CD47 and thrombospondin-1 contribute to immune evasion by Porphyromonas gingivalis

Abstract

Porphyromonas gingivalis is a gram-negative anaerobic bacterium linked to periodontal disease. Remarkably, P. gingivalis thrives in an inflamed environment rich in activated neutrophils. Toll-like receptor 2 (TLR2) recognition is required for P. gingivalis to evade innate immune killing; however, the mechanisms through which P. gingivalis uncouples host inflammation from bactericidal activity are only partially known. Since integrin activation and alternative signaling are implicated in P. gingivalis TLR2-mediated immune escape, we explored the role of CD47, a widely expressed integrin-associated protein known to suppress phagocytosis and implicated as an interacting partner with other innate immune receptors. We found that CD47 associates with TLR2, and blocking CD47 leads to decreased intracellular P. gingivalis survival in macrophages in a manner dependent on the bacterial major fimbria. In vivo, CD47 knock-out mice cleared P. gingivalis more efficiently than wild-type mice. Next, we found increased expression and secretion of the CD47 ligand thrombospondin-1 (TSP-1) following P. gingivalis infection. Secreted TSP-1 broadly protected P. gingivalis and other periodontitis-associated bacterial species from neutrophil bactericidal activity. Therefore, CD47-TLR2 cosignaling in response to P. gingivalis induces TSP-1 that in turn suppresses neutrophil activity, an effect that can explain how species such as P. gingivalis survive in an inflamed environment and cause dysbiosis.

Keywords: CD47; Porphyromonas gingivalis; immune evasion; thrombospondin-1.

Fig. 1.

CD47 promotes P. gingivalis survival in vitro and in vivo. THP-1 (A) and Raw 264.7 (B) macrophages were treated with blocking antibodies to CD47 or IgG control (10 μg/mL), followed by P. gingivalis (PG) infection (MOI 10). The intracellular bacterial survival assay was performed, and bacterial colonies were determined. Data of one representative experiment of three are shown. (C) THP-1 macrophages were pretreated with blocking antibodies to CD47 or IgG control. Cells were infected with ATCC strain 33277 (WT) or fimbria-mutant PG for 1 h. Intracellular bacterial survival was determined. (D) Graphical representation of in vivo experimental model. (E) Viable PG counts were determined following infection of WT and CD47^−/− mice. Each dot represents an individual mouse. One representative experiment of two is shown. (A and B) One-way ANOVA, (C) two-way ANOVA, and (E) two-tailed t test were performed to determine statistical significance (ns: nonsignificant, *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Fig. 2.

CD47 associates with TLR2. (A) THP-1 macrophages and (B) TLR2-Flag overexpressing THP-1 macrophages were infected with P. gingivalis (PG, MOI 10) for 30 min. TLR2 was IP with anti-TLR2 (A) or anti-Flag (B), and eluates were analyzed for CD47 (A and B), TLR2 (A), or TLR2-Flag (B) by immunoblot (IB). CD47 and TLR2-Flag protein levels in whole cell lysates (WCL) were also determined (no native TLR2 was detected in the THP-1 WCL). (C) THP-1 macrophages were infected with PG for 30 min, fixed, and stained for TLR2 (green) and CD47 (red). Nuclei were stained with DAPI (blue) and merged images were used to evaluate colocalization of TLR2 and CD47. Images were obtained using a Nikon confocal microscope at 40× magnification and colocalization coefficients were determined using ImageJ analysis software. Data are representative of two individual experiments. Two-tailed t test analysis was performed (ns: nonsignificant; *P ≤ 0.05; **P ≤ 0.01).

Fig. 3.

P. gingivalis influences CD47 expression. CD47 expression at baseline and in response to P. gingivalis (MOI 10) was measured by qRT-PCR after 4 h of challenge and flow cytometry after 24 h challenge. (A and B) THP-1, (C and D) Raw 264.7 macrophages, (E and F) human PBMC-derived macrophages, and (G and H) human peripheral blood neutrophils. Data are representative of two to four individual experiments. A two-tailed t test was used to analyze differences between groups (ns: nonsignificant; *P ≤ 0.05; **P ≤ 0.01; ***P < 0.001).

Fig. 4.

CD47 mediates P. gingivalis–induced AKT phosphorylation but not cytokine production and NFκB translocation. THP-1 (A) and Raw 264.7 (B) macrophages were pretreated with blocking antibodies to CD47 or TLR2, or IgG control. Cells were infected with P. gingivalis (PG, MOI 10) overnight, and cytokine levels were analyzed by the ELISA. (C) THP-1 macrophages were infected with PG (MOI 10) for 2 h, fixed, and stained with antibody to p65. Percentage nuclear translocation was determined per field. (D) THP-1 macrophages were infected with PG (MOI 20) for 10 min, and phosphorylated and total AKT levels were determined. (E) Densitometry graph of phospho/total ratios of D. Data are representative of three independent experiments. Statistical significance was determined using one-way ANOVA (ns: nonsignificant, *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Fig. 5.

P. gingivalis induces TSP-1 production in a CD47- and TLR2-dependent manner. SIRP-α mRNA (A) and protein levels (B) were determined after P. gingivalis (PG) infection for 4 h and 24 h, respectively. TSP-1 mRNA expression was measured in response to PG infection (4 h, MOI 10) in (C) THP-1 macrophages, (D) Raw 264.7 macrophages, (E) PBMC-derived macrophages, and (F) peripheral blood neutrophils. (G) TSP-1 protein expression was determined by the ELISA after PG infection at different time points and, (H) 24 h after infection with different PG MOI (10, 50, and 100) in THP-1 macrophages. (I) THP-1 macrophages were pretreated with blocking antibodies for CD47 or TLR2, or IgG control. TSP-1 mRNA levels were then determined by qRT-PCR after 4 h of PG infection. (J and K) THP-1 macrophages were pretreated with the NFκB inhibitor (Bay11) (J) or PI3K inhibitor (LY294; 100 μM) or its control inactive analogue (LY303; 100 μM) for 1 h. Cells were then infected with PG MOI 10 for 4 h. TSP-1 mRNA levels were determined by qRT-PCR. Data are representative of two-three individual experiments. Two-tailed t test analysis (A–D, G, and H) and one-way ANOVA (E, F, and I) were performed (ns: nonsignificant; *P ≤ 0.05; **P ≤ 0.01; ***P < 0.001; ****P < 0.0001).

Fig. 6.

TSP-1 down-regulates neutrophil bactericidal activity. (A) Peripheral blood neutrophils were pretreated with recombinant (r) TSP-1 (100 ng/mL) for 30 min and cells were then infected with FITC-labeled P. gingivalis (PG) for 30 min. Phagocytosis was determined by flow cytometry. (B and C) Neutrophils were treated with rTSP-1 (100 ng/mL) for 30 min and then infected with P. gingivalis for 1 h. Viable intracellular PG were determined after antibiotic treatment to remove extracellular bacteria (B). ROS levels were determined by Dihydrorhodamine 123 (DHR-123) (C). (D and E) Neutrophils were pretreated with TAX2 peptide or scrambled peptide control (100 µM) for 1 h prior to rTSP-1 (100 ng/mL) for 30 min and PG infection (MOI 10) for 1 h. Intracellular survival (D) and ROS levels (E) were determined. Data are representative of three individual experiments. Two-tailed t test (A and B) and one-way ANOVA (C–E) were performed to determine statistical significance (ns: nonsignificant, *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Fig. 7.

TSP-1 protects periodontal pathogens from neutrophil-mediated killing. (A) THP-1 macrophages were infected with F. nucleatum (FN, MOI 10) or T. forsythia (TF, MOI 10) for 4 h, and TSP-1 mRNA expression was determined. (B) TSP-1 protein after overnight infection of THP-1 macrophages with FN (MOI 10) or TF (MOI 10). (C) Peripheral blood neutrophils were pretreated with rTSP-1 at varying concentrations for 30 min. Cells were then infected with TF (MOI 10) for 1 h. Intracellular bacteria survival was determined. (D) Neutrophils were pretreated with rTSP-1 (100 ng/mL) for 30 min. Cells were then infected with FN (MOI 2) for 1 h and intracellular survival was determined. Data are representative of three individual experiments. Two-tailed t test (A, B, and D) and one-way ANOVA (C) were used to determine statistical significance (ns: nonsignificant, *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Fig. 8.

Schematic illustration. TLR2–CD47 signaling induces AKT phosphorylation, enhanced intracellular survival, and TSP-1 secretion in macrophages in response to P. gingivalis (A). Secreted TSP-1 down-regulates neutrophil ROS and bactericidal activity (B) that is restored by blocking the interaction of TSP-1 with CD47 (C). Figure created with BioRender.com.