Figure 1.
Influence of TLR5 on bacterial growth during experimental B. cereus endophthalmitis.
C57BL/6J wild type and TLR5−/− mouse eyes were injected with 100 CFU B. cereus. Eyes were harvested, homogenized, and analyzed for bacterial growth. Overall, B. cereus grew to similar concentrations in infected eyes of TLR5−/− and C57BL/6J mice, suggesting that the absence of TLR5 did not influence the overall growth of B. cereus in the eye. Values represent the mean log10 CFU±SD of N≥4 eyes per time point for at least 2 separate experiments. *P≤0.05.
Figure 2.
Influence of TLR5 on retinal function during experimental B. cereus endophthalmitis.
C57BL/6J wild type and TLR5−/− mouse eyes were injected with 100 CFU B. cereus. Retinal function was assessed by electroretinography (ERG). A) A-wave amplitudes were slightly greater in C57BL/6J infected eyes at 12 h postinfection (P = 0.02), while B-wave amplitudes were greater in TLR5−/− infected eyes at 8 h postinfection (P = 0.008). By 12 hours postinfection, A-wave and B-wave amplitudes retained in both groups decreased to approximately 20% or less in infected eyes, indicating significant retinal function loss in both groups of mice regardless of the presence of TLR5. Values represent the mean ±SD of N≥4 eyes per time point for at least 2 separate experiments. *P≤0.05. B) Representative averaged waveforms from wild type (WT) and TLR5−/− mice at 12 h postinfection, with one eye infected and the contralateral eye serving as the uninfected control. Representative of N≥4 eyes per time point.
Figure 3.
Whole eye histology of experimental B. cereus endophthalmitis.
C57BL/6J and TLR5−/− mouse eyes were injected with 100 CFU B. cereus. Whole globes were harvested and processed for hematoxylin and eosin staining. Infected eyes of both groups had significant inflammation by 12 h postinfection, suggesting that the absence of TLR5 did not greatly impact intraocular inflammation. Sections are representative of 4 eyes per group. Magnification, 10X.
Figure 4.
Influence of TLR5 on infiltration of PMN into mouse eyes during experimental B. cereus endophthalmitis.
C57BL/6J and TLR5−/− mouse eyes were injected with 100 CFU B. cereus. PMN infiltration was estimated by quantifying MPO in whole eyes by sandwich ELISA. MPO levels were similar in these groups at all times points postinfection (P≥0.17), suggesting that the absence of TLR5 did not alter the PMN response during infection. Values represent the mean ±SD for N≥4 per group for at least 2 separate experiments. *P≤0.05.
Figure 5.
Influence of TLR5 on proinflammatory mediator expression during experimental B. cereus endophthalmitis.
C57BL/6J and TLR5−/− mouse eyes were injected with 100 CFU B. cereus. Ocular proinflammatory cytokines and chemokines were analyzed by sandwich ELISA. Overall, similar levels of TNFα, KC, IL6, and IL1β were synthesized in infected eyes of C57BL/6J mice compared with that in infected eyes of TLR5−/− mice, suggesting that the absence of TLR5 did not alter the inflammatory mediator response during infection. Values represent the mean ±SD for N≥4 per group for at least 2 separate experiments. *P≤0.05.
Figure 6.
Role of flagellin in intraocular inflammation.
A) Purified B. cereus flagellin was injected into C57BL/6J mouse eyes. B. cereus flagellin was purified as described in the Methods and analyzed for purity on a SDS-PAGE/Coomassie gel. A single band of 29 kD was recovered (lane F). B and C) Following injection of 0.5 ng/0.5 µL of purified flagellin, eyes were analyzed by ERG and whole globe histology at 8 and 12 hours. At 8 hours, posterior segment inflammation was minimal, retinas were intact, and retinal function decreased slightly (but not significantly) from that at time 0 (Figure 6B, P≥0.4). Compared with the significant inflammation and retinal function loss observed during infection, these results suggest that B. cereus flagellin alone may not have contributed significantly to the process. Infected eyes at 12 hours (Figure 3) demonstrated much greater pathology than that seen at 12 hours in eyes injected with flagellin alone. D) TLR5 activation by purified B. cereus (BC) flagellin was compared with that of purified Salmonella typhimurium (ST) flagellin. B. cereus flagellin was a weak TLR5 agonist, resulting in significantly less NFκB activity at comparable flagellin concentrations (mean ±SD for two repeated experiments, *P≤0.001).
Figure 7.
Multiple sequence alignments of S. typhimurium, B. cereus, B. thuringiensis, and B. anthracis flagellins.
The amino acid sequences of B. cereus (Bc) and S. typhimurium (St) (top) and Bc, B. thuringiensis (Bt), and B. anthracis (Ba) (bottom) were aligned by ClustalW [71], focusing on regions important for IL8 activity (Region 1) and TLR5 stimulation and recognition (Regions 2 and 3). Asterisks and red letters identify amino acids conserved between Bc and St sequences or among Bc/Ba/Bt sequences. Blue amino acids have strongly similar properties, while green amino acids have weakly similar properties. Bc and St had 81% conserved residues in Region 1, 62.5% conserved residues in Region 2, and 25% conserved residues in Region 3. Bc, Ba, and Bt had 86% conserved residues in Region 1, 62.5% conserved residues in Region 2, and 44% conserved residues in Region 3.