As P. gingivalis is implicated as a major pathogenic bacteria for chronic periodontitis [14], we examined the effect of P. gingivalis infection on IL-33 mRNA expression in human gingival epithelial cells in culture. The IL-33 mRNA expression was increased 20-fold at 48 h after stimulation with P. gingivalis W83 in Ca9-22 cells, which is a human gingival epithelial cell line established from squamous cell carcinoma. Pretreatment of the cells with cycloheximide, a protein synthesis inhibitor, blocked the induction of IL-33 mRNA levels, suggesting that de novo protein synthesis was required for P. gingivalis-mediated IL-33 mRNA induction. Nile et al. [10] reported that P. gingivalis LPS induced IL-33 expression in human monocytes. Therefore, we examined the possible IL-33 mRNA-inducing capacity of P. gingivalis-related specimens; fimbriae (Toll-like receptor (TLR) 2 ligand [15]) and LPS (TLR2 and 4 ligand [16]) prepared from P. gingivalis, and synthetic P. gingivalis-type lipopeptide PGTP2-RL (TLR2/1 ligand [17]). Unlike P. gingivalis whole cells, fimbriae and lipopeptide were completely inactive, and LPS specimens were only weakly active in this respect. These results suggest that gingival epithelial cells respond to P. gingivalis rather than TLR ligands for induction of IL-33 mRNA.

P. gingivalis produces two types of arginine-specific cysteine proteinases (Arg-gingipains, RgpA and RgpB) and a lysine-specific cysteine proteinase (Lys-gingipain, Kgp) [18]. Gingipains are localized to a cell-associated form, a soluble form, and released as outer membrane blebs [18]. We next examined whether enzymatic activities of gingipains are involved in the IL-33-inducing capacity in Ca9-22 cells and primary oral epithelial cells. The induction of IL-33 mRNA expression by P. gingivalis W83 in the two types was completely and significantly inhibited by FPR-cmk (Rgp inhibitor) and KYT-36 (Kgp inhibitor), respectively. Furthermore, P. gingivalis KDP136, a gingipains-null mutant, unlike P. gingivalis ATCC 33277, a wild-type parent strain of KDP136, did not induce IL-33 mRNA expression in either cell type. P. gingivalis-induced IL-33 mRNA expression was abolished by heat treatment (70 °C, 1 h) to inactivate enzymatic activities of P. gingivalis. We confirmed intact proteolytic activities for Rgps and Kgp in the whole cells of P. gingivalis W83 and ATCC 33277, but not in those of P. gingivalis KDP136. These observations suggest that the proteolytic activity of gingipains is essential for the induction of IL-33 mRNA expression by P. gingivalis in human gingival epithelial cells.

Next, we examined whether P. gingivalis induced IL-33 expression in human gingival epithelial cells using an immunoblot analysis. IL-33 expression in Ca9-22 cells was increased tenfold with a peak at 4 days by stimulation with whole cell preparations of P. gingivalis. Epithelial cells constitutively express IL-33 in their nuclei under resting conditions [19]. IL-33 is a nuclear protein that is also released into the extracellular environment. To further determine whether P. gingivalis-induced IL-33 was accumulated in the nuclei or the cytoplasm, we analyzed the location of IL-33 protein in Ca9-22 cells after stimulation with P. gingivalis using immunocytochemical analysis. Although IL-33 was constitutively expressed in the resting cells, IL-33 molecules accumulated in the cytoplasm of the cells when they were stimulated with P. gingivalis for 4 days. However, the released IL-33 level was quite low (approximately 30 pg/mL), even after stimulation with P. gingivalis whole cells. IL-33 is not detected in viable human monocytic cells, even upon stimulation with P. gingivalis LPS [10]. Furthermore, IL-33 levels in gingival crevicular fluid (GCF) are not different between chronic periodontitis patients and healthy controls [11]. In fact, IL-33 is not detected in the GCF of inflamed regions from chronic periodontitis patients [13]. Further studies are required to elucidate the fate of IL-33 induced by P. gingivalis in gingival epithelial cells, which may modulate the innate immune functions of the cells in infected periodontal lesions.

Proteinase-activated receptor-2 (PAR-2) is a seven-transmembrane domain receptor family which couples to G-proteins [20]. Rgps are capable of activating PAR-2 expressed on human gingival epithelial cells, which produce proinflammatory cytokines [21]. We examined the possible involvement of PAR-2 in P. gingivalis-induced IL-33 mRNA expression. Ca9-22 cells constitutively expressed PAR-1, −2, −3, and −4 mRNA. PAR-2 mRNA expression was increased upon stimulation with P. gingivalis W38 in Ca9-22 cells. Next, we performed inhibition of PAR-2 mRNA expression by RNA interference using a PAR-2-specific small interference RNA (siRNA). The induction of IL-33 mRNA by P. gingivalis W83 was partially inhibited in PAR-2 siRNA-transfected cells. PAR-2 is activated by a tethered ligand when cleaved by protease. A PAR-2 agonist peptide based on the tethered ligand sequences can activate PAR-2 in a proteolysis-independent manner. However, the PAR-2 agonist peptide was not able to induce IL-33 mRNA in Ca9-22 cells. Pretreatment of the cells with cytochalasin D, a particle internalization inhibitor, inhibited the IL-33 mRNA induction by P. gingivalis. These findings suggested that both proteolytic activation of PAR-2 by gingipains and the endocytosis of P. gingivalis are required for the up-regulation of IL-33 expression induced by P. gingivalis in oral epithelial cells. P. gingivalis enters gingival epithelial cells by endocytosis, which mediates binding of Rgp to the cells [22]. It must be noted that gingipains are required for maturation of P. gingivalis fimbriae [23], which is essential for internalization of the bacterium into epithelial cells [24]. However, endocytosis of both PAR-2 and P. gingivalis are probably required for the up-regulation of IL-33 in gingival epithelial cells.

Because PAR-2 has been reported to be coupled to G protein, leading to activation of phospholipase C (PLC) [25, 26], we examined whether or not the PLC pathway is involved in P. gingivalis-induced IL-33 mRNA expression in Ca9-22 cells. The cells were pretreated with U-73122, a PLC inhibitor, and then stimulated P. gingivalis W83 for 48 h. The pretreatment significantly inhibited the P. gingivalis-induced IL-33 mRNA expression. In contrast, a protein kinase C (PKC) inhibitor, GF-109203X, did not inhibit either the basal level of IL-33 or the P. gingivalis-induced increase in IL-33. These findings suggest that P. gingivalis-induced up-regulation of IL-33 expression is mediated via a PAR-2-PLC-signaling pathway.

As p38 and NF-κB are implicated in PAR-2 signaling [27], we first confirmed p38 phosphorylation in Ca9-22 cells stimulated with P. gingivalis W83. Then, we demonstrated that p38 signaling is involved in P. gingivalis-induced IL-33 mRNA expression using SB203580, a p38 inhibitor. In contrast, the P. gingivalis