Differential activation of human gingival epithelial cells and monocytes by Porphyromonas gingivalis fimbriae

Abstract

Humans develop periodontitis in response to challenge by microbial dental plaque. Inflammation begins after perturbation of gingival epithelial cells by subgingival bacteria interacting through pattern-recognition receptors, including the Toll-like receptors (TLR). Porphyromonas gingivalis is a major periodontopathogen that interacts with epithelial cells through its cell surface fimbriae (FimA), leading to colonization and/or invasion. Previous work by our group has established membrane CD14 as an essential coreceptor for TLR2-mediated activation of transfected cell lines by P. gingivalis FimA. We have shown that gingival epithelial cells express TLR2 but not CD14 on their cell surfaces. We thus speculated that P. gingivalis FimA does not readily activate epithelial innate immune responses but rather functions to promote P. gingivalis colonization in the absence of a vigorous FimA-induced response. This hypothesis was verified by the findings that primary human gingival epithelial cells responded poorly to FimA in terms of interleukin (IL)-6, IL-8, granulocyte-macrophage colony-stimulating factor, and tumor necrosis factor alpha responses, in stark contrast to the marked response to other TLR2 agonists (Pam3Cys, FSL-1) that are not strictly dependent on CD14. On the other hand, CD14-expressing human primary monocytes responded with high levels of the same cytokines to both FimA and the control TLR2 agonists. The gingival epithelial cells failed to respond to FimA even in the presence of exogenously added soluble CD14. These data indicate that the gingival epithelial cell hyporesponsiveness to FimA is attributable to the lack of membrane-expressed but not soluble CD14. In conclusion, P. gingivalis FimA differentially activates human monocytes and epithelial cells, perhaps reflecting different tactics used by P. gingivalis when interacting with different host cell types or a host strategy to limit inflammation.

FIG. 1.

TLR2 and CD14 protein expression by human primary monocytes and gingival epithelial cells. HGECs (A) and monocytes (C) were stained with phycoerythrin-conjugated monoclonal anti-human CD14 or its isotype control, mouse IgG1. Isolated HGECs (B) and monocytes (D) were stained with fluorescein isothiocyanate-conjugated monoclonal anti-human TLR2 or its isotype control, IgG2a, for 20 min at 4°C. The stained cells were analyzed by flow cytometry using a BD FACSCalibur and CellQuest software.

FIG. 2.

Cytokine induction in epithelial cells challenged with P. gingivalis or TLR2 agonists (Pam3Cys or FSL-1). HGECs were treated with heat-killed P. gingivalis (P. g; MOI, 100:1), nonfimbriated P. gingivalis (Mutant P. g; MOI, 100:1), FimA (10 μg/ml), or the TLR2 agonist Pam3Cys (Pam3) or FSL-1 (both at 1 μg/ml) for 4 h. Induction of IL-8 (A), TNF-α (B), GM-CSF (C), and IL-6 (D) was determined in culture supernatant by use of Luminex 100 technology (Upstate Cell Signaling Solutions, NY). Data are presented as the means ± standard deviations of triplicate determinations, from one of three independent sets of experiments that yielded similar findings. Statistically significant (P < 0.05) induction of cytokine release is indicated by an asterisk.

FIG. 3.

Cytokine induction in the presence of 5% FBS in human gingival epithelial cells. HGECs were challenged by P. gingivalis (P. g; MOI, 100:1), FimA (10 μg/ml), or TLR2 agonists (Pam3 or FSL-1; 1 μg/ml) for 4 h at 37°C in the presence of 5% FBS. Induction of the cytokines TNF-α (A), IL-6 (B), and GM-CSF (C) was determined in culture supernatants by use of Luminex 100 technology (Upstate Cell Signaling Solutions, NY). Data are presented as the means ± standard deviations of triplicate determinations. Statistically significant (P < 0.05) induction of cytokine release compared to treatment with medium plus 5% FBS is indicated by an asterisk.

FIG. 4.

Cytokine induction in human gingival epithelial cells in the presence of soluble CD14 and/or LBP. HGECs were challenged with TLR2 agonists (Pam3 or FSL-1; 1 μg/ml) or FimA (10 μg/ml) in the presence of soluble human recombinant CD14 (0.5 μg/ml) and/or LBP (0.05 μg/ml) for 4 h at 37°C. Cell culture supernatants were assayed for IL-8 (A), IL-6 (B), and GM-CSF (C) by use of Luminex 100 technology (Upstate Cell Signaling Solutions, NY). Data are presented as the means ± standard deviations of triplicate determinations. Statistically significant (P < 0.05) induction of cytokine release is indicated by an asterisk.

FIG. 5.

Cytokine induction in human primary monocytes. Purified human primary monocytes were challenged by TLR2 agonists (Pam3 or FSL-1; 1 μg/ml), FimA (1 or 10 μg/ml) or P. gingivalis (MOI, 100:1) for 16 h at 37°C. Culture supernatants were assayed for IL-8 (A), TNF-α (B), and IL-6 (C) responses using enzyme-linked immunosorbent assay kits. Data are presented as the means ± standard deviations of triplicate determinations. Statistically significant (P < 0.05) induction of cytokine release is indicated by an asterisk.

FIG. 6.

P. gingivalis FimA does not influence TLR1, -2, and -6 expression in human gingival epithelial cells. HGECs were challenged with either P. gingivalis (P.g; MOI, 100:1) or FimA (10 μg/ml) for 4 h at 37°C. Real-time PCR was performed with an ABI 7500 system (Applied Biosystems). TaqMan probes and sense and antisense primers for gene expression of human TLR1, -2, and -6 were purchased from Applied Biosystems along with probes and primers for the human endogenous control, GAPDH. Using a Universal PCR Master Mix (Applied Biosystems), the reactions were carried out according to the manufacturer's protocol. The ratio of TLR2 (A), TLR1 (B), and TLR6 (C) mRNAs was normalized to GAPDH mRNA. Data are presented as the means ± standard deviations of triplicate determinations. NS, not statistically significant.