INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Fimbriae, hairlike microfibers, are observable on the cell surface of various bacteria by electron microscopy. Among the various cell surface components, fimbriae are known to be a specific adherence factor, or adhesin, in their microbial etiology (34). Porphyromonas gingivalis has been recognized as a major periodontopathogenic organism (40), and strains of P. gingivalis possessing virulence factors containing fimbriae have been shown to be involved in the development of periodontal diseases (49). Several studies have also described the immunobiological properties of P. gingivalis fimbriae and their active peptides (25-27, 31, 32).

Epithelial cells function as sensors of external stimuli and conduct signals to internal cells (18). Gingival epithelial cells are also thought to play an important role as a first barrier against periodontopathic organisms and their metabolic products. Several bacterial surface components, such as lipopolysaccharide (LPS) and its active center, lipid A, as well as fimbrial protein and peptidoglycan have been implicated in the development and progression of periodontal diseases (13, 51). However, the recognition mechanisms for these potentially pathogenic components in gingival epithelial cells are not well understood.

Toll-like receptors (TLRs) have been identified in monocytes and macrophages based on their homology to Drosophila protein (20, 52). Mammalian TLRs comprise a large family with extracellular leucine-rich repeats and a cytoplasmic Toll/interleukin-1 (IL-1) receptor homology domain and have been implicated in the recognition of bacterial cell wall components (22). Ten members (TLR1 to -10) have been reported (7, 8, 14, 22, 44), and among them, it was recently demonstrated that TLR4 plays an important role as a receptor of bacterial LPS and lipid A (15, 35). TLR2 is also essential for the signaling of various bacterial components, such as Staphylococcus aureus peptidoglycan (43), bacterial lipoproteins (2, 4, 19, 42), lipoteichoic acid (21), and zymosan (46). More recently, TLR9 was found to recognize bacterial DNA (14).

We report here the recognition mechanism of human gingival epithelial cells used to defend against P. gingivalis fimbriae and their peptides.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacteria and fimbrial preparation. P. gingivalis strain 381 was grown anaerobically in GAM broth (Nissui, Tokyo, Japan) supplemented with hemin and menadione for 26 h at 37°C. Fimbriae were isolated and purified as described previously (28).

Fimbrial synthetic peptide. In our previous study, we found that ALTTE, residues 69 to 73 of the fimbrillin, functions in the induction of IL-6 production in human peripheral blood mononuclear cells (PBMC) (31). This active peptide was synthesized and purified as described previously (32). The peptide specimen was dissolved in the culture medium described below before the assay.

Bacterial and synthetic components. Escherichia coli-type lipid A (compound 506), which possesses -(1-6)-linked glucosamine disaccharide 1, 4'-bisphosphate, with two acyloxyacyl groups at the 2' and 3' positions and two 3-hydroxytetradecanoyl groups at the 2 and 3 positions, was synthesized as described by Imoto et al. (17), and the lipid A specimen was dissolved at a concentration of 2 mg/ml in a 0.1% (vol/vol) triethylamine aqueous solution. A cell wall peptidoglycan specimen of Staphylocaccus aureus was prepared in our laboratory as described previously (30). N-Acetylmuramyl-L-alanyl-D-isoglutamine (MDP) was supplied by Sigma (St. Louis, Mo.). These specimens were then dissolved in a pyrogen-free cell culture medium before the assay.

Establishment of HGEC cell lines. Normal human gingival tissues (ca. 500 mg) were obtained from two patients who required tooth extractions for reasons other than periodontal disease after receiving informed consent. Human gingival epithelial cells were transfected with human papillomavirus 16 (HPV-16) E6 and E7 open reading flames (ORFs) using a retroviral system of HPV-16, which was kindly provided by M. Saitoh (Kanagawa Dental College, Yokosuka, Japan) (23). Immortalized human gingival epithelial cells (HGEC) were subsequently maintained in a long-term culture with HuMedia-KG2 (Kurabo Biomedicals, Osaka, Japan) and exhibited no morphological changes in long-term cultivation. These cells were designated HGEC-1 and HGEC-2, respectively, and used between passages 50 and 60.

Isolation of human monocytes. Heparinized venous blood drawn from healthy donors was subjected to fractionation using Histopaque-1077 (Sigma) to obtain human PBMC (3). After washing the PBMC, monocytes were isolated by a magnetic cell sorting system using a monocyte isolation kit (Miltenyi Biotech Gmbh, Bergisch Gladbach, Germany).

Immunohistochemistry. HeLa, an epithelial cell line established from a human cervical cancer specimen, and Gin-1, a gingival fibroblast cell line, were provided by Dainippon Pharmaceutical, Osaka, Japan, and maintained in Dulbecco's modified Eagle's medium (Nikken Biomedicals, Kyoto, Japan) supplemented with 10% fetal bovine serum (FBS; Sigma). HGEC, HeLa, and Gin-1 cells were seeded subconfluently onto plastic tissue culture slides (no. 177437; Nunc Inc.), after which the slides were washed with phosphate-buffered saline (PBS) and fixed in 2% paraformaldehyde in PBS for 10 min. After blocking with 5% normal horse serum in PBS, the fixed cells were incubated for 30 min with mouse monoclonal antibody to human keratin/cytokeratin (AE1/AE3; Nichirei Co., Tokyo, Japan) or mouse immunogloblin G1 (IgG1) isotype (Zymed Labs, Inc., San Francisco, Calif.) as a negative control, followed by staining with a Vectastain Elite ABC kit (Vector Labs, Inc.).

Reverse transcription-PCR. Total cellular RNA from HGEC or human monocytes was extracted with RNAzol B (Tel-Test, Friendswood, Tex.) using a single-step isolation method (6) according to the manufacturer's recommendation. RNase-free DNase (Takara Biochemicals, Shiga, Japan) was used to remove genomic DNA based on methods described previously (9). Two micrograms of extracted RNA was reverse-transcribed into first-strand cDNA at 42°C for 40 min, using 100 U/ml of reverse transcriptase (RT; Takara Biomedicals) and 0.1 µM oligo(dT) adapter primer (Takara Biomedicals) in a 50-µl reaction mixture.

Flow cytometric analysis. HGEC or human monocytes were incubated for 15 min at room temperature with mouse monoclonal antibody to human CD14 (Dako, Glostrup, Denmark). In the controls, cells were incubated with mouse IgG2b isotype (Dako). After washing with PBS containing 0.1% azide, the cells were incubated for 15 min at room temperature with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse immunogloblins (Dako). For TLR4 detection, the cells were stained with goat polyclonal antibody to human TLR4 (Santa Cruz Biotech., Inc., Santa Cruz, Calif.), followed by FITC-conjugated rabbit anti-goat IgG (Zymed Lab., Inc.). Goat IgG (heavy and light chain) (Chemicon International, Inc., Temecula, Calif.) was used for the isotype control. For TLR2 detection, the cells were stained with mouse monoclonal antibody to human TLR2 (TL2.1; Cascade Bioscience Inc., Winchester, Mass.), followed by FITC-conjugated anti-mouse immunogloblins. Mouse IgG2a (Dako) was used for the isotype control. The cells were washed with PBS-0.1% azide, and then fixed with 1% paraformaldehyde. The stained cells were analyzed with a FACSCalibur using Cell Quest software (Becton Dickinson and Co., San Jose, Calif.).

Cytokine production. HGEC were trypsinized and suspended at a cell density of 2 × 10⁵ cells/ml of HuMedia-KG2 with supplements. A single-cell suspension (2 × 10⁴ cells per well) was seeded in a 96-well flat-bottomed microtiter plate (Falcon 3072; Becton Dickinson and Co., Lincoln Park, N.J.). After incubation for 16 h at 37°C in humidified air containing 5% (vol/vol) CO₂, the monolayers were washed three times with PBS and then treated with the indicated doses of the test specimens in 200 µl of HuMedia-KG2 with or without 1% FBS, 0.5 µg of CD14 (Genzyme-Techne, Minneapolis, Minn.), and/or 0.05 µg of LBP (Genzyme-Techne) per ml at 37°C for 24 h. In some experiments, HGEC were incubated with or without 1 µg of TL2.1 or mouse IgG2a per ml for 30 min at room temperature before adding the test specimens. After incubation, the supernatants were collected and stored at 80°C until the assay for cytokine production.

Luciferase assay. A single-cell suspension (5 × 10⁴ cells per well) was seeded in a 24-well flat-bottomed microtiter plate (Falcon 3047; Becton Dickinson and Co.). After incubation for 24 h at 37°C in humidified air containing 5% (vol/vol) CO₂, the monolayers were washed three times with PBS and then transfected with 0.8 µg of pNF-B-Luc plasmid (Stratagene Co., La Jolla, Calif.) by use of TransFast transfection reagent (Promega Co., Madison, Wis.). The pFC-MEKK plasmid was used as a positive control plasmid in the assay. After an initial incubation for 24 h, the cells were incubated with the indicated doses of the test specimens in 500 µl of HuMedia-KG2, with or without 0.5 µg of CD14 and LBP per ml at 37°C for 7 h. After incubation, the cells were lysed in 80 µl of cell culture lysis reagent (Promega Co.), and then luciferase activity was determined using 20 µl of lysate and 100 µl of luciferase assay substrate (Promega Co.). The luminescence was quantified with a luminometer (Promega Co.).

Statistics. Data were analyzed by a one-way analysis of variance, using the Bonferroni or Dunn method, and the results are presented as the mean ± standard error of the mean (SEM). When an individual experiment is shown, it is representative of at least three independent experiments.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Detection of total keratin and involucrin. Expression of total keratin and involucrin in HGEC was assessed by immunohistochemical staining. HGEC-1 were clearly stained with mouse monoclonal antibody to human total keratin/cytokeratin, as shown in Fig. 1A. In contrast, these cells were not stained with mouse IgG1, an isotype control (Fig. 1B). HeLa cells, which are known sources of total keratin expression (12), also exhibited definite total keratin expression (Fig. 1C), but no staining was shown in Gin-1 (Fig. 1E). Further, HGEC-1 treated with culture medium containing 2.0 mM calcium for 48 h showed the induction of differentiation and obviously expressed involucrin (Fig. 1G), in contrast to the cells treated with 0.15 mM calcium for 24 h (Fig. 1H). HGEC-2, similar to HGEC-1, were shown to express total keratin and involucrin (data not shown).

View larger version (105K): [in this window] [in a new window]   FIG. 1.   Immunohistochemical staining with anti-total keratin and anti-involucrin antibodies. The stains were used to determine total keratin-positive and -negative cells, respectively, in HGEC-1 (A, B, G, and H), HeLa (C and D), and Gin-1 (E and F) cells. Cells were stained with mouse monoclonal antibody to human total keratin (AE1/AE3) (A, C, and E) and an isotype control, IgG1 (B, D, and F). HGEC-1 were incubated with medium containing 2.0 mM (G) or 0.15 mM (H) calcium for 24 h or with mouse monoclonal antibody to human involucrin for involucrin detection.

Expression of CD14, TLR2, and TLR4. We examined the expression of CD14, TLR2, and TLR4 mRNA in HGEC by RT-PCR (Fig. 2). Both HGEC-1 and HGEC-2 showed TLR2 mRNA expression; however, they expressed no CD14 or TLR4 mRNA. Human peripheral monocytes, as a control, markedly expressed CD14, TLR2, and TLR4 mRNA. RT-PCR analysis of -actin expression confirmed the quality of all RNA preparations used for RT-PCR. No band was detected in the non-RT sample by PCR. Further, we examined the cell surface expression of CD14, TLR2, and TLR4 gene products in HGEC by flow cytometry analysis (Fig. 3). HGEC-1 clearly expressed TLR2, and HGEC-2 showed weak TLR2 expression. These cells did not express CD14 or TLR4. Human peripheral monocytes, as a positive control, demonstrated specific cell surface CD14, TLR2, and TLR4 staining.

View larger version (43K): [in this window] [in a new window]   FIG. 2.   CD14, TLR2, and TLR4 mRNA expression in HGEC. (A) Expression of human CD14, TLR2, and TLR4 mRNA was analyzed by RT-PCR as detailed in Materials and Methods. Human monocytes were used as positive sources of CD14, TLR2, and TLR4 mRNA expression to confirm the specificity of the primers and PCR. The -actin gene was assayed as a positive control. PCR products of non-RT samples were examined as a negative control. Lane M, Size markers. (B) Ethidium bromide-stained PCR products were photographed, and then the images were digitized and analyzed. Results are expressed as the ratio of each PCR product to -actin band density. Data represent three independent experiments. PCR was performed in duplicate for each assay.

View larger version (35K): [in this window] [in a new window]   FIG. 3.   CD14, TLR2, and TLR4 expression in HGEC. Expression of CD14, TLR2, and TLR4 in HGEC was determined by staining with specific antibodies (bold line) or their isotype as a control (thin line) as detailed in Materials and Methods. Human monocytes were used as positive sources of CD14, TLR2, and TLR4.

Induction of IL-8 production. IL-8-producing activities in HGEC after stimulation with test specimens without FBS were examined (Fig. 4). P. gingivalis fimbriae as well as S. aureus peptidoglycan significantly induced IL-8 production in both HGEC-1 and HGEC-2, whereas stimulation of these cells with compound 506 resulted in no induction of IL-8 production. HGEC-1 produced more IL-8 than HGEC-2. The IL-8-producing activities coincided with the level of TLR2 expression and its gene product in each of the HGEC lines.

View larger version (17K): [in this window] [in a new window]   FIG. 4.   IL-8-producing activities in HGEC by various bacterial components. Cells were stimulated at 37°C for 24 h with 10 µg/ml of each test specimen in HuMedia-KG2 in the absence of FBS. After incubation, the supernatants were collected, and IL-8 production was determined by ELISA. Experiments were done at least three times, and representative results are presented. Each assay was done in triplicate wells, and the data are expressed as the mean ± SEM. *, significant difference between the groups with and without the test specimens (P < 0.01).

View larger version (18K): [in this window] [in a new window]   FIG. 5.   Effect of FBS on IL-8-producing activities in HGEC stimulated with S. aureus peptidoglycan (A), MDP (B), P. gingivalis fimbriae (C), ALTTE (D), or compound 506 (E). HGEC-1 were incubated at 37°C for 24 h with the indicated doses of the test specimens with (solid circles) or without (open circles) FBS. The experimental protocols were the same as described in the legend to Fig. 4. Significant differences were seen between the groups with and without the test specimens and with (*, P < 0.01) or without (, P < 0.01) FBS.

View larger version (19K): [in this window] [in a new window]   FIG. 6.   Effect of soluble CD14 and LBP on IL-8-producing activities in HGEC stimulated with (solid bars) or without (open bars) S. aureus peptidoglycan (A), MDP (B), P. gingivalis fimbriae (C), ALTTE (D), or compound 506 (E). HGEC-1 were incubated at 37°C for 24 h with 1 µg/ml of each test specimen supplemented with CD14 (0.5 µg/ml) and/or LBP (0.05 µg/ml). The experimental protocols are the same as described in the legend to Fig. 4. Significant differences were seen between the groups with and without the bacterial specimens (*, P < 0.01), and between the groups with bacterial specimens with and without CD14 and/or LBP (, P < 0.01).

NF-B activation in HGEC. Transcription factor NF-B is involved in the TLR signaling cascade (54). We examined NF-B activation in HGEC-1 after stimulation with various bacterial components along with CD14 and LBP by a luciferase assay. P. gingivalis fimbriae and the peptide, ALTTE, as well as S. aureus peptidoglycan and MDP significantly activated NF-B in HGEC-1 (Fig. 7). These specimens (10 to 100 µg/ml) also clearly activated NF-B without CD14 or LBP (data not shown). However, compound 506 expressed no activation with CD14 and LBP (Fig. 7). These findings coincided with the pattern of IL-8 production by HGEC (Fig. 6).

View larger version (18K): [in this window] [in a new window]   FIG. 7.   NF-B activation in HGEC in response to various bacterial components. HGEC-1 were transfected with 1 µg of pNK-B-Luc plasmid and then incubated with or without 1 µg/ml of each test specimen with CD14 and LBP in HuMedia-KG2 at 37°C for 7 h. After incubation, the cells were lysed, and then luciferase activity was estimated with a luminometer. The experimental protocols were the same as described in the legend to Fig. 4. Significant differences were seen between the groups with and without the test specimens (*, P < 0.01).

Inhibitory effect of mouse monoclonal antibody to human TLR2 on IL-8 production by HGEC. S. aureus peptidoglycan has been reported to be recognized by TLR2 (43). To examine the recognition of HGEC in response to bacterial components, HGEC-1 were treated with mouse monoclonal antibody to human TLR2, TL2.1, or mouse IgG2a as a control before the addition of various bacterial specimens. TL2.1 significantly inhibited IL-8 production induced by P. gingivalis fimbriae, ALTTE, S. aureus peptidoglycan, and MDP in HGEC (Fig. 8). These results indicated that these bacterial components, except for compound 506, activated HGEC through TLR2.

View larger version (30K): [in this window] [in a new window]   FIG. 8.   Effect of mouse monoclonal antibody to human TLR2 on IL-8-producing activities in HGEC stimulated with various bacterial components. HGEC-1 were incubated with 1 µg/ml of mouse monoclonal antibody to human TLR2, TL2.1, or mouse IgG2a for 30 min at room temperature before the addition of various test specimens (each 1 µg/ml) in the presence of CD14 and LBP. After 24 h of incubation, the supernatants were collected, and IL-8 production was estimated by ELISA. The experimental protocols were the same as described in the legend to Fig. 4. A significant difference was seen from the medium alone (*, P < 0.01).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Periodontal diseases are infectious, have a bacterial etiology, and cause inflammatory responses elicited by periodontopathogenic exposure in periodontal tissues. In these responses, epithelial cells are thought to play important roles as the initial point of contact with these pathogens. In the present study, we established immortalized HGEC and examined the expression of total keratin, an epithelial cell marker, by immunohistochemical staining. Oda et al. (23) reported that HPV-16-immortalized oral epithelial cells expressed positive staining with antikeratin antibodies as well as these primary cultured cells. In the present study, HGEC were clearly stained with anti-total keratin antibodies AE1 and AE3. When incubated in culture medium containing 2.0 mM calcium, they also clearly expressed involucrin (Fig. 1). These results demonstrate that immortalized HGEC remained characterized by epithelial cell markers.

Our results also indicated that HGEC express TLR2 but not CD14 or TLR4 (Fig. 2 and 3). Recently, it was shown by RT-PCR that intestinal epithelial cells do not express CD14 mRNA (5). Those cell lines also exhibited a differential expression of TLR2 and TLR4. Faure et al., on the contrary, demonstrated that human dermal umbilical vein and microvessel endothelial cells predominantly expressed TLR4 and showed very weak TLR2 expression. The cells in that study responded to LPS but not Mycobacterium tuberculosis lipoprotein (11). Therefore, it is suggested that various types of cells have different combinations of each TLR to recognize microbial pathogens.

We examined the responses of HGEC using S. aureus peptidoglycan as a TLR2 ligand (44) and E. coli-type synthetic lipid A, compound 506, as a TLR4 ligand (15) in the absence of FBS. HGEC induced IL-8 production stimulated by S. aureus peptidoglycan but not compound 506 (Fig. 4). Wang et al. showed that S. aureus peptidoglycan activated a human embryonic kidney cell line, HEK293, through TLR2 (48). These results suggest that HGEC recognize S. aureus peptidoglycan through TLR2, and TLR4-deficient HGEC are unresponsive to gram-negative bacterial lipid A. It was reported that various cytokines play important roles in the pathogenesis of periodontitis (37, 51). Gingival epithelial cells have been shown to produce cytokines, including IL-8, which are considered the important factor that participates in the initiation and maintenance of inflammatory reactions (16). IL-8 has been shown to be localized in gingival tissues of patients with periodontitis, and IL-8 mRNA was also expressed. IL-8 mRNA levels were shown to correspond to the severity of periodontitis (45).

Many studies have demonstrated that bacterial cell walls and their structural units, whether obtained by use of an enzyme or synthesized, exhibit various biological activities in a variety of host cells (29). These studies have also revealed that MDP is a common component of cell wall peptidoglycan that exhibits various immunobiological properties (41). The present study found that MDP as well as S. aureus peptidoglycan induces IL-8 production and NF-B activation in TLR2-positive HGEC (Fig. 5 and 7). Furthermore, among various cell surface components, bacterial fimbriae are suspected to be a specific adherence factor. Previously, we demonstrated that fimbriae and their synthetic peptides derived from P. gingivalis, an anaerobic gram-negative periodontopathic bacterium, bind specifically to human gingival fibroblasts (24). It has also been shown that P. gingivalis fimbriae and a synthetic peptide composed of residues 69 to 73 of the fimbrillin, ALTTE, function in the activation of human PBMC (32). In the present study, the active fimbrial peptide, ALTTE, was found to induce IL-8 production in TLR2-positive and TLR4-negative HGEC. To our knowledge, this is the first time that a bacterial fimbrial protein and its active peptide have been indicated as a TLR2 ligand.

CD14, which is present on monocytic cell surfaces, binds to LPS or gram-positive bacterial cell wall components and facilitates signaling (53). It was recently shown that TLRs act as transmembrane coreceptors to CD14 in cellular responses (1). Further, a soluble form of CD14 was found to be able to functionally replace membrane-bound CD14 in CD14-negative cells, such as endothelial cells and astrocytes (38, 39, 50). It was also reported that MDP binds to CD14 molecules (10). Our present results showed that soluble CD14 clearly enhances HGEC activation stimulated with MDP as well as S. aureus peptidoglycan (Fig. 6). Furthermore, LBP with soluble CD14 causes greater cytokine-producing activity. However, LBP alone was not found to elicit an increase of activities. These results suggest that LBP is not associated with direct cell activation. Vesy et al. also reported that LBP promoted rapid binding of purified LPS to CD14 (47).

In the present study, similar results were observed, as P. gingivalis fimbriae and their active peptides appeared to bind to CD14 molecules and augment HGEC activation. This bacterial surface component-induced IL-8 production of HGEC was inhibited by mouse monoclonal antibody to human TLR2 as well as MDP. Taken together, these results demonstrate that bacterial fimbriae and their active peptides are recognized by TLR2 on human gingival epithelial cells and induce IL-8 production via NF-B activation in HGEC. The recognition of pathogenic factors and their cell responsiveness via this pathway may play an important role in the development and progression of periodontal diseases.