY, Okuda K, Skarzynski DJ 2004 Effects of nitric oxide and tumor necrosis factor- on production of prostaglandin F2 and E2 in bovine endometrial cells. J Reprod Dev 50:333–340
Skarzynski DJ, Miyamoto Y, Okuda K 2000 Production of prostaglandin F2 by cultured bovine endometrial cells in response to tumor necrosis factor : cell type specificity and intracellular mechanisms. Biol Reprod 62:1116–1120
Lewis GS 2003 Steroidal regulation of uterine resistance to bacterial infection in livestock. Reprod Biol Endocrinol 1:117
Kaushic C, Zhou F, Murdin AD, Wira CR 2000 Effects of estradiol and progesterone on susceptibility and early immune responses to Chlamydia trachomatis infection in the female reproductive tract. Infect Immun 68:4207–4216
Kaushic C, Murdin AD, Underdown BJ, Wira CR 1998 Chlamydia trachomatis infection in the female reproductive tract of the rat: influence of progesterone on infectivity and immune response. Infect Immun 66:893–898(Shan Herath, Deborah P. Fischer, Dirk We)