Porphyromonas gingivalis is a key periodontopathogen that can seriously debilitate the immune responses and the low grade periodontal infection continuously feeds more bacteria into the blood stream . The bacterial interaction with monocytes through TLRs and PARs enables a signalling cascade leading to a release of a plethora of pro-inflammatory mediators e.g. CXCL8, which is a potent chemokine responsible for neutrophil recruitment . The aim of this study was to elucidate the mechanisms of P. gingivalis interaction with monocytes in terms of induction and expression of CXCL8. By using gingipain mutants, we found that Rgp and Kgp help P. gingivalis evade the immune system both in an antagonistic and synergistic manner. The results are consistent with the previous studies and would be a rational to direct future studies using peripheral blood monocytes.
Both viable and heat-killed W50 strains of P. gingivalis shed significant amounts of LPS during interaction with THP-1 cells and induced release of CXCL8 indicating that the TLRs are involved as LPS is heat-stable and is an agonist for TLR2 activation. We observed that the levels of CXCL8 are 30–40 times higher in THP-1 cells stimulated with heat-killed ATCC and W50 compared to cells exposed to viable bacteria. TLR2 has previously been shown to be activated by both viable and heat-killed P. gingivalis . If the bacteria-induced release of CXCL8 is based only on LPS, there would have been no difference in CXCL8 response between viable and heat- killed bacteria, however, this was not the case.
Gingipains are trypsin-like cysteine proteases that can cleave the proteins at arginine and lysine specific sites . We observed that the different strains of P. gingivalis (viable or heat-killed) upregulated CXCL8 gene several hundred-folds. Uehara and colleagues investigated the possible synergistic effects of PARs, TLRs and NODs on CXCL8 production in THP-1 cells using synthetic peptide ligands and showed that gingipains stimulate the secretion of cytokines from monocytic cells through the activation of PARs with synergistic effects by PRRs . Consequently, the inflammatory response induced by gingipains was not exclusively dependent on their catalytic activity since heat-inactivated bacterial preparations were still effective whereas proteolytic activity was absent by denaturation . The loss of protease activity on heat-treatment was validated by a cytokine degradation functional assay and this also explains the reduced levels of extracellular CXCL8 in THP-1 cells stimulated with viable bacteria. Measuring the gingipain activity in the supernatant served as a functional validation of the isogenic gingipain mutants as well as the relative expression of arginine and lysine gingipains by the wild-type strains.
Aduse-Opoku and colleagues showed that Rgp and Kgp are secreted independent of each other . It is possible that ATCC33277 and W50 could be selectively inclined to secreting Rgp in aerobic cell culture conditions. It is evident that the gene transcription patterns do not follow the protein levels and could imply that the post-secretory CXCL8 protein is being degraded by gingipains. The inflammatory character of fimbriae and bacterial DNA cannot be disregarded. Thermal inactivation of microorganisms could release bacterial DNA due to possible membrane rupture activating TLR9 and P. gingivalis fimbriae are also documented to activate TLR2 and aid in CXCL8 release in human aortic endothelial cells [30, 31].
We show that W50 cleaves CXCL8 with the same efficiency as K1A, while ATCC 33277 is less effective. It is clear that there is variation in protease activity among the various wild-type strains of P. gingivalis. The degradation assay shows that at high titres, both Rgp and Kgp were equally effective in inactivating CXCL8, whereas at moderate and low titres, Rgp was more effective at degrading CXCL8. We have previously shown that CXCL8 hydrolysis can be salvaged by using leupeptin which is an arginine-specific gingipain inhibitor . However, all cytokines are not cleaved alike. Stathopolou and colleagues have shown that the rate of IL-1β degradation is lower compared to that of IL-6 or CXCL8, which could be due to the primary, secondary and tertiary molecular structure of IL-1β which makes it relatively resistant to hydrolysis by Kgp. It has been shown that interleukin-6 is rapidly cleaved by both gingipains, although Kgp was found to be more effective [16, 32].
The degradation assay makes it evident that Rgp is a more efficient protease at cleaving CXCL8, however, P. gingivalis expressing Rgp markedly amplifies CXCL8 transcription and significantly higher level of CXCL8 was present in culture supernatants of THP-1 treated with P. gingivalis expressing only Rgp (K1A). This could be due to a complex balance and interaction between induction and degradation. Induction of the CXCL8 gene is due to LPS, fimbriae, bacterial outer membrane proteins, DNA and gingipains, whereas degradation of CXCL8 is only protease dependent. In addition, it has been shown that CXCL8 is a secondary cytokine, i.e. its release could also be mediated by primary cytokines, like IL-1β and TNF in an autocrine manner . Hence, it is obvious that CXCL8 regulation during host pathogen interaction is very elaborate and the in vivo conditions are more dynamic compared to the situation in vitro. Also, gram negative bacteria can shed LPS in to the surrounding during cell death, outer membrane vesicle biogenesis and stress [34–36].
Confocal fluorescent images showed that THP-1 cells form aggregates when treated with P. gingivalis. Besides the changes in the actin cytoskeleton, P. gingivalis probably increased the expression of cell adhesion molecules in monocytes leading to cell aggregation . THP-1cells stimulated with Kgp-deficient K1A appear to be more differentiated and aggregated, compared to the E8 (lacks Rgp)-infected THP-1 cells, which could be suggestive of a role exhibited by Rgps in modulating THP-1 cell aggregation and adhesion. Hashizume and colleagues showed that cytoadherence between THP-1 cells and endothelial cells were enhanced by treatment with P. gingivalis . P. gingivalis invades or is internalized within the cells. We were able to culture intracellular bacteria by the antibiotic protection assay after 4 hour incubation showing that P. gingivalis survives intracellularly. The ability of ATCC 33277 to invade endothelial cells has previously been demonstrated in our lab . Fimbriae possibly contributes to the intracellular presence of the ATCC strain and even sparsely fimbriated strains successfully invaded although in varying degrees indicating there could be other factors supporting adhesion and infection of cells . W50 and its mutants possess type IV fimA and previous studies have confirmed that the precursor fimA can only become mature by the action of Rgp and, consequently, the lack of Rgp in E8 likely renders fimA ineffective [29, 41]. The giant multi-nucleated cells that formed on interaction of THP-1 cells with viable or heat-killed ATCC 33277 and W50 is of additional interest and one could speculate that these giant cells could be of osteoclastic phenotype of importance in the pathogenesis of periodontitis and needs further clarification.