Sphingosine 1-Phosphate 1 and TLR4 Mediate IFN-β Expression in Human Gingival Epithelial Cells

Mehmet A. Eskan, Beate G. Rose, Manjunatha R. Benakanakere, Menq-Jer Lee and Denis F. Kinane
J Immunol February 1, 2008, 180 (3) 1818-1825; DOI: https://doi.org/10.4049/jimmunol.180.3.1818

Abstract

IFN-β production is a critical step in human innate immune responses and is primarily controlled at the transcription level by highly ordered mechanisms. IFN-β can be induced by pattern-recognition receptors such as the TLR4. S1P1 is a G protein-coupled receptor, which has a high affinity for sphingosine 1-phosphate (S1P). Although many of the receptors and signaling pathways leading to the expression of IFN-β have been identified and characterized, it is still unclear how IFN-β is regulated in primary human gingival epithelial cells (HGECs). In this study, we demonstrate that S1P1 and TLR4, acting in unison, play an important role in IFN-β expression at the protein and mRNA level in HGECs. We demonstrate that the expression of both IFN-β and IFN-inducible protein-10 (CXCL-10) is significantly up-regulated by LPS and S1P or LPS and a specific S1P1 agonist. This enhanced innate immune response is attenuated in HGECs by small interfering RNA knockdown of either TLR4 or S1P1. Moreover, we show that triggering of TLR4 results in the increased expression of S1P1 receptors. Furthermore, we found that IFN-regulatory factor 3 activation was maximized by LPS and S1P through PI3K. Our data show that triggering TLR4 increases S1P1, such that both TLR4 and S1P1 acting through PI3K enhancement of IFN-regulatory factor 3 activation increase IFN-β expression in epithelial cells. The functional association between TLR4 and the S1P1 receptor demonstrates a novel mechanism in the regulation of IFN-β and CXCL-10 in human primary gingival epithelial cells.

Oral mucosal epithelium, in addition to acting as a physiological barrier, plays a crucial role in initiating and augmenting oral defense mechanisms by releasing a variety of inflammatory molecules in response to oral microorganisms (1, 2). Recognition of microorganisms is performed by pattern-recognition receptors that recognize highly conserved molecular patterns (microbial-associated molecular patterns) (3) such as LPS of Gram-negative bacteria. This specificity allows the TLRs to detect the presence of infection and induce inflammatory, chemotactic, and antimicrobial innate immune responses. TLR4 was the first characterized TLR in humans (4) and is the receptor for bacterial LPS (5).

IFN-α and IFN-β are the predominant classes of type I IFNs (6), and both are induced by nucleic acids (dsRNA); however, IFN-β can be directly induced by nonnucleic acids (LPS) (7). Type I IFNs are multifunctional cytokines that control many different cellular functions, including antiviral, antitumor, and immunomodulatory activity (8). IFN-β has been demonstrated to control immune cell trafficking to the site of infection by mediating chemokine induction (9, 10), and also plays a role in the differentiation of CTL (8). IFN-β has an anti-inflammatory effect by inhibiting TNF-α production (11). CXCL-10 is a member of the CXC chemokine family induced by IFN-β (12), and not only regulates immune cells during inflammation through its receptor CXCR3 (13), but also reduces inflammation (14).

Evidence is accumulating that TLR signaling pathways and G protein-coupled receptor (GPCR)3 pathways may modulate each other (15). S1P1 is a GPCR and has high affinity for sphingosine 1-phosphate (S1P) (16). S1P is a bioactive sphingolipid released predominantly by activated platelets and acts on other cells through their S1P receptors to promote proliferation and healing responses (16, 17). S1P binds to all five members of the S1P family of receptors, which are numbered S1P 1 to 5 (18). S1P1 signals through the Gi/O proteins; however, S1P2 and S1P3 stimulate Gq and G12/13 (19), and S1P5 activates both G12/13 and Gi (20). S1P has been extensively studied for its healing and repair cell functions, such as cell migration (21) and tightening (22), but its role in inflammation is largely unknown particularly in epithelial cells. LPS-induced IFN-β expression (23) has been extensively studied in mammalian cells, including dendritic cells and macrophages (7); however, a detailed investigation of the regulation of IFN-β in epithelial cells is lacking. Epithelial cells can regulate inflammatory and immune cell recruitment by induction of adhesion molecules, inflammatory cytokines, and chemokines (1). In this study, we show that IFN-β and the related CXCL-10 were enhanced by LPS and S1P (or 5-(4-phenyl-5-trifluoromethylthiophen-2-yl)-3-(3-trifluoromethylphenyl)-(1,2,4)-oxadiazole (SEW), a specific S1P1 agonist) and that S1P1 expression was increased in human gingival epithelial cells (HGECs) following LPS treatment. Both effects were abrogated by gene silencing of either TLR4 or S1P1.

Expression of IFN-β is highly controlled by several molecules, including IFN-regulatory factor 3 (IRF3) (24). IRF3 undergoes serine phosphorylation of its C-terminal region, causing IRF3 to dimerize and translocate to the nucleus (24) so that it can function as a promoter and signal-specific cofactor to initiate transcription of a set of NF-κB-dependent genes (25). It has been demonstrated that IRF3 is a major activator of the IFN-β (26). More interestingly, it has been reported that IRF3 plays a crucial role in the induction of the IFN-β gene in response to LPS in dendritic cells (23) and IRF3 activation can be enhanced by PI3K (27), which can be activated by S1P1 (21). Our results show that IRF3 phosphorylation (serine 398) was maximized through PI3K in the cells challenged with LPS in the presence of S1P. We also found that PI3K and IRF3 play important roles in the regulation of IFN-β expression in human gingival epithelial cells.

Hence, this novel regulatory mechanism for the induction of IFN-β and CXCL-10 in HGECs requires both TLR4 and S1P1 receptor signaling. This is the first demonstration of the intimate cooperation between TLR4 and the S1P1 receptor in IFN-β induction, whose purpose may be to enhance oral mucosal defense mechanisms by triggering the inflammatory defensive and chemotactic responses following microbial insult.

Materials and Methods

Cell isolation and culture

HGECs were obtained with Institutional Review Board approval from healthy patients after third molar extraction. The gingiva was treated with 0.025% trypsin and 0.01% EDTA overnight at 4°C, and HGECs were isolated, as previously described (1). Briefly, the cell suspension was centrifuged at 120 × g for 5 min, and the pellet was suspended in K-SFM medium (Invitrogen Life Technologies) containing 10 μg/ml insulin, 5 μg/ml transferrin, 10 μM 2-ME, 10 μM 2-aminoethanol, 10 mM sodium selenite, 50 μg/ml bovine pituitary extract, 100 U/ml penicillin/streptomycin, and 50 ng/ml fungizone (complete medium). The cells were seeded in 60-mm plastic tissue culture plates coated with type-I collagen, and incubated in 5% CO2 and 95% air at 37°C. When the cells reached subconfluence, they were harvested and subcultured, as described (2).

Cell challenge assays

HGECs at the fourth passage were harvested, seeded at a density of 0.5 × 105 cells/6-well culture plate coated with type-I collagen, and maintained in 2 ml of medium. When they reached confluence, the cells were washed twice with fresh medium, and 1 ml of complete medium was added. The cells were challenged with Escherichia coli LPS (1 μg/ml; InvivoGen), S1P (100 nM, Biomol), SEW (3 μM; BIOMOL), or combination of LPS and S1P at the same concentrations (LPS/S1P). Production of IFN-β and CXCL-10 was determined by ELISA (PBL Biomedical Laboratories) and Luminex100 technology (Upstate Biotechnology), respectively. To examine the effect of IFN-inducible genes, the cells were pretreated with Jak2 inhibitor, AG490 (5 μM; Calbiochem), or Jak1 inhibitor (1 nM; Calbiochem) for 1.5 h before LPS and S1P stimulation. Alternatively, the cells were incubated with human rIFN-β (50 U/ml; R&D Systems) in the presence of neutralizing anti-human IFN-β (1 μg/ml; R&D Systems) or its isotype control, goat-IgG (R&D Systems), for 24 h. To block S1P1 signaling, cells were pretreated with pertussis toxin (PTx, 100 ng/ml; Calbiochem), an inhibitor of Gi/o heterotrimeric G protein, for 2 h. Subsequently, the cell medium was replaced with fresh medium and the cells were challenged with LPS/S1P in the presence of 10 ng/ml PTx for 24 h. Furthermore, to examine the PI3K activity, the cells were treated with 50 μM LY294002 (Calbiochem) 1 h before challenge with LPS/S1P for 24 h. To inhibit S1P2 signaling, the cells were treated with an S1P2-specific antagonist, JTE-013 (0.5 μM; Tocris Bioscience), then challenged with S1P for 24 h.

Real-time PCR

Total RNA was extracted from cultured cells by using TRIzol reagent (Invitrogen Life Technologies). Ten micrograms of RNA samples were used to perform first-strand cDNA synthesis using the High-Capacity cDNA Archive kit (Applied Biosystems). Real-time PCR was performed by using 50 ng of cDNA with an ABI 7500 system (Applied Biosystems) in the presence of Taqman DNA polymerase. The sense and antisense primers used to detect the gene expression of human S1P1, S1P2, IFN-β, CXCL-10, IRF3, TLR4, and GAPDH were purchased from Applied Biosystems. The quantitative PCR (qPCR) was performed by using a universal PCR Master Mix (Applied Biosystems), according to the manufacturer’s instruction.

Silencing TLR4, S1P1, or IRF3

HGECs at the fourth passage were harvested, seeded at a density of 0.5 × 105 cells/6-well culture plate coated with type I collagen, and maintained in 2 ml of medium until they reached 50–70% confluence. Subsequently, HGECs were transfected with 100 pmol of small interfering TLF4 (siTLR4; Dharmacon), siRNA to laminin (Dharmacon), 100 nM siIRF3 (Dharmacon), or 1 μg of lentiviral vector carrying siRNA to S1P1 (28) using the electroporation technique (Amaxa). Briefly, when the cells reached 70% confluence, they were collected and then 1 × 106 cells were incubated in 100 μl of Human Keratinocyte Nucleofector Solution containing siRNA to IRF3, S1P1, TLR4, or laminin for 15 min at room temperature. The transfection reaction was performed with the Nucleofector program T23 (Amaxa). The following day, the cell medium was replaced with fresh medium and the challenge assay was performed 48 h after the transfection.

Immunohistochemistry

The cells were seeded into collagen-coated glass chamber slides (Lab-Tek II Chamber Slide; Nalge Nunc International). At 100% confluence, the cells were treated with LPS (1 μg/ml; Invitrogen Life Technologies) for 24 h in 5% CO2 and 95% air at 37°C. The cells were fixed with 4% paraformaldehyde and stained with anti-human S1P1 (a gift from C.-Y. Lin, Georgetown University, Washington D.C.) overnight at 4°C, followed by Alexa Fluor anti-mouse IgG in 3% BSA (1/500; Invitrogen Life Technologies) for 1 h at room temperature. The stained images were visualized and photographed using a confocal laser scanning microscope (FV500; Olympus).

Western blot analysis

The cells were treated with LPS, S1P, LPS/S1P, or medium for 1 h. In addition, the cells were challenged with LPS/S1P after blocking PI3K with LY294002 (50 μM; Calbiochem). Subsequently, the collected cells were lysed on ice for 30 min with radioimmunoprecipitation assay buffer (Sigma-Aldrich) containing protease (Roche) and phosphatase inhibitor mixture (Sigma-Aldrich). Cell debris was pelleted by centrifugation, and the supernatants were collected and stored at −80°C until assayed. The cellular extract was suspended in lithium dodecyl sulfate (Invitrogen Life Technologies) buffer and reducing reagent (Invitrogen Life Technologies) and heated for 10 min at 70°C. The samples were resolved by 4–12% lithium dodecyl sulfate-PAGE (Invitrogen Life Technologies), and then transferred to polyvinylidene difluoride membranes using the NuPAGE system (Invitrogen Life Technologies). Primary Abs, phospho-IRF3 (serine 398; Upstate Biotechnology), or β-actin (Cell Signaling Technology) was used at 1/1000, and secondary Ab, anti-rabbit IgG (Cell Signaling Technology), was used at 1/2000 dilution after blocking the membrane with 5% nonfat milk. Membranes were incubated with primary Ab overnight at 4°C and then with HRP-conjugated secondary Ab for 1 h at room temperature, followed by four washes of 5 min each. The same membrane was stripped with Western blot stripping solution (Pierce) following the company’s protocol and reprobed with control Ab, β-actin. Protein bands were detected using the chemiluminescence kit (Pierce). Densitometer scans of the blots were performed using the Kodak Image 4000M.

Results

IFN-β is up-regulated by LPS and S1P

We examined the expression of IFN-β in human gingival epithelial cells in response to LPS, S1P, or LPS/S1P stimulation. Cells were treated with LPS in the presence or absence of S1P for 24 h. Surprisingly, the costimulation of LPS and S1P significantly induced the expression of IFN-β at the gene (Fig. 1⇓A) and protein level (Fig. 1⇓B) in the primary cultures of human gingival epithelial cells. LPS treatment did not increase either the transcription or production of IFN-β. However, S1P appears to reduce the IFN-β mRNA expression, but not IFN-β protein (Fig. 1⇓, A and B). Furthermore, we examined the precise role of TLR4 in the production of IFN-β following the concurrent challenge of LPS and S1P. HGECs were challenged with LPS plus S1P following TLR4 knockdown by RNA interference (RNAi) technology. Initially, we determined the silencing efficiency of TLR4 following RNAi. TLR4 expression was dramatically down-regulated in cells following LPS/S1P challenge (Fig. 1⇓C). Similarly, gene silencing of TLR4 was capable of attenuating the increase of IFN-β expression in cells following LPS/S1P challenge (Fig. 1⇓D). Next, we examined the reason for the reduction in IFN-β expression at the transcription level by S1P alone, as shown in Fig. 1⇓A. It has been demonstrated that the S1P2 signal typically has a negative biological effect in mammalian cells in terms of angiogenesis and morphogenesis compared with S1P1 (29). S1P1 and S1P2 appear to have opposite effects, and yet both are triggered by extracellular S1P; thus, the cellular response would appear to be dependent on the relative concentrations of these receptors. To examine S1P2 effect on IFN-β transcription level, we initially determined S1P2 expression in S1P-treated cells. We found that there was a greater increase in S1P2 expression in S1P-treated cells compared with LPS-treated cells. However, S1P1 expression was increased more in LPS-challenged cells than S1P-challenged cells (Fig. 1⇓E). Therefore, we assessed the IFN-β expression in cells challenged with S1P after blocking S1P2 signaling with JTE-013, a S1P2-specific antagonist (30). Initially, we determined the optimal dose of JTE-013. We found that 1, 2, 5, or 10 μM JTE-013 had toxic effects, but that 0.5 μM JTE-013 was optimal (data not shown). We observed that IFN-β expression, initially reduced by S1P, was restored when we used JTE-013 (Fig. 1⇓F). Together these data show clearly the novel cross-talk mechanism between TLR4 and S1P1 receptors, which is needed for the expression of IFN-β at the transcriptional and protein level. Our data also demonstrate that S1P2 signaling regulates negatively the transcription of IFN-β in the cells challenged with S1P.

FIGURE 1.
FIGURE 1.

LPS and S1P resulted in induced IFN-β expression. A, Human gingival epithelial cells were challenged with E. coli LPS (1 μg/ml), S1P (100 nM), or LPS/S1P for 24 h at 37°C. After treatment with LPS, S1P, or their combination (x-axis), mRNA was extracted and IFN-β and GAPDH expression was determined by qPCR. B, The expression of IFN-β at the protein level was determined by ELISA. C, Values are plotted as the ratio between TLR4 and GAPDH. Real-time PCR was performed with Taqman probes. TLR4 expression in HGECs was down-regulated by RNAi. D, After 48 h of silencing, the cells were challenged with LPS in the presence of S1P for 24 h. The expression of IFN-β mRNA was determined following the silencing, as in A. E, Total RNA was extracted from S1P-treated cells, and S1P1 or S1P2 gene expression level was determined by qPCR after normalizing their expression to their own GAPDH. F, Following pretreatment with JTE-013 (0.5 μM), HGECs were challenged with S1P and IFN-β expression was determined by qPCR. Data presented are the mean ± SD of triplicate determinations. The statistically significant induction or reduction (p < 0.05) of IFN-β expression is indicated by two asterisks and one asterisk, respectively.

S1P1 plays an important role in IFN-β expression

Our results (vide supra) show that members of the S1P receptor family, including S1P1 and S1P2, which can be activated by S1P (31), are expressed in gingival epithelial cells. Thus, we hypothesized that S1P1 gene and protein expression, which can be induced by LPS, is also involved in IFN-β expression in the presence of LPS. To test this hypothesis, we knocked down S1P1 expression using the RNAi technique and challenged the cells with LPS/S1P. We successfully reduced S1P1 expression at the gene (Fig. 2⇓A) and protein level (28). The IFN-β mRNA expression was induced in the cells treated with LPS/S1P after scramble gene silencing; however, its expression was down-regulated in the cells challenged with LPS/S1P after siRNA to S1P1 (Fig. 2⇓B). Moreover, it has been demonstrated that S1P stimulates the S1P1 signaling pathway through the G0/i protein, which is PTx sensitive (32, 33). Thus, the level of IFN-β expression was reduced in cells challenged with LPS/S1P after S1P1 signaling was blocked with PTx (Fig. 2⇓C). More importantly, LPS markedly increased IFN-β mRNA expression in the presence of SEW, which is a S1P1-specific agonist; however, neither SEW nor LPS alone resulted in an IFN-β mRNA expression increase (Fig. 2⇓D). Taken together, our results clearly show that S1P1 specifically cooperates with TLR4 to induce the expression of IFN-β in human epithelial cells.

FIGURE 2.
FIGURE 2.

S1P1 plays an important role in the IFN-β gene expression. A, Values are plotted as the ratio between S1P1 and GAPDH. Real-time PCR was performed with Taqman probes. S1P1 expression in HGECs was down-regulated by RNAi. B, Human gingival epithelial cells were challenged with E. coli LPS (1 μg/ml) and S1P (100 nM) for 24 h at 37°C following LPS/S1P challenge. After 48 h of silencing, the cells were challenged with LPS in the presence of S1P for 24 h. After treatment with a combination of LPS and S1P (x-axis), mRNA was extracted and IFN-β and GAPDH expression was determined by qPCR. C, The expression of IFN-β was determined by qPCR, as in B. The cells were challenged with LPS/S1P after blocking Go/i with pertussis toxin. D, Epithelial cells were challenged with SEW (3 μM), LPS, or their combination for 24 h. Following the treatment, values are plotted as the ratio of IFN-β and GAPDH. Real-time PCR was performed with Taqman probes. Data presented are the mean ± SD of triplicate determinations. Statistically significant induction or reduction (p < 0.05) of IFN-β or S1P1 expression is indicated by two asterisks or one asterisk, respectively.

S1P1 is induced by LPS in epithelial cells

Given our specific experimental observations, we next examined what effect the activation of the individual receptors might have on each other in HGECs. To this end, HGECs were challenged with purified E. coli LPS. Initially, we determined the optimal LPS concentration and time point for S1P1 induction. Epithelial cells were challenged with different concentrations of LPS, including 0.1, 0.2, 0.5, 1, or 2 μg/ml (data not shown). We found that 1 μg/ml LPS was the most efficient concentration (Fig. 3⇓A) to induce S1P1 expression in HGECs. Induction of S1P1 increased by 4 h and maximized at 24 h (data not shown). Our results showed that S1P1 specifically (Fig. 3⇓A) was increased by LPS over 24 h, but the other S1P receptors, including S1P2, 3, and 5, were not (S1P4 was not expressed in HGECs) (data not shown). We also found that TLR4 expression could not be induced by S1P itself (data not shown). We, therefore, focused on S1P1 expression induced by TLR4. We confirmed our LPS-induced S1P1 gene expression by also examining the protein expression. HGECs were treated with LPS for 24 h and stained with anti-human S1P1. LPS treatment resulted in an elevated S1P1 induction in HGECs (Fig. 3⇓B). Furthermore, knockdown of TLR4 expression in epithelial cells attenuated S1P1 expression. We also observed that activation of other TLRs, e.g., TLR2, did not result in increased S1P1 expression (data not shown). Together the data reveal that TLR4 signaling induces S1P1 expression, but not S1P2 (Fig 1⇑D), in human primary gingival epithelial cells, and by this means may modulate S1P action on HGECs.

FIGURE 3.
FIGURE 3.

S1P1 was induced by LPS in HGECs. A, Primary human gingival epithelial cells were challenged with protein-free E. coli LPS (1 μg/ml) for 24 h, and the transcription of the S1P1 gene was determined by real-time PCR. The ratio of S1P1 was normalized to GAPDH mRNA. B, HGECs were challenged with LPS (1 μg/ml) for 24 h. The induction of S1P1 protein expression was detected by immunohistochemistry and photographed using a confocal microscope.

CXCL-10 production follows IFN-β expression

IFN-β elicits multifaceted effects in the host innate immune system by inducing other molecules such as CXCL-10 via a pathway requiring Jak proteins (7), including Jak2 or Jak1 (6, 34). To examine the biological effect of IFN-β induced by LPS/S1P stimuli, we, initially, tested whether CXCL-10 expression was an IFN-β-inducible molecule in HGECs. We challenged HGECs with human rIFN-β in the presence or absence of an Ab neutralizing the biological effect of IFN-β. Transcription of the CXCL-10 gene was induced using 50 IU/ml human rIFN-β; however, CXCL-10 mRNA expression was dramatically reduced in HGECs challenged with IFN-β in the presence of its neutralizing Ab (Fig. 4⇓A). Next, we tested whether CXCL-10 expression follows IFN-β expression in the cells challenged with LPS/S1P. HGECs were thus challenged with LPS and S1P in the presence or absence of Jak1 inhibitor or AG490 (Jak2 inhibitor) for 24 h. The cells were incubated with different doses of AG490 (5, 10, or 30 μM) or Jak1 inhibitor (1, 2, 4, or 5 nM) for 24 h. We found that the optimal dose to block Jak1 or Jak2 was 1 nM or 5 μM, respectively (data not shown). The production of CXCL-10 was dramatically reduced in LPS/S1P-treated cells that were pretreated with AG490 (Fig. 4⇓B). Similarly, Jak1 inhibition also significantly attenuated CXCl-10 production following LPS/S1P challenge (Fig. 4⇓C). Our data reveal that CXCL-10 is an IFN-β-inducible gene, and the data show that the production of CXCL-10 follows that of LPS/S1P-induced IFN-β production in human primary gingival epithelial cells.

FIGURE 4.
FIGURE 4.

CXCL-10 expression is an IFN-β-inducible gene. A, Primary human gingival epithelial cells were incubated with human rIFN-β in the presence of IFN-β-neutralizing Ab or its isotype control for 24 h. The expression of CXCL-10 was determined by qPCR. The ratio of CXCL-10 was normalized to GAPDH mRNA. B, CXCL-10 protein production was determined by ELISA in HGECs that were left untreated or were preincubated with AG490 (5 μM) for 1.5 h before stimulation with LPS/S1P. C, CXCL-10 protein production was determined by Luminex100 after blocking Jak1 with Jak1 inhibitor (1 nM). The cells were challenged with LPS/S1P after blocking Jak1. LPS/S1P stimulation increased levels of CXCL-10, which when blocked with Jak1/2 inhibitors resulted in a reduction at the protein level. Data are presented as the mean ± SD of triplicate determinations. The statistically significant (p < 0.05) induction and reduction of IFN-β and CXCL-10 expression are indicated by two asterisks or one asterisks, respectively.

CXCL-10 is induced by S1P1 and TLR4 activation

We then tested whether the CXCL-10 expression was modulated in the same manner as IFN-β regulation in the cells following LPS, S1P, or LPS/S1P stimuli. Epithelial cells were challenged with LPS, S1P, or a combination of both for 24 h. The CXCL-10 mRNA expression could not be increased by LPS alone. Interestingly, CXCL-10 gene expression was inhibited by S1P-challenged cells as compared with medium alone. In contrast, the LPS and S1P combination significantly increased CXCL-10 expression at the gene level (Fig. 5⇓A). We also confirmed our results with protein quantification. Similarly to gene expression, CXCL-10 protein expression was induced in HGECs challenged with LPS/S1P (Fig. 4⇑B). Because gingival epithelial cells express all S1P receptors except S1P4 (data not shown), we examined whether S1P1 specifically cooperates with TLR4 signaling to induce the production of CXCL-10. The cells were stimulated with the S1P1-specific agonist, SEW, and LPS for 24 h. LPS or SEW stimuli alone did not result in increased CXCL-10 expression. In contrast, the LPS and SEW combination treatment induced expression of CXCL-10 in the cells (Fig. 5⇓C). To confirm our results, we blocked G0/i proteins, which are required in the S1P1 signaling pathway (16), with PTx and then challenged the cells with LPS/S1P for 24 h. CXCL-10 expression was reduced in the cells challenged with LPS/S1P after blocking G0/i proteins (Fig. 5⇓D). We found that the CXCL-10 transcription level followed a similar pattern to that of IFN-β, as shown in Fig. 1⇑A, when the cells were challenged with S1P alone. We, thus, determined CXCL-10 expression in cells treated with S1P alone after inhibiting S1P2 signaling with JTE-013. The CXCL-10 transcription level was restored when S1P2 signaling was blocked with JTE-013 in epithelial cells challenged with S1P (Fig. 5⇓E). Taken together, these results clearly show that CXCL-10 was induced by LPS and S1P stimuli in HGECs and specifically required TLR4 and S1P1 receptor activation. Our findings also reveal clearly that S1P2 signaling regulates CXCL-10 gene expression negatively, in analogous fashion to IFN-β expression.

FIGURE 5.
FIGURE 5.

CXCL-10 requires S1P1 and TLR4. A, Epithelial cells were challenged with S1P, SEW (3 μM), LPS, LPS/S1P, or SEW/LPS for 24 h. Following the treatment, values are plotted as the ratio of CXCL-10 and GAPDH. Real-time PCR was performed with Taqman probes. B, CXCL-10 production was determined by Luminex in cells challenged with LPS, S1P, or LPS/S1P. C, The transcription of CXCL-10 was determined by qPCR in the cells following treatment with LPS and SEW (S1P1-specific agonist). D, Expression of CXCL-10 was measured by qPCR, as in A, when cells were challenged with LPS/S1P after blocking Go/i with pertussis toxin. E, CXCL-10 expression in epithelial cells was challenged with S1P alone following pretreatment with JTE-013 (0.5 μM). After the assay, the expression of CXCL-10 or GAPDH was determined by qPCR. Data presented are the mean ± SD of triplicate determinations. The statistically significant induction and reduction (p < 0.05) of CXCL-10 expression are indicated by one asterisk and two asterisks, respectively.

IRF3 phosphorylation is enhanced by S1P in the cells challenged with LPS

Finally, we examined how S1P1 and TLR4 act together to elevate IFN-β expression in HGECs. One of the critical molecules involved in IFN-β expression is the IRF3 (35). IRF3 activation is maximized by PI3K (27), which can be activated by S1P1 (21). We, therefore, determined the IRF3 activation level in cells challenged with LPS, S1P, LPS/S1P, or medium only. Activation of IRF3 was slightly increased in cells challenged with LPS or S1P. However, its activation level was dramatically enhanced in cells challenged with LPS in the presence of S1P (Fig. 6⇓A). We also found that the activation of IRF3 was reduced in cells challenged with LPS/S1P after blocking PI3K activity with LY294002 (Fig. 6⇓A). Furthermore, we examined whether PI3K and IRF3 were involved in the expression of IFN-β in cells treated with LPS/S1P. Thus, we blocked PI3K activity or knocked down IRF3 expression with LY294002 or siRNA technique, respectively, before the challenge assay. Expression of IFN-β was induced in the cells challenged with LPS/S1P. In contrast, PI3K inhibition in the presence of LY294002 significantly diminished the expression of IFN-β in cells treated with LPS/S1P (Fig. 6⇓B). We next knocked down expression of IRF3 by siRNA technique in LPS/S1P-challenged cells (Fig. 6⇓C) and found that IFN-β expression was induced in the cells treated with LPS/S1P following an irrelevant gene silencing (laminin). However, IFN-β expression was attenuated by siRNA to IRF3 in the cells challenged with LSP/S1P (Fig. 6⇓D). Taken together, our data clearly reveal that the activation level of IRF3 through PI3K was maximized by S1P in LPS-challenged cells, and the expression of IFN-β is regulated by PI3K and IRF3 in human epithelial cells challenged with LPS/S1P.

FIGURE 6.
FIGURE 6.

Expression of IFN-β was regulated by IRF3 and PI3K in human epithelial cells. A, Epithelial cells were challenged with S1P, LPS, LPS/S1P, or medium for 1 h. Additionally, the cells were challenged with LPS/S1P for 24 h after blocking PI3K activity with LY294002 (50 μM). The phosphorylation level of IRF3 (serine 398) was measured by Western blotting, and the membrane was stripped and reprobed with β-actin, as an assay control. B, The cells were treated with LPS/S1P in the presence of LY294002 for 24 h. The expression of IFN-β or GAPDH, endogenous control, was determined by qPCR. C, IRF3 expression in the cells was reduced by siRNA, and its expression was normalized to GAPDH and determined by qPCR following the challenge. D, After reducing the expression of IRF3, IFN-β expression was determined by qPCR. The statistically significant induction and reduction (p < 0.05) of IFN-β expression are indicated by one asterisk and two asterisks, respectively.

Discussion

The oral epithelial surfaces is a passive barrier to microorganisms and actively secretes inflammatory cytokines such as IFN-β (8) as part of the host innate defenses. Secreted inflammatory cytokines can have paracrine or autocrine effects on surrounding cells to induce further inflammatory molecules. For example, IFN-β is induced by LPS binding to TLR4 in macrophages (7) and then induces production of CXCL-10 (12). TLR4 signaling can be activated either by MyD88-dependent or -independent pathways to induce inflammatory cytokines. MyD88-dependent TLR4 activation results in proinflammatory cytokine induction including that of IL-6. However, the MyD88-independent pathway typically induces IFN-β. Although molecules that trigger proinflammatory cytokine release such as IL-6 have been extensively studied, molecules that trigger increased IFN-β expression in human epithelial cells are not well elucidated in the literature. In this study, we demonstrate a novel interaction between TLR4 and S1P1 in inducing IFN-β expression in human gingival epithelial cells.

Recently, evidence has been accumulating that many other receptors and signaling pathways modulate IFN-β expression. For example, it has been reported that GPCRs are involved in TLR signaling (15), and the S1P receptor signaling pathway requires GPCR activation following receptor ligation. G0/i is a common protein used by S1P receptors (19, 20) and is pertussin toxin (PTx) sensitive (33). Indeed, our results showed that IFN-β expression at the transcription and protein level was increased in HGECs challenged with LPS and S1P together (Fig. 1⇑, A and B). However, blocking S1P1 signaling by PTx or siRNA to S1P1 resulted in decreased IFN-β expression in human epithelial cells challenged with LPS/S1P (Fig. 2⇑, B and C), suggesting that S1P1 signaling may be critical in the regulation of IFN-β in human epithelial cells. Furthermore, using SEW, an S1P1-specific agonist (Fig. 2⇑D), together with LPS, gave enhanced IFN-β expression, confirming the IFN-β induction effect of S1P is through S1P1. Together, our results reveal that S1P1 signaling is required for the production of IFN-β in cells challenged with LPS. More interestingly, when the cells were challenged with the specific S1P1 agonist (Fig. 2⇑D), or when S1P2 signaling was blocked (JTE-013) before the S1P challenge, IFN-β expression was increased. Parallel to these findings, we found that S1P2 expression was induced by S1P (Fig. 1⇑E), which suggests that there is a balance between the positive effect of S1P1 and the negative effect of S1P2 in the regulation of IFN-β in human epithelial cells treated with S1P. We observed the same effect on CXCL-10 expression in the cells challenged with S1P (Fig. 5⇑A). Intriguingly, only the IFN-β gene expression and not the secretion of IFN-β was reduced by S1P, and we interpret this difference as a time-related effect with changes in mRNA levels occurring more immediately than the protein.

The literature suggests that CXCL-10 expression usually follows IFN-β expression (12) presumably through an autocrine pathway. Our results are confirmatory to others in that CXCL-10 followed the expression of IFN-β in the cells challenged with LSP/S1P (Fig. 4⇑, B and C). In contrast, our data indicated that S1P1 signaling is specifically required to up-regulate the production of CXCL-10 (Fig. 5⇑, C and D). Similar to IFN-β expression in S1P-treated cells, we found that CXCL-10 expression was restored when S1P2 signaling was blocked with JTE-013 (Fig. 5⇑E). All together, S1P1 and TLR4 signaling are indispensable in the regulation of IFN-β and subsequently CXCL-10 production, and S1P2 serves as a negative control on IFN-β in these epithelial cells, hypothetically, to control excessive effects of S1P.

It has been demonstrated that S1P1 is a GPCR and can activate PI3K (21). In contrast, it has been reported that the phosphorylation of IRF3, which is an important transcription factor in the regulation of type 1 IFN (24), can be maximized by PI3K (27). Indeed, we found that activation of IRF3 was increased >1.5-fold in LPS/S1P-challenged cells as compared with S1P- or LPS-treated cells. We also found that the activation of IRF3 was attenuated when PI3K activity was blocked in cells challenged with LPS/S1P (Fig. 6⇑A). More importantly, our data revealed clearly that the expression of IFN-β was reduced in HGECs challenged with LPS/S1P after blocking PI3K activity with LY294002 (Fig. 6⇑B) or knocking down IRF3 expression by siRNA techniques (Fig. 6⇑, B–D). Our data suggest that TLR4 increases S1P1 expression relative to S1P2 expression, and then both TLR4 and S1P1 signaling activates PI3K to maximize IRF3 activation and subsequently induce IFN-β and CXCL-10 in human epithelial cells.

In summary, notable progress has been made in identifying the processes involved in increasing IFN-β expression in different cell types, but these processes are as yet unclear in epithelial cells. This study is the first to report that TLR4 and the S1P family of receptor signaling pathways cooperate with each other to induce expression of IFN-β and CXCL-10 in human primary gingival cells. The association between TLR4 and S1P1 illustrates that the innate immune response can be regulated by TLR and GPCR cooperation in oral mucosal epithelial cells.

Disclosures

The authors have no financial conflict of interest.

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 grant DE017384.

  • ↵2 Address correspondence and reprint requests to Dr. Denis F. Kinane, University of Louisville School of Dentistry, Oral Health and Systemic Disease, 501 South Preston Street, Room 204, Louisville, KY 40202. E-mail address: denis.kinane{at}louisville.edu

  • ↵3 Abbreviations used in this paper: GPCR, G protein-coupled receptor; HGEC, human gingival epithelial cell; IRF3, IFN-regulatory factor 3; PTx, pertussis toxin; qPCR, quantitative PCR; RNAi, RNA interference; S1P, sphingosine 1-phosphate; si, small interfering; SEW, 5-(4-phenyl-5-trifluoromethylthiophen-2-yl)-3-(3-trifluoromethylphenyl)-(1,2,4)-oxadiazole.

  • Received August 28, 2007.
  • Accepted November 14, 2007.

References

  1. ↵
    Kinane, D. F., H. Shiba, P. G. Stathopoulou, H. Zhao, D. F. Lappin, A. Singh, M. A. Eskan, S. Beckers, S. Waigel, B. Alpert, T. B. Knudsen. 2006. Gingival epithelial cells heterozygous for Toll-like receptor 4 polymorphisms Asp299Gly and Thr399Ile are hypo-responsive to Porphyromonas gingivalis. Genes Immun. 7: 190-200.
    OpenUrlCrossRefPubMed
  2. ↵
    Eskan, M. A., G. Hajishengallis, D. F. Kinane. 2006. Differential activation of human gingival epithelial cells and monocytes by Porphyromonas gingivalis fimbriae. Infect. Immun. 75: 892-898.
    OpenUrlPubMed
  3. ↵
    Kaisho, T., S. Akira. 2000. Critical roles of Toll-like receptors in host defense. Crit. Rev. Immunol. 20: 393-405.
    OpenUrlPubMed
  4. ↵
    Medzhitov, R., P. Preston-Hurlburt, C. A. Janeway, Jr. 1997. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity. Nature 388: 394-397.
    OpenUrlCrossRefPubMed
  5. ↵
    Takeuchi, O., K. Hoshino, T. Kawai, H. Sanjo, H. Takada, T. Ogawa, K. Takeda, S. Akira. 1999. Differential roles of TLR2 and TLR4 in recognition of Gram-negative and Gram-positive bacterial cell wall components. Immunity 11: 443-451.
    OpenUrlCrossRefPubMed
  6. ↵
    Van Boxel-Dezaire, A. H., M. R. Rani, G. R. Stark. 2006. Complex modulation of cell type-specific signaling in response to type I interferons. Immunity 25: 361-372.
    OpenUrlCrossRefPubMed
  7. ↵
    Toshchakov, V., B. W. Jones, P. Y. Perera, K. Thomas, M. J. Cody, S. Zhang, B. R. Williams, J. Major, T. A. Hamilton, M. J. Fenton, S. N. Vogel. 2002. TLR4, but not TLR2, mediates IFN-β-induced STAT1α/β-dependent gene expression in macrophages. Nat. Immunol. 3: 392-398.
    OpenUrlCrossRefPubMed
  8. ↵
    Stetson, D. B., R. Medzhitov. 2006. Type I interferons in host defense. Immunity 25: 373-381.
    OpenUrlCrossRefPubMed
  9. ↵
    Salazar-Mather, T. P., C. A. Lewis, C. A. Biron. 2002. Type I interferons regulate inflammatory cell trafficking and macrophage inflammatory protein 1 delivery to the liver. J. Clin. Invest. 110: 321-330.
    OpenUrlCrossRefPubMed
  10. ↵
    Barca, O., S. Ferre, M. Seoane, J. M. Prieto, M. Lema, R. Senaris, V. M. Arce. 2003. Interferon β promotes survival in primary astrocytes through phosphatidylinositol 3-kinase. J. Neuroimmunol. 139: 155-159.
    OpenUrlCrossRefPubMed
  11. ↵
    Teige, I., A. Treschow, A. Teige, R. Mattsson, V. Navikas, T. Leanderson, R. Holmdahl, S. Issazadeh-Navikas. 2003. IFN-β gene deletion leads to augmented and chronic demyelinating experimental autoimmune encephalomyelitis. J. Immunol. 170: 4776-4784.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Buttmann, M., F. Berberich-Siebelt, E. Serfling, P. Rieckmann. 2007. Interferon-β is a potent inducer of interferon regulatory factor-1/2-dependent IP-10/CXCL10 expression in primary human endothelial cells. J. Vasc. Res. 44: 51-60.
    OpenUrlCrossRefPubMed
  13. ↵
    Taub, D. D., A. R. Lloyd, K. Conlon, J. M. Wang, J. R. Ortalo, A. Harada, K. Matsushima, D. J. Kelvin, J. J. Oppenheim. 1993. Recombinant human interferon-inducible protein 10 is a chemoattractant for human monocytes and T lymphocytes and promotes T cell adhesion to endothelial cells. J. Exp. Med. 177: 1809-1814.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Hyun, J. G., G. Lee, J. B. Brown, G. R. Grimm, Y. Tang, N. Mittal, R. Dirisina, Z. Zhang, J. P. Fryer, J. V. Weinstock, et al 2005. Anti-interferon-inducible chemokine, CXCL10, reduces colitis by impairing T helper-1 induction and recruitment in mice. Inflamm. Bowel Dis. 11: 799-805.
    OpenUrlCrossRefPubMed
  15. ↵
    Kagan, J. C., R. Medzhitov. 2006. Phosphoinositide-mediated adaptor recruitment controls Toll-like receptor signaling. Cell 125: 943-955.
    OpenUrlCrossRefPubMed
  16. ↵
    Lee, M. J., J. R. Van Brocklyn, S. Thangada, C. H. Liu, A. Hand, R. Menzeleev, S. Spiegel, T. Hla. 1998. Sphingosine-1-phosphate as a ligand for the G protein-coupled receptor EDG-1. Science 279: 1552-1555.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Hla, T., M. J. Lee, N. Ancellin, C. H. Liu, S. Thangada, B. D. Thompson, M. Kluk. 1999. Sphingosine-1-phosphate: extracellular mediator or intracellular second messenger?. Biochem. Pharmacol. 58: 201-207.
    OpenUrlCrossRefPubMed
  18. ↵
    Hla, T., M. J. Lee, N. Ancellin, J. H. Paik, M. J. Kluk. 2001. Lysophospholipids: receptor revelations. Science 294: 1875-1878.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Meyer zu Heringdorf, D., H. Lass, I. Kuchar, M. Lipinski, R. Alemany, U. Rumenapp, K. H. Jakobs. 2001. Stimulation of intracellular sphingosine-1-phosphate production by G-protein-coupled sphingosine-1-phosphate receptors. Eur. J. Pharmacol. 414: 145-154.
    OpenUrlCrossRefPubMed
  20. ↵
    Malek, R. L., R. E. Toman, L. C. Edsall, S. Wong, J. Chiu, C. A. Letterle, J. R. Van Brocklyn, S. Milstien, S. Spiegel, N. H. Lee. 2001. Nrg-1 belongs to the endothelial differentiation gene family of G protein-coupled sphingosine-1-phosphate receptors. J. Biol. Chem. 276: 5692-5699.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Lee, M. J., S. Thangada, J. H. Paik, G. P. Sapkota, N. Ancellin, S. S. Chae, M. Wu, M. Morales-Ruiz, W. C. Sessa, D. R. Alessi, T. Hla. 2001. Akt-mediated phosphorylation of the G protein-coupled receptor EDG-1 is required for endothelial cell chemotaxis. Mol. Cell 8: 693-704.
    OpenUrlCrossRefPubMed
  22. ↵
    Lee, J. F., Q. Zeng, H. Ozaki, L. Wang, A. R. Hand, T. Hla, E. Wang, M. J. Lee. 2006. Dual roles of tight junction associated protein, zonula occludens-1, in sphingosine-1-phosphate mediated endothelial chemotaxis and barrier integrity. J. Biol. Chem. 281: 29190-29200.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    Sakaguchi, S., H. Negishi, M. Asagiri, C. Nakajima, T. Mizutani, A. Takaoka, K. Honda, T. Taniguchi. 2003. Essential role of IRF-3 in lipopolysaccharide-induced interferon-β gene expression and endotoxin shock. Biochem. Biophys. Res. Commun. 306: 860-866.
    OpenUrlCrossRefPubMed
  24. ↵
    Honda, K., T. Taniguchi. 2006. IRFs: master regulators of signalling by Toll-like receptors and cytosolic pattern-recognition receptors. Nat. Rev. Immunol. 6: 644-658.
    OpenUrlCrossRefPubMed
  25. ↵
    Wietek, C., S. M. Miggin, C. A. Jefferies, L. A. J. O’Neill. 2003. Interferon regulatory factor-3-mediated activation of the interferon-sensitive response element by Toll-like receptor (TLR) 4 but not TLR3 requires the p65 subunit of NF-κ. J. Biol. Chem. 278: 50923-50931.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    Sato, M., H. Suemori, N. Hata, M. Asagiri, K. Ogasawara, K. Nakao, T. Nakaya, M. Katsuki, S. Noguchi, N. Tanaka, T. Taniguchi. 2000. Distinct and essential roles of transcription factors IRF-3 and IRF-7 in response to viruses for IFN-α/β gene induction. Immunity 13: 539-548.
    OpenUrlCrossRefPubMed
  27. ↵
    Ruse, M., U. G. Knaus. 2006. New players in TLR-mediated innate immunity: PI3K and small Rho GTPases. Immunol. Res. 34: 33-48.
    OpenUrlCrossRefPubMed
  28. ↵
    Lee, J. F., Q. Zeng, H. Ozaki, L. Wang, A. R. Hand, T. Hla, E. Wang, M. J. Lee. 2006. Dual roles of tight junction associated protein, zonula occludens-1, in sphingosine-1-phosphate mediated endothelial chemotaxis and barrier integrity. J. Biol. Chem. 281: 29190-29200.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    Inoki, I., N. Takuwa, N. Sugimoto, K. Yoshioka, S. Takata, S. Kaneko, Y. Takuwa. 2006. Negative regulation of endothelial morphogenesis and angiogenesis by S1P2 receptor. Biochem. Biophys. Res. Commun. 346: 293-300.
    OpenUrlCrossRefPubMed
  30. ↵
    Sanchez, T., A. Skoura, M. T. Wu, B. Casserly, E. O. Harrington, T. Hla. 2007. Induction of vascular permeability by the sphingosine-1-phosphate receptor-2 (S1P2R) and its downstream effectors ROCK and PTEN. Arterioscler. Thromb. Vasc. Biol. 27: 1312-1318.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    Sanchez, T., T. Hla. 2004. Structural and functional characteristics of S1P receptors. J. Cell. Biochem. 92: 913-922.
    OpenUrlCrossRefPubMed
  32. ↵
    Pyne, S., N. J. Pyne. 1996. The differential regulation of cyclic AMP by sphingomyelin-derived lipids and the modulation of sphingolipid-stimulated extracellular signal regulated kinase-2 in airway smooth muscle. Biochem. J. 315: 917-923.
    OpenUrlPubMed
  33. ↵
    Lee, M.-J., M. Evans, T. Hla. 1996. The inducible G protein-coupled receptor edg-1 signals via the Gi/mitogen activated-protein kinase pathway. J. Biol. Chem. : 11272-11282.
  34. ↵
    Kim, H. J., K. Tsoyi, J. M. Heo, Y. J. Kang, M. K. Park, Y. S. Lee, J. H. Lee, H. G. Seo, H. S. Yun-Choi, K. C. Chang. 2007. Regulation of lipopolysaccharide-induced inducible nitric-oxide synthase expression through the nuclear factor-κB pathway and interferon-β/tyrosine kinase 2/Janus tyrosine kinase 2-signal transducer and activator of transcription-1 signaling cascades by 2-naphthylethyl-6,7-dihydroxy-1,2,3,4-tetrahydroisoquinoline (THI 53), a new synthetic isoquinoline alkaloid. J. Pharmacol. Exp. Ther. 320: 782-789.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Stockinger, S., B. Reutterer, B. Schaljo, C. Schellack, S. Brunner, T. Materna, M. Yamamoto, S. Akira, T. Taniguchi, P. J. Murray, et al 2004. IFN regulatory factor 3-dependent induction of type I IFNs by intracellular bacteria is mediated by a TLR- and Nod2-independent mechanism. J. Immunol. 173: 7416-7425.
    OpenUrlAbstract/FREE Full Text
View Abstract
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section