ExoS and ExoT ADP ribosyltransferase activities mediate Pseudomonas aeruginosa keratitis by promoting neutrophil apoptosis and bacterial survival

Abstract

Pseudomonas aeruginosa is a leading cause of blinding corneal ulcers worldwide. To determine the role of type III secretion in the pathogenesis of P. aeruginosa keratitis, corneas of C57BL/6 mice were infected with P. aeruginosa strain PAO1 or PAK, which expresses ExoS, ExoT, and ExoY, but not ExoU. PAO1- and PAK-infected corneas developed severe disease with pronounced opacification and rapid bacterial growth. In contrast, corneas infected with ΔpscD or ΔpscJ mutants that cannot assemble a type III secretion system, or with mutants lacking the translocator proteins, do not develop clinical disease, and bacteria are rapidly killed by infiltrating neutrophils. Furthermore, survival of PAO1 and PAK strains in the cornea and development of corneal disease was impaired in ΔexoS, ΔexoT, and ΔexoST mutants of both strains, but not in a ΔexoY mutant. ΔexoST mutants were also rapidly killed in neutrophils in vitro and were impaired in their ability to promote neutrophil apoptosis in vivo compared with PAO1. Point mutations in the ADP ribosyltransferase (ADPR) regions of ExoS or ExoT also impaired proapoptotic activity in infected neutrophils, and exoST(ADPR-) mutants replicated the ΔexoST phenotype in vitro and in vivo, whereas mutations in rho-GTPase-activating protein showed the same phenotype as PAO1. Together, these findings demonstrate that the pathogenesis of P. aeruginosa keratitis in ExoS- and ExoT-producing strains is almost entirely due to their ADPR activities, which subvert the host response by targeting the antibacterial activity of infiltrating neutrophils.

Figure 1. The role of Type III secretion in P. aeruginosa keratitis

Corneas of C57BL/6 mice were abraded and infected with PAO1, ΔpscD, or ΔpscD with a plasmid containing the pscD gene (ΔpscD/pscD). A. Representative images of corneal disease at 72h post infection (original magnification is x20); B,C. Quantification of corneal opacity by area (B, percent of total cornea) and average pixel intensity (C) as described in Methods. D. CFU 2h and 72h post infection. Data points represent individual eyes were analyzed by ANOVA and the overall value p value was <0.001. P values for pair-wise comparisons from the post-hoc tests are shown in the graphs. E. H&E stains of 5 μm corneal sections at 24h post infection; F. Posterior corneal stroma, corneal endothelium (endo) and anterior chamber (ac) of a ΔpscD/pscD infected cornea immunostained with anti-neutrophil antibody and Alexafluor 495 tagged secondary Ab (green fluorescence). Neutrophils are present in the stroma, attached to the endothelium, and in the anterior chamber (original magnification for B is x400 and C is x600). G. Chemokine production in corneas 3h or 24h post infection. Data are mean +/− SD of five mice per group. These experiments were repeated five times with similar results.

Figure 2. The role of Type III secretion in MyD88^−/− mice

Corneas of C57BL/6 and MyD88^−/− mice were abraded and infected with 1 × 10^5 PAO1 or ΔpscD. Corneal opacification, histology, and CFU were determined after 24h. A. Representative images of corneal disease; B. Percent corneal opacity; C. average corneal opacity; D. H&E stained corneal sections from ΔpscD infected C57BL/6 and MyD88^−/− mice (inset shows bacteria in the corneal stroma). E. CFU showing growth of ΔpscD mutants in MyD88^−/− mice. Data points represent corneas from individual mice, analyzed by ANOVA, and p values for pair-wise comparisons from the post-hoc tests are shown. The overall ANOVA value for corneal opacification and CFU was <0.001. Original magnification for A: x20, D: x400, inset: x1000) Experiments were repeated three times with similar results.

Figure 3. The role of popBD, ExoS, ExoT, and ExoY after infection with bacterial strains PAO1 or PAK

A. Representative corneas, quantitative corneal opacity and CFU in C57BL/6 mice 48h after infection with 1 × 10⁵ PAO1, ΔpscD or with mutants in the translocon apparatus (ΔpopBD); B. Corneas infected with PAO1, ΔexoS,T,Y (Δ3TOX) or ΔexoY mutants; C. Corneal infection with PAO1, Δ3TOX, ΔexoS, ΔexoT, or ΔexoST mutants; D. Corneas infected with strain PAK, type III secretion mutant ΔpscJ, or ΔexoS, ΔexoT, or ΔexoST PAK mutants. Data points represent individual corneas, which were analyzed by ANOVA, and p values for pair-wise comparisons from the post-hoc tests are shown. Overall ANOVA p value was <0.0001 for each set of studies. These experiments were repeated twice with similar results.

Figure 4. Survival of ΔexoST mutants in mice with impaired neutrophil and macrophage recruitment to the cornea

Corneas of C57BL/6, MyD88^−/−, TLR4/5^−/−, and macrophage depleted Mafia mice previously shown to have impaired cellular recruitment (8) were infected with 1 × 10⁵ PAO1 derived ΔexoST mutants, and CFU were examined after 48h. Data points represent CFU of individual corneas from one of two repeat experiments, analyzed by ANOVA, and p values for pair-wise comparisons from the post-hoc test are shown.

Figure 5. Survival of ExoS and ExoT ADP ribosyltransferase and rho-GAP mutants in neutrophils and macrophages in vitro

A,B. Protein expression in ADPR and GAP point mutants: Point mutations were generated as described in Methods. Total cell lysates and culture supernatants were processed for Western blot analysis to detect ExoS, ExoT and the cell wall associated protein RpoA. C-H. Peritoneal neutrophils or macrophages were incubated 30 min with PAO1 or with each mutant (10:1 MOI) followed by gentamicin as described in the Methods. After 90 min incubation, cells were lysed and CFU were quantified as before. C,D. Neutrophils (C) or macrophages (D) incubated with PAO1, ΔpscD, ΔexoS, ΔexoT, or ΔexoS/T. E,F. Neutrophils incubated with ExoS or ExoT ADPR and GAP mutants. G. Neutrophils were infected with the exoS(A−) mutant, exoT(A−) mutant, or the exoS(A−)/exoT(A−) double mutant. H. Neutrophils were infected with exoS(G-), exoT(G-) or exoS(G-) / exoT(G-) double mutants. Data are mean +/− SD of four wells per group, were analyzed by ANOVA, and p values for pair-wise comparisons from the post-hoc test are shown. Data are representative of three repeat experiments.

Figure 6. The role of ExoS and ExoT ADPR and rho-GAP activities in neutrophil apoptosis

Corneas of C57BL/6 mice were infected with 1 × 10⁵ PAO1 or ExoS and ExoT rho-GAP and ADPR mutants as before. After 24h, corneas were digested with collagenase, and cells were incubated with anti-neutrophil NIMPR-14 anti-neutrophil antibody and anti-annexin antibody, and examined by flow cytometry to determine the total neutrophil number and the percent apoptotic neutrophils per group. Representative FACS profiles are also shown. Total neutrophils and percent annexin positive neutrophils are shown after infection with: A: ΔexoS, ΔexoT or ΔexoST mutants; B: exoS(G-), exoS(A−), or exoS(GA-) mutants; C: exoT(G-), exoT(A−), or exoT(GA-) mutants; D: exoS(A−), exoT(A−), exoST(A−) mutants. For each set of experiments, data were analyzed by ANOVA, and p values for pair-wise comparisons from the post-hoc test are shown. These data are representative of two repeat experiments with similar results.

Figure 7. The role of ExoS and ExoT ADP ribosyltransferase in corneal disease and bacterial survival

Corneas of C57BL/6 mice were abraded and infected with 1 × 10⁵ PAO1, exoS(A−), exoT(A−), exoST(A−), or ΔexoST. A. Corneal opacification in representative mice examined 48h after infection. B. percent corneal opacity; C. average corneal opacity; D. CFU recovered from infected eyes; each data point represents an individual animal showing a significant difference between PAO1 and each of the mutants. Data were analyzed by ANOVA, and p values for pair-wise comparisons from the post-hoc test are shown. Overall p value was <0.001, although there was no significant difference between exoST(A−) and ΔexoST mutants. This experiment was repeated twice with similar results.