Fimbriae-mediated outer membrane vesicle production and invasion of Porphyromonas gingivalis

Abstract

Porphyromonas gingivalis is a keystone periopathogen that plays an essential role in the progress of periodontitis. Like other gram-negative bacteria, the ability of P. gingivalis to produce outer membrane vesicles is a strategy used to interact with, and survive within its biological niches. Here we compared the protein components associated with vesicles derived from a fimbriated strain (33277) and an afimbriated strain (W83) of P. gingivalis using proteomic analyses. Some well-known virulence factors were identified in vesicles from both strains, such as gingipains and hemagglutinin. In contrast, FimC, FimD, and FimE, minor components of long fimbriae were found exclusively in 33277 vesicles, while proteins with a tetratricopeptide repeat (TPR) domain were unique to W83 vesicles. We found that significantly more 33277 than W83 vesicles were internalized into human oral keratinocytes and gingival fibroblasts. Interestingly, FimA, a well-known adhesin responsible for the attachment and invasion of P. gingivalis into host cells, was not essential for the invasive capabilities of P. gingivalis vesicles. Rather minor components of long fimbriae were required for an efficient invasive activity of vesicles. The most striking finding was that P. gingivalis strains lacking or having a reduced FimA expression showed a significant reduction in vesiculation. These results suggest that production and pathogenicity of P. gingivalis vesicles may largely depend on expression of the fim locus, and that the integration of vesicle production and pathogenicity with fimbrial expression may allow P. gingivalis to confer upon itself certain functional advantages.

Keywords: Invasion; P. gingivalis; outer membrane vesicles.

Figure 1

Comparison of vesicle production from Porphyromonas gingivalis strains. Vesicles were isolated from the 30 mL of growth media for P. gingivalis 33277, fimA mutant (FAE), fimR mutant (FRE), and an afimbriated strain (W83). (A) Protein concentrations of vesicle suspensions were determined using a Bio-Rad Protein Assay Kit, and each bar presents the relative protein concentrations of vesicles isolated from each P. gingivalis strain compared to that of P. gingivalis 33277 (as 1 unit). Error bars represent standard deviations (n = 3). An asterisk indicates a significant difference between the protein concentration of vesicles of a given P. gingivalis strain compared to P. gingivalis 33277 (*P < 0.05 by t-test). (B) Equal amounts of vesicle suspensions were determined by averaging the number of vesicles in 10 electron microscopic images. Each bar presents an average of the number of vesicles in an area of 5.6 × 5.6 μm.

Figure 2

Expression of Hgp44 and FimA in Porphyromonas gingivalis vesicles. (A) Levels of FimA and Hgp44 of gingipains in surface extracts (SE) of P. gingivalis 33277 or an afimbriated strain (W83), and in vesicles (V) of 33277, W83, fimA mutant (FAE), and fimR mutant (FRE) were analyzed using western blot analysis with rabbit anti-FimA or rabbit anti-Hgp44 sera. (B, C) Semiquantitation of western blots was conducted with ImageJ software. Each bar represents the relative intensity of (B) Hgp44 and (C) FimA. Values are shown with their standard deviations (n = 3). An asterisk indicates a significant difference between the relative level of Hgp44 or FimA in vesicles or surface extracts of P. gingivalis strains compared to those seen in 33277 vesicles (*P < 0.05 by t-test).

Figure 3

Visualization of intracellular vesicles in human gingival fibroblasts (HGFs) with confocal microscopy. (A) HGFs were grown in a glass bottom dish for 16 h, and then infected with Porphyromonas gingivalis vesicles (0.5 μg) isolated from 33277, fimA mutant (FAE), and an afimbriated strain (W83) strains. Cells were examined by confocal microscopy at 30 min and again at 4 h. Internalized vesicles were probed with anti-P. gingivalis polyclonal antibodies, visualized by Alexa Fluor 546-conjugated anti-rabbit IgG secondary antibody, and presented by differential interference contrast (DIC) imaging. (B) HGFs were treated with or without (control) 33277 vesicles for 15 min to 24 h as shown. As in (A), vesicles were probed with anti-P. gingivalis polyclonal antibodies and visualized by Alexa Fluor 546-conjugated anti-rabbit IgG secondary antibody.

Figure 4

Invasive efficiencies of Porphyromonas gingivalis vesicles using human gingival fibroblasts (HGFs). HGFs were infected with P. gingivalis vesicles derived from 33277, fimA mutant (FAE), fimR mutant (FRE), and an afimbriated strain (W83), respectively. Four hours after an initial exposure, the infected HGFs were analyzed by flow cytometry. (A) Representative FACS scatter plots for quantification of the infected HGFs with vesicles of different P. gingivalis strains. The percent of infected HGFs is indicated. (B) Bars are shown with standard deviations (n = 3) and represent the percent of HGFs with internalized vesicles. An asterisk indicates a significant difference between the % HGFs infected with 33277 vesicles compared to the % infected using the other P. gingivalis strains (*P < 0.05 by t-test).

Figure 5

Differential expression of fimC, fimD, and fimE in Porphyromonas gingivalis strains. Total RNA was extracted from P. gingivalis 33277, fimA mutant (FAE), fimR mutant (FRE), and an afimbriated strain (W83). The expression of fimC, fimD, and fimE in P. gingivalis was measured using a quantitative reverse transcript polymerase chain reaction (qRT-PCR). Each bar represents the relative expression of the gene shown compared to that in 33277 (unit of 1.0) after normalization with the expression level of glk. An asterisk indicates a statistically significant difference in expression level of the genes between P. gingivalis FAE, FRE, or W83, and 33277 (t-test; P < 0.05).

Figure 6

Invasive activity of Porphyromonas gingivalis vesicles into human oral keratinocytes (HOKs). (A) Vesicles derived from P. gingivalis 33277, fimA mutant (FAE), fimR mutant (FRE), or an afimbriated strain (W83) were added to transwells covered with a monolayer of HOKs. After a 24 h exposure, HOKs were collected by trypsinization and grown in a glass-bottomed dish for 24 h. Porphyromonas gingivalis vesicles were stained with Alex Fluor 546 (red) and observed under a confocal microscope. Infected cells are presented using differential interference contrast (DIC) images. (B) The percent of HOKs infected with vesicles was quantitated using FACS analysis. Data are shown as means ± standard deviations (n = 3). An asterisk indicates a significant difference between the percent of HOKs with internalized vesicles of P. gingivalis strains compared to those with 33277 vesicles (*P < 0.05 by t-test).

Figure 7

Transepithelial transport of Porphyromonas gingivalis vesicles across a monolayer and invasion of vesicles into human gingival fibroblasts (HGFs) in a transwell culture system. (A) The ability of vesicles derived from 33277, fimA mutant (FAE), fimR mutant (FRE), and an afimbriated strain (W83) to translocate across a monolayer of HOKs was examined using a transwell system and quantitated with enzyme-linked immunosorbent assay (ELISA). Data are shown as means of three independent experiments with standard deviations. An asterisk indicates a significant difference between translocated vesicles of 33277 compared to those with vesicles from other strains (*P < 0.05 by t-test). (B) The percent of HGFs infected by transepithelial vesicles in the wells of a six-well plate is shown. An asterisk indicates a significant difference between the percent of HGFs with internalized vesicles of P. gingivalis strains compared to those with 33277 vesicles (*P < 0.05 by t-test).