Abstract
The objective of this study was to examine the expression of TLR by human primary uterine epithelial cells (UEC) and to determine whether exposure to the TLR agonist poly(I:C) would induce an antiviral response. The secretion of several cytokines and chemokines was examined as well as the mRNA expression of human β-defensin-1 and -2 (HBD1 and HBD2), IFN-β, and the IFN-β-stimulated genes myxovirus resistance gene 1 and 2′,5′ oligoadenylate synthetase. The expression of TLR1–9 by UEC was demonstrated by RT-PCR, with only TLR10 not expressed. Stimulation of UEC with the TLR3 agonist poly(I:C) induced the expression of the proinflammatory cytokines TNF-α, IL-6, GM-CSF, and G-CSF, as well as the chemokines CXCL8/IL-8, CCL2/MCP-1, and CCL4/MIP-1β. In addition, poly(I:C) exposure induced the mRNA expression of HBD1 and HBD2 by 6- and 4-fold, respectively. Furthermore, upon exposure to poly(I:C) UEC initiated a potent antiviral response resulting in the induction of IFN-β mRNA expression 70-fold and myxovirus resistance gene 1 and 2′,5′ oligoadenylate synthetase mRNA expression (107- and 96-fold), respectively. These results suggest that epithelial cells that line the uterine cavity are sensitive to viral infection and/or exposure to viral dsRNA released from killed epithelial cells. Not only do UEC release proinflammatory cytokines and chemokines that mediate the initiation of an inflammatory response and recruitment of immune cells to the site of infection, but they also express β-defensins, IFN-β, and IFN-β-stimulated genes that can have a direct inhibiting effect on viral replication.
Epithelial cells of the female reproductive tract (FRT)3 are the first line of defense against sexually transmitted diseases and invading pathogens. Understanding what role epithelial cells of the FRT play in immune surveillance and host defense is crucial for the development of effective mucosal vaccines as well as therapies for ongoing infections. Known to be an efficient physical barrier to infection, epithelial cells are in constant contact with the normal flora of the FRT and must discriminate between commensal organisms and pathogens (1). For the purpose of limiting the growth of commensal organisms on their external surface and defense of the underlying tissues from invading pathogens, epithelial cells have evolved innate immune antimicrobial functions as well as the ability to modulate the recruitment and activity of immune cells of both the innate and adaptive immune systems (2).
Important mediators of microbial recognition are germline-encoded receptors called TLR. TLR recognize conserved pathogen-associated molecular patterns (PAMP) synthesized by microorganisms but not by the host. Members of the TLR family, of which at least 11 TLR have been identified, recognize distinct PAMP produced by various bacterial, fungal, and viral pathogens. The recognition of bacterial PAMP, such as LPS, peptidoglycan, and flagellin, are mediated by TLR1, 2, 4, 5, and 6 (3, 4, 5, 6, 7, 8, 9). TLR7 and TLR8 recognize nucleotide derivatives, such as imiquimod, resiquimod, and loxoribine, and TLR9 binds unmethylated DNA found in bacteria (10, 11, 12, 13). In addition, recent reports have shown that TLR7 and TLR8 recognize both self and viral ssRNA (14, 15). TLR3 recognizes dsRNA, and more recently has been shown to be stimulated by cellular mRNA, potentially from cells killed by an invading pathogen (16, 17). The specific ligands for TLR10 and TLR11 have yet to be demonstrated (18, 19).
Epithelial cells throughout the body have been shown to express a wide range of TLR. Airway epithelial cells express TLR1–10, intestinal epithelial cells express TLR1–4, 6, and 9, and gastric epithelial cells express TLR2, 4, and 5 (20, 21, 22, 23, 24). Epithelial cells of the human FRT have also been shown to express a variety of TLR, including TLR1–3, 5, and 6 by vaginal and cervical epithelial cell lines and TLR1–3 and TLR6 by primary endocervical epithelial cells (25). We have previously shown that the uterine epithelial cell (UEC) line ECC-1 expressed TLR1–9 (26), but little work has been done to characterize the expression of TLR in human primary UEC.
Stimulation of human intestinal and pulmonary epithelial cells, as well as cell lines from the reproductive tract, with TLR agonists has been shown to induce proinflammatory cytokines and chemokines. TLR2 and TLR5 agonists induce the production of IL-6, CCL20/MIP-3α, CXCL8/IL-8, CXCL1/GROα, and CCL2/MCP-1 (24, 26, 27, 28, 29). TLR4 and TLR9 agonists induce epithelial cells to produce IL-6 and CXCL8/IL-8 (20, 30). Moreover, stimulation of epithelial cells has also been shown to induce the expression of small antimicrobial peptides called defensins (27, 29).
Defensins are small cysteine-rich, cationic peptides that are divided into three subfamilies: α-, β-, and θ-defensins. In humans, α-defensins are expressed in neutrophils, certain macrophage populations, and paneth cells of the small intestine and θ-defensin has been purified from the leukocytes and bone marrow of rhesus macaques (31, 32, 33, 34). In contrast, human β-defensins (HBD), of which four have been identified (HBD1–4), are largely expressed in various epithelial tissues, including skin, lung, and FRT (35, 36, 37, 38, 39, 40, 41, 42). HBD1 is generally thought to be constitutively expressed, while HBD2–4 have been shown to be highly inducible (35, 38, 39).
The roles that defensins play in defense of the host from pathogens are multifaceted, ranging from direct killing of invading microbes to linking innate and acquired immunities (31, 39, 41). Defensins are positively charged and interact with negatively charged components of microbial membranes and facilitate the permeabilization of Gram-positive and Gram-negative bacteria, fungi, and enveloped viruses (reviewed in Refs.38 and 42). Defensins are also crucial to the recruitment of immune cells to the sites of pathogenic challenge. HBD1–3 are chemotactic for immature dendritic cells (iDCs), CD4+CD45RO+ memory T lymphocytes, and mast cells, while HBD3 and HBD4 elicit monocyte chemotaxis (31, 35, 36, 43, 44). Although defensins have inhibitory effects on certain viruses, broad antiviral innate immunity is thought to be mediated by type I IFNs.
The IFN system is the first line of defense against viral pathogens and is crucial for limiting early replication and spread of viruses (45). Type I IFNs, known also as viral IFNs, exert their activity through the IFN-αβ receptor and include IFN-α, -β, -δ, -τ, -κ, and -ω. Production of type I IFNs are induced by dsRNA, which is synthesized by most viruses during their replication cycle (46), and by viral glycoproteins, including HIV gp120 (47, 48, 49, 50, 51). Recognition of dsRNA is mediated by TLR3, and expression of TLR3 is up-regulated following treatment of cells with IFN-αβ (52). Although all type I IFNs are important for an effective antiviral response, IFN-β is crucial to this process, because its absence results in the host being highly susceptible to viral infection (53).
Type I IFNs induce the synthesis of antiviral factors, such as 2′,5′-oligoadenylate synthetase (2′,5′-OAS) and myxovirus resistance gene A (MxA) (reviewed in Ref.54). Type I IFN induces the expression of the inactive form of 2′,5′-OAS, followed by activation of this enzyme in the presence of dsRNA. Once activated, 2′,5′-OAS polymerizes ATP into 2′,5′-linked oligoadenylates that are specific activators of a latent endoribonuclease, RNase L (55). RNase L degrades viral and cellular RNAs resulting in inhibition of protein synthesis (56). The 2′,5′-OAS/RNase L pathway has been shown to inhibit the replication of viruses belonging to different families, such as Picornaviridae (57, 58), Hepadnaviridae (59), Paramyxoviridae (60), Herpesviridae (61), and Retroviridae (HIV-1) (62, 63, 64). MxA belongs to the GTPase superfamily of dynamin-like GTPases (65, 66). MxA proteins associate with viral protein complexes and possess an intrinsic antiviral activity, although the mechanism by which MxA inhibits viral replication is poorly understood. MxA has been shown to be a crucial component in IFN-induced defense against several viral families, including Bunyaviridae (67, 68, 69, 70), Hepadnaviridae (71), Orthomyxoviridae (72, 73), Paramyxoviridae (74), Picornaviridae (75), Rhabdoviridae (73), Togaviridae (76), and possibly Retroviridae (77).
The goal of this study was to identify the TLR expressed by UEC and to determine whether exposure of epithelial cells to the TLR3 agonist poly(I:C) results in the secretion of proinflammatory cytokines and/or chemokines. Furthermore, the expression of antimicrobial defensins and antiviral genes was also examined following exposure of UEC to poly(I:C).
Materials and Methods
Source of uterine tissue
Uterine mucosal tissue was obtained immediately following surgery from premenopausal women who had undergone hysterectomies at Dartmouth-Hitchcock Medical Center (Lebanon, NH). Tissues used in this study were distal to the sites of pathology and were determined to be unaffected with disease upon inspection by a trained pathologist. Pathologists also determined the stage in the cycle of the premenopausal patients. Tissues were transported from Pathology on ice and procedures to prepare purified epithelial sheets began within 2 h of surgery. Approval to use tissues was previously obtained from the Committee for the Protection of Human Subjects (CPHS). Menstrual status had no influence on the experiments done in this study.
Isolation of UEC
Epithelial cells were isolated as previously described (78, 79). Briefly, tissues were minced under sterile conditions into 1- to 2-mm fragments and subjected to enzymatic digestion using a “PHC” enzyme mixture that contained final concentrations of 3.4 mg/ml pancreatin (Invitrogen Life Technologies), 0.1 mg/ml hyaluronidase (Worthington Biochemical), 1.6 mg/ml collagenase (Worthington Biochemical), and 2 mg/ml d-glucose, in 1× HBSS (Invitrogen Life Technologies) containing 50 U/ml penicillin and 50 mg/ml streptomycin. Enzymes were chosen to maximize digestion of the extracellular matrix while minimizing digestion of cell surface Ags. After incubating in PHC-HBSS for 1 h at 37°C, cells were dispersed through a 250-μm mesh screen, washed, resuspended in DMEM/F12 complete medium, and analyzed for cell number and viability. Complete medium was supplemented with 20 mM HEPES, 50 U/ml penicillin, 50 mg/ml streptomycin, 1 mg/ml fungizone, 2 mM l-glutamine (all from Invitrogen Life Technologies), and 10% defined FBS (HyClone) and did not contain phenol red.
Epithelial cell sheets were separated from stromal cells by filtration through a 20-μm nylon mesh filter (Small Parts). Epithelial sheets were retained on the 20-μm filter, while stromal cells passed through the filter. Epithelial sheets were recovered by washing and backwashing the filter with complete medium. Epithelial sheets were collected, centrifuged at 500 × g for 10 min, and resuspended in a small volume of complete medium.
Cell culture
To establish a cell culture system of polarized human UEC with both apical and basolateral compartments, the human UEC were cultured in Human Extracellular Matrix (Collaborative Biomedical Products) coated Falcon cell culture inserts in 24-well culture dishes designed for these cell inserts (Fisher Scientific). For these experiments, apical and basolateral compartments had 300 and 850 μl of complete medium, respectively. The medium was changed every 2 days. Following a 24-h incubation with various TLR agonists: ultra pure LPS from Salmonella minnesota (List Biological Laboratories) used at a final concentration of 1 μg/ml; Pam3cys-Ser-(Lys)4 (EMC Microcollections), 1 μg/ml; poly(I:C) (Invivogen), 25 μg/ml; Flagellin from Escherichia coli (Inotek Pharmaceuticals), 100 ng/ml; Zymosan from Saccharomyces cerevisiae (Invivogen), 10 μg/ml; Peptidoglycan from Staphylococcus aureus (Invivogen), 10 μg/ml: CpG oligonucleotide (Invivogen), 1 μM, UEC apical and basolateral supernatants were centrifuged for 5 min at 10,000 × g. For Ab-blocking experiments, UEC were pretreated with anti-TLR3 mAb (clone TLR3.7) (Cell Sciences) or an IgG1 isotype control (clone MOPC-21) (BD Pharmingen) at a final concentration of 20 μg/ml for 1 h at 37°C then stimulated with poly(I:C) for 24 h.
Measurement of transepithelial resistance (TER)
As an indicator of tight junction formation of epithelial cell monolayers, TER was periodically assessed using an EVOM electrode and Voltohmmeter (World Precision Instruments), as described previously (80).
RT-PCR analysis
Total RNA was isolated from cells using TRIzol Reagent according to the manufacturer’s recommendations (Invitrogen Life Technologies) and purified with RNeasy columns (Qiagen). Two micrograms of total RNA were reverse-transcribed using the Superscript First-Strand Synthesis System for RT-PCR according to the manufacturer’s recommendations (Invitrogen Life Technologies). PCR amplification was performed using Platinum PCR Supermix (Invitrogen Life Technologies) on the PTC-100 Thermal Cycler (MJ Research) for 35 cycles using the following cycling conditions: 94°C for 1 min, followed by 35 cycles of 94°C for 30 s, 55°C for 30 s, 72°C for 1 min, and then a final extension of 72°C for 2 min. Forward and reverse primer pairs for the TLR have been previously described (26). Forward and reverse primer pairs for HBD1 were HBD1 forward (5′-GTCAGCTCAGCCTCCAAAGG-3′) AND HBD1 reverse (5′CTTCTGCGTCATTTCTTCTG-3′); primer pairs for HBD2 were HBD2F2 (5′-CCTGATGCCTCTTCCAGGTG-3′) AND HBD2 reverse (5′-GAGGGAGCCCTTTCTGAATC-3′); primer pairs for HBD3 were HBD3 forward (5′-AGCCTAGCAGCTATGAGGATC-3′) AND HBD3 reverse (5′-CTTTCTTCGGCAGCATTTTC-3′). Lack of DNA contamination in the RNA preparations was verified by PCR amplification in the absence of reverse transcription.
Multiplex cytokine assays
Cytokines were measured using Bio-Plex human cytokine multiplex kits (Bio-Rad). Calibration curves from recombinant cytokine standards were prepared with serial dilutions in the same medium as the culture supernatants (RPMI 1640 medium containing 10% FBS). High and low reference points were included to determine cytokine recovery. Standards and reference points were measured in triplicate, each sample was measured once, and blank values were subtracted from all readings. All assays were conducted directly in a 96-well filtration plate (Millipore) at room temperature and protected from light. Briefly, wells were prewet with 100 μl of PBS containing 1% BSA, then beads (5000 beads per cytokine) together with either a standard, sample, reference point or blank in a final volume of 100 μl were incubated together at room temperature for 30 min with continuous shaking. Beads were washed three times with 100 μl of PBS containing 1% BSA and 0.05% Tween 20. A mixture of biotinylated Abs (50 μl/well) was added to beads for a further 30-min incubation with continuous shaking. Beads were washed three times, then streptavidin-PE was added for 10 min. Beads were again washed three times and resuspended in 125 μl of PBS containing 1% BSA and 0.05% Tween 20. The fluorescence intensity of the beads was measured using the Bio-Plex array reader. Bio-Plex Manager software with five-parametric-curve fitting (Bio-Rad technical note 2861 at <www.Bio-Rad.com>) was used for data analysis. The lower level of detection for IFN-γ was 37.5 pg/well; GM-CSF, IL-17, and CCL2/MCP-1 was 4.7 pg/well; IL-4, G-CSF, IL-6, IL-10, IL-1β, IL-12, and CCL4/MIP-1β was 2.3 pg/well.
Measurement of cytokine/chemokine and HBD2 secretion
TaqMan real-time RT-PCR
Real-time RT-PCR was done with a two-step protocol as described previously (81). Total RNA was isolated from cells using TRIzol Reagent according to the manufacturer’s recommendations (Invitrogen Life Technologies) and purified with RNeasy columns (Qiagen). Coincident with RNA purification, was on-column DNase digestion using the RNase-Free DNase set (Qiagen). For each specimen, 600 ng of total RNA was reverse-transcribed using the iScript cDNA synthesis kit according to the manufacturer’s recommendations (Bio-Rad) in a 20-ul volume. Relative expression levels of HBD1, HBD2, and HBD3 were measured using the 5′ fluorogenic nuclease assay in real-time quantitative PCR using TaqMan chemistry on the ABI 7700 Prism real-time PCR instrument (Applied Biosystems). The HBD1, HBD2, IFN-β, 2′,5′-OAS, MxA, and CD71 primer/MGB probe sets were obtained from Applied Biosystems assays-on-demand (ID nos. Hs00608345, Hs00823638, Hs00277188, Hs00242943, Hs00182073, and Hs99999911, respectively). Primers and probe used for detection of HBD3 are as previously described (82). HBD3 primers were used at 300 nM and probe at 50 nM. PCR was conducted using the following cycle parameters: 95°C, 12 min for 1 cycle (95°C, 20 s; 60°C, 1 min), for 40 cycles. Analysis was conducted using the sequence detection software supplied with the ABI 7700. The software calculates the threshold cycle (Ct) for each reaction and this was used to quantitate the amount of starting template in the reaction. The Ct values for each set of duplicate reactions were averaged for all subsequent calculations. A difference in Ct values (ΔCt) was calculated for each gene by taking the mean Ct of gene of interest and subtracting the mean Ct for CD71 for each cDNA sample. Assuming that each reaction functions at 100% PCR efficiency, a difference of one Ct represents a 2-fold difference. Relative expression levels were expressed as a fold-increase in mRNA expression and calculated using the formula 2−ΔΔCt.
Statistics
The data are presented as the mean ± SE. A two-tailed paired t test or a one-way ANOVA with Bonferonni’s posttest was performed using GraphPad InStat version 3.0a (GraphPad Software). A p value of <0.05 was taken as indicative of statistical significance.
Results
Human UEC express TLR1–9
The expression of TLR1–10 UEC was examined using RT-PCR. UEC were isolated from hysterectomy tissue samples and grown to confluence on cell inserts coated with human extracellular matrix. The formation of tight junctions, as indicated by TER measurements of >500 ohm/well or greater (control wells had TER of 130–150 ohm/well), is an indication of UEC integrity and verification that a polarized monolayer has been formed. Once sufficient TER was achieved, total RNA was isolated and mRNA expression of TLR1–10 was examined. As seen in Fig. 1⇓, UEC express mRNA for TLR1–9, while expression of TLR10 mRNA was not observed. The presence of TLR1–9 was observed in UEC preparations from eight different patients, and in all cases TLR10 mRNA was not observed.
TLR expression in UEC by RT-PCR. Lanes 1–10 correspond to TLR1–10, respectively. Total RNA was isolated from UEC and examined by RT-PCR for TLR mRNA expression.
The TLR3 agonist poly(I:C) induces the secretion of proinflammatory cytokines/chemokines by UEC
Having demonstrated that UEC express mRNA for TLR1–9, we examined whether stimulation of various TLR agonists would induce the secretion of cytokines and/or chemokines. UEC obtained from two patients (patients 2739 and 2749) were treated both apically and basolaterally with agonists to TLR2, TLR3, TLR4, TLR5, and TLR9 for a 24-h time-period, after which conditioned apical and basolateral medium were collected and screened for the expression of several cytokines and chemokines using the Luminex multiplex system. A greater amount of protein in the conditioned apical medium would signify preferential secretion into the lumen of the uterus, while a greater amount of protein in the conditioned basolateral medium would signify preferential secretion into the stromal tissues. Only the TLR3 agonist poly(I:C) stimulated UEC, while the other TLR agonists had no effect on the secretion of the cytokines and chemokines examined (data not shown). As seen in Table I⇓, the levels of IL-6 and G-CSF secretion into the apical compartment exceeded the upper limit of sensitivity. Poly(I:C) induced the apical secretion of TNF-α, GM-CSF, CCL2/MCP-1, and CCL4/MIP-1β. Poly(I:C) induced the basolateral secretion of TNF-α, IL-6, GM-CSF, G-CSF, CCL2/MCP-1, and CCL4/MIP-1β (Table I⇓). Poly(I:C) had no effect on the other cytokines examined, IFN-γ, IL-4, IL-1β, IL-10, IL-12, and IL-17, either apically or basolaterally (data not shown). These results demonstrated that UEC exposed to poly(I:C) induced the apical and/or basolateral secretion of four cytokines, TNF-α, IL-6, GM-CSF, and G-CSF and two chemokines CCL2/MCP-1 and CCL4/MIP-1β.
Cytokine secretion by primary endometrial epithelial cells treated with the TLR3 agonist poly(I:C)a
Poly(I:C) induces the secretion of proinflammatory cytokines TNF-α, IL-6, G-CSF, and GM-CSF by UEC
Having used the Luminex multiplex system to screen for differentially secreted cytokines, we focused on TNF-α, IL-6, G-CSF, and GM-CSF as being induced by poly(I:C), and sought to confirm and extend these results by ELISA using UEC isolated from six patients. The baseline apical and basolateral secretion by UEC of TNF-α, IL-6, G-CSF, and GM-CSF are shown in Table II⇓. The values are presented as the mean baseline secretion (nanograms per well) for the six patients examined as well as the range of secretion observed. TNF-α, IL-6, G-CSF, and GM-CSF were preferentially secreted into the apical compartment, with IL-6 secretion being ∼3- to 6-fold greater than TNF-α, G-CSF, and GM-CSF.
Baseline cytokine and chemokine secretion by primary UECa
To more fully define the role of poly(I:C), UEC were treated for 24 h before analysis of apical and basolateral medium for the secretion of TNF-α, IL-6, G-CSF, GM-CSF by ELISA. As seen in Fig. 2⇓, exposure of UEC to the TLR3 agonist poly(I:C) significantly induced the apical and basolateral secretion of TNF-α, IL-6, G-CSF, and GM-CSF. These results demonstrate that exposure of UEC to poly(I:C) induces the apical and basolateral secretion of proinflammatory cytokines TNF-α, IL-6, G-CSF, and GM-CSF without affecting the preferential directional cytokine release.
Apical and basolateral cytokine production by UEC treated with the TLR3 ligand poly(I:C). Cultured medium was collected following 24 h poly(I:C) stimulation and analyzed for TNF-α, IL-6, G-CSF, and GM-CSF protein secretion by ELISA. Cells were treated with poly(I:C) at a final concentration of 25 μg/ml. The p values were calculated using a two-tailed paired t test. ∗, Significantly different (p < 0.05) from control. ∗∗, Significantly different (p < 0.01) from control. (Results are mean of six patients).
Poly(I:C) induces the secretion of chemokines CCL2/MCP-1, CCL4/MIP-1β, and CXCL8/IL-8 by UEC
The Luminex multiplex system also implicated the secretion of the chemokines CCL2/MCP-1 and CCL4/MIP-1β as being induced upon exposure of UEC to poly(I:C). Although not included in the Luminex multiplex screening, the secretion of CXCL8/IL-8 was included in these analyses because previous work done using the human UEC line ECC-1 demonstrated that CXCL8/IL-8 secretion was induced upon exposure of these cells to agonists TLR2 and TLR5 (26). Using identical supernatants that were examined for cytokine secretion, the apical and basolateral release of CCL2/MCP-1, CCL4/MIP-1β, and CXCL8/IL-8 was examined by ELISA. The baseline apical and basolateral secretion by UEC for CCL2/MCP-1, CCL4/MIP-1β, and CXCL8/IL-8 are shown in Table II⇑. CCL2/MCP-1, CCL4/MIP-1β, and CXCL8/IL-8 were preferentially secreted into the apical compartment, with the secretion levels of CCL2/MCP-1 and CXCL8/IL-8 exceeding those of TNF-α, IL-6, G-CSF, and GM-CSF. For example, the apical secretion of CXCL8/IL-8 was >24 ng/well, which was nearly one to two logs greater than the secretion of TNF-α, IL-6, G-CSF, GM-CSF (Table II⇑). As seen in Fig. 3⇓, exposure of UEC to poly(I:C) significantly induced the apical and basolateral secretion of CCL2/MCP-1, CXCL8/IL-8, and CCL4/MIP-1β. These results demonstrate that exposure of UEC with the TLR3 agonist poly(I:C) induces the apical and/or basolateral secretion of the chemokines CCL2/MCP-1, CCL4/MIP-1β, and CXCL8/IL-8.
Chemokine production by UEC treated with the TLR3 ligand poly(I:C). Cultured medium was collected following 24 h poly(I:C) stimulation and analyzed for the presence of (A) CCL2/MCP-1, (B) CXCL8/IL-8, and (C) CCL4/MIP-1β protein expression by ELISA. The p values were calculated using a two-tailed paired t test. ∗∗, Significantly different (p < 0.01) from control. (Results are mean of six patients).
HBD1 and HBD2 mRNA expression is induced by UEC exposed to poly(I:C)
Because exposure of bronchial epithelial cells to poly(I:C) has previously been shown to induce β-defensin expression (49), we sought to determine whether exposure of UEC to poly(I:C) had any effect on the expression of HBD1, HBD2, and HBD3 (HBD1–3). Coincident with removal of apical and basolateral supernatants in previous experiments was the isolation of total RNA from the UEC. To determine whether UEC constitutively express HBD1–3, RT-PCR was performed. UEC were observed to express HBD1–3 mRNA (Fig. 4⇓A). Using real-time RT-PCR, we sought to determine whether poly(I:C) had an effect on the mRNA expression of HBD1–3. As shown in Fig. 4⇓B, exposure of UEC to poly(I:C) significantly induced the mRNA expression of HBD1 and HBD2 5.9- and 4.0-fold, respectively. To verify that protein secretion was also induced by poly(I:C), HBD2 secretion into the apical and basolateral supernatants from four patients was examined by ELISA. Stimulation of UEC with poly(I:C) resulted in a 45% increase in apical release of HBD2 protein (control = 42.2 ± 0.67 pg, poly(I:C) = 61.3 ± 3.8 pg; p < 0.05), but had no effect on HBD2 secretion into the basolateral compartment. Poly(I:C) had no effect on HBD3 (data not shown). Agonists to TLR2, TLR4, TLR5, and TLR9 had no effect on the expression of HBD mRNA.
Constitutive HBD mRNA expression in UEC and following treatment with the TLR3 agonist poly(I:C). A, Lane 3 corresponds to HBD1–3, respectively. Total RNA was isolated from UEC and examined by RT-PCR for HBD mRNA expression. B, Real-time RT-PCR was used to determine the relative levels of expression of HBD1 and HBD2 mRNA, normalized against an endogenous control, CD71. The data were further normalized by using values from the control for calibration. The p values were calculated using a two-tailed paired t test. ∗, Significantly different (p < 0.05) from control. ∗∗, Significantly different (p < 0.01) from control. (Results are mean of four patients).
Exposure of UEC to poly(I:C) induces mRNA expression of IFN-β and IFN-β-stimulated antiviral genes, MxA and 2′,5′-OAS
Because TLR3 binds and responds to poly(I:C), a viral dsRNA analog, we undertook to determine whether the mRNA expression of the antiviral IFN, IFN-β, was induced by poly(I:C) when added to UEC. In addition, the mRNA expression of the IFN-β-stimulated antiviral genes MxA and 2′,5′-OAS were examined. As determined by real-time RT-PCR, exposure of UEC to poly(I:C) induced IFN-β mRNA expression 70-fold and the mRNA expression of MxA and 2′,5′-OAS 107- and 96-fold, respectively (Fig. 5⇓). These results demonstrate that poly(I:C) not only induces secretion of proinflammatory cytokines and chemokines and expression of β-defensins by UEC, but also induces the expression of IFN-β and the antiviral genes MxA and 2′,5′-OAS, which are important mediators of an antiviral response. Agonists to TLR2, TLR4, TLR5, and TLR9 had no effect on the expression of IFN-β, MxA, or 2′,5′-OAS mRNA.
Expression of antiviral genes by UEC treated with the TLR3 ligand poly(I:C). Real-time RT-PCR was used to determine the relative levels of expression of (A) IFN-β, (B) MxA, and 2′,5′-OAS, normalized against an endogenous control, CD71. The data were further normalized by using values from the control for calibration. The p values were calculated using a two-tailed paired t test. ∗∗, Significantly different (p < 0.01) from control. (Results are mean of four patients).
Preincubation of UEC with anti-TLR3 mAb partially inhibits TNF-α and CCL2/MCP-1 secretion
To determine whether TLR3 expressed at the cell surface was involved in this response, UEC were preincubated with an anti-TLR3 mAb or isotype control. As shown in Fig. 6⇓, incubation of UEC with an anti-TLR3 mAb for 1 h before the addition of poly(I:C) significantly inhibited the apical and basolateral secretion of TNF-α and CCL2/MCP-1 relative to that seen with isotype control. Inhibition of IL-6, G-CSF, GM-CSF, CCL4/MIP-1β, and CXCL8/IL-8 apical and basolateral secretion were also observed, albeit to varying degrees (data not shown). In contrast, preincubation of UEC with anti-TLR3 mAb, before poly(I:C) stimulation, had no effect on HBD1, HBD2, IFN-β, MxA, or 2′,5′-OAS mRNA expression (Fig. 6⇓C). These results demonstrate that blocking TLR3 expressed at the cell surface selectively inhibits the secretion of cytokines and chemokines in response to poly(I:C), but had no effect on the transcription of β-defensin, IFN-β, and IFN-β-stimulated gene mRNA.
Pretreatment of UEC with an anti-TLR3 mAb inhibits poly(I:C)-mediated secretion of TNF-α and CCL2/MCP-1 and has no effect on IFN-β, 2′,5′-OAS, MxA, or β-defensin mRNA. UEC were pretreated with anti-TLR3 mAb or isotype mAb for 1 h and then stimulated with poly(I:C) for 24 h (four inserts/treatment group). Conditioned apical and basolateral medium were then analyzed for the presence of (A) TNF-α or (B) CCL2/MCP-1 protein expression by ELISA. C, Real-time RT-PCR was used to determine the relative levels of expression of IFN-β, MxA, 2′,5′-OAS, HBD1, and HBD2, normalized against an endogenous control, CD71. The data were further normalized by using values from the control for calibration. NS, Not significantly different. The p values were calculated using an ANOVA with Bonferonni posttest. ∗, Significantly different (p < 0.05) from control. ∗∗, Significantly different (p < 0.01) from control. (Representative of two patients).
Discussion
This study demonstrates that UEC express TLR1–9 and that only the TLR3 agonist poly(I:C) stimulates UEC to secrete several cytokines and chemokines, particularly those factors involved in proinflammatory responses. Furthermore, we show that TLR3 stimulation of UEC induces the mRNA expression of the antimicrobial peptides HBD1 and HBD2. Finally, TLR3-mediated stimulation of UEC induced the mRNA expression of IFN-β and the IFN-β-stimulated antiviral genes MxA and 2′,5′-OAS. To our knowledge, this is the first demonstration of an antiviral response generated by primary human UEC.
Once thought to be a sterile environment, the uterus is now known to be periodically exposed to organisms colonizing the vagina (83). Furthermore, the demonstration that macrospheres deposited in the cervix move through the cervix and uterus into the Fallopian tubes within minutes of deposition is supportive evidence that bacteria and viruses in the vagina and cervix routinely reach the uterus (84). Considering the importance of the uterus with respect to pregnancy and fertility, expression of a large repertoire of TLR by UEC would allow for an enhanced sensitivity to a wider range of foreign microbes to rapidly and efficiently initiate innate and adaptive immune responses. Previous studies from our laboratory have shown that mRNA from a human UEC line express TLR1–9 and that TLR expression varies throughout tissues of the FRT (26, 85). Consistent with our previous findings, we demonstrate here that UEC also express TLR1–9. Neither ECC-1 cells nor UEC express TLR10. Others have recently reported mRNA expression of TLR1–6 and TLR9 by primary UEC and the absence of TLR10 expression (86). In contrast, we demonstrate that TLR7 and TLR8 are also expressed by UEC.
Poly(I:C) is a synthetic mimic of viral dsRNA that binds TLR3. Our findings demonstrate TLR3-mediated stimulation of UEC, via poly(I:C), results in enhanced secretion of the proinflammatory cytokines and chemokines TNF-α, IL-6, G-CSF, GM-CSF, CXCL8/IL-8, CCL2/MCP-1, and CCL4/MIP-1β. We previously reported that the adenocarcinoma-derived UEC line ECC-1 was not stimulated by poly(I:C) (26). This difference is likely due to greatly reduced TLR3 expression by ECC-1 cells compared with UEC (our unpublished observation). Secretion of the CXCL8/IL-8, CCL2/MCP-1, and CCL4/MIP-1β are known to have profound effects on the recruitment of immune cells, such as macrophages, monocytes, DC, neutrophils, and T lymphocytes (57, 87, 88, 89, 90). In concert with chemokine-mediated recruitment, TNF-α, IL-6, G-CSF, and GM-CSF induce B cell responses and enhance T cell help (91, 92, 93, 94, 95), stimulate the proliferation, differentiation, maturation, and antimicrobial functions of neutrophils (96, 97), as well as augment the functions of monocytes, macrophages, and DC (98). Therefore, in response to an invading viral pathogen, UEC have the capability of recruiting and activating a variety of immune cells that are known to be essential in preventing viral infection and restricting viral replication until an adaptive immune response is generated. Additionally, baseline secretion of these cytokines and chemokines, in the absence of poly(I:C) stimulation, suggests that UEC may play an important role in maintaining immune readiness/surveillance in the absence of pathogens. Previously, we and others have shown that CD45 leukocytes are present in the uterine endometrium throughout the menstrual cycle (99). Our findings in the present study suggest that baseline secretion levels of CXCL8/IL-8, CCL2/MCP-1, and IL-6 by UEC play an important role in leukocyte migration to the uterus to both detect as well as protect against potential pathogens.
In addition to the induction of proinflammatory cytokines and chemokines, UEC in response to poly(I:C) express 6- and 4-fold higher levels of HBD1 and HBD2, respectively. To the best of our knowledge, this study is the first demonstration that UEC expression of β-defensins increase in response to the TLR3 agonist poly(I:C). Our finding that HBD1 is induced by UEC following poly(I:C) treatment was unexpected, because HBD1 is thought to be constitutively expressed and unaffected by external stimuli (100). HBD1 and HBD2 are antimicrobial peptides that are effective against nonviral agents such as Escherichia coli, Pseudomonas aeruginosa, and Staphylococcus aureus as well as exhibit antiviral activity toward adenovirus, HSV, rhinovirus, and HIV (101, 102, 103, 104, 105). Both HBD1 and HBD2 facilitate the recruitment of immune cells, such as memory T lymphocytes, iDC and mast cells (44). Although HBD1 and HBD2 have been shown to be effective at inhibiting replication of HIV, HSV-1, and HSV-2, it is unclear whether they would have an inhibitory effect on other viral pathogens of the FRT, such as EBV, hepatitis B virus, hepatitis C virus, CMV, and human papillomavirus (106, 107, 108, 109, 110, 111, 112, 113). That HBD1 mRNA expression is induced by the TLR3 agonist poly(I:C) in UEC suggests that, in addition to its known antimicrobial function, HBD1 may have a more important role in antiviral innate immunity.
We and others have previously shown that secretions from throughout the FRT exhibit both bactericidal and viricidal activity (114, 115). Acting through secreted products including secretory leukocyte protease inhibitor, β-defensins, etc., epithelial cells of the uterus have the potential to interfere with HIV as well as other viruses that are both cell-free and cell-associated in the ejaculate (116, 117, 118, 119, 120). Other investigators have previously reported the antiviral action of IFNs (reviewed in Ref.54). Building on these important observations, that cells are capable of intracellular viral protection, we found that poly(I:C) induced UEC to express high levels of IFN-β and IFN-β-stimulated genes MxA and 2′,5′-OAS. It is known that TLR3-mediated immune responses, induced by dsRNA or viral infection, result in the rapid production of IFN-β and IFN-β-stimulated genes, such as MxA and 2′,5′-OAS (121, 122, 123). Using a pulmonary epithelial cell line, others have shown that the expression of TLR3 is induced by poly(I:C) (122). We have recently confirmed this finding with UEC exposed to poly(I:C) (our unpublished observation). The present study demonstrates that UEC express IFN-β in response to the viral analog poly(I:C). To our knowledge, these studies are the first demonstration that human UEC express IFN-β, as well as antiviral molecules MxA and 2′,5′-OAS, in response to the viral dsRNA analog poly(I:C), and that this response is mediated in a TLR3-dependent manner. Beyond the induction of IFN-β and the expression of antiviral molecules, such as MxA and 2′,5′-OAS, the induction of IFN-β by UEC may have important effects on leukocytes present at or recruited to the site of viral exposure/infection. For example, IFN-β has been shown to enhance cytotoxic activity and the production of IFN-γ by NK cells, stimulate the maturation of DCs and differentiation of monocytes into DCs, rescue memory T cells from apoptosis, enhance T lymphocyte proliferation and Th1 responses, as well as up-regulate iNOS in monocytes and macrophages (124, 125, 126, 127, 128, 129, 130). Our studies suggest that by secreting IFN-β in response to viral challenge, epithelial cells act as sentinels at multiple levels to prevent and/or control viral infection
To determine whether TLR3 expressed on the cell surface is responsible for binding poly(I:C), UEC were preincubated with anti-TLR3 mAb before stimulation with poly(I:C). We found that poly(I:C)-mediated secretion of cytokines and chemokines was partially inhibited by preincubation with an anti-TLR3 mAb. These findings suggest that TLR3 expressed on the cell surface participates in the recognition of dsRNA and triggers downstream signals leading to cytokine and chemokine production. Others have shown that TLR3 is found both on the cell surface as well as within some cells (124). Moreover, neutralization of cell surface TLR3 has been shown to partially inhibit (50%) the secretion of IFN-β (121). Our findings of 25–50% inhibition of TNF-α and MCP-1 secretion by epithelial cells in response to poly(I:C) (Fig. 6⇑) are consistent with these observations. Unexpectedly, we found that preincubation of UEC with anti-TLR3 mAb had no effect on poly(I:C) induced expression of HBD1, HBD2, IFN-β, MxA, or 2′,5′-OAS when exposed to poly(I:C). The differences seen between cytokine and chemokine secretion vs mRNA expression of HBD1, HBD2, IFN-β, MxA, and 2′,5′-OAS may be due to the differential kinetics of protein synthesis and mRNA transcription over a 24-h time period. An alternative interpretation, suggested from our studies, is that in UEC stimulation through cell surface TLR3 is linked to a signaling pathway that is distinct from TLR3 expressed in subcellular compartments. Others have shown that in response to stimulation of cell surface TLR3, the MyD88-dependent signaling pathway, which is known to act through NF-κB (131), results in the production and secretion of cytokines and chemokines. In contrast, internalized dsRNA binding to TLR3 localized in endosomal compartments (124) may favor a MyD88-independent pathway, which through the activation of IRF-3 (132), has been shown to result in the expression of IFN-β and IFN-β-stimulated genes. Considering that TLR4 is also localized at the cell surface and intracellularly (133), determining whether stimulation of TLR3 and/or TLR4 at these different locales triggers different signaling pathways may further our understanding of how TLR function as part of innate immunity.
In conclusion, our studies suggest that UEC are poised to respond to viral infection at several levels. Stimulation of UEC with dsRNA, in a TLR3-dependent manner, induces the secretion of proinflammatory cytokines and chemokines that are known to facilitate the recruitment of immune cells to the site(s) of viral infection. Concurrently, TLR3 stimulation of UEC enhances the expression and possible secretion of the antimicrobial peptides HBD1 and HBD2 that would have inhibitory effects on viral entry into susceptible cells and therefore slow viral replication. At the intracellular level, IFN-β production by UEC might induce the expression of IFN-β-stimulated genes that have potent antiviral properties. Overall, UEC appear primed and ready to respond to viral exposure and/or infection of the uterus in a TLR3-mediated manner, which, in turn, leads to the stimulation of IFN-β and IFN-β-stimulated genes that are essential for inhibiting or slowing viral replication until an effective adaptive response can be mounted. Further studies are needed to extend our understanding of how UEC are involved in innate immunity, particularly antiviral responses, which are crucial for the development of mucosal vaccines targeting pathogens of the FRT, as well as therapeutics for uterine infections and treatments for infertility.
Acknowledgments
We thank Dr. Jacqueline Smith for advice and assistance on the Luminex multiplex system; Dr. David G. Ginzinger for advice with TaqMan real-time RT-PCR; Drs. Vincent Memoli and Jorge Gonzalez (Department of Pathology), Shannon Shutz, Peter Seery, Elizabeth Rizzo, and Richard Merrill for tissue preparation; Linda Hallock and Barbara Schaeffer for clinical support, and Drs. Kris Strohbehn, Leslie Demars, Paul Hanissian, Robert Porter, Benjamin Mahlab, Paul Manganiello, William Young, Roger Young, and Richard Meter (Department of Obstetrics and Gynecology) for tissue procurement.
Footnotes
-
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
-
↵1 This work was supported by National Institutes of Health Grants AI51877 (to C.R.W.), Immunology Training Grant T32 AI07363-12, and Norris Cotton Cancer Center Grant CA23108.
-
↵2 Address correspondence and reprint requests to Dr. Todd M Schaefer, Department of Physiology, Dartmouth Medical School, 710W Borwell, 1 Medical Center Drive, Lebanon, NH 03756. E-mail address: todd.m.schaefer{at}dartmouth.edu
-
↵3 Abbreviations used in this paper: FRT, female reproductive tract; PAMP, pathogen-associated molecular pattern; UEC, uterine epithelial cell; HBD, human β-defensin; iDC, immature dendritic cell; 2′,5′-OAS, 2′,5′-oligoadenylate synthetase; Mx, myxovirus resistance gene; TER, transepithelial resistance; Ct, threshold cycle.
- Received July 20, 2004.
- Accepted October 22, 2004.
- Copyright © 2005 by The American Association of Immunologists